File size: 1,510 Bytes
ab27367
 
 
 
 
 
 
cca197a
 
ab27367
bee9c4b
ab27367
34c4df4
ab27367
 
555b50d
ab27367
 
bee9c4b
ab27367
bee9c4b
ab27367
bee9c4b
ab27367
bee9c4b
ab27367
f8bd0c2
 
 
 
 
 
 
 
ab27367
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
---
license: apache-2.0
language:
- en
- th
library_name: transformers
pipeline_tag: text-generation
datasets:
- wannaphong/KhanomTanLLM-pretrained-dataset
---
# KhanomTan LLM (1B)

KhanomTan LLM is a bilingual language model trained in Thai and English from open source dataset by PyThaiNLP. We train the model from public dataset only. We public the dataset, source code, and model.


Repository: https://github.com/pythainlp/KhanomTanLLM


Codename: numfa-v2

## Model Details

### Model Description

The model was trained by [easylm](https://github.com/young-geng/EasyLM)

## Acknowledgements

Research supported with Cloud TPUs from Google's [TPU Research Cloud](https://sites.research.google/trc/about/) (TRC). We use TPU4-64 for training model about 4 days.

Thank you [TPU Research Cloud](https://sites.research.google/trc/about/) and [EasyLM project](https://github.com/young-geng/EasyLM)! We use EasyLM for pretraining model.

## How to use

**Example**

```python
# !pip install accelerate sentencepiece transformers bitsandbytes
import torch
from transformers import pipeline

pipe = pipeline("text-generation", model="numfa/numfa_v2-1b", 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

outputs = pipe("test is", max_new_tokens=300, do_sample=True, temperature=0.9, top_k=50, top_p=0.95, no_repeat_ngram_size=2,typical_p=1.)
print(outputs[0]["generated_text"])
```