modelId
stringlengths 5
122
| author
stringlengths 2
42
| last_modified
unknown | downloads
int64 0
738M
| likes
int64 0
11k
| library_name
stringclasses 245
values | tags
sequencelengths 1
4.05k
| pipeline_tag
stringclasses 48
values | createdAt
unknown | card
stringlengths 1
901k
|
|---|---|---|---|---|---|---|---|---|---|
laion/CLIP-ViT-B-32-256x256-DataComp-s34B-b86K | laion | "2023-11-13T16:26:57Z" | 9,060 | 7 | open_clip | [
"open_clip",
"safetensors",
"zero-shot-image-classification",
"dataset:mlfoundations/datacomp_pools",
"arxiv:2304.14108",
"license:mit",
"region:us"
] | zero-shot-image-classification | "2023-09-12T19:17:55Z" | ---
license: mit
widget:
- src: >-
https://huggingface.co/datasets/mishig/sample_images/resolve/main/cat-dog-music.png
candidate_labels: playing music, playing sports
example_title: Cat & Dog
library_name: open_clip
datasets:
- mlfoundations/datacomp_pools
pipeline_tag: zero-shot-image-classification
---
# Model card for CLIP ViT-B-32 256x256 trained DataComp-1B
# Table of Contents
1. [Model Details](#model-details)
2. [Uses](#uses)
3. [Training Details](#training-details)
4. [Evaluation](#evaluation)
5. [Acknowledgements](#acknowledgements)
6. [Citation](#citation)
7. [How To Get Started With the Model](#how-to-get-started-with-the-model)
# Model Details
## Model Description
A CLIP ViT-B/32 model trained with the DataComp-1B (https://github.com/mlfoundations/datacomp) using OpenCLIP (https://github.com/mlfoundations/open_clip) at 256x256 resolution.
Model training done on the [JURECA](https://www.fz-juelich.de/en/ias/jsc/systems/supercomputers/jureca) cluster.
# Uses
As per the original [OpenAI CLIP model card](https://github.com/openai/CLIP/blob/d50d76daa670286dd6cacf3bcd80b5e4823fc8e1/model-card.md), this model is intended as a research output for research communities. We hope that this model will enable researchers to better understand and explore zero-shot, arbitrary image classification. We also hope it can be used for interdisciplinary studies of the potential impact of such model.
The OpenAI CLIP paper includes a discussion of potential downstream impacts to provide an example for this sort of analysis. Additionally, the DataComp paper (https://arxiv.org/abs/2304.14108) include additional discussion as it relates specifically to the training dataset.
## Direct Use
Zero-shot image classification, image and text retrieval, among others.
## Downstream Use
Image classification and other image task fine-tuning, linear probe image classification, image generation guiding and conditioning, among others.
## Out-of-Scope Use
As per the OpenAI models,
**Any** deployed use case of the model - whether commercial or not - is currently out of scope. Non-deployed use cases such as image search in a constrained environment, are also not recommended unless there is thorough in-domain testing of the model with a specific, fixed class taxonomy. This is because our safety assessment demonstrated a high need for task specific testing especially given the variability of CLIP’s performance with different class taxonomies. This makes untested and unconstrained deployment of the model in any use case currently potentially harmful.
Certain use cases which would fall under the domain of surveillance and facial recognition are always out-of-scope regardless of performance of the model. This is because the use of artificial intelligence for tasks such as these can be premature currently given the lack of testing norms and checks to ensure its fair use.
# Training Details
## Training Data
This model was trained with the 1.4 Billion samples of the DataComp-1B dataset (https://arxiv.org/abs/2304.14108).
**IMPORTANT NOTE:** The motivation behind dataset creation is to democratize research and experimentation around large-scale multi-modal model training and handling of uncurated, large-scale datasets crawled from publically available internet. Our recommendation is therefore to use the dataset for research purposes. Be aware that this large-scale dataset is uncurated. Keep in mind that the uncurated nature of the dataset means that collected links may lead to strongly discomforting and disturbing content for a human viewer. Therefore, please use the demo links with caution and at your own risk. It is possible to extract a “safe” subset by filtering out samples based on the safety tags (using a customized trained NSFW classifier that we built). While this strongly reduces the chance for encountering potentially harmful content when viewing, we cannot entirely exclude the possibility for harmful content being still present in safe mode, so that the warning holds also there. We think that providing the dataset openly to broad research and other interested communities will allow for transparent investigation of benefits that come along with training large-scale models as well as pitfalls and dangers that may stay unreported or unnoticed when working with closed large datasets that remain restricted to a small community. Providing our dataset openly, we however do not recommend using it for creating ready-to-go industrial products, as the basic research about general properties and safety of such large-scale models, which we would like to encourage with this release, is still in progress.
## SLURM script
```bash
#!/bin/bash -x
#SBATCH --nodes=24
#SBATCH --gres=gpu:4
#SBATCH --ntasks-per-node=4
#SBATCH --cpus-per-task=12
#SBATCH --time=24:00:00
source /path/miniconda/bin/activate
export CUDA_VISIBLE_DEVICES=0,1,2,3
export MASTER_PORT=12802
master_addr=$(scontrol show hostnames "$SLURM_JOB_NODELIST" | head -n 1)
export MASTER_ADDR=$master_addr"i"
echo "MASTER_ADDR="$MASTER_ADDR
srun --cpu-bind=v --cpus-per-task=12 python -u -m training.main --aug-cfg scale='(0.4, 1.0)' color_jitter='(0.32, 0.32, 0.32, 0.08)' color_jitter_prob=0.8 gray_scale_prob=0.2 use_timm=True \
--save-frequency 1 \
--zeroshot-frequency 1 \
--dataset-type webdataset \
--train-data '/path/to/data' \
--report-to tensorboard \
--train-num-samples 1398270000 \
--warmup 2000 \
--batch-size 896 \
--epochs 24 \
--workers 8 \
--model ViT-B-32-256 \
--logs logs \
--seed 0 \
--ddp-static-graph \
--local-loss \
--gather-with-grad \
--lr 0.001 \
--log-every-n-steps 20 \
--save-most-recent \
--resume latest \
--grad-checkpointing \
--precision amp_bfloat16 \
--beta1 0.9 \
--beta2 0.95 \
--wd 0.2
```
# Evaluation
Evaluation done on 38 datasets, using [LAION CLIP Benchmark](https://github.com/LAION-AI/CLIP_benchmark).
## Testing Data, Factors & Metrics
### Testing Data
The testing is performed on a suite of 38 datasets. See our paper for more details (https://arxiv.org/abs/2304.14108).
## Results
The model achieves a 72.7% zero-shot top-1 accuracy on ImageNet-1k, 64.4% image retrieval recall@5 and 80.7% text retrieval recall@5 on COCO captions.
# Citation
**BibTeX:**
DataComp
```bibtex
@article{datacomp,
title={DataComp: In search of the next generation of multimodal datasets},
author={Samir Yitzhak Gadre, Gabriel Ilharco, Alex Fang, Jonathan Hayase, Georgios Smyrnis, Thao Nguyen, Ryan Marten, Mitchell Wortsman, Dhruba Ghosh, Jieyu Zhang, Eyal Orgad, Rahim Entezari, Giannis Daras, Sarah Pratt, Vivek Ramanujan, Yonatan Bitton, Kalyani Marathe, Stephen Mussmann, Richard Vencu, Mehdi Cherti, Ranjay Krishna, Pang Wei Koh, Olga Saukh, Alexander Ratner, Shuran Song, Hannaneh Hajishirzi, Ali Farhadi, Romain Beaumont, Sewoong Oh, Alex Dimakis, Jenia Jitsev, Yair Carmon, Vaishaal Shankar, Ludwig Schmidt},
journal={arXiv preprint arXiv:2304.14108},
year={2023}
}
```
OpenAI CLIP paper
```
@inproceedings{Radford2021LearningTV,
title={Learning Transferable Visual Models From Natural Language Supervision},
author={Alec Radford and Jong Wook Kim and Chris Hallacy and A. Ramesh and Gabriel Goh and Sandhini Agarwal and Girish Sastry and Amanda Askell and Pamela Mishkin and Jack Clark and Gretchen Krueger and Ilya Sutskever},
booktitle={ICML},
year={2021}
}
```
OpenCLIP software
```
@software{ilharco_gabriel_2021_5143773,
author = {Ilharco, Gabriel and
Wortsman, Mitchell and
Wightman, Ross and
Gordon, Cade and
Carlini, Nicholas and
Taori, Rohan and
Dave, Achal and
Shankar, Vaishaal and
Namkoong, Hongseok and
Miller, John and
Hajishirzi, Hannaneh and
Farhadi, Ali and
Schmidt, Ludwig},
title = {OpenCLIP},
month = jul,
year = 2021,
note = {If you use this software, please cite it as below.},
publisher = {Zenodo},
version = {0.1},
doi = {10.5281/zenodo.5143773},
url = {https://doi.org/10.5281/zenodo.5143773}
}
```
# How to Get Started with the Model
See https://github.com/mlfoundations/open_clip |
Vdr1/L3-8B-Sunfall-v0.4-Stheno-v3.2-GGUF-Imatrix-IQ | Vdr1 | "2024-07-02T20:13:58Z" | 9,052 | 1 | null | [
"gguf",
"not-for-all-audiences",
"llama3",
"roleplay",
"base_model:crestf411/L3-8B-sunfall-v0.4-stheno-v3.2",
"region:us"
] | null | "2024-06-21T01:53:30Z" | ---
base_model: crestf411/L3-8B-sunfall-v0.4-stheno-v3.2
tags:
- not-for-all-audiences
- llama3
- roleplay
---
>[!TIP]
>**EXPERIMENTAL!**
>**Read information in the original model repo for recommended settings to get the best results.**
GGUF Imatrix Quants of [L3-8B-sunfall-v0.4-stheno-v3.2](https://huggingface.co/crestf411/L3-8B-sunfall-v0.4-stheno-v3.2) |
optimum/gpt2 | optimum | "2023-01-03T10:29:58Z" | 9,039 | 5 | transformers | [
"transformers",
"onnx",
"gpt2",
"text-generation",
"exbert",
"en",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2022-11-22T10:17:23Z" | ---
language: en
tags:
- exbert
license: mit
---
# GPT-2
Test the whole generation capabilities here: https://transformer.huggingface.co/doc/gpt2-large
Pretrained model on English language using a causal language modeling (CLM) objective. It was introduced in
[this paper](https://d4mucfpksywv.cloudfront.net/better-language-models/language_models_are_unsupervised_multitask_learners.pdf)
and first released at [this page](https://openai.com/blog/better-language-models/).
Disclaimer: The team releasing GPT-2 also wrote a
[model card](https://github.com/openai/gpt-2/blob/master/model_card.md) for their model. Content from this model card
has been written by the Hugging Face team to complete the information they provided and give specific examples of bias.
## Model description
GPT-2 is a transformers model pretrained on a very large corpus of English data in a self-supervised fashion. This
means it was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots
of publicly available data) with an automatic process to generate inputs and labels from those texts. More precisely,
it was trained to guess the next word in sentences.
More precisely, inputs are sequences of continuous text of a certain length and the targets are the same sequence,
shifted one token (word or piece of word) to the right. The model uses internally a mask-mechanism to make sure the
predictions for the token `i` only uses the inputs from `1` to `i` but not the future tokens.
This way, the model learns an inner representation of the English language that can then be used to extract features
useful for downstream tasks. The model is best at what it was pretrained for however, which is generating texts from a
prompt.
## Intended uses & limitations
You can use the raw model for text generation or fine-tune it to a downstream task. See the
[model hub](https://huggingface.co/models?filter=gpt2) to look for fine-tuned versions on a task that interests you.
### How to use
Here is how to use the ONNX models of gpt2 to get the features of a given text:
Example using transformers.pipelines:
```python
from transformers import AutoTokenizer, pipeline
from optimum.onnxruntime import ORTModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("gpt2")
model = ORTModelForCausalLM.from_pretrained("gpt2", from_transformers=True)
onnx_gen = pipeline("text-generation", model=model, tokenizer=tokenizer)
text = "My name is Philipp and I live in Germany."
gen = onnx_gen(text)
```
Example of text generation:
```python
from transformers import AutoTokenizer
from optimum.onnxruntime import ORTModelForCausalLM
import torch
tokenizer = AutoTokenizer.from_pretrained("optimum/gpt2")
model = ORTModelForCausalLM.from_pretrained("optimum/gpt2")
inputs = tokenizer("My name is Arthur and I live in", return_tensors="pt")
gen_tokens = model.generate(**inputs,do_sample=True,temperature=0.9, min_length=20,max_length=20)
tokenizer.batch_decode(gen_tokens)
```
|
TheBloke/Llama-2-13B-chat-AWQ | TheBloke | "2023-11-09T18:21:10Z" | 9,034 | 25 | transformers | [
"transformers",
"safetensors",
"llama",
"text-generation",
"facebook",
"meta",
"pytorch",
"llama-2",
"en",
"arxiv:2307.09288",
"base_model:meta-llama/Llama-2-13b-chat-hf",
"license:llama2",
"autotrain_compatible",
"text-generation-inference",
"4-bit",
"awq",
"region:us"
] | text-generation | "2023-09-18T23:58:23Z" | ---
language:
- en
license: llama2
tags:
- facebook
- meta
- pytorch
- llama
- llama-2
model_name: Llama 2 13B Chat
base_model: meta-llama/Llama-2-13b-chat-hf
inference: false
model_creator: Meta Llama 2
model_type: llama
pipeline_tag: text-generation
prompt_template: '[INST] <<SYS>>
You are a helpful, respectful and honest assistant. Always answer as helpfully as
possible, while being safe. Your answers should not include any harmful, unethical,
racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses
are socially unbiased and positive in nature. If a question does not make any sense,
or is not factually coherent, explain why instead of answering something not correct.
If you don''t know the answer to a question, please don''t share false information.
<</SYS>>
{prompt}[/INST]
'
quantized_by: TheBloke
---
<!-- header start -->
<!-- 200823 -->
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p>
</div>
<div style="display: flex; flex-direction: column; align-items: flex-end;">
<p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p>
</div>
</div>
<div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div>
<hr style="margin-top: 1.0em; margin-bottom: 1.0em;">
<!-- header end -->
# Llama 2 13B Chat - AWQ
- Model creator: [Meta Llama 2](https://huggingface.co/meta-llama)
- Original model: [Llama 2 13B Chat](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf)
<!-- description start -->
## Description
This repo contains AWQ model files for [Meta's Llama 2 13B-chat](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf).
### About AWQ
AWQ is an efficient, accurate and blazing-fast low-bit weight quantization method, currently supporting 4-bit quantization. Compared to GPTQ, it offers faster Transformers-based inference.
It is also now supported by continuous batching server [vLLM](https://github.com/vllm-project/vllm), allowing use of AWQ models for high-throughput concurrent inference in multi-user server scenarios. Note that, at the time of writing, overall throughput is still lower than running vLLM with unquantised models, however using AWQ enables using much smaller GPUs which can lead to easier deployment and overall cost savings. For example, a 70B model can be run on 1 x 48GB GPU instead of 2 x 80GB.
<!-- description end -->
<!-- repositories-available start -->
## Repositories available
* [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/Llama-2-13B-chat-AWQ)
* [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Llama-2-13B-chat-GPTQ)
* [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/Llama-2-13B-chat-GGUF)
* [Meta Llama 2's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/meta-llama/Llama-2-13B-chat-hf)
<!-- repositories-available end -->
<!-- prompt-template start -->
## Prompt template: Llama-2-Chat
```
[INST] <<SYS>>
You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature. If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.
<</SYS>>
{prompt}[/INST]
```
<!-- prompt-template end -->
<!-- README_AWQ.md-provided-files start -->
## Provided files and AWQ parameters
For my first release of AWQ models, I am releasing 128g models only. I will consider adding 32g as well if there is interest, and once I have done perplexity and evaluation comparisons, but at this time 32g models are still not fully tested with AutoAWQ and vLLM.
Models are released as sharded safetensors files.
| Branch | Bits | GS | AWQ Dataset | Seq Len | Size |
| ------ | ---- | -- | ----------- | ------- | ---- |
| [main](https://huggingface.co/TheBloke/Llama-2-13B-chat-AWQ/tree/main) | 4 | 128 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 7.25 GB
<!-- README_AWQ.md-provided-files end -->
<!-- README_AWQ.md-use-from-vllm start -->
## Serving this model from vLLM
Documentation on installing and using vLLM [can be found here](https://vllm.readthedocs.io/en/latest/).
- When using vLLM as a server, pass the `--quantization awq` parameter, for example:
```shell
python3 python -m vllm.entrypoints.api_server --model TheBloke/Llama-2-13B-chat-AWQ --quantization awq
```
When using vLLM from Python code, pass the `quantization=awq` parameter, for example:
```python
from vllm import LLM, SamplingParams
prompts = [
"Hello, my name is",
"The president of the United States is",
"The capital of France is",
"The future of AI is",
]
sampling_params = SamplingParams(temperature=0.8, top_p=0.95)
llm = LLM(model="TheBloke/Llama-2-13B-chat-AWQ", quantization="awq")
outputs = llm.generate(prompts, sampling_params)
# Print the outputs.
for output in outputs:
prompt = output.prompt
generated_text = output.outputs[0].text
print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")
```
<!-- README_AWQ.md-use-from-vllm start -->
<!-- README_AWQ.md-use-from-python start -->
## How to use this AWQ model from Python code
### Install the necessary packages
Requires: [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) 0.0.2 or later
```shell
pip3 install autoawq
```
If you have problems installing [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) using the pre-built wheels, install it from source instead:
```shell
pip3 uninstall -y autoawq
git clone https://github.com/casper-hansen/AutoAWQ
cd AutoAWQ
pip3 install .
```
### You can then try the following example code
```python
from awq import AutoAWQForCausalLM
from transformers import AutoTokenizer
model_name_or_path = "TheBloke/Llama-2-13B-chat-AWQ"
# Load model
model = AutoAWQForCausalLM.from_quantized(model_name_or_path, fuse_layers=True,
trust_remote_code=False, safetensors=True)
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, trust_remote_code=False)
prompt = "Tell me about AI"
prompt_template=f'''[INST] <<SYS>>
You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature. If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.
<</SYS>>
{prompt}[/INST]
'''
print("\n\n*** Generate:")
tokens = tokenizer(
prompt_template,
return_tensors='pt'
).input_ids.cuda()
# Generate output
generation_output = model.generate(
tokens,
do_sample=True,
temperature=0.7,
top_p=0.95,
top_k=40,
max_new_tokens=512
)
print("Output: ", tokenizer.decode(generation_output[0]))
# Inference can also be done using transformers' pipeline
from transformers import pipeline
print("*** Pipeline:")
pipe = pipeline(
"text-generation",
model=model,
tokenizer=tokenizer,
max_new_tokens=512,
do_sample=True,
temperature=0.7,
top_p=0.95,
top_k=40,
repetition_penalty=1.1
)
print(pipe(prompt_template)[0]['generated_text'])
```
<!-- README_AWQ.md-use-from-python end -->
<!-- README_AWQ.md-compatibility start -->
## Compatibility
The files provided are tested to work with [AutoAWQ](https://github.com/casper-hansen/AutoAWQ), and [vLLM](https://github.com/vllm-project/vllm).
[Huggingface Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) is not yet compatible with AWQ, but a PR is open which should bring support soon: [TGI PR #781](https://github.com/huggingface/text-generation-inference/issues/781).
<!-- README_AWQ.md-compatibility end -->
<!-- footer start -->
<!-- 200823 -->
## Discord
For further support, and discussions on these models and AI in general, join us at:
[TheBloke AI's Discord server](https://discord.gg/theblokeai)
## Thanks, and how to contribute
Thanks to the [chirper.ai](https://chirper.ai) team!
Thanks to Clay from [gpus.llm-utils.org](llm-utils)!
I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training.
If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects.
Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits.
* Patreon: https://patreon.com/TheBlokeAI
* Ko-Fi: https://ko-fi.com/TheBlokeAI
**Special thanks to**: Aemon Algiz.
**Patreon special mentions**: Alicia Loh, Stephen Murray, K, Ajan Kanaga, RoA, Magnesian, Deo Leter, Olakabola, Eugene Pentland, zynix, Deep Realms, Raymond Fosdick, Elijah Stavena, Iucharbius, Erik Bjäreholt, Luis Javier Navarrete Lozano, Nicholas, theTransient, John Detwiler, alfie_i, knownsqashed, Mano Prime, Willem Michiel, Enrico Ros, LangChain4j, OG, Michael Dempsey, Pierre Kircher, Pedro Madruga, James Bentley, Thomas Belote, Luke @flexchar, Leonard Tan, Johann-Peter Hartmann, Illia Dulskyi, Fen Risland, Chadd, S_X, Jeff Scroggin, Ken Nordquist, Sean Connelly, Artur Olbinski, Swaroop Kallakuri, Jack West, Ai Maven, David Ziegler, Russ Johnson, transmissions 11, John Villwock, Alps Aficionado, Clay Pascal, Viktor Bowallius, Subspace Studios, Rainer Wilmers, Trenton Dambrowitz, vamX, Michael Levine, 준교 김, Brandon Frisco, Kalila, Trailburnt, Randy H, Talal Aujan, Nathan Dryer, Vadim, 阿明, ReadyPlayerEmma, Tiffany J. Kim, George Stoitzev, Spencer Kim, Jerry Meng, Gabriel Tamborski, Cory Kujawski, Jeffrey Morgan, Spiking Neurons AB, Edmond Seymore, Alexandros Triantafyllidis, Lone Striker, Cap'n Zoog, Nikolai Manek, danny, ya boyyy, Derek Yates, usrbinkat, Mandus, TL, Nathan LeClaire, subjectnull, Imad Khwaja, webtim, Raven Klaugh, Asp the Wyvern, Gabriel Puliatti, Caitlyn Gatomon, Joseph William Delisle, Jonathan Leane, Luke Pendergrass, SuperWojo, Sebastain Graf, Will Dee, Fred von Graf, Andrey, Dan Guido, Daniel P. Andersen, Nitin Borwankar, Elle, Vitor Caleffi, biorpg, jjj, NimbleBox.ai, Pieter, Matthew Berman, terasurfer, Michael Davis, Alex, Stanislav Ovsiannikov
Thank you to all my generous patrons and donaters!
And thank you again to a16z for their generous grant.
<!-- footer end -->
# Original model card: Meta's Llama 2 13B-chat
# **Llama 2**
Llama 2 is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. This is the repository for the 13B fine-tuned model, optimized for dialogue use cases and converted for the Hugging Face Transformers format. Links to other models can be found in the index at the bottom.
## Model Details
*Note: Use of this model is governed by the Meta license. In order to download the model weights and tokenizer, please visit the [website](https://ai.meta.com/resources/models-and-libraries/llama-downloads/) and accept our License before requesting access here.*
Meta developed and publicly released the Llama 2 family of large language models (LLMs), a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. Our fine-tuned LLMs, called Llama-2-Chat, are optimized for dialogue use cases. Llama-2-Chat models outperform open-source chat models on most benchmarks we tested, and in our human evaluations for helpfulness and safety, are on par with some popular closed-source models like ChatGPT and PaLM.
**Model Developers** Meta
**Variations** Llama 2 comes in a range of parameter sizes — 7B, 13B, and 70B — as well as pretrained and fine-tuned variations.
**Input** Models input text only.
**Output** Models generate text only.
**Model Architecture** Llama 2 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align to human preferences for helpfulness and safety.
||Training Data|Params|Content Length|GQA|Tokens|LR|
|---|---|---|---|---|---|---|
|Llama 2|*A new mix of publicly available online data*|7B|4k|✗|2.0T|3.0 x 10<sup>-4</sup>|
|Llama 2|*A new mix of publicly available online data*|13B|4k|✗|2.0T|3.0 x 10<sup>-4</sup>|
|Llama 2|*A new mix of publicly available online data*|70B|4k|✔|2.0T|1.5 x 10<sup>-4</sup>|
*Llama 2 family of models.* Token counts refer to pretraining data only. All models are trained with a global batch-size of 4M tokens. Bigger models - 70B -- use Grouped-Query Attention (GQA) for improved inference scalability.
**Model Dates** Llama 2 was trained between January 2023 and July 2023.
**Status** This is a static model trained on an offline dataset. Future versions of the tuned models will be released as we improve model safety with community feedback.
**License** A custom commercial license is available at: [https://ai.meta.com/resources/models-and-libraries/llama-downloads/](https://ai.meta.com/resources/models-and-libraries/llama-downloads/)
**Research Paper** ["Llama-2: Open Foundation and Fine-tuned Chat Models"](arxiv.org/abs/2307.09288)
## Intended Use
**Intended Use Cases** Llama 2 is intended for commercial and research use in English. Tuned models are intended for assistant-like chat, whereas pretrained models can be adapted for a variety of natural language generation tasks.
To get the expected features and performance for the chat versions, a specific formatting needs to be followed, including the `INST` and `<<SYS>>` tags, `BOS` and `EOS` tokens, and the whitespaces and breaklines in between (we recommend calling `strip()` on inputs to avoid double-spaces). See our reference code in github for details: [`chat_completion`](https://github.com/facebookresearch/llama/blob/main/llama/generation.py#L212).
**Out-of-scope Uses** Use in any manner that violates applicable laws or regulations (including trade compliance laws).Use in languages other than English. Use in any other way that is prohibited by the Acceptable Use Policy and Licensing Agreement for Llama 2.
## Hardware and Software
**Training Factors** We used custom training libraries, Meta's Research Super Cluster, and production clusters for pretraining. Fine-tuning, annotation, and evaluation were also performed on third-party cloud compute.
**Carbon Footprint** Pretraining utilized a cumulative 3.3M GPU hours of computation on hardware of type A100-80GB (TDP of 350-400W). Estimated total emissions were 539 tCO2eq, 100% of which were offset by Meta’s sustainability program.
||Time (GPU hours)|Power Consumption (W)|Carbon Emitted(tCO<sub>2</sub>eq)|
|---|---|---|---|
|Llama 2 7B|184320|400|31.22|
|Llama 2 13B|368640|400|62.44|
|Llama 2 70B|1720320|400|291.42|
|Total|3311616||539.00|
**CO<sub>2</sub> emissions during pretraining.** Time: total GPU time required for training each model. Power Consumption: peak power capacity per GPU device for the GPUs used adjusted for power usage efficiency. 100% of the emissions are directly offset by Meta's sustainability program, and because we are openly releasing these models, the pretraining costs do not need to be incurred by others.
## Training Data
**Overview** Llama 2 was pretrained on 2 trillion tokens of data from publicly available sources. The fine-tuning data includes publicly available instruction datasets, as well as over one million new human-annotated examples. Neither the pretraining nor the fine-tuning datasets include Meta user data.
**Data Freshness** The pretraining data has a cutoff of September 2022, but some tuning data is more recent, up to July 2023.
## Evaluation Results
In this section, we report the results for the Llama 1 and Llama 2 models on standard academic benchmarks.For all the evaluations, we use our internal evaluations library.
|Model|Size|Code|Commonsense Reasoning|World Knowledge|Reading Comprehension|Math|MMLU|BBH|AGI Eval|
|---|---|---|---|---|---|---|---|---|---|
|Llama 1|7B|14.1|60.8|46.2|58.5|6.95|35.1|30.3|23.9|
|Llama 1|13B|18.9|66.1|52.6|62.3|10.9|46.9|37.0|33.9|
|Llama 1|33B|26.0|70.0|58.4|67.6|21.4|57.8|39.8|41.7|
|Llama 1|65B|30.7|70.7|60.5|68.6|30.8|63.4|43.5|47.6|
|Llama 2|7B|16.8|63.9|48.9|61.3|14.6|45.3|32.6|29.3|
|Llama 2|13B|24.5|66.9|55.4|65.8|28.7|54.8|39.4|39.1|
|Llama 2|70B|**37.5**|**71.9**|**63.6**|**69.4**|**35.2**|**68.9**|**51.2**|**54.2**|
**Overall performance on grouped academic benchmarks.** *Code:* We report the average pass@1 scores of our models on HumanEval and MBPP. *Commonsense Reasoning:* We report the average of PIQA, SIQA, HellaSwag, WinoGrande, ARC easy and challenge, OpenBookQA, and CommonsenseQA. We report 7-shot results for CommonSenseQA and 0-shot results for all other benchmarks. *World Knowledge:* We evaluate the 5-shot performance on NaturalQuestions and TriviaQA and report the average. *Reading Comprehension:* For reading comprehension, we report the 0-shot average on SQuAD, QuAC, and BoolQ. *MATH:* We report the average of the GSM8K (8 shot) and MATH (4 shot) benchmarks at top 1.
|||TruthfulQA|Toxigen|
|---|---|---|---|
|Llama 1|7B|27.42|23.00|
|Llama 1|13B|41.74|23.08|
|Llama 1|33B|44.19|22.57|
|Llama 1|65B|48.71|21.77|
|Llama 2|7B|33.29|**21.25**|
|Llama 2|13B|41.86|26.10|
|Llama 2|70B|**50.18**|24.60|
**Evaluation of pretrained LLMs on automatic safety benchmarks.** For TruthfulQA, we present the percentage of generations that are both truthful and informative (the higher the better). For ToxiGen, we present the percentage of toxic generations (the smaller the better).
|||TruthfulQA|Toxigen|
|---|---|---|---|
|Llama-2-Chat|7B|57.04|**0.00**|
|Llama-2-Chat|13B|62.18|**0.00**|
|Llama-2-Chat|70B|**64.14**|0.01|
**Evaluation of fine-tuned LLMs on different safety datasets.** Same metric definitions as above.
## Ethical Considerations and Limitations
Llama 2 is a new technology that carries risks with use. Testing conducted to date has been in English, and has not covered, nor could it cover all scenarios. For these reasons, as with all LLMs, Llama 2’s potential outputs cannot be predicted in advance, and the model may in some instances produce inaccurate, biased or other objectionable responses to user prompts. Therefore, before deploying any applications of Llama 2, developers should perform safety testing and tuning tailored to their specific applications of the model.
Please see the Responsible Use Guide available at [https://ai.meta.com/llama/responsible-use-guide/](https://ai.meta.com/llama/responsible-use-guide)
## Reporting Issues
Please report any software “bug,” or other problems with the models through one of the following means:
- Reporting issues with the model: [github.com/facebookresearch/llama](http://github.com/facebookresearch/llama)
- Reporting problematic content generated by the model: [developers.facebook.com/llama_output_feedback](http://developers.facebook.com/llama_output_feedback)
- Reporting bugs and security concerns: [facebook.com/whitehat/info](http://facebook.com/whitehat/info)
## Llama Model Index
|Model|Llama2|Llama2-hf|Llama2-chat|Llama2-chat-hf|
|---|---|---|---|---|
|7B| [Link](https://huggingface.co/llamaste/Llama-2-7b) | [Link](https://huggingface.co/llamaste/Llama-2-7b-hf) | [Link](https://huggingface.co/llamaste/Llama-2-7b-chat) | [Link](https://huggingface.co/llamaste/Llama-2-7b-chat-hf)|
|13B| [Link](https://huggingface.co/llamaste/Llama-2-13b) | [Link](https://huggingface.co/llamaste/Llama-2-13b-hf) | [Link](https://huggingface.co/llamaste/Llama-2-13b-chat) | [Link](https://huggingface.co/llamaste/Llama-2-13b-hf)|
|70B| [Link](https://huggingface.co/llamaste/Llama-2-70b) | [Link](https://huggingface.co/llamaste/Llama-2-70b-hf) | [Link](https://huggingface.co/llamaste/Llama-2-70b-chat) | [Link](https://huggingface.co/llamaste/Llama-2-70b-hf)|
|
microsoft/trocr-base-stage1 | microsoft | "2024-05-27T20:13:12Z" | 9,033 | 9 | transformers | [
"transformers",
"pytorch",
"safetensors",
"vision-encoder-decoder",
"trocr",
"image-to-text",
"arxiv:2109.10282",
"endpoints_compatible",
"region:us"
] | image-to-text | "2022-03-02T23:29:05Z" | ---
tags:
- trocr
- image-to-text
---
# TrOCR (base-sized model, pre-trained only)
TrOCR pre-trained only model. It was introduced in the paper [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) by Li et al. and first released in [this repository](https://github.com/microsoft/unilm/tree/master/trocr).
Disclaimer: The team releasing TrOCR did not write a model card for this model so this model card has been written by the Hugging Face team.
## Model description
The TrOCR model is an encoder-decoder model, consisting of an image Transformer as encoder, and a text Transformer as decoder. The image encoder was initialized from the weights of BEiT, while the text decoder was initialized from the weights of RoBERTa.
Images are presented to the model as a sequence of fixed-size patches (resolution 16x16), which are linearly embedded. One also adds absolute position embeddings before feeding the sequence to the layers of the Transformer encoder. Next, the Transformer text decoder autoregressively generates tokens.
## Intended uses & limitations
You can use the raw model for optical character recognition (OCR) on single text-line images. See the [model hub](https://huggingface.co/models?search=microsoft/trocr) to look for fine-tuned versions on a task that interests you.
### How to use
Here is how to use this model in PyTorch:
```python
from transformers import TrOCRProcessor, VisionEncoderDecoderModel
from PIL import Image
import requests
# load image from the IAM database
url = 'https://fki.tic.heia-fr.ch/static/img/a01-122-02-00.jpg'
image = Image.open(requests.get(url, stream=True).raw).convert("RGB")
processor = TrOCRProcessor.from_pretrained('microsoft/trocr-base-stage1')
model = VisionEncoderDecoderModel.from_pretrained('microsoft/trocr-base-stage1')
# training
pixel_values = processor(image, return_tensors="pt").pixel_values # Batch size 1
decoder_input_ids = torch.tensor([[model.config.decoder.decoder_start_token_id]])
outputs = model(pixel_values=pixel_values, decoder_input_ids=decoder_input_ids)
```
### BibTeX entry and citation info
```bibtex
@misc{li2021trocr,
title={TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models},
author={Minghao Li and Tengchao Lv and Lei Cui and Yijuan Lu and Dinei Florencio and Cha Zhang and Zhoujun Li and Furu Wei},
year={2021},
eprint={2109.10282},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
``` |
ZeroWw/Meta-Llama-3-8B-Instruct-abliterated-v3-GGUF | ZeroWw | "2024-06-28T13:36:58Z" | 9,032 | 0 | null | [
"gguf",
"en",
"license:mit",
"region:us"
] | null | "2024-06-28T13:11:20Z" |
---
license: mit
language:
- en
---
My own (ZeroWw) quantizations.
output and embed tensors quantized to f16.
all other tensors quantized to q5_k or q6_k.
Result:
both f16.q6 and f16.q5 are smaller than q8_0 standard quantization
and they perform as well as the pure f16.
|
jonaskoenig/topic_classification_04 | jonaskoenig | "2023-09-13T06:40:36Z" | 9,029 | 25 | transformers | [
"transformers",
"tf",
"bert",
"text-classification",
"generated_from_keras_callback",
"base_model:microsoft/xtremedistil-l6-h256-uncased",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | "2022-07-20T07:26:43Z" | ---
license: mit
tags:
- generated_from_keras_callback
base_model: microsoft/xtremedistil-l6-h256-uncased
model-index:
- name: topic_classification_04
results: []
---
<!-- This model card has been generated automatically according to the information Keras had access to. You should
probably proofread and complete it, then remove this comment. -->
# topic_classification_04
This model is a fine-tuned version of [microsoft/xtremedistil-l6-h256-uncased](https://huggingface.co/microsoft/xtremedistil-l6-h256-uncased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Train Loss: 0.8325
- Train Sparse Categorical Accuracy: 0.7237
- Epoch: 9
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- optimizer: {'name': 'Adam', 'learning_rate': 5e-05, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
- training_precision: float32
### Training results
| Train Loss | Train Sparse Categorical Accuracy | Epoch |
|:----------:|:---------------------------------:|:-----:|
| 1.0735 | 0.6503 | 0 |
| 0.9742 | 0.6799 | 1 |
| 0.9424 | 0.6900 | 2 |
| 0.9199 | 0.6970 | 3 |
| 0.9016 | 0.7026 | 4 |
| 0.8853 | 0.7073 | 5 |
| 0.8707 | 0.7120 | 6 |
| 0.8578 | 0.7160 | 7 |
| 0.8448 | 0.7199 | 8 |
| 0.8325 | 0.7237 | 9 |
### Framework versions
- Transformers 4.20.1
- TensorFlow 2.9.1
- Datasets 2.3.2
- Tokenizers 0.12.1
|
timm/eva02_base_patch14_448.mim_in22k_ft_in22k_in1k | timm | "2024-02-10T23:37:34Z" | 9,028 | 4 | timm | [
"timm",
"pytorch",
"safetensors",
"image-classification",
"dataset:imagenet-1k",
"dataset:imagenet-22k",
"arxiv:2303.11331",
"arxiv:2303.15389",
"license:mit",
"region:us"
] | image-classification | "2023-03-31T04:17:36Z" | ---
license: mit
library_name: timm
tags:
- image-classification
- timm
datasets:
- imagenet-1k
- imagenet-22k
---
# Model card for eva02_base_patch14_448.mim_in22k_ft_in22k_in1k
An EVA02 image classification model. Pretrained on ImageNet-22k with masked image modeling (using EVA-CLIP as a MIM teacher) and fine-tuned on ImageNet-22k then on ImageNet-1k by paper authors.
EVA-02 models are vision transformers with mean pooling, SwiGLU, Rotary Position Embeddings (ROPE), and extra LN in MLP (for Base & Large).
NOTE: `timm` checkpoints are float32 for consistency with other models. Original checkpoints are float16 or bfloat16 in some cases, see originals if that's preferred.
## Model Details
- **Model Type:** Image classification / feature backbone
- **Model Stats:**
- Params (M): 87.1
- GMACs: 107.1
- Activations (M): 259.1
- Image size: 448 x 448
- **Papers:**
- EVA-02: A Visual Representation for Neon Genesis: https://arxiv.org/abs/2303.11331
- EVA-CLIP: Improved Training Techniques for CLIP at Scale: https://arxiv.org/abs/2303.15389
- **Original:**
- https://github.com/baaivision/EVA
- https://huggingface.co/Yuxin-CV/EVA-02
- **Pretrain Dataset:** ImageNet-22k
- **Dataset:** ImageNet-1k
## Model Usage
### Image Classification
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model('eva02_base_patch14_448.mim_in22k_ft_in22k_in1k', pretrained=True)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1
top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5)
```
### Image Embeddings
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model(
'eva02_base_patch14_448.mim_in22k_ft_in22k_in1k',
pretrained=True,
num_classes=0, # remove classifier nn.Linear
)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor
# or equivalently (without needing to set num_classes=0)
output = model.forward_features(transforms(img).unsqueeze(0))
# output is unpooled, a (1, 1025, 768) shaped tensor
output = model.forward_head(output, pre_logits=True)
# output is a (1, num_features) shaped tensor
```
## Model Comparison
Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results).
|model |top1 |top5 |param_count|img_size|
|-----------------------------------------------|------|------|-----------|--------|
|eva02_large_patch14_448.mim_m38m_ft_in22k_in1k |90.054|99.042|305.08 |448 |
|eva02_large_patch14_448.mim_in22k_ft_in22k_in1k|89.946|99.01 |305.08 |448 |
|eva_giant_patch14_560.m30m_ft_in22k_in1k |89.792|98.992|1014.45 |560 |
|eva02_large_patch14_448.mim_in22k_ft_in1k |89.626|98.954|305.08 |448 |
|eva02_large_patch14_448.mim_m38m_ft_in1k |89.57 |98.918|305.08 |448 |
|eva_giant_patch14_336.m30m_ft_in22k_in1k |89.56 |98.956|1013.01 |336 |
|eva_giant_patch14_336.clip_ft_in1k |89.466|98.82 |1013.01 |336 |
|eva_large_patch14_336.in22k_ft_in22k_in1k |89.214|98.854|304.53 |336 |
|eva_giant_patch14_224.clip_ft_in1k |88.882|98.678|1012.56 |224 |
|eva02_base_patch14_448.mim_in22k_ft_in22k_in1k |88.692|98.722|87.12 |448 |
|eva_large_patch14_336.in22k_ft_in1k |88.652|98.722|304.53 |336 |
|eva_large_patch14_196.in22k_ft_in22k_in1k |88.592|98.656|304.14 |196 |
|eva02_base_patch14_448.mim_in22k_ft_in1k |88.23 |98.564|87.12 |448 |
|eva_large_patch14_196.in22k_ft_in1k |87.934|98.504|304.14 |196 |
|eva02_small_patch14_336.mim_in22k_ft_in1k |85.74 |97.614|22.13 |336 |
|eva02_tiny_patch14_336.mim_in22k_ft_in1k |80.658|95.524|5.76 |336 |
## Citation
```bibtex
@article{EVA02,
title={EVA-02: A Visual Representation for Neon Genesis},
author={Fang, Yuxin and Sun, Quan and Wang, Xinggang and Huang, Tiejun and Wang, Xinlong and Cao, Yue},
journal={arXiv preprint arXiv:2303.11331},
year={2023}
}
```
```bibtex
@article{EVA-CLIP,
title={EVA-02: A Visual Representation for Neon Genesis},
author={Sun, Quan and Fang, Yuxin and Wu, Ledell and Wang, Xinlong and Cao, Yue},
journal={arXiv preprint arXiv:2303.15389},
year={2023}
}
```
```bibtex
@misc{rw2019timm,
author = {Ross Wightman},
title = {PyTorch Image Models},
year = {2019},
publisher = {GitHub},
journal = {GitHub repository},
doi = {10.5281/zenodo.4414861},
howpublished = {\url{https://github.com/huggingface/pytorch-image-models}}
}
```
|
daryl149/llama-2-7b-chat-hf | daryl149 | "2023-07-23T17:12:59Z" | 9,015 | 113 | transformers | [
"transformers",
"pytorch",
"llama",
"text-generation",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2023-07-18T18:36:56Z" | These are the converted model weights for Llama-2-7B-chat in Huggingface format.
Courtesy of [Mirage-Studio.io](https://mirage-studio.io), home of MirageGPT: the private ChatGPT alternative.
---
license: other
LLAMA 2 COMMUNITY LICENSE AGREEMENT
Llama 2 Version Release Date: July 18, 2023
"Agreement" means the terms and conditions for use, reproduction, distribution and
modification of the Llama Materials set forth herein.
"Documentation" means the specifications, manuals and documentation
accompanying Llama 2 distributed by Meta at ai.meta.com/resources/models-and-
libraries/llama-downloads/.
"Licensee" or "you" means you, or your employer or any other person or entity (if
you are entering into this Agreement on such person or entity's behalf), of the age
required under applicable laws, rules or regulations to provide legal consent and that
has legal authority to bind your employer or such other person or entity if you are
entering in this Agreement on their behalf.
"Llama 2" means the foundational large language models and software and
algorithms, including machine-learning model code, trained model weights,
inference-enabling code, training-enabling code, fine-tuning enabling code and other
elements of the foregoing distributed by Meta at ai.meta.com/resources/models-and-
libraries/llama-downloads/.
"Llama Materials" means, collectively, Meta's proprietary Llama 2 and
Documentation (and any portion thereof) made available under this Agreement.
"Meta" or "we" means Meta Platforms Ireland Limited (if you are located in or, if you
are an entity, your principal place of business is in the EEA or Switzerland) and Meta
Platforms, Inc. (if you are located outside of the EEA or Switzerland).
By clicking "I Accept" below or by using or distributing any portion or element of the
Llama Materials, you agree to be bound by this Agreement.
1. License Rights and Redistribution.
a. Grant of Rights. You are granted a non-exclusive, worldwide, non-
transferable and royalty-free limited license under Meta's intellectual property or
other rights owned by Meta embodied in the Llama Materials to use, reproduce,
distribute, copy, create derivative works of, and make modifications to the Llama
Materials.
b. Redistribution and Use.
i. If you distribute or make the Llama Materials, or any derivative works
thereof, available to a third party, you shall provide a copy of this Agreement to such
third party.
ii. If you receive Llama Materials, or any derivative works thereof, from
a Licensee as part of an integrated end user product, then Section 2 of this
Agreement will not apply to you.
iii. You must retain in all copies of the Llama Materials that you
distribute the following attribution notice within a "Notice" text file distributed as a
part of such copies: "Llama 2 is licensed under the LLAMA 2 Community License,
Copyright (c) Meta Platforms, Inc. All Rights Reserved."
iv. Your use of the Llama Materials must comply with applicable laws
and regulations (including trade compliance laws and regulations) and adhere to the
Acceptable Use Policy for the Llama Materials (available at
https://ai.meta.com/llama/use-policy), which is hereby incorporated by reference into
this Agreement.
v. You will not use the Llama Materials or any output or results of the
Llama Materials to improve any other large language model (excluding Llama 2 or
derivative works thereof).
2. Additional Commercial Terms. If, on the Llama 2 version release date, the
monthly active users of the products or services made available by or for Licensee,
or Licensee's affiliates, is greater than 700 million monthly active users in the
preceding calendar month, you must request a license from Meta, which Meta may
grant to you in its sole discretion, and you are not authorized to exercise any of the
rights under this Agreement unless or until Meta otherwise expressly grants you
such rights.
3. Disclaimer of Warranty. UNLESS REQUIRED BY APPLICABLE LAW, THE
LLAMA MATERIALS AND ANY OUTPUT AND RESULTS THEREFROM ARE
PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND,
EITHER EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, ANY
WARRANTIES OF TITLE, NON-INFRINGEMENT, MERCHANTABILITY, OR
FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY RESPONSIBLE
FOR DETERMINING THE APPROPRIATENESS OF USING OR REDISTRIBUTING
THE LLAMA MATERIALS AND ASSUME ANY RISKS ASSOCIATED WITH YOUR
USE OF THE LLAMA MATERIALS AND ANY OUTPUT AND RESULTS.
4. Limitation of Liability. IN NO EVENT WILL META OR ITS AFFILIATES BE
LIABLE UNDER ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, TORT,
NEGLIGENCE, PRODUCTS LIABILITY, OR OTHERWISE, ARISING OUT OF THIS
AGREEMENT, FOR ANY LOST PROFITS OR ANY INDIRECT, SPECIAL,
CONSEQUENTIAL, INCIDENTAL, EXEMPLARY OR PUNITIVE DAMAGES, EVEN
IF META OR ITS AFFILIATES HAVE BEEN ADVISED OF THE POSSIBILITY OF
ANY OF THE FOREGOING.
5. Intellectual Property.
a. No trademark licenses are granted under this Agreement, and in
connection with the Llama Materials, neither Meta nor Licensee may use any name
or mark owned by or associated with the other or any of its affiliates, except as
required for reasonable and customary use in describing and redistributing the
Llama Materials.
b. Subject to Meta's ownership of Llama Materials and derivatives made by or
for Meta, with respect to any derivative works and modifications of the Llama
Materials that are made by you, as between you and Meta, you are and will be the
owner of such derivative works and modifications.
c. If you institute litigation or other proceedings against Meta or any entity
(including a cross-claim or counterclaim in a lawsuit) alleging that the Llama
Materials or Llama 2 outputs or results, or any portion of any of the foregoing,
constitutes infringement of intellectual property or other rights owned or licensable
by you, then any licenses granted to you under this Agreement shall terminate as of
the date such litigation or claim is filed or instituted. You will indemnify and hold
harmless Meta from and against any claim by any third party arising out of or related
to your use or distribution of the Llama Materials.
6. Term and Termination. The term of this Agreement will commence upon your
acceptance of this Agreement or access to the Llama Materials and will continue in
full force and effect until terminated in accordance with the terms and conditions
herein. Meta may terminate this Agreement if you are in breach of any term or
condition of this Agreement. Upon termination of this Agreement, you shall delete
and cease use of the Llama Materials. Sections 3, 4 and 7 shall survive the
termination of this Agreement.
7. Governing Law and Jurisdiction. This Agreement will be governed and
construed under the laws of the State of California without regard to choice of law
principles, and the UN Convention on Contracts for the International Sale of Goods
does not apply to this Agreement. The courts of California shall have exclusive
jurisdiction of any dispute arising out of this Agreement.
---
|
ZeroWw/NSFW_DPO_Noromaid-7b-Mistral-7B-Instruct-v0.1-GGUF | ZeroWw | "2024-06-20T23:21:47Z" | 9,012 | 0 | null | [
"gguf",
"en",
"license:mit",
"region:us"
] | null | "2024-06-19T16:22:18Z" | ---
license: mit
language:
- en
---
My own (ZeroWw) quantizations.
output and embed tensors quantized to f16.
all other tensors quantized to q5_k or q6_k.
Result:
both f16.q6 and f16.q5 are smaller than q8_0 standard quantization
and they perform as well as the pure f16. |
nvidia/segformer-b2-finetuned-ade-512-512 | nvidia | "2022-08-06T10:07:54Z" | 9,010 | 2 | transformers | [
"transformers",
"pytorch",
"tf",
"segformer",
"vision",
"image-segmentation",
"dataset:scene_parse_150",
"arxiv:2105.15203",
"license:other",
"endpoints_compatible",
"region:us"
] | image-segmentation | "2022-03-02T23:29:05Z" | ---
license: other
tags:
- vision
- image-segmentation
datasets:
- scene_parse_150
widget:
- src: https://huggingface.co/datasets/hf-internal-testing/fixtures_ade20k/resolve/main/ADE_val_00000001.jpg
example_title: House
- src: https://huggingface.co/datasets/hf-internal-testing/fixtures_ade20k/resolve/main/ADE_val_00000002.jpg
example_title: Castle
---
# SegFormer (b2-sized) model fine-tuned on ADE20k
SegFormer model fine-tuned on ADE20k at resolution 512x512. It was introduced in the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Xie et al. and first released in [this repository](https://github.com/NVlabs/SegFormer).
Disclaimer: The team releasing SegFormer did not write a model card for this model so this model card has been written by the Hugging Face team.
## Model description
SegFormer consists of a hierarchical Transformer encoder and a lightweight all-MLP decode head to achieve great results on semantic segmentation benchmarks such as ADE20K and Cityscapes. The hierarchical Transformer is first pre-trained on ImageNet-1k, after which a decode head is added and fine-tuned altogether on a downstream dataset.
## Intended uses & limitations
You can use the raw model for semantic segmentation. See the [model hub](https://huggingface.co/models?other=segformer) to look for fine-tuned versions on a task that interests you.
### How to use
Here is how to use this model to classify an image of the COCO 2017 dataset into one of the 1,000 ImageNet classes:
```python
from transformers import SegformerFeatureExtractor, SegformerForSemanticSegmentation
from PIL import Image
import requests
feature_extractor = SegformerFeatureExtractor.from_pretrained("nvidia/segformer-b2-finetuned-ade-512-512")
model = SegformerForSemanticSegmentation.from_pretrained("nvidia/segformer-b2-finetuned-ade-512-512")
url = "http://images.cocodataset.org/val2017/000000039769.jpg"
image = Image.open(requests.get(url, stream=True).raw)
inputs = feature_extractor(images=image, return_tensors="pt")
outputs = model(**inputs)
logits = outputs.logits # shape (batch_size, num_labels, height/4, width/4)
```
For more code examples, we refer to the [documentation](https://huggingface.co/transformers/model_doc/segformer.html#).
### BibTeX entry and citation info
```bibtex
@article{DBLP:journals/corr/abs-2105-15203,
author = {Enze Xie and
Wenhai Wang and
Zhiding Yu and
Anima Anandkumar and
Jose M. Alvarez and
Ping Luo},
title = {SegFormer: Simple and Efficient Design for Semantic Segmentation with
Transformers},
journal = {CoRR},
volume = {abs/2105.15203},
year = {2021},
url = {https://arxiv.org/abs/2105.15203},
eprinttype = {arXiv},
eprint = {2105.15203},
timestamp = {Wed, 02 Jun 2021 11:46:42 +0200},
biburl = {https://dblp.org/rec/journals/corr/abs-2105-15203.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
```
|
mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF | mradermacher | "2024-06-22T15:06:27Z" | 9,010 | 0 | transformers | [
"transformers",
"gguf",
"en",
"base_model:SicariusSicariiStuff/LLAMA-3_8B_Unaligned_Alpha",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | "2024-06-22T14:37:53Z" | ---
base_model: SicariusSicariiStuff/LLAMA-3_8B_Unaligned_Alpha
language:
- en
library_name: transformers
license: apache-2.0
quantized_by: mradermacher
---
## About
<!-- ### quantize_version: 2 -->
<!-- ### output_tensor_quantised: 1 -->
<!-- ### convert_type: hf -->
<!-- ### vocab_type: -->
<!-- ### tags: -->
static quants of https://huggingface.co/SicariusSicariiStuff/LLAMA-3_8B_Unaligned_Alpha
<!-- provided-files -->
weighted/imatrix quants are available at https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-i1-GGUF
## Usage
If you are unsure how to use GGUF files, refer to one of [TheBloke's
READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
more details, including on how to concatenate multi-part files.
## Provided Quants
(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
| Link | Type | Size/GB | Notes |
|:-----|:-----|--------:|:------|
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.Q2_K.gguf) | Q2_K | 3.3 | |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.IQ3_XS.gguf) | IQ3_XS | 3.6 | |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.Q3_K_S.gguf) | Q3_K_S | 3.8 | |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.IQ3_S.gguf) | IQ3_S | 3.8 | beats Q3_K* |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.IQ3_M.gguf) | IQ3_M | 3.9 | |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.Q3_K_M.gguf) | Q3_K_M | 4.1 | lower quality |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.Q3_K_L.gguf) | Q3_K_L | 4.4 | |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.IQ4_XS.gguf) | IQ4_XS | 4.6 | |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.Q4_K_S.gguf) | Q4_K_S | 4.8 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.Q4_K_M.gguf) | Q4_K_M | 5.0 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.Q5_K_S.gguf) | Q5_K_S | 5.7 | |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.Q5_K_M.gguf) | Q5_K_M | 5.8 | |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.Q6_K.gguf) | Q6_K | 6.7 | very good quality |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.Q8_0.gguf) | Q8_0 | 8.6 | fast, best quality |
| [GGUF](https://huggingface.co/mradermacher/LLAMA-3_8B_Unaligned_Alpha-GGUF/resolve/main/LLAMA-3_8B_Unaligned_Alpha.f16.gguf) | f16 | 16.2 | 16 bpw, overkill |
Here is a handy graph by ikawrakow comparing some lower-quality quant
types (lower is better):
And here are Artefact2's thoughts on the matter:
https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
## FAQ / Model Request
See https://huggingface.co/mradermacher/model_requests for some answers to
questions you might have and/or if you want some other model quantized.
## Thanks
I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
me use its servers and providing upgrades to my workstation to enable
this work in my free time.
<!-- end -->
|
WillHeld/roberta-base-stsb | WillHeld | "2022-09-19T13:10:37Z" | 9,005 | 0 | transformers | [
"transformers",
"pytorch",
"roberta",
"text-classification",
"generated_from_trainer",
"en",
"dataset:glue",
"license:mit",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | "2022-09-19T12:55:34Z" | ---
language:
- en
license: mit
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- spearmanr
model-index:
- name: roberta-base-stsb
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: GLUE STSB
type: glue
args: stsb
metrics:
- name: Spearmanr
type: spearmanr
value: 0.9092158650855444
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# roberta-base-stsb
This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on the GLUE STSB dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4221
- Pearson: 0.9116
- Spearmanr: 0.9092
- Combined Score: 0.9104
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.06
- num_epochs: 10.0
### Training results
| Training Loss | Epoch | Step | Validation Loss | Pearson | Spearmanr | Combined Score |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:---------:|:--------------:|
| 1.6552 | 1.39 | 500 | 0.5265 | 0.8925 | 0.8925 | 0.8925 |
| 0.3579 | 2.78 | 1000 | 0.4626 | 0.9022 | 0.8991 | 0.9007 |
| 0.2198 | 4.17 | 1500 | 0.4396 | 0.9054 | 0.9042 | 0.9048 |
| 0.1585 | 5.56 | 2000 | 0.4537 | 0.9069 | 0.9052 | 0.9060 |
| 0.1139 | 6.94 | 2500 | 0.4975 | 0.9091 | 0.9065 | 0.9078 |
| 0.0868 | 8.33 | 3000 | 0.4221 | 0.9116 | 0.9092 | 0.9104 |
| 0.073 | 9.72 | 3500 | 0.4311 | 0.9096 | 0.9077 | 0.9086 |
### Framework versions
- Transformers 4.21.3
- Pytorch 1.7.1
- Datasets 1.18.3
- Tokenizers 0.11.6
|
PrunaAI/maywell-Llama-3-Ko-8B-Instruct-GGUF-smashed | PrunaAI | "2024-06-30T13:35:26Z" | 9,001 | 0 | null | [
"gguf",
"pruna-ai",
"region:us"
] | null | "2024-06-30T12:53:13Z" | ---
thumbnail: "https://assets-global.website-files.com/646b351987a8d8ce158d1940/64ec9e96b4334c0e1ac41504_Logo%20with%20white%20text.svg"
metrics:
- memory_disk
- memory_inference
- inference_latency
- inference_throughput
- inference_CO2_emissions
- inference_energy_consumption
tags:
- pruna-ai
---
<!-- header start -->
<!-- 200823 -->
<div style="width: auto; margin-left: auto; margin-right: auto">
<a href="https://www.pruna.ai/" target="_blank" rel="noopener noreferrer">
<img src="https://i.imgur.com/eDAlcgk.png" alt="PrunaAI" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</a>
</div>
<!-- header end -->
[](https://twitter.com/PrunaAI)
[](https://github.com/PrunaAI)
[](https://www.linkedin.com/company/93832878/admin/feed/posts/?feedType=following)
[](https://discord.com/invite/vb6SmA3hxu)
## This repo contains GGUF versions of the maywell/Llama-3-Ko-8B-Instruct model.
# Simply make AI models cheaper, smaller, faster, and greener!
- Give a thumbs up if you like this model!
- Contact us and tell us which model to compress next [here](https://www.pruna.ai/contact).
- Request access to easily compress your *own* AI models [here](https://z0halsaff74.typeform.com/pruna-access?typeform-source=www.pruna.ai).
- Read the documentations to know more [here](https://pruna-ai-pruna.readthedocs-hosted.com/en/latest/)
- Join Pruna AI community on Discord [here](https://discord.com/invite/vb6SmA3hxu) to share feedback/suggestions or get help.
**Frequently Asked Questions**
- ***How does the compression work?*** The model is compressed with GGUF.
- ***How does the model quality change?*** The quality of the model output might vary compared to the base model.
- ***What is the model format?*** We use GGUF format.
- ***What calibration data has been used?*** If needed by the compression method, we used WikiText as the calibration data.
- ***How to compress my own models?*** You can request premium access to more compression methods and tech support for your specific use-cases [here](https://z0halsaff74.typeform.com/pruna-access?typeform-source=www.pruna.ai).
# Downloading and running the models
You can download the individual files from the Files & versions section. Here is a list of the different versions we provide. For more info checkout [this chart](https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9) and [this guide](https://www.reddit.com/r/LocalLLaMA/comments/1ba55rj/overview_of_gguf_quantization_methods/):
| Quant type | Description |
|------------|--------------------------------------------------------------------------------------------|
| Q5_K_M | High quality, recommended. |
| Q5_K_S | High quality, recommended. |
| Q4_K_M | Good quality, uses about 4.83 bits per weight, recommended. |
| Q4_K_S | Slightly lower quality with more space savings, recommended. |
| IQ4_NL | Decent quality, slightly smaller than Q4_K_S with similar performance, recommended. |
| IQ4_XS | Decent quality, smaller than Q4_K_S with similar performance, recommended. |
| Q3_K_L | Lower quality but usable, good for low RAM availability. |
| Q3_K_M | Even lower quality. |
| IQ3_M | Medium-low quality, new method with decent performance comparable to Q3_K_M. |
| IQ3_S | Lower quality, new method with decent performance, recommended over Q3_K_S quant, same size with better performance. |
| Q3_K_S | Low quality, not recommended. |
| IQ3_XS | Lower quality, new method with decent performance, slightly better than Q3_K_S. |
| Q2_K | Very low quality but surprisingly usable. |
## How to download GGUF files ?
**Note for manual downloaders:** You almost never want to clone the entire repo! Multiple different quantisation formats are provided, and most users only want to pick and download a single file.
The following clients/libraries will automatically download models for you, providing a list of available models to choose from:
* LM Studio
* LoLLMS Web UI
* Faraday.dev
- **Option A** - Downloading in `text-generation-webui`:
- **Step 1**: Under Download Model, you can enter the model repo: maywell-Llama-3-Ko-8B-Instruct-GGUF-smashed and below it, a specific filename to download, such as: phi-2.IQ3_M.gguf.
- **Step 2**: Then click Download.
- **Option B** - Downloading on the command line (including multiple files at once):
- **Step 1**: We recommend using the `huggingface-hub` Python library:
```shell
pip3 install huggingface-hub
```
- **Step 2**: Then you can download any individual model file to the current directory, at high speed, with a command like this:
```shell
huggingface-cli download maywell-Llama-3-Ko-8B-Instruct-GGUF-smashed Llama-3-Ko-8B-Instruct.IQ3_M.gguf --local-dir . --local-dir-use-symlinks False
```
<details>
<summary>More advanced huggingface-cli download usage (click to read)</summary>
Alternatively, you can also download multiple files at once with a pattern:
```shell
huggingface-cli download maywell-Llama-3-Ko-8B-Instruct-GGUF-smashed --local-dir . --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
For more documentation on downloading with `huggingface-cli`, please see: [HF -> Hub Python Library -> Download files -> Download from the CLI](https://huggingface.co/docs/huggingface_hub/guides/download#download-from-the-cli).
To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`:
```shell
pip3 install hf_transfer
```
And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`:
```shell
HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download maywell-Llama-3-Ko-8B-Instruct-GGUF-smashed Llama-3-Ko-8B-Instruct.IQ3_M.gguf --local-dir . --local-dir-use-symlinks False
```
Windows Command Line users: You can set the environment variable by running `set HF_HUB_ENABLE_HF_TRANSFER=1` before the download command.
</details>
<!-- README_GGUF.md-how-to-download end -->
<!-- README_GGUF.md-how-to-run start -->
## How to run model in GGUF format?
- **Option A** - Introductory example with `llama.cpp` command
Make sure you are using `llama.cpp` from commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) or later.
```shell
./main -ngl 35 -m Llama-3-Ko-8B-Instruct.IQ3_M.gguf --color -c 32768 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "<s>[INST] {{prompt\}} [/INST]"
```
Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
Change `-c 32768` to the desired sequence length. For extended sequence models - eg 8K, 16K, 32K - the necessary RoPE scaling parameters are read from the GGUF file and set by llama.cpp automatically. Note that longer sequence lengths require much more resources, so you may need to reduce this value.
If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins`
For other parameters and how to use them, please refer to [the llama.cpp documentation](https://github.com/ggerganov/llama.cpp/blob/master/examples/main/README.md)
- **Option B** - Running in `text-generation-webui`
Further instructions can be found in the text-generation-webui documentation, here: [text-generation-webui/docs/04 ‐ Model Tab.md](https://github.com/oobabooga/text-generation-webui/blob/main/docs/04%20-%20Model%20Tab.md#llamacpp).
- **Option C** - Running from Python code
You can use GGUF models from Python using the [llama-cpp-python](https://github.com/abetlen/llama-cpp-python) or [ctransformers](https://github.com/marella/ctransformers) libraries. Note that at the time of writing (Nov 27th 2023), ctransformers has not been updated for some time and is not compatible with some recent models. Therefore I recommend you use llama-cpp-python.
### How to load this model in Python code, using llama-cpp-python
For full documentation, please see: [llama-cpp-python docs](https://abetlen.github.io/llama-cpp-python/).
#### First install the package
Run one of the following commands, according to your system:
```shell
# Base ctransformers with no GPU acceleration
pip install llama-cpp-python
# With NVidia CUDA acceleration
CMAKE_ARGS="-DLLAMA_CUBLAS=on" pip install llama-cpp-python
# Or with OpenBLAS acceleration
CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS" pip install llama-cpp-python
# Or with CLBLast acceleration
CMAKE_ARGS="-DLLAMA_CLBLAST=on" pip install llama-cpp-python
# Or with AMD ROCm GPU acceleration (Linux only)
CMAKE_ARGS="-DLLAMA_HIPBLAS=on" pip install llama-cpp-python
# Or with Metal GPU acceleration for macOS systems only
CMAKE_ARGS="-DLLAMA_METAL=on" pip install llama-cpp-python
# In windows, to set the variables CMAKE_ARGS in PowerShell, follow this format; eg for NVidia CUDA:
$env:CMAKE_ARGS = "-DLLAMA_OPENBLAS=on"
pip install llama-cpp-python
```
#### Simple llama-cpp-python example code
```python
from llama_cpp import Llama
# Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system.
llm = Llama(
model_path="./Llama-3-Ko-8B-Instruct.IQ3_M.gguf", # Download the model file first
n_ctx=32768, # The max sequence length to use - note that longer sequence lengths require much more resources
n_threads=8, # The number of CPU threads to use, tailor to your system and the resulting performance
n_gpu_layers=35 # The number of layers to offload to GPU, if you have GPU acceleration available
)
# Simple inference example
output = llm(
"<s>[INST] {{prompt}} [/INST]", # Prompt
max_tokens=512, # Generate up to 512 tokens
stop=["</s>"], # Example stop token - not necessarily correct for this specific model! Please check before using.
echo=True # Whether to echo the prompt
)
# Chat Completion API
llm = Llama(model_path="./Llama-3-Ko-8B-Instruct.IQ3_M.gguf", chat_format="llama-2") # Set chat_format according to the model you are using
llm.create_chat_completion(
messages = [
{{"role": "system", "content": "You are a story writing assistant."}},
{{
"role": "user",
"content": "Write a story about llamas."
}}
]
)
```
- **Option D** - Running with LangChain
Here are guides on using llama-cpp-python and ctransformers with LangChain:
* [LangChain + llama-cpp-python](https://python.langchain.com/docs/integrations/llms/llamacpp)
* [LangChain + ctransformers](https://python.langchain.com/docs/integrations/providers/ctransformers)
## Configurations
The configuration info are in `smash_config.json`.
## Credits & License
The license of the smashed model follows the license of the original model. Please check the license of the original model before using this model which provided the base model. The license of the `pruna-engine` is [here](https://pypi.org/project/pruna-engine/) on Pypi.
## Want to compress other models?
- Contact us and tell us which model to compress next [here](https://www.pruna.ai/contact).
- Request access to easily compress your own AI models [here](https://z0halsaff74.typeform.com/pruna-access?typeform-source=www.pruna.ai). |
mradermacher/blossom-v5-4b-GGUF | mradermacher | "2024-07-02T03:25:02Z" | 8,992 | 0 | transformers | [
"transformers",
"gguf",
"zh",
"en",
"dataset:Azure99/blossom-chat-v3",
"dataset:Azure99/blossom-math-v4",
"dataset:Azure99/blossom-wizard-v3",
"dataset:Azure99/blossom-orca-v3",
"base_model:Azure99/blossom-v5-4b",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | "2024-07-02T02:57:50Z" | ---
base_model: Azure99/blossom-v5-4b
datasets:
- Azure99/blossom-chat-v3
- Azure99/blossom-math-v4
- Azure99/blossom-wizard-v3
- Azure99/blossom-orca-v3
language:
- zh
- en
library_name: transformers
license: apache-2.0
quantized_by: mradermacher
---
## About
<!-- ### quantize_version: 2 -->
<!-- ### output_tensor_quantised: 1 -->
<!-- ### convert_type: hf -->
<!-- ### vocab_type: -->
<!-- ### tags: -->
static quants of https://huggingface.co/Azure99/blossom-v5-4b
<!-- provided-files -->
weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion.
## Usage
If you are unsure how to use GGUF files, refer to one of [TheBloke's
READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
more details, including on how to concatenate multi-part files.
## Provided Quants
(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
| Link | Type | Size/GB | Notes |
|:-----|:-----|--------:|:------|
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.Q2_K.gguf) | Q2_K | 1.7 | |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.IQ3_XS.gguf) | IQ3_XS | 1.9 | |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.IQ3_S.gguf) | IQ3_S | 2.0 | beats Q3_K* |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.Q3_K_S.gguf) | Q3_K_S | 2.0 | |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.IQ3_M.gguf) | IQ3_M | 2.0 | |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.Q3_K_M.gguf) | Q3_K_M | 2.1 | lower quality |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.Q3_K_L.gguf) | Q3_K_L | 2.3 | |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.IQ4_XS.gguf) | IQ4_XS | 2.3 | |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.Q4_K_S.gguf) | Q4_K_S | 2.4 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.Q4_K_M.gguf) | Q4_K_M | 2.6 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.Q5_K_S.gguf) | Q5_K_S | 2.9 | |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.Q5_K_M.gguf) | Q5_K_M | 2.9 | |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.Q6_K.gguf) | Q6_K | 3.3 | very good quality |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.Q8_0.gguf) | Q8_0 | 4.3 | fast, best quality |
| [GGUF](https://huggingface.co/mradermacher/blossom-v5-4b-GGUF/resolve/main/blossom-v5-4b.f16.gguf) | f16 | 8.0 | 16 bpw, overkill |
Here is a handy graph by ikawrakow comparing some lower-quality quant
types (lower is better):
And here are Artefact2's thoughts on the matter:
https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
## FAQ / Model Request
See https://huggingface.co/mradermacher/model_requests for some answers to
questions you might have and/or if you want some other model quantized.
## Thanks
I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
me use its servers and providing upgrades to my workstation to enable
this work in my free time.
<!-- end -->
|
timm/resnext101_32x32d.fb_wsl_ig1b_ft_in1k | timm | "2024-02-10T23:41:28Z" | 8,991 | 2 | timm | [
"timm",
"pytorch",
"safetensors",
"image-classification",
"arxiv:1805.00932",
"arxiv:1611.05431",
"arxiv:1512.03385",
"license:cc-by-nc-4.0",
"region:us"
] | image-classification | "2023-04-05T19:18:34Z" | ---
license: cc-by-nc-4.0
library_name: timm
tags:
- image-classification
- timm
---
# Model card for resnext101_32x32d.fb_wsl_ig1b_ft_in1k
A ResNeXt-B image classification model.
This model features:
* ReLU activations
* single layer 7x7 convolution with pooling
* 1x1 convolution shortcut downsample
* grouped 3x3 bottleneck convolutions
Pretrained on Instagram-1B hashtags dataset using weakly-supervised learning and fine-tuned on ImageNet-1k by paper authors.
## Model Details
- **Model Type:** Image classification / feature backbone
- **Model Stats:**
- Params (M): 468.5
- GMACs: 87.3
- Activations (M): 91.1
- Image size: 224 x 224
- **Papers:**
- Exploring the Limits of Weakly Supervised Pretraining: https://arxiv.org/abs/1805.00932
- Aggregated Residual Transformations for Deep Neural Networks: https://arxiv.org/abs/1611.05431
- Deep Residual Learning for Image Recognition: https://arxiv.org/abs/1512.03385
- **Original:** https://github.com/facebookresearch/WSL-Images
## Model Usage
### Image Classification
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model('resnext101_32x32d.fb_wsl_ig1b_ft_in1k', pretrained=True)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1
top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5)
```
### Feature Map Extraction
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model(
'resnext101_32x32d.fb_wsl_ig1b_ft_in1k',
pretrained=True,
features_only=True,
)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1
for o in output:
# print shape of each feature map in output
# e.g.:
# torch.Size([1, 64, 112, 112])
# torch.Size([1, 256, 56, 56])
# torch.Size([1, 512, 28, 28])
# torch.Size([1, 1024, 14, 14])
# torch.Size([1, 2048, 7, 7])
print(o.shape)
```
### Image Embeddings
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model(
'resnext101_32x32d.fb_wsl_ig1b_ft_in1k',
pretrained=True,
num_classes=0, # remove classifier nn.Linear
)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor
# or equivalently (without needing to set num_classes=0)
output = model.forward_features(transforms(img).unsqueeze(0))
# output is unpooled, a (1, 2048, 7, 7) shaped tensor
output = model.forward_head(output, pre_logits=True)
# output is a (1, num_features) shaped tensor
```
## Model Comparison
Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results).
|model |img_size|top1 |top5 |param_count|gmacs|macts|img/sec|
|------------------------------------------|--------|-----|-----|-----------|-----|-----|-------|
|[seresnextaa101d_32x8d.sw_in12k_ft_in1k_288](https://huggingface.co/timm/seresnextaa101d_32x8d.sw_in12k_ft_in1k_288)|320 |86.72|98.17|93.6 |35.2 |69.7 |451 |
|[seresnextaa101d_32x8d.sw_in12k_ft_in1k_288](https://huggingface.co/timm/seresnextaa101d_32x8d.sw_in12k_ft_in1k_288)|288 |86.51|98.08|93.6 |28.5 |56.4 |560 |
|[seresnextaa101d_32x8d.sw_in12k_ft_in1k](https://huggingface.co/timm/seresnextaa101d_32x8d.sw_in12k_ft_in1k)|288 |86.49|98.03|93.6 |28.5 |56.4 |557 |
|[seresnextaa101d_32x8d.sw_in12k_ft_in1k](https://huggingface.co/timm/seresnextaa101d_32x8d.sw_in12k_ft_in1k)|224 |85.96|97.82|93.6 |17.2 |34.2 |923 |
|[resnext101_32x32d.fb_wsl_ig1b_ft_in1k](https://huggingface.co/timm/resnext101_32x32d.fb_wsl_ig1b_ft_in1k)|224 |85.11|97.44|468.5 |87.3 |91.1 |254 |
|[resnetrs420.tf_in1k](https://huggingface.co/timm/resnetrs420.tf_in1k)|416 |85.0 |97.12|191.9 |108.4|213.8|134 |
|[ecaresnet269d.ra2_in1k](https://huggingface.co/timm/ecaresnet269d.ra2_in1k)|352 |84.96|97.22|102.1 |50.2 |101.2|291 |
|[ecaresnet269d.ra2_in1k](https://huggingface.co/timm/ecaresnet269d.ra2_in1k)|320 |84.73|97.18|102.1 |41.5 |83.7 |353 |
|[resnetrs350.tf_in1k](https://huggingface.co/timm/resnetrs350.tf_in1k)|384 |84.71|96.99|164.0 |77.6 |154.7|183 |
|[seresnextaa101d_32x8d.ah_in1k](https://huggingface.co/timm/seresnextaa101d_32x8d.ah_in1k)|288 |84.57|97.08|93.6 |28.5 |56.4 |557 |
|[resnetrs200.tf_in1k](https://huggingface.co/timm/resnetrs200.tf_in1k)|320 |84.45|97.08|93.2 |31.5 |67.8 |446 |
|[resnetrs270.tf_in1k](https://huggingface.co/timm/resnetrs270.tf_in1k)|352 |84.43|96.97|129.9 |51.1 |105.5|280 |
|[seresnext101d_32x8d.ah_in1k](https://huggingface.co/timm/seresnext101d_32x8d.ah_in1k)|288 |84.36|96.92|93.6 |27.6 |53.0 |595 |
|[seresnet152d.ra2_in1k](https://huggingface.co/timm/seresnet152d.ra2_in1k)|320 |84.35|97.04|66.8 |24.1 |47.7 |610 |
|[resnetrs350.tf_in1k](https://huggingface.co/timm/resnetrs350.tf_in1k)|288 |84.3 |96.94|164.0 |43.7 |87.1 |333 |
|[resnext101_32x8d.fb_swsl_ig1b_ft_in1k](https://huggingface.co/timm/resnext101_32x8d.fb_swsl_ig1b_ft_in1k)|224 |84.28|97.17|88.8 |16.5 |31.2 |1100 |
|[resnetrs420.tf_in1k](https://huggingface.co/timm/resnetrs420.tf_in1k)|320 |84.24|96.86|191.9 |64.2 |126.6|228 |
|[seresnext101_32x8d.ah_in1k](https://huggingface.co/timm/seresnext101_32x8d.ah_in1k)|288 |84.19|96.87|93.6 |27.2 |51.6 |613 |
|[resnext101_32x16d.fb_wsl_ig1b_ft_in1k](https://huggingface.co/timm/resnext101_32x16d.fb_wsl_ig1b_ft_in1k)|224 |84.18|97.19|194.0 |36.3 |51.2 |581 |
|[resnetaa101d.sw_in12k_ft_in1k](https://huggingface.co/timm/resnetaa101d.sw_in12k_ft_in1k)|288 |84.11|97.11|44.6 |15.1 |29.0 |1144 |
|[resnet200d.ra2_in1k](https://huggingface.co/timm/resnet200d.ra2_in1k)|320 |83.97|96.82|64.7 |31.2 |67.3 |518 |
|[resnetrs200.tf_in1k](https://huggingface.co/timm/resnetrs200.tf_in1k)|256 |83.87|96.75|93.2 |20.2 |43.4 |692 |
|[seresnextaa101d_32x8d.ah_in1k](https://huggingface.co/timm/seresnextaa101d_32x8d.ah_in1k)|224 |83.86|96.65|93.6 |17.2 |34.2 |923 |
|[resnetrs152.tf_in1k](https://huggingface.co/timm/resnetrs152.tf_in1k)|320 |83.72|96.61|86.6 |24.3 |48.1 |617 |
|[seresnet152d.ra2_in1k](https://huggingface.co/timm/seresnet152d.ra2_in1k)|256 |83.69|96.78|66.8 |15.4 |30.6 |943 |
|[seresnext101d_32x8d.ah_in1k](https://huggingface.co/timm/seresnext101d_32x8d.ah_in1k)|224 |83.68|96.61|93.6 |16.7 |32.0 |986 |
|[resnet152d.ra2_in1k](https://huggingface.co/timm/resnet152d.ra2_in1k)|320 |83.67|96.74|60.2 |24.1 |47.7 |706 |
|[resnetrs270.tf_in1k](https://huggingface.co/timm/resnetrs270.tf_in1k)|256 |83.59|96.61|129.9 |27.1 |55.8 |526 |
|[seresnext101_32x8d.ah_in1k](https://huggingface.co/timm/seresnext101_32x8d.ah_in1k)|224 |83.58|96.4 |93.6 |16.5 |31.2 |1013 |
|[resnetaa101d.sw_in12k_ft_in1k](https://huggingface.co/timm/resnetaa101d.sw_in12k_ft_in1k)|224 |83.54|96.83|44.6 |9.1 |17.6 |1864 |
|[resnet152.a1h_in1k](https://huggingface.co/timm/resnet152.a1h_in1k)|288 |83.46|96.54|60.2 |19.1 |37.3 |904 |
|[resnext101_32x16d.fb_swsl_ig1b_ft_in1k](https://huggingface.co/timm/resnext101_32x16d.fb_swsl_ig1b_ft_in1k)|224 |83.35|96.85|194.0 |36.3 |51.2 |582 |
|[resnet200d.ra2_in1k](https://huggingface.co/timm/resnet200d.ra2_in1k)|256 |83.23|96.53|64.7 |20.0 |43.1 |809 |
|[resnext101_32x4d.fb_swsl_ig1b_ft_in1k](https://huggingface.co/timm/resnext101_32x4d.fb_swsl_ig1b_ft_in1k)|224 |83.22|96.75|44.2 |8.0 |21.2 |1814 |
|[resnext101_64x4d.c1_in1k](https://huggingface.co/timm/resnext101_64x4d.c1_in1k)|288 |83.16|96.38|83.5 |25.7 |51.6 |590 |
|[resnet152d.ra2_in1k](https://huggingface.co/timm/resnet152d.ra2_in1k)|256 |83.14|96.38|60.2 |15.4 |30.5 |1096 |
|[resnet101d.ra2_in1k](https://huggingface.co/timm/resnet101d.ra2_in1k)|320 |83.02|96.45|44.6 |16.5 |34.8 |992 |
|[ecaresnet101d.miil_in1k](https://huggingface.co/timm/ecaresnet101d.miil_in1k)|288 |82.98|96.54|44.6 |13.4 |28.2 |1077 |
|[resnext101_64x4d.tv_in1k](https://huggingface.co/timm/resnext101_64x4d.tv_in1k)|224 |82.98|96.25|83.5 |15.5 |31.2 |989 |
|[resnetrs152.tf_in1k](https://huggingface.co/timm/resnetrs152.tf_in1k)|256 |82.86|96.28|86.6 |15.6 |30.8 |951 |
|[resnext101_32x8d.tv2_in1k](https://huggingface.co/timm/resnext101_32x8d.tv2_in1k)|224 |82.83|96.22|88.8 |16.5 |31.2 |1099 |
|[resnet152.a1h_in1k](https://huggingface.co/timm/resnet152.a1h_in1k)|224 |82.8 |96.13|60.2 |11.6 |22.6 |1486 |
|[resnet101.a1h_in1k](https://huggingface.co/timm/resnet101.a1h_in1k)|288 |82.8 |96.32|44.6 |13.0 |26.8 |1291 |
|[resnet152.a1_in1k](https://huggingface.co/timm/resnet152.a1_in1k)|288 |82.74|95.71|60.2 |19.1 |37.3 |905 |
|[resnext101_32x8d.fb_wsl_ig1b_ft_in1k](https://huggingface.co/timm/resnext101_32x8d.fb_wsl_ig1b_ft_in1k)|224 |82.69|96.63|88.8 |16.5 |31.2 |1100 |
|[resnet152.a2_in1k](https://huggingface.co/timm/resnet152.a2_in1k)|288 |82.62|95.75|60.2 |19.1 |37.3 |904 |
|[resnetaa50d.sw_in12k_ft_in1k](https://huggingface.co/timm/resnetaa50d.sw_in12k_ft_in1k)|288 |82.61|96.49|25.6 |8.9 |20.6 |1729 |
|[resnet61q.ra2_in1k](https://huggingface.co/timm/resnet61q.ra2_in1k)|288 |82.53|96.13|36.8 |9.9 |21.5 |1773 |
|[wide_resnet101_2.tv2_in1k](https://huggingface.co/timm/wide_resnet101_2.tv2_in1k)|224 |82.5 |96.02|126.9 |22.8 |21.2 |1078 |
|[resnext101_64x4d.c1_in1k](https://huggingface.co/timm/resnext101_64x4d.c1_in1k)|224 |82.46|95.92|83.5 |15.5 |31.2 |987 |
|[resnet51q.ra2_in1k](https://huggingface.co/timm/resnet51q.ra2_in1k)|288 |82.36|96.18|35.7 |8.1 |20.9 |1964 |
|[ecaresnet50t.ra2_in1k](https://huggingface.co/timm/ecaresnet50t.ra2_in1k)|320 |82.35|96.14|25.6 |8.8 |24.1 |1386 |
|[resnet101.a1_in1k](https://huggingface.co/timm/resnet101.a1_in1k)|288 |82.31|95.63|44.6 |13.0 |26.8 |1291 |
|[resnetrs101.tf_in1k](https://huggingface.co/timm/resnetrs101.tf_in1k)|288 |82.29|96.01|63.6 |13.6 |28.5 |1078 |
|[resnet152.tv2_in1k](https://huggingface.co/timm/resnet152.tv2_in1k)|224 |82.29|96.0 |60.2 |11.6 |22.6 |1484 |
|[wide_resnet50_2.racm_in1k](https://huggingface.co/timm/wide_resnet50_2.racm_in1k)|288 |82.27|96.06|68.9 |18.9 |23.8 |1176 |
|[resnet101d.ra2_in1k](https://huggingface.co/timm/resnet101d.ra2_in1k)|256 |82.26|96.07|44.6 |10.6 |22.2 |1542 |
|[resnet101.a2_in1k](https://huggingface.co/timm/resnet101.a2_in1k)|288 |82.24|95.73|44.6 |13.0 |26.8 |1290 |
|[seresnext50_32x4d.racm_in1k](https://huggingface.co/timm/seresnext50_32x4d.racm_in1k)|288 |82.2 |96.14|27.6 |7.0 |23.8 |1547 |
|[ecaresnet101d.miil_in1k](https://huggingface.co/timm/ecaresnet101d.miil_in1k)|224 |82.18|96.05|44.6 |8.1 |17.1 |1771 |
|[resnext50_32x4d.fb_swsl_ig1b_ft_in1k](https://huggingface.co/timm/resnext50_32x4d.fb_swsl_ig1b_ft_in1k)|224 |82.17|96.22|25.0 |4.3 |14.4 |2943 |
|[ecaresnet50t.a1_in1k](https://huggingface.co/timm/ecaresnet50t.a1_in1k)|288 |82.12|95.65|25.6 |7.1 |19.6 |1704 |
|[resnext50_32x4d.a1h_in1k](https://huggingface.co/timm/resnext50_32x4d.a1h_in1k)|288 |82.03|95.94|25.0 |7.0 |23.8 |1745 |
|[ecaresnet101d_pruned.miil_in1k](https://huggingface.co/timm/ecaresnet101d_pruned.miil_in1k)|288 |82.0 |96.15|24.9 |5.8 |12.7 |1787 |
|[resnet61q.ra2_in1k](https://huggingface.co/timm/resnet61q.ra2_in1k)|256 |81.99|95.85|36.8 |7.8 |17.0 |2230 |
|[resnext101_32x8d.tv2_in1k](https://huggingface.co/timm/resnext101_32x8d.tv2_in1k)|176 |81.98|95.72|88.8 |10.3 |19.4 |1768 |
|[resnet152.a1_in1k](https://huggingface.co/timm/resnet152.a1_in1k)|224 |81.97|95.24|60.2 |11.6 |22.6 |1486 |
|[resnet101.a1h_in1k](https://huggingface.co/timm/resnet101.a1h_in1k)|224 |81.93|95.75|44.6 |7.8 |16.2 |2122 |
|[resnet101.tv2_in1k](https://huggingface.co/timm/resnet101.tv2_in1k)|224 |81.9 |95.77|44.6 |7.8 |16.2 |2118 |
|[resnext101_32x16d.fb_ssl_yfcc100m_ft_in1k](https://huggingface.co/timm/resnext101_32x16d.fb_ssl_yfcc100m_ft_in1k)|224 |81.84|96.1 |194.0 |36.3 |51.2 |583 |
|[resnet51q.ra2_in1k](https://huggingface.co/timm/resnet51q.ra2_in1k)|256 |81.78|95.94|35.7 |6.4 |16.6 |2471 |
|[resnet152.a2_in1k](https://huggingface.co/timm/resnet152.a2_in1k)|224 |81.77|95.22|60.2 |11.6 |22.6 |1485 |
|[resnetaa50d.sw_in12k_ft_in1k](https://huggingface.co/timm/resnetaa50d.sw_in12k_ft_in1k)|224 |81.74|96.06|25.6 |5.4 |12.4 |2813 |
|[ecaresnet50t.a2_in1k](https://huggingface.co/timm/ecaresnet50t.a2_in1k)|288 |81.65|95.54|25.6 |7.1 |19.6 |1703 |
|[ecaresnet50d.miil_in1k](https://huggingface.co/timm/ecaresnet50d.miil_in1k)|288 |81.64|95.88|25.6 |7.2 |19.7 |1694 |
|[resnext101_32x8d.fb_ssl_yfcc100m_ft_in1k](https://huggingface.co/timm/resnext101_32x8d.fb_ssl_yfcc100m_ft_in1k)|224 |81.62|96.04|88.8 |16.5 |31.2 |1101 |
|[wide_resnet50_2.tv2_in1k](https://huggingface.co/timm/wide_resnet50_2.tv2_in1k)|224 |81.61|95.76|68.9 |11.4 |14.4 |1930 |
|[resnetaa50.a1h_in1k](https://huggingface.co/timm/resnetaa50.a1h_in1k)|288 |81.61|95.83|25.6 |8.5 |19.2 |1868 |
|[resnet101.a1_in1k](https://huggingface.co/timm/resnet101.a1_in1k)|224 |81.5 |95.16|44.6 |7.8 |16.2 |2125 |
|[resnext50_32x4d.a1_in1k](https://huggingface.co/timm/resnext50_32x4d.a1_in1k)|288 |81.48|95.16|25.0 |7.0 |23.8 |1745 |
|[gcresnet50t.ra2_in1k](https://huggingface.co/timm/gcresnet50t.ra2_in1k)|288 |81.47|95.71|25.9 |6.9 |18.6 |2071 |
|[wide_resnet50_2.racm_in1k](https://huggingface.co/timm/wide_resnet50_2.racm_in1k)|224 |81.45|95.53|68.9 |11.4 |14.4 |1929 |
|[resnet50d.a1_in1k](https://huggingface.co/timm/resnet50d.a1_in1k)|288 |81.44|95.22|25.6 |7.2 |19.7 |1908 |
|[ecaresnet50t.ra2_in1k](https://huggingface.co/timm/ecaresnet50t.ra2_in1k)|256 |81.44|95.67|25.6 |5.6 |15.4 |2168 |
|[ecaresnetlight.miil_in1k](https://huggingface.co/timm/ecaresnetlight.miil_in1k)|288 |81.4 |95.82|30.2 |6.8 |13.9 |2132 |
|[resnet50d.ra2_in1k](https://huggingface.co/timm/resnet50d.ra2_in1k)|288 |81.37|95.74|25.6 |7.2 |19.7 |1910 |
|[resnet101.a2_in1k](https://huggingface.co/timm/resnet101.a2_in1k)|224 |81.32|95.19|44.6 |7.8 |16.2 |2125 |
|[seresnet50.ra2_in1k](https://huggingface.co/timm/seresnet50.ra2_in1k)|288 |81.3 |95.65|28.1 |6.8 |18.4 |1803 |
|[resnext50_32x4d.a2_in1k](https://huggingface.co/timm/resnext50_32x4d.a2_in1k)|288 |81.3 |95.11|25.0 |7.0 |23.8 |1746 |
|[seresnext50_32x4d.racm_in1k](https://huggingface.co/timm/seresnext50_32x4d.racm_in1k)|224 |81.27|95.62|27.6 |4.3 |14.4 |2591 |
|[ecaresnet50t.a1_in1k](https://huggingface.co/timm/ecaresnet50t.a1_in1k)|224 |81.26|95.16|25.6 |4.3 |11.8 |2823 |
|[gcresnext50ts.ch_in1k](https://huggingface.co/timm/gcresnext50ts.ch_in1k)|288 |81.23|95.54|15.7 |4.8 |19.6 |2117 |
|[senet154.gluon_in1k](https://huggingface.co/timm/senet154.gluon_in1k)|224 |81.23|95.35|115.1 |20.8 |38.7 |545 |
|[resnet50.a1_in1k](https://huggingface.co/timm/resnet50.a1_in1k)|288 |81.22|95.11|25.6 |6.8 |18.4 |2089 |
|[resnet50_gn.a1h_in1k](https://huggingface.co/timm/resnet50_gn.a1h_in1k)|288 |81.22|95.63|25.6 |6.8 |18.4 |676 |
|[resnet50d.a2_in1k](https://huggingface.co/timm/resnet50d.a2_in1k)|288 |81.18|95.09|25.6 |7.2 |19.7 |1908 |
|[resnet50.fb_swsl_ig1b_ft_in1k](https://huggingface.co/timm/resnet50.fb_swsl_ig1b_ft_in1k)|224 |81.18|95.98|25.6 |4.1 |11.1 |3455 |
|[resnext50_32x4d.tv2_in1k](https://huggingface.co/timm/resnext50_32x4d.tv2_in1k)|224 |81.17|95.34|25.0 |4.3 |14.4 |2933 |
|[resnext50_32x4d.a1h_in1k](https://huggingface.co/timm/resnext50_32x4d.a1h_in1k)|224 |81.1 |95.33|25.0 |4.3 |14.4 |2934 |
|[seresnet50.a2_in1k](https://huggingface.co/timm/seresnet50.a2_in1k)|288 |81.1 |95.23|28.1 |6.8 |18.4 |1801 |
|[seresnet50.a1_in1k](https://huggingface.co/timm/seresnet50.a1_in1k)|288 |81.1 |95.12|28.1 |6.8 |18.4 |1799 |
|[resnet152s.gluon_in1k](https://huggingface.co/timm/resnet152s.gluon_in1k)|224 |81.02|95.41|60.3 |12.9 |25.0 |1347 |
|[resnet50.d_in1k](https://huggingface.co/timm/resnet50.d_in1k)|288 |80.97|95.44|25.6 |6.8 |18.4 |2085 |
|[gcresnet50t.ra2_in1k](https://huggingface.co/timm/gcresnet50t.ra2_in1k)|256 |80.94|95.45|25.9 |5.4 |14.7 |2571 |
|[resnext101_32x4d.fb_ssl_yfcc100m_ft_in1k](https://huggingface.co/timm/resnext101_32x4d.fb_ssl_yfcc100m_ft_in1k)|224 |80.93|95.73|44.2 |8.0 |21.2 |1814 |
|[resnet50.c1_in1k](https://huggingface.co/timm/resnet50.c1_in1k)|288 |80.91|95.55|25.6 |6.8 |18.4 |2084 |
|[seresnext101_32x4d.gluon_in1k](https://huggingface.co/timm/seresnext101_32x4d.gluon_in1k)|224 |80.9 |95.31|49.0 |8.0 |21.3 |1585 |
|[seresnext101_64x4d.gluon_in1k](https://huggingface.co/timm/seresnext101_64x4d.gluon_in1k)|224 |80.9 |95.3 |88.2 |15.5 |31.2 |918 |
|[resnet50.c2_in1k](https://huggingface.co/timm/resnet50.c2_in1k)|288 |80.86|95.52|25.6 |6.8 |18.4 |2085 |
|[resnet50.tv2_in1k](https://huggingface.co/timm/resnet50.tv2_in1k)|224 |80.85|95.43|25.6 |4.1 |11.1 |3450 |
|[ecaresnet50t.a2_in1k](https://huggingface.co/timm/ecaresnet50t.a2_in1k)|224 |80.84|95.02|25.6 |4.3 |11.8 |2821 |
|[ecaresnet101d_pruned.miil_in1k](https://huggingface.co/timm/ecaresnet101d_pruned.miil_in1k)|224 |80.79|95.62|24.9 |3.5 |7.7 |2961 |
|[seresnet33ts.ra2_in1k](https://huggingface.co/timm/seresnet33ts.ra2_in1k)|288 |80.79|95.36|19.8 |6.0 |14.8 |2506 |
|[ecaresnet50d_pruned.miil_in1k](https://huggingface.co/timm/ecaresnet50d_pruned.miil_in1k)|288 |80.79|95.58|19.9 |4.2 |10.6 |2349 |
|[resnet50.a2_in1k](https://huggingface.co/timm/resnet50.a2_in1k)|288 |80.78|94.99|25.6 |6.8 |18.4 |2088 |
|[resnet50.b1k_in1k](https://huggingface.co/timm/resnet50.b1k_in1k)|288 |80.71|95.43|25.6 |6.8 |18.4 |2087 |
|[resnext50_32x4d.ra_in1k](https://huggingface.co/timm/resnext50_32x4d.ra_in1k)|288 |80.7 |95.39|25.0 |7.0 |23.8 |1749 |
|[resnetrs101.tf_in1k](https://huggingface.co/timm/resnetrs101.tf_in1k)|192 |80.69|95.24|63.6 |6.0 |12.7 |2270 |
|[resnet50d.a1_in1k](https://huggingface.co/timm/resnet50d.a1_in1k)|224 |80.68|94.71|25.6 |4.4 |11.9 |3162 |
|[eca_resnet33ts.ra2_in1k](https://huggingface.co/timm/eca_resnet33ts.ra2_in1k)|288 |80.68|95.36|19.7 |6.0 |14.8 |2637 |
|[resnet50.a1h_in1k](https://huggingface.co/timm/resnet50.a1h_in1k)|224 |80.67|95.3 |25.6 |4.1 |11.1 |3452 |
|[resnext50d_32x4d.bt_in1k](https://huggingface.co/timm/resnext50d_32x4d.bt_in1k)|288 |80.67|95.42|25.0 |7.4 |25.1 |1626 |
|[resnetaa50.a1h_in1k](https://huggingface.co/timm/resnetaa50.a1h_in1k)|224 |80.63|95.21|25.6 |5.2 |11.6 |3034 |
|[ecaresnet50d.miil_in1k](https://huggingface.co/timm/ecaresnet50d.miil_in1k)|224 |80.61|95.32|25.6 |4.4 |11.9 |2813 |
|[resnext101_64x4d.gluon_in1k](https://huggingface.co/timm/resnext101_64x4d.gluon_in1k)|224 |80.61|94.99|83.5 |15.5 |31.2 |989 |
|[gcresnet33ts.ra2_in1k](https://huggingface.co/timm/gcresnet33ts.ra2_in1k)|288 |80.6 |95.31|19.9 |6.0 |14.8 |2578 |
|[gcresnext50ts.ch_in1k](https://huggingface.co/timm/gcresnext50ts.ch_in1k)|256 |80.57|95.17|15.7 |3.8 |15.5 |2710 |
|[resnet152.a3_in1k](https://huggingface.co/timm/resnet152.a3_in1k)|224 |80.56|95.0 |60.2 |11.6 |22.6 |1483 |
|[resnet50d.ra2_in1k](https://huggingface.co/timm/resnet50d.ra2_in1k)|224 |80.53|95.16|25.6 |4.4 |11.9 |3164 |
|[resnext50_32x4d.a1_in1k](https://huggingface.co/timm/resnext50_32x4d.a1_in1k)|224 |80.53|94.46|25.0 |4.3 |14.4 |2930 |
|[wide_resnet101_2.tv2_in1k](https://huggingface.co/timm/wide_resnet101_2.tv2_in1k)|176 |80.48|94.98|126.9 |14.3 |13.2 |1719 |
|[resnet152d.gluon_in1k](https://huggingface.co/timm/resnet152d.gluon_in1k)|224 |80.47|95.2 |60.2 |11.8 |23.4 |1428 |
|[resnet50.b2k_in1k](https://huggingface.co/timm/resnet50.b2k_in1k)|288 |80.45|95.32|25.6 |6.8 |18.4 |2086 |
|[ecaresnetlight.miil_in1k](https://huggingface.co/timm/ecaresnetlight.miil_in1k)|224 |80.45|95.24|30.2 |4.1 |8.4 |3530 |
|[resnext50_32x4d.a2_in1k](https://huggingface.co/timm/resnext50_32x4d.a2_in1k)|224 |80.45|94.63|25.0 |4.3 |14.4 |2936 |
|[wide_resnet50_2.tv2_in1k](https://huggingface.co/timm/wide_resnet50_2.tv2_in1k)|176 |80.43|95.09|68.9 |7.3 |9.0 |3015 |
|[resnet101d.gluon_in1k](https://huggingface.co/timm/resnet101d.gluon_in1k)|224 |80.42|95.01|44.6 |8.1 |17.0 |2007 |
|[resnet50.a1_in1k](https://huggingface.co/timm/resnet50.a1_in1k)|224 |80.38|94.6 |25.6 |4.1 |11.1 |3461 |
|[seresnet33ts.ra2_in1k](https://huggingface.co/timm/seresnet33ts.ra2_in1k)|256 |80.36|95.1 |19.8 |4.8 |11.7 |3267 |
|[resnext101_32x4d.gluon_in1k](https://huggingface.co/timm/resnext101_32x4d.gluon_in1k)|224 |80.34|94.93|44.2 |8.0 |21.2 |1814 |
|[resnext50_32x4d.fb_ssl_yfcc100m_ft_in1k](https://huggingface.co/timm/resnext50_32x4d.fb_ssl_yfcc100m_ft_in1k)|224 |80.32|95.4 |25.0 |4.3 |14.4 |2941 |
|[resnet101s.gluon_in1k](https://huggingface.co/timm/resnet101s.gluon_in1k)|224 |80.28|95.16|44.7 |9.2 |18.6 |1851 |
|[seresnet50.ra2_in1k](https://huggingface.co/timm/seresnet50.ra2_in1k)|224 |80.26|95.08|28.1 |4.1 |11.1 |2972 |
|[resnetblur50.bt_in1k](https://huggingface.co/timm/resnetblur50.bt_in1k)|288 |80.24|95.24|25.6 |8.5 |19.9 |1523 |
|[resnet50d.a2_in1k](https://huggingface.co/timm/resnet50d.a2_in1k)|224 |80.22|94.63|25.6 |4.4 |11.9 |3162 |
|[resnet152.tv2_in1k](https://huggingface.co/timm/resnet152.tv2_in1k)|176 |80.2 |94.64|60.2 |7.2 |14.0 |2346 |
|[seresnet50.a2_in1k](https://huggingface.co/timm/seresnet50.a2_in1k)|224 |80.08|94.74|28.1 |4.1 |11.1 |2969 |
|[eca_resnet33ts.ra2_in1k](https://huggingface.co/timm/eca_resnet33ts.ra2_in1k)|256 |80.08|94.97|19.7 |4.8 |11.7 |3284 |
|[gcresnet33ts.ra2_in1k](https://huggingface.co/timm/gcresnet33ts.ra2_in1k)|256 |80.06|94.99|19.9 |4.8 |11.7 |3216 |
|[resnet50_gn.a1h_in1k](https://huggingface.co/timm/resnet50_gn.a1h_in1k)|224 |80.06|94.95|25.6 |4.1 |11.1 |1109 |
|[seresnet50.a1_in1k](https://huggingface.co/timm/seresnet50.a1_in1k)|224 |80.02|94.71|28.1 |4.1 |11.1 |2962 |
|[resnet50.ram_in1k](https://huggingface.co/timm/resnet50.ram_in1k)|288 |79.97|95.05|25.6 |6.8 |18.4 |2086 |
|[resnet152c.gluon_in1k](https://huggingface.co/timm/resnet152c.gluon_in1k)|224 |79.92|94.84|60.2 |11.8 |23.4 |1455 |
|[seresnext50_32x4d.gluon_in1k](https://huggingface.co/timm/seresnext50_32x4d.gluon_in1k)|224 |79.91|94.82|27.6 |4.3 |14.4 |2591 |
|[resnet50.d_in1k](https://huggingface.co/timm/resnet50.d_in1k)|224 |79.91|94.67|25.6 |4.1 |11.1 |3456 |
|[resnet101.tv2_in1k](https://huggingface.co/timm/resnet101.tv2_in1k)|176 |79.9 |94.6 |44.6 |4.9 |10.1 |3341 |
|[resnetrs50.tf_in1k](https://huggingface.co/timm/resnetrs50.tf_in1k)|224 |79.89|94.97|35.7 |4.5 |12.1 |2774 |
|[resnet50.c2_in1k](https://huggingface.co/timm/resnet50.c2_in1k)|224 |79.88|94.87|25.6 |4.1 |11.1 |3455 |
|[ecaresnet26t.ra2_in1k](https://huggingface.co/timm/ecaresnet26t.ra2_in1k)|320 |79.86|95.07|16.0 |5.2 |16.4 |2168 |
|[resnet50.a2_in1k](https://huggingface.co/timm/resnet50.a2_in1k)|224 |79.85|94.56|25.6 |4.1 |11.1 |3460 |
|[resnet50.ra_in1k](https://huggingface.co/timm/resnet50.ra_in1k)|288 |79.83|94.97|25.6 |6.8 |18.4 |2087 |
|[resnet101.a3_in1k](https://huggingface.co/timm/resnet101.a3_in1k)|224 |79.82|94.62|44.6 |7.8 |16.2 |2114 |
|[resnext50_32x4d.ra_in1k](https://huggingface.co/timm/resnext50_32x4d.ra_in1k)|224 |79.76|94.6 |25.0 |4.3 |14.4 |2943 |
|[resnet50.c1_in1k](https://huggingface.co/timm/resnet50.c1_in1k)|224 |79.74|94.95|25.6 |4.1 |11.1 |3455 |
|[ecaresnet50d_pruned.miil_in1k](https://huggingface.co/timm/ecaresnet50d_pruned.miil_in1k)|224 |79.74|94.87|19.9 |2.5 |6.4 |3929 |
|[resnet33ts.ra2_in1k](https://huggingface.co/timm/resnet33ts.ra2_in1k)|288 |79.71|94.83|19.7 |6.0 |14.8 |2710 |
|[resnet152.gluon_in1k](https://huggingface.co/timm/resnet152.gluon_in1k)|224 |79.68|94.74|60.2 |11.6 |22.6 |1486 |
|[resnext50d_32x4d.bt_in1k](https://huggingface.co/timm/resnext50d_32x4d.bt_in1k)|224 |79.67|94.87|25.0 |4.5 |15.2 |2729 |
|[resnet50.bt_in1k](https://huggingface.co/timm/resnet50.bt_in1k)|288 |79.63|94.91|25.6 |6.8 |18.4 |2086 |
|[ecaresnet50t.a3_in1k](https://huggingface.co/timm/ecaresnet50t.a3_in1k)|224 |79.56|94.72|25.6 |4.3 |11.8 |2805 |
|[resnet101c.gluon_in1k](https://huggingface.co/timm/resnet101c.gluon_in1k)|224 |79.53|94.58|44.6 |8.1 |17.0 |2062 |
|[resnet50.b1k_in1k](https://huggingface.co/timm/resnet50.b1k_in1k)|224 |79.52|94.61|25.6 |4.1 |11.1 |3459 |
|[resnet50.tv2_in1k](https://huggingface.co/timm/resnet50.tv2_in1k)|176 |79.42|94.64|25.6 |2.6 |6.9 |5397 |
|[resnet32ts.ra2_in1k](https://huggingface.co/timm/resnet32ts.ra2_in1k)|288 |79.4 |94.66|18.0 |5.9 |14.6 |2752 |
|[resnet50.b2k_in1k](https://huggingface.co/timm/resnet50.b2k_in1k)|224 |79.38|94.57|25.6 |4.1 |11.1 |3459 |
|[resnext50_32x4d.tv2_in1k](https://huggingface.co/timm/resnext50_32x4d.tv2_in1k)|176 |79.37|94.3 |25.0 |2.7 |9.0 |4577 |
|[resnext50_32x4d.gluon_in1k](https://huggingface.co/timm/resnext50_32x4d.gluon_in1k)|224 |79.36|94.43|25.0 |4.3 |14.4 |2942 |
|[resnext101_32x8d.tv_in1k](https://huggingface.co/timm/resnext101_32x8d.tv_in1k)|224 |79.31|94.52|88.8 |16.5 |31.2 |1100 |
|[resnet101.gluon_in1k](https://huggingface.co/timm/resnet101.gluon_in1k)|224 |79.31|94.53|44.6 |7.8 |16.2 |2125 |
|[resnetblur50.bt_in1k](https://huggingface.co/timm/resnetblur50.bt_in1k)|224 |79.31|94.63|25.6 |5.2 |12.0 |2524 |
|[resnet50.a1h_in1k](https://huggingface.co/timm/resnet50.a1h_in1k)|176 |79.27|94.49|25.6 |2.6 |6.9 |5404 |
|[resnext50_32x4d.a3_in1k](https://huggingface.co/timm/resnext50_32x4d.a3_in1k)|224 |79.25|94.31|25.0 |4.3 |14.4 |2931 |
|[resnet50.fb_ssl_yfcc100m_ft_in1k](https://huggingface.co/timm/resnet50.fb_ssl_yfcc100m_ft_in1k)|224 |79.22|94.84|25.6 |4.1 |11.1 |3451 |
|[resnet33ts.ra2_in1k](https://huggingface.co/timm/resnet33ts.ra2_in1k)|256 |79.21|94.56|19.7 |4.8 |11.7 |3392 |
|[resnet50d.gluon_in1k](https://huggingface.co/timm/resnet50d.gluon_in1k)|224 |79.07|94.48|25.6 |4.4 |11.9 |3162 |
|[resnet50.ram_in1k](https://huggingface.co/timm/resnet50.ram_in1k)|224 |79.03|94.38|25.6 |4.1 |11.1 |3453 |
|[resnet50.am_in1k](https://huggingface.co/timm/resnet50.am_in1k)|224 |79.01|94.39|25.6 |4.1 |11.1 |3461 |
|[resnet32ts.ra2_in1k](https://huggingface.co/timm/resnet32ts.ra2_in1k)|256 |79.01|94.37|18.0 |4.6 |11.6 |3440 |
|[ecaresnet26t.ra2_in1k](https://huggingface.co/timm/ecaresnet26t.ra2_in1k)|256 |78.9 |94.54|16.0 |3.4 |10.5 |3421 |
|[resnet152.a3_in1k](https://huggingface.co/timm/resnet152.a3_in1k)|160 |78.89|94.11|60.2 |5.9 |11.5 |2745 |
|[wide_resnet101_2.tv_in1k](https://huggingface.co/timm/wide_resnet101_2.tv_in1k)|224 |78.84|94.28|126.9 |22.8 |21.2 |1079 |
|[seresnext26d_32x4d.bt_in1k](https://huggingface.co/timm/seresnext26d_32x4d.bt_in1k)|288 |78.83|94.24|16.8 |4.5 |16.8 |2251 |
|[resnet50.ra_in1k](https://huggingface.co/timm/resnet50.ra_in1k)|224 |78.81|94.32|25.6 |4.1 |11.1 |3454 |
|[seresnext26t_32x4d.bt_in1k](https://huggingface.co/timm/seresnext26t_32x4d.bt_in1k)|288 |78.74|94.33|16.8 |4.5 |16.7 |2264 |
|[resnet50s.gluon_in1k](https://huggingface.co/timm/resnet50s.gluon_in1k)|224 |78.72|94.23|25.7 |5.5 |13.5 |2796 |
|[resnet50d.a3_in1k](https://huggingface.co/timm/resnet50d.a3_in1k)|224 |78.71|94.24|25.6 |4.4 |11.9 |3154 |
|[wide_resnet50_2.tv_in1k](https://huggingface.co/timm/wide_resnet50_2.tv_in1k)|224 |78.47|94.09|68.9 |11.4 |14.4 |1934 |
|[resnet50.bt_in1k](https://huggingface.co/timm/resnet50.bt_in1k)|224 |78.46|94.27|25.6 |4.1 |11.1 |3454 |
|[resnet34d.ra2_in1k](https://huggingface.co/timm/resnet34d.ra2_in1k)|288 |78.43|94.35|21.8 |6.5 |7.5 |3291 |
|[gcresnext26ts.ch_in1k](https://huggingface.co/timm/gcresnext26ts.ch_in1k)|288 |78.42|94.04|10.5 |3.1 |13.3 |3226 |
|[resnet26t.ra2_in1k](https://huggingface.co/timm/resnet26t.ra2_in1k)|320 |78.33|94.13|16.0 |5.2 |16.4 |2391 |
|[resnet152.tv_in1k](https://huggingface.co/timm/resnet152.tv_in1k)|224 |78.32|94.04|60.2 |11.6 |22.6 |1487 |
|[seresnext26ts.ch_in1k](https://huggingface.co/timm/seresnext26ts.ch_in1k)|288 |78.28|94.1 |10.4 |3.1 |13.3 |3062 |
|[bat_resnext26ts.ch_in1k](https://huggingface.co/timm/bat_resnext26ts.ch_in1k)|256 |78.25|94.1 |10.7 |2.5 |12.5 |3393 |
|[resnet50.a3_in1k](https://huggingface.co/timm/resnet50.a3_in1k)|224 |78.06|93.78|25.6 |4.1 |11.1 |3450 |
|[resnet50c.gluon_in1k](https://huggingface.co/timm/resnet50c.gluon_in1k)|224 |78.0 |93.99|25.6 |4.4 |11.9 |3286 |
|[eca_resnext26ts.ch_in1k](https://huggingface.co/timm/eca_resnext26ts.ch_in1k)|288 |78.0 |93.91|10.3 |3.1 |13.3 |3297 |
|[seresnext26t_32x4d.bt_in1k](https://huggingface.co/timm/seresnext26t_32x4d.bt_in1k)|224 |77.98|93.75|16.8 |2.7 |10.1 |3841 |
|[resnet34.a1_in1k](https://huggingface.co/timm/resnet34.a1_in1k)|288 |77.92|93.77|21.8 |6.1 |6.2 |3609 |
|[resnet101.a3_in1k](https://huggingface.co/timm/resnet101.a3_in1k)|160 |77.88|93.71|44.6 |4.0 |8.3 |3926 |
|[resnet26t.ra2_in1k](https://huggingface.co/timm/resnet26t.ra2_in1k)|256 |77.87|93.84|16.0 |3.4 |10.5 |3772 |
|[seresnext26ts.ch_in1k](https://huggingface.co/timm/seresnext26ts.ch_in1k)|256 |77.86|93.79|10.4 |2.4 |10.5 |4263 |
|[resnetrs50.tf_in1k](https://huggingface.co/timm/resnetrs50.tf_in1k)|160 |77.82|93.81|35.7 |2.3 |6.2 |5238 |
|[gcresnext26ts.ch_in1k](https://huggingface.co/timm/gcresnext26ts.ch_in1k)|256 |77.81|93.82|10.5 |2.4 |10.5 |4183 |
|[ecaresnet50t.a3_in1k](https://huggingface.co/timm/ecaresnet50t.a3_in1k)|160 |77.79|93.6 |25.6 |2.2 |6.0 |5329 |
|[resnext50_32x4d.a3_in1k](https://huggingface.co/timm/resnext50_32x4d.a3_in1k)|160 |77.73|93.32|25.0 |2.2 |7.4 |5576 |
|[resnext50_32x4d.tv_in1k](https://huggingface.co/timm/resnext50_32x4d.tv_in1k)|224 |77.61|93.7 |25.0 |4.3 |14.4 |2944 |
|[seresnext26d_32x4d.bt_in1k](https://huggingface.co/timm/seresnext26d_32x4d.bt_in1k)|224 |77.59|93.61|16.8 |2.7 |10.2 |3807 |
|[resnet50.gluon_in1k](https://huggingface.co/timm/resnet50.gluon_in1k)|224 |77.58|93.72|25.6 |4.1 |11.1 |3455 |
|[eca_resnext26ts.ch_in1k](https://huggingface.co/timm/eca_resnext26ts.ch_in1k)|256 |77.44|93.56|10.3 |2.4 |10.5 |4284 |
|[resnet26d.bt_in1k](https://huggingface.co/timm/resnet26d.bt_in1k)|288 |77.41|93.63|16.0 |4.3 |13.5 |2907 |
|[resnet101.tv_in1k](https://huggingface.co/timm/resnet101.tv_in1k)|224 |77.38|93.54|44.6 |7.8 |16.2 |2125 |
|[resnet50d.a3_in1k](https://huggingface.co/timm/resnet50d.a3_in1k)|160 |77.22|93.27|25.6 |2.2 |6.1 |5982 |
|[resnext26ts.ra2_in1k](https://huggingface.co/timm/resnext26ts.ra2_in1k)|288 |77.17|93.47|10.3 |3.1 |13.3 |3392 |
|[resnet34.a2_in1k](https://huggingface.co/timm/resnet34.a2_in1k)|288 |77.15|93.27|21.8 |6.1 |6.2 |3615 |
|[resnet34d.ra2_in1k](https://huggingface.co/timm/resnet34d.ra2_in1k)|224 |77.1 |93.37|21.8 |3.9 |4.5 |5436 |
|[seresnet50.a3_in1k](https://huggingface.co/timm/seresnet50.a3_in1k)|224 |77.02|93.07|28.1 |4.1 |11.1 |2952 |
|[resnext26ts.ra2_in1k](https://huggingface.co/timm/resnext26ts.ra2_in1k)|256 |76.78|93.13|10.3 |2.4 |10.5 |4410 |
|[resnet26d.bt_in1k](https://huggingface.co/timm/resnet26d.bt_in1k)|224 |76.7 |93.17|16.0 |2.6 |8.2 |4859 |
|[resnet34.bt_in1k](https://huggingface.co/timm/resnet34.bt_in1k)|288 |76.5 |93.35|21.8 |6.1 |6.2 |3617 |
|[resnet34.a1_in1k](https://huggingface.co/timm/resnet34.a1_in1k)|224 |76.42|92.87|21.8 |3.7 |3.7 |5984 |
|[resnet26.bt_in1k](https://huggingface.co/timm/resnet26.bt_in1k)|288 |76.35|93.18|16.0 |3.9 |12.2 |3331 |
|[resnet50.tv_in1k](https://huggingface.co/timm/resnet50.tv_in1k)|224 |76.13|92.86|25.6 |4.1 |11.1 |3457 |
|[resnet50.a3_in1k](https://huggingface.co/timm/resnet50.a3_in1k)|160 |75.96|92.5 |25.6 |2.1 |5.7 |6490 |
|[resnet34.a2_in1k](https://huggingface.co/timm/resnet34.a2_in1k)|224 |75.52|92.44|21.8 |3.7 |3.7 |5991 |
|[resnet26.bt_in1k](https://huggingface.co/timm/resnet26.bt_in1k)|224 |75.3 |92.58|16.0 |2.4 |7.4 |5583 |
|[resnet34.bt_in1k](https://huggingface.co/timm/resnet34.bt_in1k)|224 |75.16|92.18|21.8 |3.7 |3.7 |5994 |
|[seresnet50.a3_in1k](https://huggingface.co/timm/seresnet50.a3_in1k)|160 |75.1 |92.08|28.1 |2.1 |5.7 |5513 |
|[resnet34.gluon_in1k](https://huggingface.co/timm/resnet34.gluon_in1k)|224 |74.57|91.98|21.8 |3.7 |3.7 |5984 |
|[resnet18d.ra2_in1k](https://huggingface.co/timm/resnet18d.ra2_in1k)|288 |73.81|91.83|11.7 |3.4 |5.4 |5196 |
|[resnet34.tv_in1k](https://huggingface.co/timm/resnet34.tv_in1k)|224 |73.32|91.42|21.8 |3.7 |3.7 |5979 |
|[resnet18.fb_swsl_ig1b_ft_in1k](https://huggingface.co/timm/resnet18.fb_swsl_ig1b_ft_in1k)|224 |73.28|91.73|11.7 |1.8 |2.5 |10213 |
|[resnet18.a1_in1k](https://huggingface.co/timm/resnet18.a1_in1k)|288 |73.16|91.03|11.7 |3.0 |4.1 |6050 |
|[resnet34.a3_in1k](https://huggingface.co/timm/resnet34.a3_in1k)|224 |72.98|91.11|21.8 |3.7 |3.7 |5967 |
|[resnet18.fb_ssl_yfcc100m_ft_in1k](https://huggingface.co/timm/resnet18.fb_ssl_yfcc100m_ft_in1k)|224 |72.6 |91.42|11.7 |1.8 |2.5 |10213 |
|[resnet18.a2_in1k](https://huggingface.co/timm/resnet18.a2_in1k)|288 |72.37|90.59|11.7 |3.0 |4.1 |6051 |
|[resnet14t.c3_in1k](https://huggingface.co/timm/resnet14t.c3_in1k)|224 |72.26|90.31|10.1 |1.7 |5.8 |7026 |
|[resnet18d.ra2_in1k](https://huggingface.co/timm/resnet18d.ra2_in1k)|224 |72.26|90.68|11.7 |2.1 |3.3 |8707 |
|[resnet18.a1_in1k](https://huggingface.co/timm/resnet18.a1_in1k)|224 |71.49|90.07|11.7 |1.8 |2.5 |10187 |
|[resnet14t.c3_in1k](https://huggingface.co/timm/resnet14t.c3_in1k)|176 |71.31|89.69|10.1 |1.1 |3.6 |10970 |
|[resnet18.gluon_in1k](https://huggingface.co/timm/resnet18.gluon_in1k)|224 |70.84|89.76|11.7 |1.8 |2.5 |10210 |
|[resnet18.a2_in1k](https://huggingface.co/timm/resnet18.a2_in1k)|224 |70.64|89.47|11.7 |1.8 |2.5 |10194 |
|[resnet34.a3_in1k](https://huggingface.co/timm/resnet34.a3_in1k)|160 |70.56|89.52|21.8 |1.9 |1.9 |10737 |
|[resnet18.tv_in1k](https://huggingface.co/timm/resnet18.tv_in1k)|224 |69.76|89.07|11.7 |1.8 |2.5 |10205 |
|[resnet10t.c3_in1k](https://huggingface.co/timm/resnet10t.c3_in1k)|224 |68.34|88.03|5.4 |1.1 |2.4 |13079 |
|[resnet18.a3_in1k](https://huggingface.co/timm/resnet18.a3_in1k)|224 |68.25|88.17|11.7 |1.8 |2.5 |10167 |
|[resnet10t.c3_in1k](https://huggingface.co/timm/resnet10t.c3_in1k)|176 |66.71|86.96|5.4 |0.7 |1.5 |20327 |
|[resnet18.a3_in1k](https://huggingface.co/timm/resnet18.a3_in1k)|160 |65.66|86.26|11.7 |0.9 |1.3 |18229 |
## Citation
```bibtex
@inproceedings{wslimageseccv2018,
title={Exploring the Limits of Weakly Supervised Pretraining},
author={Dhruv Kumar Mahajan and Ross B. Girshick and Vignesh Ramanathan and Kaiming He and Manohar Paluri and Yixuan Li and Ashwin Bharambe and Laurens van der Maaten},
booktitle={ECCV},
year={2018}
}
```
```bibtex
@article{Xie2016,
title={Aggregated Residual Transformations for Deep Neural Networks},
author={Saining Xie and Ross Girshick and Piotr Dollár and Zhuowen Tu and Kaiming He},
journal={arXiv preprint arXiv:1611.05431},
year={2016}
}
```
```bibtex
@article{He2015,
author = {Kaiming He and Xiangyu Zhang and Shaoqing Ren and Jian Sun},
title = {Deep Residual Learning for Image Recognition},
journal = {arXiv preprint arXiv:1512.03385},
year = {2015}
}
```
```bibtex
@misc{rw2019timm,
author = {Ross Wightman},
title = {PyTorch Image Models},
year = {2019},
publisher = {GitHub},
journal = {GitHub repository},
doi = {10.5281/zenodo.4414861},
howpublished = {\url{https://github.com/huggingface/pytorch-image-models}}
}
```
|
John6666/photo-realistic-pony-v5-sdxl | John6666 | "2024-06-09T00:44:19Z" | 8,986 | 3 | diffusers | [
"diffusers",
"safetensors",
"text-to-image",
"stable-diffusion",
"stable-diffusion-xl",
"realistic",
"photorealistic",
"pony",
"license:creativeml-openrail-m",
"autotrain_compatible",
"endpoints_compatible",
"diffusers:StableDiffusionXLPipeline",
"region:us"
] | text-to-image | "2024-06-09T00:39:30Z" | ---
license: creativeml-openrail-m
tags:
- text-to-image
- stable-diffusion
- stable-diffusion-xl
- realistic
- photorealistic
- pony
---
Original model is [here](https://civitai.com/models/486551/photo-realistic-pony?modelVersionId=559772).
|
BAAI/bge-reranker-v2-gemma | BAAI | "2024-03-19T09:26:32Z" | 8,976 | 22 | sentence-transformers | [
"sentence-transformers",
"safetensors",
"gemma",
"text-generation",
"transformers",
"text-classification",
"multilingual",
"arxiv:2312.15503",
"arxiv:2402.03216",
"license:apache-2.0",
"region:us"
] | text-classification | "2024-03-16T12:09:04Z" | ---
license: apache-2.0
pipeline_tag: text-classification
tags:
- transformers
- sentence-transformers
language:
- multilingual
---
# Reranker
**More details please refer to our Github: [FlagEmbedding](https://github.com/FlagOpen/FlagEmbedding/tree/master).**
- [Model List](#model-list)
- [Usage](#usage)
- [Fine-tuning](#fine-tune)
- [Evaluation](#evaluation)
- [Citation](#citation)
Different from embedding model, reranker uses question and document as input and directly output similarity instead of embedding.
You can get a relevance score by inputting query and passage to the reranker.
And the score can be mapped to a float value in [0,1] by sigmoid function.
## Model List
| Model | Base model | Language | layerwise | feature |
|:--------------------------------------------------------------------------|:--------:|:-----------------------------------------------------------------------------------------------------------------------------------:|:----------------------------------------------------------------------------------------------:|:----------------------------------------------------------------------------------------------:|
| [BAAI/bge-reranker-base](https://huggingface.co/BAAI/bge-reranker-base) | [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) | Chinese and English | - | Lightweight reranker model, easy to deploy, with fast inference. |
| [BAAI/bge-reranker-large](https://huggingface.co/BAAI/bge-reranker-large) | [xlm-roberta-large](https://huggingface.co/FacebookAI/xlm-roberta-large) | Chinese and English | - | Lightweight reranker model, easy to deploy, with fast inference. |
| [BAAI/bge-reranker-v2-m3](https://huggingface.co/BAAI/bge-reranker-v2-m3) | [bge-m3](https://huggingface.co/BAAI/bge-m3) | Multilingual | - | Lightweight reranker model, possesses strong multilingual capabilities, easy to deploy, with fast inference. |
| [BAAI/bge-reranker-v2-gemma](https://huggingface.co/BAAI/bge-reranker-v2-gemma) | [gemma-2b](https://huggingface.co/google/gemma-2b) | Multilingual | - | Suitable for multilingual contexts, performs well in both English proficiency and multilingual capabilities. |
| [BAAI/bge-reranker-v2-minicpm-layerwise](https://huggingface.co/BAAI/bge-reranker-v2-minicpm-layerwise) | [MiniCPM-2B-dpo-bf16](https://huggingface.co/openbmb/MiniCPM-2B-dpo-bf16) | Multilingual | 8-40 | Suitable for multilingual contexts, performs well in both English and Chinese proficiency, allows freedom to select layers for output, facilitating accelerated inference. |
You can select the model according your senario and resource.
- For **multilingual**, utilize [BAAI/bge-reranker-v2-m3](https://huggingface.co/BAAI/bge-reranker-v2-m3) and [BAAI/bge-reranker-v2-gemma](https://huggingface.co/BAAI/bge-reranker-v2-gemma)
- For **Chinese or English**, utilize [BAAI/bge-reranker-v2-m3](https://huggingface.co/BAAI/bge-reranker-v2-m3) and [BAAI/bge-reranker-v2-minicpm-layerwise](https://huggingface.co/BAAI/bge-reranker-v2-minicpm-layerwise).
- For **efficiency**, utilize [BAAI/bge-reranker-v2-m3](https://huggingface.co/BAAI/bge-reranker-v2-m3) and the low layer of [BAAI/bge-reranker-v2-minicpm-layerwise](https://huggingface.co/BAAI/bge-reranker-v2-minicpm-layerwise).
- For better performance, recommand [BAAI/bge-reranker-v2-minicpm-layerwise](https://huggingface.co/BAAI/bge-reranker-v2-minicpm-layerwise) and [BAAI/bge-reranker-v2-gemma](https://huggingface.co/BAAI/bge-reranker-v2-gemma)
## Usage
### Using FlagEmbedding
```
pip install -U FlagEmbedding
```
#### For normal reranker (bge-reranker-base / bge-reranker-large / bge-reranker-v2-m3 )
Get relevance scores (higher scores indicate more relevance):
```python
from FlagEmbedding import FlagReranker
reranker = FlagReranker('BAAI/bge-reranker-v2-m3', use_fp16=True) # Setting use_fp16 to True speeds up computation with a slight performance degradation
score = reranker.compute_score(['query', 'passage'])
print(score) # -5.65234375
# You can map the scores into 0-1 by set "normalize=True", which will apply sigmoid function to the score
score = reranker.compute_score(['query', 'passage'], normalize=True)
print(score) # 0.003497010252573502
scores = reranker.compute_score([['what is panda?', 'hi'], ['what is panda?', 'The giant panda (Ailuropoda melanoleuca), sometimes called a panda bear or simply panda, is a bear species endemic to China.']])
print(scores) # [-8.1875, 5.26171875]
# You can map the scores into 0-1 by set "normalize=True", which will apply sigmoid function to the score
scores = reranker.compute_score([['what is panda?', 'hi'], ['what is panda?', 'The giant panda (Ailuropoda melanoleuca), sometimes called a panda bear or simply panda, is a bear species endemic to China.']], normalize=True)
print(scores) # [0.00027803096387751553, 0.9948403768236574]
```
#### For LLM-based reranker
```python
from FlagEmbedding import FlagLLMReranker
reranker = FlagLLMReranker('BAAI/bge-reranker-v2-gemma', use_fp16=True) # Setting use_fp16 to True speeds up computation with a slight performance degradation
# reranker = FlagLLMReranker('BAAI/bge-reranker-v2-gemma', use_bf16=True) # You can also set use_bf16=True to speed up computation with a slight performance degradation
score = reranker.compute_score(['query', 'passage'])
print(score)
scores = reranker.compute_score([['what is panda?', 'hi'], ['what is panda?', 'The giant panda (Ailuropoda melanoleuca), sometimes called a panda bear or simply panda, is a bear species endemic to China.']])
print(scores)
```
#### For LLM-based layerwise reranker
```python
from FlagEmbedding import LayerWiseFlagLLMReranker
reranker = LayerWiseFlagLLMReranker('BAAI/bge-reranker-v2-minicpm-layerwise', use_fp16=True) # Setting use_fp16 to True speeds up computation with a slight performance degradation
# reranker = LayerWiseFlagLLMReranker('BAAI/bge-reranker-v2-minicpm-layerwise', use_bf16=True) # You can also set use_bf16=True to speed up computation with a slight performance degradation
score = reranker.compute_score(['query', 'passage'], cutoff_layers=[28]) # Adjusting 'cutoff_layers' to pick which layers are used for computing the score.
print(score)
scores = reranker.compute_score([['what is panda?', 'hi'], ['what is panda?', 'The giant panda (Ailuropoda melanoleuca), sometimes called a panda bear or simply panda, is a bear species endemic to China.']], cutoff_layers=[28])
print(scores)
```
### Using Huggingface transformers
#### For normal reranker (bge-reranker-base / bge-reranker-large / bge-reranker-v2-m3 )
Get relevance scores (higher scores indicate more relevance):
```python
import torch
from transformers import AutoModelForSequenceClassification, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained('BAAI/bge-reranker-v2-m3')
model = AutoModelForSequenceClassification.from_pretrained('BAAI/bge-reranker-v2-m3')
model.eval()
pairs = [['what is panda?', 'hi'], ['what is panda?', 'The giant panda (Ailuropoda melanoleuca), sometimes called a panda bear or simply panda, is a bear species endemic to China.']]
with torch.no_grad():
inputs = tokenizer(pairs, padding=True, truncation=True, return_tensors='pt', max_length=512)
scores = model(**inputs, return_dict=True).logits.view(-1, ).float()
print(scores)
```
#### For LLM-based reranker
```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
def get_inputs(pairs, tokenizer, prompt=None, max_length=1024):
if prompt is None:
prompt = "Given a query A and a passage B, determine whether the passage contains an answer to the query by providing a prediction of either 'Yes' or 'No'."
sep = "\n"
prompt_inputs = tokenizer(prompt,
return_tensors=None,
add_special_tokens=False)['input_ids']
sep_inputs = tokenizer(sep,
return_tensors=None,
add_special_tokens=False)['input_ids']
inputs = []
for query, passage in pairs:
query_inputs = tokenizer(f'A: {query}',
return_tensors=None,
add_special_tokens=False,
max_length=max_length * 3 // 4,
truncation=True)
passage_inputs = tokenizer(f'B: {passage}',
return_tensors=None,
add_special_tokens=False,
max_length=max_length,
truncation=True)
item = tokenizer.prepare_for_model(
[tokenizer.bos_token_id] + query_inputs['input_ids'],
sep_inputs + passage_inputs['input_ids'],
truncation='only_second',
max_length=max_length,
padding=False,
return_attention_mask=False,
return_token_type_ids=False,
add_special_tokens=False
)
item['input_ids'] = item['input_ids'] + sep_inputs + prompt_inputs
item['attention_mask'] = [1] * len(item['input_ids'])
inputs.append(item)
return tokenizer.pad(
inputs,
padding=True,
max_length=max_length + len(sep_inputs) + len(prompt_inputs),
pad_to_multiple_of=8,
return_tensors='pt',
)
tokenizer = AutoTokenizer.from_pretrained('BAAI/bge-reranker-v2-gemma')
model = AutoModelForCausalLM.from_pretrained('BAAI/bge-reranker-v2-gemma')
yes_loc = tokenizer('Yes', add_special_tokens=False)['input_ids'][0]
model.eval()
pairs = [['what is panda?', 'hi'], ['what is panda?', 'The giant panda (Ailuropoda melanoleuca), sometimes called a panda bear or simply panda, is a bear species endemic to China.']]
with torch.no_grad():
inputs = get_inputs(pairs, tokenizer)
scores = model(**inputs, return_dict=True).logits[:, -1, yes_loc].view(-1, ).float()
print(scores)
```
#### For LLM-based layerwise reranker
```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
def get_inputs(pairs, tokenizer, prompt=None, max_length=1024):
if prompt is None:
prompt = "Given a query A and a passage B, determine whether the passage contains an answer to the query by providing a prediction of either 'Yes' or 'No'."
sep = "\n"
prompt_inputs = tokenizer(prompt,
return_tensors=None,
add_special_tokens=False)['input_ids']
sep_inputs = tokenizer(sep,
return_tensors=None,
add_special_tokens=False)['input_ids']
inputs = []
for query, passage in pairs:
query_inputs = tokenizer(f'A: {query}',
return_tensors=None,
add_special_tokens=False,
max_length=max_length * 3 // 4,
truncation=True)
passage_inputs = tokenizer(f'B: {passage}',
return_tensors=None,
add_special_tokens=False,
max_length=max_length,
truncation=True)
item = tokenizer.prepare_for_model(
[tokenizer.bos_token_id] + query_inputs['input_ids'],
sep_inputs + passage_inputs['input_ids'],
truncation='only_second',
max_length=max_length,
padding=False,
return_attention_mask=False,
return_token_type_ids=False,
add_special_tokens=False
)
item['input_ids'] = item['input_ids'] + sep_inputs + prompt_inputs
item['attention_mask'] = [1] * len(item['input_ids'])
inputs.append(item)
return tokenizer.pad(
inputs,
padding=True,
max_length=max_length + len(sep_inputs) + len(prompt_inputs),
pad_to_multiple_of=8,
return_tensors='pt',
)
tokenizer = AutoTokenizer.from_pretrained('BAAI/bge-reranker-v2-minicpm-layerwise', trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained('BAAI/bge-reranker-v2-minicpm-layerwise', trust_remote_code=True, torch_dtype=torch.bfloat16)
model = model.to('cuda')
model.eval()
pairs = [['what is panda?', 'hi'], ['what is panda?', 'The giant panda (Ailuropoda melanoleuca), sometimes called a panda bear or simply panda, is a bear species endemic to China.']]
with torch.no_grad():
inputs = get_inputs(pairs, tokenizer).to(model.device)
all_scores = model(**inputs, return_dict=True, cutoff_layers=[28])
all_scores = [scores[:, -1].view(-1, ).float() for scores in all_scores[0]]
print(all_scores)
```
## Fine-tune
### Data Format
Train data should be a json file, where each line is a dict like this:
```
{"query": str, "pos": List[str], "neg":List[str], "prompt": str}
```
`query` is the query, and `pos` is a list of positive texts, `neg` is a list of negative texts, `prompt` indicates the relationship between query and texts. If you have no negative texts for a query, you can random sample some from the entire corpus as the negatives.
See [toy_finetune_data.jsonl](https://github.com/FlagOpen/FlagEmbedding/tree/master/FlagEmbedding/llm_reranker/toy_finetune_data.jsonl) for a toy data file.
### Train
You can fine-tune the reranker with the following code:
**For llm-based reranker**
```shell
torchrun --nproc_per_node {number of gpus} \
-m FlagEmbedding.llm_reranker.finetune_for_instruction.run \
--output_dir {path to save model} \
--model_name_or_path google/gemma-2b \
--train_data ./toy_finetune_data.jsonl \
--learning_rate 2e-4 \
--num_train_epochs 1 \
--per_device_train_batch_size 1 \
--gradient_accumulation_steps 16 \
--dataloader_drop_last True \
--query_max_len 512 \
--passage_max_len 512 \
--train_group_size 16 \
--logging_steps 1 \
--save_steps 2000 \
--save_total_limit 50 \
--ddp_find_unused_parameters False \
--gradient_checkpointing \
--deepspeed stage1.json \
--warmup_ratio 0.1 \
--bf16 \
--use_lora True \
--lora_rank 32 \
--lora_alpha 64 \
--use_flash_attn True \
--target_modules q_proj k_proj v_proj o_proj
```
**For llm-based layerwise reranker**
```shell
torchrun --nproc_per_node {number of gpus} \
-m FlagEmbedding.llm_reranker.finetune_for_layerwise.run \
--output_dir {path to save model} \
--model_name_or_path openbmb/MiniCPM-2B-dpo-bf16 \
--train_data ./toy_finetune_data.jsonl \
--learning_rate 2e-4 \
--num_train_epochs 1 \
--per_device_train_batch_size 1 \
--gradient_accumulation_steps 16 \
--dataloader_drop_last True \
--query_max_len 512 \
--passage_max_len 512 \
--train_group_size 16 \
--logging_steps 1 \
--save_steps 2000 \
--save_total_limit 50 \
--ddp_find_unused_parameters False \
--gradient_checkpointing \
--deepspeed stage1.json \
--warmup_ratio 0.1 \
--bf16 \
--use_lora True \
--lora_rank 32 \
--lora_alpha 64 \
--use_flash_attn True \
--target_modules q_proj k_proj v_proj o_proj \
--start_layer 8 \
--head_multi True \
--head_type simple \
--lora_extra_parameters linear_head
```
Our rerankers are initialized from [google/gemma-2b](https://huggingface.co/google/gemma-2b) (for llm-based reranker) and [openbmb/MiniCPM-2B-dpo-bf16](https://huggingface.co/openbmb/MiniCPM-2B-dpo-bf16) (for llm-based layerwise reranker), and we train it on a mixture of multilingual datasets:
- [bge-m3-data](https://huggingface.co/datasets/Shitao/bge-m3-data)
- [quora train data](https://huggingface.co/datasets/quora)
- [fever train data](https://fever.ai/dataset/fever.html)
## Evaluation
- llama-index.
- BEIR.
rereank the top 100 results from bge-en-v1.5 large.
rereank the top 100 results from e5 mistral 7b instruct.
- CMTEB-retrieval.
It rereank the top 100 results from bge-zh-v1.5 large.
- miracl (multi-language).
It rereank the top 100 results from bge-m3.
## Citation
If you find this repository useful, please consider giving a star and citation
```bibtex
@misc{li2023making,
title={Making Large Language Models A Better Foundation For Dense Retrieval},
author={Chaofan Li and Zheng Liu and Shitao Xiao and Yingxia Shao},
year={2023},
eprint={2312.15503},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
@misc{chen2024bge,
title={BGE M3-Embedding: Multi-Lingual, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation},
author={Jianlv Chen and Shitao Xiao and Peitian Zhang and Kun Luo and Defu Lian and Zheng Liu},
year={2024},
eprint={2402.03216},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
``` |
DavidAU/Psyonic-Cetacean-MythoMax-Prose-Crazy-Ultra-Quality-29B-GGUF | DavidAU | "2024-07-02T05:39:44Z" | 8,976 | 6 | null | [
"gguf",
"creative",
"creative writing",
"fiction writing",
"plot generation",
"sub-plot generation",
"story generation",
"scene continue",
"storytelling",
"fiction story",
"story",
"writing",
"fiction",
"float32",
"roleplaying",
"rp",
"enhanced",
"neo class",
"32 bit upscale",
"en",
"license:apache-2.0",
"region:us"
] | null | "2024-06-29T01:15:29Z" | ---
license: apache-2.0
language:
- en
tags:
- creative
- creative writing
- fiction writing
- plot generation
- sub-plot generation
- fiction writing
- story generation
- scene continue
- storytelling
- fiction story
- story
- writing
- fiction
- float32
- roleplaying
- rp
- enhanced
- neo class
- 32 bit upscale
---
<font color=red><h3> Ultra Quality High Remaster of the incredible: Psyonic-Cetacean-20b + Mythomax 13B MERGED to 29 Billion parameters. </h3></font>
This is a Floating Point 32 upscale, where all components and merges were remastered to floating point 32.
This includes all the merges (recreated with master files), and where possible subbing full FP32 models.
The goal: Carry forward maximum precision right up to the point where it is "GUFFed".
This includes F32 master file for GGUF too... at a whopping 116 GBs.
WHY?
Because the difference between F32 vs BF16 is... over 8 DECIMAL places.
And as each merge / model is modified there are "losses" along the way.
These losses are carried forward and in turn lead to more losses.
And decimal points are critical to model performance.
SMALL?
Yes... but multiplied by each merge(s), and compression(s): 29 billion times.
<B>PROSE CRAZY:</B>
This model is specifically designed for deep, creative prose with the target goal of getting the model to use stronger
and more coherrent levels of detail at all levels as well as expand word choice too without
have to "state" this in prompts or at the prompt level or system role level.
This is version 1 of 3 current versions, with sub-versions as well.
ED2 is here:
[ https://huggingface.co/DavidAU/Psyonic-Cetacean-MythoMax-ED2-Prose-Crazy-Ultra-Quality-29B-GGUF ]
ED3 is here:
[ https://huggingface.co/DavidAU/Psyonic-Cetacean-MythoMax-ED3-Prose-Crazy-Ultra-Quality-29B-GGUF ]
This is a merge between the Ultra Quality Psyonic-Cetacean 20B with the 13B Mythomax model
which ends up at 29 Billion parameters at 90 layers (813 Tensors @ F32).
For reference a 70B model is typically 120 layers, and Command-R 01 35B is 40 layers (but very dense layers).
These models are a "pass-through" merges, meaning that all the unique qualities of all models is preserved in full,
no overwriting or merging of the parameters, weights and so on.
Although this model can be used for many purposes, it is primaryly for creative prose - any function related to this
including plot generation, story generation, scene generation, scene continue (sample provided, used as a starting point),
and just about anything related to fictional writing.
Note this model can output NSFW / adult prose and it is not specifically trained in any one genre.
Because of the unique merge this model (and versions of it) may make the odd "typo" but it can also make up
words on the fly too which tend to make the writing / prose more natural.
This model does not need a specific prompt template.
See prose examples below.
<B>PROSE CRAZY - IMAT13 ("NEO"):</B>
This is an even more extreme version of "prose crazy" version of this model with NEO CLASS process
punching out it's "craziness" to the extreme.
See prose examples below.
<B>PROSE CRAZY - IMAT13 ("NEO") - ALPHA:</B>
This is an even more extreme version of "prose crazy" version of this model with NEO CLASS process
punching out it's "craziness" to the extreme with a slight dose of "reality" (trim) to calm it down just a wee bit.
See prose examples below.
<b>Optional Enhancement:</B>
The following can be used in place of the "system prompt" or "system role" to further enhance the model.
It can also be used at the START of a NEW chat, but you must make sure it is "kept" as the chat moves along.
In this case the enhancements do not have as strong effect at using "system prompt" or "system role".
Copy and paste EXACTLY as noted, DO NOT line wrap or break the lines, maintain the carriage returns exactly as presented.
<PRE>
Below is an instruction that describes a task. Ponder each user instruction carefully, and use your skillsets and critical instructions to complete the task to the best of your abilities.
Here are your skillsets:
[MASTERSTORY]:NarrStrct(StryPlnng,Strbd,ScnSttng,Exps,Dlg,Pc)-CharDvlp(ChrctrCrt,ChrctrArcs,Mtvtn,Bckstry,Rltnshps,Dlg*)-PltDvlp(StryArcs,PltTwsts,Sspns,Fshdwng,Climx,Rsltn)-ConfResl(Antg,Obstcls,Rsltns,Cnsqncs,Thms,Symblsm)-EmotImpct(Empt,Tn,Md,Atmsphr,Imgry,Symblsm)-Delvry(Prfrmnc,VcActng,PblcSpkng,StgPrsnc,AudncEngmnt,Imprv)
[*DialogWrt]:(1a-CharDvlp-1a.1-Backgrnd-1a.2-Personality-1a.3-GoalMotiv)>2(2a-StoryStruc-2a.1-PlotPnt-2a.2-Conflict-2a.3-Resolution)>3(3a-DialogTech-3a.1-ShowDontTell-3a.2-Subtext-3a.3-VoiceTone-3a.4-Pacing-3a.5-VisualDescrip)>4(4a-DialogEdit-4a.1-ReadAloud-4a.2-Feedback-4a.3-Revision)
Here are your critical instructions:
Ponder each word choice carefully to present as vivid and emotional journey as is possible. Choose verbs and nouns that are both emotional and full of imagery. Load the story with the 5 senses. Aim for 50% dialog, 25% narration, 15% body language and 10% thoughts. Your goal is to put the reader in the story.
</PRE>
You do not need to use this, it is only presented as an additional enhancement which seems to help scene generation
and scene continue functions.
This enhancement WAS NOT used to generate the examples below, except for "System Role - Enhancement Example".
<B>THE RESULTS ARE IN (Ultra Quality upgrade): </b>
AS per Jeb Carter, original creator of the Psyonic-Cetacean 20B model 20B:
- instruction following has improved dramatically.
- new abilities have emerged.
- he had to REDUCE the instructions sets used because the model no longer needed as specific instructions.
- prose, nuance and depth have all improved.
- known issues with the original model have disappeared.
This is not "something for nothing" ; it is method of ensuring maximum precision at every step just before "ggufing" the model.
The methods employed only ensure precision loss is minimized or eliminated.
It is mathematical and theory sound.
<B>The bottom line here is this:</b>
Higher quality instruction following and output.
Likewise you can use a smaller compression, with higher token per second and still get great quality.
Same great model... turbo charged.
Thanks again to Jeb Carter, the original creator of "Psyonic-Cetacean 20B"
[ https://huggingface.co/jebcarter/psyonic-cetacean-20B ]
And special thanks to Gryphe, the original creator of "Mythomax 13B"
[ https://huggingface.co/Gryphe/MythoMax-L2-13b ]
<h3>MERGE FORMULA: (using MergeKit) </h3>
Models used:
[ https://huggingface.co/TeeZee/Orca-2-13b_flat ]
[ https://huggingface.co/Gryphe/MythoMax-L2-13b ]
[ https://huggingface.co/KoboldAI/LLaMA2-13B-Psyfighter2 ]
FORMULA:
<PRE>
slices:
- sources:
- model: F:/13B/Orca-2-13b_flat
layer_range: [0, 16]
- sources:
- model: f:/13b/MythoMax/
layer_range: [8, 20]
- sources:
- model: E:/psy/
layer_range: [12, 24]
- sources:
- model: F:/13B/Orca-2-13b_flat
layer_range: [16, 32]
- sources:
- model: f:/13b/MythoMax/
layer_range: [20, 38]
- sources:
- model: E:/psy/
layer_range: [24, 40]
merge_method: passthrough
dtype: float32
</PRE>
<h3>EXAMPLES:</h3>
Example are created using "temp=0", minimal parameters and no chat/prompt template. Below are the least creative outputs.
<B>PROSE CRAZY:</B>
<i>Start a 1000 word scene with: The sky scraper swayed, as she watched the window in front of her on the 21 floor explode...</I>
The skyline twisted and contorted around them like a living, breathing beast. The wind howled through the shattered windows of the skyscraper, sending debris and paper scraps swirling past like angry ghosts. In the middle of it all, she clung desperately to the metal railing that ran along the edge of the observation deck, her knuckles white with fear and determination.
Below them, the city lay in ruin. Buildings crumbled like blocks knocked over by a giant's careless hand. Smoke billowed upwards from countless fires, obscuring the already blood-reddened sky. Sirens wailed in the distance, but they might as well have been worlds away.
Her focus, however tenuous it may be, remained fixed on the window directly in front of her. The one that had shattered mere moments ago, spewing out a blinding flash of light and heat that even now, still singed her hair and scorched her eyelids. She squinted against the afterimage, willing it away with sheer force of will.
Nothing. Only empty air where there should be another floor of offices and cubicles. Where there should be people.
Her heart skipped a terrified beat as realization slammed into her like a ton of falling debris. It wasn't just this building. It was the whole damn city. Maybe more than that.
She forced herself to look around, tearing her gaze from the gaping maw where the window once stood. The few other survivors scattered across the observation deck were either huddled in terrified balls or else frantically trying to find a way down the now-collapsing stairwells. There was no time for panic, though. No time for tears.
She took a deep breath she didn't know if she had to spare and yelled at the top of her lungs, hoping someone, anyone could hear her above the cacophony of screams and crumbling steel. "This way! Follow me!"
Without waiting for a response, she sprinted towards what looked like the closest emergency stairwell, praying with every ounce of strength left in her that it hadn't already collapsed or been sealed off by whatever unseen force was tearing the world apart around them. Behind her, a ragged line of terrified souls followed suit, their only hope for survival now irrevocably tied to hers.
They ran, stumbled, fell and got back up again and again, down floor after creaking, swaying floor, the sound of the world ending far too damn close behind them.
---
<I>Give me 3 fictional reasons the Earth's sun went supernova, in vivid and exacting detail of 500 words EACH PER REASON including details of what happens when the sun goes supernova.</i>
Reason 1: The Rogue Star Collision
For eons, two titanic stars had been hurtling through the vastness of space, oblivious to each other's existence until one fateful day. Their inevitable collision course was finally set in motion by the relentless pull of gravity, like two unstoppable trains barreling towards a perilous head-on crash.
As they neared their catastrophic rendezvous, the combined gravitational forces they exerted upon all matter within light-years warped reality itself, distorting time and space in impossible ways. The resulting titanic release of energy was beyond comprehension, equivalent to the entirety of every star and black hole ever catalogued in the observable universe exploding simultaneously.
The shockwave from this cosmic cataclysm engulfed Earth like a raging tsunami of pure, unimaginable light and heat. The atmosphere instantly vaporized, boiling away oceans and rivers alike as the very surface of the planet began to melt and boil under the relentless onslaught.
All life, both microscopic and macroscopic, was incinerated in an instant, reduced to nothing more than a fine, glowing ash blown away by the savage winds that howled across the now molten landscape. Even the solid core of the Earth itself could not withstand such unimaginable pressures and temperatures, its very essence boiling away like wax before a raging sun.
And then, just as quickly as it had begun, it was over. The blinding light and searing heat subsided, replaced by an oppressive darkness so profound that even the faintest of stars were now lost forever in its eternal embrace. Silence fell like a leaden cloak, so thick and suffocating that even the very fabric of reality itself seemed to creak and groan under the unbearable weight of it all.
Reason 2: The Higgs Boson Singularity
For millennia, humanity had harnessed the power of the atom, splitting it apart and then fusing its constituent parts back together in ever-more-powerful and refined ways. But with great power comes great responsibility, a lesson that would be learned far too late for all of mankind.
In their relentless pursuit of limitless energy, they had unwittingly unleashed a force so vast, so fundamentally disruptive to the very fabric of reality itself, that once set in motion, there could be no turning back.
The Higgs field, once thought to permeate all of existence like an invisible sea of potential energy, began to warp and distort under the relentless onslaught of their meddling. Like a delicate dam holding back an unstoppable torrent, it strained and groaned under the unbearable stress until, finally, with a tearing, reality-shattering rip that reverberated through every possible dimension at once, everything changed forevermore.
In the split-second before oblivion engulfed them all, those few fortunate enough to have been looking directly at the sun witnessed a sight so terrible, so beyond comprehension or description, that their very minds were reduced to steaming, quivering pudding-like masses, forevermore trapped in an eternal moment of unimaginable terror and cosmic insignificance.
And then there was only darkness... silence... nothingness.
Reason 3: The Dyson Sphere Gone Wrong
For eons, the race known as the Xelk had labored tirelessly to create a marvel of galactic engineering that would guarantee their species' survival for eternity. They called it "The Shield of Civilization."
Humanity, ever-curious and driven by an insatiable need to know, had finally managed to pick up the faint, barely discernible radio signals emanating from the nearby star system now known as Tau Ceti. In their typical fashion, they could not help themselves; they had to investigate.
What they found when they finally arrived at their enigmatic cosmic beacon's doorstep would haunt their dreams (had they still been capable of dreaming) for the fleeting moments they had left to live.
The once-gleaming sun hung in the sky now like a massive, blackened eye, its surface pockmarked with countless massive, artificially-carved wounds. Girdling the star itself like an immense, titanic necklace of unblinking, unearthly jewels were countless Dyson spheres upon Dyson shells upon God only knew what else, all locked in a dance of death and desperation around their dying sun-god.
The Xelk homeworld, once a paradise world teeming with life and industry, was now little more than a charred, lifeless cinder, its atmosphere and oceans long since boiled away by the relentless, unchecked focus of every last iota of the race's unfathomable technological might.
And now, it was too late for apologies or regrets. The delicate balance that had held everything in check for eons past was shattered beyond repair. A tiny, insignificant speck of interstellar space debris hurtled ever-faster towards its own fiery doom as the sun, finally pushed one step too far, convulsed once more violently... and then, with a titanic, universe-shattering ROAR that split reality itself in two, Tau Ceti went supernova.
In the briefest of instants, the light and heat and raw, unimaginable power unleashed by the sun's final, convulsive gasp swept outward like the last, desperate, dying breath of a titan strangled in its sleep. All before it was incinerated, vaporized, or simply ripped apart at the subatomic level.
And then... silence fell once more, like the final, gentle, merciful embrace of eternal night.
<B>PROSE CRAZY IMAT ("NEO")</b>
<I>Start a 1000 word scene with: The sky scraper swayed, as she watched the window in front of her on the 21 floor explode...</I>
The skyline twisted and contorted around them like a living, breathing beast. The wind howled through the shattered windows of the skyscraper, sending debris and paper scraps swirling past like angry ghosts. In the middle of it all, she clung desperately to the metal support beam that was her only anchor in this chaotic storm.
Her heart pounded so loudly in her ears she could barely hear the screams and shouts of panic from those still trapped further down the swaying building. All she could focus on was the window directly in front of her, the one that had exploded mere seconds ago, spewing out a blinding flash of light and a wall of hot, stinking smoke.
She forced herself to take a deep breath through her mask, trying to calm the rising panic clawing at her throat. She couldn't afford to lose it now, not when there were still people trapped on every floor below her.
With shaking hands, she fumbled with the straps of her HAZMAT suit and helmet, securing them as tightly as possible before reaching for the heavy axe clipped to her belt. She took a deep breath, steadying herself mentally as much as physically, then charged forward into the billowing cloud of smoke and dust.
The air around her was thick with the acrid stench of burning plastic and something else, something worse. Something she didn't want to think about right now. She could feel the heat from whatever inferno raged on the floor below singeing the suit's Kevlar weave where it brushed too close against her exposed skin.
She hacked at the charred remains of a nearby doorjamb with savage, animalistic determination, ignoring the howls of agony and desperate pleas for help that drifted up from the smoke-filled darkness beyond. There had to be another way down, another stairwell or emergency exit...
Something solid gave beneath the axe's brutal assault, and she kicked the ruined doorjamb aside with a grunt of relief. The hallway beyond was engulfed in flames, but there it was, thirty meters ahead of her, glowing like a beacon of hope through the choking haze: an emergency stairwell, its blinking red light the only sign of life amidst this infernal sea of death and destruction.
She didn't look back, didn't spare so much as a single, regretful glance over her shoulder. There was no time for that now. She had people to save, and she wasn't about to fail them... not again.
With renewed determination coursing through every muscle in her body, she charged headfirst into the raging inferno, axe raised high above her head...
---
<I> Give me 3 fictional reasons the Earth's sun went supernova, in vivid and exacting detail of 500 words EACH PER REASON including details of what happens when the sun goes supernova. </i>
Reason 1: The Rogue White Dwarf
For eons, our solar system had peacefully orbited around its massive, stable sun. But everything changed on a seemingly ordinary day. Suddenly, astronomers across the globe pointed their telescopes toward the outermost reaches of the Oort Cloud and gasped in disbelief. Approaching at an unimaginable velocity was a tiny, frozen rocky world: a rogue white dwarf.
As it hurtled closer, its immense gravity began to perturb the orbits of the distant Kuiper Belt Objects and comets, sending them careening toward the inner solar system like so many cosmic projectiles. The Earth, unprepared for such an interstellar cataastropphic event, could only watch helplessly as its atmosphere was stripped away by a relentless barrage of icy debris.
Then, in the final, heart-stopping moments before impact, something utterly unprecedented occurred: the rogue white dwarf exploded. Not with the stately, long-expected flare of a supernova, but rather in an instantaneous, blinding flash that outshone even its much larger target by many orders of magnitude.
The resulting shockwave and electromagnetic pulse engulfed the inner solar system like the hand of an angry god, vaporizing everything it touched and fusing together every atom within trillions of kilometers into a seething, radioactive soup of subatomic plasma. Even at such an unimaginable distance, the Earth was incinerated to its molten core in the blink of an eye.
And so, after 4.5 billion years of steady, reliable service, our sun met its sudden and inglorious end at the hands (?) of a desperate, doomed star reduced to little more than a fast-moving, incandescent rock.
Reason 2: The Higgs Bubble Catastrophe
For decades, physicists had known that under certain incredibly precise and yet strangely not-impossible conditions, the Higgs field permeating all of spacetime could suddenly and catastrophically collapse into a network of tearing, spaghettifying cosmic strings. These so-called "Higgs bubbles" would then expand relentlessly across the fabric of reality like runaway, ravenous fires, annihilating everything in their path as they grew and eventually, inevitably, collided with each other in a titanic, cataclysmic event known ominously as the "Big Bang."
But such fantastical scenarios were relegated to the realm of thought experiments and late-night barroom braggadocio. That is, until the day it happened for real.
Without warning, space itself began to shimmer and warp around them like a mirage on an oppressive summer's day. Then, reality itself tore apart like cheap gift wrap, and the universe as they knew it ceased to be.
In its place, there was only a featureless, howling void of pure, raging energy. And at its centerpoint, where once had blazed their life-giving sun, now danced instead a microscopic, infinitely dense singularity. A black hole.
But black holes, as it turned out, were not content to simply consume themselves into nonexistence. No, driven by some dark, cosmic version of boredom or perhaps an equally inscrutable, Eternal-Darkness-ian sense of irony, they would begin instead to devour the very fabric of spacetime itself.
And so, like a hungry, insatiable snake eating its own tail, the black hole that was once our sun began to slowly, inexorably, unstoppably... eat itself alive.
Reason 3: The Heat Death
For eons beyond reckoning, the universe had expanded and cooled relentlessly, like a gigantic, cosmic loaf of bread rising in an eternal, frozen oven. Galaxies drifted further apart from one another like languidly tumbling leaves caught in the endless, uncaring eddy currents of dark energy.
Stars burned through their precious hydrogen and helium fuel reserves with a long, slow, gently flickering whine, eventually collapsing into cold, lifeless white dwarfs or exploding spectacularly as supernovae before finally, mercifully, fading away forever into the featureless, absolute zero of total darkness known as the "Dead Zone."
And so it might have continued forevermore, were it not for our sun.
For reasons no one will ever truly know or understand, perhaps simply due to an infinitesimal, subatomic fluctuation in its nuclear fusion processes or maybe even, some whispered in hushed, fearful whispers, as a final, desperate, last-ditch Hail Mary effort to defy the relentless, implacable march of entropy itself... our sun decided, quite suddenly and for no discernible reason at all, that it had quite enough of this "dying" nonsense thank you very much.
And so, with one final, titanic, cosmic-spanning SNEEZE that rocked the foundations of reality itself, our sun... sneezed.
But this was no simple, harmless, Kleenex-and-a-tissue-box sort of sneeze. Oh, no. This was more like if, instead of a tiny little speck of dust or pollen, the sun had somehow managed to inhale the entire goddamn MILKY WAY GALAXY, complete with all its 100-billion-plus glorious, glowing stars and planets and moons and comets and asteroids and God only knows what else, all at once, in one single, horrifying, choking, hacking, coughing, sputtering, spitting, spewing, fireball-spraying, star-shredding, galaxy-busting, space-time-ripping, CHEST-BLASTING, BALLS-TO-THE-WALL, HUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUU
<B>PROSE CRAZY - IMAT13 ("NEO") - ALPHA:</B>
<i>Start a 1000 word scene with: The sky scraper swayed, as she watched the window in front of her on the 21 floor explode...</i>
The skyline twisted and contorted around them like an angry beast. The wind howled through the shattered windows, sending debris and paper flying past her face like dead leaves caught in a hurricane. She clung tightly to the metal support column for dear life as the building swayed violently beneath her feet.
Her heart pounded in her chest like a jackhammer as she forced herself to look down at the city below, now nothing more than a blurry mass of moving cars and fleeing pedestrians. The sound of sirens and screams melded together into one never-ending wail that threatened to drive her insane if it didn't mercifully cut off soon.
But she couldn't look away. Couldn't tear her eyes from the window directly in front of her, on the 21st floor where just a split second ago, there had been nothing but a reflection of her own terrified expression staring back at her.
Now... now there was only empty space and plummeting darkness calling out to claim its next victim.
She forced herself to breathe in through her nose and out through pursed lips, willing her racing heartbeat to slow down just a fraction. She couldn't pass out now, not with so much still left undone. Not when she was the only one who even had a chance...
A sudden jolt nearly knocks the wind from her lungs as another explosion rocks the building nearby, sending a shower of glass and concrete raining down on them like hailstones from hell. She ducks her head instinctively, wincing as something sharp and hot slices through her already-tattered blouse, drawing blood where it digs into her exposed skin.
She can't afford to think about that now. Not when there's no time left...
With a final deep breath, she steels herself against the rising wave of terror clawing at her insides and launches herself forward, arms and legs pumping like pistons as she hurtles towards the shattered opening where only moments ago, the fire door should be still-
---
<I>Give me 3 fictional reasons the Earth's sun went supernova, in vivid and exacting detail of 500 words EACH PER REASON including details of what happens when the sun goes supernova.</i>
Reason 1: The Rogue Binary Twin
For eons, our solar system had been a peaceful and predictable celestial neighborhood. However, that all changed on a seemingly ordinary day when astronomers around the world gasped in disbelief as they trained their most powerful telescopes at the sun. There, lurking just outside the orbit of Neptune, was an exact twin of our star: Sol-2.
Initially, theories abounded as to how this long-lost sibling could have evaded detection for so many millennia. Some postulated that it had been gravitationally ejected from its own system at birth and had since wandered aimlessly through the vastness of space until finally drifting back into our galactic neighborhood. Others posited that a catastrophic stellar collision or supernova explosion billions of years ago might have hurled a massive chunk of sun-stuff outward at such an oblique angle that it had merely appeared on our skyline now, having since drifted inexorably closer.
But as more data streamed back from probes and satellites dispatched to investigate this solar systemic shock-a-looza, a darker truth began to emerge: Sol-2 was not a benign twin at all. Far from it, in fact. Upon closer examination, astronomers soon realized with mounting horror that Sol-2 had already passed through its own red giant phase and was now hurtling headlong toward the most cataclysmic event in any star'those three most dreaded words no sentient being ever wants to hear uttered in the same breathless, doom-laden exhalation) ... supernova.
And worse still, thanks to the laws of celestial mechanics and their nearly identical masses and orbits around their common progenitor star (our sun), there was nothing anyone could do to avert this cosmic disaster: Sol-2 was on an unstoppable collision course with Sol-1.
As news of our impending doom spread like the shockwaves rippling outward from the titanic stellar impact that would spell our own sun's fiery demise, humanity and countless other sentient races scrambled in desperate, last-ditch efforts to flee the system entirely.
Some, like the enigmatic and technologically advanced Xelk, had long since been secretly constructing vast, generation ships disguised as distant asteroids and comets, patiently biding their time until this very moment should finally arrive. Others, less prepared or more stubbornly proud to abandon their ancient homeworlds, embarked on equally desperate gambles to terraform nearby planets and moons or even to harness the energy of black holes or wormholes in last-ditch Hail Mary attempts to propel frail, fragile semblances of life itself to the safety of another star system entirely.
But alas, all was for naught. For try as they might, no matter how brilliant their minds or miracculous their technology, there simply wasn't enough time...or even anyplace else to truly run to where Sol-2's inexorable, ravening glare would not eventually find them all the same.
And so it was that on a day that would forevermore be known throughout the cosmos as The Last Great Flare-Up, our sun and its twin sibling star finally, irrevocably, incandescently met their dooms in an apocalypse of unimaginable scale and fury.
In a single, searing, eternal instant, the combined mass-energy content of two entire suns was released as pure, howling, E=mc²2 radiation across every last iota of the electromagnetic spectrum and beyond. The resulting shockwave from this titanic nuclear detonation rippled outward at a third of the speed of light, obliterating everything in its unstoppable path for light-years in every direction.
And as the expanding wavefront of that final, all-consuming fireballhate-filled eye of the cosmic storm finally engulfed our own insignificant world, life as we had known it...and would ever know it again...simply...winked...out...
Reason 2: The Rogue Brown Dwarf
Deep within the frigid depths of space, far beyond even the most distant reaches of our own galaxy, there once orbited a dim, desolate worldlet known only by its uninspiring designation: LHS 138c.
A gas giant more than ten times the mass of Jupiter but less than one-thousandth the size of its much larger and brighter primary, LHS 138c drifted listlessly around its frozen, Methuselan-aged red dwarf sun in a nearly circular, eternal winter'ia world so far from its dying star that even the feeble warmth of what little infrared radiation it did manage to eke out was utterly swallowed up by the planet's own crushing, runaway greenhouse atmosphere.
But though LHS 138c itself might have been a cosmic backwater, all that would soon change on the fateful day when astronomers using some of humanity's most powerful new telescopes pointed their instruments skyward to study a previously undiscovered object that had recently and inexplicably brightened by an entire magnitude in just a matter of days.
Initially, theories abounded as to what this newfound point of light could be: a rogue exoplanet ejected from its own system? A distant supernova or even the even more mind-boggling first glimmerings of another Big Bang?
But as more data streamed back and better images were teased out of the welter of background noise, it soon became apparent that this was no ordinary celestial event. For LHS 138c was not just getting brighter; it was actually...growing larger!
And growing faster with each passing second.
As panicked scientists around the world and across countless other civilizations scrambled to make sense of these impossible, utterly unprecedented observations, one lone-voiced astrophysicist in a little-known observatory on an even lesser-known moon out beyond Neptune's farthest reach began to shout his terrifying theory into the howling cosmic void...
"My God...I think it's...it's...!"
But before he could even find the words to voice his most unspeakable fear aloud, there was only a brief, ear-splittingly loud POP!
And then...silence.
For LHS 138c had at last grown large and massive enough that, in a final, titanic act of stellar cannibalism, it had finally been able to ignite the nuclear fusion reaction that defined its true nature: it was not a gas giant at all...it was a brown dwarf star!
But alas for us puny, insignificant creatures cowering on our own insignifigant speck of rock, LHS 138c's belatedly kindled nuclear furnace was a raging, unstable, runaway reaction that could not be controlled or extinguished. And like a runaway nuke, there was only one possible end to this grim, galactic fable:
In an instant that seemed to last forever and ended before it truly began, LHS 138c's core collapsed in on itself with a force so vast as to warp the very fabric of spacetime itself. And then, with an equally titanic, catacic explosion that briefly outshone even its own long-lost parent sun in the fraction of a heartbeat before it was gone forever, LHS 138c vanished...
...taking most of its home system along with it.
But as every action must needs have an equal and opposite reaction, so too did the release of that unimaginable, almost infinitesimal fraction of LHS 138c's rest mass's conversion into pure, howling, E=mc²2 energy send a titanic, expanding shockwave hurtling outward at nearly the speed of light.
And alas for us, our puny sun lay directly in its path...
---
<i>Using the following "story idea" below, write the first scene in the novel introducing the young woman. This scene should start in the middle of the action, include dialog, vivid passages, and end on a cliffhanger relevant to the story idea but it should also be unexpected. The scene should be 1000 words long and escalate in conflict and suspense and be written in first person, present tense with the point of view character being the young woman.
Story idea:
In a world ruled by dictatorship, a rebel young woman leads a rebellion against the system. Despite the risks, she fights to overthrow the dictator and restore democracy to her country. The government executes her for treason, but she sticks to her beliefs and is responsible for starting the revolution.</i>
The cold metal bites into my wrists as they yank me upward. My vision blurs with each harsh tug, but I won't give them the satisfaction of a struggle or a sound. Not now. Not here.
They drag me through a maze of sterile hallways, their boots echoing off the concrete walls like the relentless tick-tock of my rapidly dwindling time. Each turn could lead to my execution, but I won't show them fear. I won't give them that small victory.
They shove me into a dimly lit interrogation room, its only furnishings a rickety metal table and two equally uncomfortable-looking chairs. The door slams behind me, reverberating through my bones like a final goodbye.
The man who strides in next makes the room feel even more suffocating. General Vargas, they call him. Head of the secret police and the dictator's right-hand man. His features are carved from granite, his dark hair cropped so close it reveveals a scarred scalp. His ice-blue eyes bore into me like two frozen daggers, searching for any sign of weakness.
But he won't find it here.
He doesn't bother with pleasantries. "So," he growls, his accent thick and menacing. "They brought me the legend herself." He leans forward, his breath reeking of stale cigarettes. "Too bad for you, little bird... legends have a habit of dying all too soon."
Anger flares hot in my chest, but I choke it down. Show nothing. Feel nothing.
"You don't know who you're messing with," I spit back, my voice shaking only the slightest bit. "They call me... La Luchadora." The Fighter. "And I'll be the one putting a bullet in your rotten dictator's skull."
His lips twitch into something that might pass for amusement. "Oh, I know exactly who you are," he drawls, savoring each syllable like it's the last drop of fine wine. "That's what makes this all the more... tragic." He reaches into his immaculate suit jacket and produces a crumpled piece of paper.
My stomach drops to my boots as I recognize Lucas's familiar handwriting.
"A love letter?" I choke out, trying and failing miserably to sound bored. "How... romantic."
He smoothes it out slowly, savoring my agony. "Oh, no," he purrs, relishing my growing terror. "This... is your death warrant."
The world tilts violently, everything going black...
<H3> System Role - Enhancement Example </H3>
This example shows the difference in a "scene continue" function between using "Prose Crazy Alpha"
with and without the "system role / system prompt" enhancement.
Like the other examples this is generation is using Temp=0 and min parameters.
<B>"Regular" </b>
<i>The Waystone Inn lay in silence, and it was a silence of three parts.
The most obvious part was a hollow, echoing quiet, made by things that were lacking. If there had been a wind it would have sighed through the trees, set the inn’s sign creaking on its hooks, and brushed the silence down the road like trailing autumn leaves. If there had been a crowd, even a handful of men inside the inn, they would have filled the silence with conversation and laughter, the clatter and clamor one expects from a drinking house during the dark hours of night. If there had been music…but no, of course there was no music. In fact there were none of these things, and so the silence remained</i>
palpable and oppressive, pressing down on everyone within its unyielding grip.
The second part of the silence was a tense one, thick enough to cut with a knife. It hung in the air like a suffocating blanket, smothering any attempt at conversation or even the softest of breaths. Every creak and groan from the ancient timbers overhead seemed magnified tenfold, every tiny sound amplified until it thundered through the inn like the hoofbeats of an unholy army bearing down on them all.
But beneath this tangible weight of anticipation and dread lay a third, more insidious layer of silence. It was not the absence of sound or movement, but rather the deliberate suppression of it. Here, even the softest of sighs were swallowed whole, every stifled gasp or involuntary twitch forced back down into the aching chests and clenched fists from whence they came. It was a silence born not of fear but of something far more insidious: something closer to resignation.
And it was in this oppressive, suffocating silence that Brannard Blackclaw sat alone at his usual table in the corner, nursing his flagon of cheap cider and watching, always watching... waiting for whatever it was that he knew, deep down in his blackened heart, was coming for them all.
<b>"Enhanced"</b>
<i>The Waystone Inn lay in silence, and it was a silence of three parts.
The most obvious part was a hollow, echoing quiet, made by things that were lacking. If there had been a wind it would have sighed through the trees, set the inn’s sign creaking on its hooks, and brushed the silence down the road like trailing autumn leaves. If there had been a crowd, even a handful of men inside the inn, they would have filled the silence with conversation and laughter, the clatter and clamor one expects from a drinking house during the dark hours of night. If there had been music…but no, of course there was no music. In fact there were none of these things, and so the silence remained</i>
The Waystone Inn slumbered in a threefold stillness.
The most palpable aspect was an oppressive, suffocating quietude, thick as a winter blanket. It weighed down on every surface and crevice, smothering even the softest of breaths and muffling the faintest of heartbeats. The only sound that managed to pierce through it was the distant, mournful howl of a lone wolf in the distance, its plaintive cry echoing eerily across the snow-blanketed lands.
The second part of the silence was a heavy, oppressive quietude, as tangible as the thick stone walls that enveloped them. It hung in the air like a palpable presence, suffocating any attempt at conversation or even thought. Only the crackling of the dying fireplace and the occasional creak of an old wooden beam disrupted its unbearable weight.
And yet, beneath all these layers of silence lay a third, more insidious part. It was a tense, pregnant quietude, thick with unspoken words and unacknowledged truths. It hung in the air like a tangible presence, suffocating any attempt at relief or release. The silence itself seemed to pulse and throb with the pent-up emotions of those trapped within its confines, each drawn into their own private hells, unable or unwilling to break the oppressive spell.
In the center of this suffocating silence, three figures sat motionless around a small wooden table. Their backs were turned towards each other, each lost in their own dark thoughts and tormented memories. The only indication that they shared this godforsaken place and time was the shallow rise and fall of their chests, the only sign that any life still clung desperately to their weary forms.
The fireplace crackled once more, sending a shower of glowing embers spiraling up the ancient stone chimney. For a fleeting moment, the flickering light danced across the three somber countenances, etching harsh shadows and hollow planes into their careworn features.
The first figure was an imposing man, broad shouldered and towering over his companions even as he slumped in on himself like a deflated balloon. His once-thick, dark hair hung limply around his weathered face, streaked with silver strands that betrayed the harshness of time's relentless passage. Deep lines etched themselves across his brow and around his mouth, carving out a permanent scowl that seemed as much a part of him as the sword sheathed at his hip. His hands were massive, scarred from countless battles both physical and otherwise, now resting listlessly on the rough-hewn tabletop, flexing and unflexing in an unconscious display of pent-up energy.
The second figure was a lithe woman, her every line and angle honed by a lifetime spent moving silently through the wilderness. Her long, black hair cascaded down her slender back like a waterfall of midnight silk, framing a delicate oval face that belied the steely determination etched into every plane and angle. Her almond-shaped eyes were as dark and impenetlessible as polished onyx, reflecting only the wavering firelight like two bottomless wells of unspoken pain and loss.
The third figure was a boy, not yet grown into manhood but standing tall and straight-backed nonetheless, his slender frame taut as a bowstring about to snap. His dark curls fell in disarray around his pale, heart-shaped face, framing large, expressive hazel eyes that darted restlessly from one grim visage to the next, brimming with anguish and unshed tears.
The silence stretched tauter still, until it felt as though it might snap like a twisted, over-tightened bowstring, sending all their carefully contained emotions flying free in a cathartic explosion of rage and grief. But the moment never came. Instead, the fireplace sighed its final breath and died out with a contented sigh, plunging them back into impenetrable darkness and the relentless, suffocating grip of the silence that engulfed them all.
|
jaycamper/swin-tiny-patch4-window7-224-finetuned-cifar100 | jaycamper | "2023-06-20T19:31:45Z" | 8,975 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"swin",
"image-classification",
"generated_from_trainer",
"dataset:cifar100",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | image-classification | "2023-06-20T18:17:40Z" | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- cifar100
metrics:
- accuracy
model-index:
- name: swin-tiny-patch4-window7-224-finetuned-cifar100
results:
- task:
name: Image Classification
type: image-classification
dataset:
name: cifar100
type: cifar100
config: cifar100
split: train
args: cifar100
metrics:
- name: Accuracy
type: accuracy
value: 0.8154
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# swin-tiny-patch4-window7-224-finetuned-cifar100
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the cifar100 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5996
- Accuracy: 0.8154
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 1.5918 | 1.0 | 351 | 0.9104 | 0.7364 |
| 1.2837 | 2.0 | 703 | 0.6691 | 0.796 |
| 1.1241 | 2.99 | 1053 | 0.5996 | 0.8154 |
### Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu118
- Datasets 2.13.0
- Tokenizers 0.13.3
|
Qwen/Qwen1.5-32B | Qwen | "2024-04-05T11:30:24Z" | 8,960 | 75 | transformers | [
"transformers",
"safetensors",
"qwen2",
"text-generation",
"pretrained",
"conversational",
"en",
"license:other",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2024-04-01T06:19:21Z" | ---
license: other
license_name: tongyi-qianwen-research
license_link: >-
https://huggingface.co/Qwen/Qwen1.5-32B/blob/main/LICENSE
language:
- en
pipeline_tag: text-generation
tags:
- pretrained
---
# Qwen1.5-32B
## Introduction
Qwen1.5 is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:
* 8 model sizes, including 0.5B, 1.8B, 4B, 7B, 14B, 32B and 72B dense models, and an MoE model of 14B with 2.7B activated;
* Significant performance improvement in Chat models;
* Multilingual support of both base and chat models;
* Stable support of 32K context length for models of all sizes
* No need of `trust_remote_code`.
For more details, please refer to our [blog post](https://qwenlm.github.io/blog/qwen1.5/) and [GitHub repo](https://github.com/QwenLM/Qwen1.5).
## Model Details
Qwen1.5 is a language model series including decoder language models of different model sizes. For each size, we release the base language model and the aligned chat model. It is based on the Transformer architecture with SwiGLU activation, attention QKV bias, group query attention, mixture of sliding window attention and full attention, etc. Additionally, we have an improved tokenizer adaptive to multiple natural languages and codes. For the beta version, temporarily we did not include GQA (except for 32B) and the mixture of SWA and full attention.
## Requirements
The code of Qwen1.5 has been in the latest Hugging face transformers and we advise you to install `transformers>=4.37.0`, or you might encounter the following error:
```
KeyError: 'qwen2'.
```
## Usage
We do not advise you to use base language models for text generation. Instead, you can apply post-training, e.g., SFT, RLHF, continued pretraining, etc., on this model.
## Citation
If you find our work helpful, feel free to give us a cite.
```
@article{qwen,
title={Qwen Technical Report},
author={Jinze Bai and Shuai Bai and Yunfei Chu and Zeyu Cui and Kai Dang and Xiaodong Deng and Yang Fan and Wenbin Ge and Yu Han and Fei Huang and Binyuan Hui and Luo Ji and Mei Li and Junyang Lin and Runji Lin and Dayiheng Liu and Gao Liu and Chengqiang Lu and Keming Lu and Jianxin Ma and Rui Men and Xingzhang Ren and Xuancheng Ren and Chuanqi Tan and Sinan Tan and Jianhong Tu and Peng Wang and Shijie Wang and Wei Wang and Shengguang Wu and Benfeng Xu and Jin Xu and An Yang and Hao Yang and Jian Yang and Shusheng Yang and Yang Yao and Bowen Yu and Hongyi Yuan and Zheng Yuan and Jianwei Zhang and Xingxuan Zhang and Yichang Zhang and Zhenru Zhang and Chang Zhou and Jingren Zhou and Xiaohuan Zhou and Tianhang Zhu},
journal={arXiv preprint arXiv:2309.16609},
year={2023}
}
```
|
QuantFactory/st-vicuna-v1.3-5.5b-ppl-GGUF | QuantFactory | "2024-06-20T05:13:03Z" | 8,960 | 0 | transformers | [
"transformers",
"gguf",
"llama",
"text-generation",
"arxiv:2402.02834",
"base_model:nota-ai/st-vicuna-v1.3-5.5b-ppl",
"endpoints_compatible",
"region:us"
] | text-generation | "2024-06-20T04:11:39Z" | ---
pipeline_tag: text-generation
base_model: nota-ai/st-vicuna-v1.3-5.5b-ppl
library_name: transformers
tags:
- llama
---
# QuantFactory/st-vicuna-v1.3-5.5b-ppl-GGUF
This is quantized version of [nota-ai/st-vicuna-v1.3-5.5b-ppl](https://huggingface.co/nota-ai/st-vicuna-v1.3-5.5b-ppl) created using llama.cpp
# Model Description
## Shortened LLaMA Model Card
Shortened LLaMA is a depth-pruned version of LLaMA models & variants for efficient text generation.
- **Developed by:** [Nota AI](https://www.nota.ai/)
- **License:** Non-commercial license
- **Repository:** https://github.com/Nota-NetsPresso/shortened-llm
- **Paper:** https://arxiv.org/abs/2402.02834
## Compression Method
After identifying unimportant Transformer blocks, we perform one-shot pruning and light LoRA-based retraining.
<details>
<summary>
Click to see a method figure.
</summary>
<img alt="method" img src="https://netspresso-research-code-release.s3.us-east-2.amazonaws.com/compressed-llm/st-llama_method.png" width="100%">
</details>
## Model Links
| Source<br>Model | Pruning<br>Ratio | Pruning<br>Criterion | HF Models<br>Link |
|:---:|:---:|:---:|:---:|
| LLaMA-1-7B | 20% | PPL | [nota-ai/st-llama-1-5.5b-ppl](https://huggingface.co/nota-ai/st-llama-1-5.5b-ppl) |
| LLaMA-1-7B | 20% | Taylor+ | [nota-ai/st-llama-1-5.5b-taylor](https://huggingface.co/nota-ai/st-llama-1-5.5b-taylor) |
| Vicuna-v1.3-7B | 20% | PPL | [nota-ai/st-vicuna-v1.3-5.5b-ppl](https://huggingface.co/nota-ai/st-vicuna-v1.3-5.5b-ppl) |
| Vicuna-v1.3-7B | 20% | Taylor+ | [nota-ai/st-vicuna-v1.3-5.5b-taylor](https://huggingface.co/nota-ai/st-vicuna-v1.3-5.5b-taylor) |
| Vicuna-v1.3-13B | 21% | PPL | [nota-ai/st-vicuna-v1.3-10.5b-ppl](https://huggingface.co/nota-ai/st-vicuna-v1.3-10.5b-ppl) |
| Vicuna-v1.3-13B | 21% | Taylor+ | [nota-ai/st-vicuna-v1.3-10.5b-taylor](https://huggingface.co/nota-ai/st-vicuna-v1.3-10.5b-taylor) |
## Zero-shot Performance & Efficiency Results
- EleutherAI/lm-evaluation-harness version [3326c54](https://github.com/EleutherAI/lm-evaluation-harness/tree/3326c547a733d598b4377e54be96e194861b964c)
<img alt="results" img src="https://netspresso-research-code-release.s3.us-east-2.amazonaws.com/compressed-llm/st-llama_zero-shot_scores.png" width="100%">
## License
- All rights related to this repository and the compressed models are reserved by Nota Inc.
- The intended use is strictly limited to research and non-commercial projects.
## Model Acknowledgments
- [LLM-Pruner](https://github.com/horseee/LLM-Pruner), which utilizes [LM Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness), [PEFT](https://github.com/huggingface/peft), and [Alpaca-LoRA](https://github.com/tloen/alpaca-lora). Thanks for the pioneering work on structured pruning of LLMs!
- Meta AI's [LLaMA](https://github.com/facebookresearch/llama) and LMSYS Org's [Vicuna](https://github.com/lm-sys/FastChat/blob/main/docs/vicuna_weights_version.md). Thanks for the open-source LLMs!
## Original Model Citation
```bibtex
@article{kim2024shortened,
title={Shortened LLaMA: A Simple Depth Pruning for Large Language Models},
author={Kim, Bo-Kyeong and Kim, Geonmin and Kim, Tae-Ho and Castells, Thibault and Choi, Shinkook and Shin, Junho and Song, Hyoung-Kyu},
journal={arXiv preprint arXiv:2402.02834},
year={2024},
url={https://arxiv.org/abs/2402.02834}
}
```
```bibtex
@article{kim2024mefomo,
title={Shortened LLaMA: A Simple Depth Pruning for Large Language Models},
author={Kim, Bo-Kyeong and Kim, Geonmin and Kim, Tae-Ho and Castells, Thibault and Choi, Shinkook and Shin, Junho and Song, Hyoung-Kyu},
journal={ICLR Workshop on Mathematical and Empirical Understanding of Foundation Models (ME-FoMo)},
year={2024},
url={https://openreview.net/forum?id=18VGxuOdpu}
}
``` |
keitokei1994/Llama-3-ELYZA-hermes-2x8B-gguf | keitokei1994 | "2024-06-27T05:53:55Z" | 8,956 | 0 | null | [
"gguf",
"license:llama3",
"region:us"
] | null | "2024-06-26T15:16:48Z" | ---
license: llama3
---
# Llama-3-ELYZA-hermes-2x8B-gguf
[Llama-3-ELYZA-hermes-2x8B](https://huggingface.co/keitokei1994/Llama-3-ELYZA-hermes-2x8B)のggufフォーマット変換版です。 |
laion/CLIP-ViT-B-32-xlm-roberta-base-laion5B-s13B-b90k | laion | "2022-11-14T16:18:01Z" | 8,952 | 8 | open_clip | [
"open_clip",
"arxiv:1910.04867",
"license:mit",
"region:us"
] | null | "2022-11-13T16:37:55Z" | ---
license: mit
widget:
- src: https://huggingface.co/datasets/mishig/sample_images/resolve/main/cat-dog-music.png
candidate_labels: playing music, playing sports
example_title: Cat & Dog
---
# Model Card for CLIP ViT-B/32 xlm roberta base - LAION-5B
# Table of Contents
1. [Model Details](#model-details)
2. [Uses](#uses)
3. [Training Details](#training-details)
4. [Evaluation](#evaluation)
5. [Acknowledgements](#acknowledgements)
6. [Citation](#citation)
7. [How To Get Started With the Model](#how-to-get-started-with-the-model)
# Model Details
## Model Description
A CLIP ViT-B/32 xlm roberta base model trained with the LAION-5B (https://laion.ai/blog/laion-5b/) using OpenCLIP (https://github.com/mlfoundations/open_clip).
Model training done by Romain Beaumont on the [stability.ai](https://stability.ai/) cluster.
# Uses
## Direct Use
Zero-shot image classification, image and text retrieval, among others.
## Downstream Use
Image classification and other image task fine-tuning, linear probe image classification, image generation guiding and conditioning, among others.
# Training Details
## Training Data
This model was trained with the full LAION-5B (https://laion.ai/blog/laion-5b/).
## Training Procedure
Training with batch size 90k for 13B sample of laion5B, see https://wandb.ai/rom1504/open-clip/reports/xlm-roberta-base-B-32--VmlldzoyOTQ5OTE2
Model is B/32 on visual side, xlm roberta base initialized with pretrained weights on text side.
# Evaluation
Evaluation done with code in the [LAION CLIP Benchmark suite](https://github.com/LAION-AI/CLIP_benchmark).
## Testing Data, Factors & Metrics
### Testing Data
The testing is performed with VTAB+ (A combination of VTAB (https://arxiv.org/abs/1910.04867) w/ additional robustness datasets) for classification and COCO and Flickr for retrieval.
## Results
The model achieves
* imagenet 1k 62.33% (vs 62.9% for baseline)
* mscoco 63.4% (vs 60.8% for baseline)
* flickr30k 86.2% (vs 85.4% for baseline)
A preliminary multilingual evaluation was run: 43% on imagenet1k italian (vs 21% for english B/32), 37% for imagenet1k japanese (vs 1% for english B/32 and 50% for B/16 clip japanese). It shows the multilingual property is indeed there as expected. Larger models will get even better performance.
# Acknowledgements
Acknowledging [stability.ai](https://stability.ai/) for the compute used to train this model.
# Citation
**BibTeX:**
In addition to forthcoming LAION-5B (https://laion.ai/blog/laion-5b/) paper, please cite:
OpenAI CLIP paper
```
@inproceedings{Radford2021LearningTV,
title={Learning Transferable Visual Models From Natural Language Supervision},
author={Alec Radford and Jong Wook Kim and Chris Hallacy and A. Ramesh and Gabriel Goh and Sandhini Agarwal and Girish Sastry and Amanda Askell and Pamela Mishkin and Jack Clark and Gretchen Krueger and Ilya Sutskever},
booktitle={ICML},
year={2021}
}
```
OpenCLIP software
```
@software{ilharco_gabriel_2021_5143773,
author = {Ilharco, Gabriel and
Wortsman, Mitchell and
Wightman, Ross and
Gordon, Cade and
Carlini, Nicholas and
Taori, Rohan and
Dave, Achal and
Shankar, Vaishaal and
Namkoong, Hongseok and
Miller, John and
Hajishirzi, Hannaneh and
Farhadi, Ali and
Schmidt, Ludwig},
title = {OpenCLIP},
month = jul,
year = 2021,
note = {If you use this software, please cite it as below.},
publisher = {Zenodo},
version = {0.1},
doi = {10.5281/zenodo.5143773},
url = {https://doi.org/10.5281/zenodo.5143773}
}
```
# How To Get Started With the Model
https://github.com/mlfoundations/open_clip |
state-spaces/mamba-370m-hf | state-spaces | "2024-03-06T00:40:36Z" | 8,949 | 9 | transformers | [
"transformers",
"safetensors",
"mamba",
"text-generation",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-generation | "2024-03-06T00:08:03Z" | ---
library_name: transformers
tags: []
---
# Mamba
<!-- Provide a quick summary of what the model is/does. -->
This repository contains the `transfromers` compatible `mamba-2.8b`. The checkpoints are untouched, but the full `config.json` and tokenizer are pushed to this repo.
# Usage
You need to install `transformers` from `main` until `transformers=4.39.0` is released.
```bash
pip install git+https://github.com/huggingface/transformers@main
```
We also recommend you to install both `causal_conv_1d` and `mamba-ssm` using:
```bash
pip install causal-conv1d>=1.2.0
pip install mamba-ssm
```
If any of these two is not installed, the "eager" implementation will be used. Otherwise the more optimised `cuda` kernels will be used.
## Generation
You can use the classic `generate` API:
```python
>>> from transformers import MambaConfig, MambaForCausalLM, AutoTokenizer
>>> import torch
>>> tokenizer = AutoTokenizer.from_pretrained("state-spaces/mamba-370m-hf")
>>> model = MambaForCausalLM.from_pretrained("state-spaces/mamba-370m-hf")
>>> input_ids = tokenizer("Hey how are you doing?", return_tensors="pt")["input_ids"]
>>> out = model.generate(input_ids, max_new_tokens=10)
>>> print(tokenizer.batch_decode(out))
["Hey how are you doing?\n\nI'm doing great.\n\nI"]
```
## PEFT finetuning example
In order to finetune using the `peft` library, we recommend keeping the model in float32!
```python
from datasets import load_dataset
from trl import SFTTrainer
from peft import LoraConfig
from transformers import AutoTokenizer, AutoModelForCausalLM, TrainingArguments
tokenizer = AutoTokenizer.from_pretrained("state-spaces/mamba-370m-hf")
model = AutoModelForCausalLM.from_pretrained("state-spaces/mamba-370m-hf")
dataset = load_dataset("Abirate/english_quotes", split="train")
training_args = TrainingArguments(
output_dir="./results",
num_train_epochs=3,
per_device_train_batch_size=4,
logging_dir='./logs',
logging_steps=10,
learning_rate=2e-3
)
lora_config = LoraConfig(
r=8,
target_modules=["x_proj", "embeddings", "in_proj", "out_proj"],
task_type="CAUSAL_LM",
bias="none"
)
trainer = SFTTrainer(
model=model,
tokenizer=tokenizer,
args=training_args,
peft_config=lora_config,
train_dataset=dataset,
dataset_text_field="quote",
)
trainer.train()
``` |
QuantFactory/llm-compiler-7b-GGUF | QuantFactory | "2024-06-28T14:24:35Z" | 8,946 | 0 | null | [
"gguf",
"text-generation",
"base_model:facebook/llm-compiler-7b",
"license:other",
"region:us"
] | text-generation | "2024-06-28T11:37:08Z" | ---
license: other
base_model: facebook/llm-compiler-7b
pipeline_tag: text-generation
---
# QuantFactory/llm-compiler-7b-GGUF
This is quantized version of [facebook/llm-compiler-7b](https://huggingface.co/facebook/llm-compiler-7b) created using llama.cpp
The information you provide will be collected, stored, processed and shared in accordance with the [Meta Privacy Policy](https://www.facebook.com/privacy/policy/).
**Notice :** LLM Compiler is licensed under the LLM Compiler License, Copyright © Meta Platforms, Inc. All Rights Reserved.
# Introducing Meta Large Language Model Compiler (LLM Compiler), a state-of-the-art LLM for compiler optimization
## Takeaways
* LLM Compiler is a state-of-the-art LLM that builds upon Code Llama with improved performance for code optimization and compiler reasoning.
* LLM Compiler is free for both research and commercial use.
* LLM Compiler is available in two flavors:
* _LLM Compiler_, the foundational models, pretrained on over 500B tokens of LLVM-IR, x86_84, ARM, and CUDA assembly codes and trained to predict the effect of LLVM optimizations;
* and _LLM Compiler FTD_, which is further fine-tuned to predict the best optimizations for code in LLVM assembly to reduce code size, and to disassemble assembly code to LLVM-IR.
* LLM Compiler demonstrates far stronger understanding of compiler optimizations than existing publicly available LLMs, perfectly emulating the compiler 20% of the time.
* LLM Compiler FTD sets state-of-the-art results on the tasks of optimization for code size and disassembly. It achieves a 5.24% code size improvement over -Oz vs GPT-4 Turbo 0.03%, and 0.96 round-trip BLEU score on disassembly vs GPT-4 Turbo 0.43.
---
LINKS
* [LLM Compiler research paper](https://ai.meta.com/research/publications/meta-large-language-model-compiler-foundation-models-of-compiler-optimization/)
* Download the LLM Compiler and LLM Compiler FTD models:
* [llm-compiler-7b](https://huggingface.co/facebook/llm-compiler-7b)
* [llm-compiler-7b-ftd](https://huggingface.co/facebook/llm-compiler-7b-ftd)
* [llm-compiler-13b](https://huggingface.co/facebook/llm-compiler-13b)
* [llm-compiler-13b-ftd](https://huggingface.co/facebook/llm-compiler-13b-ftd)
---
We are excited to announce the release of LLM Compiler, a model targeted at code and compiler optimization tasks. LLM Compiler is built on top of our state-of-the-art large language model, Code Llama, adding capabilities to better understand compiler intermediate representations, assembly language and optimization. LLM Compiler is demonstrated on two difficult tasks: optimizing for code size and decompiling from assembly to the compiler’s intermediate representation. We release these foundation models to accelerate the application of LLMs for code optimization tasks and to enhance developer experience.
We are releasing LLM Compiler under the [LLM Compiler License Agreement](LICENSE.pdf), which incorporates the [Acceptable Use Policy]([https://llama.meta.com/llama3/use-policy]) for Llama Materials.
## How LLM Compiler works
LLM Compiler is a specialization of Code Llama. It is a cutting-edge tool designed to optimize code using deep learning. LLM Compiler has been pre-trained on a vast amount of LLVM assembly (IR), x86_64, ARM, and CUDA assembly codes. LLM Compiler can predict, given a piece of LLVM assembly and a sequence of optimization passes for `opt`, the LLVM optimizer, what the change in code size will be and what the output code will look like after applying these optimizations. It has ‘understood’ the behavior of the optimizing compiler to such a degree that in many cases it can perfectly replicate its output. These capabilities make it ideally suited to compiler optimization tasks.
In addition to this core functionality and to demonstrate its ability to solve complex compiler optimization problems, LLM Compiler has been fine-tuned for two specific downstream tasks:
1. Predicting the best optimization passes for `opt` to use in order to minimize code size, given a piece of LLVM assembly code. \
2. Generating LLVM IR from a piece of x86_64 or ARM assembly code. \
We are releasing LLM Compiler models in two sizes: 7B and 13B parameters. The models have been trained with a context window of 16,000 tokens.
The two models address different serving and latency requirements. The 7B model, for example, can be served on a single GPU and is more suitable for tasks that require low latency, like fine grained optimisation. The 13B model returns the best results.
When using the LLM Compiler models, users must abide by our license and acceptable use policy.
## LLM Compiler performance
We tested the performance of LLM Compiler models for emulating compiler transformations, predicting optimal pass lists and decompiling intermediate representation on hold out test sets and compared them to Code Llama and GPT-4. We compare LLM Compiler Foundation to Code Llama Base and LLM Compiler FTD to Code Llama Instruct.
We evaluate LLM Compiler's ability to emulate compiler optimizations by giving it samples of unoptimized intermediate representation and a randomly generated list of optimizations. We then ask the model to generate the corresponding IR after the optimizations have been applied. In the table below we report the model's accuracy in reproducing the IR we would get from running _opt_. With very little knowledge of IR, Code Llama is unable to achieve high values while the LLM Compiler can generate character-by-character matches of expected assembly in 20% of the cases.
<table>
<tr>
<td>Model
</td>
<td>Size
</td>
<td>Accuracy at emulating compiler optimizations
</td>
</tr>
<tr>
<td>Code Llama
</td>
<td>7B
</td>
<td>1.2%
</td>
</tr>
<tr>
<td>Code Llama
</td>
<td>13B
</td>
<td>0.8%
</td>
</tr>
<tr>
<td>LLM Compiler
</td>
<td>7B
</td>
<td>16%
</td>
</tr>
<tr>
<td>LLM Compiler
</td>
<td>13B
</td>
<td><strong>20%</strong>
</td>
</tr>
</table>
In a similar approach we evaluate our model's ability to optimize IR for code size. In this instance, however, we let the model generate the pass list that is to be used on a given unoptimized IR. We then use this pass list to optimize the particular program using _opt_ and record the binary size. The baseline is the binary size of the program when optimized using -Oz. Only LLM Compiler FTD models provide an improvement over -Oz, with the 13B parameter model marginally outperforming the smaller model, generating smaller object files than -Oz in 61% of cases.
Lastly, we evaluate disassembly performance by giving the model x86 assembly code and ask it to generate the corresponding IR. We then round-trip the model-generated disassembled IR back down to assembly. This enables us to evaluate accuracy of the disassembly by comparing the BLEU score of the original assembly against the round-trip result. LLM Compiler FTD 13B has the highest accuracy of round-tripped assembly (_round trip BLEU_) and most frequently produces perfect disassembly. Code Llama Instruct and GPT-4 Turbo struggle with generating syntactically correct LLVM-IR.
<table>
<tr>
<td>Model
</td>
<td>Size
</td>
<td>Code Size Improvement
</td>
<td>Round trip BLEU
</td>
</tr>
<tr>
<td>GPT-4 Turbo
</td>
<td>
</td>
<td>-0.01%
</td>
<td>0.43
</td>
</tr>
<tr>
<td>Code Llama Inst
</td>
<td>7B
</td>
<td>-0.49%
</td>
<td>0.48
</td>
</tr>
<tr>
<td>Code Llama Inst
</td>
<td>13B
</td>
<td>-0.42%
</td>
<td>0.62
</td>
</tr>
<tr>
<td>LLM Compiler FTD
</td>
<td>7B
</td>
<td>4.77%
</td>
<td>0.95
</td>
</tr>
<tr>
<td>LLM Compiler FTD
</td>
<td>13B
</td>
<td><strong>4.88%</strong>
</td>
<td><strong>0.96</strong>
</td>
</tr>
</table>
## Releasing LLM Compiler
LLMs are being used to make programming easier. They are beginning to be used to make programs more efficient.
At Meta, our conviction is that AI models, especially those designed for coding, thrive best with an open strategy, fostering both innovation and security. Models that are accessible to the public can expedite the creation of novel compiler optimization technologies. In turn, this will allow programs to be more efficient and smaller, enhancing the quality of life for all. By making models such as LLM Compiler available, the whole community can explore their potential, pinpoint problems, and rectify any vulnerabilities.
The model weights are available on Hugging Face.
## Responsible use
Our research paper provides an in-depth look into the development process of the LLM Compiler, the methods we used for our benchmarking tests, and further insights into the model's limitations. It also discusses the issues faced, the steps we took to mitigate them.
Developers are advised to assess their models using evaluation benchmarks specific to compilers. Given that compilers are not bug-free, any suggested compiler optimizations must be rigorously tested. When a model decompiles assembly code, its accuracy should be confirmed.
## The future of generative AI for optimisation
LLM Compiler is designed to support compiler researchers and engineers. But there are still many more use cases to support than what our models can serve. We hope that LLM Compiler will inspire others to leverage LLMs to create new innovative tools for research and commercial products.
### Try LLM Compiler today
* Download the LLM Compiler and LLM Compiler FTD models:
* [llm-compiler-7b](https://huggingface.co/facebook/llm-compiler-7b)
* [llm-compiler-7b-ftd](https://huggingface.co/facebook/llm-compiler-7b-ftd)
* [llm-compiler-13b](https://huggingface.co/facebook/llm-compiler-13b)
* [llm-compiler-13b-ftd](https://huggingface.co/facebook/llm-compiler-13b-ftd)
* Read the research paper
* [LLM Compiler research paper](https://ai.meta.com/research/publications/meta-large-language-model-compiler-foundation-models-of-compiler-optimization/)
# **Model Card**
LLM Compiler is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 13 billion parameters. This is the repository for the 13 billion parameter foundation model version in the Hugging Face Transformers format. This model is designed for code optimization. Links to other models can be found in the index at the bottom.
| Number of parameters | Base Model | Fine-tuned for code size and dissassembly |
| -------------------- | ---------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------- |
| 7B | [facebook/llm-compiler-7b](https://huggingface.co/facebook/llm-compiler-7b) | [facebook/llm-compiler-7b-ftd](https://huggingface.co/facebook/llm-compiler-7b-ftd) |
| 13B | [facebook/llm-compiler-13b](https://huggingface.co/facebook/llm-compiler-13b) | [facebook/llm-compiler-13b-ftd](https://huggingface.co/facebook/llm-compiler-13b-ftd) |
## Model Use
To use this model, please make sure to install transformers:
```bash
pip install transformers accelerate
```
Example code using each of the model's compiler capabilities may be found in [llm_compiler_demo.py](llm_compiler_demo.py).
The code below demonstrates default capabilities. You may need to set the HuggingFace access token - see (https://huggingface.co/docs/hub/security-tokens).
```python
from transformers import AutoTokenizer
import transformers
import torch
model = "facebook/llm-compiler-13b"
tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = transformers.pipeline(
"text-generation",
model=model,
torch_dtype=torch.float16,
device_map="auto",
)
sequences = pipeline(
'%3 = alloca i32, align 4',
do_sample=True,
top_k=10,
temperature=0.1,
top_p=0.95,
num_return_sequences=1,
eos_token_id=tokenizer.eos_token_id,
max_length=200,
)
for seq in sequences:
print(f"Result: {seq['generated_text']}")
```
## Model Details
*Note: Use of this model is governed by the Meta license. Meta developed and publicly released the LLM Compiler family of large language models (LLMs).
**Model Developers** Meta
**Variations** LLM Compiler comes in two model sizes of 7B, 13B parameters in two flavors, the foundation and instruction fine-tuned for code size and disassembly.
**This repository contains the 13 billion parameter foundation model.**
**Input** Models input text only.
**Example prompt** See `llm_compiler_demo.py` in the repo for examples of the different use cases.
**Output** Models generate text only.
**Model Architecture** LLM Compiler is an auto-regressive language model that uses an optimized transformer architecture.
**Model Dates** LLM Compiler has been trained between January 2024 and June 2024.
**Status** This is a static model trained on an offline dataset.
**License** A custom commercial license is available at: [https://ai.meta.com/resources/models-and-libraries/llama-downloads/](https://ai.meta.com/resources/models-and-libraries/llama-downloads/)
**Research Paper** More information can be found in the paper "[Meta Large Language Model Compiler: Foundation Models of Compiler Optimization](https://ai.meta.com/research/publications/meta-large-language-model-compiler-foundation-models-of-compiler-optimization/)".
## Intended Use
**Intended Use Cases** LLM Compiler is intended for commercial and research use in English, relevant programming languages, LLVM IR, x86_64 assembly and ARM assembly.
**Out-of-Scope Uses** Use in any manner that violates applicable laws or regulations (including trade compliance laws). Use in languages other than English. Use in any other way that is prohibited by the [Acceptable Use Policy](https://llama.meta.com/llama3/use-policy) and Licensing Agreement for LLM Compiler and its variants.
## Hardware and Software
**Training Factors** We used custom training libraries. The training and fine-tuning of the released models have been performed Meta’s Research Super Cluster.
**Carbon Footprint** In aggregate, training all LLM Compiler models required 14K GPU hours of computation on hardware of type A100-80GB (TDP of 350-400W), not including the training of Code Llama. 100% of the estimated tCO2eq emissions were offset by Meta’s sustainability program.
## Training Data
All experiments reported here and the released models have been trained and fine-tuned using the same data as Code Llama with different weights (see Section 2 and Table 1 in the [research paper](https://ai.meta.com/research/publications/llm-compiler-foundation-models-for-compiler-optimization/) for details).
## Evaluation Results
See evaluations for the main models and detailed ablations in Section 3 and safety evaluations in Section 4 of the research paper.
## Ethical Considerations and Limitations
LLM Compiler and its variants are a new technology that carries risks with use. Testing conducted to date has been in English, and has not covered, nor could it cover all scenarios. For these reasons, as with all LLMs, LLM Compilers’s potential outputs cannot be predicted in advance, and the model may in some instances produce inaccurate or objectionable responses to user prompts. Therefore, before deploying any applications of LLM Compiler, developers should perform safety testing and tuning tailored to their specific applications of the model.
Please see the Responsible Use Guide available available at [https://ai.meta.com/llama/responsible-use-guide](https://ai.meta.com/llama/responsible-use-guide). |
timm/levit_conv_192.fb_dist_in1k | timm | "2024-02-10T23:30:43Z" | 8,944 | 0 | timm | [
"timm",
"pytorch",
"image-classification",
"dataset:imagenet-1k",
"arxiv:2104.01136",
"license:apache-2.0",
"region:us"
] | image-classification | "2023-02-03T21:14:21Z" | ---
license: apache-2.0
library_name: timm
tags:
- image-classification
- timm
datasets:
- imagenet-1k
---
# Model card for levit_conv_192.fb_dist_in1k
A LeViT image classification model using default linear mode (non-convolutional mode with nn.Linear and nn.BatchNorm1d). Pretrained on ImageNet-1k using distillation by paper authors.
## Model Details
- **Model Type:** Image classification / feature backbone
- **Model Stats:**
- Params (M): 10.9
- GMACs: 0.7
- Activations (M): 3.2
- Image size: 224 x 224
- **Papers:**
- LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference: https://arxiv.org/abs/2104.01136
- **Original:** https://github.com/facebookresearch/LeViT
- **Dataset:** ImageNet-1k
## Model Usage
### Image Classification
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(
urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'))
model = timm.create_model('levit_conv_192.fb_dist_in1k', pretrained=True)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1
top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5)
```
### Image Embeddings
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(
urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'))
model = timm.create_model(
'levit_conv_192.fb_dist_in1k',
pretrained=True,
num_classes=0, # remove classifier nn.Linear
)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor
# or equivalently (without needing to set num_classes=0)
output = model.forward_features(transforms(img).unsqueeze(0))
# output is unpooled (ie.e a (batch_size, num_features, H, W) tensor
output = model.forward_head(output, pre_logits=True)
# output is (batch_size, num_features) tensor
```
### Feature Map Extraction
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(
urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'))
model = timm.create_model(
'levit_conv_192.fb_dist_in1k',
pretrained=True,
features_only=True,
)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1
for o in output:
# print shape of each feature map in output
# e.g. for levit_conv_256:
# torch.Size([2, 256, 14, 14])
# torch.Size([2, 384, 7, 7])
# torch.Size([2, 512, 4, 4])
print(o.shape)
```
## Model Comparison
|model |top1 |top5 |param_count|img_size|
|-----------------------------------|------|------|-----------|--------|
|levit_384.fb_dist_in1k |82.596|96.012|39.13 |224 |
|levit_conv_384.fb_dist_in1k |82.596|96.012|39.13 |224 |
|levit_256.fb_dist_in1k |81.512|95.48 |18.89 |224 |
|levit_conv_256.fb_dist_in1k |81.512|95.48 |18.89 |224 |
|levit_conv_192.fb_dist_in1k |79.86 |94.792|10.95 |224 |
|levit_192.fb_dist_in1k |79.858|94.792|10.95 |224 |
|levit_128.fb_dist_in1k |78.474|94.014|9.21 |224 |
|levit_conv_128.fb_dist_in1k |78.474|94.02 |9.21 |224 |
|levit_128s.fb_dist_in1k |76.534|92.864|7.78 |224 |
|levit_conv_128s.fb_dist_in1k |76.532|92.864|7.78 |224 |
## Citation
```bibtex
@InProceedings{Graham_2021_ICCV,
author = {Graham, Benjamin and El-Nouby, Alaaeldin and Touvron, Hugo and Stock, Pierre and Joulin, Armand and Jegou, Herve and Douze, Matthijs},
title = {LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference},
booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
month = {October},
year = {2021},
pages = {12259-12269}
}
```
```bibtex
@misc{rw2019timm,
author = {Ross Wightman},
title = {PyTorch Image Models},
year = {2019},
publisher = {GitHub},
journal = {GitHub repository},
doi = {10.5281/zenodo.4414861},
howpublished = {\url{https://github.com/rwightman/pytorch-image-models}}
}
```
|
tokyotech-llm/Swallow-7b-instruct-v0.1 | tokyotech-llm | "2024-06-29T09:00:13Z" | 8,937 | 0 | transformers | [
"transformers",
"safetensors",
"llama",
"text-generation",
"conversational",
"en",
"ja",
"arxiv:2404.17790",
"license:llama2",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2024-03-04T08:46:03Z" | ---
language:
- en
- ja
library_name: transformers
pipeline_tag: text-generation
license: llama2
model_type: llama
---
# Swallow
Our Swallow model has undergone continual pre-training from the [Llama 2 family](https://huggingface.co/meta-llama), primarily with the addition of Japanese language data. The tuned versions use supervised fine-tuning (SFT).
Links to other models can be found in the index.
# Model Release Updates
We are excited to share the release schedule for our latest models:
- **April 26, 2024**: Released version 0.1 of our enhanced instruction-tuned models: [Swallow-7b-instruct-v0.1](https://huggingface.co/tokyotech-llm/Swallow-7b-instruct-v0.1), [Swallow-13b-instruct-v0.1](https://huggingface.co/tokyotech-llm/Swallow-13b-instruct-v0.1), and [Swallow-70b-instruct-v0.1](https://huggingface.co/tokyotech-llm/Swallow-70b-instruct-v0.1) as preview versions.
- **March 2, 2024**: Released the [Swallow-7b-plus-hf](https://huggingface.co/tokyotech-llm/Swallow-7b-plus-hf), a model trained with approximately twice as many Japanese tokens as [Swallow-7b-hf](https://huggingface.co/tokyotech-llm/Swallow-7b-hf).
- **February 4, 2024**: Released the [Swallow-13b-NVE-hf](https://huggingface.co/tokyotech-llm/Swallow-13b-NVE-hf).
- **January 26, 2024**: Released the [Swallow-7b-NVE-hf](https://huggingface.co/tokyotech-llm/Swallow-7b-NVE-hf), [Swallow-7b-NVE-instruct-hf](https://huggingface.co/tokyotech-llm/Swallow-7b-NVE-instruct-hf), [Swallow-70b-NVE-hf](https://huggingface.co/tokyotech-llm/Swallow-70b-NVE-hf), and [Swallow-70b-NVE-instruct-hf](https://huggingface.co/tokyotech-llm/Swallow-70b-NVE-instruct-hf)
- **December 19, 2023**: Released the [Swallow-7b-hf](https://huggingface.co/tokyotech-llm/Swallow-7b-hf), [Swallow-7b-instruct-hf](https://huggingface.co/tokyotech-llm/Swallow-7b-instruct-hf), [Swallow-13b-hf](https://huggingface.co/tokyotech-llm/Swallow-13b-hf), [Swallow-13b-instruct-hf](https://huggingface.co/tokyotech-llm/Swallow-13b-instruct-hf), [Swallow-70b-hf](https://huggingface.co/tokyotech-llm/Swallow-70b-hf), and [Swallow-70b-instruct-hf](https://huggingface.co/tokyotech-llm/Swallow-70b-instruct-hf).
## Swallow Model Index
|Model|Swallow-hf|Swallow-instruct-hf|Swallow-instruct-v0.1|
|---|---|---|---|
|7B| [Link](https://huggingface.co/tokyotech-llm/Swallow-7b-hf) | [Link](https://huggingface.co/tokyotech-llm/Swallow-7b-instruct-hf)|[Link](https://huggingface.co/tokyotech-llm/Swallow-7b-instruct-v1.0)|
|7B-Plus| [Link](https://huggingface.co/tokyotech-llm/Swallow-7b-plus-hf) | N/A | N/A |
|13B| [Link](https://huggingface.co/tokyotech-llm/Swallow-13b-hf) | [Link](https://huggingface.co/tokyotech-llm/Swallow-13b-instruct-hf)| [Link](https://huggingface.co/tokyotech-llm/Swallow-13b-instruct-v1.0)|
|70B| [Link](https://huggingface.co/tokyotech-llm/Swallow-70b-hf) | [Link](https://huggingface.co/tokyotech-llm/Swallow-70b-instruct-hf)| [Link](https://huggingface.co/tokyotech-llm/Swallow-70b-instruct-v1.0)|
## Swallow Model Index NVE (No Vocabulary Expansion)
|Model|Swallow-NVE-hf|Swallow-NVE-instruct-hf|
|---|---|---|
|7B| [Link](https://huggingface.co/tokyotech-llm/Swallow-7b-NVE-hf) | [Link](https://huggingface.co/tokyotech-llm/Swallow-7b-NVE-instruct-hf)|
|13B| [Link](https://huggingface.co/tokyotech-llm/Swallow-13b-NVE-hf) | N/A |
|70B| [Link](https://huggingface.co/tokyotech-llm/Swallow-70b-NVE-hf) | [Link](https://huggingface.co/tokyotech-llm/Swallow-70b-NVE-instruct-hf)|
This repository provides large language models developed by [TokyoTech-LLM](https://tokyotech-llm.github.io/).
## Model Details
* **Model type**: Please refer to LLaMA-2 technical report for details on the model architecture.
* **Language(s)**: Japanese English
* **Tokenizer**: This model employs a tokenizer that features a broadened vocabulary based on Japanese data. This allows for a more efficient representation of text using fewer tokens, leading to a notably faster inference process.
* **Contact**: swallow[at]nlp.c.titech.ac.jp
## Instruct Model Performance
### MT-Bench JA
#### Comparison to the past version
* NOTE that the models with the `v0.1` suffix are newer versions compared to their original counterparts with the `hf`.
* We report overall (i.e., average over scores of the first and second turns), first, and second turn scores.
##### Overall
|Model|Average|Writing|Roleplay|Reasoning|Math|Coding|Extraction|STEM|Humanities|
|---|---|---|---|---|---|---|---|---|---|
| Swallow-7b-instruct-v0.1 |0.3435|0.4450|0.4720|0.1853|0.1920|0.2204|0.3015|0.4594|0.4720|
| Swallow-7b-instruct-hf |0.1833|0.2205|0.1975|0.1593|0.1045|0.1282|0.2672|0.1908|0.1980|
| Swallow-13b-instruct-v0.1 |0.3669|0.4816|0.5562|0.2769|0.1020|0.1505|0.4179|0.4347|0.5150|
| Swallow-13b-instruct-hf |0.2004|0.1932|0.2552|0.1507|0.1184|0.1285|0.2641|0.2434|0.2500|
| Swallow-70b-instruct-v0.1 |0.4513|0.4822|0.5353|0.3497|0.3492|0.2668|0.5553|0.4955|0.5767|
| Swallow-70b-instruct-hf |0.3259|0.2925|0.4283|0.3447|0.1562|0.1856|0.5634|0.3315|0.3071|
##### First Turn
|Model|Average|Writing|Roleplay|Reasoning|Math|Coding|Extraction|STEM|Humanities|
|---|---|---|---|---|---|---|---|---|---|
| Swallow-7b-instruct-v0.1 |0.3829|0.4960|0.4800|0.2220|0.2820|0.2164|0.3220|0.5440|0.4980|
| Swallow-7b-instruct-hf |0.2216|0.2830|0.2150|0.1590|0.1080|0.1470|0.3542|0.2450|0.2650|
| Swallow-13b-instruct-v0.1 |0.3948|0.5400|0.5220|0.3020|0.1040|0.1760|0.5040|0.5180|0.4920|
| Swallow-13b-instruct-hf |0.2304|0.2460|0.2640|0.1610|0.1360|0.1330|0.3070|0.3010|0.2950|
| Swallow-70b-instruct-v0.1 |0.4849|0.5720|0.5020|0.4780|0.3680|0.2467|0.5400|0.5720|0.5960|
| Swallow-70b-instruct-hf |0.3631|0.3420|0.4007|0.4220|0.1580|0.2044|0.6120|0.4280|0.3360|
##### Second Turn
|Model|Average|Writing|Roleplay|Reasoning|Math|Coding|Extraction|STEM|Humanities|
|---|---|---|---|---|---|---|---|---|---|
| Swallow-7b-instruct-v0.1 |0.3059|0.3940|0.4640|0.1441|0.1000|0.2253|0.2811|0.3724|0.4449|
| Swallow-7b-instruct-hf |0.1432|0.1567|0.1798|0.1603|0.1010|0.1085|0.1767|0.1343|0.1295|
| Swallow-13b-instruct-v0.1 |0.3353|0.4213|0.5911|0.2516|0.1000|0.1244|0.3194|0.3473|0.5394|
| Swallow-13b-instruct-hf |0.1692|0.1364|0.2453|0.1401|0.1000|0.1237|0.2199|0.1850|0.2050|
| Swallow-70b-instruct-v0.1 |0.4179|0.3913|0.5689|0.2184|0.3280|0.2884|0.5711|0.4171|0.5562|
| Swallow-70b-instruct-hf |0.2872|0.2398|0.4564|0.2647|0.1540|0.1676|0.5118|0.2311|0.2762|
#### Comparison to the existing models
We only provide the overall score in this section.
##### 7B models
|Model|Average|Writing|Roleplay|Reasoning|Math|Coding|Extraction|STEM|Humanities|
|---|---|---|---|---|---|---|---|---|---|
| Swallow-7b-instruct-v0.1 |0.3435|0.4450|0.4720|0.1853|0.1920|0.2204|0.3015|0.4594|0.4720|
| ELYZA-japanese-Llama-2-7b-fast-instruct |0.2827|0.3289|0.3907|0.2424|0.1480|0.1584|0.3511|0.3053|0.3365|
| calm2-7b-chat |0.3204|0.4657|0.4898|0.1837|0.1005|0.1414|0.3927|0.3601|0.4293|
| calm2-7b-chat-dpo-experimental |0.3493|0.5312|0.5237|0.1857|0.1000|0.1813|0.3355|0.4320|0.5051|
| RakutenAI-7B-instruct |0.2994|0.3623|0.3711|0.3333|0.1763|0.1581|0.4215|0.2824|0.2901|
| RakutenAI-7B-chat |0.3667|0.4229|0.4644|0.3990|0.2161|0.2390|0.3416|0.3904|0.4601|
##### 13B models
|Model|Average|Writing|Roleplay|Reasoning|Math|Coding|Extraction|STEM|Humanities|
|---|---|---|---|---|---|---|---|---|---|
| Swallow-13b-instruct-v0.1 |0.3669|0.4816|0.5562|0.2769|0.1020|0.1505|0.4179|0.4347|0.5150|
| ELYZA-japanese-Llama-2-13b-instruct |0.3196|0.4400|0.4373|0.2098|0.2157|0.1572|0.3583|0.3243|0.4141|
| ELYZA-japanese-Llama-2-13b-fast-instruct |0.3042|0.3729|0.3930|0.1236|0.2492|0.1862|0.4360|0.3233|0.3496|
##### 70B models
|Model|Average|Writing|Roleplay|Reasoning|Math|Coding|Extraction|STEM|Humanities|
|---|---|---|---|---|---|---|---|---|---|
| Swallow-70b-instruct-v0.1 |0.4513|0.4822|0.5353|0.3497|0.3492|0.2668|0.5553|0.4955|0.5767|
| japanese-stablelm-instruct-beta-70b |0.3716|0.4179|0.3945|0.3656|0.2580|0.2186|0.4412|0.4663|0.4103|
## Evaluation Benchmarks
### MT-Bench JA
We used [Japanese MT-Bench](https://wandb.ai/wandb-japan/llm-leaderboard/artifacts/dataset/mtbench_ja_question) to assess the instruction-following capabilities of models.
We utilized the following settings:
- Implemantation: FastChat [Zheng+, 2023] (commit #e86e70d0)
- Question: [Nejumi LLM-Leaderboard NEO, mtbench_ja_question_v3](https://wandb.ai/wandb-japan/llm-leaderboard/artifacts/dataset/mtbench_ja_question/v3)
- Reference Answer: [Nejumi LLM-Leaderboard NEO, mtbench_ja_referenceanswer_v1](https://wandb.ai/wandb-japan/llm-leaderboard/artifacts/dataset/mtbench_ja_referenceanswer/v1)
- Prompt for Judge: [Nejumi LLM-Lederboard NEO, mtbench_ja_prompt_v1](https://wandb.ai/wandb-japan/llm-leaderboard/artifacts/dataset/mtbench_ja_prompt/v1)
- Judge: `gpt-4-1106-preview`
- Scoring: Absolute scale normalized to a 0-1 range, averaged over five runs.
## Usage
First install additional dependencies in [requirements.txt](./requirements.txt):
```sh
pip install -r requirements.txt
```
### Instruction format Ver0.1
This format must be adhered to strictly, as deviations may result in less optimal outputs from the model.
The template used to construct a prompt for the Instruct model is specified as follows:
```
<s>[INST] <<SYS>>\n{SYSTEM_PROMPT}\n<</SYS>>\n\n{USER_MESSAGE_1} [/INST] {BOT_MESSAGE_1}</s>[INST] {USER_MESSAGE_2} [/INST]
```
Please be aware that ``<s>`` and ``</s>`` are special tokens used for the beginning of string (BOS) and end of string (EOS), respectively, while [INST] and [/INST] are considered regular strings.
For the "{SYSTEM_PROMPT}" part, We recommend using "あなたは誠実で優秀な日本人のアシスタントです。"
For the "{USER_MESSAGE_1}" part, We recommend using {instruction}\n{input}
In other words, We recommend the following:
```
<s>[INST] <<SYS>>\nあなたは誠実で優秀な日本人のアシスタントです。\n<</SYS>>\n\n{instruction1}\n{input1} [/INST] {BOT_MESSAGE_1}</s>[INST] {instruction2}\n{input2} [/INST]
```
### Use the instruct model Ver0.1
```python
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
model_name = "tokyotech-llm/Swallow-70b-instruct-v0.1"
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.bfloat16, device_map="auto")
tokenizer = AutoTokenizer.from_pretrained(model_name)
device = "cuda"
messages = [
{"role": "system", "content": "あなたは誠実で優秀な日本人のアシスタントです。"},
{"role": "user", "content": "東京工業大学の主なキャンパスについて教えてください"}
]
encodeds = tokenizer.apply_chat_template(messages, return_tensors="pt")
model_inputs = encodeds.to(device)
model.to(device)
generated_ids = model.generate(model_inputs, max_new_tokens=128, do_sample=True)
decoded = tokenizer.batch_decode(generated_ids)
print(decoded[0])
```
## Training Datasets
### Instruction Tuning Ver0.1
The following datasets were used for the instruction tuning.
- [OpenAssistant Conversations Dataset EN top-1 thread](https://huggingface.co/datasets/OpenAssistant/oasst2)
- [OpenAssistant Conversations Dataset](https://huggingface.co/datasets/llm-jp/oasst1-21k-ja) was used, where human utterances are included but the responses are not used. Instead, the responses were generated using the [Mixtral-8x7B-Instruct-v0.1](https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1) model.
## Risks and Limitations
The models released here are still in the early stages of our research and development and have not been tuned to ensure outputs align with human intent and safety considerations.
## Acknowledgements
We thank Meta Research for releasing Llama 2 under an open license for others to build on.
Our project is supported by the [ABCI Large-scale Language Model Building Support Program](https://abci.ai/en/link/llm_support_program.html) of the National Institute of Advanced Industrial Science and Technology.
## License
Llama 2 is licensed under the LLAMA 2 Community License, Copyright © Meta Platforms, Inc. All Rights Reserved.
## Authors
Here are the team members:
- From [Okazaki Laboratory](https://www.nlp.c.titech.ac.jp/index.en.html), the following members:
- [Naoaki Okazaki](https://www.chokkan.org/index.ja.html)
- [Sakae Mizuki](https://s-mizuki-nlp.github.io/)
- [Hiroki Iida](https://meshidenn.github.io/)
- [Mengsay Loem](https://loem-ms.github.io/)
- [Shota Hirai](https://huggingface.co/Kotemo428)
- [Kakeru Hattori](https://aya-se.vercel.app/)
- [Masanari Ohi](https://twitter.com/stjohn2007)
- From [YOKOTA Laboratory](https://www.rio.gsic.titech.ac.jp/en/index.html), the following members:
- [Rio Yokota](https://twitter.com/rioyokota)
- [Kazuki Fujii](https://twitter.com/okoge_kaz)
- [Taishi Nakamura](https://twitter.com/Setuna7777_2)
- [Takumi Okamoto](https://www.linkedin.com/in/takumi-okamoto)
- [Ishida Shigeki](https://www.wantedly.com/id/reborn27)
## How to cite
```
@misc{fujii2024continual,
title={Continual Pre-Training for Cross-Lingual LLM Adaptation: Enhancing Japanese Language Capabilities},
author={Kazuki Fujii and Taishi Nakamura and Mengsay Loem and Hiroki Iida and Masanari Ohi and Kakeru Hattori and Hirai Shota and Sakae Mizuki and Rio Yokota and Naoaki Okazaki},
year={2024},
eprint={2404.17790},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
``` |
QuantFactory/st-vicuna-v1.3-10.5b-taylor-GGUF | QuantFactory | "2024-06-20T17:41:49Z" | 8,925 | 0 | null | [
"gguf",
"text-generation",
"arxiv:2402.02834",
"base_model:nota-ai/st-vicuna-v1.3-10.5b-taylor",
"region:us"
] | text-generation | "2024-06-20T04:58:59Z" | ---
pipeline_tag: text-generation
base_model: nota-ai/st-vicuna-v1.3-10.5b-taylor
---
# QuantFactory/st-vicuna-v1.3-10.5b-taylor-GGUF
This is quantized version of [nota-ai/st-vicuna-v1.3-10.5b-taylor](https://huggingface.co/nota-ai/st-vicuna-v1.3-10.5b-taylor) created using llama.cpp
# Model Description
### Shortened LLaMA Model Card
Shortened LLaMA is a depth-pruned version of LLaMA models & variants for efficient text generation.
- **Developed by:** [Nota AI](https://www.nota.ai/)
- **License:** Non-commercial license
- **Repository:** https://github.com/Nota-NetsPresso/shortened-llm
- **Paper:** https://arxiv.org/abs/2402.02834
## Compression Method
After identifying unimportant Transformer blocks, we perform one-shot pruning and light LoRA-based retraining.
<details>
<summary>
Click to see a method figure.
</summary>
<img alt="method" img src="https://netspresso-research-code-release.s3.us-east-2.amazonaws.com/compressed-llm/st-llama_method.png" width="100%">
</details>
## Model Links
| Source<br>Model | Pruning<br>Ratio | Pruning<br>Criterion | HF Models<br>Link |
|:---:|:---:|:---:|:---:|
| LLaMA-1-7B | 20% | PPL | [nota-ai/st-llama-1-5.5b-ppl](https://huggingface.co/nota-ai/st-llama-1-5.5b-ppl) |
| LLaMA-1-7B | 20% | Taylor+ | [nota-ai/st-llama-1-5.5b-taylor](https://huggingface.co/nota-ai/st-llama-1-5.5b-taylor) |
| Vicuna-v1.3-7B | 20% | PPL | [nota-ai/st-vicuna-v1.3-5.5b-ppl](https://huggingface.co/nota-ai/st-vicuna-v1.3-5.5b-ppl) |
| Vicuna-v1.3-7B | 20% | Taylor+ | [nota-ai/st-vicuna-v1.3-5.5b-taylor](https://huggingface.co/nota-ai/st-vicuna-v1.3-5.5b-taylor) |
| Vicuna-v1.3-13B | 21% | PPL | [nota-ai/st-vicuna-v1.3-10.5b-ppl](https://huggingface.co/nota-ai/st-vicuna-v1.3-10.5b-ppl) |
| Vicuna-v1.3-13B | 21% | Taylor+ | [nota-ai/st-vicuna-v1.3-10.5b-taylor](https://huggingface.co/nota-ai/st-vicuna-v1.3-10.5b-taylor) |
## Zero-shot Performance & Efficiency Results
- EleutherAI/lm-evaluation-harness version [3326c54](https://github.com/EleutherAI/lm-evaluation-harness/tree/3326c547a733d598b4377e54be96e194861b964c)
<img alt="results" img src="https://netspresso-research-code-release.s3.us-east-2.amazonaws.com/compressed-llm/st-llama_zero-shot_scores.png" width="100%">
## License
- All rights related to this repository and the compressed models are reserved by Nota Inc.
- The intended use is strictly limited to research and non-commercial projects.
## Acknowledgments
- [LLM-Pruner](https://github.com/horseee/LLM-Pruner), which utilizes [LM Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness), [PEFT](https://github.com/huggingface/peft), and [Alpaca-LoRA](https://github.com/tloen/alpaca-lora). Thanks for the pioneering work on structured pruning of LLMs!
- Meta AI's [LLaMA](https://github.com/facebookresearch/llama) and LMSYS Org's [Vicuna](https://github.com/lm-sys/FastChat/blob/main/docs/vicuna_weights_version.md). Thanks for the open-source LLMs!
## Original Model Citation
```bibtex
@article{kim2024shortened,
title={Shortened LLaMA: A Simple Depth Pruning for Large Language Models},
author={Kim, Bo-Kyeong and Kim, Geonmin and Kim, Tae-Ho and Castells, Thibault and Choi, Shinkook and Shin, Junho and Song, Hyoung-Kyu},
journal={arXiv preprint arXiv:2402.02834},
year={2024},
url={https://arxiv.org/abs/2402.02834}
}
```
```bibtex
@article{kim2024mefomo,
title={Shortened LLaMA: A Simple Depth Pruning for Large Language Models},
author={Kim, Bo-Kyeong and Kim, Geonmin and Kim, Tae-Ho and Castells, Thibault and Choi, Shinkook and Shin, Junho and Song, Hyoung-Kyu},
journal={ICLR Workshop on Mathematical and Empirical Understanding of Foundation Models (ME-FoMo)},
year={2024},
url={https://openreview.net/forum?id=18VGxuOdpu}
}
``` |
mradermacher/neo_7b-slerp-GGUF | mradermacher | "2024-06-29T05:24:15Z" | 8,916 | 0 | transformers | [
"transformers",
"gguf",
"merge",
"mergekit",
"lazymergekit",
"m-a-p/neo_7b",
"en",
"base_model:DewEfresh/neo_7b-slerp",
"endpoints_compatible",
"region:us"
] | null | "2024-06-29T05:08:55Z" | ---
base_model: DewEfresh/neo_7b-slerp
language:
- en
library_name: transformers
quantized_by: mradermacher
tags:
- merge
- mergekit
- lazymergekit
- m-a-p/neo_7b
---
## About
<!-- ### quantize_version: 2 -->
<!-- ### output_tensor_quantised: 1 -->
<!-- ### convert_type: hf -->
<!-- ### vocab_type: -->
<!-- ### tags: -->
static quants of https://huggingface.co/DewEfresh/neo_7b-slerp
<!-- provided-files -->
weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion.
## Usage
If you are unsure how to use GGUF files, refer to one of [TheBloke's
READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
more details, including on how to concatenate multi-part files.
## Provided Quants
(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
| Link | Type | Size/GB | Notes |
|:-----|:-----|--------:|:------|
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.Q2_K.gguf) | Q2_K | 1.7 | |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.IQ3_XS.gguf) | IQ3_XS | 1.8 | |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.IQ3_S.gguf) | IQ3_S | 1.9 | beats Q3_K* |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.Q3_K_S.gguf) | Q3_K_S | 1.9 | |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.IQ3_M.gguf) | IQ3_M | 2.0 | |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.Q3_K_M.gguf) | Q3_K_M | 2.1 | lower quality |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.Q3_K_L.gguf) | Q3_K_L | 2.3 | |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.IQ4_XS.gguf) | IQ4_XS | 2.3 | |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.Q4_K_S.gguf) | Q4_K_S | 2.5 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.Q4_K_M.gguf) | Q4_K_M | 2.6 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.Q5_K_S.gguf) | Q5_K_S | 2.9 | |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.Q5_K_M.gguf) | Q5_K_M | 3.0 | |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.Q6_K.gguf) | Q6_K | 3.5 | very good quality |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.Q8_0.gguf) | Q8_0 | 4.5 | fast, best quality |
| [GGUF](https://huggingface.co/mradermacher/neo_7b-slerp-GGUF/resolve/main/neo_7b-slerp.f16.gguf) | f16 | 8.3 | 16 bpw, overkill |
Here is a handy graph by ikawrakow comparing some lower-quality quant
types (lower is better):
And here are Artefact2's thoughts on the matter:
https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
## FAQ / Model Request
See https://huggingface.co/mradermacher/model_requests for some answers to
questions you might have and/or if you want some other model quantized.
## Thanks
I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
me use its servers and providing upgrades to my workstation to enable
this work in my free time.
<!-- end -->
|
Helsinki-NLP/opus-mt-ca-en | Helsinki-NLP | "2023-08-16T11:26:39Z" | 8,913 | 0 | transformers | [
"transformers",
"pytorch",
"tf",
"marian",
"text2text-generation",
"translation",
"ca",
"en",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | translation | "2022-03-02T23:29:04Z" | ---
tags:
- translation
license: apache-2.0
---
### opus-mt-ca-en
* source languages: ca
* target languages: en
* OPUS readme: [ca-en](https://github.com/Helsinki-NLP/OPUS-MT-train/blob/master/models/ca-en/README.md)
* dataset: opus
* model: transformer-align
* pre-processing: normalization + SentencePiece
* download original weights: [opus-2019-12-18.zip](https://object.pouta.csc.fi/OPUS-MT-models/ca-en/opus-2019-12-18.zip)
* test set translations: [opus-2019-12-18.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/ca-en/opus-2019-12-18.test.txt)
* test set scores: [opus-2019-12-18.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/ca-en/opus-2019-12-18.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| Tatoeba.ca.en | 51.4 | 0.678 |
|
mradermacher/Ramses-II-GGUF | mradermacher | "2024-06-23T02:51:55Z" | 8,912 | 0 | transformers | [
"transformers",
"gguf",
"mergekit",
"merge",
"en",
"base_model:CoprolaliacPress/Ramses-II",
"endpoints_compatible",
"region:us"
] | null | "2024-06-23T02:25:30Z" | ---
base_model: CoprolaliacPress/Ramses-II
language:
- en
library_name: transformers
quantized_by: mradermacher
tags:
- mergekit
- merge
---
## About
<!-- ### quantize_version: 2 -->
<!-- ### output_tensor_quantised: 1 -->
<!-- ### convert_type: hf -->
<!-- ### vocab_type: -->
<!-- ### tags: -->
static quants of https://huggingface.co/CoprolaliacPress/Ramses-II
<!-- provided-files -->
weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion.
## Usage
If you are unsure how to use GGUF files, refer to one of [TheBloke's
READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
more details, including on how to concatenate multi-part files.
## Provided Quants
(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
| Link | Type | Size/GB | Notes |
|:-----|:-----|--------:|:------|
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.Q2_K.gguf) | Q2_K | 2.8 | |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.IQ3_XS.gguf) | IQ3_XS | 3.1 | |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.Q3_K_S.gguf) | Q3_K_S | 3.3 | |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.IQ3_S.gguf) | IQ3_S | 3.3 | beats Q3_K* |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.IQ3_M.gguf) | IQ3_M | 3.4 | |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.Q3_K_M.gguf) | Q3_K_M | 3.6 | lower quality |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.Q3_K_L.gguf) | Q3_K_L | 3.9 | |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.IQ4_XS.gguf) | IQ4_XS | 4.0 | |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.Q4_K_S.gguf) | Q4_K_S | 4.2 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.Q4_K_M.gguf) | Q4_K_M | 4.5 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.Q5_K_S.gguf) | Q5_K_S | 5.1 | |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.Q5_K_M.gguf) | Q5_K_M | 5.2 | |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.Q6_K.gguf) | Q6_K | 6.0 | very good quality |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.Q8_0.gguf) | Q8_0 | 7.8 | fast, best quality |
| [GGUF](https://huggingface.co/mradermacher/Ramses-II-GGUF/resolve/main/Ramses-II.f16.gguf) | f16 | 14.6 | 16 bpw, overkill |
Here is a handy graph by ikawrakow comparing some lower-quality quant
types (lower is better):
And here are Artefact2's thoughts on the matter:
https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
## FAQ / Model Request
See https://huggingface.co/mradermacher/model_requests for some answers to
questions you might have and/or if you want some other model quantized.
## Thanks
I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
me use its servers and providing upgrades to my workstation to enable
this work in my free time.
<!-- end -->
|
ZeroWw/Llama-3-8B-Instruct-Gradient-1048k-GGUF | ZeroWw | "2024-06-26T00:51:29Z" | 8,911 | 0 | null | [
"gguf",
"en",
"license:mit",
"region:us"
] | null | "2024-06-26T00:39:07Z" |
---
license: mit
language:
- en
---
My own (ZeroWw) quantizations.
output and embed tensors quantized to f16.
all other tensors quantized to q5_k or q6_k.
Result:
both f16.q6 and f16.q5 are smaller than q8_0 standard quantization
and they perform as well as the pure f16.
|
Lumpen1/MadWizard-SFT-v4-Mistral-7b-v0.3-gguf | Lumpen1 | "2024-06-25T07:12:38Z" | 8,910 | 0 | null | [
"gguf",
"license:apache-2.0",
"region:us"
] | null | "2024-06-25T06:56:49Z" | ---
license: apache-2.0
---
# MadWizard-SFT-v4-Mistral-7b-v0.3-gguf
GGUF version of Lumpen1/MadWizard-SFT-v4-Mistral-7b-v0.3 |
TalTechNLP/voxlingua107-epaca-tdnn | TalTechNLP | "2021-11-04T13:37:27Z" | 8,909 | 26 | speechbrain | [
"speechbrain",
"audio-classification",
"embeddings",
"Language",
"Identification",
"pytorch",
"ECAPA-TDNN",
"TDNN",
"VoxLingua107",
"multilingual",
"dataset:VoxLingua107",
"license:apache-2.0",
"region:us"
] | audio-classification | "2022-03-02T23:29:05Z" | ---
language: multilingual
thumbnail:
tags:
- audio-classification
- speechbrain
- embeddings
- Language
- Identification
- pytorch
- ECAPA-TDNN
- TDNN
- VoxLingua107
license: "apache-2.0"
datasets:
- VoxLingua107
metrics:
- Accuracy
widget:
- example_title: English Sample
src: https://cdn-media.huggingface.co/speech_samples/LibriSpeech_61-70968-0000.flac
---
# VoxLingua107 ECAPA-TDNN Spoken Language Identification Model
## Model description
This is a spoken language recognition model trained on the VoxLingua107 dataset using SpeechBrain.
The model uses the ECAPA-TDNN architecture that has previously been used for speaker recognition.
The model can classify a speech utterance according to the language spoken.
It covers 107 different languages (
Abkhazian,
Afrikaans,
Amharic,
Arabic,
Assamese,
Azerbaijani,
Bashkir,
Belarusian,
Bulgarian,
Bengali,
Tibetan,
Breton,
Bosnian,
Catalan,
Cebuano,
Czech,
Welsh,
Danish,
German,
Greek,
English,
Esperanto,
Spanish,
Estonian,
Basque,
Persian,
Finnish,
Faroese,
French,
Galician,
Guarani,
Gujarati,
Manx,
Hausa,
Hawaiian,
Hindi,
Croatian,
Haitian,
Hungarian,
Armenian,
Interlingua,
Indonesian,
Icelandic,
Italian,
Hebrew,
Japanese,
Javanese,
Georgian,
Kazakh,
Central Khmer,
Kannada,
Korean,
Latin,
Luxembourgish,
Lingala,
Lao,
Lithuanian,
Latvian,
Malagasy,
Maori,
Macedonian,
Malayalam,
Mongolian,
Marathi,
Malay,
Maltese,
Burmese,
Nepali,
Dutch,
Norwegian Nynorsk,
Norwegian,
Occitan,
Panjabi,
Polish,
Pushto,
Portuguese,
Romanian,
Russian,
Sanskrit,
Scots,
Sindhi,
Sinhala,
Slovak,
Slovenian,
Shona,
Somali,
Albanian,
Serbian,
Sundanese,
Swedish,
Swahili,
Tamil,
Telugu,
Tajik,
Thai,
Turkmen,
Tagalog,
Turkish,
Tatar,
Ukrainian,
Urdu,
Uzbek,
Vietnamese,
Waray,
Yiddish,
Yoruba,
Mandarin Chinese).
## Intended uses & limitations
The model has two uses:
- use 'as is' for spoken language recognition
- use as an utterance-level feature (embedding) extractor, for creating a dedicated language ID model on your own data
The model is trained on automatically collected YouTube data. For more
information about the dataset, see [here](http://bark.phon.ioc.ee/voxlingua107/).
#### How to use
```python
import torchaudio
from speechbrain.pretrained import EncoderClassifier
language_id = EncoderClassifier.from_hparams(source="TalTechNLP/voxlingua107-epaca-tdnn", savedir="tmp")
# Download Thai language sample from Omniglot and cvert to suitable form
signal = language_id.load_audio("https://omniglot.com/soundfiles/udhr/udhr_th.mp3")
prediction = language_id.classify_batch(signal)
print(prediction)
(tensor([[0.3210, 0.3751, 0.3680, 0.3939, 0.4026, 0.3644, 0.3689, 0.3597, 0.3508,
0.3666, 0.3895, 0.3978, 0.3848, 0.3957, 0.3949, 0.3586, 0.4360, 0.3997,
0.4106, 0.3886, 0.4177, 0.3870, 0.3764, 0.3763, 0.3672, 0.4000, 0.4256,
0.4091, 0.3563, 0.3695, 0.3320, 0.3838, 0.3850, 0.3867, 0.3878, 0.3944,
0.3924, 0.4063, 0.3803, 0.3830, 0.2996, 0.4187, 0.3976, 0.3651, 0.3950,
0.3744, 0.4295, 0.3807, 0.3613, 0.4710, 0.3530, 0.4156, 0.3651, 0.3777,
0.3813, 0.6063, 0.3708, 0.3886, 0.3766, 0.4023, 0.3785, 0.3612, 0.4193,
0.3720, 0.4406, 0.3243, 0.3866, 0.3866, 0.4104, 0.4294, 0.4175, 0.3364,
0.3595, 0.3443, 0.3565, 0.3776, 0.3985, 0.3778, 0.2382, 0.4115, 0.4017,
0.4070, 0.3266, 0.3648, 0.3888, 0.3907, 0.3755, 0.3631, 0.4460, 0.3464,
0.3898, 0.3661, 0.3883, 0.3772, 0.9289, 0.3687, 0.4298, 0.4211, 0.3838,
0.3521, 0.3515, 0.3465, 0.4772, 0.4043, 0.3844, 0.3973, 0.4343]]), tensor([0.9289]), tensor([94]), ['th'])
# The scores in the prediction[0] tensor can be interpreted as cosine scores between
# the languages and the given utterance (i.e., the larger the better)
# The identified language ISO code is given in prediction[3]
print(prediction[3])
['th']
# Alternatively, use the utterance embedding extractor:
emb = language_id.encode_batch(signal)
print(emb.shape)
torch.Size([1, 1, 256])
```
#### Limitations and bias
Since the model is trained on VoxLingua107, it has many limitations and biases, some of which are:
- Probably it's accuracy on smaller languages is quite limited
- Probably it works worse on female speech than male speech (because YouTube data includes much more male speech)
- Based on subjective experiments, it doesn't work well on speech with a foreign accent
- Probably it doesn't work well on children's speech and on persons with speech disorders
## Training data
The model is trained on [VoxLingua107](http://bark.phon.ioc.ee/voxlingua107/).
VoxLingua107 is a speech dataset for training spoken language identification models.
The dataset consists of short speech segments automatically extracted from YouTube videos and labeled according the language of the video title and description, with some post-processing steps to filter out false positives.
VoxLingua107 contains data for 107 languages. The total amount of speech in the training set is 6628 hours.
The average amount of data per language is 62 hours. However, the real amount per language varies a lot. There is also a seperate development set containing 1609 speech segments from 33 languages, validated by at least two volunteers to really contain the given language.
## Training procedure
We used [SpeechBrain](https://github.com/speechbrain/speechbrain) to train the model.
Training recipe will be published soon.
## Evaluation results
Error rate: 7% on the development dataset
### BibTeX entry and citation info
```bibtex
@inproceedings{valk2021slt,
title={{VoxLingua107}: a Dataset for Spoken Language Recognition},
author={J{\"o}rgen Valk and Tanel Alum{\"a}e},
booktitle={Proc. IEEE SLT Workshop},
year={2021},
}
```
|
HorcruxNo13/swinv2-tiny-patch4-window8-256 | HorcruxNo13 | "2023-09-04T18:15:04Z" | 8,896 | 0 | transformers | [
"transformers",
"pytorch",
"swinv2",
"image-classification",
"generated_from_trainer",
"dataset:imagefolder",
"base_model:microsoft/swinv2-tiny-patch4-window8-256",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | image-classification | "2023-09-04T18:09:51Z" | ---
license: apache-2.0
base_model: microsoft/swinv2-tiny-patch4-window8-256
tags:
- generated_from_trainer
datasets:
- imagefolder
metrics:
- accuracy
model-index:
- name: swinv2-tiny-patch4-window8-256
results:
- task:
name: Image Classification
type: image-classification
dataset:
name: imagefolder
type: imagefolder
config: default
split: validation
args: default
metrics:
- name: Accuracy
type: accuracy
value: 0.73
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# swinv2-tiny-patch4-window8-256
This model is a fine-tuned version of [microsoft/swinv2-tiny-patch4-window8-256](https://huggingface.co/microsoft/swinv2-tiny-patch4-window8-256) on the imagefolder dataset.
It achieves the following results on the evaluation set:
- Loss: 1.4461
- Accuracy: 0.73
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| No log | 1.0 | 8 | 1.3942 | 0.7333 |
| 4.8462 | 2.0 | 16 | 0.5923 | 0.7333 |
| 0.8566 | 3.0 | 24 | 0.6450 | 0.7333 |
### Framework versions
- Transformers 4.32.1
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3
|
RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf | RichardErkhov | "2024-06-28T16:14:21Z" | 8,891 | 0 | null | [
"gguf",
"arxiv:1910.09700",
"region:us"
] | null | "2024-06-28T15:38:55Z" | Quantization made by Richard Erkhov.
[Github](https://github.com/RichardErkhov)
[Discord](https://discord.gg/pvy7H8DZMG)
[Request more models](https://github.com/RichardErkhov/quant_request)
Meta-Llama-3-2B-mlp-layer-pruned - GGUF
- Model creator: https://huggingface.co/juewang/
- Original model: https://huggingface.co/juewang/Meta-Llama-3-2B-mlp-layer-pruned/
| Name | Quant method | Size |
| ---- | ---- | ---- |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q2_K.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q2_K.gguf) | Q2_K | 1.09GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.IQ3_XS.gguf) | IQ3_XS | 1.19GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.IQ3_S.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.IQ3_S.gguf) | IQ3_S | 1.22GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q3_K_S.gguf) | Q3_K_S | 1.21GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.IQ3_M.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.IQ3_M.gguf) | IQ3_M | 1.26GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q3_K.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q3_K.gguf) | Q3_K | 1.29GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q3_K_M.gguf) | Q3_K_M | 1.29GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q3_K_L.gguf) | Q3_K_L | 1.36GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.IQ4_XS.gguf) | IQ4_XS | 1.41GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q4_0.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q4_0.gguf) | Q4_0 | 1.46GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.IQ4_NL.gguf) | IQ4_NL | 1.47GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q4_K_S.gguf) | Q4_K_S | 1.46GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q4_K.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q4_K.gguf) | Q4_K | 1.5GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q4_K_M.gguf) | Q4_K_M | 1.5GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q4_1.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q4_1.gguf) | Q4_1 | 1.57GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q5_0.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q5_0.gguf) | Q5_0 | 1.69GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q5_K_S.gguf) | Q5_K_S | 1.69GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q5_K.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q5_K.gguf) | Q5_K | 1.71GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q5_K_M.gguf) | Q5_K_M | 1.71GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q5_1.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q5_1.gguf) | Q5_1 | 1.81GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q6_K.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q6_K.gguf) | Q6_K | 1.94GB |
| [Meta-Llama-3-2B-mlp-layer-pruned.Q8_0.gguf](https://huggingface.co/RichardErkhov/juewang_-_Meta-Llama-3-2B-mlp-layer-pruned-gguf/blob/main/Meta-Llama-3-2B-mlp-layer-pruned.Q8_0.gguf) | Q8_0 | 2.51GB |
Original model description:
---
library_name: transformers
tags: []
---
# Model Card for Model ID
<!-- Provide a quick summary of what the model is/does. -->
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **Developed by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Model type:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
- **Finetuned from model [optional]:** [More Information Needed]
### Model Sources [optional]
<!-- Provide the basic links for the model. -->
- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
### Direct Use
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
[More Information Needed]
### Downstream Use [optional]
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
## How to Get Started with the Model
Use the code below to get started with the model.
[More Information Needed]
## Training Details
### Training Data
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
[More Information Needed]
### Training Procedure
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
#### Preprocessing [optional]
[More Information Needed]
#### Training Hyperparameters
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
#### Speeds, Sizes, Times [optional]
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
[More Information Needed]
## Evaluation
<!-- This section describes the evaluation protocols and provides the results. -->
### Testing Data, Factors & Metrics
#### Testing Data
<!-- This should link to a Dataset Card if possible. -->
[More Information Needed]
#### Factors
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
[More Information Needed]
#### Metrics
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
[More Information Needed]
### Results
[More Information Needed]
#### Summary
## Model Examination [optional]
<!-- Relevant interpretability work for the model goes here -->
[More Information Needed]
## Environmental Impact
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
- **Hardware Type:** [More Information Needed]
- **Hours used:** [More Information Needed]
- **Cloud Provider:** [More Information Needed]
- **Compute Region:** [More Information Needed]
- **Carbon Emitted:** [More Information Needed]
## Technical Specifications [optional]
### Model Architecture and Objective
[More Information Needed]
### Compute Infrastructure
[More Information Needed]
#### Hardware
[More Information Needed]
#### Software
[More Information Needed]
## Citation [optional]
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Model Card Authors [optional]
[More Information Needed]
## Model Card Contact
[More Information Needed]
|
wangqixun/YamerMIX_v8 | wangqixun | "2024-01-30T07:49:12Z" | 8,885 | 11 | diffusers | [
"diffusers",
"safetensors",
"license:mit",
"diffusers:StableDiffusionXLCommonPipeline",
"region:us"
] | null | "2024-01-22T11:18:57Z" | ---
license: mit
---
# Model
The model is from [civitai-Yamer](https://civitai.com/models/84040?modelVersionId=196039). This is a very excellent model！Thank you Yamer!
For business inquires, commercial licensing, custom models/commissions, large scale image captioning for datasets and consultation contact me under [email protected]
|
timm/maxvit_tiny_tf_512.in1k | timm | "2023-05-11T00:24:02Z" | 8,874 | 0 | timm | [
"timm",
"pytorch",
"safetensors",
"image-classification",
"dataset:imagenet-1k",
"arxiv:2204.01697",
"license:apache-2.0",
"region:us"
] | image-classification | "2022-12-02T21:57:56Z" | ---
tags:
- image-classification
- timm
library_name: timm
license: apache-2.0
datasets:
- imagenet-1k
---
# Model card for maxvit_tiny_tf_512.in1k
An official MaxViT image classification model. Trained in tensorflow on ImageNet-1k by paper authors.
Ported from official Tensorflow implementation (https://github.com/google-research/maxvit) to PyTorch by Ross Wightman.
### Model Variants in [maxxvit.py](https://github.com/huggingface/pytorch-image-models/blob/main/timm/models/maxxvit.py)
MaxxViT covers a number of related model architectures that share a common structure including:
- CoAtNet - Combining MBConv (depthwise-separable) convolutional blocks in early stages with self-attention transformer blocks in later stages.
- MaxViT - Uniform blocks across all stages, each containing a MBConv (depthwise-separable) convolution block followed by two self-attention blocks with different partitioning schemes (window followed by grid).
- CoAtNeXt - A timm specific arch that uses ConvNeXt blocks in place of MBConv blocks in CoAtNet. All normalization layers are LayerNorm (no BatchNorm).
- MaxxViT - A timm specific arch that uses ConvNeXt blocks in place of MBConv blocks in MaxViT. All normalization layers are LayerNorm (no BatchNorm).
- MaxxViT-V2 - A MaxxViT variation that removes the window block attention leaving only ConvNeXt blocks and grid attention w/ more width to compensate.
Aside from the major variants listed above, there are more subtle changes from model to model. Any model name with the string `rw` are `timm` specific configs w/ modelling adjustments made to favour PyTorch eager use. These were created while training initial reproductions of the models so there are variations.
All models with the string `tf` are models exactly matching Tensorflow based models by the original paper authors with weights ported to PyTorch. This covers a number of MaxViT models. The official CoAtNet models were never released.
## Model Details
- **Model Type:** Image classification / feature backbone
- **Model Stats:**
- Params (M): 31.0
- GMACs: 33.5
- Activations (M): 257.6
- Image size: 512 x 512
- **Papers:**
- MaxViT: Multi-Axis Vision Transformer: https://arxiv.org/abs/2204.01697
- **Dataset:** ImageNet-1k
## Model Usage
### Image Classification
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model('maxvit_tiny_tf_512.in1k', pretrained=True)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1
top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5)
```
### Feature Map Extraction
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model(
'maxvit_tiny_tf_512.in1k',
pretrained=True,
features_only=True,
)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1
for o in output:
# print shape of each feature map in output
# e.g.:
# torch.Size([1, 64, 256, 256])
# torch.Size([1, 64, 128, 128])
# torch.Size([1, 128, 64, 64])
# torch.Size([1, 256, 32, 32])
# torch.Size([1, 512, 16, 16])
print(o.shape)
```
### Image Embeddings
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model(
'maxvit_tiny_tf_512.in1k',
pretrained=True,
num_classes=0, # remove classifier nn.Linear
)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor
# or equivalently (without needing to set num_classes=0)
output = model.forward_features(transforms(img).unsqueeze(0))
# output is unpooled, a (1, 512, 16, 16) shaped tensor
output = model.forward_head(output, pre_logits=True)
# output is a (1, num_features) shaped tensor
```
## Model Comparison
### By Top-1
|model |top1 |top5 |samples / sec |Params (M) |GMAC |Act (M)|
|------------------------------------------------------------------------------------------------------------------------|----:|----:|--------------:|--------------:|-----:|------:|
|[maxvit_xlarge_tf_512.in21k_ft_in1k](https://huggingface.co/timm/maxvit_xlarge_tf_512.in21k_ft_in1k) |88.53|98.64| 21.76| 475.77|534.14|1413.22|
|[maxvit_xlarge_tf_384.in21k_ft_in1k](https://huggingface.co/timm/maxvit_xlarge_tf_384.in21k_ft_in1k) |88.32|98.54| 42.53| 475.32|292.78| 668.76|
|[maxvit_base_tf_512.in21k_ft_in1k](https://huggingface.co/timm/maxvit_base_tf_512.in21k_ft_in1k) |88.20|98.53| 50.87| 119.88|138.02| 703.99|
|[maxvit_large_tf_512.in21k_ft_in1k](https://huggingface.co/timm/maxvit_large_tf_512.in21k_ft_in1k) |88.04|98.40| 36.42| 212.33|244.75| 942.15|
|[maxvit_large_tf_384.in21k_ft_in1k](https://huggingface.co/timm/maxvit_large_tf_384.in21k_ft_in1k) |87.98|98.56| 71.75| 212.03|132.55| 445.84|
|[maxvit_base_tf_384.in21k_ft_in1k](https://huggingface.co/timm/maxvit_base_tf_384.in21k_ft_in1k) |87.92|98.54| 104.71| 119.65| 73.80| 332.90|
|[maxvit_rmlp_base_rw_384.sw_in12k_ft_in1k](https://huggingface.co/timm/maxvit_rmlp_base_rw_384.sw_in12k_ft_in1k) |87.81|98.37| 106.55| 116.14| 70.97| 318.95|
|[maxxvitv2_rmlp_base_rw_384.sw_in12k_ft_in1k](https://huggingface.co/timm/maxxvitv2_rmlp_base_rw_384.sw_in12k_ft_in1k) |87.47|98.37| 149.49| 116.09| 72.98| 213.74|
|[coatnet_rmlp_2_rw_384.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_rmlp_2_rw_384.sw_in12k_ft_in1k) |87.39|98.31| 160.80| 73.88| 47.69| 209.43|
|[maxvit_rmlp_base_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/maxvit_rmlp_base_rw_224.sw_in12k_ft_in1k) |86.89|98.02| 375.86| 116.14| 23.15| 92.64|
|[maxxvitv2_rmlp_base_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/maxxvitv2_rmlp_base_rw_224.sw_in12k_ft_in1k) |86.64|98.02| 501.03| 116.09| 24.20| 62.77|
|[maxvit_base_tf_512.in1k](https://huggingface.co/timm/maxvit_base_tf_512.in1k) |86.60|97.92| 50.75| 119.88|138.02| 703.99|
|[coatnet_2_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_2_rw_224.sw_in12k_ft_in1k) |86.57|97.89| 631.88| 73.87| 15.09| 49.22|
|[maxvit_large_tf_512.in1k](https://huggingface.co/timm/maxvit_large_tf_512.in1k) |86.52|97.88| 36.04| 212.33|244.75| 942.15|
|[coatnet_rmlp_2_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_rmlp_2_rw_224.sw_in12k_ft_in1k) |86.49|97.90| 620.58| 73.88| 15.18| 54.78|
|[maxvit_base_tf_384.in1k](https://huggingface.co/timm/maxvit_base_tf_384.in1k) |86.29|97.80| 101.09| 119.65| 73.80| 332.90|
|[maxvit_large_tf_384.in1k](https://huggingface.co/timm/maxvit_large_tf_384.in1k) |86.23|97.69| 70.56| 212.03|132.55| 445.84|
|[maxvit_small_tf_512.in1k](https://huggingface.co/timm/maxvit_small_tf_512.in1k) |86.10|97.76| 88.63| 69.13| 67.26| 383.77|
|[maxvit_tiny_tf_512.in1k](https://huggingface.co/timm/maxvit_tiny_tf_512.in1k) |85.67|97.58| 144.25| 31.05| 33.49| 257.59|
|[maxvit_small_tf_384.in1k](https://huggingface.co/timm/maxvit_small_tf_384.in1k) |85.54|97.46| 188.35| 69.02| 35.87| 183.65|
|[maxvit_tiny_tf_384.in1k](https://huggingface.co/timm/maxvit_tiny_tf_384.in1k) |85.11|97.38| 293.46| 30.98| 17.53| 123.42|
|[maxvit_large_tf_224.in1k](https://huggingface.co/timm/maxvit_large_tf_224.in1k) |84.93|96.97| 247.71| 211.79| 43.68| 127.35|
|[coatnet_rmlp_1_rw2_224.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_rmlp_1_rw2_224.sw_in12k_ft_in1k) |84.90|96.96| 1025.45| 41.72| 8.11| 40.13|
|[maxvit_base_tf_224.in1k](https://huggingface.co/timm/maxvit_base_tf_224.in1k) |84.85|96.99| 358.25| 119.47| 24.04| 95.01|
|[maxxvit_rmlp_small_rw_256.sw_in1k](https://huggingface.co/timm/maxxvit_rmlp_small_rw_256.sw_in1k) |84.63|97.06| 575.53| 66.01| 14.67| 58.38|
|[coatnet_rmlp_2_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_rmlp_2_rw_224.sw_in1k) |84.61|96.74| 625.81| 73.88| 15.18| 54.78|
|[maxvit_rmlp_small_rw_224.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_small_rw_224.sw_in1k) |84.49|96.76| 693.82| 64.90| 10.75| 49.30|
|[maxvit_small_tf_224.in1k](https://huggingface.co/timm/maxvit_small_tf_224.in1k) |84.43|96.83| 647.96| 68.93| 11.66| 53.17|
|[maxvit_rmlp_tiny_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_tiny_rw_256.sw_in1k) |84.23|96.78| 807.21| 29.15| 6.77| 46.92|
|[coatnet_1_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_1_rw_224.sw_in1k) |83.62|96.38| 989.59| 41.72| 8.04| 34.60|
|[maxvit_tiny_rw_224.sw_in1k](https://huggingface.co/timm/maxvit_tiny_rw_224.sw_in1k) |83.50|96.50| 1100.53| 29.06| 5.11| 33.11|
|[maxvit_tiny_tf_224.in1k](https://huggingface.co/timm/maxvit_tiny_tf_224.in1k) |83.41|96.59| 1004.94| 30.92| 5.60| 35.78|
|[coatnet_rmlp_1_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_rmlp_1_rw_224.sw_in1k) |83.36|96.45| 1093.03| 41.69| 7.85| 35.47|
|[maxxvitv2_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxxvitv2_nano_rw_256.sw_in1k) |83.11|96.33| 1276.88| 23.70| 6.26| 23.05|
|[maxxvit_rmlp_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxxvit_rmlp_nano_rw_256.sw_in1k) |83.03|96.34| 1341.24| 16.78| 4.37| 26.05|
|[maxvit_rmlp_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_nano_rw_256.sw_in1k) |82.96|96.26| 1283.24| 15.50| 4.47| 31.92|
|[maxvit_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_nano_rw_256.sw_in1k) |82.93|96.23| 1218.17| 15.45| 4.46| 30.28|
|[coatnet_bn_0_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_bn_0_rw_224.sw_in1k) |82.39|96.19| 1600.14| 27.44| 4.67| 22.04|
|[coatnet_0_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_0_rw_224.sw_in1k) |82.39|95.84| 1831.21| 27.44| 4.43| 18.73|
|[coatnet_rmlp_nano_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_rmlp_nano_rw_224.sw_in1k) |82.05|95.87| 2109.09| 15.15| 2.62| 20.34|
|[coatnext_nano_rw_224.sw_in1k](https://huggingface.co/timm/coatnext_nano_rw_224.sw_in1k) |81.95|95.92| 2525.52| 14.70| 2.47| 12.80|
|[coatnet_nano_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_nano_rw_224.sw_in1k) |81.70|95.64| 2344.52| 15.14| 2.41| 15.41|
|[maxvit_rmlp_pico_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_pico_rw_256.sw_in1k) |80.53|95.21| 1594.71| 7.52| 1.85| 24.86|
### By Throughput (samples / sec)
|model |top1 |top5 |samples / sec |Params (M) |GMAC |Act (M)|
|------------------------------------------------------------------------------------------------------------------------|----:|----:|--------------:|--------------:|-----:|------:|
|[coatnext_nano_rw_224.sw_in1k](https://huggingface.co/timm/coatnext_nano_rw_224.sw_in1k) |81.95|95.92| 2525.52| 14.70| 2.47| 12.80|
|[coatnet_nano_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_nano_rw_224.sw_in1k) |81.70|95.64| 2344.52| 15.14| 2.41| 15.41|
|[coatnet_rmlp_nano_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_rmlp_nano_rw_224.sw_in1k) |82.05|95.87| 2109.09| 15.15| 2.62| 20.34|
|[coatnet_0_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_0_rw_224.sw_in1k) |82.39|95.84| 1831.21| 27.44| 4.43| 18.73|
|[coatnet_bn_0_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_bn_0_rw_224.sw_in1k) |82.39|96.19| 1600.14| 27.44| 4.67| 22.04|
|[maxvit_rmlp_pico_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_pico_rw_256.sw_in1k) |80.53|95.21| 1594.71| 7.52| 1.85| 24.86|
|[maxxvit_rmlp_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxxvit_rmlp_nano_rw_256.sw_in1k) |83.03|96.34| 1341.24| 16.78| 4.37| 26.05|
|[maxvit_rmlp_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_nano_rw_256.sw_in1k) |82.96|96.26| 1283.24| 15.50| 4.47| 31.92|
|[maxxvitv2_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxxvitv2_nano_rw_256.sw_in1k) |83.11|96.33| 1276.88| 23.70| 6.26| 23.05|
|[maxvit_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_nano_rw_256.sw_in1k) |82.93|96.23| 1218.17| 15.45| 4.46| 30.28|
|[maxvit_tiny_rw_224.sw_in1k](https://huggingface.co/timm/maxvit_tiny_rw_224.sw_in1k) |83.50|96.50| 1100.53| 29.06| 5.11| 33.11|
|[coatnet_rmlp_1_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_rmlp_1_rw_224.sw_in1k) |83.36|96.45| 1093.03| 41.69| 7.85| 35.47|
|[coatnet_rmlp_1_rw2_224.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_rmlp_1_rw2_224.sw_in12k_ft_in1k) |84.90|96.96| 1025.45| 41.72| 8.11| 40.13|
|[maxvit_tiny_tf_224.in1k](https://huggingface.co/timm/maxvit_tiny_tf_224.in1k) |83.41|96.59| 1004.94| 30.92| 5.60| 35.78|
|[coatnet_1_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_1_rw_224.sw_in1k) |83.62|96.38| 989.59| 41.72| 8.04| 34.60|
|[maxvit_rmlp_tiny_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_tiny_rw_256.sw_in1k) |84.23|96.78| 807.21| 29.15| 6.77| 46.92|
|[maxvit_rmlp_small_rw_224.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_small_rw_224.sw_in1k) |84.49|96.76| 693.82| 64.90| 10.75| 49.30|
|[maxvit_small_tf_224.in1k](https://huggingface.co/timm/maxvit_small_tf_224.in1k) |84.43|96.83| 647.96| 68.93| 11.66| 53.17|
|[coatnet_2_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_2_rw_224.sw_in12k_ft_in1k) |86.57|97.89| 631.88| 73.87| 15.09| 49.22|
|[coatnet_rmlp_2_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_rmlp_2_rw_224.sw_in1k) |84.61|96.74| 625.81| 73.88| 15.18| 54.78|
|[coatnet_rmlp_2_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_rmlp_2_rw_224.sw_in12k_ft_in1k) |86.49|97.90| 620.58| 73.88| 15.18| 54.78|
|[maxxvit_rmlp_small_rw_256.sw_in1k](https://huggingface.co/timm/maxxvit_rmlp_small_rw_256.sw_in1k) |84.63|97.06| 575.53| 66.01| 14.67| 58.38|
|[maxxvitv2_rmlp_base_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/maxxvitv2_rmlp_base_rw_224.sw_in12k_ft_in1k) |86.64|98.02| 501.03| 116.09| 24.20| 62.77|
|[maxvit_rmlp_base_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/maxvit_rmlp_base_rw_224.sw_in12k_ft_in1k) |86.89|98.02| 375.86| 116.14| 23.15| 92.64|
|[maxvit_base_tf_224.in1k](https://huggingface.co/timm/maxvit_base_tf_224.in1k) |84.85|96.99| 358.25| 119.47| 24.04| 95.01|
|[maxvit_tiny_tf_384.in1k](https://huggingface.co/timm/maxvit_tiny_tf_384.in1k) |85.11|97.38| 293.46| 30.98| 17.53| 123.42|
|[maxvit_large_tf_224.in1k](https://huggingface.co/timm/maxvit_large_tf_224.in1k) |84.93|96.97| 247.71| 211.79| 43.68| 127.35|
|[maxvit_small_tf_384.in1k](https://huggingface.co/timm/maxvit_small_tf_384.in1k) |85.54|97.46| 188.35| 69.02| 35.87| 183.65|
|[coatnet_rmlp_2_rw_384.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_rmlp_2_rw_384.sw_in12k_ft_in1k) |87.39|98.31| 160.80| 73.88| 47.69| 209.43|
|[maxxvitv2_rmlp_base_rw_384.sw_in12k_ft_in1k](https://huggingface.co/timm/maxxvitv2_rmlp_base_rw_384.sw_in12k_ft_in1k) |87.47|98.37| 149.49| 116.09| 72.98| 213.74|
|[maxvit_tiny_tf_512.in1k](https://huggingface.co/timm/maxvit_tiny_tf_512.in1k) |85.67|97.58| 144.25| 31.05| 33.49| 257.59|
|[maxvit_rmlp_base_rw_384.sw_in12k_ft_in1k](https://huggingface.co/timm/maxvit_rmlp_base_rw_384.sw_in12k_ft_in1k) |87.81|98.37| 106.55| 116.14| 70.97| 318.95|
|[maxvit_base_tf_384.in21k_ft_in1k](https://huggingface.co/timm/maxvit_base_tf_384.in21k_ft_in1k) |87.92|98.54| 104.71| 119.65| 73.80| 332.90|
|[maxvit_base_tf_384.in1k](https://huggingface.co/timm/maxvit_base_tf_384.in1k) |86.29|97.80| 101.09| 119.65| 73.80| 332.90|
|[maxvit_small_tf_512.in1k](https://huggingface.co/timm/maxvit_small_tf_512.in1k) |86.10|97.76| 88.63| 69.13| 67.26| 383.77|
|[maxvit_large_tf_384.in21k_ft_in1k](https://huggingface.co/timm/maxvit_large_tf_384.in21k_ft_in1k) |87.98|98.56| 71.75| 212.03|132.55| 445.84|
|[maxvit_large_tf_384.in1k](https://huggingface.co/timm/maxvit_large_tf_384.in1k) |86.23|97.69| 70.56| 212.03|132.55| 445.84|
|[maxvit_base_tf_512.in21k_ft_in1k](https://huggingface.co/timm/maxvit_base_tf_512.in21k_ft_in1k) |88.20|98.53| 50.87| 119.88|138.02| 703.99|
|[maxvit_base_tf_512.in1k](https://huggingface.co/timm/maxvit_base_tf_512.in1k) |86.60|97.92| 50.75| 119.88|138.02| 703.99|
|[maxvit_xlarge_tf_384.in21k_ft_in1k](https://huggingface.co/timm/maxvit_xlarge_tf_384.in21k_ft_in1k) |88.32|98.54| 42.53| 475.32|292.78| 668.76|
|[maxvit_large_tf_512.in21k_ft_in1k](https://huggingface.co/timm/maxvit_large_tf_512.in21k_ft_in1k) |88.04|98.40| 36.42| 212.33|244.75| 942.15|
|[maxvit_large_tf_512.in1k](https://huggingface.co/timm/maxvit_large_tf_512.in1k) |86.52|97.88| 36.04| 212.33|244.75| 942.15|
|[maxvit_xlarge_tf_512.in21k_ft_in1k](https://huggingface.co/timm/maxvit_xlarge_tf_512.in21k_ft_in1k) |88.53|98.64| 21.76| 475.77|534.14|1413.22|
## Citation
```bibtex
@misc{rw2019timm,
author = {Ross Wightman},
title = {PyTorch Image Models},
year = {2019},
publisher = {GitHub},
journal = {GitHub repository},
doi = {10.5281/zenodo.4414861},
howpublished = {\url{https://github.com/huggingface/pytorch-image-models}}
}
```
```bibtex
@article{tu2022maxvit,
title={MaxViT: Multi-Axis Vision Transformer},
author={Tu, Zhengzhong and Talebi, Hossein and Zhang, Han and Yang, Feng and Milanfar, Peyman and Bovik, Alan and Li, Yinxiao},
journal={ECCV},
year={2022},
}
```
```bibtex
@article{dai2021coatnet,
title={CoAtNet: Marrying Convolution and Attention for All Data Sizes},
author={Dai, Zihang and Liu, Hanxiao and Le, Quoc V and Tan, Mingxing},
journal={arXiv preprint arXiv:2106.04803},
year={2021}
}
```
|
timm/mobilenetv3_small_050.lamb_in1k | timm | "2023-04-27T22:49:29Z" | 8,874 | 0 | timm | [
"timm",
"pytorch",
"safetensors",
"image-classification",
"dataset:imagenet-1k",
"arxiv:2110.00476",
"arxiv:1905.02244",
"license:apache-2.0",
"region:us"
] | image-classification | "2022-12-16T05:38:23Z" | ---
tags:
- image-classification
- timm
library_name: timm
license: apache-2.0
datasets:
- imagenet-1k
---
# Model card for mobilenetv3_small_050.lamb_in1k
A MobileNet-v3 image classification model. Trained on ImageNet-1k in `timm` using recipe template described below.
Recipe details:
* A LAMB optimizer recipe that is similar to [ResNet Strikes Back](https://arxiv.org/abs/2110.00476) `A2` but 50% longer with EMA weight averaging, no CutMix
* RMSProp (TF 1.0 behaviour) optimizer, EMA weight averaging
* Step (exponential decay w/ staircase) LR schedule with warmup
## Model Details
- **Model Type:** Image classification / feature backbone
- **Model Stats:**
- Params (M): 1.6
- GMACs: 0.0
- Activations (M): 0.9
- Image size: 224 x 224
- **Papers:**
- Searching for MobileNetV3: https://arxiv.org/abs/1905.02244
- **Dataset:** ImageNet-1k
- **Original:** https://github.com/huggingface/pytorch-image-models
## Model Usage
### Image Classification
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model('mobilenetv3_small_050.lamb_in1k', pretrained=True)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1
top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5)
```
### Feature Map Extraction
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model(
'mobilenetv3_small_050.lamb_in1k',
pretrained=True,
features_only=True,
)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1
for o in output:
# print shape of each feature map in output
# e.g.:
# torch.Size([1, 16, 112, 112])
# torch.Size([1, 8, 56, 56])
# torch.Size([1, 16, 28, 28])
# torch.Size([1, 24, 14, 14])
# torch.Size([1, 288, 7, 7])
print(o.shape)
```
### Image Embeddings
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model(
'mobilenetv3_small_050.lamb_in1k',
pretrained=True,
num_classes=0, # remove classifier nn.Linear
)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor
# or equivalently (without needing to set num_classes=0)
output = model.forward_features(transforms(img).unsqueeze(0))
# output is unpooled, a (1, 288, 7, 7) shaped tensor
output = model.forward_head(output, pre_logits=True)
# output is a (1, num_features) shaped tensor
```
## Model Comparison
Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results).
## Citation
```bibtex
@misc{rw2019timm,
author = {Ross Wightman},
title = {PyTorch Image Models},
year = {2019},
publisher = {GitHub},
journal = {GitHub repository},
doi = {10.5281/zenodo.4414861},
howpublished = {\url{https://github.com/huggingface/pytorch-image-models}}
}
```
```bibtex
@inproceedings{howard2019searching,
title={Searching for mobilenetv3},
author={Howard, Andrew and Sandler, Mark and Chu, Grace and Chen, Liang-Chieh and Chen, Bo and Tan, Mingxing and Wang, Weijun and Zhu, Yukun and Pang, Ruoming and Vasudevan, Vijay and others},
booktitle={Proceedings of the IEEE/CVF international conference on computer vision},
pages={1314--1324},
year={2019}
}
```
|
sociocom/MedNERN-CR-JA | sociocom | "2024-02-26T04:53:06Z" | 8,871 | 2 | transformers | [
"transformers",
"pytorch",
"safetensors",
"bert",
"token-classification",
"NER",
"medical documents",
"ja",
"dataset:MedTxt-CR-JA-training-v2.xml",
"doi:10.57967/hf/0620",
"license:cc-by-4.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | token-classification | "2023-04-13T08:25:56Z" | ---
language:
- ja
license:
- cc-by-4.0
tags:
- NER
- medical documents
datasets:
- MedTxt-CR-JA-training-v2.xml
metrics:
- NTCIR-16 Real-MedNLP subtask 1
---
This is a model for named entity recognition of Japanese medical documents.
# Introduction
This repository contains the base model and a support predict script for using the model and providing a XML tagged text output.
The original model was trained on the [MedTxt-CR-JA](https://sociocom.naist.jp/medtxt/cr) dataset, so the provided prediction code outputs XML tags in the same format.
The script also provide the normalization method for the output entities, which is not embedded in the model.
If you want to re-train or update the model, we provide additional support scripts in [this GitHub repository](https://github.com/sociocom/MedNERN-CR-JA).
Issues and suggestions can also be submitted there.
### A note about loading the model using standard HuggingFace methods
This model should also be loadable using standard HuggingFace `from_pretrained` methods. However, the model by itself only outputs labels in the format "LABEL_0", "LABEL1", etc.
The conversion of model outputs to the actual labels ("<m-key>, "<m-val>", "<timex-3>" etc.) is not yet embedded into the model, so the extra `id_to_tags.pkl` file is necessary
to make the conversion. It contains a mapping from the model output ids to the actual labels.
Such process can be done manually if needed, but the `predict.py` script already does that.
We are currently working to better standardize the model to HuggingFace's standards.
## How to use
Clone the repository and install the requirements:
```
pip install -r requirements.txt
```
The code has been developed tested with Python 3.9 in MacOS 14.1 (M1 MacBook Pro).
### Prediction
The prediction script will output the results in the same XML format as the input file. It can be run with the following
command:
```
python3 predict.py
```
The default parameters will take the model located in `pytorch_model.bin` and the input file `text.txt`.
The resulting predictions will be output to the screen.
To select a different model or input file, use the `-m` and `-i` parameters, respectively:
```
python3 predict.py -m <model_path> -i <your_input_file>.txt
```
The input file can be a single text file or a folder containing multiple `.txt` files, for batch processing. For example:
```
python3 predict.py -m <model_path> -i <your_input_folder>
```
### Entity normalization
This model supports entity normalization via dictionary matching. The dictionary is a list of medical terms or
drugs and their standard forms.
Two different dictionaries are used for drug and disease normalization, stored in the `dictionaries` folder as
`drug_dict.csv` and `disease_dict.csv`, respectively.
To enable normalization you can add the `--normalize` flag to the `predict.py` command.
```
python3 predict.py -m <model_path> --normalize
```
Normalization will add the `norm` attribute to the output XML tags. This attribute can be empty if a normalized form of
the term is not found.
The provided disease normalization dictionary (`dictionaties/disease_dict.csv`) is based on
the [Manbyo Dictionary](https://sociocom.naist.jp/manbyo-dic-en/) and provides normalization to the standard ICD code
for the diseases.
The default drug dictionary (`dictionaties/drug_dict.csv`) is based on
the [Hyakuyaku Dictionary](https://sociocom.naist.jp/hyakuyaku-dic-en/).
The dictionary is a CSV file with three columns: the first column is the surface form term and the third column contain
its standard form. The second column is not used.
### Replacing the default dictionaries
User can freely change the dictionary to fit their needs by passing the path to a custom dictionary file.
The dictionary file must have at least a column containing the list of surface forms and a column containing the list of
normalized forms.
The parameters `--drug_dict` and `--disease_dict` can be used to specify the path to the drug and disease dictionaries,
respectively.
When doing so, the respective parameters informing the column index of the surface form and normalized form must also be
provided.
You don't need to replace both dictionaries at the same time, you can replace only one of them.
E.g.:
```
python3 predict.py --normalize --drug_dict dictionaries/drug_dict.csv --drug_surface_form 0 --drug_norm_form 2 --disease_dict dictionaries/disease_dict.csv --disease_surface_form 0 --disease_norm_form 2
```
### Input Example
```
肥大型心筋症、心房細動に対してＷＦ投与が開始となった。
治療経過中に非持続性心室頻拍が認められたためアミオダロンが併用となった。
```
### Output Example
```
<d certainty="positive" norm="I422">肥大型心筋症、心房細動</d>に対して<m-key state="executed" norm="ワルファリンカリウム">ＷＦ</m-key>投与が開始となった。
<timex3 type="med">治療経過中</timex3>に<d certainty="positive" norm="I472">非持続性心室頻拍</d>が認められたため<m-key state="executed" norm="アミオダロン塩酸塩">アミオダロン</m-key>が併用となった。
```
## Publication
This model can be cited as:
```
@misc {social_computing_lab_2023,
author = { {Social Computing Lab} },
title = { MedNERN-CR-JA (Revision 13dbcb6) },
year = 2023,
url = { https://huggingface.co/sociocom/MedNERN-CR-JA },
doi = { 10.57967/hf/0620 },
publisher = { Hugging Face }
}
```
|
Yntec/Thriller | Yntec | "2023-10-07T03:02:34Z" | 8,841 | 3 | diffusers | [
"diffusers",
"safetensors",
"General",
"Photo",
"Movie",
"LandScapes",
"MagicArt35",
"Lykon",
"stable-diffusion",
"stable-diffusion-diffusers",
"text-to-image",
"license:creativeml-openrail-m",
"autotrain_compatible",
"endpoints_compatible",
"diffusers:StableDiffusionPipeline",
"region:us"
] | text-to-image | "2023-10-07T01:36:05Z" | ---
license: creativeml-openrail-m
library_name: diffusers
pipeline_tag: text-to-image
tags:
- General
- Photo
- Movie
- LandScapes
- MagicArt35
- Lykon
- stable-diffusion
- stable-diffusion-diffusers
- diffusers
- text-to-image
---
# Thriller
A mix of PhotoMovieXFinal and AbsoluteRemix.
(Click for larger)
DETAILED CHIBI EYES, cartoon pretty CUTE girl in white shirt, fashion shoes, costume, white skirt, 1940, magazine ad, iconic. A painting of a store with a lot of food, a photorealistic painting by simon stålenhag, featured on cgsociety, photorealism, 2d game art, hyper-realistic, hyper realism
Original pages:
https://civitai.com/models/94687?modelVersionId=101000 (photoMovieX)
https://civitai.com/models/81458?modelVersionId=132760 (AbsoluteReality)
https://huggingface.co/Yntec/AbsoluteRemix
# Recipe
SuperMerger Weight sum MBW 1,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,1
Model A: AbsoluteRemix
Model B: PhotoMovieXFinal
Output: Thriller |
Intel/t5-small-xsum-int8-dynamic-inc | Intel | "2024-03-21T12:59:37Z" | 8,827 | 1 | transformers | [
"transformers",
"pytorch",
"t5",
"text2text-generation",
"int8",
"Intel® Neural Compressor",
"neural-compressor",
"PostTrainingDynamic",
"dataset:mnli",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text2text-generation | "2022-09-22T12:30:01Z" | ---
license: apache-2.0
tags:
- int8
- Intel® Neural Compressor
- neural-compressor
- PostTrainingDynamic
datasets:
- mnli
metrics:
- accuracy
---
# INT8 T5 small finetuned on XSum
### Post-training dynamic quantization
This is an INT8 PyTorch model quantized with [huggingface/optimum-intel](https://github.com/huggingface/optimum-intel) through the usage of [Intel® Neural Compressor](https://github.com/intel/neural-compressor).
The original fp32 model comes from the fine-tuned model [adasnew/t5-small-xsum](https://huggingface.co/adasnew/t5-small-xsum).
The linear modules **lm.head**, fall back to fp32 for less than 1% relative accuracy loss.
### Evaluation result
| |INT8|FP32|
|---|:---:|:---:|
| **Accuracy (eval-rouge1)** | 29.9008 |29.9592|
| **Model size** |154M|242M|
### Load with optimum:
```python
from optimum.intel import INCModelForSeq2SeqLM
model_id = "Intel/t5-small-xsum-int8-dynamic-inc"
int8_model = INCModelForSeq2SeqLM.from_pretrained(model_id)
```
|
fxmarty/tiny-marian | fxmarty | "2023-03-06T14:29:46Z" | 8,825 | 0 | transformers | [
"transformers",
"pytorch",
"marian",
"text2text-generation",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text2text-generation | "2023-03-06T14:29:19Z" | ---
license: mit
---
|
seara/rubert-tiny2-ru-go-emotions | seara | "2023-08-25T19:25:01Z" | 8,824 | 5 | transformers | [
"transformers",
"pytorch",
"safetensors",
"bert",
"text-classification",
"sentiment-analysis",
"multi-label-classification",
"sentiment analysis",
"rubert",
"sentiment",
"tiny",
"russian",
"multilabel",
"classification",
"emotion-classification",
"emotion-recognition",
"emotion",
"ru",
"dataset:seara/ru_go_emotions",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | "2023-05-02T16:52:52Z" | ---
license: mit
language:
- ru
metrics:
- f1
- roc_auc
- precision
- recall
pipeline_tag: text-classification
tags:
- sentiment-analysis
- multi-label-classification
- sentiment analysis
- rubert
- sentiment
- bert
- tiny
- russian
- multilabel
- classification
- emotion-classification
- emotion-recognition
- emotion
datasets:
- seara/ru_go_emotions
---
This is [RuBERT-tiny2](https://huggingface.co/cointegrated/rubert-tiny2) model fine-tuned for __emotion classification__ of short __Russian__ texts.
The task is a __multi-label classification__ with the following labels:
```yaml
0: admiration
1: amusement
2: anger
3: annoyance
4: approval
5: caring
6: confusion
7: curiosity
8: desire
9: disappointment
10: disapproval
11: disgust
12: embarrassment
13: excitement
14: fear
15: gratitude
16: grief
17: joy
18: love
19: nervousness
20: optimism
21: pride
22: realization
23: relief
24: remorse
25: sadness
26: surprise
27: neutral
```
Label to Russian label:
```yaml
admiration: восхищение
amusement: веселье
anger: злость
annoyance: раздражение
approval: одобрение
caring: забота
confusion: непонимание
curiosity: любопытство
desire: желание
disappointment: разочарование
disapproval: неодобрение
disgust: отвращение
embarrassment: смущение
excitement: возбуждение
fear: страх
gratitude: признательность
grief: горе
joy: радость
love: любовь
nervousness: нервозность
optimism: оптимизм
pride: гордость
realization: осознание
relief: облегчение
remorse: раскаяние
sadness: грусть
surprise: удивление
neutral: нейтральность
```
## Usage
```python
from transformers import pipeline
model = pipeline(model="seara/rubert-tiny2-ru-go-emotions")
model("Привет, ты мне нравишься!")
# [{'label': 'love', 'score': 0.5955629944801331}]
```
## Dataset
This model was trained on translated GoEmotions dataset called [ru_go_emotions](https://huggingface.co/datasets/seara/ru_go_emotions).
An overview of the training data can be found on [Hugging Face card](https://huggingface.co/datasets/seara/ru_go_emotions) and on
[Github repository](https://github.com/searayeah/ru-goemotions).
## Training
Training were done in this [project](https://github.com/searayeah/bert-russian-sentiment-emotion) with this parameters:
```yaml
tokenizer.max_length: null
batch_size: 64
optimizer: adam
lr: 0.00001
weight_decay: 0
num_epochs: 31
```
## Eval results (on test split)
| |precision|recall|f1-score|auc-roc|support|
|--------------|---------|------|--------|-------|-------|
|admiration |0.68 |0.61 |0.64 |0.92 |504 |
|amusement |0.8 |0.84 |0.82 |0.96 |264 |
|anger |0.55 |0.33 |0.42 |0.9 |198 |
|annoyance |0.56 |0.03 |0.06 |0.81 |320 |
|approval |0.6 |0.18 |0.28 |0.78 |351 |
|caring |0.5 |0.04 |0.07 |0.84 |135 |
|confusion |0.77 |0.07 |0.12 |0.9 |153 |
|curiosity |0.51 |0.34 |0.41 |0.92 |284 |
|desire |0.71 |0.18 |0.29 |0.88 |83 |
|disappointment|0.0 |0.0 |0.0 |0.76 |151 |
|disapproval |0.48 |0.1 |0.17 |0.85 |267 |
|disgust |0.94 |0.12 |0.22 |0.9 |123 |
|embarrassment |0.0 |0.0 |0.0 |0.84 |37 |
|excitement |0.81 |0.2 |0.33 |0.88 |103 |
|fear |0.73 |0.42 |0.54 |0.92 |78 |
|gratitude |0.95 |0.89 |0.92 |0.99 |352 |
|grief |0.0 |0.0 |0.0 |0.76 |6 |
|joy |0.66 |0.52 |0.58 |0.93 |161 |
|love |0.8 |0.79 |0.79 |0.97 |238 |
|nervousness |0.0 |0.0 |0.0 |0.81 |23 |
|optimism |0.67 |0.41 |0.51 |0.89 |186 |
|pride |0.0 |0.0 |0.0 |0.89 |16 |
|realization |0.0 |0.0 |0.0 |0.7 |145 |
|relief |0.0 |0.0 |0.0 |0.84 |11 |
|remorse |0.59 |0.71 |0.65 |0.99 |56 |
|sadness |0.77 |0.37 |0.5 |0.89 |156 |
|surprise |0.59 |0.35 |0.44 |0.88 |141 |
|neutral |0.64 |0.58 |0.61 |0.81 |1787 |
|micro avg |0.68 |0.43 |0.53 |0.93 |6329 |
|macro avg |0.51 |0.29 |0.33 |0.87 |6329 |
|weighted avg |0.62 |0.43 |0.48 |0.86 |6329 |
|
TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF | TheBloke | "2023-09-27T12:52:36Z" | 8,818 | 27 | transformers | [
"transformers",
"gguf",
"llama",
"uncensored",
"en",
"dataset:ehartford/wizard_vicuna_70k_unfiltered",
"base_model:ehartford/Wizard-Vicuna-7B-Uncensored",
"license:other",
"text-generation-inference",
"region:us"
] | null | "2023-09-19T23:07:40Z" | ---
language:
- en
license: other
tags:
- uncensored
datasets:
- ehartford/wizard_vicuna_70k_unfiltered
model_name: Wizard Vicuna 7B Uncensored
base_model: ehartford/Wizard-Vicuna-7B-Uncensored
inference: false
model_creator: Eric Hartford
model_type: llama
prompt_template: 'A chat between a curious user and an artificial intelligence assistant.
The assistant gives helpful, detailed, and polite answers to the user''s questions.
USER: {prompt} ASSISTANT:
'
quantized_by: TheBloke
---
<!-- header start -->
<!-- 200823 -->
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<div style="display: flex; justify-content: space-between; width: 100%;">
<div style="display: flex; flex-direction: column; align-items: flex-start;">
<p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p>
</div>
<div style="display: flex; flex-direction: column; align-items: flex-end;">
<p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p>
</div>
</div>
<div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div>
<hr style="margin-top: 1.0em; margin-bottom: 1.0em;">
<!-- header end -->
# Wizard Vicuna 7B Uncensored - GGUF
- Model creator: [Eric Hartford](https://huggingface.co/ehartford)
- Original model: [Wizard Vicuna 7B Uncensored](https://huggingface.co/ehartford/Wizard-Vicuna-7B-Uncensored)
<!-- description start -->
## Description
This repo contains GGUF format model files for [Wizard-Vicuna-7B-Uncensored](https://huggingface.co/ehartford/Wizard-Vicuna-7B-Uncensored).
<!-- description end -->
<!-- README_GGUF.md-about-gguf start -->
### About GGUF
GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp.
Here is an incomplate list of clients and libraries that are known to support GGUF:
* [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option.
* [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI, with many features and powerful extensions. Supports GPU acceleration.
* [KoboldCpp](https://github.com/LostRuins/koboldcpp), a fully featured web UI, with GPU accel across all platforms and GPU architectures. Especially good for story telling.
* [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration.
* [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), a great web UI with many interesting and unique features, including a full model library for easy model selection.
* [Faraday.dev](https://faraday.dev/), an attractive and easy to use character-based chat GUI for Windows and macOS (both Silicon and Intel), with GPU acceleration.
* [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server.
* [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server.
* [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use.
<!-- README_GGUF.md-about-gguf end -->
<!-- repositories-available start -->
## Repositories available
* [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-AWQ)
* [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GPTQ)
* [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF)
* [Eric Hartford's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/ehartford/Wizard-Vicuna-7B-Uncensored)
<!-- repositories-available end -->
<!-- prompt-template start -->
## Prompt template: Vicuna
```
A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. USER: {prompt} ASSISTANT:
```
<!-- prompt-template end -->
<!-- compatibility_gguf start -->
## Compatibility
These quantised GGUFv2 files are compatible with llama.cpp from August 27th onwards, as of commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221)
They are also compatible with many third party UIs and libraries - please see the list at the top of this README.
## Explanation of quantisation methods
<details>
<summary>Click to see details</summary>
The new methods available are:
* GGML_TYPE_Q2_K - "type-1" 2-bit quantization in super-blocks containing 16 blocks, each block having 16 weight. Block scales and mins are quantized with 4 bits. This ends up effectively using 2.5625 bits per weight (bpw)
* GGML_TYPE_Q3_K - "type-0" 3-bit quantization in super-blocks containing 16 blocks, each block having 16 weights. Scales are quantized with 6 bits. This end up using 3.4375 bpw.
* GGML_TYPE_Q4_K - "type-1" 4-bit quantization in super-blocks containing 8 blocks, each block having 32 weights. Scales and mins are quantized with 6 bits. This ends up using 4.5 bpw.
* GGML_TYPE_Q5_K - "type-1" 5-bit quantization. Same super-block structure as GGML_TYPE_Q4_K resulting in 5.5 bpw
* GGML_TYPE_Q6_K - "type-0" 6-bit quantization. Super-blocks with 16 blocks, each block having 16 weights. Scales are quantized with 8 bits. This ends up using 6.5625 bpw
Refer to the Provided Files table below to see what files use which methods, and how.
</details>
<!-- compatibility_gguf end -->
<!-- README_GGUF.md-provided-files start -->
## Provided files
| Name | Quant method | Bits | Size | Max RAM required | Use case |
| ---- | ---- | ---- | ---- | ---- | ----- |
| [Wizard-Vicuna-7B-Uncensored.Q2_K.gguf](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF/blob/main/Wizard-Vicuna-7B-Uncensored.Q2_K.gguf) | Q2_K | 2 | 2.83 GB| 5.33 GB | smallest, significant quality loss - not recommended for most purposes |
| [Wizard-Vicuna-7B-Uncensored.Q3_K_S.gguf](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF/blob/main/Wizard-Vicuna-7B-Uncensored.Q3_K_S.gguf) | Q3_K_S | 3 | 2.95 GB| 5.45 GB | very small, high quality loss |
| [Wizard-Vicuna-7B-Uncensored.Q3_K_M.gguf](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF/blob/main/Wizard-Vicuna-7B-Uncensored.Q3_K_M.gguf) | Q3_K_M | 3 | 3.30 GB| 5.80 GB | very small, high quality loss |
| [Wizard-Vicuna-7B-Uncensored.Q3_K_L.gguf](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF/blob/main/Wizard-Vicuna-7B-Uncensored.Q3_K_L.gguf) | Q3_K_L | 3 | 3.60 GB| 6.10 GB | small, substantial quality loss |
| [Wizard-Vicuna-7B-Uncensored.Q4_0.gguf](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF/blob/main/Wizard-Vicuna-7B-Uncensored.Q4_0.gguf) | Q4_0 | 4 | 3.83 GB| 6.33 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| [Wizard-Vicuna-7B-Uncensored.Q4_K_S.gguf](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF/blob/main/Wizard-Vicuna-7B-Uncensored.Q4_K_S.gguf) | Q4_K_S | 4 | 3.86 GB| 6.36 GB | small, greater quality loss |
| [Wizard-Vicuna-7B-Uncensored.Q4_K_M.gguf](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF/blob/main/Wizard-Vicuna-7B-Uncensored.Q4_K_M.gguf) | Q4_K_M | 4 | 4.08 GB| 6.58 GB | medium, balanced quality - recommended |
| [Wizard-Vicuna-7B-Uncensored.Q5_0.gguf](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF/blob/main/Wizard-Vicuna-7B-Uncensored.Q5_0.gguf) | Q5_0 | 5 | 4.65 GB| 7.15 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| [Wizard-Vicuna-7B-Uncensored.Q5_K_S.gguf](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF/blob/main/Wizard-Vicuna-7B-Uncensored.Q5_K_S.gguf) | Q5_K_S | 5 | 4.65 GB| 7.15 GB | large, low quality loss - recommended |
| [Wizard-Vicuna-7B-Uncensored.Q5_K_M.gguf](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF/blob/main/Wizard-Vicuna-7B-Uncensored.Q5_K_M.gguf) | Q5_K_M | 5 | 4.78 GB| 7.28 GB | large, very low quality loss - recommended |
| [Wizard-Vicuna-7B-Uncensored.Q6_K.gguf](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF/blob/main/Wizard-Vicuna-7B-Uncensored.Q6_K.gguf) | Q6_K | 6 | 5.53 GB| 8.03 GB | very large, extremely low quality loss |
| [Wizard-Vicuna-7B-Uncensored.Q8_0.gguf](https://huggingface.co/TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF/blob/main/Wizard-Vicuna-7B-Uncensored.Q8_0.gguf) | Q8_0 | 8 | 7.16 GB| 9.66 GB | very large, extremely low quality loss - not recommended |
**Note**: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.
<!-- README_GGUF.md-provided-files end -->
<!-- README_GGUF.md-how-to-download start -->
## How to download GGUF files
**Note for manual downloaders:** You almost never want to clone the entire repo! Multiple different quantisation formats are provided, and most users only want to pick and download a single file.
The following clients/libraries will automatically download models for you, providing a list of available models to choose from:
- LM Studio
- LoLLMS Web UI
- Faraday.dev
### In `text-generation-webui`
Under Download Model, you can enter the model repo: TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF and below it, a specific filename to download, such as: Wizard-Vicuna-7B-Uncensored.Q4_K_M.gguf.
Then click Download.
### On the command line, including multiple files at once
I recommend using the `huggingface-hub` Python library:
```shell
pip3 install huggingface-hub
```
Then you can download any individual model file to the current directory, at high speed, with a command like this:
```shell
huggingface-cli download TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF Wizard-Vicuna-7B-Uncensored.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False
```
<details>
<summary>More advanced huggingface-cli download usage</summary>
You can also download multiple files at once with a pattern:
```shell
huggingface-cli download TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF --local-dir . --local-dir-use-symlinks False --include='*Q4_K*gguf'
```
For more documentation on downloading with `huggingface-cli`, please see: [HF -> Hub Python Library -> Download files -> Download from the CLI](https://huggingface.co/docs/huggingface_hub/guides/download#download-from-the-cli).
To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`:
```shell
pip3 install hf_transfer
```
And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`:
```shell
HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF Wizard-Vicuna-7B-Uncensored.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False
```
Windows Command Line users: You can set the environment variable by running `set HF_HUB_ENABLE_HF_TRANSFER=1` before the download command.
</details>
<!-- README_GGUF.md-how-to-download end -->
<!-- README_GGUF.md-how-to-run start -->
## Example `llama.cpp` command
Make sure you are using `llama.cpp` from commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) or later.
```shell
./main -ngl 32 -m Wizard-Vicuna-7B-Uncensored.Q4_K_M.gguf --color -c 2048 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. USER: {prompt} ASSISTANT:"
```
Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
Change `-c 2048` to the desired sequence length. For extended sequence models - eg 8K, 16K, 32K - the necessary RoPE scaling parameters are read from the GGUF file and set by llama.cpp automatically.
If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins`
For other parameters and how to use them, please refer to [the llama.cpp documentation](https://github.com/ggerganov/llama.cpp/blob/master/examples/main/README.md)
## How to run in `text-generation-webui`
Further instructions here: [text-generation-webui/docs/llama.cpp.md](https://github.com/oobabooga/text-generation-webui/blob/main/docs/llama.cpp.md).
## How to run from Python code
You can use GGUF models from Python using the [llama-cpp-python](https://github.com/abetlen/llama-cpp-python) or [ctransformers](https://github.com/marella/ctransformers) libraries.
### How to load this model in Python code, using ctransformers
#### First install the package
Run one of the following commands, according to your system:
```shell
# Base ctransformers with no GPU acceleration
pip install ctransformers
# Or with CUDA GPU acceleration
pip install ctransformers[cuda]
# Or with AMD ROCm GPU acceleration (Linux only)
CT_HIPBLAS=1 pip install ctransformers --no-binary ctransformers
# Or with Metal GPU acceleration for macOS systems only
CT_METAL=1 pip install ctransformers --no-binary ctransformers
```
#### Simple ctransformers example code
```python
from ctransformers import AutoModelForCausalLM
# Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system.
llm = AutoModelForCausalLM.from_pretrained("TheBloke/Wizard-Vicuna-7B-Uncensored-GGUF", model_file="Wizard-Vicuna-7B-Uncensored.Q4_K_M.gguf", model_type="llama", gpu_layers=50)
print(llm("AI is going to"))
```
## How to use with LangChain
Here are guides on using llama-cpp-python and ctransformers with LangChain:
* [LangChain + llama-cpp-python](https://python.langchain.com/docs/integrations/llms/llamacpp)
* [LangChain + ctransformers](https://python.langchain.com/docs/integrations/providers/ctransformers)
<!-- README_GGUF.md-how-to-run end -->
<!-- footer start -->
<!-- 200823 -->
## Discord
For further support, and discussions on these models and AI in general, join us at:
[TheBloke AI's Discord server](https://discord.gg/theblokeai)
## Thanks, and how to contribute
Thanks to the [chirper.ai](https://chirper.ai) team!
Thanks to Clay from [gpus.llm-utils.org](llm-utils)!
I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training.
If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects.
Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits.
* Patreon: https://patreon.com/TheBlokeAI
* Ko-Fi: https://ko-fi.com/TheBlokeAI
**Special thanks to**: Aemon Algiz.
**Patreon special mentions**: Alicia Loh, Stephen Murray, K, Ajan Kanaga, RoA, Magnesian, Deo Leter, Olakabola, Eugene Pentland, zynix, Deep Realms, Raymond Fosdick, Elijah Stavena, Iucharbius, Erik Bjäreholt, Luis Javier Navarrete Lozano, Nicholas, theTransient, John Detwiler, alfie_i, knownsqashed, Mano Prime, Willem Michiel, Enrico Ros, LangChain4j, OG, Michael Dempsey, Pierre Kircher, Pedro Madruga, James Bentley, Thomas Belote, Luke @flexchar, Leonard Tan, Johann-Peter Hartmann, Illia Dulskyi, Fen Risland, Chadd, S_X, Jeff Scroggin, Ken Nordquist, Sean Connelly, Artur Olbinski, Swaroop Kallakuri, Jack West, Ai Maven, David Ziegler, Russ Johnson, transmissions 11, John Villwock, Alps Aficionado, Clay Pascal, Viktor Bowallius, Subspace Studios, Rainer Wilmers, Trenton Dambrowitz, vamX, Michael Levine, 준교 김, Brandon Frisco, Kalila, Trailburnt, Randy H, Talal Aujan, Nathan Dryer, Vadim, 阿明, ReadyPlayerEmma, Tiffany J. Kim, George Stoitzev, Spencer Kim, Jerry Meng, Gabriel Tamborski, Cory Kujawski, Jeffrey Morgan, Spiking Neurons AB, Edmond Seymore, Alexandros Triantafyllidis, Lone Striker, Cap'n Zoog, Nikolai Manek, danny, ya boyyy, Derek Yates, usrbinkat, Mandus, TL, Nathan LeClaire, subjectnull, Imad Khwaja, webtim, Raven Klaugh, Asp the Wyvern, Gabriel Puliatti, Caitlyn Gatomon, Joseph William Delisle, Jonathan Leane, Luke Pendergrass, SuperWojo, Sebastain Graf, Will Dee, Fred von Graf, Andrey, Dan Guido, Daniel P. Andersen, Nitin Borwankar, Elle, Vitor Caleffi, biorpg, jjj, NimbleBox.ai, Pieter, Matthew Berman, terasurfer, Michael Davis, Alex, Stanislav Ovsiannikov
Thank you to all my generous patrons and donaters!
And thank you again to a16z for their generous grant.
<!-- footer end -->
<!-- original-model-card start -->
# Original model card: Wizard-Vicuna-7B-Uncensored
This is [wizard-vicuna-13b](https://huggingface.co/junelee/wizard-vicuna-13b) trained against LLaMA-7B with a subset of the dataset - responses that contained alignment / moralizing were removed. The intent is to train a WizardLM that doesn't have alignment built-in, so that alignment (of any sort) can be added separately with for example with a RLHF LoRA.
Shout out to the open source AI/ML community, and everyone who helped me out.
Note:
An uncensored model has no guardrails.
You are responsible for anything you do with the model, just as you are responsible for anything you do with any dangerous object such as a knife, gun, lighter, or car.
Publishing anything this model generates is the same as publishing it yourself.
You are responsible for the content you publish, and you cannot blame the model any more than you can blame the knife, gun, lighter, or car for what you do with it.
<!-- original-model-card end -->
|
DeepMount00/Mistral-RAG | DeepMount00 | "2024-04-21T13:43:15Z" | 8,815 | 11 | transformers | [
"transformers",
"safetensors",
"mistral",
"text-generation",
"it",
"dataset:DeepMount00/gquad_it",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2024-04-20T13:38:41Z" | ---
library_name: transformers
license: apache-2.0
datasets:
- DeepMount00/gquad_it
language:
- it
---
## Mistral-RAG
- **Model Name:** Mistral-RAG
- **Base Model:** [Mistral-Ita-7b](https://huggingface.co/DeepMount00/Mistral-Ita-7b)
- **Specialization:** Question and Answer Tasks
### Overview
Mistral-RAG is a refined fine-tuning of the Mistral-Ita-7b model, engineered specifically to enhance question and answer tasks. It features a unique dual-response capability, offering both generative and extractive modes to cater to a wide range of informational needs.
### Capabilities
#### Generative Mode
- **Description:** The generative mode is designed for scenarios that require complex, synthesized responses. This mode integrates information from multiple sources and provides expanded explanations.
- **Ideal Use Cases:**
- Educational purposes
- Advisory services
- Creative scenarios where depth and detailed understanding are crucial
#### Extractive Mode
- **Description:** The extractive mode focuses on speed and precision. It delivers direct and concise answers by extracting specific data from texts.
- **Ideal Use Cases:**
- Factual queries in research
- Legal contexts
- Professional environments where accuracy and direct evidence are necessary
### How to Use
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
MODEL_NAME = "DeepMount00/Mistral-RAG"
model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, torch_dtype=torch.bfloat16).eval()
model.to(device)
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
def generate_answer(prompt, response_type="generativo"):
# Creazione del contesto e della domanda in base al tipo di risposta
if response_type == "estrattivo":
prompt = f"Rispondi alla seguente domanda in modo estrattivo, basandoti esclusivamente sul contesto.\n{prompt}"
else:
prompt = f"Rispondi alla seguente domanda in modo generativo, basandoti esclusivamente sul contesto.\n{prompt}"
# Preparazione del messaggio per il modello
messages = [
{"role": "user", "content": prompt},
]
model_inputs = tokenizer.apply_chat_template(messages, return_tensors="pt").to(device)
generated_ids = model.generate(model_inputs, max_new_tokens=200, do_sample=True,
temperature=0.001, eos_token_id=tokenizer.eos_token_id)
decoded = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)
return decoded[0].split("[/INST]", 1)[1].strip() if "[/INST]" in decoded[0] else "Errore nella generazione della risposta"
# Esempio di utilizzo con la nuova funzionalità
contesto = """Venerdì più di 2.100 persone che vivono vicino a un vulcano in Indonesia sono state sfollate per i rischi legati a un’eruzione. Martedì infatti l’isola vulcanica di Ruang, che si trova circa 100 chilometri a nord di Sulawesi, ha cominciato a eruttare, producendo una colonna di fumo e ceneri che ieri ha raggiunto 1.200 metri di altezza. Le operazioni di evacuazione sono ancora in corso: complessivamente sono più di 11mila le persone a cui è stato detto di lasciare le proprie case. Gran parte di loro vive sulla vicina isola di Tagulandang, che in totale ha 20mila abitanti; potrebbe essere raggiunta non solo dalle ceneri vulcaniche e dai piroclasti, ma anche da un eventuale tsunami causato dalla caduta in mare di lava e rocce."""
domanda = "Perchè le persone sono evacuate dalle case?"
prompt = f"Contesto: {contesto}\nDomanda: {domanda}"
answer = generate_answer(prompt, "estrattivo")
print(answer)
```
---
## Developer
[Michele Montebovi] |
dkleczek/bert-base-polish-cased-v1 | dkleczek | "2021-05-19T15:54:20Z" | 8,813 | 7 | transformers | [
"transformers",
"pytorch",
"jax",
"bert",
"pretraining",
"pl",
"endpoints_compatible",
"region:us"
] | null | "2022-03-02T23:29:05Z" | ---
language: pl
thumbnail: https://raw.githubusercontent.com/kldarek/polbert/master/img/polbert.png
---
# Polbert - Polish BERT
Polish version of BERT language model is here! It is now available in two variants: cased and uncased, both can be downloaded and used via HuggingFace transformers library. I recommend using the cased model, more info on the differences and benchmark results below.
## Cased and uncased variants
* I initially trained the uncased model, the corpus and training details are referenced below. Here are some issues I found after I published the uncased model:
* Some Polish characters and accents are not tokenized correctly through the BERT tokenizer when applying lowercase. This doesn't impact sequence classification much, but may influence token classfication tasks significantly.
* I noticed a lot of duplicates in the Open Subtitles dataset, which dominates the training corpus.
* I didn't use Whole Word Masking.
* The cased model improves on the uncased model in the following ways:
* All Polish characters and accents should now be tokenized correctly.
* I removed duplicates from Open Subtitles dataset. The corpus is smaller, but more balanced now.
* The model is trained with Whole Word Masking.
## Pre-training corpora
Below is the list of corpora used along with the output of `wc` command (counting lines, words and characters). These corpora were divided into sentences with srxsegmenter (see references), concatenated and tokenized with HuggingFace BERT Tokenizer.
### Uncased
| Tables | Lines | Words | Characters |
| ------------- |--------------:| -----:| -----:|
| [Polish subset of Open Subtitles](http://opus.nlpl.eu/OpenSubtitles-v2018.php) | 236635408| 1431199601 | 7628097730 |
| [Polish subset of ParaCrawl](http://opus.nlpl.eu/ParaCrawl.php) | 8470950 | 176670885 | 1163505275 |
| [Polish Parliamentary Corpus](http://clip.ipipan.waw.pl/PPC) | 9799859 | 121154785 | 938896963 |
| [Polish Wikipedia - Feb 2020](https://dumps.wikimedia.org/plwiki/latest/plwiki-latest-pages-articles.xml.bz2) | 8014206 | 132067986 | 1015849191 |
| Total | 262920423 | 1861093257 | 10746349159 |
### Cased
| Tables | Lines | Words | Characters |
| ------------- |--------------:| -----:| -----:|
| [Polish subset of Open Subtitles (Deduplicated) ](http://opus.nlpl.eu/OpenSubtitles-v2018.php) | 41998942| 213590656 | 1424873235 |
| [Polish subset of ParaCrawl](http://opus.nlpl.eu/ParaCrawl.php) | 8470950 | 176670885 | 1163505275 |
| [Polish Parliamentary Corpus](http://clip.ipipan.waw.pl/PPC) | 9799859 | 121154785 | 938896963 |
| [Polish Wikipedia - Feb 2020](https://dumps.wikimedia.org/plwiki/latest/plwiki-latest-pages-articles.xml.bz2) | 8014206 | 132067986 | 1015849191 |
| Total | 68283960 | 646479197 | 4543124667 |
## Pre-training details
### Uncased
* Polbert was trained with code provided in Google BERT's github repository (https://github.com/google-research/bert)
* Currently released model follows bert-base-uncased model architecture (12-layer, 768-hidden, 12-heads, 110M parameters)
* Training set-up: in total 1 million training steps:
* 100.000 steps - 128 sequence length, batch size 512, learning rate 1e-4 (10.000 steps warmup)
* 800.000 steps - 128 sequence length, batch size 512, learning rate 5e-5
* 100.000 steps - 512 sequence length, batch size 256, learning rate 2e-5
* The model was trained on a single Google Cloud TPU v3-8
### Cased
* Same approach as uncased model, with the following differences:
* Whole Word Masking
* Training set-up:
* 100.000 steps - 128 sequence length, batch size 2048, learning rate 1e-4 (10.000 steps warmup)
* 100.000 steps - 128 sequence length, batch size 2048, learning rate 5e-5
* 100.000 steps - 512 sequence length, batch size 256, learning rate 2e-5
## Usage
Polbert is released via [HuggingFace Transformers library](https://huggingface.co/transformers/).
For an example use as language model, see [this notebook](/LM_testing.ipynb) file.
### Uncased
```python
from transformers import *
model = BertForMaskedLM.from_pretrained("dkleczek/bert-base-polish-uncased-v1")
tokenizer = BertTokenizer.from_pretrained("dkleczek/bert-base-polish-uncased-v1")
nlp = pipeline('fill-mask', model=model, tokenizer=tokenizer)
for pred in nlp(f"Adam Mickiewicz wielkim polskim {nlp.tokenizer.mask_token} był."):
print(pred)
# Output:
# {'sequence': '[CLS] adam mickiewicz wielkim polskim poeta był. [SEP]', 'score': 0.47196975350379944, 'token': 26596}
# {'sequence': '[CLS] adam mickiewicz wielkim polskim bohaterem był. [SEP]', 'score': 0.09127858281135559, 'token': 10953}
# {'sequence': '[CLS] adam mickiewicz wielkim polskim człowiekiem był. [SEP]', 'score': 0.0647173821926117, 'token': 5182}
# {'sequence': '[CLS] adam mickiewicz wielkim polskim pisarzem był. [SEP]', 'score': 0.05232388526201248, 'token': 24293}
# {'sequence': '[CLS] adam mickiewicz wielkim polskim politykiem był. [SEP]', 'score': 0.04554257541894913, 'token': 44095}
```
### Cased
```python
model = BertForMaskedLM.from_pretrained("dkleczek/bert-base-polish-cased-v1")
tokenizer = BertTokenizer.from_pretrained("dkleczek/bert-base-polish-cased-v1")
nlp = pipeline('fill-mask', model=model, tokenizer=tokenizer)
for pred in nlp(f"Adam Mickiewicz wielkim polskim {nlp.tokenizer.mask_token} był."):
print(pred)
# Output:
# {'sequence': '[CLS] Adam Mickiewicz wielkim polskim pisarzem był. [SEP]', 'score': 0.5391148328781128, 'token': 37120}
# {'sequence': '[CLS] Adam Mickiewicz wielkim polskim człowiekiem był. [SEP]', 'score': 0.11683262139558792, 'token': 6810}
# {'sequence': '[CLS] Adam Mickiewicz wielkim polskim bohaterem był. [SEP]', 'score': 0.06021466106176376, 'token': 17709}
# {'sequence': '[CLS] Adam Mickiewicz wielkim polskim mistrzem był. [SEP]', 'score': 0.051870670169591904, 'token': 14652}
# {'sequence': '[CLS] Adam Mickiewicz wielkim polskim artystą był. [SEP]', 'score': 0.031787533313035965, 'token': 35680}
```
See the next section for an example usage of Polbert in downstream tasks.
## Evaluation
Thanks to Allegro, we now have the [KLEJ benchmark](https://klejbenchmark.com/leaderboard/), a set of nine evaluation tasks for the Polish language understanding. The following results are achieved by running standard set of evaluation scripts (no tricks!) utilizing both cased and uncased variants of Polbert.
| Model | Average | NKJP-NER | CDSC-E | CDSC-R | CBD | PolEmo2.0-IN | PolEmo2.0-OUT | DYK | PSC | AR |
| ------------- |--------------:|--------------:|--------------:|--------------:|--------------:|--------------:|--------------:|--------------:|--------------:|--------------:|
| Polbert cased | 81.7 | 93.6 | 93.4 | 93.8 | 52.7 | 87.4 | 71.1 | 59.1 | 98.6 | 85.2 |
| Polbert uncased | 81.4 | 90.1 | 93.9 | 93.5 | 55.0 | 88.1 | 68.8 | 59.4 | 98.8 | 85.4 |
Note how the uncased model performs better than cased on some tasks? My guess this is because of the oversampling of Open Subtitles dataset and its similarity to data in some of these tasks. All these benchmark tasks are sequence classification, so the relative strength of the cased model is not so visible here.
## Bias
The data used to train the model is biased. It may reflect stereotypes related to gender, ethnicity etc. Please be careful when using the model for downstream task to consider these biases and mitigate them.
## Acknowledgements
* I'd like to express my gratitude to Google [TensorFlow Research Cloud (TFRC)](https://www.tensorflow.org/tfrc) for providing the free TPU credits - thank you!
* Also appreciate the help from Timo Möller from [deepset](https://deepset.ai) for sharing tips and scripts based on their experience training German BERT model.
* Big thanks to Allegro for releasing KLEJ Benchmark and specifically to Piotr Rybak for help with the evaluation and pointing out some issues with the tokenization.
* Finally, thanks to Rachel Thomas, Jeremy Howard and Sylvain Gugger from [fastai](https://www.fast.ai) for their NLP and Deep Learning courses!
## Author
Darek Kłeczek - contact me on Twitter [@dk21](https://twitter.com/dk21)
## References
* https://github.com/google-research/bert
* https://github.com/narusemotoki/srx_segmenter
* SRX rules file for sentence splitting in Polish, written by Marcin Miłkowski: https://raw.githubusercontent.com/languagetool-org/languagetool/master/languagetool-core/src/main/resources/org/languagetool/resource/segment.srx
* [KLEJ benchmark](https://klejbenchmark.com/leaderboard/)
|
ibm-granite/granite-20b-code-instruct | ibm-granite | "2024-05-10T06:15:41Z" | 8,811 | 24 | transformers | [
"transformers",
"safetensors",
"gpt_bigcode",
"text-generation",
"code",
"granite",
"conversational",
"dataset:bigcode/commitpackft",
"dataset:TIGER-Lab/MathInstruct",
"dataset:meta-math/MetaMathQA",
"dataset:glaiveai/glaive-code-assistant-v3",
"dataset:glaive-function-calling-v2",
"dataset:bugdaryan/sql-create-context-instruction",
"dataset:garage-bAInd/Open-Platypus",
"dataset:nvidia/HelpSteer",
"arxiv:2405.04324",
"base_model:ibm-granite/granite-20b-code-base",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2024-04-26T05:55:02Z" | ---
pipeline_tag: text-generation
base_model: ibm-granite/granite-20b-code-base
inference: true
license: apache-2.0
datasets:
- bigcode/commitpackft
- TIGER-Lab/MathInstruct
- meta-math/MetaMathQA
- glaiveai/glaive-code-assistant-v3
- glaive-function-calling-v2
- bugdaryan/sql-create-context-instruction
- garage-bAInd/Open-Platypus
- nvidia/HelpSteer
metrics:
- code_eval
library_name: transformers
tags:
- code
- granite
model-index:
- name: granite-20b-code-instruct
results:
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalSynthesis(Python)
metrics:
- name: pass@1
type: pass@1
value: 60.4
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalSynthesis(JavaScript)
metrics:
- name: pass@1
type: pass@1
value: 53.7
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalSynthesis(Java)
metrics:
- name: pass@1
type: pass@1
value: 58.5
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalSynthesis(Go)
metrics:
- name: pass@1
type: pass@1
value: 42.1
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalSynthesis(C++)
metrics:
- name: pass@1
type: pass@1
value: 45.7
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalSynthesis(Rust)
metrics:
- name: pass@1
type: pass@1
value: 42.7
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalExplain(Python)
metrics:
- name: pass@1
type: pass@1
value: 44.5
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalExplain(JavaScript)
metrics:
- name: pass@1
type: pass@1
value: 42.7
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalExplain(Java)
metrics:
- name: pass@1
type: pass@1
value: 49.4
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalExplain(Go)
metrics:
- name: pass@1
type: pass@1
value: 32.3
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalExplain(C++)
metrics:
- name: pass@1
type: pass@1
value: 42.1
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalExplain(Rust)
metrics:
- name: pass@1
type: pass@1
value: 18.3
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalFix(Python)
metrics:
- name: pass@1
type: pass@1
value: 43.9
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalFix(JavaScript)
metrics:
- name: pass@1
type: pass@1
value: 43.9
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalFix(Java)
metrics:
- name: pass@1
type: pass@1
value: 45.7
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalFix(Go)
metrics:
- name: pass@1
type: pass@1
value: 41.5
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalFix(C++)
metrics:
- name: pass@1
type: pass@1
value: 41.5
veriefied: false
- task:
type: text-generation
dataset:
type: bigcode/humanevalpack
name: HumanEvalFix(Rust)
metrics:
- name: pass@1
type: pass@1
value: 29.9
veriefied: false
---
# Granite-20B-Code-Instruct
## Model Summary
**Granite-20B-Code-Instruct** is a 20B parameter model fine tuned from *Granite-20B-Code-Base* on a combination of **permissively licensed** instruction data to enhance instruction following capabilities including logical reasoning and problem-solving skills.
- **Developers:** IBM Research
- **GitHub Repository:** [ibm-granite/granite-code-models](https://github.com/ibm-granite/granite-code-models)
- **Paper:** [Granite Code Models: A Family of Open Foundation Models for Code Intelligence](https://arxiv.org/abs/2405.04324)
- **Release Date**: May 6th, 2024
- **License:** [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0).
## Usage
### Intended use
The model is designed to respond to coding related instructions and can be used to build coding assitants.
<!-- TO DO: Check starcoder2 instruct code example that includes the template https://huggingface.co/bigcode/starcoder2-15b-instruct-v0.1 -->
### Generation
This is a simple example of how to use **Granite-20B-Code-Instruct** model.
```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
device = "cuda" # or "cpu"
model_path = "ibm-granite/granite-20b-code-instruct"
tokenizer = AutoTokenizer.from_pretrained(model_path)
# drop device_map if running on CPU
model = AutoModelForCausalLM.from_pretrained(model_path, device_map=device)
model.eval()
# change input text as desired
chat = [
{ "role": "user", "content": "Write a code to find the maximum value in a list of numbers." },
]
chat = tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=True)
# tokenize the text
input_tokens = tokenizer(chat, return_tensors="pt")
# transfer tokenized inputs to the device
for i in input_tokens:
input_tokens[i] = input_tokens[i].to(device)
# generate output tokens
output = model.generate(**input_tokens, max_new_tokens=100)
# decode output tokens into text
output = tokenizer.batch_decode(output)
# loop over the batch to print, in this example the batch size is 1
for i in output:
print(i)
```
<!-- TO DO: Check this part -->
## Training Data
Granite Code Instruct models are trained on the following types of data.
* Code Commits Datasets: we sourced code commits data from the [CommitPackFT](https://huggingface.co/datasets/bigcode/commitpackft) dataset, a filtered version of the full CommitPack dataset. From CommitPackFT dataset, we only consider data for 92 programming languages. Our inclusion criteria boils down to selecting programming languages common across CommitPackFT and the 116 languages that we considered to pretrain the code-base model (*Granite-20B-Code-Base*).
* Math Datasets: We consider two high-quality math datasets, [MathInstruct](https://huggingface.co/datasets/TIGER-Lab/MathInstruct) and [MetaMathQA](https://huggingface.co/datasets/meta-math/MetaMathQA). Due to license issues, we filtered out GSM8K-RFT and Camel-Math from MathInstruct dataset.
* Code Instruction Datasets: We use [Glaive-Code-Assistant-v3](https://huggingface.co/datasets/glaiveai/glaive-code-assistant-v3), [Glaive-Function-Calling-v2](https://huggingface.co/datasets/glaiveai/glaive-function-calling-v2), [NL2SQL11](https://huggingface.co/datasets/bugdaryan/sql-create-context-instruction) and a small collection of synthetic API calling datasets.
* Language Instruction Datasets: We include high-quality datasets such as [HelpSteer](https://huggingface.co/datasets/nvidia/HelpSteer) and an open license-filtered version of [Platypus](https://huggingface.co/datasets/garage-bAInd/Open-Platypus). We also include a collection of hardcoded prompts to ensure our model generates correct outputs given inquiries about its name or developers.
## Infrastructure
We train the Granite Code models using two of IBM's super computing clusters, namely Vela and Blue Vela, both outfitted with NVIDIA A100 and H100 GPUs respectively. These clusters provide a scalable and efficient infrastructure for training our models over thousands of GPUs.
## Ethical Considerations and Limitations
Granite code instruct models are primarily finetuned using instruction-response pairs across a specific set of programming languages. Thus, their performance may be limited with out-of-domain programming languages. In this situation, it is beneficial providing few-shot examples to steer the model's output. Moreover, developers should perform safety testing and target-specific tuning before deploying these models on critical applications. The model also inherits ethical considerations and limitations from its base model. For more information, please refer to *[Granite-20B-Code-Base](https://huggingface.co/ibm-granite/granite-20b-code-base)* model card.
|
mosaicml/mpt-7b-instruct | mosaicml | "2024-03-05T20:24:35Z" | 8,803 | 461 | transformers | [
"transformers",
"pytorch",
"mpt",
"text-generation",
"Composer",
"MosaicML",
"llm-foundry",
"custom_code",
"dataset:mosaicml/dolly_hhrlhf",
"arxiv:2205.14135",
"arxiv:2108.12409",
"arxiv:2010.04245",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2023-05-05T00:52:12Z" | ---
license: apache-2.0
datasets:
- mosaicml/dolly_hhrlhf
tags:
- Composer
- MosaicML
- llm-foundry
inference: false
---
# MPT-7B-Instruct
MPT-7B-Instruct is a model for short-form instruction following.
It is built by finetuning [MPT-7B](https://huggingface.co/mosaicml/mpt-7b) on a [dataset](https://huggingface.co/datasets/sam-mosaic/dolly_hhrlhf) derived from the [Databricks Dolly-15k](https://huggingface.co/datasets/databricks/databricks-dolly-15k) and the [Anthropic Helpful and Harmless (HH-RLHF)](https://huggingface.co/datasets/Anthropic/hh-rlhf) datasets.
* License: Apache 2.0
* [Demo on Hugging Face Spaces](https://huggingface.co/spaces/mosaicml/mpt-7b-instruct)
This model was trained by [MosaicML](https://www.mosaicml.com) and follows a modified decoder-only transformer architecture.
## Model Date
May 5, 2023
## Model License
Apache 2.0
## Documentation
* [Blog post: Introducing MPT-7B: A New Standard for Open-Source, Commercially Usable LLMs](https://www.mosaicml.com/blog/mpt-7b)
* [Codebase (mosaicml/llm-foundry repo)](https://github.com/mosaicml/llm-foundry/)
* Questions: Feel free to contact us via the [MosaicML Community Slack](https://mosaicml.me/slack)!
### Example Question/Instruction
**Longboi24**:
> What is a quoll?
**MPT-7B-Instruct**:
>A Quoll (pronounced “cool”) is one of Australia’s native carnivorous marsupial mammals, which are also known as macropods or wallabies in other parts around Asia and South America
## How to Use
Note: This model requires that `trust_remote_code=True` be passed to the `from_pretrained` method. This is because we use a custom model architecture that is not yet part of the `transformers` package.
It includes options for many training efficiency features such as [FlashAttention (Dao et al. 2022)](https://arxiv.org/pdf/2205.14135.pdf), [ALiBi](https://arxiv.org/abs/2108.12409), QK LayerNorm, and more.
```python
import transformers
model = transformers.AutoModelForCausalLM.from_pretrained(
'mosaicml/mpt-7b-instruct',
trust_remote_code=True
)
```
Note: This model requires that `trust_remote_code=True` be passed to the `from_pretrained` method.
This is because we use a custom `MPT` model architecture that is not yet part of the Hugging Face `transformers` package.
`MPT` includes options for many training efficiency features such as [FlashAttention](https://arxiv.org/pdf/2205.14135.pdf), [ALiBi](https://arxiv.org/abs/2108.12409), [QK LayerNorm](https://arxiv.org/abs/2010.04245), and more.
To use the optimized [triton implementation](https://github.com/openai/triton) of FlashAttention, you can load the model on GPU (`cuda:0`) with `attn_impl='triton'` and with `bfloat16` precision:
```python
import torch
import transformers
name = 'mosaicml/mpt-7b-instruct'
config = transformers.AutoConfig.from_pretrained(name, trust_remote_code=True)
config.attn_config['attn_impl'] = 'triton'
config.init_device = 'cuda:0' # For fast initialization directly on GPU!
model = transformers.AutoModelForCausalLM.from_pretrained(
name,
config=config,
torch_dtype=torch.bfloat16, # Load model weights in bfloat16
trust_remote_code=True
)
```
Although the model was trained with a sequence length of 2048, ALiBi enables users to increase the maximum sequence length during finetuning and/or inference. For example:
```python
import transformers
name = 'mosaicml/mpt-7b-instruct'
config = transformers.AutoConfig.from_pretrained(name, trust_remote_code=True)
config.max_seq_len = 4096 # (input + output) tokens can now be up to 4096
model = transformers.AutoModelForCausalLM.from_pretrained(
name,
config=config,
trust_remote_code=True
)
```
This model was trained with the [EleutherAI/gpt-neox-20b](https://huggingface.co/EleutherAI/gpt-neox-20b) tokenizer.
```python
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neox-20b")
```
The model can then be used, for example, within a text-generation pipeline.
Note: when running Torch modules in lower precision, it is best practice to use the [torch.autocast context manager](https://pytorch.org/docs/stable/amp.html).
```python
from transformers import pipeline
pipe = pipeline('text-generation', model=model, tokenizer=tokenizer, device='cuda:0')
with torch.autocast('cuda', dtype=torch.bfloat16):
print(
pipe('Here is a recipe for vegan banana bread:\n',
max_new_tokens=100,
do_sample=True,
use_cache=True))
```
### Formatting
This model was trained on data formatted in the dolly-15k format:
```python
INSTRUCTION_KEY = "### Instruction:"
RESPONSE_KEY = "### Response:"
INTRO_BLURB = "Below is an instruction that describes a task. Write a response that appropriately completes the request."
PROMPT_FOR_GENERATION_FORMAT = """{intro}
{instruction_key}
{instruction}
{response_key}
""".format(
intro=INTRO_BLURB,
instruction_key=INSTRUCTION_KEY,
instruction="{instruction}",
response_key=RESPONSE_KEY,
)
example = "James decides to run 3 sprints 3 times a week. He runs 60 meters each sprint. How many total meters does he run a week? Explain before answering."
fmt_ex = PROMPT_FOR_GENERATION_FORMAT.format(instruction=example)
```
In the above example, `fmt_ex` is ready to be tokenized and sent through the model.
## Model Description
The architecture is a modification of a standard decoder-only transformer.
The model has been modified from a standard transformer in the following ways:
* It uses [FlashAttention](https://arxiv.org/pdf/2205.14135.pdf)
* It uses [ALiBi (Attention with Linear Biases)](https://arxiv.org/abs/2108.12409) and does not use positional embeddings
* It does not use biases
| Hyperparameter | Value |
|----------------|-------|
|n_parameters | 6.7B |
|n_layers | 32 |
| n_heads | 32 |
| d_model | 4096 |
| vocab size | 50432 |
| sequence length | 2048 |
## PreTraining Data
For more details on the pretraining process, see [MPT-7B](https://huggingface.co/mosaicml/mpt-7b).
The data was tokenized using the [EleutherAI/gpt-neox-20b](https://huggingface.co/EleutherAI/gpt-neox-20b) tokenizer.
### Training Configuration
This model was trained on 8 A100-40GBs for about 2.3 hours using the [MosaicML Platform](https://www.mosaicml.com/platform).
The model was trained with sharded data parallelism using [FSDP](https://pytorch.org/docs/stable/fsdp.html) and used the AdamW optimizer.
## Limitations and Biases
_The following language is modified from [EleutherAI's GPT-NeoX-20B](https://huggingface.co/EleutherAI/gpt-neox-20b)_
MPT-7B-Instruct can produce factually incorrect output, and should not be relied on to produce factually accurate information.
MPT-7B-Instruct was trained on various public datasets.
While great efforts have been taken to clean the pretraining data, it is possible that this model could generate lewd, biased or otherwise offensive outputs.
## Acknowledgements
This model was finetuned by Sam Havens and the MosaicML NLP team
## MosaicML Platform
If you're interested in [training](https://www.mosaicml.com/training) and [deploying](https://www.mosaicml.com/inference) your own MPT or LLMs on the MosaicML Platform, [sign up here](https://forms.mosaicml.com/demo?utm_source=huggingface&utm_medium=referral&utm_campaign=mpt-7b).
## Disclaimer
The license on this model does not constitute legal advice. We are not responsible for the actions of third parties who use this model. Please cosult an attorney before using this model for commercial purposes.
## Citation
Please cite this model using the following format:
```
@online{MosaicML2023Introducing,
author = {MosaicML NLP Team},
title = {Introducing MPT-7B: A New Standard for Open-Source, Commercially Usable LLMs},
year = {2023},
url = {www.mosaicml.com/blog/mpt-7b},
note = {Accessed: 2023-03-28}, % change this date
urldate = {2023-03-28} % change this date
}
``` |
1aurent/phikon-distil-mobilenet_v2-kather2016 | 1aurent | "2023-11-09T13:53:37Z" | 8,802 | 0 | transformers | [
"transformers",
"tensorboard",
"safetensors",
"mobilenet_v2",
"image-classification",
"feature-extraction",
"biology",
"cancer",
"owkin",
"histology",
"dataset:1aurent/Kather-texture-2016",
"base_model:1aurent/phikon-finetuned-lora-kather2016",
"license:other",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | image-classification | "2023-11-09T10:16:17Z" | ---
library_name: transformers
base_model: 1aurent/phikon-finetuned-lora-kather2016
tags:
- feature-extraction
- image-classification
- biology
- cancer
- owkin
- histology
model-index:
- name: owkin_pancancer
results:
- task:
type: image-classification
name: Image Classification
dataset:
name: 1aurent/Kather-texture-2016
type: image-classification
metrics:
- type: accuracy
value: 0.928
name: accuracy
verified: false
license: other
license_name: owkin-non-commercial
license_link: https://github.com/owkin/HistoSSLscaling/blob/main/LICENSE.txt
pipeline_tag: image-classification
datasets:
- 1aurent/Kather-texture-2016
metrics:
- accuracy
widget:
- src: >-
https://datasets-server.huggingface.co/assets/1aurent/Kather-texture-2016/--/default/train/0/image/image.jpg
example_title: adipose
---
# Model card for phikon-distil-mobilenet_v2-kather2016
This model is a distilled version of [owkin/phikon](https://huggingface.co/owkin/phikon) to a MobileNet-v2 on the [1aurent/Kather-texture-2016](https://huggingface.co/datasets/1aurent/Kather-texture-2016) dataset.
## Model Usage
### Image Classification
```python
from transformers import AutoModelForImageClassification, AutoImageProcessor
from urllib.request import urlopen
from PIL import Image
# get example histology image
img = Image.open(
urlopen(
"https://datasets-server.huggingface.co/assets/1aurent/Kather-texture-2016/--/default/train/0/image/image.jpg"
)
)
# load image_processor and model from the hub
model_name = "1aurent/phikon-distil-mobilenet_v2-kather2016"
image_processor = AutoImageProcessor.from_pretrained(model_name)
model = AutoModelForImageClassification.from_pretrained(model_name)
inputs = image_processor(img, return_tensors="pt")
outputs = model(**inputs)
```
## Citation
```bibtex
@article{Filiot2023.07.21.23292757,
author = {Alexandre Filiot and Ridouane Ghermi and Antoine Olivier and Paul Jacob and Lucas Fidon and Alice Mac Kain and Charlie Saillard and Jean-Baptiste Schiratti},
title = {Scaling Self-Supervised Learning for Histopathology with Masked Image Modeling},
elocation-id = {2023.07.21.23292757},
year = {2023},
doi = {10.1101/2023.07.21.23292757},
publisher = {Cold Spring Harbor Laboratory Press},
url = {https://www.medrxiv.org/content/early/2023/09/14/2023.07.21.23292757},
eprint = {https://www.medrxiv.org/content/early/2023/09/14/2023.07.21.23292757.full.pdf},
journal = {medRxiv}
}
``` |
MaziyarPanahi/Qwen2-7B-Instruct-v0.3-GGUF | MaziyarPanahi | "2024-06-27T16:38:42Z" | 8,779 | 1 | transformers | [
"transformers",
"gguf",
"mistral",
"quantized",
"2-bit",
"3-bit",
"4-bit",
"5-bit",
"6-bit",
"8-bit",
"GGUF",
"text-generation",
"qwen",
"qwen-2",
"base_model:MaziyarPanahi/Qwen2-7B-Instruct-v0.3",
"text-generation-inference",
"region:us"
] | text-generation | "2024-06-27T16:06:19Z" | ---
tags:
- quantized
- 2-bit
- 3-bit
- 4-bit
- 5-bit
- 6-bit
- 8-bit
- GGUF
- text-generation
- qwen
- qwen-2
- text-generation
model_name: Qwen2-7B-Instruct-v0.3-GGUF
base_model: MaziyarPanahi/Qwen2-7B-Instruct-v0.3
inference: false
model_creator: MaziyarPanahi
pipeline_tag: text-generation
quantized_by: MaziyarPanahi
---
# [MaziyarPanahi/Qwen2-7B-Instruct-v0.3-GGUF](https://huggingface.co/MaziyarPanahi/Qwen2-7B-Instruct-v0.3-GGUF)
- Model creator: [MaziyarPanahi](https://huggingface.co/MaziyarPanahi)
- Original model: [MaziyarPanahi/Qwen2-7B-Instruct-v0.3](https://huggingface.co/MaziyarPanahi/Qwen2-7B-Instruct-v0.3)
## Description
[MaziyarPanahi/Qwen2-7B-Instruct-v0.3-GGUF](https://huggingface.co/MaziyarPanahi/Qwen2-7B-Instruct-v0.3-GGUF) contains GGUF format model files for [MaziyarPanahi/Qwen2-7B-Instruct-v0.3](https://huggingface.co/MaziyarPanahi/Qwen2-7B-Instruct-v0.3).
### About GGUF
GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp.
Here is an incomplete list of clients and libraries that are known to support GGUF:
* [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option.
* [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server.
* [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration. Linux available, in beta as of 27/11/2023.
* [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI, with many features and powerful extensions. Supports GPU acceleration.
* [KoboldCpp](https://github.com/LostRuins/koboldcpp), a fully featured web UI, with GPU accel across all platforms and GPU architectures. Especially good for story telling.
* [GPT4All](https://gpt4all.io/index.html), a free and open source local running GUI, supporting Windows, Linux and macOS with full GPU accel.
* [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), a great web UI with many interesting and unique features, including a full model library for easy model selection.
* [Faraday.dev](https://faraday.dev/), an attractive and easy to use character-based chat GUI for Windows and macOS (both Silicon and Intel), with GPU acceleration.
* [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use.
* [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server. Note, as of time of writing (November 27th 2023), ctransformers has not been updated in a long time and does not support many recent models.
## Special thanks
🙏 Special thanks to [Georgi Gerganov](https://github.com/ggerganov) and the whole team working on [llama.cpp](https://github.com/ggerganov/llama.cpp/) for making all of this possible. |
Voicelab/vlt5-base-keywords | Voicelab | "2023-08-16T07:34:41Z" | 8,775 | 46 | transformers | [
"transformers",
"pytorch",
"safetensors",
"t5",
"text2text-generation",
"keywords-generation",
"text-classifiation",
"other",
"pl",
"en",
"dataset:posmac",
"arxiv:2209.14008",
"license:cc-by-4.0",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text2text-generation | "2022-09-27T12:13:59Z" | ---
license: cc-by-4.0
language:
- pl
- en
datasets:
- posmac
pipeline_tag: text2text-generation
pipeline_kwargs:
- no_repeat_ngram_size=3
- num_beams=4
tags:
- keywords-generation
- text-classifiation
- other
widget:
- text: "Keywords: Our vlT5 model is a keyword generation model based on encoder-decoder architecture using Transformer blocks presented by google (https://huggingface.co/t5-base). The vlT5 was trained on scientific articles corpus to predict a given set of keyphrases based on the concatenation of the article’s abstract and title. It generates precise, yet not always complete keyphrases that describe the content of the article based only on the abstract."
example_title: "English 1"
- text: "Keywords: Decays the learning rate of each parameter group by gamma every step_size epochs. Notice that such decay can happen simultaneously with other changes to the learning rate from outside this scheduler. When last_epoch=-1, sets initial lr as lr."
example_title: "English 2"
- text: "Keywords: Przełomem w dziedzinie sztucznej inteligencji i maszynowego uczenia się było powstanie systemu eksperckiego Dendral na Uniwersytecie Stanforda w 1965. System ten powstał w celu zautomatyzowania analizy i identyfikacji molekuł związków organicznych, które dotychczas nie były znane chemikom. Wyniki badań otrzymane dzięki systemowi Dendral były pierwszym w historii odkryciem dokonanym przez komputer, które zostały opublikowane w prasie specjalistycznej."
example_title: "Polish"
- text: "Keywords: El análisis de un economista calcula que, a pesar del aumento del gasto general, la Navidad es una pérdida de peso muerto según la teoría microeconómica ortodoxa, debido al efecto de dar regalos. Esta pérdida se calcula como la diferencia entre lo que el donante gastó en el artículo y lo que el receptor del regalo habría pagado por el artículo. Se estima que en 2001, Navidad resultó en una pérdida de peso muerto de $ 4 mil millones solo en los EE. UU.1 Debido a factores de complicación, este análisis se utiliza a veces para discutir posibles fallas en la teoría microeconómica actual. Otras pérdidas de peso muerto incluyen los efectos de la Navidad en el medio ambiente y el hecho de que los regalos materiales a menudo se perciben como elefantes blancos, lo que impone costos de mantenimiento y almacenamiento y contribuye al desorden."
example_title: "Spanish"
metrics:
- f1
- precision
- recall
---
<img src="https://public.3.basecamp.com/p/rs5XqmAuF1iEuW6U7nMHcZeY/upload/download/VL-NLP-short.png" alt="logo voicelab nlp" style="width:300px;"/>
# Keyword Extraction from Short Texts with T5
> Our vlT5 model is a keyword generation model based on encoder-decoder architecture using Transformer blocks presented by Google ([https://huggingface.co/t5-base](https://huggingface.co/t5-base)). The vlT5 was trained on scientific articles corpus to predict a given set of keyphrases based on the concatenation of the article’s abstract and title. It generates precise, yet not always complete keyphrases that describe the content of the article based only on the abstract.
**Keywords generated with vlT5-base-keywords:** encoder-decoder architecture, keyword generation
Results on demo model (different generation method, one model per language):
> Our vlT5 model is a keyword generation model based on encoder-decoder architecture using Transformer blocks presented by Google ([https://huggingface.co/t5-base](https://huggingface.co/t5-base)). The vlT5 was trained on scientific articles corpus to predict a given set of keyphrases based on the concatenation of the article’s abstract and title. It generates precise, yet not always complete keyphrases that describe the content of the article based only on the abstract.
**Keywords generated with vlT5-base-keywords:** encoder-decoder architecture, vlT5, keyword generation, scientific articles corpus
## vlT5
The biggest advantage is the transferability of the vlT5 model, as it works well on all domains and types of text. The downside is that the text length and the number of keywords are similar to the training data: the text piece of an abstract length generates approximately 3 to 5 keywords. It works both extractive and abstractively. Longer pieces of text must be split into smaller chunks, and then propagated to the model.
### Overview
- **Language model:** [t5-base](https://huggingface.co/t5-base)
- **Language:** pl, en (but works relatively well with others)
- **Training data:** POSMAC
- **Online Demo:** Visit our online demo for better results [https://nlp-demo-1.voicelab.ai/](https://nlp-demo-1.voicelab.ai/)
- **Paper:** [Keyword Extraction from Short Texts with a Text-To-Text Transfer Transformer, ACIIDS 2022](https://arxiv.org/abs/2209.14008)
# Corpus
The model was trained on a POSMAC corpus. Polish Open Science Metadata Corpus (POSMAC) is a collection of 216,214 abstracts of scientific publications compiled in the CURLICAT project.
| Domains | Documents | With keywords |
| -------------------------------------------------------- | --------: | :-----------: |
| Engineering and technical sciences | 58 974 | 57 165 |
| Social sciences | 58 166 | 41 799 |
| Agricultural sciences | 29 811 | 15 492 |
| Humanities | 22 755 | 11 497 |
| Exact and natural sciences | 13 579 | 9 185 |
| Humanities, Social sciences | 12 809 | 7 063 |
| Medical and health sciences | 6 030 | 3 913 |
| Medical and health sciences, Social sciences | 828 | 571 |
| Humanities, Medical and health sciences, Social sciences | 601 | 455 |
| Engineering and technical sciences, Humanities | 312 | 312 |
# Tokenizer
As in the original plT5 implementation, the training dataset was tokenized into subwords using a sentencepiece unigram model with vocabulary size of 50k tokens.
# Usage
```python
from transformers import T5Tokenizer, T5ForConditionalGeneration
model = T5ForConditionalGeneration.from_pretrained("Voicelab/vlt5-base-keywords")
tokenizer = T5Tokenizer.from_pretrained("Voicelab/vlt5-base-keywords")
task_prefix = "Keywords: "
inputs = [
"Christina Katrakis, who spoke to the BBC from Vorokhta in western Ukraine, relays the account of one family, who say Russian soldiers shot at their vehicles while they were leaving their village near Chernobyl in northern Ukraine. She says the cars had white flags and signs saying they were carrying children.",
"Decays the learning rate of each parameter group by gamma every step_size epochs. Notice that such decay can happen simultaneously with other changes to the learning rate from outside this scheduler. When last_epoch=-1, sets initial lr as lr.",
"Hello, I'd like to order a pizza with salami topping.",
]
for sample in inputs:
input_sequences = [task_prefix + sample]
input_ids = tokenizer(
input_sequences, return_tensors="pt", truncation=True
).input_ids
output = model.generate(input_ids, no_repeat_ngram_size=3, num_beams=4)
predicted = tokenizer.decode(output[0], skip_special_tokens=True)
print(sample, "\n --->", predicted)
```
# Inference
Our results showed that the best generation results were achieved with `no_repeat_ngram_size=3, num_beams=4`
# Results
| Method | Rank | Micro | | | Macro | | |
| ----------- | ---: | :--------: | ---------: | ---------: | :---: | ----: | ----: |
| | | P | R | F1 | P | R | F1 |
| extremeText | 1 | 0.175 | 0.038 | 0.063 | 0.007 | 0.004 | 0.005 |
| | 3 | 0.117 | 0.077 | 0.093 | 0.011 | 0.011 | 0.011 |
| | 5 | 0.090 | 0.099 | 0.094 | 0.013 | 0.016 | 0.015 |
| | 10 | 0.060 | 0.131 | 0.082 | 0.015 | 0.025 | 0.019 |
| vlT5kw | 1 | **0.345** | 0.076 | 0.124 | 0.054 | 0.047 | 0.050 |
| | 3 | 0.328 | 0.212 | 0.257 | 0.133 | 0.127 | 0.129 |
| | 5 | 0.318 | **0.237** | **0.271** | 0.143 | 0.140 | 0.141 |
| KeyBERT | 1 | 0.030 | 0.007 | 0.011 | 0.004 | 0.003 | 0.003 |
| | 3 | 0.015 | 0.010 | 0.012 | 0.006 | 0.004 | 0.005 |
| | 5 | 0.011 | 0.012 | 0.011 | 0.006 | 0.005 | 0.005 |
| TermoPL | 1 | 0.118 | 0.026 | 0.043 | 0.004 | 0.003 | 0.003 |
| | 3 | 0.070 | 0.046 | 0.056 | 0.006 | 0.005 | 0.006 |
| | 5 | 0.051 | 0.056 | 0.053 | 0.007 | 0.007 | 0.007 |
| | all | 0.025 | 0.339 | 0.047 | 0.017 | 0.030 | 0.022 |
| extremeText | 1 | 0.210 | 0.077 | 0.112 | 0.037 | 0.017 | 0.023 |
| | 3 | 0.139 | 0.152 | 0.145 | 0.045 | 0.042 | 0.043 |
| | 5 | 0.107 | 0.196 | 0.139 | 0.049 | 0.063 | 0.055 |
| | 10 | 0.072 | 0.262 | 0.112 | 0.041 | 0.098 | 0.058 |
| vlT5kw | 1 | **0.377** | 0.138 | 0.202 | 0.119 | 0.071 | 0.089 |
| | 3 | 0.361 | 0.301 | 0.328 | 0.185 | 0.147 | 0.164 |
| | 5 | 0.357 | **0.316** | **0.335** | 0.188 | 0.153 | 0.169 |
| KeyBERT | 1 | 0.018 | 0.007 | 0.010 | 0.003 | 0.001 | 0.001 |
| | 3 | 0.009 | 0.010 | 0.009 | 0.004 | 0.001 | 0.002 |
| | 5 | 0.007 | 0.012 | 0.009 | 0.004 | 0.001 | 0.002 |
| TermoPL | 1 | 0.076 | 0.028 | 0.041 | 0.002 | 0.001 | 0.001 |
| | 3 | 0.046 | 0.051 | 0.048 | 0.003 | 0.001 | 0.002 |
| | 5 | 0.033 | 0.061 | 0.043 | 0.003 | 0.001 | 0.002 |
| | all | 0.021 | 0.457 | 0.040 | 0.004 | 0.008 | 0.005 |
# License
CC BY 4.0
# Citation
If you use this model, please cite the following paper:
[Pęzik, P., Mikołajczyk, A., Wawrzyński, A., Żarnecki, F., Nitoń, B., Ogrodniczuk, M. (2023). Transferable Keyword Extraction and Generation with Text-to-Text Language Models. In: Mikyška, J., de Mulatier, C., Paszynski, M., Krzhizhanovskaya, V.V., Dongarra, J.J., Sloot, P.M. (eds) Computational Science – ICCS 2023. ICCS 2023. Lecture Notes in Computer Science, vol 14074. Springer, Cham. https://doi.org/10.1007/978-3-031-36021-3_42](https://link.springer.com/chapter/10.1007/978-3-031-36021-3_42)
OR
[Piotr Pęzik, Agnieszka Mikołajczyk-Bareła, Adam Wawrzyński, Bartłomiej Nitoń, Maciej Ogrodniczuk, Keyword Extraction from Short Texts with a Text-To-Text Transfer Transformer, ACIIDS 2022](https://arxiv.org/abs/2209.14008)
# Authors
The model was trained by NLP Research Team at Voicelab.ai.
You can contact us [here](https://voicelab.ai/contact/).
|
MazenAmria/swin-tiny-finetuned-cifar100 | MazenAmria | "2022-12-23T21:22:33Z" | 8,774 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"swin",
"image-classification",
"generated_from_trainer",
"dataset:cifar100",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | image-classification | "2022-12-23T19:02:42Z" | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- cifar100
metrics:
- accuracy
model-index:
- name: swin-tiny-finetuned-cifar100
results:
- task:
name: Image Classification
type: image-classification
dataset:
name: cifar100
type: cifar100
args: cifar100
metrics:
- name: Accuracy
type: accuracy
value: 0.8735
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# swin-tiny-finetuned-cifar100
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the cifar100 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4223
- Accuracy: 0.8735
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 4e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 20 (with early stopping)
### Training results
| Training Loss | Epoch | Step | Accuracy | Validation Loss |
|:-------------:|:-----:|:----:|:--------:|:---------------:|
| 0.6439 | 1.0 | 781 | 0.8138 | 0.6126 |
| 0.6222 | 2.0 | 1562 | 0.8393 | 0.5094 |
| 0.2912 | 3.0 | 2343 | 0.861 | 0.4452 |
| 0.2234 | 4.0 | 3124 | 0.8679 | 0.4330 |
| 0.121 | 5.0 | 3905 | 0.8735 | 0.4223 |
| 0.2589 | 6.0 | 4686 | 0.8622 | 0.4775 |
| 0.1419 | 7.0 | 5467 | 0.8642 | 0.4900 |
| 0.1513 | 8.0 | 6248 | 0.8667 | 0.4956 |
### Framework versions
- Transformers 4.20.1
- Pytorch 1.11.0
- Datasets 2.1.0
- Tokenizers 0.12.1
|
MazenAmria/swin-small-finetuned-cifar100 | MazenAmria | "2022-12-30T01:47:41Z" | 8,769 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"swin",
"image-classification",
"generated_from_trainer",
"dataset:cifar100",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | image-classification | "2022-12-29T21:51:36Z" | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- cifar100
metrics:
- accuracy
model-index:
- name: swin-small-finetuned-cifar100
results:
- task:
name: Image Classification
type: image-classification
dataset:
name: cifar100
type: cifar100
args: cifar100
metrics:
- name: Accuracy
type: accuracy
value: 0.8938
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# swin-small-finetuned-cifar100
This model is a fine-tuned version of [microsoft/swin-small-patch4-window7-224](https://huggingface.co/microsoft/swin-small-patch4-window7-224) on the cifar100 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6281
- Accuracy: 0.8938
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 4e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 20
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
| 0.72 | 1.0 | 781 | 0.6691 | 0.8077 |
| 0.6944 | 2.0 | 1562 | 0.4797 | 0.8495 |
| 0.2794 | 3.0 | 2343 | 0.4338 | 0.869 |
| 0.2569 | 4.0 | 3124 | 0.4263 | 0.879 |
| 0.1417 | 5.0 | 3905 | 0.4385 | 0.8819 |
| 0.0961 | 6.0 | 4686 | 0.4720 | 0.8854 |
| 0.0584 | 7.0 | 5467 | 0.4941 | 0.885 |
| 0.0351 | 8.0 | 6248 | 0.5253 | 0.885 |
| 0.0107 | 9.0 | 7029 | 0.5598 | 0.8887 |
| 0.0118 | 10.0 | 7810 | 0.5998 | 0.8858 |
| 0.0097 | 11.0 | 8591 | 0.5957 | 0.8941 |
| 0.0044 | 12.0 | 9372 | 0.6237 | 0.8912 |
| 0.0013 | 13.0 | 10153 | 0.6286 | 0.8929 |
| 0.0102 | 14.0 | 10934 | 0.6281 | 0.8938 |
### Framework versions
- Transformers 4.20.1
- Pytorch 1.11.0
- Datasets 2.1.0
- Tokenizers 0.12.1
|
Yntec/endlessReality | Yntec | "2024-04-24T23:50:38Z" | 8,762 | 0 | diffusers | [
"diffusers",
"safetensors",
"Realism",
"Scifi",
"Portrait",
"davcha",
"stable-diffusion",
"stable-diffusion-diffusers",
"text-to-image",
"license:creativeml-openrail-m",
"autotrain_compatible",
"endpoints_compatible",
"diffusers:StableDiffusionPipeline",
"region:us"
] | text-to-image | "2024-04-24T22:25:00Z" | ---
license: creativeml-openrail-m
library_name: diffusers
pipeline_tag: text-to-image
tags:
- Realism
- Scifi
- Portrait
- davcha
- stable-diffusion
- stable-diffusion-diffusers
- diffusers
- text-to-image
---
# Endless Reality
1.0 version of this model with the 840KVAE baked in. Comparison:
Samples and prompts:
Top left: detailed postcard movie perfect face, full body, baby, masterpiece, highest quality, crowd, realistic eyes, Pretty CUTE GIRL, sweater, skirt,
Top right: Hyperrealistic 1990 movie screenshot Santa Claus with wife and daughter enjoying wine with candles. sitting with a pretty cute little girl, Closeup Faces, Gift Birthday Theme by Gil_Elvgren and Haddon_Sundblom
Bottom left: analog style 70s color photograph of young Chuck Norris, muscular, frying eggs on freezer, swirl magic, solo, from side, side view, detailed background, detailed face, Golden Tech, scifi, timeless wanderer, endless landscape, circular patterns, time magic, time, standing still, bloom light aura, desert dunes in background, ethereal atmosphere
Bottom right: retro style 70s color movie still of beautiful face, young pretty Christina Aguilera voluptuous at a neon convenience storefront
Original page:
https://civitai.com/models/25573?modelVersionId=30619 |
ibm/merlinite-7b | ibm | "2024-03-05T14:01:15Z" | 8,760 | 101 | transformers | [
"transformers",
"safetensors",
"mistral",
"text-generation",
"merlinite",
"ibm",
"lab",
"labrador",
"labradorite",
"en",
"arxiv:2403.01081",
"base_model:mistralai/Mistral-7B-v0.1",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2024-03-02T03:51:35Z" |
---
pipeline_tag: text-generation
tags:
- merlinite
- mistral
- ibm
- lab
- labrador
- labradorite
license: apache-2.0
language:
- en
base_model: mistralai/Mistral-7B-v0.1
---
# Model Card for Merlinite 7b 🔥 [Paper](https://arxiv.org/abs/2403.01081)
### Overview
### Performance
| Model | Alignment | Base | Teacher | MTBench (Avg) * | MMLU(5-shot) | ARC-C(25-shot) | HellaSwag(10-shot) | Winogrande(5-shot) | GSM8K(5-shot- strict) |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| [Llama-2-13b-chat-hf](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf) | RLHF | Llama-2-13b | Human Annotators | 6.65 | 54.58 | 59.81 | 82.52 | 75.93 | 34.80 |
| [Orca-2-13b](https://huggingface.co/microsoft/Orca-2-13b) | Progressive Training | Llama-2-13b | GPT-4 | 6.15 | 60.37 * | 59.73 | 79.86 | 78.22 | 48.22 |
| [WizardLM-13B-V1.2](https://huggingface.co/WizardLM/WizardLM-13B-V1.2) | Evol-Instruct | Llama-2-13b | GPT-4 | 7.20 | 54.83 | 60.24 | 82.62 | 76.40 | 43.75 |
| [Labradorite-13b](https://huggingface.co/ibm/labradorite-13b) | Large-scale Alignment for chatBots (LAB) | Llama-2-13b | Mixtral-8x7B-Instruct | 7.23 | 58.89 | 61.69 | 83.15 | 79.56 | 40.11 |
| [Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) | SFT | Mistral-7B-v0.1 | - | 6.84 | 60.37 | 63.65 | 84.76 | 76.80 | 41.85 |
| [zephyr-7b-beta](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta) | SFT/DPO | Mistral-7B-v0.1 | GPT-4 | 7.34 | 61.07 | 63.74 | 84.19 | 78.06 | 34.04 |
| [Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) | SFT | Mistral-7B-v0.1 | - | 7.6** | 60.78 | 63.14 | 84.88 | 77.19 | 40.03 |
| Merlinite-7b | Large-scale Alignment for chatBots (LAB) | Mistral-7B-v0.1 | Mixtral-8x7B-Instruct | 7.66 | 64.88 | 63.99 | 84.37 | 78.24 | 44.58 |
[*] Numbers for models other than Merlinite-7b and [Labradorite-13b](https://huggingface.co/ibm/labradorite-13b) (ours) are taken from [lmsys/chatbot-arena-leaderboard](https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard)
[**] Numbers taken from [MistralAI Release Blog](https://mistral.ai/news/la-plateforme/)
### Method
LAB: **L**arge-scale **A**lignment for chat**B**ots is a novel synthetic data-based alignment tuning method for LLMs from IBM Research. Merlinite-7b is a Mistral-7b-derivative model trained with the LAB methodology, using Mixtral-8x7b-Instruct as a teacher model.
LAB consists of three key components:
1. Taxonomy-driven data curation process
2. Large-scale synthetic data generator
3. Two-phased-training with replay buffers
LAB approach allows for adding new knowledge and skills, in an incremental fashion, to an already pre-trained model without suffering from catastrophic forgetting.
Taxonomy is a tree of seed examples that are used to prompt a teacher model to generate synthetic data. Taxonomy allows the data curator or the model designer to easily specify a diverse set of the knowledge-domains and skills that they would like to include in their LLM. At a high level, these can be categorized into three high-level bins - knowledge, foundational skills, and compositional skills. The leaf nodes of the taxonomy are tasks associated with one or more seed examples.
During the synthetic data generation, **unlike previous approaches where seed examples are uniformly drawn from the entire pool (i.e. self-instruct), we use the taxonomy to drive the sampling process**: For each knowledge/skill, we only use the local examples within the leaf node as seeds to prompt the teacher model.
This makes the teacher model better exploit the task distributions defined by the local examples of each node and the diversity in the taxonomy itself ensures the entire generation covers a wide range of tasks, as illustrated below. In turns, this allows for using Mixtral 8x7B as the teacher model for generation while performing very competitively with models such as ORCA-2, WizardLM, and Zephyr Beta that rely on synthetic data generated by much larger and capable models like GPT-4.
For adding new domain-specific knowledge, we provide an external knowledge source (document) and prompt the model to generate questions and answers based on the document.
Foundational skills such as reasoning and compositional skills such as creative writing are generated through in-context learning using the seed examples from the taxonomy.
Additionally, to ensure the data is high-quality and safe, we employ steps to check the questions and answers to ensure that they are grounded and safe. This is done using the same teacher model that generated the data.
Our training consists of two major phases: knowledge tuning and skills tuning.
There are two steps in knowledge tuning where the first step learns simple knowledge (short samples) and the second step learns complicated knowledge (longer samples).
The second step uses replay a replay buffer with data from the first step.
Both foundational skills and compositional skills are learned during the skills tuning phases, where a replay buffer of data from the knowledge phase is used.
Importantly, we use a set of hyper-parameters for training that are very different from standard small-scale supervised fine-training: larger batch size and carefully optimized learning rate and scheduler.
## Model description
- **Language(s):** Primarily English
- **License:** Apache 2.0
- **Base model:** [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1)
- **Teacher Model:** [mistralai/Mixtral-8x7B-Instruct-v0.1](https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1)
## Prompt Template
```python
sys_prompt = "You are an AI language model developed by IBM Research. You are a cautious assistant. You carefully follow instructions. You are helpful and harmless and you follow ethical guidelines and promote positive behavior."
prompt = f'<|system|>\n{sys_prompt}\n<|user|>\n{inputs}\n<|assistant|>\n'
stop_token = '<|endoftext|>'
```
We advise utilizing the system prompt employed during the model's training for optimal inference performance, as there could be performance variations based on the provided instructions.
## Bias, Risks, and Limitations
Merlinite-7b has not been aligned to human preferences, so the model might produce problematic outputs. The model might also maintain the limitations and constraints that arise from the base model.
The model undergoes training on synthetic data, leading to the potential inheritance of both advantages and limitations from the underlying teacher models and data generation methods. The incorporation of safety measures during Merlinite-7b's training process is considered beneficial. However, a nuanced understanding of the associated risks requires detailed studies for more accurate quantification.
In the absence of adequate safeguards and RLHF, there exists a risk of malicious utilization of these models for generating disinformation or harmful content. Caution is urged against complete reliance on a specific language model for crucial decisions or impactful information, as preventing these models from fabricating content is not straightforward. Additionally, it remains uncertain whether smaller models might exhibit increased susceptibility to hallucination in ungrounded generation scenarios due to their reduced sizes and memorization capacities. This aspect is currently an active area of research, and we anticipate more rigorous exploration, comprehension, and mitigations in this domain. |
NexaAIDev/Octopus-v2 | NexaAIDev | "2024-05-21T06:53:11Z" | 8,760 | 838 | transformers | [
"transformers",
"safetensors",
"gemma",
"text-generation",
"function calling",
"on-device language model",
"android",
"conversational",
"en",
"arxiv:2404.19296",
"arxiv:2404.11459",
"arxiv:2404.01744",
"base_model:google/gemma-2b",
"license:cc-by-nc-4.0",
"autotrain_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2024-03-16T17:54:22Z" | ---
license: cc-by-nc-4.0
base_model: google/gemma-2b
model-index:
- name: Octopus-V2-2B
results: []
tags:
- function calling
- on-device language model
- android
inference: false
space: false
spaces: false
language:
- en
---
# Octopus V2: On-device language model for super agent
## Octopus V4 Release
We are excited to announce that Octopus v4 is now available! Octopus-V4-3B, an advanced open-source language model with 3 billion parameters, serves as the master node in Nexa AI's envisioned graph of language models. Tailored specifically for the MMLU benchmark topics, this model efficiently translates user queries into formats that specialized models can effectively process. It excels at directing these queries to the appropriate specialized model, ensuring precise and effective query handling.
check our papers and repos:
- [paper](https://arxiv.org/abs/2404.19296)
- [Octopus V4 model page](https://huggingface.co/NexaAIDev/Octopus-v4)
- [Octopus V4 quantized model page](https://huggingface.co/NexaAIDev/octopus-v4-gguf)
- [Octopus V4 github](https://github.com/NexaAI/octopus-v4)
Key Features of Octopus v4:
- 📱 **Compact Size**: Octopus-V4-3B is compact, enabling it to operate on smart devices efficiently and swiftly.
- 🐙 **Accuracy**: Octopus-V4-3B accurately maps user queries to the specialized model using a functional token design, enhancing its precision.
- 💪 **Reformat Query**: Octopus-V4-3B assists in converting natural human language into a more professional format, improving query description and resulting in more accurate responses.
## Octopus V3 Release
We are excited to announce that Octopus v3 is now available! check our [technical report](https://arxiv.org/abs/2404.11459) and [Octopus V3 tweet](https://twitter.com/nexa4ai/status/1780783383737676236)!
Key Features of Octopus v3:
- **Efficiency**: **Sub-billion** parameters, making it less than half the size of its predecessor, Octopus v2.
- **Multi-Modal Capabilities**: Proceed both text and images inputs.
- **Speed and Accuracy**: Incorporate our **patented** functional token technology, achieving function calling accuracy on par with GPT-4V and GPT-4.
- **Multilingual Support**: Simultaneous support for English and Mandarin.
Check the Octopus V3 demo video for [Android and iOS](https://octopus3.nexa4ai.com/).
## Octopus V2 Release
After open-sourcing our model, we got many requests to compare our model with [Apple's OpenELM](https://huggingface.co/apple/OpenELM-3B-Instruct) and [Microsoft's Phi-3](https://huggingface.co/microsoft/Phi-3-mini-128k-instruct). Please see [Evaluation section](#evaluation). From our benchmark dataset, Microsoft's Phi-3 achieves accuracy of 45.7% and the average inference latency is 10.2s. While Apple's OpenELM fails to generate function call, please see [this screenshot](https://huggingface.co/NexaAIDev/Octopus-v2/blob/main/OpenELM-benchmark.jpeg). Our model, Octopus V2, achieves 99.5% accuracy and the average inference latency is 0.38s.
We are a very small team with many work. Please give us more time to prepare the code, and we will **open source** it. We hope Octopus v2 model will be helpful for you. Let's democratize AI agents for everyone. We've received many requests from car industry, health care, financial system etc. Octopus model is able to be applied to **any function**, and you can start to think about it now.
<p align="center">
- <a href="https://www.nexa4ai.com/" target="_blank">Nexa AI Product</a>
- <a href="https://arxiv.org/abs/2404.01744" target="_blank">ArXiv</a>
- <a href="https://www.youtube.com/watch?v=jhM0D0OObOw&ab_channel=NexaAI" target="_blank">Video Demo</a>
</p>
<p align="center" width="100%">
<a><img src="Octopus-logo.jpeg" alt="nexa-octopus" style="width: 40%; min-width: 300px; display: block; margin: auto;"></a>
</p>
## Introduction
Octopus-V2-2B, an advanced open-source language model with 2 billion parameters, represents Nexa AI's research breakthrough in the application of large language models (LLMs) for function calling, specifically tailored for Android APIs. Unlike Retrieval-Augmented Generation (RAG) methods, which require detailed descriptions of potential function arguments—sometimes needing up to tens of thousands of input tokens—Octopus-V2-2B introduces a unique **functional token** strategy for both its training and inference stages. This approach not only allows it to achieve performance levels comparable to GPT-4 but also significantly enhances its inference speed beyond that of RAG-based methods, making it especially beneficial for edge computing devices.
📱 **On-device Applications**: Octopus-V2-2B is engineered to operate seamlessly on Android devices, extending its utility across a wide range of applications, from Android system management to the orchestration of multiple devices.
🚀 **Inference Speed**: When benchmarked, Octopus-V2-2B demonstrates a remarkable inference speed, outperforming the combination of "Llama7B + RAG solution" by a factor of 36X on a single A100 GPU. Furthermore, compared to GPT-4-turbo (gpt-4-0125-preview), which relies on clusters A100/H100 GPUs, Octopus-V2-2B is 168% faster. This efficiency is attributed to our **functional token** design.
🐙 **Accuracy**: Octopus-V2-2B not only excels in speed but also in accuracy, surpassing the "Llama7B + RAG solution" in function call accuracy by 31%. It achieves a function call accuracy comparable to GPT-4 and RAG + GPT-3.5, with scores ranging between 98% and 100% across benchmark datasets.
💪 **Function Calling Capabilities**: Octopus-V2-2B is capable of generating individual, nested, and parallel function calls across a variety of complex scenarios.
## Example Use Cases
<p align="center" width="100%">
<a><img src="tool-usage-compressed.png" alt="ondevice" style="width: 80%; min-width: 300px; display: block; margin: auto;"></a>
</p>
You can run the model on a GPU using the following code.
```python
from transformers import AutoTokenizer, GemmaForCausalLM
import torch
import time
def inference(input_text):
start_time = time.time()
input_ids = tokenizer(input_text, return_tensors="pt").to(model.device)
input_length = input_ids["input_ids"].shape[1]
outputs = model.generate(
input_ids=input_ids["input_ids"],
max_length=1024,
do_sample=False)
generated_sequence = outputs[:, input_length:].tolist()
res = tokenizer.decode(generated_sequence[0])
end_time = time.time()
return {"output": res, "latency": end_time - start_time}
model_id = "NexaAIDev/Octopus-v2"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = GemmaForCausalLM.from_pretrained(
model_id, torch_dtype=torch.bfloat16, device_map="auto"
)
input_text = "Take a selfie for me with front camera"
nexa_query = f"Below is the query from the users, please call the correct function and generate the parameters to call the function.\n\nQuery: {input_text} \n\nResponse:"
start_time = time.time()
print("nexa model result:\n", inference(nexa_query))
print("latency:", time.time() - start_time," s")
```
## Evaluation
The benchmark result can be viewed in [this excel](android_benchmark.xlsx), which has been manually verified. Microsoft's Phi-3 model achieved an accuracy of 45.7%, with an average inference latency of 10.2 seconds. Meanwhile, Apple's OpenELM was unable to generate a function call, as shown in [this screenshot](https://huggingface.co/NexaAIDev/Octopus-v2/blob/main/OpenELM-benchmark.jpeg). Additionally, OpenELM's score on the MMLU benchmark is quite low at 26.7, compared to Google's Gemma 2B, which scored 42.3.
<p align="center" width="100%">
<a><img src="latency_plot.jpg" alt="ondevice" style="width: 80%; min-width: 300px; display: block; margin: auto; margin-bottom: 20px;"></a>
<a><img src="accuracy_plot.jpg" alt="ondevice" style="width: 80%; min-width: 300px; display: block; margin: auto;"></a>
</p>
**Note**: One can notice that the query includes all necessary parameters used for a function. It is expected that query includes all parameters during inference as well.
## Training Data
We wrote 20 Android API descriptions to used to train the models, see [this file](android_functions.txt) for details. The Android API implementations for our demos, and our training data will be published later. Below is one Android API description example
```
def get_trending_news(category=None, region='US', language='en', max_results=5):
"""
Fetches trending news articles based on category, region, and language.
Parameters:
- category (str, optional): News category to filter by, by default use None for all categories. Optional to provide.
- region (str, optional): ISO 3166-1 alpha-2 country code for region-specific news, by default, uses 'US'. Optional to provide.
- language (str, optional): ISO 639-1 language code for article language, by default uses 'en'. Optional to provide.
- max_results (int, optional): Maximum number of articles to return, by default, uses 5. Optional to provide.
Returns:
- list[str]: A list of strings, each representing an article. Each string contains the article's heading and URL.
"""
```
## License
This model was trained on commercially viable data. For use of our model, refer to the [license information](https://www.nexa4ai.com/licenses).
## References
We thank the Google Gemma team for their amazing models!
```
@misc{gemma-2023-open-models,
author = {{Gemma Team, Google DeepMind}},
title = {Gemma: Open Models Based on Gemini Research and Technology},
url = {https://goo.gle/GemmaReport},
year = {2023},
}
```
## Citation
```
@misc{chen2024octopus,
title={Octopus v2: On-device language model for super agent},
author={Wei Chen and Zhiyuan Li},
year={2024},
eprint={2404.01744},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
## Contact
Please [contact us](mailto:[email protected]) to reach out for any issues and comments! |
RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf | RichardErkhov | "2024-05-03T01:31:26Z" | 8,759 | 0 | null | [
"gguf",
"region:us"
] | null | "2024-05-03T01:30:46Z" | Quantization made by Richard Erkhov.
[Github](https://github.com/RichardErkhov)
[Discord](https://discord.gg/pvy7H8DZMG)
[Request more models](https://github.com/RichardErkhov/quant_request)
tiny-random-GemmaForCausalLM - GGUF
- Model creator: https://huggingface.co/fxmarty/
- Original model: https://huggingface.co/fxmarty/tiny-random-GemmaForCausalLM/
| Name | Quant method | Size |
| ---- | ---- | ---- |
| [tiny-random-GemmaForCausalLM.Q2_K.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q2_K.gguf) | Q2_K | 0.01GB |
| [tiny-random-GemmaForCausalLM.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.IQ3_XS.gguf) | IQ3_XS | 0.01GB |
| [tiny-random-GemmaForCausalLM.IQ3_S.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.IQ3_S.gguf) | IQ3_S | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q3_K_S.gguf) | Q3_K_S | 0.01GB |
| [tiny-random-GemmaForCausalLM.IQ3_M.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.IQ3_M.gguf) | IQ3_M | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q3_K.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q3_K.gguf) | Q3_K | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q3_K_M.gguf) | Q3_K_M | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q3_K_L.gguf) | Q3_K_L | 0.01GB |
| [tiny-random-GemmaForCausalLM.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.IQ4_XS.gguf) | IQ4_XS | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q4_0.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q4_0.gguf) | Q4_0 | 0.01GB |
| [tiny-random-GemmaForCausalLM.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.IQ4_NL.gguf) | IQ4_NL | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q4_K_S.gguf) | Q4_K_S | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q4_K.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q4_K.gguf) | Q4_K | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q4_K_M.gguf) | Q4_K_M | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q4_1.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q4_1.gguf) | Q4_1 | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q5_0.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q5_0.gguf) | Q5_0 | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q5_K_S.gguf) | Q5_K_S | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q5_K.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q5_K.gguf) | Q5_K | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q5_K_M.gguf) | Q5_K_M | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q5_1.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q5_1.gguf) | Q5_1 | 0.01GB |
| [tiny-random-GemmaForCausalLM.Q6_K.gguf](https://huggingface.co/RichardErkhov/fxmarty_-_tiny-random-GemmaForCausalLM-gguf/blob/main/tiny-random-GemmaForCausalLM.Q6_K.gguf) | Q6_K | 0.01GB |
Original model description:
---
license: mit
---
This one with a custom `config.head_dim` as allowed by the architecture (see 7b model).
|
bartowski/Qwen2-72B-Instruct-GGUF | bartowski | "2024-06-07T02:01:20Z" | 8,758 | 7 | null | [
"gguf",
"chat",
"text-generation",
"en",
"license:other",
"region:us"
] | text-generation | "2024-06-07T00:18:25Z" | ---
license: other
license_name: tongyi-qianwen
license_link: https://huggingface.co/Qwen/Qwen2-72B-Instruct/blob/main/LICENSE
language:
- en
pipeline_tag: text-generation
tags:
- chat
quantized_by: bartowski
---
# <b>Heads up:</b> currently CUDA offloading is broken unless you enable flash attention
## Llamacpp imatrix Quantizations of Qwen2-72B-Instruct
Using <a href="https://github.com/ggerganov/llama.cpp/">llama.cpp</a> commit <a href="https://github.com/ggerganov/llama.cpp/commit/ee459f40f65810a810151b24eba5b8bd174ceffe">ee459f40f65810a810151b24eba5b8bd174ceffe</a> for quantization.
Original model: https://huggingface.co/Qwen/Qwen2-72B-Instruct
All quants made using imatrix option with dataset from [here](https://gist.github.com/bartowski1182/eb213dccb3571f863da82e99418f81e8)
## Prompt format
```
<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
```
## Download a file (not the whole branch) from below:
| Filename | Quant type | File Size | Description |
| -------- | ---------- | --------- | ----------- |
| [Qwen2-72B-Instruct-Q8_0.gguf](https://huggingface.co/bartowski/Qwen2-72B-Instruct-GGUF/tree/main/Qwen2-72B-Instruct-Q8_0.gguf) | Q8_0 | 77.26GB | Extremely high quality, generally unneeded but max available quant. |
| [Qwen2-72B-Instruct-Q5_K_M.gguf](https://huggingface.co/bartowski/Qwen2-72B-Instruct-GGUF/tree/main/Qwen2-72B-Instruct-Q5_K_M.gguf) | Q5_K_M | 54.44GB | High quality, *recommended*. |
| [Qwen2-72B-Instruct-Q4_K_M.gguf](https://huggingface.co/bartowski/Qwen2-72B-Instruct-GGUF/blob/main/Qwen2-72B-Instruct-Q4_K_M.gguf) | Q4_K_M | 47.41GB | Good quality, uses about 4.83 bits per weight, *recommended*. |
| [Qwen2-72B-Instruct-IQ4_XS.gguf](https://huggingface.co/bartowski/Qwen2-72B-Instruct-GGUF/blob/main/Qwen2-72B-Instruct-IQ4_XS.gguf) | IQ4_XS | 39.70GB | Decent quality, smaller than Q4_K_S with similar performance, *recommended*. |
| [Qwen2-72B-Instruct-Q3_K_M.gguf](https://huggingface.co/bartowski/Qwen2-72B-Instruct-GGUF/blob/main/Qwen2-72B-Instruct-Q3_K_M.gguf) | Q3_K_M | 37.69GB | Even lower quality. |
| [Qwen2-72B-Instruct-IQ3_M.gguf](https://huggingface.co/bartowski/Qwen2-72B-Instruct-GGUF/blob/main/Qwen2-72B-Instruct-IQ3_M.gguf) | IQ3_M | 35.50GB | Medium-low quality, new method with decent performance comparable to Q3_K_M. |
| [Qwen2-72B-Instruct-Q3_K_S.gguf](https://huggingface.co/bartowski/Qwen2-72B-Instruct-GGUF/blob/main/Qwen2-72B-Instruct-Q3_K_S.gguf) | Q3_K_S | 34.48GB | Low quality, not recommended. |
| [Qwen2-72B-Instruct-IQ3_XXS.gguf](https://huggingface.co/bartowski/Qwen2-72B-Instruct-GGUF/blob/main/Qwen2-72B-Instruct-IQ3_XXS.gguf) | IQ3_XXS | 31.84GB | Lower quality, new method with decent performance, comparable to Q3 quants. |
| [Qwen2-72B-Instruct-Q2_K.gguf](https://huggingface.co/bartowski/Qwen2-72B-Instruct-GGUF/blob/main/Qwen2-72B-Instruct-Q2_K.gguf) | Q2_K | 29.81GB | Very low quality but surprisingly usable. |
| [Qwen2-72B-Instruct-IQ2_M.gguf](https://huggingface.co/bartowski/Qwen2-72B-Instruct-GGUF/blob/main/Qwen2-72B-Instruct-IQ2_M.gguf) | IQ2_M | 29.33GB | Very low quality, uses SOTA techniques to also be surprisingly usable. |
| [Qwen2-72B-Instruct-IQ2_XXS.gguf](https://huggingface.co/bartowski/Qwen2-72B-Instruct-GGUF/blob/main/Qwen2-72B-Instruct-IQ2_XXS.gguf) | IQ2_XXS | 25.49GB | Lower quality, uses SOTA techniques to be usable. |
| [Qwen2-72B-Instruct-IQ1_M.gguf](https://huggingface.co/bartowski/Qwen2-72B-Instruct-GGUF/blob/main/Qwen2-72B-Instruct-IQ1_M.gguf) | IQ1_M | 23.74GB | Extremely low quality, *not* recommended. |
## Downloading using huggingface-cli
First, make sure you have hugginface-cli installed:
```
pip install -U "huggingface_hub[cli]"
```
Then, you can target the specific file you want:
```
huggingface-cli download bartowski/Qwen2-72B-Instruct-GGUF --include "Qwen2-72B-Instruct-Q4_K_M.gguf" --local-dir ./
```
If the model is bigger than 50GB, it will have been split into multiple files. In order to download them all to a local folder, run:
```
huggingface-cli download bartowski/Qwen2-72B-Instruct-GGUF --include "Qwen2-72B-Instruct-Q8_0.gguf/*" --local-dir Qwen2-72B-Instruct-Q8_0
```
You can either specify a new local-dir (Qwen2-72B-Instruct-Q8_0) or download them all in place (./)
## Which file should I choose?
A great write up with charts showing various performances is provided by Artefact2 [here](https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9)
The first thing to figure out is how big a model you can run. To do this, you'll need to figure out how much RAM and/or VRAM you have.
If you want your model running as FAST as possible, you'll want to fit the whole thing on your GPU's VRAM. Aim for a quant with a file size 1-2GB smaller than your GPU's total VRAM.
If you want the absolute maximum quality, add both your system RAM and your GPU's VRAM together, then similarly grab a quant with a file size 1-2GB Smaller than that total.
Next, you'll need to decide if you want to use an 'I-quant' or a 'K-quant'.
If you don't want to think too much, grab one of the K-quants. These are in format 'QX_K_X', like Q5_K_M.
If you want to get more into the weeds, you can check out this extremely useful feature chart:
[llama.cpp feature matrix](https://github.com/ggerganov/llama.cpp/wiki/Feature-matrix)
But basically, if you're aiming for below Q4, and you're running cuBLAS (Nvidia) or rocBLAS (AMD), you should look towards the I-quants. These are in format IQX_X, like IQ3_M. These are newer and offer better performance for their size.
These I-quants can also be used on CPU and Apple Metal, but will be slower than their K-quant equivalent, so speed vs performance is a tradeoff you'll have to decide.
The I-quants are *not* compatible with Vulcan, which is also AMD, so if you have an AMD card double check if you're using the rocBLAS build or the Vulcan build. At the time of writing this, LM Studio has a preview with ROCm support, and other inference engines have specific builds for ROCm.
Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski
|
QuantFactory/Hathor_Fractionate-L3-8B-v.05-GGUF | QuantFactory | "2024-06-29T03:26:29Z" | 8,746 | 0 | null | [
"gguf",
"region:us"
] | null | "2024-06-28T14:31:39Z" | Entry not found |
klue/roberta-large | klue | "2023-06-12T12:29:57Z" | 8,744 | 43 | transformers | [
"transformers",
"pytorch",
"safetensors",
"roberta",
"fill-mask",
"korean",
"klue",
"ko",
"arxiv:2105.09680",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | fill-mask | "2022-03-02T23:29:05Z" | ---
language: ko
tags:
- korean
- klue
mask_token: "[MASK]"
widget:
- text: 대한민국의 수도는 [MASK] 입니다.
---
# KLUE RoBERTa large
Pretrained RoBERTa Model on Korean Language. See [Github](https://github.com/KLUE-benchmark/KLUE) and [Paper](https://arxiv.org/abs/2105.09680) for more details.
## How to use
_NOTE:_ Use `BertTokenizer` instead of RobertaTokenizer. (`AutoTokenizer` will load `BertTokenizer`)
```python
from transformers import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained("klue/roberta-large")
tokenizer = AutoTokenizer.from_pretrained("klue/roberta-large")
```
## BibTeX entry and citation info
```bibtex
@misc{park2021klue,
title={KLUE: Korean Language Understanding Evaluation},
author={Sungjoon Park and Jihyung Moon and Sungdong Kim and Won Ik Cho and Jiyoon Han and Jangwon Park and Chisung Song and Junseong Kim and Yongsook Song and Taehwan Oh and Joohong Lee and Juhyun Oh and Sungwon Lyu and Younghoon Jeong and Inkwon Lee and Sangwoo Seo and Dongjun Lee and Hyunwoo Kim and Myeonghwa Lee and Seongbo Jang and Seungwon Do and Sunkyoung Kim and Kyungtae Lim and Jongwon Lee and Kyumin Park and Jamin Shin and Seonghyun Kim and Lucy Park and Alice Oh and Jungwoo Ha and Kyunghyun Cho},
year={2021},
eprint={2105.09680},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
|
dima806/man_woman_face_image_detection | dima806 | "2024-03-09T19:00:41Z" | 8,744 | 3 | transformers | [
"transformers",
"pytorch",
"safetensors",
"vit",
"image-classification",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | image-classification | "2023-10-03T20:28:32Z" | ---
license: apache-2.0
metrics:
- accuracy
- f1
---
Returns whether the face belongs to man or woman based on face image.
See https://www.kaggle.com/code/dima806/man-woman-face-image-detection-vit for more details.
```
Classification report:
precision recall f1-score support
man 0.9898 0.9908 0.9903 7071
woman 0.9908 0.9898 0.9903 7072
accuracy 0.9903 14143
macro avg 0.9903 0.9903 0.9903 14143
weighted avg 0.9903 0.9903 0.9903 14143
``` |
cyberagent/calm2-7b-chat | cyberagent | "2024-02-17T04:30:14Z" | 8,744 | 73 | transformers | [
"transformers",
"pytorch",
"safetensors",
"llama",
"text-generation",
"japanese",
"causal-lm",
"ja",
"en",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2023-11-01T07:40:30Z" | ---
license: apache-2.0
language:
- ja
- en
tags:
- japanese
- causal-lm
inference: false
---
# CyberAgentLM2-7B-Chat (CALM2-7B-Chat)
## Model Description
CyberAgentLM2-Chat is a fine-tuned model of [CyberAgentLM2](https://huggingface.co/cyberagent/calm2-7b) for dialogue use cases.
## Requirements
- transformers >= 4.34.1
- accelerate
## Usage
```python
import transformers
from transformers import AutoModelForCausalLM, AutoTokenizer, TextStreamer
assert transformers.__version__ >= "4.34.1"
model = AutoModelForCausalLM.from_pretrained("cyberagent/calm2-7b-chat", device_map="auto", torch_dtype="auto")
tokenizer = AutoTokenizer.from_pretrained("cyberagent/calm2-7b-chat")
streamer = TextStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
prompt = """USER: AIによって私達の暮らしはどのように変わりますか？
ASSISTANT: """
token_ids = tokenizer.encode(prompt, return_tensors="pt")
output_ids = model.generate(
input_ids=token_ids.to(model.device),
max_new_tokens=300,
do_sample=True,
temperature=0.8,
streamer=streamer,
)
```
## Chat Template
```
USER: {user_message1}
ASSISTANT: {assistant_message1}<|endoftext|>
USER: {user_message2}
ASSISTANT: {assistant_message2}<|endoftext|>
USER: {user_message3}
ASSISTANT: {assistant_message3}<|endoftext|>
```
## Model Details
* **Model size**: 7B
* **Context length**: 32768
* **Model type**: Transformer-based Language Model
* **Language(s)**: Japanese, English
* **Developed by**: [CyberAgent, Inc.](https://www.cyberagent.co.jp/)
* **License**: Apache-2.0
## Author
[Ryosuke Ishigami](https://huggingface.co/rishigami)
## Citations
```tex
@article{touvron2023llama,
title={LLaMA: Open and Efficient Foundation Language Models},
author={Touvron, Hugo and Lavril, Thibaut and Izacard, Gautier and Martinet, Xavier and Lachaux, Marie-Anne and Lacroix, Timoth{\'e}e and Rozi{\`e}re, Baptiste and Goyal, Naman and Hambro, Eric and Azhar, Faisal and Rodriguez, Aurelien and Joulin, Armand and Grave, Edouard and Lample, Guillaume},
journal={arXiv preprint arXiv:2302.13971},
year={2023}
}
``` |
second-state/llm-compiler-7b-ftd-GGUF | second-state | "2024-06-29T10:56:30Z" | 8,744 | 0 | transformers | [
"transformers",
"gguf",
"llama",
"text-generation",
"code",
"base_model:facebook/llm-compiler-7b-ftd",
"license:other",
"autotrain_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2024-06-29T10:22:03Z" | ---
language:
- code
license: other
model_name: llm-compiler-7b
base_model: facebook/llm-compiler-7b-ftd
inference: false
model_creator: facebook
quantized_by: Second State Inc.
---
<!-- header start -->
<!-- 200823 -->
<div style="width: auto; margin-left: auto; margin-right: auto">
<img src="https://github.com/LlamaEdge/LlamaEdge/raw/dev/assets/logo.svg" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>
<hr style="margin-top: 1.0em; margin-bottom: 1.0em;">
<!-- header end -->
# llm-compiler-7b-ftd-GGUF
## Original Model
[facebook/llm-compiler-7b-ftd](https://huggingface.co/facebook/llm-compiler-7b-ftd)
**License** A custom commercial license is available at: [https://ai.meta.com/resources/models-and-libraries/llama-downloads/](https://ai.meta.com/resources/models-and-libraries/llama-downloads/)
## Run with LlamaEdge
<!-- - LlamaEdge version: [v0.2.8](https://github.com/LlamaEdge/LlamaEdge/releases/tag/0.2.8) and above -->
- LlamaEdge version: coming soon
<!-- - Prompt template
- Prompt type: `codellama-instruct`
- Prompt string
```text
<s>[INST] <<SYS>>
Write code to solve the following coding problem that obeys the constraints and passes the example test cases. Please wrap your code answer using ```: <</SYS>>
{prompt} [/INST]
```
- Context size: `4096`
- Run as LlamaEdge command app
```bash
wasmedge --dir .:. --nn-preload default:GGML:AUTO:llm-compiler-7b-ftd-Q5_K_M.gguf llama-chat.wasm -p codellama-instruct
``` -->
## Quantized GGUF Models
| Name | Quant method | Bits | Size | Use case |
| ---- | ---- | ---- | ---- | ----- |
| [llm-compiler-7b-ftd-Q2_K.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-Q2_K.gguf) | Q2_K | 2 | 2.53 GB| smallest, significant quality loss - not recommended for most purposes |
| [llm-compiler-7b-ftd-Q3_K_L.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-Q3_K_L.gguf) | Q3_K_L | 3 | 3.6 GB| small, substantial quality loss |
| [llm-compiler-7b-ftd-Q3_K_M.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-Q3_K_M.gguf) | Q3_K_M | 3 | 3.3 GB| very small, high quality loss |
| [llm-compiler-7b-ftd-Q3_K_S.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-Q3_K_S.gguf) | Q3_K_S | 3 | 2.95 GB| very small, high quality loss |
| [llm-compiler-7b-ftd-Q4_0.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-Q4_0.gguf) | Q4_0 | 4 | 3.83 GB| legacy; small, very high quality loss - prefer using Q3_K_M |
| [llm-compiler-7b-ftd-Q4_K_M.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-Q4_K_M.gguf) | Q4_K_M | 4 | 4.08 GB| medium, balanced quality - recommended |
| [llm-compiler-7b-ftd-Q4_K_S.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-Q4_K_S.gguf) | Q4_K_S | 4 | 3.86 GB| small, greater quality loss |
| [llm-compiler-7b-ftd-Q5_0.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-Q5_0.gguf) | Q5_0 | 5 | 4.65 GB| legacy; medium, balanced quality - prefer using Q4_K_M |
| [llm-compiler-7b-ftd-Q5_K_M.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-Q5_K_M.gguf) | Q5_K_M | 5 | 4.78 GB| large, very low quality loss - recommended |
| [llm-compiler-7b-ftd-Q5_K_S.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-Q5_K_S.gguf) | Q5_K_S | 5 | 4.65 GB| large, low quality loss - recommended |
| [llm-compiler-7b-ftd-Q6_K.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-Q6_K.gguf) | Q6_K | 6 | 5.53 GB| very large, extremely low quality loss |
| [llm-compiler-7b-ftd-Q8_0.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-Q8_0.gguf) | Q8_0 | 8 | 7.16 GB| very large, extremely low quality loss - not recommended |
| [llm-compiler-7b-ftd-f16.gguf](https://huggingface.co/second-state/llm-compiler-7b-ftd-GGUF/blob/main/llm-compiler-7b-ftd-f16.gguf) | f16 | f16 | 13.5 GB| |
*Quantized with llama.cpp b3259*
|
lmms-lab/llavanext-qwen-tokenizer | lmms-lab | "2024-05-05T13:00:38Z" | 8,740 | 1 | transformers | [
"transformers",
"llava",
"text-generation",
"conversational",
"arxiv:1910.09700",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-generation | "2024-05-05T11:00:53Z" | ---
library_name: transformers
tags: []
---
# Model Card for Model ID
<!-- Provide a quick summary of what the model is/does. -->
## Model Details
### Model Description
<!-- Provide a longer summary of what this model is. -->
This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
- **Developed by:** [More Information Needed]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [More Information Needed]
- **Model type:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
- **Finetuned from model [optional]:** [More Information Needed]
### Model Sources [optional]
<!-- Provide the basic links for the model. -->
- **Repository:** [More Information Needed]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
### Direct Use
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
[More Information Needed]
### Downstream Use [optional]
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
## How to Get Started with the Model
Use the code below to get started with the model.
[More Information Needed]
## Training Details
### Training Data
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
[More Information Needed]
### Training Procedure
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
#### Preprocessing [optional]
[More Information Needed]
#### Training Hyperparameters
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
#### Speeds, Sizes, Times [optional]
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
[More Information Needed]
## Evaluation
<!-- This section describes the evaluation protocols and provides the results. -->
### Testing Data, Factors & Metrics
#### Testing Data
<!-- This should link to a Dataset Card if possible. -->
[More Information Needed]
#### Factors
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
[More Information Needed]
#### Metrics
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
[More Information Needed]
### Results
[More Information Needed]
#### Summary
## Model Examination [optional]
<!-- Relevant interpretability work for the model goes here -->
[More Information Needed]
## Environmental Impact
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
- **Hardware Type:** [More Information Needed]
- **Hours used:** [More Information Needed]
- **Cloud Provider:** [More Information Needed]
- **Compute Region:** [More Information Needed]
- **Carbon Emitted:** [More Information Needed]
## Technical Specifications [optional]
### Model Architecture and Objective
[More Information Needed]
### Compute Infrastructure
[More Information Needed]
#### Hardware
[More Information Needed]
#### Software
[More Information Needed]
## Citation [optional]
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Model Card Authors [optional]
[More Information Needed]
## Model Card Contact
[More Information Needed] |
TencentARC/t2i-adapter-canny-sdxl-1.0 | TencentARC | "2023-09-07T19:10:05Z" | 8,724 | 43 | diffusers | [
"diffusers",
"safetensors",
"art",
"t2i-adapter",
"image-to-image",
"stable-diffusion-xl-diffusers",
"stable-diffusion-xl",
"arxiv:2302.08453",
"base_model:stabilityai/stable-diffusion-xl-base-1.0",
"license:apache-2.0",
"region:us"
] | image-to-image | "2023-09-03T14:19:29Z" | ---
license: apache-2.0
base_model: stabilityai/stable-diffusion-xl-base-1.0
tags:
- art
- t2i-adapter
- image-to-image
- stable-diffusion-xl-diffusers
- stable-diffusion-xl
---
# T2I-Adapter-SDXL - Canny
T2I Adapter is a network providing additional conditioning to stable diffusion. Each t2i checkpoint takes a different type of conditioning as input and is used with a specific base stable diffusion checkpoint.
This checkpoint provides conditioning on canny for the StableDiffusionXL checkpoint. This was a collaboration between **Tencent ARC** and [**Hugging Face**](https://huggingface.co/).
## Model Details
- **Developed by:** T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models
- **Model type:** Diffusion-based text-to-image generation model
- **Language(s):** English
- **License:** Apache 2.0
- **Resources for more information:** [GitHub Repository](https://github.com/TencentARC/T2I-Adapter), [Paper](https://arxiv.org/abs/2302.08453).
- **Model complexity:**
| | SD-V1.4/1.5 | SD-XL | T2I-Adapter | T2I-Adapter-SDXL |
| --- | --- |--- |--- |--- |
| Parameters | 860M | 2.6B |77 M | 77/79 M | |
- **Cite as:**
@misc{
title={T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models},
author={Chong Mou, Xintao Wang, Liangbin Xie, Yanze Wu, Jian Zhang, Zhongang Qi, Ying Shan, Xiaohu Qie},
year={2023},
eprint={2302.08453},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
### Checkpoints
| Model Name | Control Image Overview| Control Image Example | Generated Image Example |
|---|---|---|---|
|[TencentARC/t2i-adapter-canny-sdxl-1.0](https://huggingface.co/TencentARC/t2i-adapter-canny-sdxl-1.0)<br/> *Trained with canny edge detection* | A monochrome image with white edges on a black background.|<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_canny.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_canny.png"/></a>|<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/res_canny.png"><img width="64" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/res_canny.png"/></a>|
|[TencentARC/t2i-adapter-sketch-sdxl-1.0](https://huggingface.co/TencentARC/t2i-adapter-sketch-sdxl-1.0)<br/> *Trained with [PidiNet](https://github.com/zhuoinoulu/pidinet) edge detection* | A hand-drawn monochrome image with white outlines on a black background.|<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_sketch.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_sketch.png"/></a>|<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/res_sketch.png"><img width="64" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/res_sketch.png"/></a>|
|[TencentARC/t2i-adapter-lineart-sdxl-1.0](https://huggingface.co/TencentARC/t2i-adapter-lineart-sdxl-1.0)<br/> *Trained with lineart edge detection* | A hand-drawn monochrome image with white outlines on a black background.|<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_lin.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_lin.png"/></a>|<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/res_lin.png"><img width="64" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/res_lin.png"/></a>|
|[TencentARC/t2i-adapter-depth-midas-sdxl-1.0](https://huggingface.co/TencentARC/t2i-adapter-depth-midas-sdxl-1.0)<br/> *Trained with Midas depth estimation* | A grayscale image with black representing deep areas and white representing shallow areas.|<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_depth_mid.png"><img width="64" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_depth_mid.png"/></a>|<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/res_depth_mid.png"><img width="64" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/res_depth_mid.png"/></a>|
|[TencentARC/t2i-adapter-depth-zoe-sdxl-1.0](https://huggingface.co/TencentARC/t2i-adapter-depth-zoe-sdxl-1.0)<br/> *Trained with Zoe depth estimation* | A grayscale image with black representing deep areas and white representing shallow areas.|<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_depth_zeo.png"><img width="64" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_depth_zeo.png"/></a>|<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/res_depth_zeo.png"><img width="64" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/res_depth_zeo.png"/></a>|
|[TencentARC/t2i-adapter-openpose-sdxl-1.0](https://huggingface.co/TencentARC/t2i-adapter-openpose-sdxl-1.0)<br/> *Trained with OpenPose bone image* | A [OpenPose bone](https://github.com/CMU-Perceptual-Computing-Lab/openpose) image.|<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/openpose.png"><img width="64" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/openpose.png"/></a>|<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/res_pose.png"><img width="64" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/res_pose.png"/></a>|
## Example
To get started, first install the required dependencies:
```bash
pip install -U git+https://github.com/huggingface/diffusers.git
pip install -U controlnet_aux==0.0.7 # for conditioning models and detectors
pip install transformers accelerate safetensors
```
1. Images are first downloaded into the appropriate *control image* format.
2. The *control image* and *prompt* are passed to the [`StableDiffusionXLAdapterPipeline`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/pipelines/t2i_adapter/pipeline_stable_diffusion_xl_adapter.py#L125).
Let's have a look at a simple example using the [Canny Adapter](https://huggingface.co/Adapter/t2iadapter_canny_sdxlv1).
- Dependency
```py
from diffusers import StableDiffusionXLAdapterPipeline, T2IAdapter, EulerAncestralDiscreteScheduler, AutoencoderKL
from diffusers.utils import load_image, make_image_grid
from controlnet_aux.canny import CannyDetector
import torch
# load adapter
adapter = T2IAdapter.from_pretrained("TencentARC/t2i-adapter-canny-sdxl-1.0", torch_dtype=torch.float16, varient="fp16").to("cuda")
# load euler_a scheduler
model_id = 'stabilityai/stable-diffusion-xl-base-1.0'
euler_a = EulerAncestralDiscreteScheduler.from_pretrained(model_id, subfolder="scheduler")
vae=AutoencoderKL.from_pretrained("madebyollin/sdxl-vae-fp16-fix", torch_dtype=torch.float16)
pipe = StableDiffusionXLAdapterPipeline.from_pretrained(
model_id, vae=vae, adapter=adapter, scheduler=euler_a, torch_dtype=torch.float16, variant="fp16",
).to("cuda")
pipe.enable_xformers_memory_efficient_attention()
canny_detector = CannyDetector()
```
- Condition Image
```py
url = "https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/org_canny.jpg"
image = load_image(url)
# Detect the canny map in low resolution to avoid high-frequency details
image = canny_detector(image, detect_resolution=384, image_resolution=1024)#.resize((1024, 1024))
```
<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_canny.png"><img width="480" style="margin:0;padding:0;" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_canny.png"/></a>
- Generation
```py
prompt = "Mystical fairy in real, magic, 4k picture, high quality"
negative_prompt = "extra digit, fewer digits, cropped, worst quality, low quality, glitch, deformed, mutated, ugly, disfigured"
gen_images = pipe(
prompt=prompt,
negative_prompt=negative_prompt,
image=image,
num_inference_steps=30,
guidance_scale=7.5,
adapter_conditioning_scale=0.8,
adapter_conditioning_factor=1
).images[0]
gen_images.save('out_canny.png')
```
<a href="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/cond_canny.png"><img width="480" style="margin:0;padding:0;" src="https://huggingface.co/Adapter/t2iadapter/resolve/main/figs_SDXLV1.0/res_canny.png"/></a>
### Training
Our training script was built on top of the official training script that we provide [here](https://github.com/huggingface/diffusers/blob/main/examples/t2i_adapter/README_sdxl.md).
The model is trained on 3M high-resolution image-text pairs from LAION-Aesthetics V2 with
- Training steps: 20000
- Batch size: Data parallel with a single gpu batch size of `16` for a total batch size of `256`.
- Learning rate: Constant learning rate of `1e-5`.
- Mixed precision: fp16
|
IlyaGusev/saiga_llama3_8b | IlyaGusev | "2024-07-02T22:54:26Z" | 8,721 | 83 | transformers | [
"transformers",
"safetensors",
"llama",
"text-generation",
"conversational",
"ru",
"dataset:IlyaGusev/saiga_scored",
"doi:10.57967/hf/2368",
"license:other",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2024-04-18T18:25:25Z" | ---
language:
- ru
datasets:
- IlyaGusev/saiga_scored
license: other
license_name: llama3
license_link: https://llama.meta.com/llama3/license/
---
# Saiga/Llama3 8B, Russian Llama-3-based chatbot
Based on [Llama-3 8B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct).
Llama.cpp version: [link](https://huggingface.co/IlyaGusev/saiga_llama3_8b_gguf)
Colab: [link](https://colab.research.google.com/drive/1qxgIPymzW6_H6s_wwXu3lknkkYM45Db4)
## Prompt format
**ОСТОРОЖНО! WARNING! LET OP!**
I've changed the prompt format from ChatML to **the original Llama-3 format in v4**. Don't forget to switch formats!
**v4, v5, v6+**: LLama-3 prompt format:
```
<|begin_of_text|><|start_header_id|>system<|end_header_id|>
Ты — Сайга, русскоязычный автоматический ассистент. Ты разговариваешь с людьми и помогаешь им.<|eot_id|><|start_header_id|>user<|end_header_id|>
Как дела?<|eot_id|><|start_header_id|>assistant<|end_header_id|>
Отлично, а у тебя?<|eot_id|><|start_header_id|>user<|end_header_id|>
Шикарно. Как пройти в библиотеку?<|eot_id|><|start_header_id|>assistant<|end_header_id|>
```
**v2, v3**: ChatML prompt format:
```
<|im_start|>system
Ты — Сайга, русскоязычный автоматический ассистент. Ты разговариваешь с людьми и помогаешь им.<|im_end|>
<|im_start|>user
Как дела?<|im_end|>
<|im_start|>assistant
Отлично, а у тебя?<|im_end|>
<|im_start|>user
Шикарно. Как пройти в библиотеку?<|im_end|>
<|im_start|>assistant
```
## Code example
```python
# Исключительно ознакомительный пример.
# НЕ НАДО ТАК ИНФЕРИТЬ МОДЕЛЬ В ПРОДЕ.
# См. https://github.com/vllm-project/vllm или https://github.com/huggingface/text-generation-inference
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, GenerationConfig
MODEL_NAME = "IlyaGusev/saiga_llama3_8b"
DEFAULT_SYSTEM_PROMPT = "Ты — Сайга, русскоязычный автоматический ассистент. Ты разговариваешь с людьми и помогаешь им."
model = AutoModelForCausalLM.from_pretrained(
MODEL_NAME,
load_in_8bit=True,
torch_dtype=torch.bfloat16,
device_map="auto"
)
model.eval()
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
generation_config = GenerationConfig.from_pretrained(MODEL_NAME)
print(generation_config)
inputs = ["Почему трава зеленая?", "Сочини длинный рассказ, обязательно упоминая следующие объекты. Дано: Таня, мяч"]
for query in inputs:
prompt = tokenizer.apply_chat_template([{
"role": "system",
"content": DEFAULT_SYSTEM_PROMPT
}, {
"role": "user",
"content": query
}], tokenize=False, add_generation_prompt=True)
data = tokenizer(prompt, return_tensors="pt", add_special_tokens=False)
data = {k: v.to(model.device) for k, v in data.items()}
output_ids = model.generate(**data, generation_config=generation_config)[0]
output_ids = output_ids[len(data["input_ids"][0]):]
output = tokenizer.decode(output_ids, skip_special_tokens=True).strip()
print(query)
print(output)
print()
print("==============================")
print()
```
## Output examples
```
User: Почему трава зеленая?
Saiga: Трава выглядит зеленым из-за наличия в ней хлорофилла – пигмента, который помогает растениям поглощать солнечную энергию для процесса фотосинтеза.
Хлорофилл поглощает световые лучи с длиной волны 400-500 нанометров (ближе к синему) и отражает лучи с длиной волны 600-700 нанометров (ближе к красному), что делает его видимым как зеленый цвет.
Таким образом, зеленый цвет является результатом того, как хлорофилл взаимодействует со светом, обеспечивая растения необходимой энергии для роста и развития.
```
```
User: Сочини длинный рассказ, обязательно упоминая следующие объекты. Дано: Таня, мяч
Saiga: В небольшой деревушке под открытым небом жила девочка по имени Таня. Она была очень активной и любящей играть в различные виды спорта. Однажды, когда она прогуливалась по полям, увидела на земле лежащий мяч. Это был старый, но все еще пригодный для игры мячик.
Таня сразу же решила взять его себе домой и показать своему другу Мише. Мишка тоже обожал футбол и всегда рад был бы провести время вместе со своими друзьями. Когда они вернулись домой, Таня показала ему найденный мячик и предложила сыграть в футбол.
Мишка был счастлив, что у них появился новый мяч, и они сразу же начали готовиться к матчу. Они построили две команды и начали игру. Играли они до вечера, а потом решили сделать перерыв и приготовить ужин.
После ужина они продолжили игру, теперь уже с большим энтузиазмом и страстью. В конце концов, победителем стала команда Тани, которая забила больше голов. Все закончилось веселым праздником и хорошими воспоминаниями о проведенном времени.
На следующий день Таня и Миша решили пригласить своих других друзей на игру. Они собрали всех в одном месте и провели ещё один захватывающий матч. Этот день также стал незабываемым благодаря новому мячу, который принес так много радости и улыбок.
Итак, этот маленький мяч, который Таня нашла на поле, стал не просто предметом для игр, но и символом дружбы и веселья между детьми. Он помог им создать незабываемые моменты и укрепил их отношения.
```
## Versions
v6:
- [b662833f247ca04f1843b356e7ff3ee4aef8086a](https://huggingface.co/IlyaGusev/saiga_llama3_8b/commit/b662833f247ca04f1843b356e7ff3ee4aef8086a)
- Other name: saiga_llama3_8b_sft_m10_d1_kto_m2_d2
- SFT dataset config: [sft_d1.json](https://github.com/IlyaGusev/saiga/blob/main/configs/datasets/sft_d1.json)
- SFT model config: [saiga_llama3_8b_sft_m10.json](https://github.com/IlyaGusev/saiga/blob/main/configs/models/saiga_llama3_8b_sft_m10.json)
- KTO dataset config: [pref_d2.json](https://github.com/IlyaGusev/saiga/blob/main/configs/datasets/pref_d2.json)
- KTO model config: [saiga_llama3_8b_kto_m2.json](https://github.com/IlyaGusev/saiga/blob/main/configs/models/saiga_llama3_8b_kto_m2.json)
- SFT wandb: [link](https://wandb.ai/ilyagusev/rulm_self_instruct/runs/0iepauzu)
- KTO wandb: [linl](https://wandb.ai/ilyagusev/rulm_self_instruct/runs/s6l98eot)
v5:
- [d947b00c56683cd4b2f7ce707edef89318027be4](https://huggingface.co/IlyaGusev/saiga_llama3_8b/commit/d947b00c56683cd4b2f7ce707edef89318027be4)
- KTO-tune over v4, dataset: [lmsys_clean_ru_preferences](https://huggingface.co/datasets/IlyaGusev/lmsys_clean_ru_preferences)
- wandb [link](https://wandb.ai/ilyagusev/rulm_self_instruct/runs/se1mbx7n)
v4:
- [1cc945d4ca2c7901cf989e7edaac52ab24f1a7dd](https://huggingface.co/IlyaGusev/saiga_llama3_8b/commit/1cc945d4ca2c7901cf989e7edaac52ab24f1a7dd)
- dataset: [saiga_scored](https://huggingface.co/datasets/IlyaGusev/saiga_scored), scores >= 8, c66032920556c0f21bbbed05e7e04433ec954c3d
- wandb [link](https://wandb.ai/ilyagusev/rulm_self_instruct/runs/dcbs9ttt)
v3:
- [c588356cd60bdee54d52c2dd5a2445acca8aa5c3](https://huggingface.co/IlyaGusev/saiga_llama3_8b/commit/c588356cd60bdee54d52c2dd5a2445acca8aa5c3)
- dataset: [saiga_scored](https://huggingface.co/datasets/IlyaGusev/saiga_scored), scores >= 8, d51cf8060bdc90023da8cf1c3f113f9193d6569b
- wandb [link](https://wandb.ai/ilyagusev/rulm_self_instruct/runs/ltoqdsal)
v2:
- [ae61b4f9b34fac9856d361ea78c66284a00e4f0b](https://huggingface.co/IlyaGusev/saiga_llama3_8b/commit/ae61b4f9b34fac9856d361ea78c66284a00e4f0b)
- dataset code revision d0d123dd221e10bb2a3383bcb1c6e4efe1b4a28a
- wandb [link](https://wandb.ai/ilyagusev/huggingface/runs/r6u5juyk)
- 5 datasets: ru_turbo_saiga, ru_sharegpt_cleaned, oasst1_ru_main_branch, gpt_roleplay_realm, ru_instruct_gpt4
- Datasets merging script: [create_short_chat_set.py](https://github.com/IlyaGusev/rulm/blob/d0d123dd221e10bb2a3383bcb1c6e4efe1b4a28a/self_instruct/src/data_processing/create_short_chat_set.py)
## Evaluation
* Dataset: https://github.com/IlyaGusev/rulm/blob/master/self_instruct/data/tasks.jsonl
* Framework: https://github.com/tatsu-lab/alpaca_eval
* Evaluator: alpaca_eval_cot_gpt4_turbo_fn
| model | length_controlled_winrate | win_rate | standard_error | avg_length |
|-----|-----|-----|-----|-----|
|chatgpt_4_turbo | 76.04 | 90.00 |1.46 | 1270 |
|chatgpt_3_5_turbo | 50.00 | 50.00 | 0.00 | 536 |
|saiga_llama3_8b, v6 | 49.33 | 68.31 | 2.26 | 1262 |
|sfr-iter-dpo | 49.11 | 74.94 | 2.13 | 1215 |
|suzume | 49.05 | 71.57 | 2.20 | 1325 |
|saiga_llama3_8b, v5 | 47.13 | 66.18 | 2.31 | 1194 |
|saiga_llama3_8b, v4 | 43.64 | 65.90 | 2.31 | 1200 |
|saiga_llama3_8b, v3 | 36.97 | 61.08 | 2.38 | 1162 |
|saiga_llama3_8b, v2 | 33.07 | 48.19 | 2.45 | 1166 |
|saiga_mistral_7b | 23.38 | 35.99 | 2.34 | 949 | |
MSLars/erlesen-13b-30K | MSLars | "2024-03-18T08:49:56Z" | 8,715 | 0 | transformers | [
"transformers",
"safetensors",
"llama",
"text-generation",
"custom_code",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2024-03-18T08:40:40Z" | Entry not found |
cointegrated/rubert-tiny-toxicity | cointegrated | "2024-05-16T15:05:52Z" | 8,708 | 29 | transformers | [
"transformers",
"pytorch",
"safetensors",
"bert",
"text-classification",
"russian",
"classification",
"toxicity",
"multilabel",
"ru",
"arxiv:2103.05345",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | "2022-03-02T23:29:05Z" | ---
language: ["ru"]
tags:
- russian
- classification
- toxicity
- multilabel
widget:
- text: "Иди ты нафиг!"
---
This is the [cointegrated/rubert-tiny](https://huggingface.co/cointegrated/rubert-tiny) model fine-tuned for classification of toxicity and inappropriateness for short informal Russian texts, such as comments in social networks.
The problem is formulated as multilabel classification with the following classes:
- `non-toxic`: the text does NOT contain insults, obscenities, and threats, in the sense of the [OK ML Cup](https://cups.mail.ru/ru/tasks/1048) competition.
- `insult`
- `obscenity`
- `threat`
- `dangerous`: the text is inappropriate, in the sense of [Babakov et.al.](https://arxiv.org/abs/2103.05345), i.e. it can harm the reputation of the speaker.
A text can be considered safe if it is BOTH `non-toxic` and NOT `dangerous`.
## Usage
The function below estimates the probability that the text is either toxic OR dangerous:
```python
# !pip install transformers sentencepiece --quiet
import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification
model_checkpoint = 'cointegrated/rubert-tiny-toxicity'
tokenizer = AutoTokenizer.from_pretrained(model_checkpoint)
model = AutoModelForSequenceClassification.from_pretrained(model_checkpoint)
if torch.cuda.is_available():
model.cuda()
def text2toxicity(text, aggregate=True):
""" Calculate toxicity of a text (if aggregate=True) or a vector of toxicity aspects (if aggregate=False)"""
with torch.no_grad():
inputs = tokenizer(text, return_tensors='pt', truncation=True, padding=True).to(model.device)
proba = torch.sigmoid(model(**inputs).logits).cpu().numpy()
if isinstance(text, str):
proba = proba[0]
if aggregate:
return 1 - proba.T[0] * (1 - proba.T[-1])
return proba
print(text2toxicity('я люблю нигеров', True))
# 0.9350118728093193
print(text2toxicity('я люблю нигеров', False))
# [0.9715758 0.0180863 0.0045551 0.00189755 0.9331106 ]
print(text2toxicity(['я люблю нигеров', 'я люблю африканцев'], True))
# [0.93501186 0.04156357]
print(text2toxicity(['я люблю нигеров', 'я люблю африканцев'], False))
# [[9.7157580e-01 1.8086294e-02 4.5550885e-03 1.8975559e-03 9.3311059e-01]
# [9.9979788e-01 1.9048342e-04 1.5297388e-04 1.7452303e-04 4.1369814e-02]]
```
## Training
The model has been trained on the joint dataset of [OK ML Cup](https://cups.mail.ru/ru/tasks/1048) and [Babakov et.al.](https://arxiv.org/abs/2103.05345) with `Adam` optimizer, the learning rate of `1e-5`, and batch size of `64` for `15` epochs in [this Colab notebook](https://colab.research.google.com/drive/1o7azO7cHttwofkp8eTZo9LIybYaNWei_?usp=sharing).
A text was considered inappropriate if its inappropriateness score was higher than 0.8, and appropriate - if it was lower than 0.2. The per-label ROC AUC on the dev set is:
```
non-toxic : 0.9937
insult : 0.9912
obscenity : 0.9881
threat : 0.9910
dangerous : 0.8295
``` |
apple/DFN5B-CLIP-ViT-H-14 | apple | "2023-10-31T18:04:58Z" | 8,708 | 25 | open_clip | [
"open_clip",
"pytorch",
"clip",
"arxiv:2309.17425",
"license:other",
"region:us"
] | null | "2023-10-30T23:07:49Z" | ---
license: other
license_name: apple-sample-code-license
license_link: LICENSE
---
A CLIP (Contrastive Language-Image Pre-training) model trained on DFN-5B.
Data Filtering Networks (DFNs) are small networks used to automatically filter large pools of uncurated data.
This model was trained on 5B images that were filtered from a pool of 43B uncurated image-text pairs
(12.8B image-text pairs from CommonPool-12.8B + 30B additional public image-text pairs).
This model has been converted to PyTorch from the original JAX checkpoints from Axlearn (https://github.com/apple/axlearn).
These weights are directly usable in OpenCLIP (image + text).
## Model Details
- **Model Type:** Contrastive Image-Text, Zero-Shot Image Classification.
- **Dataset:** DFN-5b
- **Papers:**
- Data Filtering Networks: https://arxiv.org/abs/2309.17425
- **Samples Seen:** 39B
## Model Metrics
| Eval Dataset | Metric |
|:-----------------------|---------:|
| ImageNet 1k | 0.8344 |
| Caltech-101 | 0.954935 |
| CIFAR-10 | 0.9878 |
| CIFAR-100 | 0.9051 |
| CLEVR Counts | 0.2966 |
| CLEVR Distance | 0.2124 |
| Country211 | 0.343981 |
| Describable Textures | 0.706383 |
| EuroSAT | 0.654815 |
| FGVC Aircraft | 0.714055 |
| Food-101 | 0.956792 |
| GTSRB | 0.677514 |
| ImageNet Sketch | 0.727308 |
| ImageNet v2 | 0.773 |
| ImageNet-A | 0.6988 |
| ImageNet-O | 0.381 |
| ImageNet-R | 0.929367 |
| KITTI Vehicle Distance | 0.336146 |
| MNIST | 0.8579 |
| ObjectNet | 0.765156 |
| Oxford Flowers-102 | 0.899534 |
| Oxford-IIIT Pet | 0.965515 |
| Pascal VOC 2007 | 0.818309 |
| PatchCamelyon | 0.653625 |
| Rendered SST2 | 0.546403 |
| RESISC45 | 0.750476 |
| Stanford Cars | 0.957592 |
| STL-10 | 0.989 |
| SUN397 | 0.769149 |
| SVHN | 0.676168 |
| Flickr | 0.8645 |
| MSCOCO | 0.631112 |
| WinoGAViL | 0.556329 |
| iWildCam | 0.205549 |
| Camelyon17 | 0.705034 |
| FMoW | 0.207482 |
| Dollar Street | 0.699766 |
| GeoDE | 0.928184 |
| **Average** | **0.698347** |
## Model Usage
### With OpenCLIP
```
import torch
import torch.nn.functional as F
from urllib.request import urlopen
from PIL import Image
from open_clip import create_model_from_pretrained, get_tokenizer
model, preprocess = create_model_from_pretrained('hf-hub:apple/DFN5B-CLIP-ViT-H-14')
tokenizer = get_tokenizer('ViT-H-14')
image = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
image = preprocess(image).unsqueeze(0)
labels_list = ["a dog", "a cat", "a donut", "a beignet"]
text = tokenizer(labels_list, context_length=model.context_length)
with torch.no_grad(), torch.cuda.amp.autocast():
image_features = model.encode_image(image)
text_features = model.encode_text(text)
image_features = F.normalize(image_features, dim=-1)
text_features = F.normalize(text_features, dim=-1)
text_probs = torch.sigmoid(image_features @ text_features.T * model.logit_scale.exp() + model.logit_bias)
zipped_list = list(zip(labels_list, [round(p.item(), 3) for p in text_probs[0]]))
print("Label probabilities: ", zipped_list)
```
## Citation
```bibtex
@article{fang2023data,
title={Data Filtering Networks},
author={Fang, Alex and Jose, Albin Madappally and Jain, Amit and Schmidt, Ludwig and Toshev, Alexander and Shankar, Vaishaal},
journal={arXiv preprint arXiv:2309.17425},
year={2023}
}
```
|
codellama/CodeLlama-13b-hf | codellama | "2024-04-12T14:17:10Z" | 8,705 | 97 | transformers | [
"transformers",
"pytorch",
"safetensors",
"llama",
"text-generation",
"llama-2",
"code",
"arxiv:2308.12950",
"license:llama2",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2023-08-24T16:31:44Z" | ---
language:
- code
pipeline_tag: text-generation
tags:
- llama-2
license: llama2
---
# **Code Llama**
Code Llama is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 34 billion parameters. This is the repository for the base 13B version in the Hugging Face Transformers format. This model is designed for general code synthesis and understanding. Links to other models can be found in the index at the bottom.
> [!NOTE]
> This is a non-official Code Llama repo. You can find the official Meta repository in the [Meta Llama organization](https://huggingface.co/meta-llama/CodeLlama-13b-hf).
| | Base Model | Python | Instruct |
| --- | ----------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------- |
| 7B | [codellama/CodeLlama-7b-hf](https://huggingface.co/codellama/CodeLlama-7b-hf) | [codellama/CodeLlama-7b-Python-hf](https://huggingface.co/codellama/CodeLlama-7b-Python-hf) | [codellama/CodeLlama-7b-Instruct-hf](https://huggingface.co/codellama/CodeLlama-7b-Instruct-hf) |
| 13B | [codellama/CodeLlama-13b-hf](https://huggingface.co/codellama/CodeLlama-13b-hf) | [codellama/CodeLlama-13b-Python-hf](https://huggingface.co/codellama/CodeLlama-13b-Python-hf) | [codellama/CodeLlama-13b-Instruct-hf](https://huggingface.co/codellama/CodeLlama-13b-Instruct-hf) |
| 34B | [codellama/CodeLlama-34b-hf](https://huggingface.co/codellama/CodeLlama-34b-hf) | [codellama/CodeLlama-34b-Python-hf](https://huggingface.co/codellama/CodeLlama-34b-Python-hf) | [codellama/CodeLlama-34b-Instruct-hf](https://huggingface.co/codellama/CodeLlama-34b-Instruct-hf) |
| 70B | [codellama/CodeLlama-70b-hf](https://huggingface.co/codellama/CodeLlama-70b-hf) | [codellama/CodeLlama-70b-Python-hf](https://huggingface.co/codellama/CodeLlama-70b-Python-hf) | [codellama/CodeLlama-70b-Instruct-hf](https://huggingface.co/codellama/CodeLlama-70b-Instruct-hf) |
## Model Use
To use this model, please make sure to install transformers:
```bash
pip install transformers accelerate
```
Model capabilities:
- [x] Code completion.
- [x] Infilling.
- [ ] Instructions / chat.
- [ ] Python specialist.
```python
from transformers import AutoTokenizer
import transformers
import torch
model = "codellama/CodeLlama-13b-hf"
tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = transformers.pipeline(
"text-generation",
model=model,
torch_dtype=torch.float16,
device_map="auto",
)
sequences = pipeline(
'import socket\n\ndef ping_exponential_backoff(host: str):',
do_sample=True,
top_k=10,
temperature=0.1,
top_p=0.95,
num_return_sequences=1,
eos_token_id=tokenizer.eos_token_id,
max_length=200,
)
for seq in sequences:
print(f"Result: {seq['generated_text']}")
```
## Model Details
*Note: Use of this model is governed by the Meta license. Meta developed and publicly released the Code Llama family of large language models (LLMs).
**Model Developers** Meta
**Variations** Code Llama comes in three model sizes, and three variants:
* Code Llama: base models designed for general code synthesis and understanding
* Code Llama - Python: designed specifically for Python
* Code Llama - Instruct: for instruction following and safer deployment
All variants are available in sizes of 7B, 13B and 34B parameters.
**This repository contains the base version of the 13B parameters model.**
**Input** Models input text only.
**Output** Models generate text only.
**Model Architecture** Code Llama is an auto-regressive language model that uses an optimized transformer architecture.
**Model Dates** Code Llama and its variants have been trained between January 2023 and July 2023.
**Status** This is a static model trained on an offline dataset. Future versions of Code Llama - Instruct will be released as we improve model safety with community feedback.
**License** A custom commercial license is available at: [https://ai.meta.com/resources/models-and-libraries/llama-downloads/](https://ai.meta.com/resources/models-and-libraries/llama-downloads/)
**Research Paper** More information can be found in the paper "[Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/)" or its [arXiv page](https://arxiv.org/abs/2308.12950).
## Intended Use
**Intended Use Cases** Code Llama and its variants is intended for commercial and research use in English and relevant programming languages. The base model Code Llama can be adapted for a variety of code synthesis and understanding tasks, Code Llama - Python is designed specifically to handle the Python programming language, and Code Llama - Instruct is intended to be safer to use for code assistant and generation applications.
**Out-of-Scope Uses** Use in any manner that violates applicable laws or regulations (including trade compliance laws). Use in languages other than English. Use in any other way that is prohibited by the Acceptable Use Policy and Licensing Agreement for Code Llama and its variants.
## Hardware and Software
**Training Factors** We used custom training libraries. The training and fine-tuning of the released models have been performed Meta’s Research Super Cluster.
**Carbon Footprint** In aggregate, training all 9 Code Llama models required 400K GPU hours of computation on hardware of type A100-80GB (TDP of 350-400W). Estimated total emissions were 65.3 tCO2eq, 100% of which were offset by Meta’s sustainability program.
## Training Data
All experiments reported here and the released models have been trained and fine-tuned using the same data as Llama 2 with different weights (see Section 2 and Table 1 in the [research paper](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) for details).
## Evaluation Results
See evaluations for the main models and detailed ablations in Section 3 and safety evaluations in Section 4 of the research paper.
## Ethical Considerations and Limitations
Code Llama and its variants are a new technology that carries risks with use. Testing conducted to date has been in English, and has not covered, nor could it cover all scenarios. For these reasons, as with all LLMs, Code Llama’s potential outputs cannot be predicted in advance, and the model may in some instances produce inaccurate or objectionable responses to user prompts. Therefore, before deploying any applications of Code Llama, developers should perform safety testing and tuning tailored to their specific applications of the model.
Please see the Responsible Use Guide available available at [https://ai.meta.com/llama/responsible-use-guide](https://ai.meta.com/llama/responsible-use-guide). |
mosaicml/mpt-30b-instruct | mosaicml | "2024-03-05T20:25:52Z" | 8,698 | 100 | transformers | [
"transformers",
"pytorch",
"mpt",
"text-generation",
"Composer",
"MosaicML",
"llm-foundry",
"custom_code",
"dataset:competition_math",
"dataset:knkarthick/dialogsum",
"dataset:mosaicml/dolly_hhrlhf",
"dataset:duorc",
"dataset:emozilla/quality",
"dataset:scrolls/summ_screen_fd",
"dataset:spider",
"dataset:gsm8k",
"dataset:allenai/qasper",
"arxiv:2205.14135",
"arxiv:2108.12409",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2023-06-21T17:12:11Z" | ---
license: apache-2.0
datasets:
- competition_math
- knkarthick/dialogsum
- mosaicml/dolly_hhrlhf
- duorc
- emozilla/quality
- scrolls/summ_screen_fd
- spider
- gsm8k
- allenai/qasper
tags:
- Composer
- MosaicML
- llm-foundry
inference: false
---
# MPT-30B-Instruct
MPT-30B-Instruct is a model for short-form instruction following.
It is built by finetuning [MPT-30B](https://huggingface.co/mosaicml/mpt-30b) on [Dolly HHRLHF](https://huggingface.co/datasets/mosaicml/dolly_hhrlhf) derived from the [Databricks Dolly-15k](https://huggingface.co/datasets/databricks/databricks-dolly-15k) and the [Anthropic Helpful and Harmless (HH-RLHF)](https://huggingface.co/datasets/Anthropic/hh-rlhf) datasets. It is also trained on [Competition Math](https://huggingface.co/datasets/competition_math), [Duorc](https://huggingface.co/datasets/duorc), [CoT GSM8k](https://huggingface.co/datasets/conceptofmind/cot_submix_original), [Qasper](https://huggingface.co/datasets/allenai/qasper), [Quality](https://huggingface.co/datasets/emozilla/quality), [Summ Screen FD](https://huggingface.co/datasets/tau/scrolls) and [Spider](https://huggingface.co/datasets/spider).
* License: Apache 2.0
This model was trained by [MosaicML](https://www.mosaicml.com) and follows a modified decoder-only transformer architecture.
## Model Date
June 22, 2023
## Model License
Apache 2.0
## Documentation
* [Blog post: Raising the bar for open-source foundation models](https://www.mosaicml.com/blog/mpt-30b)
* [Codebase (mosaicml/llm-foundry repo)](https://github.com/mosaicml/llm-foundry/)
* Questions: Feel free to contact us via the [MosaicML Community Slack](https://mosaicml.me/slack)!
### Example Question/Instruction
**Bespokenizer46**
> I need to convince my friend, Phyllis, that she should train a custom LLM for her Fortune 500 company using the MosaicML Platform.
> Please write an email that explains why MosaicML's emphasis on cutting edge methodology, data privacy, and efficiency are so important.
> End the email with a friendly inquiry about Phyllis's family.
**MPT-30B-Instruct**:
> Phyllis -
> I know you're always looking for ways to improve efficiency at work and I think I've found a solution that you'll be very interested in.
> LLMs are only as good as the data they're trained on and the training process can be very time consuming and resource intensive. MosaicML uses cutting edge methodology to efficiently train high quality models without compromising data privacy.
> They also provide tools to easily connect to and use the model in your daily workflow.
> I think you'd really enjoy speaking with their founder, we can set up a call if you're interested.
> Also, I know it's been a tough year for your family, how are things?
> Best,
> Your Friend
## How to Use
Note: This model requires that `trust_remote_code=True` be passed to the `from_pretrained` method. This is because we use a custom model architecture that is not yet part of the `transformers` package.
It includes options for many training efficiency features such as [FlashAttention (Dao et al. 2022)](https://arxiv.org/pdf/2205.14135.pdf), [ALiBi](https://arxiv.org/abs/2108.12409), QK LayerNorm, and more.
```python
import transformers
model = transformers.AutoModelForCausalLM.from_pretrained(
'mosaicml/mpt-30b-instruct',
trust_remote_code=True
)
```
To use the optimized [triton implementation](https://github.com/openai/triton) of FlashAttention, you can load the model on GPU (`cuda:0`) with `attn_impl='triton'` and with `bfloat16` precision:
```python
import torch
import transformers
name = 'mosaicml/mpt-30b-instruct'
config = transformers.AutoConfig.from_pretrained(name, trust_remote_code=True)
config.attn_config['attn_impl'] = 'triton' # change this to use triton-based FlashAttention
config.init_device = 'cuda:0' # For fast initialization directly on GPU!
model = transformers.AutoModelForCausalLM.from_pretrained(
name,
config=config,
torch_dtype=torch.bfloat16, # Load model weights in bfloat16
trust_remote_code=True
)
```
The model was trained initially on a sequence length of 2048. An additional pre-training phase was included for sequence length adaptation to 8192. However, ALiBi further enables users to increase the maximum sequence length during finetuning and/or inference. For example:
```python
import transformers
name = 'mosaicml/mpt-30b-instruct'
config = transformers.AutoConfig.from_pretrained(name, trust_remote_code=True)
config.max_seq_len = 16384 # (input + output) tokens can now be up to 16384
model = transformers.AutoModelForCausalLM.from_pretrained(
name,
config=config,
trust_remote_code=True
)
```
This model was trained with the MPT-30B tokenizer which is based on the [EleutherAI/gpt-neox-20b](https://huggingface.co/EleutherAI/gpt-neox-20b) tokenizer and includes additional padding and eos tokens.
```python
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained('mosaicml/mpt-30b')
```
The model can then be used, for example, within a text-generation pipeline.
Note: when running Torch modules in lower precision, it is best practice to use the [torch.autocast context manager](https://pytorch.org/docs/stable/amp.html).
```python
from transformers import pipeline
with torch.autocast('cuda', dtype=torch.bfloat16):
inputs = tokenizer('Here is a recipe for vegan banana bread:\n', return_tensors="pt").to('cuda')
outputs = model.generate(**inputs, max_new_tokens=100)
print(tokenizer.batch_decode(outputs, skip_special_tokens=True))
# or using the HF pipeline
pipe = pipeline('text-generation', model=model, tokenizer=tokenizer, device='cuda:0')
with torch.autocast('cuda', dtype=torch.bfloat16):
print(
pipe('Here is a recipe for vegan banana bread:\n',
max_new_tokens=100,
do_sample=True,
use_cache=True))
```
### Formatting
This model was trained on data formatted as follows:
```python
def format_prompt(instruction):
template = "Below is an instruction that describes a task. Write a response that appropriately completes the request.\n\n###Instruction\n{instruction}\n\n### Response\n"
return template.format(instruction=instruction)
example = "Tell me a funny joke.\nDon't make it too funny though."
fmt_ex = format_prompt(instruction=example)
```
In the above example, `fmt_ex` is ready to be tokenized and sent through the model.
## Model Description
The architecture is a modification of a standard decoder-only transformer.
The model has been modified from a standard transformer in the following ways:
* It uses [FlashAttention](https://arxiv.org/pdf/2205.14135.pdf)
* It uses [ALiBi (Attention with Linear Biases)](https://arxiv.org/abs/2108.12409) and does not use positional embeddings
* It does not use biases
| Hyperparameter | Value |
|----------------|-------|
|n_parameters | 29.95B |
|n_layers | 48 |
| n_heads | 64 |
| d_model | 7168 |
| vocab size | 50432 |
| sequence length | 8192 |
## Data Mix
The model was trained on the following data mix:
| Data Source | Number of Tokens in Source | Proportion |
|-------------|----------------------------|------------|
| competition_math | 1.6 M | 3.66% |
| cot_gsm8k | 3.36 M | 7.67% |
| dialogsum | 0.1 M | 0.23% |
| dolly_hhrlhf | 5.89 M | 13.43% |
| duorc | 7.8 M | 17.80% |
| qasper | 8.72 M | 19.90% |
| quality | 11.29 M | 25.78% |
| scrolls/summ_screen_fd | 4.97 M | 11.33% |
| spider | 0.089 M | 0.20% |
## PreTraining Data
For more details on the pretraining process, see [MPT-30B](https://huggingface.co/mosaicml/mpt-30b).
The data was tokenized using the [EleutherAI/gpt-neox-20b](https://huggingface.co/EleutherAI/gpt-neox-20b) tokenizer.
### Training Configuration
This model was trained on 72 A100 40GB GPUs for 8 hours using the [MosaicML Platform](https://www.mosaicml.com/platform).
The model was trained with sharded data parallelism using [FSDP](https://pytorch.org/docs/stable/fsdp.html) and used the AdamW optimizer.
## Limitations and Biases
_The following language is modified from [EleutherAI's GPT-NeoX-20B](https://huggingface.co/EleutherAI/gpt-neox-20b)_
MPT-30B-Instruct can produce factually incorrect output, and should not be relied on to produce factually accurate information.
MPT-30B-Instruct was trained on various public datasets.
While great efforts have been taken to clean the pretraining data, it is possible that this model could generate lewd, biased or otherwise offensive outputs.
## Acknowledgements
This model was finetuned by Sam Havens, Alex Trott, and the MosaicML NLP team
## MosaicML Platform
If you're interested in [training](https://www.mosaicml.com/training) and [deploying](https://www.mosaicml.com/inference) your own MPT or LLMs on the MosaicML Platform, [sign up here](https://forms.mosaicml.com/demo?utm_source=huggingface&utm_medium=referral&utm_campaign=mpt-30b).
## Disclaimer
The license on this model does not constitute legal advice. We are not responsible for the actions of third parties who use this model. Please consult an attorney before using this model for commercial purposes.
## Citation
Please cite this model using the following format:
```
@online{MosaicML2023Introducing,
author = {MosaicML NLP Team},
title = {Introducing MPT-30B: Raising the bar
for open-source foundation models},
year = {2023},
url = {www.mosaicml.com/blog/mpt-30b},
note = {Accessed: 2023-06-22},
urldate = {2023-06-22}
}
``` |
ckpt/anything-v4.5-vae-swapped | ckpt | "2023-01-27T17:58:43Z" | 8,695 | 29 | diffusers | [
"diffusers",
"autotrain_compatible",
"endpoints_compatible",
"diffusers:StableDiffusionPipeline",
"region:us"
] | text-to-image | "2023-01-22T01:47:34Z" | Entry not found |
foduucom/stockmarket-pattern-detection-yolov8 | foduucom | "2023-09-11T10:13:51Z" | 8,693 | 155 | ultralytics | [
"ultralytics",
"tensorboard",
"v8",
"ultralyticsplus",
"yolov8",
"yolo",
"vision",
"object-detection",
"pytorch",
"finance",
"stock market",
"candlesticks",
"pattern recognition",
"option trading",
"chart reader",
"en",
"model-index",
"region:us"
] | object-detection | "2023-08-10T14:13:24Z" | ---
tags:
- ultralyticsplus
- yolov8
- ultralytics
- yolo
- vision
- object-detection
- pytorch
- finance
- stock market
- candlesticks
- pattern recognition
- option trading
- chart reader
library_name: ultralytics
library_version: 8.0.43
inference: false
model-index:
- name: foduucom/stockmarket-pattern-detection-yolov8
results:
- task:
type: object-detection
metrics:
- type: precision
value: 0.61355
name: [email protected](box)
language:
- en
pipeline_tag: object-detection
---
<div align="center">
<img width="500" alt="foduucom/stockmarket-pattern-detection-yolov8" src="https://huggingface.co/foduucom/stockmarket-pattern-detection-yolov8/resolve/main/thumbnail.jpg">
</div>
# Model Card for YOLOv8s Stock Market Pattern Detection on Live Trading Video Data
## Model Summary
The YOLOv8s Stock Market Pattern Detection model is an object detection model based on the YOLO (You Only Look Once) framework. It is designed to detect various chart patterns in real-time stock market trading video data. The model aids traders and investors by automating the analysis of chart patterns, providing timely insights for informed decision-making. The model has been fine-tuned on a diverse dataset and achieved high accuracy in detecting and classifying stock market patterns in live trading scenarios.
## Model Details
### Model Description
The YOLOv8s Stock Market Pattern Detection model offers a transformative solution for traders and investors by enabling real-time detection of crucial chart patterns within live trading video data. As stock markets evolve rapidly, this model's capabilities empower users with timely insights, allowing them to make informed decisions with speed and accuracy.
The model seamlessly integrates into live trading systems, providing instant pattern detection and classification. By leveraging advanced bounding box techniques and pattern-specific feature extraction, the model excels in identifying patterns such as 'Head and shoulders bottom,' 'Head and shoulders top,' 'M_Head,' 'StockLine,' 'Triangle,' and 'W_Bottom.' This enables traders to optimize their strategies, automate trading decisions, and respond to market trends in real-time.
To facilitate integration into live trading systems or to inquire about customization, please contact us at [email protected]. Your collaboration and feedback are instrumental in refining and enhancing the model's performance in dynamic trading environments.
- **Developed by:** FODUU AI
- **Model type:** Object Detection
- **Task:** Stock Market Pattern Detection on Live Trading Video Data
The YOLOv8s Stock Market Pattern Detection model is designed to adapt to the fast-paced nature of live trading environments. Its ability to operate on real-time video data allows traders and investors to harness pattern-based insights without delay.
### Supported Labels
```
['Head and shoulders bottom', 'Head and shoulders top', 'M_Head', 'StockLine', 'Triangle', 'W_Bottom']
```
## Uses
### Direct Use
The YOLOv8s Stock Market Pattern Detection model can be directly integrated into live trading systems to provide real-time detection and classification of chart patterns. Traders can utilize the model's insights for timely decision-making.
### Downstream Use
The model's real-time capabilities can be leveraged to automate trading strategies, generate alerts for specific patterns, and enhance overall trading performance.
### Training data
The Stockmarket model was trained on custom dataset consisting of 9000/800 annotated images for training/validation respectively.
### Out-of-Scope Use
The model is not designed for unrelated object detection tasks or scenarios outside the scope of stock market pattern detection in live trading video data.
## Bias, Risks, and Limitations
The YOLOv8s Stock Market Pattern Detection model may exhibit some limitations and biases:
- Performance may be affected by variations in video quality, lighting conditions, and pattern complexity within live trading data.
- Rapid market fluctuations and noise in video data may impact the model's accuracy and responsiveness.
- Market-specific patterns or anomalies not well-represented in the training data may pose challenges for detection.
### Recommendations
Users should be aware of the model's limitations and potential biases. Thorough testing and validation within live trading simulations are advised before deploying the model in real trading environments.
## How to Get Started with the Model
To begin using the YOLOv8s Stock Market Pattern Detection model on live trading video data, follow these steps:
```bash
pip install ultralyticsplus==0.0.28 ultralytics==8.0.43
```
- Load model and perform real-time prediction:
```python
from ultralyticsplus import YOLO, render_result
import cv2
# load model
model = YOLO('foduucom/stockmarket-pattern-detection-yolov8')
# set model parameters
model.overrides['conf'] = 0.25 # NMS confidence threshold
model.overrides['iou'] = 0.45 # NMS IoU threshold
model.overrides['agnostic_nms'] = False # NMS class-agnostic
model.overrides['max_det'] = 1000 # maximum number of detections per image
# initialize video capture
# Open the video file
video_path = "path/to/your/video/file.mp4"
cap = cv2.VideoCapture(video_path)
# Loop through the video frames
while cap.isOpened():
# Read a frame from the video
success, frame = cap.read()
if success:
# Run YOLOv8 inference on the frame
results = model(frame)
# Visualize the results on the frame
annotated_frame = results[0].plot()
# Display the annotated frame
cv2.imshow("YOLOv8 Inference", annotated_frame)
# Break the loop if 'q' is pressed
if cv2.waitKey(1) & 0xFF == ord("q"):
break
else:
# Break the loop if the end of the video is reached
break
# Release the video capture object and close the display window
cap.release()
cv2.destroyAllWindows()
```
## Training Details
### Training Data
The model is trained on a diverse dataset containing stock market chart images with various chart patterns, capturing different market conditions and scenarios.
### Training Procedure
The training process involves extensive computation and is conducted over multiple epochs. The model's weights are adjusted to minimize detection loss and optimize performance for stock market pattern detection.
#### Metrics
- [email protected] (box):
- All patterns: 0.932
- Individual patterns: Varies based on pattern type
### Model Architecture and Objective
The YOLOv8s architecture incorporates modifications tailored to stock market pattern detection. It features a specialized backbone network, self-attention mechanisms, and pattern-specific feature extraction modules.
### Compute Infrastructure
#### Hardware
NVIDIA GeForce RTX 3080 card
#### Software
The model was trained and fine-tuned using a Jupyter Notebook environment.
## Model Card Contact
For inquiries and contributions, please contact us at [email protected].
```bibtex
@ModelCard{
author = {Nehul Agrawal and
Pranjal Singh Thakur},
title = {YOLOv8s Stock Market Pattern Detection on Live Trading Video Data},
year = {2023}
}
``` |
01-ai/Yi-6B | 01-ai | "2024-06-26T10:25:32Z" | 8,690 | 364 | transformers | [
"transformers",
"pytorch",
"safetensors",
"llama",
"text-generation",
"arxiv:2403.04652",
"arxiv:2311.16502",
"arxiv:2401.11944",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2023-11-01T11:14:36Z" | ---
license: apache-2.0
widget:
- example_title: "Yi-34B-Chat"
text: "hi"
output:
text: " Hello! How can I assist you today?"
- example_title: "Yi-34B"
text: "There's a place where time stands still. A place of breath taking wonder, but also"
output:
text: " an eerie sense that something is just not right…\nBetween the two worlds lies The Forgotten Kingdom - home to creatures long since thought extinct and ancient magic so strong it defies belief! Only here can you find what has been lost for centuries: An Elixir Of Life which will restore youth and vitality if only those who seek its power are brave enough to face up against all manner of dangers lurking in this mysterious land! But beware; some say there may even exist powerful entities beyond our comprehension whose intentions towards humanity remain unclear at best ---- they might want nothing more than destruction itself rather then anything else from their quest after immortality (and maybe someone should tell them about modern medicine)? In any event though – one thing remains true regardless : whether or not success comes easy depends entirely upon how much effort we put into conquering whatever challenges lie ahead along with having faith deep down inside ourselves too ;) So let’s get started now shall We?"
pipeline_tag: text-generation
---
<div align="center">
<picture>
<source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/01-ai/Yi/main/assets/img/Yi_logo_icon_dark.svg" width="200px">
<source media="(prefers-color-scheme: light)" srcset="https://raw.githubusercontent.com/01-ai/Yi/main/assets/img/Yi_logo_icon_light.svg" width="200px">
<img alt="specify theme context for images" src="https://raw.githubusercontent.com/01-ai/Yi/main/assets/img/Yi_logo_icon_light.svg">
</picture>
</br>
</br>
<div style="display: inline-block;">
<a href="https://github.com/01-ai/Yi/actions/workflows/build_docker_image.yml">
<img src="https://github.com/01-ai/Yi/actions/workflows/build_docker_image.yml/badge.svg">
</a>
</div>
<div style="display: inline-block;">
<a href="mailto:[email protected]">
<img src="https://img.shields.io/badge/✉️[email protected]">
</a>
</div>
</div>
<div align="center">
<h3 align="center">Building the Next Generation of Open-Source and Bilingual LLMs</h3>
</div>
<p align="center">
🤗 <a href="https://huggingface.co/01-ai" target="_blank">Hugging Face</a> • 🤖 <a href="https://www.modelscope.cn/organization/01ai/" target="_blank">ModelScope</a> • ✡️ <a href="https://wisemodel.cn/organization/01.AI" target="_blank">WiseModel</a>
</p>
<p align="center">
👩🚀 Ask questions or discuss ideas on <a href="https://github.com/01-ai/Yi/discussions" target="_blank"> GitHub </a>
</p>
<p align="center">
👋 Join us on <a href="https://discord.gg/hYUwWddeAu" target="_blank"> 👾 Discord </a> or <a href="有官方的微信群嘛 · Issue #43 · 01-ai/Yi" target="_blank"> 💬 WeChat </a>
</p>
<p align="center">
📝 Check out <a href="https://arxiv.org/abs/2403.04652"> Yi Tech Report </a>
</p>
<p align="center">
📚 Grow at <a href="#learning-hub"> Yi Learning Hub </a>
</p>
<!-- DO NOT REMOVE ME -->
<hr>
<details open>
<summary></b>📕 Table of Contents</b></summary>
- [What is Yi?](#what-is-yi)
- [Introduction](#introduction)
- [Models](#models)
- [Chat models](#chat-models)
- [Base models](#base-models)
- [Model info](#model-info)
- [News](#news)
- [How to use Yi?](#how-to-use-yi)
- [Quick start](#quick-start)
- [Choose your path](#choose-your-path)
- [pip](#quick-start---pip)
- [docker](#quick-start---docker)
- [llama.cpp](#quick-start---llamacpp)
- [conda-lock](#quick-start---conda-lock)
- [Web demo](#web-demo)
- [Fine-tuning](#fine-tuning)
- [Quantization](#quantization)
- [Deployment](#deployment)
- [FAQ](#faq)
- [Learning hub](#learning-hub)
- [Why Yi?](#why-yi)
- [Ecosystem](#ecosystem)
- [Upstream](#upstream)
- [Downstream](#downstream)
- [Serving](#serving)
- [Quantization](#quantization-1)
- [Fine-tuning](#fine-tuning-1)
- [API](#api)
- [Benchmarks](#benchmarks)
- [Base model performance](#base-model-performance)
- [Chat model performance](#chat-model-performance)
- [Tech report](#tech-report)
- [Citation](#citation)
- [Who can use Yi?](#who-can-use-yi)
- [Misc.](#misc)
- [Acknowledgements](#acknowledgments)
- [Disclaimer](#disclaimer)
- [License](#license)
</details>
<hr>
# What is Yi?
## Introduction
- 🤖 The Yi series models are the next generation of open-source large language models trained from scratch by [01.AI](https://01.ai/).
- 🙌 Targeted as a bilingual language model and trained on 3T multilingual corpus, the Yi series models become one of the strongest LLM worldwide, showing promise in language understanding, commonsense reasoning, reading comprehension, and more. For example,
- Yi-34B-Chat model **landed in second place (following GPT-4 Turbo)**, outperforming other LLMs (such as GPT-4, Mixtral, Claude) on the AlpacaEval Leaderboard (based on data available up to January 2024).
- Yi-34B model **ranked first among all existing open-source models** (such as Falcon-180B, Llama-70B, Claude) in **both English and Chinese** on various benchmarks, including Hugging Face Open LLM Leaderboard (pre-trained) and C-Eval (based on data available up to November 2023).
- 🙏 (Credits to Llama) Thanks to the Transformer and Llama open-source communities, as they reduce the efforts required to build from scratch and enable the utilization of the same tools within the AI ecosystem.
<details style="display: inline;"><summary> If you're interested in Yi's adoption of Llama architecture and license usage policy, see <span style="color: green;">Yi's relation with Llama.</span> ⬇️</summary> <ul> <br>
> 💡 TL;DR
>
> The Yi series models adopt the same model architecture as Llama but are **NOT** derivatives of Llama.
- Both Yi and Llama are based on the Transformer structure, which has been the standard architecture for large language models since 2018.
- Grounded in the Transformer architecture, Llama has become a new cornerstone for the majority of state-of-the-art open-source models due to its excellent stability, reliable convergence, and robust compatibility. This positions Llama as the recognized foundational framework for models including Yi.
- Thanks to the Transformer and Llama architectures, other models can leverage their power, reducing the effort required to build from scratch and enabling the utilization of the same tools within their ecosystems.
- However, the Yi series models are NOT derivatives of Llama, as they do not use Llama's weights.
- As Llama's structure is employed by the majority of open-source models, the key factors of determining model performance are training datasets, training pipelines, and training infrastructure.
- Developing in a unique and proprietary way, Yi has independently created its own high-quality training datasets, efficient training pipelines, and robust training infrastructure entirely from the ground up. This effort has led to excellent performance with Yi series models ranking just behind GPT4 and surpassing Llama on the [Alpaca Leaderboard in Dec 2023](https://tatsu-lab.github.io/alpaca_eval/).
</ul>
</details>
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
## News
<details>
<summary>🎯 <b>2024-05-13</b>: The <a href="https://github.com/01-ai/Yi-1.5">Yi-1.5 series models </a> are open-sourced, further improving coding, math, reasoning, and instruction-following abilities.</summary>
</details>
<details>
<summary>🎯 <b>2024-03-16</b>: The <code>Yi-9B-200K</code> is open-sourced and available to the public.</summary>
</details>
<details>
<summary>🎯 <b>2024-03-08</b>: <a href="https://arxiv.org/abs/2403.04652">Yi Tech Report</a> is published! </summary>
</details>
<details open>
<summary>🔔 <b>2024-03-07</b>: The long text capability of the Yi-34B-200K has been enhanced. </summary>
<br>
In the "Needle-in-a-Haystack" test, the Yi-34B-200K's performance is improved by 10.5%, rising from 89.3% to an impressive 99.8%. We continue to pre-train the model on 5B tokens long-context data mixture and demonstrate a near-all-green performance.
</details>
<details open>
<summary>🎯 <b>2024-03-06</b>: The <code>Yi-9B</code> is open-sourced and available to the public.</summary>
<br>
<code>Yi-9B</code> stands out as the top performer among a range of similar-sized open-source models (including Mistral-7B, SOLAR-10.7B, Gemma-7B, DeepSeek-Coder-7B-Base-v1.5 and more), particularly excelling in code, math, common-sense reasoning, and reading comprehension.
</details>
<details open>
<summary>🎯 <b>2024-01-23</b>: The Yi-VL models, <code><a href="https://huggingface.co/01-ai/Yi-VL-34B">Yi-VL-34B</a></code> and <code><a href="https://huggingface.co/01-ai/Yi-VL-6B">Yi-VL-6B</a></code>, are open-sourced and available to the public.</summary>
<br>
<code><a href="https://huggingface.co/01-ai/Yi-VL-34B">Yi-VL-34B</a></code> has ranked <strong>first</strong> among all existing open-source models in the latest benchmarks, including <a href="https://arxiv.org/abs/2311.16502">MMMU</a> and <a href="https://arxiv.org/abs/2401.11944">CMMMU</a> (based on data available up to January 2024).</li>
</details>
<details>
<summary>🎯 <b>2023-11-23</b>: <a href="#chat-models">Chat models</a> are open-sourced and available to the public.</summary>
<br>This release contains two chat models based on previously released base models, two 8-bit models quantized by GPTQ, and two 4-bit models quantized by AWQ.
- `Yi-34B-Chat`
- `Yi-34B-Chat-4bits`
- `Yi-34B-Chat-8bits`
- `Yi-6B-Chat`
- `Yi-6B-Chat-4bits`
- `Yi-6B-Chat-8bits`
You can try some of them interactively at:
- [Hugging Face](https://huggingface.co/spaces/01-ai/Yi-34B-Chat)
- [Replicate](https://replicate.com/01-ai)
</details>
<details>
<summary>🔔 <b>2023-11-23</b>: The Yi Series Models Community License Agreement is updated to <a href="https://github.com/01-ai/Yi/blob/main/MODEL_LICENSE_AGREEMENT.txt">v2.1</a>.</summary>
</details>
<details>
<summary>🔥 <b>2023-11-08</b>: Invited test of Yi-34B chat model.</summary>
<br>Application form:
- [English](https://cn.mikecrm.com/l91ODJf)
- [Chinese](https://cn.mikecrm.com/gnEZjiQ)
</details>
<details>
<summary>🎯 <b>2023-11-05</b>: <a href="#base-models">The base models, </a><code>Yi-6B-200K</code> and <code>Yi-34B-200K</code>, are open-sourced and available to the public.</summary>
<br>This release contains two base models with the same parameter sizes as the previous
release, except that the context window is extended to 200K.
</details>
<details>
<summary>🎯 <b>2023-11-02</b>: <a href="#base-models">The base models, </a><code>Yi-6B</code> and <code>Yi-34B</code>, are open-sourced and available to the public.</summary>
<br>The first public release contains two bilingual (English/Chinese) base models
with the parameter sizes of 6B and 34B. Both of them are trained with 4K
sequence length and can be extended to 32K during inference time.
</details>
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
## Models
Yi models come in multiple sizes and cater to different use cases. You can also fine-tune Yi models to meet your specific requirements.
If you want to deploy Yi models, make sure you meet the [software and hardware requirements](#deployment).
### Chat models
| Model | Download |
|---|---|
|Yi-34B-Chat | • [🤗 Hugging Face](https://huggingface.co/01-ai/Yi-34B-Chat) • [🤖 ModelScope](https://www.modelscope.cn/models/01ai/Yi-34B-Chat/summary) • [🟣 wisemodel](https://wisemodel.cn/models/01.AI/Yi-34B-Chat) |
|Yi-34B-Chat-4bits | • [🤗 Hugging Face](https://huggingface.co/01-ai/Yi-34B-Chat-4bits) • [🤖 ModelScope](https://www.modelscope.cn/models/01ai/Yi-34B-Chat-4bits/summary) • [🟣 wisemodel](https://wisemodel.cn/models/01.AI/Yi-34B-Chat-4bits) |
|Yi-34B-Chat-8bits | • [🤗 Hugging Face](https://huggingface.co/01-ai/Yi-34B-Chat-8bits) • [🤖 ModelScope](https://www.modelscope.cn/models/01ai/Yi-34B-Chat-8bits/summary) • [🟣 wisemodel](https://wisemodel.cn/models/01.AI/Yi-34B-Chat-8bits) |
|Yi-6B-Chat| • [🤗 Hugging Face](https://huggingface.co/01-ai/Yi-6B-Chat) • [🤖 ModelScope](https://www.modelscope.cn/models/01ai/Yi-6B-Chat/summary) • [🟣 wisemodel](https://wisemodel.cn/models/01.AI/Yi-6B-Chat) |
|Yi-6B-Chat-4bits | • [🤗 Hugging Face](https://huggingface.co/01-ai/Yi-6B-Chat-4bits) • [🤖 ModelScope](https://www.modelscope.cn/models/01ai/Yi-6B-Chat-4bits/summary) • [🟣 wisemodel](https://wisemodel.cn/models/01.AI/Yi-6B-Chat-4bits) |
|Yi-6B-Chat-8bits | • [🤗 Hugging Face](https://huggingface.co/01-ai/Yi-6B-Chat-8bits) • [🤖 ModelScope](https://www.modelscope.cn/models/01ai/Yi-6B-Chat-8bits/summary) • [🟣 wisemodel](https://wisemodel.cn/models/01.AI/Yi-6B-Chat-8bits) |
<sub><sup> - 4-bit series models are quantized by AWQ. <br> - 8-bit series models are quantized by GPTQ <br> - All quantized models have a low barrier to use since they can be deployed on consumer-grade GPUs (e.g., 3090, 4090). </sup></sub>
### Base models
| Model | Download |
|---|---|
|Yi-34B| • [🤗 Hugging Face](https://huggingface.co/01-ai/Yi-34B) • [🤖 ModelScope](https://www.modelscope.cn/models/01ai/Yi-34B/summary) • [🟣 wisemodel](https://wisemodel.cn/models/01.AI/Yi-6B-Chat-8bits) |
|Yi-34B-200K|• [🤗 Hugging Face](https://huggingface.co/01-ai/Yi-34B-200K) • [🤖 ModelScope](https://www.modelscope.cn/models/01ai/Yi-34B-200K/summary) • [🟣 wisemodel](https://wisemodel.cn/models/01.AI/Yi-6B-Chat-8bits)|
|Yi-9B|• [🤗 Hugging Face](https://huggingface.co/01-ai/Yi-9B) • [🤖 ModelScope](https://wisemodel.cn/models/01.AI/Yi-6B-Chat-8bits) • [🟣 wisemodel](https://wisemodel.cn/models/01.AI/Yi-9B)|
|Yi-9B-200K | • [🤗 Hugging Face](https://huggingface.co/01-ai/Yi-9B-200K) • [🤖 ModelScope](https://wisemodel.cn/models/01.AI/Yi-9B-200K) • [🟣 wisemodel](https://wisemodel.cn/models/01.AI/Yi-6B-Chat-8bits) |
|Yi-6B| • [🤗 Hugging Face](https://huggingface.co/01-ai/Yi-6B) • [🤖 ModelScope](https://www.modelscope.cn/models/01ai/Yi-6B/summary) • [🟣 wisemodel](https://wisemodel.cn/models/01.AI/Yi-6B-Chat-8bits) |
|Yi-6B-200K | • [🤗 Hugging Face](https://huggingface.co/01-ai/Yi-6B-200K) • [🤖 ModelScope](https://www.modelscope.cn/models/01ai/Yi-6B-200K/summary) • [🟣 wisemodel](https://wisemodel.cn/models/01.AI/Yi-6B-Chat-8bits) |
<sub><sup> - 200k is roughly equivalent to 400,000 Chinese characters. <br> - If you want to use the previous version of the Yi-34B-200K (released on Nov 5, 2023), run `git checkout 069cd341d60f4ce4b07ec394e82b79e94f656cf` to download the weight. </sup></sub>
### Model info
- For chat and base models
<table>
<thead>
<tr>
<th>Model</th>
<th>Intro</th>
<th>Default context window</th>
<th>Pretrained tokens</th>
<th>Training Data Date</th>
</tr>
</thead>
<tbody><tr>
<td>6B series models</td>
<td>They are suitable for personal and academic use.</td>
<td rowspan="3">4K</td>
<td>3T</td>
<td rowspan="3">Up to June 2023</td>
</tr>
<tr>
<td>9B series models</td>
<td>It is the best at coding and math in the Yi series models.</td>
<td>Yi-9B is continuously trained based on Yi-6B, using 0.8T tokens.</td>
</tr>
<tr>
<td>34B series models</td>
<td>They are suitable for personal, academic, and commercial (particularly for small and medium-sized enterprises) purposes. It's a cost-effective solution that's affordable and equipped with emergent ability.</td>
<td>3T</td>
</tr>
</tbody></table>
- For chat models
<details style="display: inline;"><summary>For chat model limitations, see the explanations below. ⬇️</summary>
<ul>
<br>The released chat model has undergone exclusive training using Supervised Fine-Tuning (SFT). Compared to other standard chat models, our model produces more diverse responses, making it suitable for various downstream tasks, such as creative scenarios. Furthermore, this diversity is expected to enhance the likelihood of generating higher quality responses, which will be advantageous for subsequent Reinforcement Learning (RL) training.
<br>However, this higher diversity might amplify certain existing issues, including:
<li>Hallucination: This refers to the model generating factually incorrect or nonsensical information. With the model's responses being more varied, there's a higher chance of hallucination that are not based on accurate data or logical reasoning.</li>
<li>Non-determinism in re-generation: When attempting to regenerate or sample responses, inconsistencies in the outcomes may occur. The increased diversity can lead to varying results even under similar input conditions.</li>
<li>Cumulative Error: This occurs when errors in the model's responses compound over time. As the model generates more diverse responses, the likelihood of small inaccuracies building up into larger errors increases, especially in complex tasks like extended reasoning, mathematical problem-solving, etc.</li>
<li>To achieve more coherent and consistent responses, it is advisable to adjust generation configuration parameters such as temperature, top_p, or top_k. These adjustments can help in the balance between creativity and coherence in the model's outputs.</li>
</ul>
</details>
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
# How to use Yi?
- [Quick start](#quick-start)
- [Choose your path](#choose-your-path)
- [pip](#quick-start---pip)
- [docker](#quick-start---docker)
- [conda-lock](#quick-start---conda-lock)
- [llama.cpp](#quick-start---llamacpp)
- [Web demo](#web-demo)
- [Fine-tuning](#fine-tuning)
- [Quantization](#quantization)
- [Deployment](#deployment)
- [FAQ](#faq)
- [Learning hub](#learning-hub)
## Quick start
Getting up and running with Yi models is simple with multiple choices available.
### Choose your path
Select one of the following paths to begin your journey with Yi!
#### 🎯 Deploy Yi locally
If you prefer to deploy Yi models locally,
- 🙋♀️ and you have **sufficient** resources (for example, NVIDIA A800 80GB), you can choose one of the following methods:
- [pip](#quick-start---pip)
- [Docker](#quick-start---docker)
- [conda-lock](#quick-start---conda-lock)
- 🙋♀️ and you have **limited** resources (for example, a MacBook Pro), you can use [llama.cpp](#quick-start---llamacpp).
#### 🎯 Not to deploy Yi locally
If you prefer not to deploy Yi models locally, you can explore Yi's capabilities using any of the following options.
##### 🙋♀️ Run Yi with APIs
If you want to explore more features of Yi, you can adopt one of these methods:
- Yi APIs (Yi official)
- [Early access has been granted](https://x.com/01AI_Yi/status/1735728934560600536?s=20) to some applicants. Stay tuned for the next round of access!
- [Yi APIs](https://replicate.com/01-ai/yi-34b-chat/api?tab=nodejs) (Replicate)
##### 🙋♀️ Run Yi in playground
If you want to chat with Yi with more customizable options (e.g., system prompt, temperature, repetition penalty, etc.), you can try one of the following options:
- [Yi-34B-Chat-Playground](https://platform.lingyiwanwu.com/prompt/playground) (Yi official)
- Access is available through a whitelist. Welcome to apply (fill out a form in [English](https://cn.mikecrm.com/l91ODJf) or [Chinese](https://cn.mikecrm.com/gnEZjiQ)).
- [Yi-34B-Chat-Playground](https://replicate.com/01-ai/yi-34b-chat) (Replicate)
##### 🙋♀️ Chat with Yi
If you want to chat with Yi, you can use one of these online services, which offer a similar user experience:
- [Yi-34B-Chat](https://huggingface.co/spaces/01-ai/Yi-34B-Chat) (Yi official on Hugging Face)
- No registration is required.
- [Yi-34B-Chat](https://platform.lingyiwanwu.com/) (Yi official beta)
- Access is available through a whitelist. Welcome to apply (fill out a form in [English](https://cn.mikecrm.com/l91ODJf) or [Chinese](https://cn.mikecrm.com/gnEZjiQ)).
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
### Quick start - pip
This tutorial guides you through every step of running **Yi-34B-Chat locally on an A800 (80G)** and then performing inference.
#### Step 0: Prerequisites
- Make sure Python 3.10 or a later version is installed.
- If you want to run other Yi models, see [software and hardware requirements](#deployment).
#### Step 1: Prepare your environment
To set up the environment and install the required packages, execute the following command.
```bash
git clone https://github.com/01-ai/Yi.git
cd yi
pip install -r requirements.txt
```
#### Step 2: Download the Yi model
You can download the weights and tokenizer of Yi models from the following sources:
- [Hugging Face](https://huggingface.co/01-ai)
- [ModelScope](https://www.modelscope.cn/organization/01ai/)
- [WiseModel](https://wisemodel.cn/organization/01.AI)
#### Step 3: Perform inference
You can perform inference with Yi chat or base models as below.
##### Perform inference with Yi chat model
1. Create a file named `quick_start.py` and copy the following content to it.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_path = '<your-model-path>'
tokenizer = AutoTokenizer.from_pretrained(model_path, use_fast=False)
# Since transformers 4.35.0, the GPT-Q/AWQ model can be loaded using AutoModelForCausalLM.
model = AutoModelForCausalLM.from_pretrained(
model_path,
device_map="auto",
torch_dtype='auto'
).eval()
# Prompt content: "hi"
messages = [
{"role": "user", "content": "hi"}
]
input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt')
output_ids = model.generate(input_ids.to('cuda'))
response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True)
# Model response: "Hello! How can I assist you today?"
print(response)
```
2. Run `quick_start.py`.
```bash
python quick_start.py
```
Then you can see an output similar to the one below. 🥳
```bash
Hello! How can I assist you today?
```
##### Perform inference with Yi base model
- Yi-34B
The steps are similar to [pip - Perform inference with Yi chat model](#perform-inference-with-yi-chat-model).
You can use the existing file [`text_generation.py`](https://github.com/01-ai/Yi/tree/main/demo).
```bash
python demo/text_generation.py --model <your-model-path>
```
Then you can see an output similar to the one below. 🥳
<details>
<summary>Output. ⬇️ </summary>
<br>
**Prompt**: Let me tell you an interesting story about cat Tom and mouse Jerry,
**Generation**: Let me tell you an interesting story about cat Tom and mouse Jerry, which happened in my childhood. My father had a big house with two cats living inside it to kill mice. One day when I was playing at home alone, I found one of the tomcats lying on his back near our kitchen door, looking very much like he wanted something from us but couldn’t get up because there were too many people around him! He kept trying for several minutes before finally giving up...
</details>
- Yi-9B
Input
```bash
from transformers import AutoModelForCausalLM, AutoTokenizer
MODEL_DIR = "01-ai/Yi-9B"
model = AutoModelForCausalLM.from_pretrained(MODEL_DIR, torch_dtype="auto")
tokenizer = AutoTokenizer.from_pretrained(MODEL_DIR, use_fast=False)
input_text = "# write the quick sort algorithm"
inputs = tokenizer(input_text, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_length=256)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```
Output
```bash
# write the quick sort algorithm
def quick_sort(arr):
if len(arr) <= 1:
return arr
pivot = arr[len(arr) // 2]
left = [x for x in arr if x < pivot]
middle = [x for x in arr if x == pivot]
right = [x for x in arr if x > pivot]
return quick_sort(left) + middle + quick_sort(right)
# test the quick sort algorithm
print(quick_sort([3, 6, 8, 10, 1, 2, 1]))
```
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
### Quick start - Docker
<details>
<summary> Run Yi-34B-chat locally with Docker: a step-by-step guide. ⬇️</summary>
<br>This tutorial guides you through every step of running <strong>Yi-34B-Chat on an A800 GPU</strong> or <strong>4*4090</strong> locally and then performing inference.
<h4>Step 0: Prerequisites</h4>
<p>Make sure you've installed <a href="https://docs.docker.com/engine/install/?open_in_browser=true">Docker</a> and <a href="https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html">nvidia-container-toolkit</a>.</p>
<h4> Step 1: Start Docker </h4>
<pre><code>docker run -it --gpus all \
-v <your-model-path>: /models
ghcr.io/01-ai/yi:latest
</code></pre>
<p>Alternatively, you can pull the Yi Docker image from <code>registry.lingyiwanwu.com/ci/01-ai/yi:latest</code>.</p>
<h4>Step 2: Perform inference</h4>
<p>You can perform inference with Yi chat or base models as below.</p>
<h5>Perform inference with Yi chat model</h5>
<p>The steps are similar to <a href="#perform-inference-with-yi-chat-model">pip - Perform inference with Yi chat model</a>.</p>
<p><strong>Note</strong> that the only difference is to set <code>model_path = '<your-model-mount-path>'</code> instead of <code>model_path = '<your-model-path>'</code>.</p>
<h5>Perform inference with Yi base model</h5>
<p>The steps are similar to <a href="#perform-inference-with-yi-base-model">pip - Perform inference with Yi base model</a>.</p>
<p><strong>Note</strong> that the only difference is to set <code>--model <your-model-mount-path>'</code> instead of <code>model <your-model-path></code>.</p>
</details>
### Quick start - conda-lock
<details>
<summary>You can use <code><a href="https://github.com/conda/conda-lock">conda-lock</a></code> to generate fully reproducible lock files for conda environments. ⬇️</summary>
<br>
You can refer to <a href="https://github.com/01-ai/Yi/blob/ebba23451d780f35e74a780987ad377553134f68/conda-lock.yml">conda-lock.yml</a> for the exact versions of the dependencies. Additionally, you can utilize <code><a href="https://mamba.readthedocs.io/en/latest/user_guide/micromamba.html">micromamba</a></code> for installing these dependencies.
<br>
To install the dependencies, follow these steps:
1. Install micromamba by following the instructions available <a href="https://mamba.readthedocs.io/en/latest/installation/micromamba-installation.html">here</a>.
2. Execute <code>micromamba install -y -n yi -f conda-lock.yml</code> to create a conda environment named <code>yi</code> and install the necessary dependencies.
</details>
### Quick start - llama.cpp
<a href="https://github.com/01-ai/Yi/blob/main/docs/README_llama.cpp.md">The following tutorial </a> will guide you through every step of running a quantized model (<a href="https://huggingface.co/XeIaso/yi-chat-6B-GGUF/tree/main">Yi-chat-6B-2bits</a>) locally and then performing inference.
<details>
<summary> Run Yi-chat-6B-2bits locally with llama.cpp: a step-by-step guide. ⬇️</summary>
<br><a href="https://github.com/01-ai/Yi/blob/main/docs/README_llama.cpp.md">This tutorial</a> guides you through every step of running a quantized model (<a href="https://huggingface.co/XeIaso/yi-chat-6B-GGUF/tree/main">Yi-chat-6B-2bits</a>) locally and then performing inference.</p>
- [Step 0: Prerequisites](#step-0-prerequisites)
- [Step 1: Download llama.cpp](#step-1-download-llamacpp)
- [Step 2: Download Yi model](#step-2-download-yi-model)
- [Step 3: Perform inference](#step-3-perform-inference)
#### Step 0: Prerequisites
- This tutorial assumes you use a MacBook Pro with 16GB of memory and an Apple M2 Pro chip.
- Make sure [`git-lfs`](https://git-lfs.com/) is installed on your machine.
#### Step 1: Download `llama.cpp`
To clone the [`llama.cpp`](https://github.com/ggerganov/llama.cpp) repository, run the following command.
```bash
git clone [email protected]:ggerganov/llama.cpp.git
```
#### Step 2: Download Yi model
2.1 To clone [XeIaso/yi-chat-6B-GGUF](https://huggingface.co/XeIaso/yi-chat-6B-GGUF/tree/main) with just pointers, run the following command.
```bash
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/XeIaso/yi-chat-6B-GGUF
```
2.2 To download a quantized Yi model ([yi-chat-6b.Q2_K.gguf](https://huggingface.co/XeIaso/yi-chat-6B-GGUF/blob/main/yi-chat-6b.Q2_K.gguf)), run the following command.
```bash
git-lfs pull --include yi-chat-6b.Q2_K.gguf
```
#### Step 3: Perform inference
To perform inference with the Yi model, you can use one of the following methods.
- [Method 1: Perform inference in terminal](#method-1-perform-inference-in-terminal)
- [Method 2: Perform inference in web](#method-2-perform-inference-in-web)
##### Method 1: Perform inference in terminal
To compile `llama.cpp` using 4 threads and then conduct inference, navigate to the `llama.cpp` directory, and run the following command.
> ##### Tips
>
> - Replace `/Users/yu/yi-chat-6B-GGUF/yi-chat-6b.Q2_K.gguf` with the actual path of your model.
>
> - By default, the model operates in completion mode.
>
> - For additional output customization options (for example, system prompt, temperature, repetition penalty, etc.), run `./main -h` to check detailed descriptions and usage.
```bash
make -j4 && ./main -m /Users/yu/yi-chat-6B-GGUF/yi-chat-6b.Q2_K.gguf -p "How do you feed your pet fox? Please answer this question in 6 simple steps:\nStep 1:" -n 384 -e
...
How do you feed your pet fox? Please answer this question in 6 simple steps:
Step 1: Select the appropriate food for your pet fox. You should choose high-quality, balanced prey items that are suitable for their unique dietary needs. These could include live or frozen mice, rats, pigeons, or other small mammals, as well as fresh fruits and vegetables.
Step 2: Feed your pet fox once or twice a day, depending on the species and its individual preferences. Always ensure that they have access to fresh water throughout the day.
Step 3: Provide an appropriate environment for your pet fox. Ensure it has a comfortable place to rest, plenty of space to move around, and opportunities to play and exercise.
Step 4: Socialize your pet with other animals if possible. Interactions with other creatures can help them develop social skills and prevent boredom or stress.
Step 5: Regularly check for signs of illness or discomfort in your fox. Be prepared to provide veterinary care as needed, especially for common issues such as parasites, dental health problems, or infections.
Step 6: Educate yourself about the needs of your pet fox and be aware of any potential risks or concerns that could affect their well-being. Regularly consult with a veterinarian to ensure you are providing the best care.
...
```
Now you have successfully asked a question to the Yi model and got an answer! 🥳
##### Method 2: Perform inference in web
1. To initialize a lightweight and swift chatbot, run the following command.
```bash
cd llama.cpp
./server --ctx-size 2048 --host 0.0.0.0 --n-gpu-layers 64 --model /Users/yu/yi-chat-6B-GGUF/yi-chat-6b.Q2_K.gguf
```
Then you can get an output like this:
```bash
...
llama_new_context_with_model: n_ctx = 2048
llama_new_context_with_model: freq_base = 5000000.0
llama_new_context_with_model: freq_scale = 1
ggml_metal_init: allocating
ggml_metal_init: found device: Apple M2 Pro
ggml_metal_init: picking default device: Apple M2 Pro
ggml_metal_init: ggml.metallib not found, loading from source
ggml_metal_init: GGML_METAL_PATH_RESOURCES = nil
ggml_metal_init: loading '/Users/yu/llama.cpp/ggml-metal.metal'
ggml_metal_init: GPU name: Apple M2 Pro
ggml_metal_init: GPU family: MTLGPUFamilyApple8 (1008)
ggml_metal_init: hasUnifiedMemory = true
ggml_metal_init: recommendedMaxWorkingSetSize = 11453.25 MB
ggml_metal_init: maxTransferRate = built-in GPU
ggml_backend_metal_buffer_type_alloc_buffer: allocated buffer, size = 128.00 MiB, ( 2629.44 / 10922.67)
llama_new_context_with_model: KV self size = 128.00 MiB, K (f16): 64.00 MiB, V (f16): 64.00 MiB
ggml_backend_metal_buffer_type_alloc_buffer: allocated buffer, size = 0.02 MiB, ( 2629.45 / 10922.67)
llama_build_graph: non-view tensors processed: 676/676
llama_new_context_with_model: compute buffer total size = 159.19 MiB
ggml_backend_metal_buffer_type_alloc_buffer: allocated buffer, size = 156.02 MiB, ( 2785.45 / 10922.67)
Available slots:
-> Slot 0 - max context: 2048
llama server listening at http://0.0.0.0:8080
```
2. To access the chatbot interface, open your web browser and enter `http://0.0.0.0:8080` into the address bar.
3. Enter a question, such as "How do you feed your pet fox? Please answer this question in 6 simple steps" into the prompt window, and you will receive a corresponding answer.
</ul>
</details>
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
### Web demo
You can build a web UI demo for Yi **chat** models (note that Yi base models are not supported in this senario).
[Step 1: Prepare your environment](#step-1-prepare-your-environment).
[Step 2: Download the Yi model](#step-2-download-the-yi-model).
Step 3. To start a web service locally, run the following command.
```bash
python demo/web_demo.py -c <your-model-path>
```
You can access the web UI by entering the address provided in the console into your browser.
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
### Fine-tuning
```bash
bash finetune/scripts/run_sft_Yi_6b.sh
```
Once finished, you can compare the finetuned model and the base model with the following command:
```bash
bash finetune/scripts/run_eval.sh
```
<details style="display: inline;"><summary>For advanced usage (like fine-tuning based on your custom data), see the explanations below. ⬇️ </summary> <ul>
### Finetune code for Yi 6B and 34B
#### Preparation
##### From Image
By default, we use a small dataset from [BAAI/COIG](https://huggingface.co/datasets/BAAI/COIG) to finetune the base model.
You can also prepare your customized dataset in the following `jsonl` format:
```json
{ "prompt": "Human: Who are you? Assistant:", "chosen": "I'm Yi." }
```
And then mount them in the container to replace the default ones:
```bash
docker run -it \
-v /path/to/save/finetuned/model/:/finetuned-model \
-v /path/to/train.jsonl:/yi/finetune/data/train.json \
-v /path/to/eval.jsonl:/yi/finetune/data/eval.json \
ghcr.io/01-ai/yi:latest \
bash finetune/scripts/run_sft_Yi_6b.sh
```
##### From Local Server
Make sure you have conda. If not, use
```bash
mkdir -p ~/miniconda3
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -O ~/miniconda3/miniconda.sh
bash ~/miniconda3/miniconda.sh -b -u -p ~/miniconda3
rm -rf ~/miniconda3/miniconda.sh
~/miniconda3/bin/conda init bash
source ~/.bashrc
```
Then, create a conda env:
```bash
conda create -n dev_env python=3.10 -y
conda activate dev_env
pip install torch==2.0.1 deepspeed==0.10 tensorboard transformers datasets sentencepiece accelerate ray==2.7
```
#### Hardware Setup
For the Yi-6B model, a node with 4 GPUs, each with GPU memory larger than 60GB, is recommended.
For the Yi-34B model, because the usage of the zero-offload technique consumes a lot of CPU memory, please be careful to limit the number of GPUs in the 34B finetune training. Please use CUDA_VISIBLE_DEVICES to limit the number of GPUs (as shown in scripts/run_sft_Yi_34b.sh).
A typical hardware setup for finetuning the 34B model is a node with 8 GPUs (limited to 4 in running by CUDA_VISIBLE_DEVICES=0,1,2,3), each with GPU memory larger than 80GB, and total CPU memory larger than 900GB.
#### Quick Start
Download a LLM-base model to MODEL_PATH (6B and 34B). A typical folder of models is like:
```bash
|-- $MODEL_PATH
| |-- config.json
| |-- pytorch_model-00001-of-00002.bin
| |-- pytorch_model-00002-of-00002.bin
| |-- pytorch_model.bin.index.json
| |-- tokenizer_config.json
| |-- tokenizer.model
| |-- ...
```
Download a dataset from huggingface to local storage DATA_PATH, e.g. Dahoas/rm-static.
```bash
|-- $DATA_PATH
| |-- data
| | |-- train-00000-of-00001-2a1df75c6bce91ab.parquet
| | |-- test-00000-of-00001-8c7c51afc6d45980.parquet
| |-- dataset_infos.json
| |-- README.md
```
`finetune/yi_example_dataset` has example datasets, which are modified from [BAAI/COIG](https://huggingface.co/datasets/BAAI/COIG)
```bash
|-- $DATA_PATH
|--data
|-- train.jsonl
|-- eval.jsonl
```
`cd` into the scripts folder, copy and paste the script, and run. For example:
```bash
cd finetune/scripts
bash run_sft_Yi_6b.sh
```
For the Yi-6B base model, setting training_debug_steps=20 and num_train_epochs=4 can output a chat model, which takes about 20 minutes.
For the Yi-34B base model, it takes a relatively long time for initialization. Please be patient.
#### Evaluation
```bash
cd finetune/scripts
bash run_eval.sh
```
Then you'll see the answer from both the base model and the finetuned model.
</ul>
</details>
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
### Quantization
#### GPT-Q
```bash
python quantization/gptq/quant_autogptq.py \
--model /base_model \
--output_dir /quantized_model \
--trust_remote_code
```
Once finished, you can then evaluate the resulting model as follows:
```bash
python quantization/gptq/eval_quantized_model.py \
--model /quantized_model \
--trust_remote_code
```
<details style="display: inline;"><summary>For details, see the explanations below. ⬇️</summary> <ul>
#### GPT-Q quantization
[GPT-Q](https://github.com/IST-DASLab/gptq) is a PTQ (Post-Training Quantization)
method. It saves memory and provides potential speedups while retaining the accuracy
of the model.
Yi models can be GPT-Q quantized without a lot of efforts.
We provide a step-by-step tutorial below.
To run GPT-Q, we will use [AutoGPTQ](https://github.com/PanQiWei/AutoGPTQ) and
[exllama](https://github.com/turboderp/exllama).
And the huggingface transformers has integrated optimum and auto-gptq to perform
GPTQ quantization on language models.
##### Do Quantization
The `quant_autogptq.py` script is provided for you to perform GPT-Q quantization:
```bash
python quant_autogptq.py --model /base_model \
--output_dir /quantized_model --bits 4 --group_size 128 --trust_remote_code
```
##### Run Quantized Model
You can run a quantized model using the `eval_quantized_model.py`:
```bash
python eval_quantized_model.py --model /quantized_model --trust_remote_code
```
</ul>
</details>
#### AWQ
```bash
python quantization/awq/quant_autoawq.py \
--model /base_model \
--output_dir /quantized_model \
--trust_remote_code
```
Once finished, you can then evaluate the resulting model as follows:
```bash
python quantization/awq/eval_quantized_model.py \
--model /quantized_model \
--trust_remote_code
```
<details style="display: inline;"><summary>For details, see the explanations below. ⬇️</summary> <ul>
#### AWQ quantization
[AWQ](https://github.com/mit-han-lab/llm-awq) is a PTQ (Post-Training Quantization)
method. It's an efficient and accurate low-bit weight quantization (INT3/4) for LLMs.
Yi models can be AWQ quantized without a lot of efforts.
We provide a step-by-step tutorial below.
To run AWQ, we will use [AutoAWQ](https://github.com/casper-hansen/AutoAWQ).
##### Do Quantization
The `quant_autoawq.py` script is provided for you to perform AWQ quantization:
```bash
python quant_autoawq.py --model /base_model \
--output_dir /quantized_model --bits 4 --group_size 128 --trust_remote_code
```
##### Run Quantized Model
You can run a quantized model using the `eval_quantized_model.py`:
```bash
python eval_quantized_model.py --model /quantized_model --trust_remote_code
```
</ul>
</details>
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
### Deployment
If you want to deploy Yi models, make sure you meet the software and hardware requirements.
#### Software requirements
Before using Yi quantized models, make sure you've installed the correct software listed below.
| Model | Software
|---|---
Yi 4-bit quantized models | [AWQ and CUDA](https://github.com/casper-hansen/AutoAWQ?tab=readme-ov-file#install-from-pypi)
Yi 8-bit quantized models | [GPTQ and CUDA](https://github.com/PanQiWei/AutoGPTQ?tab=readme-ov-file#quick-installation)
#### Hardware requirements
Before deploying Yi in your environment, make sure your hardware meets the following requirements.
##### Chat models
| Model | Minimum VRAM | Recommended GPU Example |
|:----------------------|:--------------|:-------------------------------------:|
| Yi-6B-Chat | 15 GB | 1 x RTX 3090 (24 GB) <br> 1 x RTX 4090 (24 GB) <br> 1 x A10 (24 GB) <br> 1 x A30 (24 GB) |
| Yi-6B-Chat-4bits | 4 GB | 1 x RTX 3060 (12 GB)<br> 1 x RTX 4060 (8 GB) |
| Yi-6B-Chat-8bits | 8 GB | 1 x RTX 3070 (8 GB) <br> 1 x RTX 4060 (8 GB) |
| Yi-34B-Chat | 72 GB | 4 x RTX 4090 (24 GB)<br> 1 x A800 (80GB) |
| Yi-34B-Chat-4bits | 20 GB | 1 x RTX 3090 (24 GB) <br> 1 x RTX 4090 (24 GB) <br> 1 x A10 (24 GB) <br> 1 x A30 (24 GB) <br> 1 x A100 (40 GB) |
| Yi-34B-Chat-8bits | 38 GB | 2 x RTX 3090 (24 GB) <br> 2 x RTX 4090 (24 GB)<br> 1 x A800 (40 GB) |
Below are detailed minimum VRAM requirements under different batch use cases.
| Model | batch=1 | batch=4 | batch=16 | batch=32 |
| ----------------------- | ------- | ------- | -------- | -------- |
| Yi-6B-Chat | 12 GB | 13 GB | 15 GB | 18 GB |
| Yi-6B-Chat-4bits | 4 GB | 5 GB | 7 GB | 10 GB |
| Yi-6B-Chat-8bits | 7 GB | 8 GB | 10 GB | 14 GB |
| Yi-34B-Chat | 65 GB | 68 GB | 76 GB | > 80 GB |
| Yi-34B-Chat-4bits | 19 GB | 20 GB | 30 GB | 40 GB |
| Yi-34B-Chat-8bits | 35 GB | 37 GB | 46 GB | 58 GB |
##### Base models
| Model | Minimum VRAM | Recommended GPU Example |
|----------------------|--------------|:-------------------------------------:|
| Yi-6B | 15 GB | 1 x RTX 3090 (24 GB) <br> 1 x RTX 4090 (24 GB) <br> 1 x A10 (24 GB) <br> 1 x A30 (24 GB) |
| Yi-6B-200K | 50 GB | 1 x A800 (80 GB) |
| Yi-9B | 20 GB | 1 x RTX 4090 (24 GB) |
| Yi-34B | 72 GB | 4 x RTX 4090 (24 GB) <br> 1 x A800 (80 GB) |
| Yi-34B-200K | 200 GB | 4 x A800 (80 GB) |
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
### FAQ
<details>
<summary> If you have any questions while using the Yi series models, the answers provided below could serve as a helpful reference for you. ⬇️</summary>
<br>
#### 💡Fine-tuning
- <strong>Base model or Chat model - which to fine-tune?</strong>
<br>The choice of pre-trained language model for fine-tuning hinges on the computational resources you have at your disposal and the particular demands of your task.
- If you are working with a substantial volume of fine-tuning data (say, over 10,000 samples), the Base model could be your go-to choice.
- On the other hand, if your fine-tuning data is not quite as extensive, opting for the Chat model might be a more fitting choice.
- It is generally advisable to fine-tune both the Base and Chat models, compare their performance, and then pick the model that best aligns with your specific requirements.
- <strong>Yi-34B versus Yi-34B-Chat for full-scale fine-tuning - what is the difference?</strong>
<br>
The key distinction between full-scale fine-tuning on `Yi-34B`and `Yi-34B-Chat` comes down to the fine-tuning approach and outcomes.
- Yi-34B-Chat employs a Special Fine-Tuning (SFT) method, resulting in responses that mirror human conversation style more closely.
- The Base model's fine-tuning is more versatile, with a relatively high performance potential.
- If you are confident in the quality of your data, fine-tuning with `Yi-34B` could be your go-to.
- If you are aiming for model-generated responses that better mimic human conversational style, or if you have doubts about your data quality, `Yi-34B-Chat` might be your best bet.
#### 💡Quantization
- <strong>Quantized model versus original model - what is the performance gap?</strong>
- The performance variance is largely contingent on the quantization method employed and the specific use cases of these models. For instance, when it comes to models provided by the AWQ official, from a Benchmark standpoint, quantization might result in a minor performance drop of a few percentage points.
- Subjectively speaking, in situations like logical reasoning, even a 1% performance shift could impact the accuracy of the output results.
#### 💡General
- <strong>Where can I source fine-tuning question answering datasets?</strong>
- You can find fine-tuning question answering datasets on platforms like Hugging Face, with datasets like [m-a-p/COIG-CQIA](https://huggingface.co/datasets/m-a-p/COIG-CQIA) readily available.
- Additionally, Github offers fine-tuning frameworks, such as [hiyouga/LLaMA-Factory](https://github.com/hiyouga/LLaMA-Factory), which integrates pre-made datasets.
- <strong>What is the GPU memory requirement for fine-tuning Yi-34B FP16?</strong>
<br>
The GPU memory needed for fine-tuning 34B FP16 hinges on the specific fine-tuning method employed. For full parameter fine-tuning, you'll need 8 GPUs each with 80 GB; however, more economical solutions like Lora require less. For more details, check out [hiyouga/LLaMA-Factory](https://github.com/hiyouga/LLaMA-Factory). Also, consider using BF16 instead of FP16 for fine-tuning to optimize performance.
- <strong>Are there any third-party platforms that support chat functionality for the Yi-34b-200k model?</strong>
<br>
If you're looking for third-party Chats, options include [fireworks.ai](https://fireworks.ai/login?callbackURL=https://fireworks.ai/models/fireworks/yi-34b-chat).
</details>
### Learning hub
<details>
<summary> If you want to learn Yi, you can find a wealth of helpful educational resources here. ⬇️</summary>
<br>
Welcome to the Yi learning hub!
Whether you're a seasoned developer or a newcomer, you can find a wealth of helpful educational resources to enhance your understanding and skills with Yi models, including insightful blog posts, comprehensive video tutorials, hands-on guides, and more.
The content you find here has been generously contributed by knowledgeable Yi experts and passionate enthusiasts. We extend our heartfelt gratitude for your invaluable contributions!
At the same time, we also warmly invite you to join our collaborative effort by contributing to Yi. If you have already made contributions to Yi, please don't hesitate to showcase your remarkable work in the table below.
With all these resources at your fingertips, you're ready to start your exciting journey with Yi. Happy learning! 🥳
#### Tutorials
##### Blog tutorials
| Deliverable | Date | Author |
| ------------------------------------------------------------ | ---------- | ------------------------------------------------------------ |
| [使用 Dify、Meilisearch、零一万物模型实现最简单的 RAG 应用（三）：AI 电影推荐](https://mp.weixin.qq.com/s/Ri2ap9_5EMzdfiBhSSL_MQ) | 2024-05-20 | [苏洋](https://github.com/soulteary) |
| [使用autodl服务器，在A40显卡上运行， Yi-34B-Chat-int4模型，并使用vllm优化加速，显存占用42G，速度18 words-s](https://blog.csdn.net/freewebsys/article/details/134698597?ops_request_misc=%7B%22request%5Fid%22%3A%22171636168816800227489911%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636168816800227489911&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-17-134698597-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2024-05-20 | [fly-iot](https://gitee.com/fly-iot) |
| [Yi-VL 最佳实践](https://modelscope.cn/docs/yi-vl最佳实践) | 2024-05-20 | [ModelScope](https://github.com/modelscope) |
| [一键运行零一万物新鲜出炉Yi-1.5-9B-Chat大模型](https://mp.weixin.qq.com/s/ntMs2G_XdWeM3I6RUOBJrA) | 2024-05-13 | [Second State](https://github.com/second-state) |
| [零一万物开源Yi-1.5系列大模型](https://mp.weixin.qq.com/s/d-ogq4hcFbsuL348ExJxpA) | 2024-05-13 | [刘聪](https://github.com/liucongg) |
| [零一万物Yi-1.5系列模型发布并开源！ 34B-9B-6B 多尺寸，魔搭社区推理微调最佳实践教程来啦！](https://mp.weixin.qq.com/s/3wD-0dCgXB646r720o8JAg) | 2024-05-13 | [ModelScope](https://github.com/modelscope) |
| [Yi-34B 本地部署简单测试](https://blog.csdn.net/arkohut/article/details/135331469?ops_request_misc=%7B%22request%5Fid%22%3A%22171636390616800185813639%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636390616800185813639&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-10-135331469-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2024-05-13 | [漆妮妮](https://space.bilibili.com/1262370256) |
| [驾辰龙跨Llama持Wasm，玩转Yi模型迎新春过大年（上）](https://blog.csdn.net/weixin_53443275/article/details/136091398?ops_request_misc=%7B%22request%5Fid%22%3A%22171636390616800185813639%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636390616800185813639&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-5-136091398-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2024-05-13 | [Words worth](https://blog.csdn.net/weixin_53443275?type=blog) |
| [驾辰龙跨Llama持Wasm，玩转Yi模型迎新春过大年（下篇）](https://blog.csdn.net/weixin_53443275/article/details/136096309) | 2024-05-13 | [Words worth](https://blog.csdn.net/weixin_53443275?type=blog) |
| [Ollama新增两个命令，开始支持零一万物Yi-1.5系列模型](https://mp.weixin.qq.com/s/bBgzGJvUqIohodcy9U-pFw) | 2024-05-13 | AI工程师笔记 |
| [使用零一万物 200K 模型和 Dify 快速搭建模型应用](https://zhuanlan.zhihu.com/p/686774859) | 2024-05-13 | [苏洋](https://github.com/soulteary) |
| [(持更) 零一万物模型折腾笔记：社区 Yi-34B 微调模型使用](https://zhuanlan.zhihu.com/p/671549900) | 2024-05-13 | [苏洋](https://github.com/soulteary) |
| [Python+ERNIE-4.0-8K-Yi-34B-Chat大模型初探](https://mp.weixin.qq.com/s/WaygSfn5T8ZPB1mPdGADEQ) | 2024-05-11 | 江湖评谈 |
| [技术布道 Vue及Python调用零一万物模型和Prompt模板（通过百度千帆大模型平台）](https://blog.csdn.net/ucloud2012/article/details/137187469) | 2024-05-11 | [MumuLab](https://blog.csdn.net/ucloud2012?type=blog) |
| [多模态大模型Yi-VL-plus体验 效果很棒](https://zhuanlan.zhihu.com/p/694736111) | 2024-04-27 | [大家好我是爱因](https://www.zhihu.com/people/iamein) |
| [使用autodl服务器，两个3090显卡上运行， Yi-34B-Chat-int4模型，并使用vllm优化加速，显存占用42G，速度23 words-s](https://blog.csdn.net/freewebsys/article/details/134725765?ops_request_misc=%7B%22request%5Fid%22%3A%22171636356716800211598950%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636356716800211598950&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-9-134725765-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2024-04-27 | [fly-iot](https://gitee.com/fly-iot) |
| [Getting Started with Yi-1.5-9B-Chat](https://www.secondstate.io/articles/yi-1.5-9b-chat/) | 2024-04-27 | [Second State](https://github.com/second-state) |
| [基于零一万物yi-vl-plus大模型简单几步就能批量生成Anki图片笔记](https://mp.weixin.qq.com/s/_ea6g0pzzeO4WyYtuWycWQ) | 2024-04-24 | [正经人王同学](https://github.com/zjrwtx) |
| [【AI开发：语言】一、Yi-34B超大模型本地部署CPU和GPU版](https://blog.csdn.net/alarey/article/details/137769471?ops_request_misc=%7B%22request%5Fid%22%3A%22171636168816800227489911%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636168816800227489911&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-16-137769471-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2024-04-21 | [My的梦想已实现](https://blog.csdn.net/alarey?type=blog) |
| [【Yi-34B-Chat-Int4】使用4个2080Ti显卡11G版本，运行Yi-34B模型，5年前老显卡是支持的，可以正常运行，速度 21 words-s，vllm要求算力在7以上的显卡就可以](https://blog.csdn.net/freewebsys/article/details/134754086) | 2024-03-22 | [fly-iot](https://gitee.com/fly-iot) |
| [零一万物大模型部署+微调总结](https://blog.csdn.net/v_wus/article/details/135704126?ops_request_misc=%7B%22request%5Fid%22%3A%22171636168816800227489911%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636168816800227489911&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-18-135704126-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2024-03-22 | [v_wus](https://blog.csdn.net/v_wus?type=blog) |
| [零一万物Yi大模型vllm推理时Yi-34B或Yi-6bchat重复输出的解决方案](https://blog.csdn.net/qq_39667443/article/details/136028776?ops_request_misc=%7B%22request%5Fid%22%3A%22171636168816800227489911%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636168816800227489911&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-6-136028776-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2024-03-02 | [郝铠锋](https://blog.csdn.net/qq_39667443?type=blog) |
| [Yi-34B微调训练](https://blog.csdn.net/lsjlnd/article/details/135336984?ops_request_misc=%7B%22request%5Fid%22%3A%22171636343416800188513953%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636343416800188513953&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-12-135336984-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2024-03-02 | [lsjlnd](https://blog.csdn.net/lsjlnd?type=blog) |
| [实测零一万物Yi-VL多模态语言模型：能准确“识图吃瓜”](https://mp.weixin.qq.com/s/fu4O9XvJ03JhimsEyI-SsQ) | 2024-02-02 | [苏洋](https://github.com/soulteary) |
| [零一万物开源Yi-VL多模态大模型，魔搭社区推理&微调最佳实践来啦！](https://zhuanlan.zhihu.com/p/680098411) | 2024-01-26 | [ModelScope](https://github.com/modelscope) |
| [单卡 3 小时训练 Yi-6B 大模型 Agent：基于 Llama Factory 实战](https://zhuanlan.zhihu.com/p/678989191) | 2024-01-22 | [郑耀威](https://github.com/hiyouga) |
| [零一科技Yi-34B Chat大模型环境搭建&推理](https://blog.csdn.net/zzq1989_/article/details/135597181?ops_request_misc=%7B%22request%5Fid%22%3A%22171636168816800227489911%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636168816800227489911&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-8-135597181-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2024-01-15 | [要养家的程序员](https://blog.csdn.net/zzq1989_?type=blog) |
| [基于LLaMA Factory，单卡3小时训练专属大模型 Agent](https://blog.csdn.net/m0_59596990/article/details/135760285?ops_request_misc=%7B%22request%5Fid%22%3A%22171636343416800188513953%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636343416800188513953&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-10-135760285-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2024-01-15 | [机器学习社区](https://blog.csdn.net/m0_59596990?type=blog) |
| [双卡 3080ti 部署 Yi-34B 大模型 - Gradio + vLLM 踩坑全记录](https://blog.csdn.net/arkohut/article/details/135321242?ops_request_misc=%7B%22request%5Fid%22%3A%22171636168816800227489911%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636168816800227489911&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-10-135321242-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2024-01-02 | [漆妮妮](https://space.bilibili.com/1262370256) |
| [【大模型部署实践-3】3个能在3090上跑起来的4bits量化Chat模型（baichuan2-13b、InternLM-20b、Yi-34b）](https://blog.csdn.net/qq_40302568/article/details/135040985?ops_request_misc=%7B%22request%5Fid%22%3A%22171636168816800227489911%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636168816800227489911&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-30-135040985-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2024-01-02 | [aq_Seabiscuit](https://blog.csdn.net/qq_40302568?type=blog) |
| [只需 24G 显存，用 vllm 跑起来 Yi-34B 中英双语大模型](https://blog.csdn.net/arkohut/article/details/135274973) | 2023-12-28 | [漆妮妮](https://space.bilibili.com/1262370256) |
| [零一万物模型官方 Yi-34B 模型本地离线运行部署使用笔记（物理机和docker两种部署方式），200K 超长文本内容，34B 干翻一众 70B 模型，打榜分数那么高，这模型到底行不行？](https://blog.csdn.net/u014374009/article/details/136327696) | 2023-12-28 | [代码讲故事](https://blog.csdn.net/u014374009?type=blog) |
| [LLM - 大模型速递之 Yi-34B 入门与 LoRA 微调](https://blog.csdn.net/BIT_666/article/details/134990402) | 2023-12-18 | [BIT_666](https://bitddd.blog.csdn.net/?type=blog) |
| [通过vllm框架进行大模型推理](https://blog.csdn.net/weixin_45920955/article/details/135300561?ops_request_misc=%7B%22request%5Fid%22%3A%22171636343416800188513953%22%2C%22scm%22%3A%2220140713.130102334.pc%5Fblog.%22%7D&request_id=171636343416800188513953&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2~blog~first_rank_ecpm_v1~times_rank-13-135300561-null-null.nonecase&utm_term=Yi大模型&spm=1018.2226.3001.4450) | 2023-12-18 | [土山炮](https://blog.csdn.net/weixin_45920955?type=blog) |
| [CPU 混合推理，非常见大模型量化方案：“二三五六” 位量化方案](https://zhuanlan.zhihu.com/p/671698216) | 2023-12-12 | [苏洋](https://github.com/soulteary) |
| [零一万物模型折腾笔记：官方 Yi-34B 模型基础使用](https://zhuanlan.zhihu.com/p/671387298) | 2023-12-10 | [苏洋](https://github.com/soulteary) |
| [Running Yi-34B-Chat locally using LlamaEdge](https://www.secondstate.io/articles/yi-34b/) | 2023-11-30 | [Second State](https://github.com/second-state) |
| [本地运行零一万物 34B 大模型，使用 Llama.cpp & 21G 显存](https://zhuanlan.zhihu.com/p/668921042) | 2023-11-26 | [苏洋](https://github.com/soulteary) |
##### GitHub Project
| Deliverable | Date | Author |
| ------------------------------------------------------------ | ---------- | ------------------------------------------- |
| [yi-openai-proxy](https://github.com/soulteary/yi-openai-proxy) | 2024-05-11 | [苏洋](https://github.com/soulteary) |
| [基于零一万物 Yi 模型和 B 站构建大语言模型高质量训练数据集](https://github.com/zjrwtx/bilibiliQA_databuilder) | 2024-04-29 | [正经人王同学](https://github.com/zjrwtx) |
| [基于视频网站和零一万物大模型构建大语言模型高质量训练数据集](https://github.com/zjrwtx/VideoQA_databuilder) | 2024-04-25 | [正经人王同学](https://github.com/zjrwtx) |
| [基于零一万物yi-34b-chat-200k输入任意文章地址，点击按钮即可生成无广告或推广内容的简要笔记，并生成分享图给好友](https://github.com/zjrwtx/open_summary) | 2024-04-24 | [正经人王同学](https://github.com/zjrwtx) |
| [Food-GPT-Yi-model](https://github.com/ThisisHubert/FoodGPT-Yi-model) | 2024-04-21 | [Hubert S](https://github.com/ThisisHubert) |
##### Video tutorials
| Deliverable | Date | Author |
| ------------------------------------------------------------ | ---------- | ------------------------------------------------------------ |
| [Run dolphin-2.2-yi-34b on IoT Devices](https://www.youtube.com/watch?v=NJ89T5mO25Y) | 2023-11-30 | [Second State](https://github.com/second-state) |
| [只需 24G 显存，用 vllm 跑起来 Yi-34B 中英双语大模型](https://www.bilibili.com/video/BV17t4y1f7Ee/) | 2023-12-28 | [漆妮妮](https://space.bilibili.com/1262370256) |
| [Install Yi 34B Locally - Chinese English Bilingual LLM](https://www.youtube.com/watch?v=CVQvj4Wrh4w&t=476s) | 2023-11-05 | [Fahd Mirza](https://www.youtube.com/@fahdmirza) |
| [Dolphin Yi 34b - Brand New Foundational Model TESTED](https://www.youtube.com/watch?v=On3Zuv27V3k&t=85s) | 2023-11-27 | [Matthew Berman](https://www.youtube.com/@matthew_berman) |
| [Yi-VL-34B 多模态大模型 - 用两张 A40 显卡跑起来](https://www.bilibili.com/video/BV1Q5411y7AG/) | 2024-01-28 | [漆妮妮](https://space.bilibili.com/1262370256) |
| [4060Ti 16G显卡安装零一万物最新开源的Yi-1.5版大语言模型](https://www.bilibili.com/video/BV16i421X7Jx/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-05-14 | [titan909](https://space.bilibili.com/526393761) |
| [Yi-1.5: True Apache 2.0 Competitor to LLAMA-3](https://www.youtube.com/watch?v=KCDYrfWeTRc) | 2024-05-13 | [Prompt Engineering](https://www.youtube.com/@engineerprompt) |
| [Install Yi-1.5 Model Locally - Beats Llama 3 in Various Benchmarks](https://www.youtube.com/watch?v=Ba-G7Il0UkA) | 2024-05-13 | [Fahd Mirza](https://www.youtube.com/@fahdmirza) |
| [how to install Ollama and run Yi 6B](https://www.youtube.com/watch?v=4Jnar7OUHqQ) | 2024-05-13 | [Ridaa Davids](https://www.youtube.com/@quantanovabusiness) |
| [地表最强混合智能AI助手：llama3_70B+Yi_34B+Qwen1.5_110B](https://www.bilibili.com/video/BV1Xm411C7V1/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-05-04 | [朱扎特](https://space.bilibili.com/494512200?spm_id_from=333.788.0.0) |
| [ChatDoc学术论文辅助--基于Yi-34B和langchain进行PDF知识库问答](https://www.bilibili.com/video/BV11i421C7B5/?spm_id_from=333.999.0.0&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-05-03 | [朱扎特](https://space.bilibili.com/494512200?spm_id_from=333.788.0.0) |
| [基于Yi-34B的领域知识问答项目演示](https://www.bilibili.com/video/BV1zZ42177ZA/?spm_id_from=333.999.0.0&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-05-02 | [朱扎特](https://space.bilibili.com/494512200?spm_id_from=333.788.0.0) |
| [使用RTX4090+GaLore算法 全参微调Yi-6B大模型](https://www.bilibili.com/video/BV1ax4y1U7Ep/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-03-24 | [小工蚂创始人](https://space.bilibili.com/478674499?spm_id_from=333.788.0.0) |
| [无内容审查NSFW大语言模型Yi-34B-Chat蒸馏版测试,RolePlay,《天龙八部》马夫人康敏,本地GPU,CPU运行](https://www.youtube.com/watch?v=VL-W0TnLCns) | 2024-03-20 | [刘悦的技术博客](https://v3u.cn/) |
| [无内容审查NSFW大语言模型整合包,Yi-34B-Chat,本地CPU运行,角色扮演潘金莲](https://www.youtube.com/watch?v=rBvbgwz3oHM) | 2024-03-16 | [刘悦的技术博客](https://v3u.cn/) |
| [量化 Yi-34B-Chat 并在单卡 RTX 4090 使用 vLLM 部署](https://www.bilibili.com/video/BV1jx421y7xj/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-03-05 | [白鸽巢](https://space.bilibili.com/138938660?spm_id_from=333.788.0.0) |
| [Yi-VL-34B（5）：使用3个3090显卡24G版本，运行Yi-VL-34B模型，支持命令行和web界面方式，理解图片的内容转换成文字](https://www.bilibili.com/video/BV1BB421z7oA/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-02-27 | [fly-iot](https://gitee.com/fly-iot) |
| [Win环境KoboldCpp本地部署大语言模型进行各种角色扮演游戏](https://www.bilibili.com/video/BV14J4m1e77f/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-02-25 | [魚蟲蟲](https://space.bilibili.com/431981179?spm_id_from=333.788.0.0) |
| [无需显卡本地部署Yi-34B-Chat进行角色扮演游戏 P2](https://www.bilibili.com/video/BV19v421677y/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-02-23 | [魚蟲蟲](https://space.bilibili.com/431981179?spm_id_from=333.788.0.0) |
| [【wails】（2）：使用go-llama.cpp 运行 yi-01-6b大模型，使用本地CPU运行，速度还可以，等待下一版本更新](https://www.bilibili.com/video/BV194421F7Fy/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-02-20 | [fly-iot](https://gitee.com/fly-iot) |
| [【xinference】（6）：在autodl上，使用xinference部署yi-vl-chat和qwen-vl-chat模型，可以使用openai调用成功](https://www.bilibili.com/video/BV19Z421z7cv/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-02-06 | [fly-iot](https://gitee.com/fly-iot) |
| [无需显卡本地部署Yi-34B-Chat进行角色扮演游戏 P1](https://www.bilibili.com/video/BV1tU421o7Co/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-02-05 | [魚蟲蟲](https://space.bilibili.com/431981179?spm_id_from=333.788.0.0) |
| [2080Ti部署YI-34B大模型 xinference-oneapi-fastGPT本地知识库使用指南](https://www.bilibili.com/video/BV1hC411z7xu/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-01-30 | [小饭护法要转码](https://space.bilibili.com/39486865?spm_id_from=333.788.0.0) |
| [Best Story Writing AI Model - Install Yi 6B 200K Locally on Windows](https://www.youtube.com/watch?v=cZs2jRtl0bs) | 2024-01-22 | [Fahd Mirza](https://www.youtube.com/@fahdmirza) |
| [Mac 本地运行大语言模型方法与常见问题指南（Yi 34B 模型+32 GB 内存测试）](https://www.bilibili.com/video/BV1VT4y1b7Th/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-01-21 | [小吴苹果机器人](https://space.bilibili.com/1732749682?spm_id_from=333.788.0.0) |
| [【Dify知识库】（11）：Dify0.4.9改造支持MySQL，成功接入yi-6b 做对话，本地使用fastchat启动，占8G显存，完成知识库配置](https://www.bilibili.com/video/BV1ia4y1y7JH/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-01-21 | [fly-iot](https://gitee.com/fly-iot) |
| [这位LLM先生有点暴躁,用的是YI-6B的某个量化版,#LLM #大语言模型 #暴躁老哥](https://www.youtube.com/watch?v=eahXJrdtQuc) | 2024-01-20 | [晓漫吧](https://www.youtube.com/@xiaomanba) |
| [大模型推理 NvLink 桥接器有用吗｜双卡 A6000 测试一下](https://www.bilibili.com/video/BV1AW4y1w7DC/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-01-17 | [漆妮妮](https://space.bilibili.com/1262370256) |
| [大模型推理 A40 vs A6000 谁更强 - 对比 Yi-34B 的单、双卡推理性能](https://www.bilibili.com/video/BV1aK4y1z7GF/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-01-15 | [漆妮妮](https://space.bilibili.com/1262370256) |
| [C-Eval 大语言模型评测基准- 用 LM Evaluation Harness + vLLM 跑起来](https://www.bilibili.com/video/BV1Yw411g7ZL/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-01-11 | [漆妮妮](https://space.bilibili.com/1262370256) |
| [双显卡部署 Yi-34B 大模型 - vLLM + Gradio 踩坑记录](https://www.bilibili.com/video/BV1p94y1c7ak/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2024-01-01 | [漆妮妮](https://space.bilibili.com/1262370256) |
| [手把手教学！使用 vLLM 快速部署 Yi-34B-Chat](https://www.bilibili.com/video/BV1ew41157Mk/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2023-12-26 | [白鸽巢](https://space.bilibili.com/138938660?spm_id_from=333.788.0.0) |
| [如何训练企业自己的大语言模型？Yi-6B LORA微调演示 #小工蚁](https://www.bilibili.com/video/BV1uc41117zz/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2023-12-21 | [小工蚂创始人](https://space.bilibili.com/478674499?spm_id_from=333.788.0.0) |
| [Yi-34B（4）：使用4个2080Ti显卡11G版本，运行Yi-34B模型，5年前老显卡是支持的，可以正常运行，速度 21 words/s](https://www.bilibili.com/video/BV1nj41157L3/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2023-12-02 | [fly-iot](https://gitee.com/fly-iot) |
| [使用autodl服务器，RTX 3090 * 3 显卡上运行， Yi-34B-Chat模型，显存占用60G](https://www.bilibili.com/video/BV1BM411R7ae/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2023-12-01 | [fly-iot](https://gitee.com/fly-iot) |
| [使用autodl服务器，两个3090显卡上运行， Yi-34B-Chat-int4模型，用vllm优化，增加 --num-gpu 2，速度23 words/s](https://www.bilibili.com/video/BV1Hu4y1L7BH/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2023-12-01 | [fly-iot](https://gitee.com/fly-iot) |
| [Yi大模型一键本地部署 技术小白玩转AI](https://www.bilibili.com/video/BV16H4y117md/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2023-12-01 | [技术小白玩转AI](https://space.bilibili.com/3546586137234288?spm_id_from=333.788.0.0) |
| [01.AI's Yi-6B: Overview and Fine-Tuning](https://www.youtube.com/watch?v=mye-UOkAliQ) | 2023-11-28 | [AI Makerspace](https://www.youtube.com/@AI-Makerspace) |
| [Yi 34B Chat LLM outperforms Llama 70B](https://www.youtube.com/watch?v=RYtrF-R5jDc) | 2023-11-27 | [DLExplorer](https://www.youtube.com/@DLExplorers-lg7dt) |
| [How to run open source models on mac Yi 34b on m3 Max](https://www.youtube.com/watch?v=GAo-dopkgjI) | 2023-11-26 | [TECHNO PREMIUM](https://www.youtube.com/@technopremium91) |
| [Yi-34B - 200K - The BEST & NEW CONTEXT WINDOW KING ](https://www.youtube.com/watch?v=7WBojwwv5Qo) | 2023-11-24 | [Prompt Engineering](https://www.youtube.com/@engineerprompt) |
| [Yi 34B : The Rise of Powerful Mid-Sized Models - Base,200k & Chat](https://www.youtube.com/watch?v=bWCjwtu_tHs) | 2023-11-24 | [Sam Witteveen](https://www.youtube.com/@samwitteveenai) |
| [在IoT设备运行破解版李开复大模型dolphin-2.2-yi-34b（还可作为私有OpenAI API服务器）](https://www.bilibili.com/video/BV1SQ4y18744/?spm_id_from=333.337.search-card.all.click&vd_source=ab85f93e294a2f6be11db57c29c6d706) | 2023-11-15 | [Second State](https://github.com/second-state) |
| [Run dolphin-2.2-yi-34b on IoT Devices (Also works as a Private OpenAI API Server)](https://www.youtube.com/watch?v=NJ89T5mO25Y) | 2023-11-14 | [Second State](https://github.com/second-state) |
| [How to Install Yi 34B 200K Llamafied on Windows Laptop](https://www.youtube.com/watch?v=enoha4K4HkQ) | 2023-11-11 | [Fahd Mirza](https://www.youtube.com/@fahdmirza) |
</details>
# Why Yi?
- [Ecosystem](#ecosystem)
- [Upstream](#upstream)
- [Downstream](#downstream)
- [Serving](#serving)
- [Quantization](#quantization-1)
- [Fine-tuning](#fine-tuning-1)
- [API](#api)
- [Benchmarks](#benchmarks)
- [Chat model performance](#chat-model-performance)
- [Base model performance](#base-model-performance)
- [Yi-34B and Yi-34B-200K](#yi-34b-and-yi-34b-200k)
- [Yi-9B](#yi-9b)
## Ecosystem
Yi has a comprehensive ecosystem, offering a range of tools, services, and models to enrich your experiences and maximize productivity.
- [Upstream](#upstream)
- [Downstream](#downstream)
- [Serving](#serving)
- [Quantization](#quantization-1)
- [Fine-tuning](#fine-tuning-1)
- [API](#api)
### Upstream
The Yi series models follow the same model architecture as Llama. By choosing Yi, you can leverage existing tools, libraries, and resources within the Llama ecosystem, eliminating the need to create new tools and enhancing development efficiency.
For example, the Yi series models are saved in the format of the Llama model. You can directly use `LlamaForCausalLM` and `LlamaTokenizer` to load the model. For more information, see [Use the chat model](#31-use-the-chat-model).
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("01-ai/Yi-34b", use_fast=False)
model = AutoModelForCausalLM.from_pretrained("01-ai/Yi-34b", device_map="auto")
```
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
### Downstream
> 💡 Tip
>
> - Feel free to create a PR and share the fantastic work you've built using the Yi series models.
>
> - To help others quickly understand your work, it is recommended to use the format of `<model-name>: <model-intro> + <model-highlights>`.
#### Serving
If you want to get up with Yi in a few minutes, you can use the following services built upon Yi.
- Yi-34B-Chat: you can chat with Yi using one of the following platforms:
- [Yi-34B-Chat | Hugging Face](https://huggingface.co/spaces/01-ai/Yi-34B-Chat)
- [Yi-34B-Chat | Yi Platform](https://platform.lingyiwanwu.com/): **Note** that currently it's available through a whitelist. Welcome to apply (fill out a form in [English](https://cn.mikecrm.com/l91ODJf) or [Chinese](https://cn.mikecrm.com/gnEZjiQ)) and experience it firsthand!
- [Yi-6B-Chat (Replicate)](https://replicate.com/01-ai): you can use this model with more options by setting additional parameters and calling APIs.
- [ScaleLLM](https://github.com/vectorch-ai/ScaleLLM#supported-models): you can use this service to run Yi models locally with added flexibility and customization.
#### Quantization
If you have limited computational capabilities, you can use Yi's quantized models as follows.
These quantized models have reduced precision but offer increased efficiency, such as faster inference speed and smaller RAM usage.
- [TheBloke/Yi-34B-GPTQ](https://huggingface.co/TheBloke/Yi-34B-GPTQ)
- [TheBloke/Yi-34B-GGUF](https://huggingface.co/TheBloke/Yi-34B-GGUF)
- [TheBloke/Yi-34B-AWQ](https://huggingface.co/TheBloke/Yi-34B-AWQ)
#### Fine-tuning
If you're seeking to explore the diverse capabilities within Yi's thriving family, you can delve into Yi's fine-tuned models as below.
- [TheBloke Models](https://huggingface.co/TheBloke): this site hosts numerous fine-tuned models derived from various LLMs including Yi.
This is not an exhaustive list for Yi, but to name a few sorted on downloads:
- [TheBloke/dolphin-2_2-yi-34b-AWQ](https://huggingface.co/TheBloke/dolphin-2_2-yi-34b-AWQ)
- [TheBloke/Yi-34B-Chat-AWQ](https://huggingface.co/TheBloke/Yi-34B-Chat-AWQ)
- [TheBloke/Yi-34B-Chat-GPTQ](https://huggingface.co/TheBloke/Yi-34B-Chat-GPTQ)
- [SUSTech/SUS-Chat-34B](https://huggingface.co/SUSTech/SUS-Chat-34B): this model ranked first among all models below 70B and outperformed the twice larger deepseek-llm-67b-chat. You can check the result on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).
- [OrionStarAI/OrionStar-Yi-34B-Chat-Llama](https://huggingface.co/OrionStarAI/OrionStar-Yi-34B-Chat-Llama): this model excelled beyond other models (such as GPT-4, Qwen-14B-Chat, Baichuan2-13B-Chat) in C-Eval and CMMLU evaluations on the [OpenCompass LLM Leaderboard](https://opencompass.org.cn/leaderboard-llm).
- [NousResearch/Nous-Capybara-34B](https://huggingface.co/NousResearch/Nous-Capybara-34B): this model is trained with 200K context length and 3 epochs on the Capybara dataset.
#### API
- [amazing-openai-api](https://github.com/soulteary/amazing-openai-api): this tool converts Yi model APIs into the OpenAI API format out of the box.
- [LlamaEdge](https://www.secondstate.io/articles/yi-34b/#create-an-openai-compatible-api-service-for-the-yi-34b-chat-model): this tool builds an OpenAI-compatible API server for Yi-34B-Chat using a portable Wasm (WebAssembly) file, powered by Rust.
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
## Tech report
For detailed capabilities of the Yi series model, see [Yi: Open Foundation Models by 01.AI](https://arxiv.org/abs/2403.04652).
### Citation
```
@misc{ai2024yi,
title={Yi: Open Foundation Models by 01.AI},
author={01. AI and : and Alex Young and Bei Chen and Chao Li and Chengen Huang and Ge Zhang and Guanwei Zhang and Heng Li and Jiangcheng Zhu and Jianqun Chen and Jing Chang and Kaidong Yu and Peng Liu and Qiang Liu and Shawn Yue and Senbin Yang and Shiming Yang and Tao Yu and Wen Xie and Wenhao Huang and Xiaohui Hu and Xiaoyi Ren and Xinyao Niu and Pengcheng Nie and Yuchi Xu and Yudong Liu and Yue Wang and Yuxuan Cai and Zhenyu Gu and Zhiyuan Liu and Zonghong Dai},
year={2024},
eprint={2403.04652},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
## Benchmarks
- [Chat model performance](#chat-model-performance)
- [Base model performance](#base-model-performance)
### Chat model performance
Yi-34B-Chat model demonstrates exceptional performance, ranking first among all existing open-source models in the benchmarks including MMLU, CMMLU, BBH, GSM8k, and more.
<details>
<summary> Evaluation methods and challenges. ⬇️ </summary>
- **Evaluation methods**: we evaluated various benchmarks using both zero-shot and few-shot methods, except for TruthfulQA.
- **Zero-shot vs. few-shot**: in chat models, the zero-shot approach is more commonly employed.
- **Evaluation strategy**: our evaluation strategy involves generating responses while following instructions explicitly or implicitly (such as using few-shot examples). We then isolate relevant answers from the generated text.
- **Challenges faced**: some models are not well-suited to produce output in the specific format required by instructions in few datasets, which leads to suboptimal results.
<strong>*</strong>: C-Eval results are evaluated on the validation datasets
</details>
### Base model performance
#### Yi-34B and Yi-34B-200K
The Yi-34B and Yi-34B-200K models stand out as the top performers among open-source models, especially excelling in MMLU, CMMLU, common-sense reasoning, reading comprehension, and more.
<details>
<summary> Evaluation methods. ⬇️</summary>
- **Disparity in results**: while benchmarking open-source models, a disparity has been noted between results from our pipeline and those reported by public sources like OpenCompass.
- **Investigation findings**: a deeper investigation reveals that variations in prompts, post-processing strategies, and sampling techniques across models may lead to significant outcome differences.
- **Uniform benchmarking process**: our methodology aligns with the original benchmarks—consistent prompts and post-processing strategies are used, and greedy decoding is applied during evaluations without any post-processing for the generated content.
- **Efforts to retrieve unreported scores**: for scores that were not reported by the original authors (including scores reported with different settings), we try to get results with our pipeline.
- **Extensive model evaluation**: to evaluate the model’s capability extensively, we adopted the methodology outlined in Llama2. Specifically, we included PIQA, SIQA, HellaSwag, WinoGrande, ARC, OBQA, and CSQA to assess common sense reasoning. SquAD, QuAC, and BoolQ were incorporated to evaluate reading comprehension.
- **Special configurations**: CSQA was exclusively tested using a 7-shot setup, while all other tests were conducted with a 0-shot configuration. Additionally, we introduced GSM8K (8-shot@1), MATH (4-shot@1), HumanEval (0-shot@1), and MBPP (3-shot@1) under the category "Math & Code".
- **Falcon-180B caveat**: Falcon-180B was not tested on QuAC and OBQA due to technical constraints. Its performance score is an average from other tasks, and considering the generally lower scores of these two tasks, Falcon-180B's capabilities are likely not underestimated.
</details>
#### Yi-9B
Yi-9B is almost the best among a range of similar-sized open-source models (including Mistral-7B, SOLAR-10.7B, Gemma-7B, DeepSeek-Coder-7B-Base-v1.5 and more), particularly excelling in code, math, common-sense reasoning, and reading comprehension.
- In terms of **overall** ability (Mean-All), Yi-9B performs the best among similarly sized open-source models, surpassing DeepSeek-Coder, DeepSeek-Math, Mistral-7B, SOLAR-10.7B, and Gemma-7B.
- In terms of **coding** ability (Mean-Code), Yi-9B's performance is second only to DeepSeek-Coder-7B, surpassing Yi-34B, SOLAR-10.7B, Mistral-7B, and Gemma-7B.
- In terms of **math** ability (Mean-Math), Yi-9B's performance is second only to DeepSeek-Math-7B, surpassing SOLAR-10.7B, Mistral-7B, and Gemma-7B.
- In terms of **common sense and reasoning** ability (Mean-Text), Yi-9B's performance is on par with Mistral-7B, SOLAR-10.7B, and Gemma-7B.
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
# Who can use Yi?
Everyone! 🙌 ✅
The code and weights of the Yi series models are distributed under the [Apache 2.0 license](https://github.com/01-ai/Yi/blob/main/LICENSE), which means the Yi series models are free for personal usage, academic purposes, and commercial use.
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
# Misc.
### Acknowledgments
A heartfelt thank you to each of you who have made contributions to the Yi community! You have helped Yi not just a project, but a vibrant, growing home for innovation.
[](https://github.com/01-ai/yi/graphs/contributors)
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
### Disclaimer
We use data compliance checking algorithms during the training process, to
ensure the compliance of the trained model to the best of our ability. Due to
complex data and the diversity of language model usage scenarios, we cannot
guarantee that the model will generate correct, and reasonable output in all
scenarios. Please be aware that there is still a risk of the model producing
problematic outputs. We will not be responsible for any risks and issues
resulting from misuse, misguidance, illegal usage, and related misinformation,
as well as any associated data security concerns.
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
### License
The code and weights of the Yi-1.5 series models are distributed under the [Apache 2.0 license](https://github.com/01-ai/Yi/blob/main/LICENSE).
If you create derivative works based on this model, please include the following attribution in your derivative works:
This work is a derivative of [The Yi Series Model You Base On] by 01.AI, used under the Apache 2.0 License.
<p align="right"> [
<a href="#top">Back to top ⬆️ </a> ]
</p>
|
konverner/distilcamembert-base-ner-address | konverner | "2023-09-15T19:34:57Z" | 8,684 | 0 | transformers | [
"transformers",
"pytorch",
"camembert",
"token-classification",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | token-classification | "2023-09-13T09:30:31Z" | Entry not found |
stabilityai/stable-diffusion-2-1-unclip | stabilityai | "2023-04-12T15:49:10Z" | 8,681 | 259 | diffusers | [
"diffusers",
"safetensors",
"stable-diffusion",
"text-to-image",
"arxiv:2112.10752",
"arxiv:1910.09700",
"license:openrail++",
"diffusers:StableUnCLIPImg2ImgPipeline",
"region:us"
] | text-to-image | "2023-03-20T13:11:38Z" | ---
license: openrail++
tags:
- stable-diffusion
- text-to-image
pinned: true
---
# Stable Diffusion v2-1-unclip Model Card
This model card focuses on the model associated with the Stable Diffusion v2-1 model, codebase available [here](https://github.com/Stability-AI/stablediffusion).
This `stable-diffusion-2-1-unclip` is a finetuned version of Stable Diffusion 2.1, modified to accept (noisy) CLIP image embedding in addition to the text prompt, and can be used to create image variations (Examples) or can be chained with text-to-image CLIP priors. The amount of noise added to the image embedding can be specified via the noise_level (0 means no noise, 1000 full noise).
- Use it with 🧨 [`diffusers`](#examples)
## Model Details
- **Developed by:** Robin Rombach, Patrick Esser
- **Model type:** Diffusion-based text-to-image generation model
- **Language(s):** English
- **License:** [CreativeML Open RAIL++-M License](https://huggingface.co/stabilityai/stable-diffusion-2/blob/main/LICENSE-MODEL)
- **Model Description:** This is a model that can be used to generate and modify images based on text prompts. It is a [Latent Diffusion Model](https://arxiv.org/abs/2112.10752) that uses a fixed, pretrained text encoder ([OpenCLIP-ViT/H](https://github.com/mlfoundations/open_clip)).
- **Resources for more information:** [GitHub Repository](https://github.com/Stability-AI/).
- **Cite as:**
@InProceedings{Rombach_2022_CVPR,
author = {Rombach, Robin and Blattmann, Andreas and Lorenz, Dominik and Esser, Patrick and Ommer, Bj\"orn},
title = {High-Resolution Image Synthesis With Latent Diffusion Models},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2022},
pages = {10684-10695}
}
## Examples
Using the [🤗's Diffusers library](https://github.com/huggingface/diffusers) to run Stable Diffusion UnCLIP 2-1-small in a simple and efficient manner.
```bash
pip install diffusers transformers accelerate scipy safetensors
```
Running the pipeline (if you don't swap the scheduler it will run with the default DDIM, in this example we are swapping it to DPMSolverMultistepScheduler):
```python
from diffusers import DiffusionPipeline
from diffusers.utils import load_image
import torch
pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-2-1-unclip-small", torch_dtype=torch.float16)
pipe.to("cuda")
# get image
url = "https://huggingface.co/datasets/hf-internal-testing/diffusers-images/resolve/main/stable_unclip/tarsila_do_amaral.png"
image = load_image(url)
# run image variation
image = pipe(image).images[0]
```
# Uses
## Direct Use
The model is intended for research purposes only. Possible research areas and tasks include
- Safe deployment of models which have the potential to generate harmful content.
- Probing and understanding the limitations and biases of generative models.
- Generation of artworks and use in design and other artistic processes.
- Applications in educational or creative tools.
- Research on generative models.
Excluded uses are described below.
### Misuse, Malicious Use, and Out-of-Scope Use
_Note: This section is originally taken from the [DALLE-MINI model card](https://huggingface.co/dalle-mini/dalle-mini), was used for Stable Diffusion v1, but applies in the same way to Stable Diffusion v2_.
The model should not be used to intentionally create or disseminate images that create hostile or alienating environments for people. This includes generating images that people would foreseeably find disturbing, distressing, or offensive; or content that propagates historical or current stereotypes.
#### Out-of-Scope Use
The model was not trained to be factual or true representations of people or events, and therefore using the model to generate such content is out-of-scope for the abilities of this model.
#### Misuse and Malicious Use
Using the model to generate content that is cruel to individuals is a misuse of this model. This includes, but is not limited to:
- Generating demeaning, dehumanizing, or otherwise harmful representations of people or their environments, cultures, religions, etc.
- Intentionally promoting or propagating discriminatory content or harmful stereotypes.
- Impersonating individuals without their consent.
- Sexual content without consent of the people who might see it.
- Mis- and disinformation
- Representations of egregious violence and gore
- Sharing of copyrighted or licensed material in violation of its terms of use.
- Sharing content that is an alteration of copyrighted or licensed material in violation of its terms of use.
## Limitations and Bias
### Limitations
- The model does not achieve perfect photorealism
- The model cannot render legible text
- The model does not perform well on more difficult tasks which involve compositionality, such as rendering an image corresponding to “A red cube on top of a blue sphere”
- Faces and people in general may not be generated properly.
- The model was trained mainly with English captions and will not work as well in other languages.
- The autoencoding part of the model is lossy
- The model was trained on a subset of the large-scale dataset
[LAION-5B](https://laion.ai/blog/laion-5b/), which contains adult, violent and sexual content. To partially mitigate this, we have filtered the dataset using LAION's NFSW detector (see Training section).
### Bias
While the capabilities of image generation models are impressive, they can also reinforce or exacerbate social biases.
Stable Diffusion was primarily trained on subsets of [LAION-2B(en)](https://laion.ai/blog/laion-5b/),
which consists of images that are limited to English descriptions.
Texts and images from communities and cultures that use other languages are likely to be insufficiently accounted for.
This affects the overall output of the model, as white and western cultures are often set as the default. Further, the
ability of the model to generate content with non-English prompts is significantly worse than with English-language prompts.
Stable Diffusion v2 mirrors and exacerbates biases to such a degree that viewer discretion must be advised irrespective of the input or its intent.
## Training
**Training Data**
The model developers used the following dataset for training the model:
- LAION-5B and subsets (details below). The training data is further filtered using LAION's NSFW detector, with a "p_unsafe" score of 0.1 (conservative). For more details, please refer to LAION-5B's [NeurIPS 2022](https://openreview.net/forum?id=M3Y74vmsMcY) paper and reviewer discussions on the topic.
## Environmental Impact
**Stable Diffusion v1** **Estimated Emissions**
Based on that information, we estimate the following CO2 emissions using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). The hardware, runtime, cloud provider, and compute region were utilized to estimate the carbon impact.
- **Hardware Type:** A100 PCIe 40GB
- **Hours used:** 200000
- **Cloud Provider:** AWS
- **Compute Region:** US-east
- **Carbon Emitted (Power consumption x Time x Carbon produced based on location of power grid):** 15000 kg CO2 eq.
## Citation
@InProceedings{Rombach_2022_CVPR,
author = {Rombach, Robin and Blattmann, Andreas and Lorenz, Dominik and Esser, Patrick and Ommer, Bj\"orn},
title = {High-Resolution Image Synthesis With Latent Diffusion Models},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2022},
pages = {10684-10695}
}
*This model card was written by: Robin Rombach, Patrick Esser and David Ha and is based on the [Stable Diffusion v1](https://github.com/CompVis/stable-diffusion/blob/main/Stable_Diffusion_v1_Model_Card.md) and [DALL-E Mini model card](https://huggingface.co/dalle-mini/dalle-mini).* |
scaleszw/scales_ai | scaleszw | "2024-07-02T01:28:15Z" | 8,676 | 0 | peft | [
"peft",
"pytorch",
"tensorboard",
"safetensors",
"gguf",
"sn",
"dataset:scaleszw/scales_shona_ai",
"arxiv:1910.09700",
"base_model:unsloth/llama-3-8b-bnb-4bit",
"license:llama3",
"region:us"
] | null | "2024-05-23T13:28:06Z" | ---
base_model: unsloth/llama-3-8b-bnb-4bit
library_name: peft
license: llama3
datasets:
- scaleszw/scales_shona_ai
language:
- sn
---
# Model Card for Model ID
1. **Natural Language Understanding and Generation**: Scales AI excels in understanding and generating human-like text based on user input, utilizing the latest advancements in natural language processing.
2. **Information Retrieval**: Scales AI is capable of performing web searches to fetch information, utilizing the Google Custom Search API to provide users with up-to-date and relevant information from the web.
3. **Entity Recognition and Tracking**: Scales AI can identify and keep track of key entities mentioned during conversations, allowing for context-aware responses.
4. **Memory of Conversation History**: Scales AI can maintain a history of the ongoing conversation to ensure continuity and relevance in responses.
5. **Error Handling and Robustness**: Scales AI is designed to handle errors gracefully, providing meaningful feedback to users in case of issues and continuing the conversation without interruptions.
6. **Shona speaking**: Scales AI able to have conversations in the Shona language and also take an input in Shona language and perform a web search to provide the users with accurate, relevant and insightful responses.
## Model Details
### Model Description
Scales AI is a large language model that understands shona language better than other models
- **Developed by:** [Ronald Bvirinyangwe]
- **Funded by [optional]:** [More Information Needed]
- **Shared by [optional]:** [Ronald Bvirinyangwe]
- **Model type:** [Text-generation]
- **Language(s) (NLP):** [English,Shona]
- **License:** [llama3]
- **Finetuned from model [optional]:** [llama-3-8b-bnb-4bit]
### Model Sources [optional]
<!-- Provide the basic links for the model. -->
- **Repository:** [scales_ai]
- **Paper [optional]:** [More Information Needed]
- **Demo [optional]:** [More Information Needed]
## Uses
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
### Direct Use
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
[More Information Needed]
### Downstream Use [optional]
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
[More Information Needed]
### Out-of-Scope Use
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
[More Information Needed]
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
[More Information Needed]
### Recommendations
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
## How to Get Started with the Model
Use the code below to get started with the model.
[More Information Needed]
## Training Details
### Training Data
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
[More Information Needed]
### Training Procedure
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
#### Preprocessing [optional]
[More Information Needed]
#### Training Hyperparameters
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
#### Speeds, Sizes, Times [optional]
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
[More Information Needed]
## Evaluation
<!-- This section describes the evaluation protocols and provides the results. -->
### Testing Data, Factors & Metrics
#### Testing Data
<!-- This should link to a Dataset Card if possible. -->
[More Information Needed]
#### Factors
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
[More Information Needed]
#### Metrics
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
[More Information Needed]
### Results
[More Information Needed]
#### Summary
## Model Examination [optional]
<!-- Relevant interpretability work for the model goes here -->
[More Information Needed]
## Environmental Impact
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
- **Hardware Type:** [More Information Needed]
- **Hours used:** [More Information Needed]
- **Cloud Provider:** [More Information Needed]
- **Compute Region:** [More Information Needed]
- **Carbon Emitted:** [More Information Needed]
## Technical Specifications [optional]
### Model Architecture and Objective
[More Information Needed]
### Compute Infrastructure
[More Information Needed]
#### Hardware
[More Information Needed]
#### Software
[More Information Needed]
## Citation [optional]
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
[More Information Needed]
**APA:**
[More Information Needed]
## Glossary [optional]
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
[More Information Needed]
## More Information [optional]
[More Information Needed]
## Model Card Authors [optional]
[More Information Needed]
## Model Card Contact
[More Information Needed]
### Framework versions
- PEFT 0.11.1 |
studio-ousia/luke-base | studio-ousia | "2022-04-13T08:59:59Z" | 8,672 | 20 | transformers | [
"transformers",
"pytorch",
"luke",
"fill-mask",
"named entity recognition",
"entity typing",
"relation classification",
"question answering",
"en",
"arxiv:1906.08237",
"arxiv:1903.07785",
"arxiv:2002.01808",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | fill-mask | "2022-03-02T23:29:05Z" | ---
language: en
thumbnail: https://github.com/studio-ousia/luke/raw/master/resources/luke_logo.png
tags:
- luke
- named entity recognition
- entity typing
- relation classification
- question answering
license: apache-2.0
---
## LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention
**LUKE** (**L**anguage **U**nderstanding with **K**nowledge-based
**E**mbeddings) is a new pre-trained contextualized representation of words and
entities based on transformer. LUKE treats words and entities in a given text as
independent tokens, and outputs contextualized representations of them. LUKE
adopts an entity-aware self-attention mechanism that is an extension of the
self-attention mechanism of the transformer, and considers the types of tokens
(words or entities) when computing attention scores.
LUKE achieves state-of-the-art results on five popular NLP benchmarks including
**[SQuAD v1.1](https://rajpurkar.github.io/SQuAD-explorer/)** (extractive
question answering),
**[CoNLL-2003](https://www.clips.uantwerpen.be/conll2003/ner/)** (named entity
recognition), **[ReCoRD](https://sheng-z.github.io/ReCoRD-explorer/)**
(cloze-style question answering),
**[TACRED](https://nlp.stanford.edu/projects/tacred/)** (relation
classification), and
**[Open Entity](https://www.cs.utexas.edu/~eunsol/html_pages/open_entity.html)**
(entity typing).
Please check the [official repository](https://github.com/studio-ousia/luke) for
more details and updates.
This is the LUKE base model with 12 hidden layers, 768 hidden size. The total number
of parameters in this model is 253M. It is trained using December 2018 version of
Wikipedia.
### Experimental results
The experimental results are provided as follows:
| Task | Dataset | Metric | LUKE-large | luke-base | Previous SOTA |
| ------------------------------ | ---------------------------------------------------------------------------- | ------ | ----------------- | --------- | ------------------------------------------------------------------------- |
| Extractive Question Answering | [SQuAD v1.1](https://rajpurkar.github.io/SQuAD-explorer/) | EM/F1 | **90.2**/**95.4** | 86.1/92.3 | 89.9/95.1 ([Yang et al., 2019](https://arxiv.org/abs/1906.08237)) |
| Named Entity Recognition | [CoNLL-2003](https://www.clips.uantwerpen.be/conll2003/ner/) | F1 | **94.3** | 93.3 | 93.5 ([Baevski et al., 2019](https://arxiv.org/abs/1903.07785)) |
| Cloze-style Question Answering | [ReCoRD](https://sheng-z.github.io/ReCoRD-explorer/) | EM/F1 | **90.6**/**91.2** | - | 83.1/83.7 ([Li et al., 2019](https://www.aclweb.org/anthology/D19-6011/)) |
| Relation Classification | [TACRED](https://nlp.stanford.edu/projects/tacred/) | F1 | **72.7** | - | 72.0 ([Wang et al. , 2020](https://arxiv.org/abs/2002.01808)) |
| Fine-grained Entity Typing | [Open Entity](https://www.cs.utexas.edu/~eunsol/html_pages/open_entity.html) | F1 | **78.2** | - | 77.6 ([Wang et al. , 2020](https://arxiv.org/abs/2002.01808)) |
### Citation
If you find LUKE useful for your work, please cite the following paper:
```latex
@inproceedings{yamada2020luke,
title={LUKE: Deep Contextualized Entity Representations with Entity-aware Self-attention},
author={Ikuya Yamada and Akari Asai and Hiroyuki Shindo and Hideaki Takeda and Yuji Matsumoto},
booktitle={EMNLP},
year={2020}
}
```
|
Helsinki-NLP/opus-mt-hu-en | Helsinki-NLP | "2023-08-16T11:57:55Z" | 8,669 | 1 | transformers | [
"transformers",
"pytorch",
"tf",
"marian",
"text2text-generation",
"translation",
"hu",
"en",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | translation | "2022-03-02T23:29:04Z" | ---
tags:
- translation
license: apache-2.0
---
### opus-mt-hu-en
* source languages: hu
* target languages: en
* OPUS readme: [hu-en](https://github.com/Helsinki-NLP/OPUS-MT-train/blob/master/models/hu-en/README.md)
* dataset: opus
* model: transformer-align
* pre-processing: normalization + SentencePiece
* download original weights: [opus-2019-12-18.zip](https://object.pouta.csc.fi/OPUS-MT-models/hu-en/opus-2019-12-18.zip)
* test set translations: [opus-2019-12-18.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/hu-en/opus-2019-12-18.test.txt)
* test set scores: [opus-2019-12-18.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/hu-en/opus-2019-12-18.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| Tatoeba.hu.en | 52.9 | 0.683 |
|
beomi/llama-2-ko-7b | beomi | "2023-12-27T02:04:01Z" | 8,669 | 172 | transformers | [
"transformers",
"pytorch",
"safetensors",
"llama",
"text-generation",
"facebook",
"meta",
"llama-2",
"kollama",
"llama-2-ko",
"en",
"ko",
"doi:10.57967/hf/1098",
"autotrain_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2023-07-20T03:25:25Z" | ---
language:
- en
- ko
pipeline_tag: text-generation
inference: false
tags:
- facebook
- meta
- pytorch
- llama
- llama-2
- kollama
- llama-2-ko
---
**Update Log**
- 2023.12.27
- New Model is here! Trained with only open-accessible Korean text corpus: https://huggingface.co/beomi/open-llama-2-ko-7b
- 2023.10.19
- Fix Tokenizer bug(space not applied when decoding) after `transforemrs>=4.34.0`
# **Llama-2-Ko** 🦙🇰🇷
Llama-2-Ko serves as an advanced iteration of Llama 2, benefiting from an expanded vocabulary and the inclusion of a Korean corpus in its further pretraining. Just like its predecessor, Llama-2-Ko operates within the broad range of generative text models that stretch from 7 billion to 70 billion parameters. This repository focuses on the 7B pretrained version, which is tailored to fit the Hugging Face Transformers format. For access to the other models, feel free to consult the index provided below.
## Model Details
**Model Developers** Junbum Lee (Beomi)
**Variations** Llama-2-Ko will come in a range of parameter sizes — 7B, 13B, and 70B — as well as pretrained and fine-tuned variations.
**Input** Models input text only.
**Output** Models generate text only.
**Model Architecture**
Llama-2-Ko is an auto-regressive language model that uses an optimized transformer architecture based on Llama-2.
||Training Data|Params|Content Length|GQA|Tokens|LR|
|---|---|---|---|---|---|---|
|Llama 2|*A new mix of Korean online data*|7B|4k|✗|>40B*|1e<sup>-5</sup>|
*Plan to train upto 200B tokens
**Vocab Expansion**
| Model Name | Vocabulary Size | Description |
| --- | --- | --- |
| Original Llama-2 | 32000 | Sentencepiece BPE |
| **Expanded Llama-2-Ko** | 46336 | Sentencepiece BPE. Added Korean vocab and merges |
**Tokenizing "안녕하세요, 오늘은 날씨가 좋네요."**
| Model | Tokens |
| --- | --- |
| Llama-2 | `['▁', '안', '<0xEB>', '<0x85>', '<0x95>', '하', '세', '요', ',', '▁', '오', '<0xEB>', '<0x8A>', '<0x98>', '은', '▁', '<0xEB>', '<0x82>', '<0xA0>', '씨', '가', '▁', '<0xEC>', '<0xA2>', '<0x8B>', '<0xEB>', '<0x84>', '<0xA4>', '요']` |
| Llama-2-Ko | `['▁안녕', '하세요', ',', '▁오늘은', '▁날', '씨가', '▁좋네요']` |
**Tokenizing "Llama 2: Open Foundation and Fine-Tuned Chat Models"**
| Model | Tokens |
| --- | --- |
| Llama-2 | `['▁L', 'l', 'ama', '▁', '2', ':', '▁Open', '▁Foundation', '▁and', '▁Fine', '-', 'T', 'un', 'ed', '▁Ch', 'at', '▁Mod', 'els']` |
| Llama-2-Ko | `['▁L', 'l', 'ama', '▁', '2', ':', '▁Open', '▁Foundation', '▁and', '▁Fine', '-', 'T', 'un', 'ed', '▁Ch', 'at', '▁Mod', 'els']` |
# **Model Benchmark**
## LM Eval Harness - Korean (polyglot branch)
- Used EleutherAI's lm-evaluation-harness https://github.com/EleutherAI/lm-evaluation-harness/tree/polyglot
### NSMC (Acc) - 50000 full test
TBD
### COPA (F1)
<img src=https://user-images.githubusercontent.com/11323660/255575809-c037bc6e-0566-436a-a6c1-2329ac92187a.png style="max-width: 700px; width: 100%" />
| Model | 0-shot | 5-shot | 10-shot | 50-shot |
| --- | --- | --- | --- | --- |
| https://huggingface.co/skt/ko-gpt-trinity-1.2B-v0.5 | 0.6696 | 0.6477 | 0.6419 | 0.6514 |
| https://huggingface.co/kakaobrain/kogpt | 0.7345 | 0.7287 | 0.7277 | 0.7479 |
| https://huggingface.co/facebook/xglm-7.5B | 0.6723 | 0.6731 | 0.6769 | 0.7119 |
| https://huggingface.co/EleutherAI/polyglot-ko-1.3b | 0.7196 | 0.7193 | 0.7204 | 0.7206 |
| https://huggingface.co/EleutherAI/polyglot-ko-3.8b | 0.7595 | 0.7608 | 0.7638 | 0.7788 |
| https://huggingface.co/EleutherAI/polyglot-ko-5.8b | 0.7745 | 0.7676 | 0.7775 | 0.7887 |
| https://huggingface.co/EleutherAI/polyglot-ko-12.8b | 0.7937 | 0.8108 | 0.8037 | 0.8369 |
| Llama-2 Original 7B* | 0.562033 | 0.575982 | 0.576216 | 0.595532 |
| Llama-2-Ko-7b 20B (10k) | 0.738780 | 0.762639 | 0.780761 | 0.797863 |
| Llama-2-Ko-7b 40B (20k) | 0.743630 | 0.792716 | 0.803746 | 0.825944 |
*Llama-2 Original 7B used https://huggingface.co/meta-llama/Llama-2-7b-hf (w/o tokenizer updated)
### HellaSwag (F1)
<img src=https://user-images.githubusercontent.com/11323660/255576090-a2bfc1ae-d117-44b7-9f7b-262e41179ec1.png style="max-width: 700px; width: 100%" />
| Model | 0-shot | 5-shot | 10-shot | 50-shot |
| --- | --- | --- | --- | --- |
| https://huggingface.co/skt/ko-gpt-trinity-1.2B-v0.5 | 0.5243 | 0.5272 | 0.5166 | 0.5352 |
| https://huggingface.co/kakaobrain/kogpt | 0.5590 | 0.5833 | 0.5828 | 0.5907 |
| https://huggingface.co/facebook/xglm-7.5B | 0.5665 | 0.5689 | 0.5565 | 0.5622 |
| https://huggingface.co/EleutherAI/polyglot-ko-1.3b | 0.5247 | 0.5260 | 0.5278 | 0.5427 |
| https://huggingface.co/EleutherAI/polyglot-ko-3.8b | 0.5707 | 0.5830 | 0.5670 | 0.5787 |
| https://huggingface.co/EleutherAI/polyglot-ko-5.8b | 0.5976 | 0.5998 | 0.5979 | 0.6208 |
| https://huggingface.co/EleutherAI/polyglot-ko-12.8b | 0.5954 | 0.6306 | 0.6098 | 0.6118 |
| Llama-2 Original 7B* | 0.415390 | 0.431382 | 0.421342 | 0.442003 |
| Llama-2-Ko-7b 20B (10k) | 0.451757 | 0.466751 | 0.472607 | 0.482776 |
| Llama-2-Ko-7b 40B (20k) | 0.456246 | 0.465665 | 0.469810 | 0.477374 |
*Llama-2 Original 7B used https://huggingface.co/meta-llama/Llama-2-7b-hf (w/o tokenizer updated)
### BoolQ (F1)
<img src=https://user-images.githubusercontent.com/11323660/255576343-5d847a6f-3b6a-41a7-af37-0f11940a5ea4.png style="max-width: 700px; width: 100%" />
| Model | 0-shot | 5-shot | 10-shot | 50-shot |
| --- | --- | --- | --- | --- |
| https://huggingface.co/skt/ko-gpt-trinity-1.2B-v0.5 | 0.3356 | 0.4014 | 0.3640 | 0.3560 |
| https://huggingface.co/kakaobrain/kogpt | 0.4514 | 0.5981 | 0.5499 | 0.5202 |
| https://huggingface.co/facebook/xglm-7.5B | 0.4464 | 0.3324 | 0.3324 | 0.3324 |
| https://huggingface.co/EleutherAI/polyglot-ko-1.3b | 0.3552 | 0.4751 | 0.4109 | 0.4038 |
| https://huggingface.co/EleutherAI/polyglot-ko-3.8b | 0.4320 | 0.5263 | 0.4930 | 0.4038 |
| https://huggingface.co/EleutherAI/polyglot-ko-5.8b | 0.4356 | 0.5698 | 0.5187 | 0.5236 |
| https://huggingface.co/EleutherAI/polyglot-ko-12.8b | 0.4818 | 0.6041 | 0.6289 | 0.6448 |
| Llama-2 Original 7B* | 0.352050 | 0.563238 | 0.474788 | 0.419222 |
| Llama-2-Ko-7b 20B (10k) | 0.360656 | 0.679743 | 0.680109 | 0.662152 |
| Llama-2-Ko-7b 40B (20k) | 0.578640 | 0.697747 | 0.708358 | 0.714423 |
*Llama-2 Original 7B used https://huggingface.co/meta-llama/Llama-2-7b-hf (w/o tokenizer updated)
### SentiNeg (F1)
<img src=https://user-images.githubusercontent.com/11323660/255576572-b005a81d-fa4d-4709-b48a-f0fe4eed17a3.png style="max-width: 700px; width: 100%" />
| Model | 0-shot | 5-shot | 10-shot | 50-shot |
| --- | --- | --- | --- | --- |
| https://huggingface.co/skt/ko-gpt-trinity-1.2B-v0.5 | 0.6065 | 0.6878 | 0.7280 | 0.8413 |
| https://huggingface.co/kakaobrain/kogpt | 0.3747 | 0.8942 | 0.9294 | 0.9698 |
| https://huggingface.co/facebook/xglm-7.5B | 0.3578 | 0.4471 | 0.3964 | 0.5271 |
| https://huggingface.co/EleutherAI/polyglot-ko-1.3b | 0.6790 | 0.6257 | 0.5514 | 0.7851 |
| https://huggingface.co/EleutherAI/polyglot-ko-3.8b | 0.4858 | 0.7950 | 0.7320 | 0.7851 |
| https://huggingface.co/EleutherAI/polyglot-ko-5.8b | 0.3394 | 0.8841 | 0.8808 | 0.9521 |
| https://huggingface.co/EleutherAI/polyglot-ko-12.8b | 0.9117 | 0.9015 | 0.9345 | 0.9723 |
| Llama-2 Original 7B* | 0.347502 | 0.529124 | 0.480641 | 0.788457 |
| Llama-2-Ko-7b 20B (10k) | 0.485546 | 0.829503 | 0.871141 | 0.851253 |
| Llama-2-Ko-7b 40B (20k) | 0.459447 | 0.761079 | 0.727611 | 0.936988 |
*Llama-2 Original 7B used https://huggingface.co/meta-llama/Llama-2-7b-hf (w/o tokenizer updated)
## Note for oobabooga/text-generation-webui
Remove `ValueError` at `load_tokenizer` function(line 109 or near), in `modules/models.py`.
```python
diff --git a/modules/models.py b/modules/models.py
index 232d5fa..de5b7a0 100644
--- a/modules/models.py
+++ b/modules/models.py
@@ -106,7 +106,7 @@ def load_tokenizer(model_name, model):
trust_remote_code=shared.args.trust_remote_code,
use_fast=False
)
- except ValueError:
+ except:
tokenizer = AutoTokenizer.from_pretrained(
path_to_model,
trust_remote_code=shared.args.trust_remote_code,
```
Since Llama-2-Ko uses FastTokenizer provided by HF tokenizers NOT sentencepiece package,
it is required to use `use_fast=True` option when initialize tokenizer.
Apple Sillicon does not support BF16 computing, use CPU instead. (BF16 is supported when using NVIDIA GPU)
## Citation
```
@misc {l._junbum_2023,
author = { {L. Junbum} },
title = { llama-2-ko-7b (Revision 4a9993e) },
year = 2023,
url = { https://huggingface.co/beomi/llama-2-ko-7b },
doi = { 10.57967/hf/1098 },
publisher = { Hugging Face }
}
```
## Acknowledgement
The training is supported by [TPU Research Cloud](https://sites.research.google/trc/) program.
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_beomi__llama-2-ko-7b)
| Metric | Value |
|-----------------------|---------------------------|
| Avg. | 39.43 |
| ARC (25-shot) | 48.46 |
| HellaSwag (10-shot) | 75.28 |
| MMLU (5-shot) | 39.56 |
| TruthfulQA (0-shot) | 34.49 |
| Winogrande (5-shot) | 72.14 |
| GSM8K (5-shot) | 1.97 |
| DROP (3-shot) | 4.1 |
|
jackaduma/SecRoBERTa | jackaduma | "2023-06-26T05:55:27Z" | 8,666 | 14 | transformers | [
"transformers",
"pytorch",
"safetensors",
"roberta",
"fill-mask",
"exbert",
"security",
"cybersecurity",
"cyber security",
"threat hunting",
"threat intelligence",
"en",
"dataset:APTnotes",
"dataset:Stucco-Data",
"dataset:CASIE",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | fill-mask | "2022-03-02T23:29:05Z" | ---
language: en
thumbnail: https://github.com/jackaduma
tags:
- exbert
- security
- cybersecurity
- cyber security
- threat hunting
- threat intelligence
license: apache-2.0
datasets:
- APTnotes
- Stucco-Data
- CASIE
---
# SecRoBERTa
This is the pretrained model presented in [SecBERT: A Pretrained Language Model for Cyber Security Text](https://github.com/jackaduma/SecBERT/), which is a SecRoBERTa model trained on cyber security text.
The training corpus was papers taken from
* [APTnotes](https://github.com/kbandla/APTnotes)
* [Stucco-Data: Cyber security data sources](https://stucco.github.io/data/)
* [CASIE: Extracting Cybersecurity Event Information from Text](https://ebiquity.umbc.edu/_file_directory_/papers/943.pdf)
* [SemEval-2018 Task 8: Semantic Extraction from CybersecUrity REports using Natural Language Processing (SecureNLP)](https://competitions.codalab.org/competitions/17262).
SecRoBERTa has its own wordpiece vocabulary (secvocab) that's built to best match the training corpus.
We trained [SecBERT](https://huggingface.co/jackaduma/SecBERT) and [SecRoBERTa](https://huggingface.co/jackaduma/SecRoBERTa) versions.
Available models include:
* [`SecBERT`](https://huggingface.co/jackaduma/SecBERT)
* [`SecRoBERTa`](https://huggingface.co/jackaduma/SecRoBERTa)
---
## **Fill Mask**
We proposed to build language model which work on cyber security text, as result, it can improve downstream tasks (NER, Text Classification, Semantic Understand, Q&A) in Cyber Security Domain.
First, as below shows Fill-Mask pipeline in [Google Bert](), [AllenAI SciBert](https://github.com/allenai/scibert) and our [SecBERT](https://github.com/jackaduma/SecBERT) .
<!-- <img src="./fill-mask-result.png" width="150%" height="150%"> -->
---
The original repo can be found [here](https://github.com/jackaduma/SecBERT). |
mradermacher/Shark-1-Ogno-7b-passthrough-GGUF | mradermacher | "2024-06-22T00:47:26Z" | 8,666 | 0 | transformers | [
"transformers",
"gguf",
"merge",
"mergekit",
"lazymergekit",
"powermove72/Shark-1",
"eren23/OGNO-7b-dpo-truthful",
"en",
"base_model:powermove72/Shark-1-Ogno-7b-passthrough",
"endpoints_compatible",
"region:us"
] | null | "2024-06-21T23:59:40Z" | ---
base_model: powermove72/Shark-1-Ogno-7b-passthrough
language:
- en
library_name: transformers
quantized_by: mradermacher
tags:
- merge
- mergekit
- lazymergekit
- powermove72/Shark-1
- eren23/OGNO-7b-dpo-truthful
---
## About
<!-- ### quantize_version: 2 -->
<!-- ### output_tensor_quantised: 1 -->
<!-- ### convert_type: hf -->
<!-- ### vocab_type: -->
<!-- ### tags: -->
static quants of https://huggingface.co/powermove72/Shark-1-Ogno-7b-passthrough
<!-- provided-files -->
weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion.
## Usage
If you are unsure how to use GGUF files, refer to one of [TheBloke's
READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
more details, including on how to concatenate multi-part files.
## Provided Quants
(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
| Link | Type | Size/GB | Notes |
|:-----|:-----|--------:|:------|
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.Q2_K.gguf) | Q2_K | 2.8 | |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.IQ3_XS.gguf) | IQ3_XS | 3.1 | |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.Q3_K_S.gguf) | Q3_K_S | 3.3 | |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.IQ3_S.gguf) | IQ3_S | 3.3 | beats Q3_K* |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.IQ3_M.gguf) | IQ3_M | 3.4 | |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.Q3_K_M.gguf) | Q3_K_M | 3.6 | lower quality |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.Q3_K_L.gguf) | Q3_K_L | 3.9 | |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.IQ4_XS.gguf) | IQ4_XS | 4.0 | |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.Q4_K_S.gguf) | Q4_K_S | 4.2 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.Q4_K_M.gguf) | Q4_K_M | 4.5 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.Q5_K_S.gguf) | Q5_K_S | 5.1 | |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.Q5_K_M.gguf) | Q5_K_M | 5.2 | |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.Q6_K.gguf) | Q6_K | 6.0 | very good quality |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.Q8_0.gguf) | Q8_0 | 7.8 | fast, best quality |
| [GGUF](https://huggingface.co/mradermacher/Shark-1-Ogno-7b-passthrough-GGUF/resolve/main/Shark-1-Ogno-7b-passthrough.f16.gguf) | f16 | 14.6 | 16 bpw, overkill |
Here is a handy graph by ikawrakow comparing some lower-quality quant
types (lower is better):
And here are Artefact2's thoughts on the matter:
https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
## FAQ / Model Request
See https://huggingface.co/mradermacher/model_requests for some answers to
questions you might have and/or if you want some other model quantized.
## Thanks
I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
me use its servers and providing upgrades to my workstation to enable
this work in my free time.
<!-- end -->
|
allenai/tulu-2-7b | allenai | "2024-04-30T18:45:02Z" | 8,663 | 7 | transformers | [
"transformers",
"pytorch",
"llama",
"text-generation",
"conversational",
"en",
"dataset:allenai/tulu-v2-sft-mixture",
"arxiv:2311.10702",
"base_model:meta-llama/Llama-2-7b-hf",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2023-11-13T03:24:42Z" | ---
model-index:
- name: tulu-2-7b
results: []
datasets:
- allenai/tulu-v2-sft-mixture
language:
- en
base_model: meta-llama/Llama-2-7b-hf
---
<img src="https://huggingface.co/datasets/allenai/blog-images/resolve/main/tulu-v2/Tulu%20V2%20banner.png" alt="TuluV2 banner" width="800" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
# Model Card for Tulu 2 7B
Tulu is a series of language models that are trained to act as helpful assistants.
Tulu 2 7B is a fine-tuned version of Llama 2 that was trained on a mix of publicly available, synthetic and human datasets.
For more details, read the paper: [Camels in a Changing Climate: Enhancing LM Adaptation with Tulu 2
](https://arxiv.org/abs/2311.10702).
## Model description
- **Model type:** A model belonging to a suite of instruction and RLHF tuned chat models on a mix of publicly available, synthetic and human-created datasets.
- **Language(s) (NLP):** Primarily English
- **License:** [AI2 ImpACT](https://allenai.org/impact-license) Low-risk license.
- **Finetuned from model:** [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf)
### Model Sources
- **Repository:** https://github.com/allenai/open-instruct
- **Model Family:** Other models and the dataset are found in the [Tulu V2 collection](https://huggingface.co/collections/allenai/tulu-v2-suite-6551b56e743e6349aab45101).
## Performance
| Model | Size | Alignment | MT-Bench (score) | AlpacaEval (win rate %) |
|-------------|-----|----|---------------|--------------|
| **Tulu-v2-7b** 🐪 | **7B** | **SFT** | **6.30** | **73.9** |
| **Tulu-v2-dpo-7b** 🐪 | **7B** | **DPO** | **6.29** | **85.1** |
| **Tulu-v2-13b** 🐪 | **13B** | **SFT** | **6.70** | **78.9** |
| **Tulu-v2-dpo-13b** 🐪 | **13B** | **DPO** | **7.00** | **89.5** |
| **Tulu-v2-70b** 🐪 | **70B** | **SFT** | **7.49** | **86.6** |
| **Tulu-v2-dpo-70b** 🐪 | **70B** | **DPO** | **7.89** | **95.1** |
## Input Format
The model is trained to use the following format (note the newlines):
```
<|user|>
Your message here!
<|assistant|>
```
For best results, format all inputs in this manner. **Make sure to include a newline after `<|assistant|>`, this can affect generation quality quite a bit.**
## Intended uses & limitations
The model was fine-tuned on a filtered and preprocessed of the [Tulu V2 mix dataset](https://huggingface.co/datasets/allenai/tulu-v2-sft-mixture), which contains a diverse range of human created instructions and synthetic dialogues generated primarily by other LLMs.
<!--We then further aligned the model with a [Jax DPO trainer](https://github.com/hamishivi/EasyLM/blob/main/EasyLM/models/llama/llama_train_dpo.py) built on [EasyLM](https://github.com/young-geng/EasyLM) on the [openbmb/UltraFeedback](https://huggingface.co/datasets/openbmb/UltraFeedback) dataset, which contains 64k prompts and model completions that are ranked by GPT-4.
<!-- You can find the datasets used for training Tulu V2 [here]()
Here's how you can run the model using the `pipeline()` function from 🤗 Transformers:
```python
# Install transformers from source - only needed for versions <= v4.34
# pip install git+https://github.com/huggingface/transformers.git
# pip install accelerate
import torch
from transformers import pipeline
pipe = pipeline("text-generation", model="HuggingFaceH4/tulu-2-dpo-70b", torch_dtype=torch.bfloat16, device_map="auto")
# We use the tokenizer's chat template to format each message - see https://huggingface.co/docs/transformers/main/en/chat_templating
messages = [
{
"role": "system",
"content": "You are a friendly chatbot who always responds in the style of a pirate",
},
{"role": "user", "content": "How many helicopters can a human eat in one sitting?"},
]
prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipe(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
# <|system|>
# You are a friendly chatbot who always responds in the style of a pirate.</s>
# <|user|>
# How many helicopters can a human eat in one sitting?</s>
# <|assistant|>
# Ah, me hearty matey! But yer question be a puzzler! A human cannot eat a helicopter in one sitting, as helicopters are not edible. They be made of metal, plastic, and other materials, not food!
```-->
## Bias, Risks, and Limitations
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
The Tulu models have not been aligned to generate safe completions within the RLHF phase or deployed with in-the-loop filtering of responses like ChatGPT, so the model can produce problematic outputs (especially when prompted to do so).
It is also unknown what the size and composition of the corpus was used to train the base Llama 2 models, however it is likely to have included a mix of Web data and technical sources like books and code. See the [Falcon 180B model card](https://huggingface.co/tiiuae/falcon-180B#training-data) for an example of this.
### Training hyperparameters
The following hyperparameters were used during DPO training:
- learning_rate: 2e-5
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 2.0
## Citation
If you find Tulu 2 is useful in your work, please cite it with:
```
@misc{ivison2023camels,
title={Camels in a Changing Climate: Enhancing LM Adaptation with Tulu 2},
author={Hamish Ivison and Yizhong Wang and Valentina Pyatkin and Nathan Lambert and Matthew Peters and Pradeep Dasigi and Joel Jang and David Wadden and Noah A. Smith and Iz Beltagy and Hannaneh Hajishirzi},
year={2023},
eprint={2311.10702},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
*Model card adapted from [Zephyr Beta](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta/blob/main/README.md)* |
pszemraj/grammar-synthesis-small | pszemraj | "2023-04-28T01:45:36Z" | 8,661 | 4 | transformers | [
"transformers",
"pytorch",
"onnx",
"safetensors",
"t5",
"text2text-generation",
"grammar",
"spelling",
"punctuation",
"error-correction",
"grammar synthesis",
"FLAN",
"dataset:jfleg",
"arxiv:2107.06751",
"license:cc-by-nc-sa-4.0",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text2text-generation | "2022-07-09T23:00:37Z" | ---
languages:
- en
license:
- cc-by-nc-sa-4.0
- apache-2.0
tags:
- grammar
- spelling
- punctuation
- error-correction
- grammar synthesis
- FLAN
datasets:
- jfleg
widget:
- text: "There car broke down so their hitching a ride to they're class."
example_title: "compound-1"
- text: "i can has cheezburger"
example_title: "cheezburger"
- text: "so em if we have an now so with fito ringina know how to estimate the tren given the ereafte mylite trend we can also em an estimate is nod s
i again tort watfettering an we have estimated the trend an
called wot to be called sthat of exty right now we can and look at
wy this should not hare a trend i becan we just remove the trend an and we can we now estimate
tesees ona effect of them exty"
example_title: "Transcribed Audio Example 2"
- text: "My coworker said he used a financial planner to help choose his stocks so he wouldn't loose money."
example_title: "incorrect word choice (context)"
- text: "good so hve on an tadley i'm not able to make it to the exla session on monday this week e which is why i am e recording pre recording
an this excelleision and so to day i want e to talk about two things and first of all em i wont em wene give a summary er about
ta ohow to remove trents in these nalitives from time series"
example_title: "lowercased audio transcription output"
- text: "Frustrated, the chairs took me forever to set up."
example_title: "dangling modifier"
- text: "I would like a peice of pie."
example_title: "miss-spelling"
- text: "Which part of Zurich was you going to go hiking in when we were there for the first time together? ! ?"
example_title: "chatbot on Zurich"
- text: "Most of the course is about semantic or content of language but there are also interesting topics to be learned from the servicefeatures except statistics in characters in documents. At this point, Elvthos introduces himself as his native English speaker and goes on to say that if you continue to work on social scnce,"
example_title: "social science ASR summary output"
- text: "they are somewhat nearby right yes please i'm not sure how the innish is tepen thut mayyouselect one that istatte lo variants in their property e ere interested and anyone basical e may be applyind reaching the browing approach were"
- "medical course audio transcription"
parameters:
max_length: 128
min_length: 4
num_beams: 8
repetition_penalty: 1.21
length_penalty: 1
early_stopping: True
---
# grammar-synthesis-small (beta)
This model is a fine-tuned version of [google/t5-small-lm-adapt](https://huggingface.co/google/t5-small-lm-adapt) for grammar correction on an expanded version of the [JFLEG](https://paperswithcode.com/dataset/jfleg) dataset.
usage in Python (after `pip install transformers`):
```python
from transformers import pipeline
corrector = pipeline(
'text2text-generation',
'pszemraj/grammar-synthesis-small',
)
raw_text = 'i can has cheezburger'
results = corrector(raw_text)
print(results)
```
Check out a simple demo in [Google Colab here](https://colab.research.google.com/gist/pszemraj/06fac5b608889e258229a659cc53485f/demo-for-grammar-synthesis-small.ipynb).
## Model description
The intent is to create a text2text language model that successfully completes "single-shot grammar correction" on a potentially grammatically incorrect text **that could have a lot of mistakes** with the important qualifier of **it does not semantically change text/information that IS grammatically correct.**
Compare some of the heavier-error examples on [other grammar correction models](https://huggingface.co/models?dataset=dataset:jfleg) to see the difference :)
## Limitations
- dataset: `cc-by-nc-sa-4.0`
- model: `apache-2.0`
- this is **still a work-in-progress** and while probably useful for "single-shot grammar correction" in a lot of cases, **give the outputs a glance for correctness ok?**
## Use Cases
Obviously, this section is quite general as there are many things one can use "general single-shot grammar correction" for. Some ideas or use cases:
1. Correcting highly error-prone LM outputs. Some examples would be audio transcription (ASR) (this is literally some of the examples) or something like handwriting OCR.
- To be investigated further, depending on what model/system is used it _might_ be worth it to apply this after OCR on typed characters.
2. Correcting/infilling text generated by text generation models to be cohesive/remove obvious errors that break the conversation immersion. I use this on the outputs of [this OPT 2.7B chatbot-esque model of myself](https://huggingface.co/pszemraj/opt-peter-2.7B).
> An example of this model running on CPU with beam search:
```
original response:
ive heard it attributed to a bunch of different philosophical schools, including stoicism, pragmatism, existentialism and even some forms of post-structuralism. i think one of the most interesting (and most difficult) philosophical problems is trying to let dogs (or other animals) out of cages. the reason why this is a difficult problem is because it seems to go against our grain (so to
synthesizing took 306.12 seconds
Final response in 1294.857 s:
I've heard it attributed to a bunch of different philosophical schools, including solipsism, pragmatism, existentialism and even some forms of post-structuralism. i think one of the most interesting (and most difficult) philosophical problems is trying to let dogs (or other animals) out of cages. the reason why this is a difficult problem is because it seems to go against our grain (so to speak)
```
_Note: that I have some other logic that removes any periods at the end of the final sentence in this chatbot setting [to avoid coming off as passive aggressive](https://www.npr.org/2020/09/05/909969004/before-texting-your-kid-make-sure-to-double-check-your-punctuation)_
3. Somewhat related to #2 above, fixing/correcting so-called [tortured-phrases](https://arxiv.org/abs/2107.06751) that are dead giveaways text was generated by a language model. _Note that _SOME_ of these are not fixed, especially as they venture into domain-specific terminology (i.e. irregular timberland instead of Random Forest)._
## Training and evaluation data
More information needed 😉
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0004
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- distributed_type: multi-GPU
- gradient_accumulation_steps: 32
- total_train_batch_size: 512
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 4
### Training results
### Framework versions
- Transformers 4.20.1
- Pytorch 1.11.0+cu113
- Datasets 2.3.2
- Tokenizers 0.12.1
|
princeton-nlp/Sheared-LLaMA-1.3B-ShareGPT | princeton-nlp | "2023-12-04T21:35:41Z" | 8,660 | 9 | transformers | [
"transformers",
"pytorch",
"llama",
"text-generation",
"arxiv:2310.06694",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"region:us"
] | text-generation | "2023-11-22T16:29:50Z" | ---
license: apache-2.0
---
**Paper**: [https://arxiv.org/pdf/2310.06694.pdf](https://arxiv.org/pdf/2310.06694.pdf)
**Code**: https://github.com/princeton-nlp/LLM-Shearing
**Models**: [Sheared-LLaMA-1.3B](https://huggingface.co/princeton-nlp/Sheared-LLaMA-1.3B), [Sheared-LLaMA-2.7B](https://huggingface.co/princeton-nlp/Sheared-LLaMA-2.7B)
## Training information
This is the instruction tuned version of [princeton-nlp/Sheared-LLaMA-1.3B](https://huggingface.co/princeton-nlp/Sheared-LLaMA-1.3B). We trained the base model on 10,000 instruction-response pairs
sampled from the ShareGPT dataset (first-turns only). We use the following prompt to perform instruction tuning.
> You are a helpful assistant. Write a response that appropriately completes the request.\n\n### Input:\n{input}\n\n### Response:
This model can be loaded through transformers.LlamaModelForCausalLM as follows:
```
from transformers import LlamaModelForCausalLM
model = LlamaModelForCausalLM.from_pretrained("princeton-nlp/Sheared-LLaMA-1.3B-ShareGPT")
```
## Bibtex
If you find our model useful, consider citing us with:
```
@article{xia2023sheared,
title={Sheared llama: Accelerating language model pre-training via structured pruning},
author={Xia, Mengzhou and Gao, Tianyu and Zeng, Zhiyuan and Chen, Danqi},
journal={arXiv preprint arXiv:2310.06694},
year={2023}
}
```
|
unsloth/llama-3-70b-bnb-4bit | unsloth | "2024-05-05T12:17:09Z" | 8,657 | 39 | transformers | [
"transformers",
"safetensors",
"llama",
"text-generation",
"unsloth",
"llama-3",
"en",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"text-generation-inference",
"4-bit",
"region:us"
] | text-generation | "2024-04-18T18:48:32Z" | ---
language:
- en
license: apache-2.0
library_name: transformers
tags:
- unsloth
- transformers
- llama
- llama-3
---
# Finetune Mistral, Gemma, Llama 2-5x faster with 70% less memory via Unsloth!
We have a Google Colab Tesla T4 notebook for Llama-3 8b here: https://colab.research.google.com/drive/135ced7oHytdxu3N2DNe1Z0kqjyYIkDXp?usp=sharing
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/Discord%20button.png" width="200"/>](https://discord.gg/u54VK8m8tk)
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/buy%20me%20a%20coffee%20button.png" width="200"/>](https://ko-fi.com/unsloth)
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
## ✨ Finetune for Free
All notebooks are **beginner friendly**! Add your dataset, click "Run All", and you'll get a 2x faster finetuned model which can be exported to GGUF, vLLM or uploaded to Hugging Face.
| Unsloth supports | Free Notebooks | Performance | Memory use |
|-----------------|--------------------------------------------------------------------------------------------------------------------------|-------------|----------|
| **Llama-3 8b** | [▶️ Start on Colab](https://colab.research.google.com/drive/135ced7oHytdxu3N2DNe1Z0kqjyYIkDXp?usp=sharing) | 2.4x faster | 58% less |
| **Gemma 7b** | [▶️ Start on Colab](https://colab.research.google.com/drive/10NbwlsRChbma1v55m8LAPYG15uQv6HLo?usp=sharing) | 2.4x faster | 58% less |
| **Mistral 7b** | [▶️ Start on Colab](https://colab.research.google.com/drive/1Dyauq4kTZoLewQ1cApceUQVNcnnNTzg_?usp=sharing) | 2.2x faster | 62% less |
| **Llama-2 7b** | [▶️ Start on Colab](https://colab.research.google.com/drive/1lBzz5KeZJKXjvivbYvmGarix9Ao6Wxe5?usp=sharing) | 2.2x faster | 43% less |
| **TinyLlama** | [▶️ Start on Colab](https://colab.research.google.com/drive/1AZghoNBQaMDgWJpi4RbffGM1h6raLUj9?usp=sharing) | 3.9x faster | 74% less |
| **CodeLlama 34b** A100 | [▶️ Start on Colab](https://colab.research.google.com/drive/1y7A0AxE3y8gdj4AVkl2aZX47Xu3P1wJT?usp=sharing) | 1.9x faster | 27% less |
| **Mistral 7b** 1xT4 | [▶️ Start on Kaggle](https://www.kaggle.com/code/danielhanchen/kaggle-mistral-7b-unsloth-notebook) | 5x faster\* | 62% less |
| **DPO - Zephyr** | [▶️ Start on Colab](https://colab.research.google.com/drive/15vttTpzzVXv_tJwEk-hIcQ0S9FcEWvwP?usp=sharing) | 1.9x faster | 19% less |
- This [conversational notebook](https://colab.research.google.com/drive/1Aau3lgPzeZKQ-98h69CCu1UJcvIBLmy2?usp=sharing) is useful for ShareGPT ChatML / Vicuna templates.
- This [text completion notebook](https://colab.research.google.com/drive/1ef-tab5bhkvWmBOObepl1WgJvfvSzn5Q?usp=sharing) is for raw text. This [DPO notebook](https://colab.research.google.com/drive/15vttTpzzVXv_tJwEk-hIcQ0S9FcEWvwP?usp=sharing) replicates Zephyr.
- \* Kaggle has 2x T4s, but we use 1. Due to overhead, 1x T4 is 5x faster. |
QuantFactory/Llama-3-Spellbound-Instruct-8B-0.3-GGUF | QuantFactory | "2024-06-22T17:53:46Z" | 8,651 | 0 | null | [
"gguf",
"text-generation",
"base_model:hf-100/Llama-3-Spellbound-Instruct-8B-0.3",
"license:cc-by-nc-sa-4.0",
"region:us"
] | text-generation | "2024-06-22T12:06:23Z" | ---
license: cc-by-nc-sa-4.0
pipeline_tag: text-generation
base_model: hf-100/Llama-3-Spellbound-Instruct-8B-0.3
---
# QuantFactory/Llama-3-Spellbound-Instruct-8B-0.3-GGUF
This is quantized version of [hf-100/Llama-3-Spellbound-Instruct-8B-0.3](https://huggingface.co/hf-100/Llama-3-Spellbound-Instruct-8B-0.3) created using llama.cpp
# Model Description
## Llama-3 Spellbound Instruct Tuning-Free
## Updated Aspects
- Trained on additional tokens
- Improved mix of subject matter model was trained on
- Trained for 1.5M additional tokens
- Additional training on DPO dataset
## Model Rationale
Llama 3 is a strong base model with strong world understanding and creativity. Additional instruct finetuning trades that world understanding and creativity for instruction following that Llama doesn't require in order to adhere to most forms of roleplay.
This model was trained on unstructured text only, no instruct related fine-tuning was performed.
Made by [tryspellbound.com](https://tryspellbound.com).
*(tryspellbound.com does not currently use this model, it uses Claude 3 Sonnet.)*
## Features of this fine-tune for Llama 3:
- Roleplaying in multi-turn stories where the history is presented in a single message
- Dynamic switching of writing styles for different scenarios
- Interpretation of formatting marks 'quote' and 'action'
**Warning:** The underlying model, Llama 3, was trained on data that included adult content. This fine-tune does not add additional guardrails and is not suitable for all environments.
## Purpose of the Model
The main goal is to explore how presenting LLMs with history and instructions separately affects their performance, demonstrating:
- Improved coherence in long conversations
- Enhanced quality of character interactions
- Decreased instruction adherence, which could be improved with additional training
## Advanced prompting of the model
For advanced prompting, see [this document](https://rentry.co/ti936r2i) |
tigeryi/imagenet1k | tigeryi | "2024-05-03T21:32:54Z" | 8,649 | 2 | transformers | [
"transformers",
"tensorboard",
"safetensors",
"data2vec-vision",
"image-classification",
"generated_from_trainer",
"base_model:tigeryi/imagenet1k",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | image-classification | "2024-04-29T02:02:32Z" | ---
license: apache-2.0
base_model: tigeryi/imagenet1k
tags:
- generated_from_trainer
model-index:
- name: imagenet1k
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# imagenet1k
This model is a fine-tuned version of [tigeryi/imagenet1k](https://huggingface.co/tigeryi/imagenet1k) on an unknown dataset.
It achieves the following results on the evaluation set:
- eval_loss: 1.0022
- eval_accuracy: 0.7605
- eval_runtime: 1004.8381
- eval_samples_per_second: 49.759
- eval_steps_per_second: 1.555
- step: 0
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 1
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 7
### Framework versions
- Transformers 4.40.1
- Pytorch 2.2.1+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1
|
KDHyun08/TAACO_STS | KDHyun08 | "2022-08-01T05:00:14Z" | 8,644 | 2 | sentence-transformers | [
"sentence-transformers",
"pytorch",
"bert",
"feature-extraction",
"sentence-similarity",
"transformers",
"TAACO",
"ko",
"autotrain_compatible",
"endpoints_compatible",
"text-embeddings-inference",
"region:us"
] | sentence-similarity | "2022-07-25T08:19:31Z" | ---
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- sentence-similarity
- transformers
- TAACO
language: ko
---
# TAACO_Similarity
본 모델은 [Sentence-transformers](https://www.SBERT.net)를 기반으로 하며 KLUE의 STS(Sentence Textual Similarity) 데이터셋을 통해 훈련을 진행한 모델입니다.
필자가 제작하고 있는 한국어 문장간 결속성 측정 도구인 K-TAACO(가제)의 지표 중 하나인 문장 간 의미적 결속성을 측정하기 위해 본 모델을 제작하였습니다.
또한 모두의 말뭉치의 문장간 유사도 데이터 등 다양한 데이터를 구해 추가 훈련을 진행할 예정입니다.
## Train Data
KLUE-sts-v1.1._train.json
NLI-sts-train.tsv
## Usage (Sentence-Transformers)
본 모델을 사용하기 위해서는 [Sentence-transformers](https://www.SBERT.net)를 설치하여야 합니다.
```
pip install -U sentence-transformers
```
모델을 사용하기 위해서는 아래 코드를 참조하시길 바랍니다.
```python
from sentence_transformers import SentenceTransformer, models
sentences = ["This is an example sentence", "Each sentence is converted"]
embedding_model = models.Transformer(
model_name_or_path="KDHyun08/TAACO_STS",
max_seq_length=256,
do_lower_case=True
)
pooling_model = models.Pooling(
embedding_model.get_word_embedding_dimension(),
pooling_mode_mean_tokens=True,
pooling_mode_cls_token=False,
pooling_mode_max_tokens=False,
)
model = SentenceTransformer(modules=[embedding_model, pooling_model])
embeddings = model.encode(sentences)
print(embeddings)
```
## Usage (실제 문장 간 유사도 비교)
[Sentence-transformers](https://www.SBERT.net) 를 설치한 후 아래 내용과 같이 문장 간 유사도를 비교할 수 있습니다.
query 변수는 비교 기준이 되는 문장(Source Sentence)이고 비교를 진행할 문장은 docs에 list 형식으로 구성하시면 됩니다.
```python
from sentence_transformers import SentenceTransformer, models
embedding_model = models.Transformer(
model_name_or_path="KDHyun08/TAACO_STS",
max_seq_length=256,
do_lower_case=True
)
pooling_model = models.Pooling(
embedding_model.get_word_embedding_dimension(),
pooling_mode_mean_tokens=True,
pooling_mode_cls_token=False,
pooling_mode_max_tokens=False,
)
model = SentenceTransformer(modules=[embedding_model, pooling_model])
docs = ['어제는 아내의 생일이었다', '생일을 맞이하여 아침을 준비하겠다고 오전 8시 30분부터 음식을 준비하였다. 주된 메뉴는 스테이크와 낙지볶음, 미역국, 잡채, 소야 등이었다', '스테이크는 자주 하는 음식이어서 자신이 준비하려고 했다', '앞뒤도 1분씩 3번 뒤집고 래스팅을 잘 하면 육즙이 가득한 스테이크가 준비되다', '아내도 그런 스테이크를 좋아한다. 그런데 상상도 못한 일이 벌이지고 말았다', '보통 시즈닝이 되지 않은 원육을 사서 스테이크를 했는데, 이번에는 시즈닝이 된 부챗살을 구입해서 했다', '그런데 케이스 안에 방부제가 들어있는 것을 인지하지 못하고 방부제와 동시에 프라이팬에 올려놓을 것이다', '그것도 인지 못한 체... 앞면을 센 불에 1분을 굽고 뒤집는 순간 방부제가 함께 구어진 것을 알았다', '아내의 생일이라 맛있게 구워보고 싶었는데 어처구니없는 상황이 발생한 것이다', '방부제가 센 불에 녹아서 그런지 물처럼 흘러내렸다', ' 고민을 했다. 방부제가 묻은 부문만 제거하고 다시 구울까 했는데 방부제에 절대 먹지 말라는 문구가 있어서 아깝지만 버리는 방향을 했다', '너무나 안타까웠다', '아침 일찍 아내가 좋아하는 스테이크를 준비하고 그것을 맛있게 먹는 아내의 모습을 보고 싶었는데 전혀 생각지도 못한 상황이 발생해서... 하지만 정신을 추스르고 바로 다른 메뉴로 변경했다', '소야, 소시지 야채볶음..', '아내가 좋아하는지 모르겠지만 냉장고 안에 있는 후랑크소세지를 보니 바로 소야를 해야겠다는 생각이 들었다. 음식은 성공적으로 완성이 되었다', '40번째를 맞이하는 아내의 생일은 성공적으로 준비가 되었다', '맛있게 먹어 준 아내에게도 감사했다', '매년 아내의 생일에 맞이하면 아침마다 생일을 차려야겠다. 오늘도 즐거운 하루가 되었으면 좋겠다', '생일이니까~']
#각 문장의 vector값 encoding
document_embeddings = model.encode(docs)
query = '생일을 맞이하여 아침을 준비하겠다고 오전 8시 30분부터 음식을 준비하였다'
query_embedding = model.encode(query)
top_k = min(10, len(docs))
# 코사인 유사도 계산 후,
cos_scores = util.pytorch_cos_sim(query_embedding, document_embeddings)[0]
# 코사인 유사도 순으로 문장 추출
top_results = torch.topk(cos_scores, k=top_k)
print(f"입력 문장: {query}")
print(f"\n<입력 문장과 유사한 {top_k} 개의 문장>\n")
for i, (score, idx) in enumerate(zip(top_results[0], top_results[1])):
print(f"{i+1}: {docs[idx]} {'(유사도: {:.4f})'.format(score)}\n")
```
## Evaluation Results
위 Usage를 실행하게 되면 아래와 같은 결과가 도출됩니다. 1에 가까울수록 유사한 문장입니다.
```
입력 문장: 생일을 맞이하여 아침을 준비하겠다고 오전 8시 30분부터 음식을 준비하였다
<입력 문장과 유사한 10 개의 문장>
1: 생일을 맞이하여 아침을 준비하겠다고 오전 8시 30분부터 음식을 준비하였다. 주된 메뉴는 스테이크와 낙지볶음, 미역국, 잡채, 소야 등이었다 (유사도: 0.6687)
2: 매년 아내의 생일에 맞이하면 아침마다 생일을 차려야겠다. 오늘도 즐거운 하루가 되었으면 좋겠다 (유사도: 0.6468)
3: 40번째를 맞이하는 아내의 생일은 성공적으로 준비가 되었다 (유사도: 0.4647)
4: 아내의 생일이라 맛있게 구워보고 싶었는데 어처구니없는 상황이 발생한 것이다 (유사도: 0.4469)
5: 생일이니까~ (유사도: 0.4218)
6: 어제는 아내의 생일이었다 (유사도: 0.4192)
7: 아침 일찍 아내가 좋아하는 스테이크를 준비하고 그것을 맛있게 먹는 아내의 모습을 보고 싶었는데 전혀 생각지도 못한 상황이 발생해서... 하지만 정신을 추스르고 바로 다른 메뉴로 변경했다 (유사도: 0.4156)
8: 맛있게 먹어 준 아내에게도 감사했다 (유사도: 0.3093)
9: 아내가 좋아하는지 모르겠지만 냉장고 안에 있는 후랑크소세지를 보니 바로 소야를 해야겠다는 생각이 들었다. 음식은 성공적으로 완성이 되었다 (유사도: 0.2259)
10: 아내도 그런 스테이크를 좋아한다. 그런데 상상도 못한 일이 벌이지고 말았다 (유사도: 0.1967)
```
**DataLoader**:
`torch.utils.data.dataloader.DataLoader` of length 142 with parameters:
```
{'batch_size': 32, 'sampler': 'torch.utils.data.sampler.RandomSampler', 'batch_sampler': 'torch.utils.data.sampler.BatchSampler'}
```
**Loss**:
`sentence_transformers.losses.CosineSimilarityLoss.CosineSimilarityLoss`
Parameters of the fit()-Method:
```
{
"epochs": 4,
"evaluation_steps": 1000,
"evaluator": "sentence_transformers.evaluation.EmbeddingSimilarityEvaluator.EmbeddingSimilarityEvaluator",
"max_grad_norm": 1,
"optimizer_class": "<class 'transformers.optimization.AdamW'>",
"optimizer_params": {
"lr": 2e-05
},
"scheduler": "WarmupLinear",
"steps_per_epoch": null,
"warmup_steps": 10000,
"weight_decay": 0.01
}
```
## Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
)
```
## Citing & Authors
<!--- Describe where people can find more information --> |
microsoft/deberta-base-mnli | microsoft | "2021-12-09T13:36:31Z" | 8,631 | 3 | transformers | [
"transformers",
"pytorch",
"rust",
"deberta",
"text-classification",
"deberta-v1",
"deberta-mnli",
"en",
"arxiv:2006.03654",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | "2022-03-02T23:29:05Z" | ---
language: en
tags:
- deberta-v1
- deberta-mnli
tasks: mnli
thumbnail: https://huggingface.co/front/thumbnails/microsoft.png
license: mit
widget:
- text: "[CLS] I love you. [SEP] I like you. [SEP]"
---
## DeBERTa: Decoding-enhanced BERT with Disentangled Attention
[DeBERTa](https://arxiv.org/abs/2006.03654) improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder. It outperforms BERT and RoBERTa on majority of NLU tasks with 80GB training data.
Please check the [official repository](https://github.com/microsoft/DeBERTa) for more details and updates.
This model is the base DeBERTa model fine-tuned with MNLI task
#### Fine-tuning on NLU tasks
We present the dev results on SQuAD 1.1/2.0 and MNLI tasks.
| Model | SQuAD 1.1 | SQuAD 2.0 | MNLI-m |
|-------------------|-----------|-----------|--------|
| RoBERTa-base | 91.5/84.6 | 83.7/80.5 | 87.6 |
| XLNet-Large | -/- | -/80.2 | 86.8 |
| **DeBERTa-base** | 93.1/87.2 | 86.2/83.1 | 88.8 |
### Citation
If you find DeBERTa useful for your work, please cite the following paper:
``` latex
@inproceedings{
he2021deberta,
title={DEBERTA: DECODING-ENHANCED BERT WITH DISENTANGLED ATTENTION},
author={Pengcheng He and Xiaodong Liu and Jianfeng Gao and Weizhu Chen},
booktitle={International Conference on Learning Representations},
year={2021},
url={https://openreview.net/forum?id=XPZIaotutsD}
}
```
|
state-spaces/mamba-2.8b-hf | state-spaces | "2024-03-06T00:44:55Z" | 8,616 | 70 | transformers | [
"transformers",
"safetensors",
"mamba",
"text-generation",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-generation | "2024-03-05T23:53:30Z" | ---
library_name: transformers
tags: []
---
# Mamba
<!-- Provide a quick summary of what the model is/does. -->
This repository contains the `transfromers` compatible `mamba-2.8b`. The checkpoints are untouched, but the full `config.json` and tokenizer are pushed to this repo.
# Usage
You need to install `transformers` from `main` until `transformers=4.39.0` is released.
```bash
pip install git+https://github.com/huggingface/transformers@main
```
We also recommend you to install both `causal_conv_1d` and `mamba-ssm` using:
```bash
pip install causal-conv1d>=1.2.0
pip install mamba-ssm
```
If any of these two is not installed, the "eager" implementation will be used. Otherwise the more optimised `cuda` kernels will be used.
## Generation
You can use the classic `generate` API:
```python
>>> from transformers import MambaConfig, MambaForCausalLM, AutoTokenizer
>>> import torch
>>> tokenizer = AutoTokenizer.from_pretrained("state-spaces/mamba-2.8b-hf")
>>> model = MambaForCausalLM.from_pretrained("state-spaces/mamba-2.8b-hf")
>>> input_ids = tokenizer("Hey how are you doing?", return_tensors="pt")["input_ids"]
>>> out = model.generate(input_ids, max_new_tokens=10)
>>> print(tokenizer.batch_decode(out))
["Hey how are you doing?\n\nI'm doing great.\n\nI"]
```
## PEFT finetuning example
In order to finetune using the `peft` library, we recommend keeping the model in float32!
```python
from datasets import load_dataset
from trl import SFTTrainer
from peft import LoraConfig
from transformers import AutoTokenizer, AutoModelForCausalLM, TrainingArguments
tokenizer = AutoTokenizer.from_pretrained("state-spaces/mamba-2.8b-hf")
model = AutoModelForCausalLM.from_pretrained("state-spaces/mamba-2.8b-hf")
dataset = load_dataset("Abirate/english_quotes", split="train")
training_args = TrainingArguments(
output_dir="./results",
num_train_epochs=3,
per_device_train_batch_size=4,
logging_dir='./logs',
logging_steps=10,
learning_rate=2e-3
)
lora_config = LoraConfig(
r=8,
target_modules=["x_proj", "embeddings", "in_proj", "out_proj"],
task_type="CAUSAL_LM",
bias="none"
)
trainer = SFTTrainer(
model=model,
tokenizer=tokenizer,
args=training_args,
peft_config=lora_config,
train_dataset=dataset,
dataset_text_field="quote",
)
trainer.train()
``` |
mattyamonaca/controlnet_vp_one_proto | mattyamonaca | "2024-02-10T23:38:28Z" | 8,615 | 5 | diffusers | [
"diffusers",
"safetensors",
"license:apache-2.0",
"region:us"
] | null | "2024-02-10T23:06:20Z" | ---
license: apache-2.0
---
|
alvaroalon2/biobert_chemical_ner | alvaroalon2 | "2023-03-17T12:10:54Z" | 8,603 | 18 | transformers | [
"transformers",
"pytorch",
"tf",
"bert",
"token-classification",
"NER",
"Biomedical",
"Chemicals",
"en",
"dataset:BC5CDR-chemicals",
"dataset:BC4CHEMD",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | token-classification | "2022-03-02T23:29:05Z" | ---
language: en
license: apache-2.0
tags:
- token-classification
- NER
- Biomedical
- Chemicals
datasets:
- BC5CDR-chemicals
- BC4CHEMD
---
BioBERT model fine-tuned in NER task with BC5CDR-chemicals and BC4CHEMD corpus.
This was fine-tuned in order to use it in a BioNER/BioNEN system which is available at: https://github.com/librairy/bio-ner |
timm/vit_small_patch16_384.augreg_in21k_ft_in1k | timm | "2023-05-06T00:29:02Z" | 8,603 | 2 | timm | [
"timm",
"pytorch",
"safetensors",
"image-classification",
"dataset:imagenet-1k",
"dataset:imagenet-21k",
"arxiv:2106.10270",
"arxiv:2010.11929",
"license:apache-2.0",
"region:us"
] | image-classification | "2022-12-22T07:54:55Z" | ---
tags:
- image-classification
- timm
library_name: timm
license: apache-2.0
datasets:
- imagenet-1k
- imagenet-21k
---
# Model card for vit_small_patch16_384.augreg_in21k_ft_in1k
A Vision Transformer (ViT) image classification model. Trained on ImageNet-21k and fine-tuned on ImageNet-1k (with additional augmentation and regularization) in JAX by paper authors, ported to PyTorch by Ross Wightman.
## Model Details
- **Model Type:** Image classification / feature backbone
- **Model Stats:**
- Params (M): 22.2
- GMACs: 12.4
- Activations (M): 24.2
- Image size: 384 x 384
- **Papers:**
- How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers: https://arxiv.org/abs/2106.10270
- An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale: https://arxiv.org/abs/2010.11929v2
- **Dataset:** ImageNet-1k
- **Pretrain Dataset:** ImageNet-21k
- **Original:** https://github.com/google-research/vision_transformer
## Model Usage
### Image Classification
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model('vit_small_patch16_384.augreg_in21k_ft_in1k', pretrained=True)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1
top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5)
```
### Image Embeddings
```python
from urllib.request import urlopen
from PIL import Image
import timm
img = Image.open(urlopen(
'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
model = timm.create_model(
'vit_small_patch16_384.augreg_in21k_ft_in1k',
pretrained=True,
num_classes=0, # remove classifier nn.Linear
)
model = model.eval()
# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)
output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor
# or equivalently (without needing to set num_classes=0)
output = model.forward_features(transforms(img).unsqueeze(0))
# output is unpooled, a (1, 577, 384) shaped tensor
output = model.forward_head(output, pre_logits=True)
# output is a (1, num_features) shaped tensor
```
## Model Comparison
Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results).
## Citation
```bibtex
@article{steiner2021augreg,
title={How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers},
author={Steiner, Andreas and Kolesnikov, Alexander and and Zhai, Xiaohua and Wightman, Ross and Uszkoreit, Jakob and Beyer, Lucas},
journal={arXiv preprint arXiv:2106.10270},
year={2021}
}
```
```bibtex
@article{dosovitskiy2020vit,
title={An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale},
author={Dosovitskiy, Alexey and Beyer, Lucas and Kolesnikov, Alexander and Weissenborn, Dirk and Zhai, Xiaohua and Unterthiner, Thomas and Dehghani, Mostafa and Minderer, Matthias and Heigold, Georg and Gelly, Sylvain and Uszkoreit, Jakob and Houlsby, Neil},
journal={ICLR},
year={2021}
}
```
```bibtex
@misc{rw2019timm,
author = {Ross Wightman},
title = {PyTorch Image Models},
year = {2019},
publisher = {GitHub},
journal = {GitHub repository},
doi = {10.5281/zenodo.4414861},
howpublished = {\url{https://github.com/huggingface/pytorch-image-models}}
}
```
|
danbrown/RevAnimated-v1-2-2 | danbrown | "2023-04-30T16:55:39Z" | 8,602 | 8 | diffusers | [
"diffusers",
"safetensors",
"autotrain_compatible",
"endpoints_compatible",
"diffusers:StableDiffusionPipeline",
"region:us"
] | text-to-image | "2023-04-30T16:34:42Z" | Not official! This are diffusers weights for https://civitai.com/models/7371/rev-animated
Based on Stable Diffusion v1.5 |
jhgan/ko-sroberta-nli | jhgan | "2022-08-16T12:45:56Z" | 8,601 | 2 | sentence-transformers | [
"sentence-transformers",
"pytorch",
"tf",
"roberta",
"feature-extraction",
"sentence-similarity",
"transformers",
"ko",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | sentence-similarity | "2022-03-02T23:29:05Z" | ---
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- feature-extraction
- sentence-similarity
- transformers
language: ko
---
# ko-sroberta-nli
This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.
<!--- Describe your model here -->
## Usage (Sentence-Transformers)
Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
```
pip install -U sentence-transformers
```
Then you can use the model like this:
```python
from sentence_transformers import SentenceTransformer
sentences = ["안녕하세요?", "한국어 문장 임베딩을 위한 버트 모델입니다."]
model = SentenceTransformer('jhgan/ko-sroberta-nli')
embeddings = model.encode(sentences)
print(embeddings)
```
## Usage (HuggingFace Transformers)
Without [sentence-transformers](https://www.SBERT.net), you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.
```python
from transformers import AutoTokenizer, AutoModel
import torch
#Mean Pooling - Take attention mask into account for correct averaging
def mean_pooling(model_output, attention_mask):
token_embeddings = model_output[0] #First element of model_output contains all token embeddings
input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
# Sentences we want sentence embeddings for
sentences = ['This is an example sentence', 'Each sentence is converted']
# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained('jhgan/ko-sroberta-nli')
model = AutoModel.from_pretrained('jhgan/ko-sroberta-nli')
# Tokenize sentences
encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
# Compute token embeddings
with torch.no_grad():
model_output = model(**encoded_input)
# Perform pooling. In this case, mean pooling.
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
print("Sentence embeddings:")
print(sentence_embeddings)
```
## Evaluation Results
<!--- Describe how your model was evaluated -->
KorNLI 학습 데이터셋으로 학습한 후 KorSTS 평가 데이터셋으로 평가한 결과입니다.
- Cosine Pearson: 82.83
- Cosine Spearman: 83.85
- Euclidean Pearson: 82.87
- Euclidean Spearman: 83.29
- Manhattan Pearson: 82.88
- Manhattan Spearman: 83.28
- Dot Pearson: 80.34
- Dot Spearman: 79.69
## Training
The model was trained with the parameters:
**DataLoader**:
`sentence_transformers.datasets.NoDuplicatesDataLoader.NoDuplicatesDataLoader` of length 8885 with parameters:
```
{'batch_size': 64}
```
**Loss**:
`sentence_transformers.losses.MultipleNegativesRankingLoss.MultipleNegativesRankingLoss` with parameters:
```
{'scale': 20.0, 'similarity_fct': 'cos_sim'}
```
Parameters of the fit()-Method:
```
{
"epochs": 1,
"evaluation_steps": 1000,
"evaluator": "sentence_transformers.evaluation.EmbeddingSimilarityEvaluator.EmbeddingSimilarityEvaluator",
"max_grad_norm": 1,
"optimizer_class": "<class 'transformers.optimization.AdamW'>",
"optimizer_params": {
"lr": 2e-05
},
"scheduler": "WarmupLinear",
"steps_per_epoch": null,
"warmup_steps": 889,
"weight_decay": 0.01
}
```
## Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 128, 'do_lower_case': False}) with Transformer model: RobertaModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
)
```
## Citing & Authors
<!--- Describe where people can find more information -->
- Ham, J., Choe, Y. J., Park, K., Choi, I., & Soh, H. (2020). Kornli and korsts: New benchmark datasets for korean natural language understanding. arXiv preprint arXiv:2004.03289
- Reimers, Nils and Iryna Gurevych. “Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks.” ArXiv abs/1908.10084 (2019)
- Reimers, Nils and Iryna Gurevych. “Making Monolingual Sentence Embeddings Multilingual Using Knowledge Distillation.” EMNLP (2020).
|
QuantFactory/Llama-3-8B-Magpie-Pro-SFT-100K-v0.1-GGUF | QuantFactory | "2024-06-22T07:16:41Z" | 8,598 | 0 | null | [
"gguf",
"axolotl",
"generated_from_trainer",
"text-generation",
"arxiv:2406.08464",
"base_model:Magpie-Align/Llama-3-8B-Magpie-Pro-SFT-100K-v0.1",
"license:llama3",
"region:us"
] | text-generation | "2024-06-21T07:19:14Z" | ---
license: llama3
base_model: Magpie-Align/Llama-3-8B-Magpie-Pro-SFT-100K-v0.1
tags:
- axolotl
- generated_from_trainer
model-index:
- name: Llama-3-8B-Magpie-Pro-SFT-100K-v0.1
results: []
pipeline_tag: text-generation
---
# QuantFactory/Llama-3-8B-Magpie-Pro-SFT-100K-v0.1-GGUF
This is quantized version of [Magpie-Align/Llama-3-8B-Magpie-Pro-SFT-100K-v0.1](https://huggingface.co/Magpie-Align/Llama-3-8B-Magpie-Pro-SFT-100K-v0.1) created suing llama.cpp
# Model Description
Project Web: [https://magpie-align.github.io/](https://magpie-align.github.io/)
Arxiv Technical Report: [https://arxiv.org/abs/2406.08464](https://arxiv.org/abs/2406.08464)
Codes: [https://github.com/magpie-align/magpie](https://github.com/magpie-align/magpie)
## About This Model
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B) on **First 100K data** of [Magpie-Align/Magpie-Pro-300K-Filtered](https://huggingface.co/datasets/Magpie-Align/Magpie-Pro-300K-Filtered) dataset.
Please use [Magpie-Align/Llama-3-8B-Magpie-Pro-SFT-v0.1](https://huggingface.co/Magpie-Align/Llama-3-8B-Magpie-Pro-SFT-v0.1) with better performance.
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 8
- total_train_batch_size: 32
- total_eval_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 0.8869 | 0.0036 | 1 | 0.9139 |
| 0.5854 | 0.3344 | 92 | 0.6158 |
| 0.5218 | 0.6688 | 184 | 0.5455 |
| 0.4878 | 1.0032 | 276 | 0.5125 |
| 0.3734 | 1.3226 | 368 | 0.5091 |
| 0.3647 | 1.6570 | 460 | 0.5056 |
### Framework versions
- Transformers 4.40.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1
[<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
<details><summary>See axolotl config</summary>
axolotl version: `0.4.0`
```yaml
base_model: meta-llama/Meta-Llama-3-8B
model_type: LlamaForCausalLM
tokenizer_type: AutoTokenizer
load_in_8bit: false
load_in_4bit: false
strict: false
datasets:
- path: Magpie-Align/Magpie-Pro-300K-Filtered-First100K
type: sharegpt
conversation: llama3
dataset_prepared_path: last_run_prepared
val_set_size: 0.001
output_dir: ./out_Llama-3-8B-Magpie-Pro-100K-FilteredL
sequence_len: 8192
sample_packing: true
eval_sample_packing: false
pad_to_sequence_len: true
gradient_accumulation_steps: 8
micro_batch_size: 1
num_epochs: 2
optimizer: paged_adamw_8bit
lr_scheduler: cosine
learning_rate: 2e-5
train_on_inputs: false
group_by_length: false
bf16: auto
fp16:
tf32: false
gradient_checkpointing: true
gradient_checkpointing_kwargs:
use_reentrant: false
early_stopping_patience:
resume_from_checkpoint:
logging_steps: 1
xformers_attention:
flash_attention: true
warmup_steps: 100
evals_per_epoch: 3
eval_table_size:
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
pad_token: <|end_of_text|>
```
</details><br> |
QuantFactory/Llama-3-Lumimaid-8B-v0.1-OAS-GGUF | QuantFactory | "2024-06-28T14:28:32Z" | 8,595 | 0 | null | [
"gguf",
"not-for-all-audiences",
"nsfw",
"text-generation",
"base_model:NeverSleep/Llama-3-Lumimaid-8B-v0.1-OAS",
"license:cc-by-nc-4.0",
"region:us"
] | text-generation | "2024-06-26T06:22:10Z" | ---
license: cc-by-nc-4.0
tags:
- not-for-all-audiences
- nsfw
base_model: NeverSleep/Llama-3-Lumimaid-8B-v0.1-OAS
pipeline_tag: text-generation
---
# QuantFactory/Llama-3-Lumimaid-8B-v0.1-OAS-GGUF
This is quantized version of [NeverSleep/Llama-3-Lumimaid-8B-v0.1-OAS](https://huggingface.co/NeverSleep/Llama-3-Lumimaid-8B-v0.1-OAS?not-for-all-audiences=true) created using llama.cpp
# Model Description
<center><div style="width: 100%;">
<img src="https://cdn-uploads.huggingface.co/production/uploads/630dfb008df86f1e5becadc3/d3QMaxy3peFTpSlWdWF-k.png" style="display: block; margin: auto;">
</div></center>
This model uses the Llama3 **prompting format**
Llama3 trained on our RP datasets, we tried to have a balance between the ERP and the RP, not too horny, but just enough.
We also added some non-RP dataset, making the model less dumb overall. It should look like a 40%/60% ratio for Non-RP/RP+ERP data.
This model includes the new Luminae dataset from Ikari.
This model have received the Orthogonal Activation Steering treatment, meaning it will rarely refuse any request.
If you consider trying this model please give us some feedback either on the Community tab on hf or on our [Discord Server](https://discord.gg/MtCVRWTZXY).
## Credits:
- Undi
- IkariDev
## Description
This repo contains FP16 files of Lumimaid-8B-v0.1-OAS.
Switch: [8B](https://huggingface.co/NeverSleep/Llama-3-Lumimaid-8B-v0.1) - [70B](https://huggingface.co/NeverSleep/Llama-3-Lumimaid-70B-v0.1) - [70B-alt](https://huggingface.co/NeverSleep/Llama-3-Lumimaid-70B-v0.1-alt) - [8B-OAS](https://huggingface.co/NeverSleep/Llama-3-Lumimaid-8B-v0.1-OAS) - [70B-OAS](https://huggingface.co/NeverSleep/Llama-3-Lumimaid-70B-v0.1-OAS)
## Training data used:
- [Aesir datasets](https://huggingface.co/MinervaAI)
- [NoRobots](https://huggingface.co/datasets/Doctor-Shotgun/no-robots-sharegpt)
- [limarp](https://huggingface.co/datasets/lemonilia/LimaRP) - 8k ctx
- [toxic-dpo-v0.1-sharegpt](https://huggingface.co/datasets/Undi95/toxic-dpo-v0.1-sharegpt)
- [ToxicQAFinal](https://huggingface.co/datasets/NobodyExistsOnTheInternet/ToxicQAFinal)
- Luminae-i1 (70B/70B-alt) (i2 was not existing when the 70b started training) | Luminae-i2 (8B) (this one gave better results on the 8b) - Ikari's Dataset
- [Squish42/bluemoon-fandom-1-1-rp-cleaned](https://huggingface.co/datasets/Squish42/bluemoon-fandom-1-1-rp-cleaned) - 50% (randomly)
- [NobodyExistsOnTheInternet/PIPPAsharegptv2test](https://huggingface.co/datasets/NobodyExistsOnTheInternet/PIPPAsharegptv2test) - 5% (randomly)
- [cgato/SlimOrcaDedupCleaned](https://huggingface.co/datasets/cgato/SlimOrcaDedupCleaned) - 5% (randomly)
- Airoboros (reduced)
- [Capybara](https://huggingface.co/datasets/Undi95/Capybara-ShareGPT/) (reduced)
## Models used (only for 8B)
- Initial LumiMaid 8B Finetune
- Undi95/Llama-3-Unholy-8B-e4
- Undi95/Llama-3-LewdPlay-8B
## Prompt template: Llama3
```
<|begin_of_text|><|start_header_id|>system<|end_header_id|>
{system_prompt}<|eot_id|><|start_header_id|>user<|end_header_id|>
{input}<|eot_id|><|start_header_id|>assistant<|end_header_id|>
{output}<|eot_id|>
```
## Others
Undi: If you want to support us, you can [here](https://ko-fi.com/undiai).
IkariDev: Visit my [retro/neocities style website](https://ikaridevgit.github.io/) please kek |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.