modelId
stringlengths 4
81
| tags
list | pipeline_tag
stringclasses 17
values | config
dict | downloads
int64 0
59.7M
| first_commit
timestamp[ns, tz=UTC] | card
stringlengths 51
438k
|
|---|---|---|---|---|---|---|
Declan/CNN_model_v2 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 5 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Declan/CNN_model_v3 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
tags:
- generated_from_keras_callback
model-index:
- name: dung1308/RM_system_not_mixed__NLP_model_90_10_CPU_2_epochs
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. -->
# dung1308/RM_system_not_mixed__NLP_model_90_10_CPU_2_epochs
This model is a fine-tuned version of [vinai/phobert-base](https://huggingface.co/vinai/phobert-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Train Loss: 4.2989
- Validation Loss: 4.2424
- Epoch: 1
## 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': 'AdamWeightDecay', 'learning_rate': {'class_name': 'WarmUp', 'config': {'initial_learning_rate': 2e-05, 'decay_schedule_fn': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': -275, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, '__passive_serialization__': True}, 'warmup_steps': 1000, 'power': 1.0, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}
- training_precision: float32
### Training results
| Train Loss | Validation Loss | Epoch |
|:----------:|:---------------:|:-----:|
| 5.1315 | 4.5299 | 0 |
| 4.2989 | 4.2424 | 1 |
### Framework versions
- Transformers 4.18.0
- TensorFlow 2.8.0
- Datasets 2.7.0
- Tokenizers 0.11.0
|
Declan/CNN_model_v4 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Declan/CNN_model_v5 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Declan/ChicagoTribune_model_v7 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 7 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Declan/ChicagoTribune_model_v8 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 7 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Declan/FoxNews_model_v2 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Declan/FoxNews_model_v3 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 7 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Declan/HuffPost_model_v5 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Declan/HuffPost_model_v6 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 9 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Declan/NPR_model_v6 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | 2022-12-06T10:03:14Z | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Declan/NewYorkTimes_model_v1 | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | 2022-12-06T10:04:19Z | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Declan/NewYorkTimes_model_v2 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 7 | 2022-12-06T10:04:46Z | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeepBasak/Slack | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | 2022-12-06T10:13:31Z | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeepChem/ChemBERTa-10M-MLM | [
"pytorch",
"roberta",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"RobertaForMaskedLM"
],
"model_type": "roberta",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 90 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeepChem/ChemBERTa-5M-MLM | [
"pytorch",
"roberta",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"RobertaForMaskedLM"
],
"model_type": "roberta",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 29 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeepPavlov/distilrubert-tiny-cased-conversational-v1 | [
"pytorch",
"distilbert",
"ru",
"arxiv:2205.02340",
"transformers"
]
| null | {
"architectures": null,
"model_type": "distilbert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 9,141 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeepPavlov/rubert-base-cased-sentence | [
"pytorch",
"jax",
"bert",
"feature-extraction",
"ru",
"arxiv:1508.05326",
"arxiv:1809.05053",
"arxiv:1908.10084",
"transformers",
"has_space"
]
| feature-extraction | {
"architectures": [
"BertModel"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 46,991 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeepPavlov/rubert-base-cased | [
"pytorch",
"jax",
"bert",
"feature-extraction",
"ru",
"arxiv:1905.07213",
"transformers",
"has_space"
]
| feature-extraction | {
"architectures": [
"BertModel"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 148,127 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeepPavlov/xlm-roberta-large-en-ru-mnli | [
"pytorch",
"xlm-roberta",
"text-classification",
"en",
"ru",
"dataset:glue",
"dataset:mnli",
"transformers",
"xlm-roberta-large",
"xlm-roberta-large-en-ru",
"xlm-roberta-large-en-ru-mnli",
"has_space"
]
| text-classification | {
"architectures": [
"XLMRobertaForSequenceClassification"
],
"model_type": "xlm-roberta",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 227 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeepPavlov/xlm-roberta-large-en-ru | [
"pytorch",
"xlm-roberta",
"feature-extraction",
"en",
"ru",
"transformers"
]
| feature-extraction | {
"architectures": [
"XLMRobertaModel"
],
"model_type": "xlm-roberta",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 190 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeividasM/wav2vec2-large-xlsr-53-lithuanian | [
"pytorch",
"jax",
"wav2vec2",
"automatic-speech-recognition",
"lt",
"dataset:common_voice",
"transformers",
"audio",
"speech",
"xlsr-fine-tuning-week",
"license:apache-2.0",
"model-index"
]
| automatic-speech-recognition | {
"architectures": [
"Wav2Vec2ForCTC"
],
"model_type": "wav2vec2",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 7 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeltaHub/adapter_t5-3b_cola | [
"pytorch",
"transformers"
]
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeltaHub/adapter_t5-3b_mrpc | [
"pytorch",
"transformers"
]
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeltaHub/adapter_t5-3b_qnli | [
"pytorch",
"transformers"
]
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeltaHub/lora_t5-base_mrpc | [
"pytorch",
"transformers"
]
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DemangeJeremy/4-sentiments-with-flaubert | [
"pytorch",
"flaubert",
"text-classification",
"fr",
"transformers",
"sentiments",
"french",
"flaubert-large"
]
| text-classification | {
"architectures": [
"FlaubertForSequenceClassification"
],
"model_type": "flaubert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 226 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Denilson/gbert-base-germaner | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Deniskin/essays_small_2000i | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Deniskin/gpt3_medium | [
"pytorch",
"gpt2",
"text-generation",
"transformers",
"has_space"
]
| text-generation | {
"architectures": [
"GPT2LMHeadModel"
],
"model_type": "gpt2",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 52 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Denny29/DialoGPT-medium-asunayuuki | [
"pytorch",
"gpt2",
"text-generation",
"transformers",
"conversational"
]
| conversational | {
"architectures": [
"GPT2LMHeadModel"
],
"model_type": "gpt2",
"task_specific_params": {
"conversational": {
"max_length": 1000
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 9 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeskDown/MarianMixFT_en-fil | [
"pytorch",
"marian",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"MarianMTModel"
],
"model_type": "marian",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeskDown/MarianMixFT_en-hi | [
"pytorch",
"marian",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"MarianMTModel"
],
"model_type": "marian",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeskDown/MarianMixFT_en-id | [
"pytorch",
"marian",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"MarianMTModel"
],
"model_type": "marian",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeskDown/MarianMixFT_en-ja | [
"pytorch",
"marian",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"MarianMTModel"
],
"model_type": "marian",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 9 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeskDown/MarianMixFT_en-my | [
"pytorch",
"marian",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"MarianMTModel"
],
"model_type": "marian",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 7 | null | ---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
metrics:
- type: mean_reward
value: 268.09 +/- 15.44
name: mean_reward
verified: false
---
# **PPO** Agent playing **LunarLander-v2**
This is a trained model of a **PPO** agent playing **LunarLander-v2**
using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3).
## Usage (with Stable-baselines3)
TODO: Add your code
```python
from stable_baselines3 import ...
from huggingface_sb3 import load_from_hub
...
```
|
DeskDown/MarianMixFT_en-th | [
"pytorch",
"marian",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"MarianMTModel"
],
"model_type": "marian",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeskDown/MarianMixFT_en-vi | [
"pytorch",
"marian",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"MarianMTModel"
],
"model_type": "marian",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 5 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeskDown/MarianMix_en-ja-10 | [
"pytorch",
"tensorboard",
"marian",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"MarianMTModel"
],
"model_type": "marian",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 1 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeskDown/MarianMix_en-zh-10 | [
"pytorch",
"tensorboard",
"marian",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"MarianMTModel"
],
"model_type": "marian",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DeskDown/MarianMix_en-zh_to_vi-ms-hi-ja | [
"pytorch",
"tensorboard",
"marian",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"MarianMTModel"
],
"model_type": "marian",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 5 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Despin89/test | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Dev-DGT/food-dbert-multiling | [
"pytorch",
"distilbert",
"token-classification",
"transformers",
"autotrain_compatible"
]
| token-classification | {
"architectures": [
"DistilBertForTokenClassification"
],
"model_type": "distilbert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 17 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Devid/DialoGPT-small-Miku | [
"pytorch",
"gpt2",
"text-generation",
"transformers",
"conversational"
]
| conversational | {
"architectures": [
"GPT2LMHeadModel"
],
"model_type": "gpt2",
"task_specific_params": {
"conversational": {
"max_length": 1000
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 10 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Devmapall/paraphrase-quora | [
"pytorch",
"jax",
"t5",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"T5ForConditionalGeneration"
],
"model_type": "t5",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": true,
"length_penalty": 2,
"max_length": 200,
"min_length": 30,
"no_repeat_ngram_size": 3,
"num_beams": 4,
"prefix": "summarize: "
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": true,
"max_length": 300,
"num_beams": 4,
"prefix": "translate English to German: "
},
"translation_en_to_fr": {
"early_stopping": true,
"max_length": 300,
"num_beams": 4,
"prefix": "translate English to French: "
},
"translation_en_to_ro": {
"early_stopping": true,
"max_length": 300,
"num_beams": 4,
"prefix": "translate English to Romanian: "
}
}
} | 3 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Devrim/prism-default | [
"license:mit"
]
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DevsIA/Devs_IA | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DevsIA/imagenes | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DewiBrynJones/wav2vec2-large-xlsr-welsh | [
"cy",
"dataset:common_voice",
"audio",
"automatic-speech-recognition",
"speech",
"xlsr-fine-tuning-week",
"license:apache-2.0",
"model-index"
]
| automatic-speech-recognition | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
DheerajPranav/Dialo-GPT-Rick-bot | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language: en
tags:
- exbert
license: mit
---
# ColD Fusion model
Finetuned model that aims to be a great base model. It improves over RoBERTa base, trained on 35 datasets.
Full details at [this paper](https://arxiv.org/abs/2212.01378).
## Paper Abstract:
Pretraining has been shown to scale well with compute, data size and data diversity. Multitask learning trains on a
mixture of supervised datasets and produces improved performance compared to self-supervised pretraining. Until now,
massively multitask learning required simultaneous access to all datasets in the mixture and heavy compute resources
that are only available to well-resourced teams.
In this paper, we propose ColD Fusion, a method that provides the benefits of multitask learning but leverages distributed
computation and requires limited communication and no sharing of data. Consequentially, ColD Fusion can create a synergistic
loop, where finetuned models can be recycled to continually improve the pretrained model they are based on. We show that
ColD Fusion yields comparable benefits to multitask pretraining by producing a model that (a) attains strong performance on
all of the datasets it was multitask trained on and (b) is a better starting point for finetuning on unseen datasets. We find
ColD Fusion outperforms RoBERTa and even previous multitask models. Specifically, when training and testing on 35 diverse datasets,
ColD Fusion-based model outperforms RoBERTa by 2.45 points in average without any changes to the architecture.
### How to use
Best way to use is to finetune on your own task, but you can also extract features directly.
To get the features of a given text in PyTorch:
```python
from transformers import RobertaTokenizer, RobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = RobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
and in TensorFlow:
```python
from transformers import RobertaTokenizer, TFRobertaModel
tokenizer = RobertaTokenizer.from_pretrained('ibm/ColD-Fusion')
model = TFRobertaModel.from_pretrained('ibm/ColD-Fusion')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='tf')
output = model(encoded_input)
```
## Evaluation results
See full evaluation results of this model and many more [here](https://ibm.github.io/model-recycling/roberta-base_table.html)
When fine-tuned on downstream tasks, this model achieves the following results:
### BibTeX entry and citation info
```bibtex
@article{ColDFusion,
author = {Shachar Don-Yehiya, Elad Venezian, Colin Raffel, Noam Slonim, Yoav Katz, Leshem ChoshenYinhan Liu and},
title = {ColD Fusion: Collaborative Descent for Distributed Multitask Finetuning},
journal = {CoRR},
volume = {abs/2212.01378},
year = {2022},
url = {https://arxiv.org/abs/2212.01378},
archivePrefix = {arXiv},
eprint = {2212.01378},
}
```
<a href="https://huggingface.co/exbert/?model=ibm/ColD-Fusion">
<img width="300px" src="https://cdn-media.huggingface.co/exbert/button.png">
</a>
|
Dhito/am | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null |
---
tags:
- unity-ml-agents
- ml-agents
- deep-reinforcement-learning
- reinforcement-learning
- ML-Agents-Pyramids
library_name: ml-agents
---
# **ppo** Agent playing **Pyramids**
This is a trained model of a **ppo** agent playing **Pyramids** using the [Unity ML-Agents Library](https://github.com/Unity-Technologies/ml-agents).
## Usage (with ML-Agents)
The Documentation: https://github.com/huggingface/ml-agents#get-started
We wrote a complete tutorial to learn to train your first agent using ML-Agents and publish it to the Hub:
### Resume the training
```
mlagents-learn <your_configuration_file_path.yaml> --run-id=<run_id> --resume
```
### Watch your Agent play
You can watch your agent **playing directly in your browser:**.
1. Go to https://huggingface.co/spaces/unity/ML-Agents-Pyramids
2. Step 1: Write your model_id: sun1638650145/ML-Agents-Pyramids
3. Step 2: Select your *.nn /*.onnx file
4. Click on Watch the agent play 👀
|
DicoTiar/wisdomfiy | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3 | null | ---
language:
- zh
license: apache-2.0
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
metrics:
- wer
model-index:
- name: Whisper Small zh - howl
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 11.0
type: mozilla-foundation/common_voice_11_0
config: zh-CN
split: test
args: 'config: zh, split: test'
metrics:
- name: Wer
type: wer
value: 75.2976752976753
---
<!-- 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. -->
# Whisper Small zh - howl
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Common Voice 11.0 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3644
- Wer: 75.2977
## 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: 1e-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_steps: 500
- training_steps: 2000
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 0.2309 | 1.51 | 1000 | 0.3694 | 76.4411 |
| 0.1069 | 3.02 | 2000 | 0.3644 | 75.2977 |
### Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu116
- Datasets 2.7.1
- Tokenizers 0.13.2
|
DiegoAlysson/opus-mt-en-ro-finetuned-en-to-ro | [
"pytorch",
"tensorboard",
"marian",
"text2text-generation",
"dataset:wmt16",
"transformers",
"generated_from_trainer",
"license:apache-2.0",
"model-index",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"MarianMTModel"
],
"model_type": "marian",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 1 | null | ---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
model-index:
- name: distilbert-base-uncased-finetuned-emotion
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. -->
# distilbert-base-uncased-finetuned-emotion
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2203
- Accuracy: 0.9245
- F1: 0.9246
## 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: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.8038 | 1.0 | 250 | 0.3028 | 0.913 | 0.9115 |
| 0.246 | 2.0 | 500 | 0.2203 | 0.9245 | 0.9246 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.5.1
- Tokenizers 0.11.6
|
DimaOrekhov/cubert-method-name | [
"pytorch",
"encoder-decoder",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"EncoderDecoderModel"
],
"model_type": "encoder-decoder",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 10 | null | ---
language: en
license: mit
tags:
- vision
model_name: microsoft/git-base-vqav2
inference: false
pipeline_tag: visual-question-answering
---
# GIT (GenerativeImage2Text), base-sized, fine-tuned on VQAv2
GIT (short for GenerativeImage2Text) model, base-sized version, fine-tuned on VQAv2. It was introduced in the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Wang et al. and first released in [this repository](https://github.com/microsoft/GenerativeImage2Text).
Disclaimer: The team releasing GIT did not write a model card for this model so this model card has been written by the Hugging Face team.
## Model description
GIT is a Transformer decoder conditioned on both CLIP image tokens and text tokens. The model is trained using "teacher forcing" on a lot of (image, text) pairs.
The goal for the model is simply to predict the next text token, giving the image tokens and previous text tokens.
The model has full access to (i.e. a bidirectional attention mask is used for) the image patch tokens, but only has access to the previous text tokens (i.e. a causal attention mask is used for the text tokens) when predicting the next text token.
This allows the model to be used for tasks like:
- image and video captioning
- visual question answering (VQA) on images and videos
- even image classification (by simply conditioning the model on the image and asking it to generate a class for it in text).
## Intended uses & limitations
You can use the raw model for visual question answering (VQA). See the [model hub](https://huggingface.co/models?search=microsoft/git) to look for
fine-tuned versions on a task that interests you.
### How to use
For code examples, we refer to the [documentation](https://huggingface.co/docs/transformers/main/model_doc/git#transformers.GitForCausalLM.forward.example-2).
## Training data
From the paper:
> We collect 0.8B image-text pairs for pre-training, which include COCO (Lin et al., 2014), Conceptual Captions
(CC3M) (Sharma et al., 2018), SBU (Ordonez et al., 2011), Visual Genome (VG) (Krishna et al., 2016),
Conceptual Captions (CC12M) (Changpinyo et al., 2021), ALT200M (Hu et al., 2021a), and an extra 0.6B
data following a similar collection procedure in Hu et al. (2021a).
=> however this is for the model referred to as "GIT" in the paper, which is not open-sourced.
This checkpoint is "GIT-base", which is a smaller variant of GIT trained on 10 million image-text pairs.
Next, the model was fine-tuned on VQAv2.
See table 11 in the [paper](https://arxiv.org/abs/2205.14100) for more details.
### Preprocessing
We refer to the original repo regarding details for preprocessing during training.
During validation, one resizes the shorter edge of each image, after which center cropping is performed to a fixed-size resolution. Next, frames are normalized across the RGB channels with the ImageNet mean and standard deviation.
## Evaluation results
For evaluation results, we refer readers to the [paper](https://arxiv.org/abs/2205.14100). |
DimaOrekhov/transformer-method-name | [
"pytorch",
"encoder-decoder",
"text2text-generation",
"transformers",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"EncoderDecoderModel"
],
"model_type": "encoder-decoder",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 8 | 2022-12-06T11:08:00Z | ---
language:
- hi
license: apache-2.0
tags:
- whisper-event
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
metrics:
- wer
model-index:
- name: Whisper Small Hindi
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: mozilla-foundation/common_voice_11_0 hi
type: mozilla-foundation/common_voice_11_0
config: hi
split: test
args: hi
metrics:
- name: Wer
type: wer
value: 22.429210134128166
---
<!-- 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. -->
# Whisper Small Hindi
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the mozilla-foundation/common_voice_11_0 hi dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6260
- Wer: 22.4292
## 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: 7e-06
- train_batch_size: 64
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 3000
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 0.0176 | 7.01 | 500 | 0.4165 | 22.5066 |
| 0.0015 | 14.01 | 1000 | 0.5186 | 22.2573 |
| 0.0004 | 21.02 | 1500 | 0.5741 | 22.2401 |
| 0.0002 | 28.02 | 2000 | 0.6025 | 22.3834 |
| 0.0002 | 36.01 | 2500 | 0.6197 | 22.3977 |
| 0.0002 | 43.01 | 3000 | 0.6260 | 22.4292 |
### Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu117
- Datasets 2.7.1.dev0
- Tokenizers 0.13.2
|
DivyanshuSheth/T5-Seq2Seq-Final | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language: en
license: mit
tags:
- vision
model_name: microsoft/git-base-textvqa
inference: false
pipeline_tag: visual-question-answering
---
# GIT (GenerativeImage2Text), base-sized, fine-tuned on TextVQA
GIT (short for GenerativeImage2Text) model, base-sized version, fine-tuned on TextVQA. It was introduced in the paper [GIT: A Generative Image-to-text Transformer for Vision and Language](https://arxiv.org/abs/2205.14100) by Wang et al. and first released in [this repository](https://github.com/microsoft/GenerativeImage2Text).
Disclaimer: The team releasing GIT did not write a model card for this model so this model card has been written by the Hugging Face team.
## Model description
GIT is a Transformer decoder conditioned on both CLIP image tokens and text tokens. The model is trained using "teacher forcing" on a lot of (image, text) pairs.
The goal for the model is simply to predict the next text token, giving the image tokens and previous text tokens.
The model has full access to (i.e. a bidirectional attention mask is used for) the image patch tokens, but only has access to the previous text tokens (i.e. a causal attention mask is used for the text tokens) when predicting the next text token.
This allows the model to be used for tasks like:
- image and video captioning
- visual question answering (VQA) on images and videos
- even image classification (by simply conditioning the model on the image and asking it to generate a class for it in text).
## Intended uses & limitations
You can use the raw model for visual question answering (VQA). See the [model hub](https://huggingface.co/models?search=microsoft/git) to look for
fine-tuned versions on a task that interests you.
### How to use
For code examples, we refer to the [documentation](https://huggingface.co/transformers/main/model_doc/git.html).
## Training data
From the paper:
> We collect 0.8B image-text pairs for pre-training, which include COCO (Lin et al., 2014), Conceptual Captions
(CC3M) (Sharma et al., 2018), SBU (Ordonez et al., 2011), Visual Genome (VG) (Krishna et al., 2016),
Conceptual Captions (CC12M) (Changpinyo et al., 2021), ALT200M (Hu et al., 2021a), and an extra 0.6B
data following a similar collection procedure in Hu et al. (2021a).
=> however this is for the model referred to as "GIT" in the paper, which is not open-sourced.
This checkpoint is "GIT-base", which is a smaller variant of GIT trained on 10 million image-text pairs.
Next, the model was fine-tuned on TextVQA.
See table 11 in the [paper](https://arxiv.org/abs/2205.14100) for more details.
### Preprocessing
We refer to the original repo regarding details for preprocessing during training.
During validation, one resizes the shorter edge of each image, after which center cropping is performed to a fixed-size resolution. Next, frames are normalized across the RGB channels with the ImageNet mean and standard deviation.
## Evaluation results
For evaluation results, we refer readers to the [paper](https://arxiv.org/abs/2205.14100). |
Dmitriiserg/Pxd | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language: et
license: cc-by-4.0
datasets:
- ERRnews
---
# mBART ERRnews
Pretrained mbart-large-cc25 model finetuned on ERRnews Estonian news story dataset.
## How to use
Here is how to use this model to get a summary of a given text in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("TalTechNLP/mBART-ERRnews")
model = AutoModelForSeq2SeqLM.from_pretrained("TalTechNLP/mBART-ERRnews")
text = "Riigikogu rahanduskomisjon võttis esmaspäeval maha riigieelarvesse esitatud investeeringuettepanekutest siseministeeriumi investeeringud koolidele ja lasteaedadele, sest komisjoni hinnangul ei peaks siseministeerium tegelema investeeringutega väljaspoole oma vastutusala. Komisjoni esimees Aivar Kokk ütles, et komisjon lähtus otsuse tegemisel riigikontrolör Janar Holmi soovitusest ja seadustest."
inputs = tokenizer(text, return_tensors='pt', max_length=1024)
summary_ids = model.generate(inputs['input_ids'])
summary = [tokenizer.decode(g, skip_special_tokens=True, clean_up_tokenization_spaces=False) for g in summary_ids]
```
## Training data
The mBART model was finetuned on [ERRnews](https://huggingface.co/datasets/TalTechNLP/ERRnews), a dataset consisting of 10 420
Estonian news story transcripts and summaries.
### Training
The model was trained on 2 cloud GPUs with a batch size of 16 for 16 epochs. The optimizer
used is Adam with a learning rate of 5e-05, betas of 0.9 and 0.999.
## Evaluation results
This model achieves the following results:
| Dataset | ROUGE-1 | ROUGE-2 | ROUGE-L | ROUGE-L-SUM |
|:-------:|:-------:|:-------:|:-------:|:-----------:|
| ERRnews | 19.2 | 6.7 | 16.1 | 17.4 |
### BibTeX entry and citation info
```bibtex
article{henryabstractive,
title={Abstractive Summarization of Broadcast News Stories for {Estonian}},
author={Henry, H{\"a}rm and Tanel, Alum{\"a}e},
journal={Baltic J. Modern Computing},
volume={10},
number={3},
pages={511-524},
year={2022}
}
``` |
Doiman/DialoGPT-medium-harrypotter | [
"pytorch",
"gpt2",
"text-generation",
"transformers",
"conversational"
]
| conversational | {
"architectures": [
"GPT2LMHeadModel"
],
"model_type": "gpt2",
"task_specific_params": {
"conversational": {
"max_length": 1000
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 13 | null | ---
language:
- sv
license: apache-2.0
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
model-index:
- name: my_tuned_whisper_cn
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. -->
# my_tuned_whisper_cn
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Common Voice 11.0 dataset.
It achieves the following results on the evaluation set:
- eval_loss: 0.5297
- eval_wer: 80.2457
- eval_runtime: 457.7207
- eval_samples_per_second: 2.311
- eval_steps_per_second: 0.291
- epoch: 2.02
- step: 1000
## 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: 1e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 4000
- mixed_precision_training: Native AMP
### Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2
|
DongHyoungLee/distilbert-base-uncased-finetuned-cola | [
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"dataset:glue",
"transformers",
"generated_from_trainer",
"license:apache-2.0",
"model-index"
]
| text-classification | {
"architectures": [
"DistilBertForSequenceClassification"
],
"model_type": "distilbert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 27 | null | ---
language:
- it
license: apache-2.0
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
metrics:
- wer
model-index:
- name: Whisper Tiny It 3 - Gianluca Ruberto
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 11.0
type: mozilla-foundation/common_voice_11_0
config: it
split: test[:10%]
args: 'config: hi, split: test'
metrics:
- name: Wer
type: wer
value: 43.233499722684414
---
# Whisper Tiny It 3 - Gianluca Ruberto
This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the Common Voice 11.0 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.711673
- Wer: 43.233500
## Model description
This model is the openai whisper small transformer adapted for Italian audio to text transcription. This model has weight decay set to 0.1 to cope with overfitting.
## Intended uses & limitations
The model is available through its [HuggingFace web app](https://huggingface.co/spaces/GIanlucaRub/whisper-it)
## Training and evaluation data
Data used for training is the initial 10% of train and validation of [Italian Common Voice](https://huggingface.co/datasets/mozilla-foundation/common_voice_11_0/viewer/it/train) 11.0 from Mozilla Foundation.
The dataset used for evaluation is the initial 10% of test of Italian Common Voice.
Weight decay showed to have slightly better result also on the evaluation dataset.
## Training procedure
After loading the pre trained model, it has been trained on the dataset.
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1e-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_steps: 500
- training_steps: 4000
- mixed_precision_training: Native AMP
- weight_decay: 0.1
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 0.5837 | 0.95 | 1000 | 0.790374 | 50.2981 |
| 0.4183 | 1.91 | 2000 | 0.730100 | 45.4174 |
| 0.3147 | 2.86 | 3000 | 0.713152 | 44.3150 |
| 0.2670 | 3.82 | 4000 | 0.711673 | 43.2335 |
### Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2
|
Doogie/Waynehills-KE-T5-doogie | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
license: mit
tags:
- pytorch
- diffusers
- unconditional-image-generation
- diffusion-models-class
---
# Model Card for Unit 1 of the [Diffusion Models Class 🧨](https://github.com/huggingface/diffusion-models-class)
This model is a diffusion model for unconditional image generation of cute 🦋.
## Usage
```python
from diffusers import DDPMPipeline
pipeline = DDPMPipeline.from_pretrained('fluorine/sd-class-butterflies-32')
image = pipeline().images[0]
image
```
|
Waynehillsdev/Waynehills_summary_tensorflow | [
"tf",
"t5",
"text2text-generation",
"transformers",
"generated_from_keras_callback",
"autotrain_compatible"
]
| text2text-generation | {
"architectures": [
"T5ForConditionalGeneration"
],
"model_type": "t5",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 5 | null | ---
license: mit
tags:
- generated_from_trainer
model-index:
- name: recipe-nlg-gpt2-ingredient-to-recipe-model
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. -->
# recipe-nlg-gpt2-ingredient-to-recipe-model
This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on the None dataset.
## 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: 8
- 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_steps: 200
- num_epochs: 1
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0
- Datasets 2.7.1
- Tokenizers 0.13.2
|
Doquey/DialoGPT-small-Luisbot1 | [
"pytorch",
"gpt2",
"text-generation",
"transformers",
"conversational"
]
| conversational | {
"architectures": [
"GPT2LMHeadModel"
],
"model_type": "gpt2",
"task_specific_params": {
"conversational": {
"max_length": 1000
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 7 | null | ---
tags:
- LunarLander-v2
- ppo
- deep-reinforcement-learning
- reinforcement-learning
- custom-implementation
- deep-rl-course
model-index:
- name: PPO
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
metrics:
- type: mean_reward
value: -204.64 +/- 88.46
name: mean_reward
verified: false
---
# PPO Agent Playing LunarLander-v2
This is a trained model of a PPO agent playing LunarLander-v2.
# Hyperparameters
```python
{'exp_name': 'ppo.py'
'gym_id': 'LunarLander-v2'
'seed': 1
'learning_rate': 0.00025
'total_timesteps': 25000
'torch_deterministic': True
'cuda': True
'track': False
'wandb_project_name': 'cleanRL'
'wandb_entity': None
'capture_video': False
'env_id': 'LunarLander-v2'
'num_envs': 4
'num_steps': 128
'anneal_lr': True
'gae': True
'gamma': 0.99
'gae_lambda': 0.95
'num_minibatches': 4
'update_epochs': 4
'norm_adv': True
'clip_coef': 0.2
'clip_vloss': True
'ent_coef': 0.01
'vf_coef': 0.5
'max_grad_norm': 0.5
'target_kl': None
'repo_id': 'lsaulier/ppo-LunarLander-v2'
'batch_size': 512
'minibatch_size': 128}
```
|
Doxophobia/DialoGPT-medium-celeste | [
"pytorch",
"gpt2",
"text-generation",
"transformers",
"conversational"
]
| conversational | {
"architectures": [
"GPT2LMHeadModel"
],
"model_type": "gpt2",
"task_specific_params": {
"conversational": {
"max_length": 1000
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 11 | null | ---
license: mit
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- accuracy
model-index:
- name: roberta-large-finetuned-mnli-batch_size_4_100000_samples
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: glue
type: glue
config: mnli
split: train
args: mnli
metrics:
- name: Accuracy
type: accuracy
value: 0.3544574630667346
---
<!-- 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-large-finetuned-mnli-batch_size_4_100000_samples
This model is a fine-tuned version of [roberta-large](https://huggingface.co/roberta-large) on the glue dataset.
It achieves the following results on the evaluation set:
- Loss: 1.0980
- Accuracy: 0.3545
## 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: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
| 1.1026 | 1.0 | 25000 | 1.0980 | 0.3545 |
### Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2
|
DoyyingFace/bert-asian-hate-tweets-asian-unclean-freeze-8 | [
"pytorch",
"bert",
"text-classification",
"transformers"
]
| text-classification | {
"architectures": [
"BertForSequenceClassification"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 30 | null | ---
license: creativeml-openrail-m
---
Science Fiction/Horror monster textual embedding for Stable Diffusion 2.0.
This embedding is trained initially on 49 images from Tod Ryan's Artstation (https://www.artstation.com/todryan), then further tuned with an expanded dataset that includes 119 additional images generated with the initial embedding alongside specific prompting tailored to improving the quality. These generated training images were color graded collectively to mimic the visual aesthetic of modern horror media.
I have also included the initial version of the embedding that circulated on the Stable Diffusion discord. It is excellent for disgusting (but repetitive) monster/grossness.
Example generations:
_Prompt: Macro Terror, Steps: 15, Sampler: DPM++ SDE Karras, CFG scale: 3.5, Seed: 2889499141, Size: 768x768, Model hash: 2c02b20a_
_Prompt: Macro Terror, Steps: 15, Sampler: DPM++ SDE Karras, CFG scale: 7, Seed: 2324809867, Size: 768x768, Model hash: 2c02b20a_
_Prompt: Macro Terror, Steps: 15, Sampler: DPM++ 2S a, CFG scale: 5, Seed: 2276531391, Size: 768x768, Model hash: 2c02b20a_ |
DoyyingFace/bert-asian-hate-tweets-asian-unclean-warmup-25 | [
"pytorch",
"bert",
"text-classification",
"transformers"
]
| text-classification | {
"architectures": [
"BertForSequenceClassification"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 30 | null | ---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
metrics:
- type: mean_reward
value: 248.96 +/- 25.61
name: mean_reward
verified: false
---
# **PPO** Agent playing **LunarLander-v2**
This is a trained model of a **PPO** agent playing **LunarLander-v2**
using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3).
## Usage (with Stable-baselines3)
TODO: Add your code
```python
from stable_baselines3 import ...
from huggingface_sb3 import load_from_hub
...
```
|
DoyyingFace/bert-asian-hate-tweets-concat-clean-with-unclean-valid | [
"pytorch",
"bert",
"text-classification",
"transformers"
]
| text-classification | {
"architectures": [
"BertForSequenceClassification"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 25 | null | ---
license: mit
tags:
- generated_from_trainer
model-index:
- name: camembert-base-squad-fr
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. -->
# camembert-base-squad-fr
This model is a fine-tuned version of [camembert-base](https://huggingface.co/camembert-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.5182
## 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: 3e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.2
- num_epochs: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 1.7504 | 1.0 | 3581 | 1.6470 |
| 1.4776 | 2.0 | 7162 | 1.5182 |
### Framework versions
- Transformers 4.24.0
- Pytorch 1.13.0+cu117
- Datasets 2.7.1
- Tokenizers 0.13.2
|
DoyyingFace/bert-asian-hate-tweets-concat-clean | [
"pytorch",
"bert",
"text-classification",
"transformers"
]
| text-classification | {
"architectures": [
"BertForSequenceClassification"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 25 | null | ---
license: mit
tags:
- generated_from_trainer
model-index:
- name: bert-base-historic-multilingual-cased-squad-fr
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. -->
# bert-base-historic-multilingual-cased-squad-fr
This model is a fine-tuned version of [dbmdz/bert-base-historic-multilingual-cased](https://huggingface.co/dbmdz/bert-base-historic-multilingual-cased) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.7001
## 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: 3e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.2
- num_epochs: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 1.9769 | 1.0 | 3660 | 1.8046 |
| 1.6309 | 2.0 | 7320 | 1.7001 |
### Framework versions
- Transformers 4.24.0
- Pytorch 1.13.0+cu117
- Datasets 2.7.1
- Tokenizers 0.13.2
|
albert-base-v1 | [
"pytorch",
"tf",
"safetensors",
"albert",
"fill-mask",
"en",
"dataset:bookcorpus",
"dataset:wikipedia",
"arxiv:1909.11942",
"transformers",
"exbert",
"license:apache-2.0",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"AlbertForMaskedLM"
],
"model_type": "albert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 38,156 | 2022-12-06T13:03:24Z | ---
license: mit
tags:
- generated_from_trainer
model-index:
- name: bert-base-french-europeana-cased-squad-fr
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. -->
# bert-base-french-europeana-cased-squad-fr
This model is a fine-tuned version of [dbmdz/bert-base-french-europeana-cased](https://huggingface.co/dbmdz/bert-base-french-europeana-cased) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.7031
## 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: 3e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.2
- num_epochs: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 1.9069 | 1.0 | 3539 | 1.7853 |
| 1.6263 | 2.0 | 7078 | 1.7031 |
### Framework versions
- Transformers 4.24.0
- Pytorch 1.13.0+cu117
- Datasets 2.7.1
- Tokenizers 0.13.2
|
albert-xlarge-v1 | [
"pytorch",
"tf",
"albert",
"fill-mask",
"en",
"dataset:bookcorpus",
"dataset:wikipedia",
"arxiv:1909.11942",
"transformers",
"license:apache-2.0",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"AlbertForMaskedLM"
],
"model_type": "albert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 341 | 2022-12-06T13:09:03Z | ---
license: wtfpl
---
Cat picture embedding for 2.0. Trained on high quality Unsplash images, so it tends to prefer photorealism.
Warning: the weights are quite strong. But, when tamed, it works great with stylistic embeddings like the last couple of images!
Trained for 1500 steps, but added the 1000 steps one as well which also works pretty decently and is a bit less strong.
![03171-4193474301-cute [kittyhelper] wearing a tuxedo, sharp focus, photohelper, fur, extremely detailed.png](https://s3.amazonaws.com/moonup/production/uploads/1670332296701-6312579fc7577b68d90a7646.png)
![03164-4193474294-cute [kittyhelper] wearing a tuxedo, sharp focus, photohelper, fur, extremely detailed.png](https://s3.amazonaws.com/moonup/production/uploads/1670332307002-6312579fc7577b68d90a7646.png)
 |
bert-base-cased | [
"pytorch",
"tf",
"jax",
"safetensors",
"bert",
"fill-mask",
"en",
"dataset:bookcorpus",
"dataset:wikipedia",
"arxiv:1810.04805",
"transformers",
"exbert",
"license:apache-2.0",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 8,621,271 | 2022-12-06T13:28:42Z | ---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: Proximal Policy Optimisation (PPO)
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
metrics:
- type: mean_reward
value: 290.88 +/- 17.28
name: mean_reward
verified: false
---
# **Proximal Policy Optimisation (PPO)** Agent playing **LunarLander-v2**
This is a trained model of a **Proximal Policy Optimisation (PPO)** agent playing **LunarLander-v2**
using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3).
## Usage (with Stable-baselines3)
TODO: Add your code
```python
from stable_baselines3 import ...
from huggingface_sb3 import load_from_hub
...
```
|
bert-base-chinese | [
"pytorch",
"tf",
"jax",
"safetensors",
"bert",
"fill-mask",
"zh",
"arxiv:1810.04805",
"transformers",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 3,377,486 | 2022-12-06T13:30:14Z | ---
license: apache-2.0
tags:
- generated_from_keras_callback
model-index:
- name: Gorenzelg/bert-finetuned-squad11
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. -->
# Gorenzelg/bert-finetuned-squad11
This model is a fine-tuned version of [bert-base-cased](https://huggingface.co/bert-base-cased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Train Loss: 1.0664
- Epoch: 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:
- optimizer: {'inner_optimizer': {'class_name': 'AdamWeightDecay', 'config': {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 55450, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}}, 'dynamic': True, 'initial_scale': 32768.0, 'dynamic_growth_steps': 2000}
- training_precision: mixed_float16
### Training results
| Train Loss | Epoch |
|:----------:|:-----:|
| 1.0664 | 0 |
### Framework versions
- Transformers 4.24.0
- TensorFlow 2.10.1
- Datasets 2.6.1
- Tokenizers 0.11.0
|
bert-base-german-cased | [
"pytorch",
"tf",
"jax",
"safetensors",
"bert",
"fill-mask",
"de",
"transformers",
"exbert",
"license:mit",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 175,983 | 2022-12-06T13:32:44Z | ---
language: ru
datasets:
- bond005/sberdevices_golos_10h_crowd
- bond005/sberdevices_golos_100h_farfield
- common_voice
- bond005/sova_rudevices
- bond005/rulibrispeech
metrics:
- wer
- cer
tags:
- audio
- automatic-speech-recognition
- speech
- common_voice
- SberDevices/Golos
- sova_rudevices
- rulibrispeech
license: apache-2.0
widget:
- example_title: test sound with Russian speech
src: https://huggingface.co/bond005/wav2vec2-mbart50-ru/resolve/main/test_sound.wav
model-index:
- name: Wav2Vec2-mBART-50 for speech-to-text in Russian by Ivan Bondarenko
results:
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: Sberdevices Golos (crowd)
type: SberDevices/Golos
args: ru
metrics:
- name: Test WER
type: wer
value: 13.204
- name: Test CER
type: cer
value: 4.157
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: Sberdevices Golos (farfield)
type: SberDevices/Golos
args: ru
metrics:
- name: Test WER
type: wer
value: 17.681
- name: Test CER
type: cer
value: 6.773
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice ru
type: common_voice
args: ru
metrics:
- name: Test WER
type: wer
value: 14.693
- name: Test CER
type: cer
value: 5.765
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: Sova RuDevices
type: sova_rudevices
args: ru
metrics:
- name: Test WER
type: wer
value: 22.727
- name: Test CER
type: cer
value: 9.183
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: Russian Librispeech
type: rulibrispeech
args: ru
metrics:
- name: Test WER
type: wer
value: 32.540
- name: Test CER
type: cer
value: 10.369
---
# Wav2Vec2-mBART-50-Ru
Wav2Vec2-mBART-50-Ru is a speech-sequence-to-text-sequence model, which can convert an input audio with Russian speech into a text with punctuation, capitalization and so on.
Wav2Vec2-mBART-50-Ru is the [SpeechEncoderDecoderModel](https://huggingface.co/docs/transformers/model_doc/speech-encoder-decoder), which was initialized with [Wav2Vec2-Large-Ru-Golos](https://huggingface.co/bond005/wav2vec2-large-ru-golos) as the encoder and [mBART-large-50](https://huggingface.co/facebook/mbart-large-50) as the decoder. After its initialization the model was fine-tuned using the training parts of several annotated speech corpora:
- [the 10 hours crowd subset of SberDevices Golos](https://huggingface.co/datasets/bond005/sberdevices_golos_10h_crowd)
- [the 100 hours farfield subset of SberDevices Golos](https://huggingface.co/datasets/bond005/sberdevices_golos_100h_farfield)
- [the Russian subset of Common Voice 6.0](https://huggingface.co/datasets/common_voice)
- [Sova RuDevices](https://huggingface.co/datasets/bond005/sova_rudevices)
- 15% part of the training subset of [Russian Librispeech](https://huggingface.co/datasets/bond005/rulibrispeech)
CommonVoice 6.0 contains "rich" text annotations with punctuation and capitalization, but other speech corpora includes plain texts only. Therefore, text annotations of these corpora were riched automatically using the [Silero text enhancement model](https://github.com/snakers4/silero-models#text-enhancement).
## Usage
When using this model, make sure that your speech input is sampled at 16kHz.
You can use this model by writing your own inference script:
```python
import os
import warnings
import torch
from datasets import load_dataset
from datasets.features import Audio
from transformers import SpeechEncoderDecoderModel, Wav2Vec2Processor
LANG_ID = "ru"
MODEL_ID = "bond005/wav2vec2-mbart50-ru"
SAMPLES = 30
num_processes = max(1, os.cpu_count())
processor = Wav2Vec2Processor.from_pretrained(MODEL_ID)
model = SpeechEncoderDecoderModel.from_pretrained(MODEL_ID)
test_dataset = load_dataset("common_voice", LANG_ID, split=f"test[:{SAMPLES}]")
if test_dataset.features['audio'].sampling_rate != 16_000:
test_dataset = test_dataset.cast_column(
'audio',
Audio(sampling_rate=16_000)
)
audio_data = [test_dataset[i]['audio']['array'] for i in range(SAMPLES)]
processed = processor(audio_data, sampling_rate=16_000,
return_tensors="pt", padding='longest')
with torch.no_grad():
predicted_ids = model.generate(**processed)
predicted_sentences = processor.batch_decode(
predicted_ids,
num_processes=num_processes,
skip_special_tokens=True
)
with warnings.catch_warnings():
warnings.simplefilter("ignore")
for i, predicted_sentence in enumerate(predicted_sentences):
print("-" * 100)
print("Reference: ", test_dataset[i]["sentence"])
print("Prediction:", predicted_sentence)
```
```text
----------------------------------------------------------------------------------------------------
Reference: Я беру маленький кусочек бумажки.
Prediction: Я беру маленькие кусочек бумажки.
----------------------------------------------------------------------------------------------------
Reference: О потерях пока не сообщается.
Prediction: А потеря их пока не сообщается.
----------------------------------------------------------------------------------------------------
Reference: Ваша воля.
Prediction: Ваша воля.
----------------------------------------------------------------------------------------------------
Reference: Мы высоко ценим ее роль в этом отношении.
Prediction: Мы высоко ценим ее роль в этом отношении.
----------------------------------------------------------------------------------------------------
Reference: Вот это вызывало у нас жуткое отторжение.
Prediction: Вот это вызвало у нас жуткое отвержение.
----------------------------------------------------------------------------------------------------
Reference: Он положил ей букет на книгу.
Prediction: Он положил ее букет на книгу.
----------------------------------------------------------------------------------------------------
Reference: Ну и положу, – обиделась Женя.
Prediction: – Ну и положи, – обиделась Женя.
----------------------------------------------------------------------------------------------------
Reference: Благодарю представителя Австралии за ее заявление.
Prediction: Благодарю представителя Австралии за ее заявление.
----------------------------------------------------------------------------------------------------
Reference: Для меня это не было неожиданностью.
Prediction: Для меня это не было неожиданностью.
----------------------------------------------------------------------------------------------------
Reference: Поздняя ночь.
Prediction: Поздняя ночь.
----------------------------------------------------------------------------------------------------
Reference: Тем не менее нужно вновь вычленить некоторые элементы наших политических установок.
Prediction: Тем не менее нужно назвать нищие нынешние элементы наших политических устоков.
----------------------------------------------------------------------------------------------------
Reference: Мы не можем позволить себе упустить эту возможность.
Prediction: Мы не можем позволить себе упустить эту возможность.
----------------------------------------------------------------------------------------------------
Reference: В предстоящие месяцы Суд примет решение по ордеру на арест министра обороны Хусейна.
Prediction: В предстоящие месяцы Суд примет решение по оратору на орифлейм министра иностранных дел Кубы.
----------------------------------------------------------------------------------------------------
Reference: Валерия живет в старом панельном доме советских времён.
Prediction: Валерия живет в старом Баньяном, да не советских временах.
----------------------------------------------------------------------------------------------------
Reference: Я вернусь скоро.
Prediction: Я вернусь скоро...
----------------------------------------------------------------------------------------------------
Reference: Слово предоставляется Его Превосходительству принцу Зайду.
Prediction: Слово предоставляется Его Превосходительству Пан Ги Муну.
----------------------------------------------------------------------------------------------------
Reference: Ну конечно, тебе бы этого хотелось.
Prediction: Ну, конечно, тебе бы этого хотелось.
----------------------------------------------------------------------------------------------------
Reference: Общественные объединения равны перед законом.
Prediction: Общественные объединения равны перед законом.
----------------------------------------------------------------------------------------------------
Reference: Ну, что же, нету этики, эстетики.
Prediction: Ну что же, не туда зайти? Не туда зайти?
----------------------------------------------------------------------------------------------------
Reference: Сразу же она легла в постель.
Prediction: Сразу же она легла в постель.
----------------------------------------------------------------------------------------------------
Reference: Сейчас я сделаю заявление в своем национальном качестве.
Prediction: Сейчас я сделаю заявление в своем национальном качестве.
----------------------------------------------------------------------------------------------------
Reference: Что там сейчас происходит в Твиттере?
Prediction: Что там сейчас происходит в Твиттере?
----------------------------------------------------------------------------------------------------
Reference: Ну хорошо, что револьвер был заряжен холостыми.
Prediction: Ну хорошо, что Револьвер был заряжен холостыми.
----------------------------------------------------------------------------------------------------
Reference: А потом дальше может проходить работа такая.
Prediction: А потом дальше может проходить работа такая.
----------------------------------------------------------------------------------------------------
Reference: Из Microsoft написали что на текущий момент у них нет открытых вакансий.
Prediction: Из моих красотов написали, что на текущий момент у них нет открытых вакансий.
----------------------------------------------------------------------------------------------------
Reference: Мы добились многого, но сейчас не время терять набранную динамику.
Prediction: Мы добились многого, но сейчас не время терять набранную динамику.
----------------------------------------------------------------------------------------------------
Reference: Мы внимательно проанализировали документ и содержащиеся в нем выводы и рекомендации.
Prediction: Мы внимательно проанализировали документ, содержащийся в нем, выводы рекомендаций.
----------------------------------------------------------------------------------------------------
Reference: А сейчас слово имеет представитель Соединенных Штатов Америки.
Prediction: А сейчас слово имеет представитель Соединенных Штатов Америки.
----------------------------------------------------------------------------------------------------
Reference: Обстоятельства изменились, и мы должны учитывать это.
Prediction: Обстоятельно изменились и мы должны учитывать это.
----------------------------------------------------------------------------------------------------
Reference: На этом принципе основывается и наша позиция по Фолклендским островам.
Prediction: На этом принципе основывается и наша позиция по Фолклендским островам.
```
The Google Colab version of [this script](https://colab.research.google.com/drive/1VlTrsc9d9wyzLPAWagpXLzoDLn2PRvZA?usp=sharing) is available too.
## Evaluation
This model was evaluated on the test subsets of [SberDevices Golos](https://huggingface.co/datasets/SberDevices/Golos), [Common Voice 6.0](https://huggingface.co/datasets/common_voice) (Russian part), and [Sova RuDevices](https://huggingface.co/datasets/bond005/sova_rudevices).
The evaluation script [wav2vec2_mbart50_ru_eval](https://www.kaggle.com/code/bond005/wav2vec2-mbart50-ru-eval) is available for checking and reproducibility.
## Citation
If you want to cite this model you can use this:
```bibtex
@misc{bondarenko2023-wav2vec2-mbart50-ru,
title={Wav2Vec2-mBART-50 for speech-to-text in Russian by Ivan Bondarenko},
author={Bondarenko, Ivan},
publisher={Hugging Face},
journal={Hugging Face Hub},
howpublished={\url{https://huggingface.co/bond005/wav2vec2-mbart50-ru}},
year={2023}
}
```
|
bert-base-german-dbmdz-cased | [
"pytorch",
"jax",
"bert",
"fill-mask",
"de",
"transformers",
"license:mit",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 1,814 | 2022-12-06T13:36:21Z | ---
license: creativeml-openrail-m
tags:
- text-to-image
- stable-diffusion
widget:
- text: "a photo of dpkbjwn and mnlrvr at a christmas market"
---
### Deepika and Manuel Simulator
classifiers:
"dpkbjwn" for Deepika
"mnlrvr" for Manuel
Example prompt: a photo of dpkbjwn and mnlrvr at a christmas market
Sample pictures:
|
bert-base-multilingual-cased | [
"pytorch",
"tf",
"jax",
"safetensors",
"bert",
"fill-mask",
"multilingual",
"af",
"sq",
"ar",
"an",
"hy",
"ast",
"az",
"ba",
"eu",
"bar",
"be",
"bn",
"inc",
"bs",
"br",
"bg",
"my",
"ca",
"ceb",
"ce",
"zh",
"cv",
"hr",
"cs",
"da",
"nl",
"en",
"et",
"fi",
"fr",
"gl",
"ka",
"de",
"el",
"gu",
"ht",
"he",
"hi",
"hu",
"is",
"io",
"id",
"ga",
"it",
"ja",
"jv",
"kn",
"kk",
"ky",
"ko",
"la",
"lv",
"lt",
"roa",
"nds",
"lm",
"mk",
"mg",
"ms",
"ml",
"mr",
"mn",
"min",
"ne",
"new",
"nb",
"nn",
"oc",
"fa",
"pms",
"pl",
"pt",
"pa",
"ro",
"ru",
"sco",
"sr",
"scn",
"sk",
"sl",
"aze",
"es",
"su",
"sw",
"sv",
"tl",
"tg",
"th",
"ta",
"tt",
"te",
"tr",
"uk",
"ud",
"uz",
"vi",
"vo",
"war",
"cy",
"fry",
"pnb",
"yo",
"dataset:wikipedia",
"arxiv:1810.04805",
"transformers",
"license:apache-2.0",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 4,749,504 | 2022-12-06T13:37:38Z | ---
license: mit
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
model-index:
- name: finetuned-im-rahmen-der-rechtlichen-und-ethischen-bestimmungen-arbeiten
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. -->
# finetuned-im-rahmen-der-rechtlichen-und-ethischen-bestimmungen-arbeiten
This model is a fine-tuned version of [bert-base-german-cased](https://huggingface.co/bert-base-german-cased) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2919
- Accuracy: 0.8970
- F1: 0.8843
## 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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.2946 | 1.0 | 1365 | 0.2791 | 0.8992 | 0.8829 |
| 0.2204 | 2.0 | 2730 | 0.2919 | 0.8970 | 0.8843 |
### Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1+cu113
- Datasets 2.3.2
- Tokenizers 0.13.2
|
bert-base-multilingual-uncased | [
"pytorch",
"tf",
"jax",
"safetensors",
"bert",
"fill-mask",
"multilingual",
"af",
"sq",
"ar",
"an",
"hy",
"ast",
"az",
"ba",
"eu",
"bar",
"be",
"bn",
"inc",
"bs",
"br",
"bg",
"my",
"ca",
"ceb",
"ce",
"zh",
"cv",
"hr",
"cs",
"da",
"nl",
"en",
"et",
"fi",
"fr",
"gl",
"ka",
"de",
"el",
"gu",
"ht",
"he",
"hi",
"hu",
"is",
"io",
"id",
"ga",
"it",
"ja",
"jv",
"kn",
"kk",
"ky",
"ko",
"la",
"lv",
"lt",
"roa",
"nds",
"lm",
"mk",
"mg",
"ms",
"ml",
"mr",
"min",
"ne",
"new",
"nb",
"nn",
"oc",
"fa",
"pms",
"pl",
"pt",
"pa",
"ro",
"ru",
"sco",
"sr",
"scn",
"sk",
"sl",
"aze",
"es",
"su",
"sw",
"sv",
"tl",
"tg",
"ta",
"tt",
"te",
"tr",
"uk",
"ud",
"uz",
"vi",
"vo",
"war",
"cy",
"fry",
"pnb",
"yo",
"dataset:wikipedia",
"arxiv:1810.04805",
"transformers",
"license:apache-2.0",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 328,585 | 2022-12-06T13:42:40Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- imdb
metrics:
- accuracy
- f1
model-index:
- name: finetuning-sentiment-model-Test
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: imdb
type: imdb
config: plain_text
split: train
args: plain_text
metrics:
- name: Accuracy
type: accuracy
value: 0.902
- name: F1
type: f1
value: 0.9037328094302554
---
<!-- 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. -->
# finetuning-sentiment-model-Test
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the imdb dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2793
- Accuracy: 0.902
- F1: 0.9037
## 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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
### Training results
### Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2
|
bert-base-uncased | [
"pytorch",
"tf",
"jax",
"rust",
"safetensors",
"bert",
"fill-mask",
"en",
"dataset:bookcorpus",
"dataset:wikipedia",
"arxiv:1810.04805",
"transformers",
"exbert",
"license:apache-2.0",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 59,663,489 | 2022-12-06T13:42:46Z | ---
language:
- en
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- top_v2
model-index:
- name: t5-base-pointer-top_v2
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. -->
# t5-base-pointer-top_v2
This model is a fine-tuned version of [google/mt5-base](https://huggingface.co/google/mt5-base) on the top_v2 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0256
- Exact Match: 0.8517
## 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: 0.001
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 128
- total_train_batch_size: 512
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 3000
### Training results
| Training Loss | Epoch | Step | Validation Loss | Exact Match |
|:-------------:|:-----:|:----:|:---------------:|:-----------:|
| 1.4545 | 0.82 | 200 | 0.2542 | 0.1294 |
| 0.1878 | 1.65 | 400 | 0.0668 | 0.2128 |
| 0.0796 | 2.47 | 600 | 0.0466 | 0.2276 |
| 0.0536 | 3.29 | 800 | 0.0356 | 0.2309 |
| 0.0424 | 4.12 | 1000 | 0.0317 | 0.2328 |
| 0.0356 | 4.94 | 1200 | 0.0295 | 0.2340 |
| 0.0306 | 5.76 | 1400 | 0.0288 | 0.2357 |
| 0.0277 | 6.58 | 1600 | 0.0271 | 0.2351 |
| 0.0243 | 7.41 | 1800 | 0.0272 | 0.2351 |
| 0.0225 | 8.23 | 2000 | 0.0272 | 0.2353 |
| 0.0206 | 9.05 | 2200 | 0.0267 | 0.2368 |
| 0.0187 | 9.88 | 2400 | 0.0260 | 0.2367 |
| 0.0173 | 10.7 | 2600 | 0.0256 | 0.2383 |
| 0.0161 | 11.52 | 2800 | 0.0260 | 0.2383 |
| 0.0153 | 12.35 | 3000 | 0.0257 | 0.2377 |
### Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0+cu117
- Datasets 2.7.1
- Tokenizers 0.13.2
|
bert-large-uncased-whole-word-masking | [
"pytorch",
"tf",
"jax",
"safetensors",
"bert",
"fill-mask",
"en",
"dataset:bookcorpus",
"dataset:wikipedia",
"arxiv:1810.04805",
"transformers",
"license:apache-2.0",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 76,685 | 2022-12-06T13:49:42Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- squad
model-index:
- name: Bert-test-model
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. -->
# Bert-test-model
This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the squad dataset.
It achieves the following results on the evaluation set:
- Loss: 1.3708
## 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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| No log | 1.0 | 250 | 1.7369 |
| 2.2639 | 2.0 | 500 | 1.3940 |
| 2.2639 | 3.0 | 750 | 1.3708 |
### Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2
|
bert-large-uncased | [
"pytorch",
"tf",
"jax",
"safetensors",
"bert",
"fill-mask",
"en",
"dataset:bookcorpus",
"dataset:wikipedia",
"arxiv:1810.04805",
"transformers",
"license:apache-2.0",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 1,058,496 | 2022-12-06T13:51:46Z | ---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: idrak_wav2vec_timit_subsample
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. -->
# idrak_wav2vec_timit_subsample
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
## 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: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 100
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu113
- Datasets 1.18.3
- Tokenizers 0.10.3
|
camembert-base | [
"pytorch",
"tf",
"safetensors",
"camembert",
"fill-mask",
"fr",
"dataset:oscar",
"arxiv:1911.03894",
"transformers",
"license:mit",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"CamembertForMaskedLM"
],
"model_type": "camembert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 1,440,898 | 2022-12-06T13:59:52Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- squad
model-index:
- name: distilbert-base-uncased-finetuned-squad
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. -->
# distilbert-base-uncased-finetuned-squad
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the squad dataset.
## 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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2
|
distilbert-base-cased-distilled-squad | [
"pytorch",
"tf",
"rust",
"safetensors",
"openvino",
"distilbert",
"question-answering",
"en",
"dataset:squad",
"arxiv:1910.01108",
"arxiv:1910.09700",
"transformers",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"has_space"
]
| question-answering | {
"architectures": [
"DistilBertForQuestionAnswering"
],
"model_type": "distilbert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 257,745 | null | Access to model syndikatet/kaia is restricted and you are not in the authorized list. Visit https://huggingface.co/syndikatet/kaia to ask for access. |
distilbert-base-multilingual-cased | [
"pytorch",
"tf",
"onnx",
"safetensors",
"distilbert",
"fill-mask",
"multilingual",
"af",
"sq",
"ar",
"an",
"hy",
"ast",
"az",
"ba",
"eu",
"bar",
"be",
"bn",
"inc",
"bs",
"br",
"bg",
"my",
"ca",
"ceb",
"ce",
"zh",
"cv",
"hr",
"cs",
"da",
"nl",
"en",
"et",
"fi",
"fr",
"gl",
"ka",
"de",
"el",
"gu",
"ht",
"he",
"hi",
"hu",
"is",
"io",
"id",
"ga",
"it",
"ja",
"jv",
"kn",
"kk",
"ky",
"ko",
"la",
"lv",
"lt",
"roa",
"nds",
"lm",
"mk",
"mg",
"ms",
"ml",
"mr",
"mn",
"min",
"ne",
"new",
"nb",
"nn",
"oc",
"fa",
"pms",
"pl",
"pt",
"pa",
"ro",
"ru",
"sco",
"sr",
"scn",
"sk",
"sl",
"aze",
"es",
"su",
"sw",
"sv",
"tl",
"tg",
"th",
"ta",
"tt",
"te",
"tr",
"uk",
"ud",
"uz",
"vi",
"vo",
"war",
"cy",
"fry",
"pnb",
"yo",
"dataset:wikipedia",
"arxiv:1910.01108",
"arxiv:1910.09700",
"transformers",
"license:apache-2.0",
"autotrain_compatible",
"has_space"
]
| fill-mask | {
"architectures": [
"DistilBertForMaskedLM"
],
"model_type": "distilbert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 8,339,633 | 2022-12-06T14:05:41Z | ---
language:
- en
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- cstop_artificial
model-index:
- name: t5-base-pointer-cstop_artificial
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. -->
# t5-base-pointer-cstop_artificial
This model is a fine-tuned version of [google/mt5-base](https://huggingface.co/google/mt5-base) on the cstop_artificial dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0776
- Exact Match: 0.7746
## 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: 0.001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 64
- total_train_batch_size: 512
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 3000
### Training results
| Training Loss | Epoch | Step | Validation Loss | Exact Match |
|:-------------:|:------:|:----:|:---------------:|:-----------:|
| 1.7482 | 28.5 | 200 | 0.2505 | 0.1020 |
| 0.1366 | 57.13 | 400 | 0.0776 | 0.3238 |
| 0.0275 | 85.63 | 600 | 0.0881 | 0.3381 |
| 0.0114 | 114.25 | 800 | 0.0990 | 0.3399 |
| 0.0064 | 142.75 | 1000 | 0.1120 | 0.3417 |
| 0.0045 | 171.38 | 1200 | 0.1081 | 0.3435 |
| 0.0036 | 199.88 | 1400 | 0.1230 | 0.3435 |
| 0.0025 | 228.5 | 1600 | 0.1211 | 0.3399 |
| 0.002 | 257.13 | 1800 | 0.1367 | 0.3399 |
| 0.0016 | 285.63 | 2000 | 0.1324 | 0.3435 |
| 0.0013 | 314.25 | 2200 | 0.1340 | 0.3470 |
| 0.001 | 342.75 | 2400 | 0.1374 | 0.3435 |
| 0.0009 | 371.38 | 2600 | 0.1384 | 0.3417 |
| 0.0007 | 399.88 | 2800 | 0.1422 | 0.3435 |
| 0.0006 | 428.5 | 3000 | 0.1452 | 0.3417 |
### Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0+cu117
- Datasets 2.7.1
- Tokenizers 0.13.2
|
Akshay-Vs/AI | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | Access to model rybread01/email-ds-bert is restricted and you are not in the authorized list. Visit https://huggingface.co/rybread01/email-ds-bert to ask for access. |
ASCCCCCCCC/distilbert-base-chinese-amazon_zh_20000 | [
"pytorch",
"tensorboard",
"bert",
"text-classification",
"transformers",
"generated_from_trainer"
]
| text-classification | {
"architectures": [
"BertForSequenceClassification"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 93 | 2022-12-06T19:14:49Z | ---
license: apache-2.0
tags:
- text-classification
- generated_from_trainer
datasets:
- paws-x
metrics:
- accuracy
model-index:
- name: paws_x_m_bert_only_ko
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: paws-x
type: paws-x
config: ko
split: train
args: ko
metrics:
- name: Accuracy
type: accuracy
value: 0.8215
---
<!-- 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. -->
# paws_x_m_bert_only_ko
This model is a fine-tuned version of [bert-base-multilingual-cased](https://huggingface.co/bert-base-multilingual-cased) on the paws-x dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7649
- Accuracy: 0.8215
## 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: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- num_epochs: 10
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 0.5446 | 1.0 | 386 | 0.4837 | 0.768 |
| 0.3443 | 2.0 | 772 | 0.4530 | 0.8125 |
| 0.258 | 3.0 | 1158 | 0.4496 | 0.8145 |
| 0.2023 | 4.0 | 1544 | 0.4944 | 0.81 |
| 0.1581 | 5.0 | 1930 | 0.5040 | 0.814 |
| 0.1263 | 6.0 | 2316 | 0.5937 | 0.8145 |
| 0.1041 | 7.0 | 2702 | 0.6578 | 0.8115 |
| 0.0828 | 8.0 | 3088 | 0.6841 | 0.8215 |
| 0.0697 | 9.0 | 3474 | 0.7239 | 0.82 |
| 0.0596 | 10.0 | 3860 | 0.7649 | 0.8215 |
### Framework versions
- Transformers 4.24.0
- Pytorch 1.13.0
- Datasets 2.6.1
- Tokenizers 0.13.1
|
ASCCCCCCCC/distilbert-base-multilingual-cased-amazon_zh_20000 | [
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"transformers",
"generated_from_trainer",
"license:apache-2.0"
]
| text-classification | {
"architectures": [
"DistilBertForSequenceClassification"
],
"model_type": "distilbert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 39 | null | ---
language:
- mn
license: apache-2.0
tags:
- whisper-event
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
- google/fleurs
- bayartsogt/ulaanbal-v0
metrics:
- wer
model-index:
- name: whisper-medium-mn-5
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 11.0
type: mozilla-foundation/common_voice_11_0
config: mn
split: test
metrics:
- name: Wer
type: wer
value: 24.7268953462967
---
<!-- 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. -->
# whisper-medium-mn-4
This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the Common Voice 11.0 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3396
- Wer: 24.7268
- Cer: 8.6712
## 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: 1e-05
- train_batch_size: 32
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 12000
- mixed_precision_training: Native AMP
### Training results
```
{'eval_loss': 0.3396347761154175, 'eval_wer': 24.7268953462967, 'eval_cer': 8.671234994074913, 'eval_runtime': 2202.1539, 'eval_samples_per_second': 0.856, 'eval_steps_per_second': 0.027, 'epoch': 7
.3}
```
### Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu117
- Datasets 2.7.1.dev0
- Tokenizers 0.13.2
|
Aeroxas/Botroxas-small | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
metrics:
- type: mean_reward
value: 257.25 +/- 21.61
name: mean_reward
verified: false
---
# **PPO** Agent playing **LunarLander-v2**
This is a trained model of a **PPO** agent playing **LunarLander-v2**
using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3).
## Usage (with Stable-baselines3)
TODO: Add your code
```python
from stable_baselines3 import ...
from huggingface_sb3 import load_from_hub
...
```
|
AetherIT/DialoGPT-small-Hal | [
"conversational"
]
| conversational | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: wav2vec2-large-teacher-base-student-en-asr-timit
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. -->
# wav2vec2-large-teacher-base-student-en-asr-timit
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 73.5882
- Wer: 0.3422
## 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: 0.0001
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 30
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 920.6083 | 3.17 | 200 | 1256.0675 | 1.0 |
| 660.5993 | 6.35 | 400 | 717.6098 | 0.9238 |
| 336.5288 | 9.52 | 600 | 202.0025 | 0.5306 |
| 131.3178 | 12.7 | 800 | 108.0701 | 0.4335 |
| 73.4232 | 15.87 | 1000 | 90.2797 | 0.3728 |
| 54.9439 | 19.05 | 1200 | 76.9043 | 0.3636 |
| 44.6595 | 22.22 | 1400 | 79.2443 | 0.3550 |
| 38.6381 | 25.4 | 1600 | 73.6277 | 0.3493 |
| 35.074 | 28.57 | 1800 | 73.5882 | 0.3422 |
### Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0+cu116
- Datasets 1.18.3
- Tokenizers 0.13.2
|
AethiQs-Max/aethiqs-base_bertje-data_rotterdam-epochs_10 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 9 | null | ---
license: apache-2.0
tags:
- translation
- generated_from_trainer
metrics:
- bleu
model-index:
- name: krirk-finetuned-Helsinki-NLP_opus-mt-ar-en
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. -->
# krirk-finetuned-Helsinki-NLP_opus-mt-ar-en
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-ar-en](https://huggingface.co/Helsinki-NLP/opus-mt-ar-en) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.3665
- Bleu: 35.0219
## 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: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 1.4469 | 1.0 | 32 | 1.3744 | 34.9616 |
| 1.2938 | 2.0 | 64 | 1.3674 | 34.9145 |
| 1.2582 | 3.0 | 96 | 1.3665 | 35.0219 |
### Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2
|
AethiQs-Max/aethiqs-base_bertje-data_rotterdam-epochs_30-epoch_30 | [
"pytorch",
"bert",
"fill-mask",
"transformers",
"autotrain_compatible"
]
| fill-mask | {
"architectures": [
"BertForMaskedLM"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 8 | null | ---
language: en
license: apache-2.0
library_name: diffusers
tags: []
datasets: huggan/smithsonian_butterflies_subset
metrics: []
---
<!-- This model card has been generated automatically according to the information the training script had access to. You
should probably proofread and complete it, then remove this comment. -->
# ddpm-butterflies-128
## Model description
This diffusion model is trained with the [🤗 Diffusers](https://github.com/huggingface/diffusers) library
on the `huggan/smithsonian_butterflies_subset` dataset.
## Intended uses & limitations
#### How to use
```python
from diffusers import DDPMPipeline
model_id = "hjjeon/ddpm-butterflies-128"
# load model and scheduler
pipeline = DDPMPipeline.from_pretrained(model_id)
# run pipeline in inference
image = pipeline()["sample"]
# save image
image[0].save("butterfly.png")
```
#### Limitations and bias
[TODO: provide examples of latent issues and potential remediations]
## Training data
[TODO: describe the data used to train the model]
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0001
- train_batch_size: 16
- eval_batch_size: 16
- gradient_accumulation_steps: 1
- optimizer: AdamW with betas=(None, None), weight_decay=None and epsilon=None
- lr_scheduler: None
- lr_warmup_steps: 500
- ema_inv_gamma: None
- ema_inv_gamma: None
- ema_inv_gamma: None
- mixed_precision: fp16
### Training results
📈 [TensorBoard logs](https://huggingface.co/hjjeon/ddpm-butterflies-128/tensorboard?#scalars)
|
Ahmedahmed/Wewe | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language:
- vi
---
## Introduction
This model was initialized from [vinai/bartpho-word-base](https://huggingface.co/vinai/bartpho-word-base) and converted to [Allenai's Longformer Encoder-Decoder (LED)](https://github.com/allenai/longformer#longformer) based on [Longformer: The Long-Document Transformer](https://arxiv.org/pdf/2004.05150.pdf).
To be able to process 16K tokens, *bartpho-word-base*'s position embedding matrix was simply copied 16 times.
This model is especially interesting for long-range summarization and question answering.
## Fine-tuning for down-stream task
[This notebook](https://colab.research.google.com/drive/12LjJazBl7Gam0XBPy_y0CTOJZeZ34c2v?usp=sharing) shows how led model can effectively be fine-tuned on a downstream task. |
Akash7897/test-clm | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | Access to model DocPIXL/DOCPICL is restricted and you are not in the authorized list. Visit https://huggingface.co/DocPIXL/DOCPICL to ask for access. |
Akashpb13/Hausa_xlsr | [
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"ha",
"dataset:mozilla-foundation/common_voice_8_0",
"transformers",
"generated_from_trainer",
"hf-asr-leaderboard",
"model_for_talk",
"mozilla-foundation/common_voice_8_0",
"robust-speech-event",
"license:apache-2.0",
"model-index",
"has_space"
]
| automatic-speech-recognition | {
"architectures": [
"Wav2Vec2ForCTC"
],
"model_type": "wav2vec2",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 31 | null | ---
tags:
- mteb
model-index:
- name: e5-small
results:
- task:
type: Classification
dataset:
type: mteb/amazon_counterfactual
name: MTEB AmazonCounterfactualClassification (en)
config: en
split: test
revision: e8379541af4e31359cca9fbcf4b00f2671dba205
metrics:
- type: accuracy
value: 76.22388059701493
- type: ap
value: 40.27466219523129
- type: f1
value: 70.60533006025108
- task:
type: Classification
dataset:
type: mteb/amazon_polarity
name: MTEB AmazonPolarityClassification
config: default
split: test
revision: e2d317d38cd51312af73b3d32a06d1a08b442046
metrics:
- type: accuracy
value: 87.525775
- type: ap
value: 83.51063993897611
- type: f1
value: 87.49342736805572
- task:
type: Classification
dataset:
type: mteb/amazon_reviews_multi
name: MTEB AmazonReviewsClassification (en)
config: en
split: test
revision: 1399c76144fd37290681b995c656ef9b2e06e26d
metrics:
- type: accuracy
value: 42.611999999999995
- type: f1
value: 42.05088045932892
- task:
type: Retrieval
dataset:
type: arguana
name: MTEB ArguAna
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 23.826
- type: map_at_10
value: 38.269
- type: map_at_100
value: 39.322
- type: map_at_1000
value: 39.344
- type: map_at_3
value: 33.428000000000004
- type: map_at_5
value: 36.063
- type: mrr_at_1
value: 24.253
- type: mrr_at_10
value: 38.425
- type: mrr_at_100
value: 39.478
- type: mrr_at_1000
value: 39.5
- type: mrr_at_3
value: 33.606
- type: mrr_at_5
value: 36.195
- type: ndcg_at_1
value: 23.826
- type: ndcg_at_10
value: 46.693
- type: ndcg_at_100
value: 51.469
- type: ndcg_at_1000
value: 52.002
- type: ndcg_at_3
value: 36.603
- type: ndcg_at_5
value: 41.365
- type: precision_at_1
value: 23.826
- type: precision_at_10
value: 7.383000000000001
- type: precision_at_100
value: 0.9530000000000001
- type: precision_at_1000
value: 0.099
- type: precision_at_3
value: 15.268
- type: precision_at_5
value: 11.479000000000001
- type: recall_at_1
value: 23.826
- type: recall_at_10
value: 73.82600000000001
- type: recall_at_100
value: 95.306
- type: recall_at_1000
value: 99.431
- type: recall_at_3
value: 45.804
- type: recall_at_5
value: 57.397
- task:
type: Clustering
dataset:
type: mteb/arxiv-clustering-p2p
name: MTEB ArxivClusteringP2P
config: default
split: test
revision: a122ad7f3f0291bf49cc6f4d32aa80929df69d5d
metrics:
- type: v_measure
value: 44.13995374767436
- task:
type: Clustering
dataset:
type: mteb/arxiv-clustering-s2s
name: MTEB ArxivClusteringS2S
config: default
split: test
revision: f910caf1a6075f7329cdf8c1a6135696f37dbd53
metrics:
- type: v_measure
value: 37.13950072624313
- task:
type: Reranking
dataset:
type: mteb/askubuntudupquestions-reranking
name: MTEB AskUbuntuDupQuestions
config: default
split: test
revision: 2000358ca161889fa9c082cb41daa8dcfb161a54
metrics:
- type: map
value: 59.35843292105327
- type: mrr
value: 73.72312359846987
- task:
type: STS
dataset:
type: mteb/biosses-sts
name: MTEB BIOSSES
config: default
split: test
revision: d3fb88f8f02e40887cd149695127462bbcf29b4a
metrics:
- type: cos_sim_pearson
value: 84.55140418324174
- type: cos_sim_spearman
value: 84.21637675860022
- type: euclidean_pearson
value: 81.26069614610006
- type: euclidean_spearman
value: 83.25069210421785
- type: manhattan_pearson
value: 80.17441422581014
- type: manhattan_spearman
value: 81.87596198487877
- task:
type: Classification
dataset:
type: mteb/banking77
name: MTEB Banking77Classification
config: default
split: test
revision: 0fd18e25b25c072e09e0d92ab615fda904d66300
metrics:
- type: accuracy
value: 81.87337662337661
- type: f1
value: 81.76647866926402
- task:
type: Clustering
dataset:
type: mteb/biorxiv-clustering-p2p
name: MTEB BiorxivClusteringP2P
config: default
split: test
revision: 65b79d1d13f80053f67aca9498d9402c2d9f1f40
metrics:
- type: v_measure
value: 35.80600542614507
- task:
type: Clustering
dataset:
type: mteb/biorxiv-clustering-s2s
name: MTEB BiorxivClusteringS2S
config: default
split: test
revision: 258694dd0231531bc1fd9de6ceb52a0853c6d908
metrics:
- type: v_measure
value: 31.86321613256603
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackAndroidRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 32.054
- type: map_at_10
value: 40.699999999999996
- type: map_at_100
value: 41.818
- type: map_at_1000
value: 41.959999999999994
- type: map_at_3
value: 37.742
- type: map_at_5
value: 39.427
- type: mrr_at_1
value: 38.769999999999996
- type: mrr_at_10
value: 46.150000000000006
- type: mrr_at_100
value: 46.865
- type: mrr_at_1000
value: 46.925
- type: mrr_at_3
value: 43.705
- type: mrr_at_5
value: 45.214999999999996
- type: ndcg_at_1
value: 38.769999999999996
- type: ndcg_at_10
value: 45.778
- type: ndcg_at_100
value: 50.38
- type: ndcg_at_1000
value: 52.922999999999995
- type: ndcg_at_3
value: 41.597
- type: ndcg_at_5
value: 43.631
- type: precision_at_1
value: 38.769999999999996
- type: precision_at_10
value: 8.269
- type: precision_at_100
value: 1.278
- type: precision_at_1000
value: 0.178
- type: precision_at_3
value: 19.266
- type: precision_at_5
value: 13.705
- type: recall_at_1
value: 32.054
- type: recall_at_10
value: 54.947
- type: recall_at_100
value: 74.79599999999999
- type: recall_at_1000
value: 91.40899999999999
- type: recall_at_3
value: 42.431000000000004
- type: recall_at_5
value: 48.519
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackEnglishRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 29.035
- type: map_at_10
value: 38.007000000000005
- type: map_at_100
value: 39.125
- type: map_at_1000
value: 39.251999999999995
- type: map_at_3
value: 35.77
- type: map_at_5
value: 37.057
- type: mrr_at_1
value: 36.497
- type: mrr_at_10
value: 44.077
- type: mrr_at_100
value: 44.743
- type: mrr_at_1000
value: 44.79
- type: mrr_at_3
value: 42.123
- type: mrr_at_5
value: 43.308
- type: ndcg_at_1
value: 36.497
- type: ndcg_at_10
value: 42.986000000000004
- type: ndcg_at_100
value: 47.323
- type: ndcg_at_1000
value: 49.624
- type: ndcg_at_3
value: 39.805
- type: ndcg_at_5
value: 41.286
- type: precision_at_1
value: 36.497
- type: precision_at_10
value: 7.8340000000000005
- type: precision_at_100
value: 1.269
- type: precision_at_1000
value: 0.178
- type: precision_at_3
value: 19.023
- type: precision_at_5
value: 13.248
- type: recall_at_1
value: 29.035
- type: recall_at_10
value: 51.06
- type: recall_at_100
value: 69.64099999999999
- type: recall_at_1000
value: 84.49
- type: recall_at_3
value: 41.333999999999996
- type: recall_at_5
value: 45.663
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackGamingRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 37.239
- type: map_at_10
value: 47.873
- type: map_at_100
value: 48.842999999999996
- type: map_at_1000
value: 48.913000000000004
- type: map_at_3
value: 45.050000000000004
- type: map_at_5
value: 46.498
- type: mrr_at_1
value: 42.508
- type: mrr_at_10
value: 51.44
- type: mrr_at_100
value: 52.087
- type: mrr_at_1000
value: 52.129999999999995
- type: mrr_at_3
value: 49.164
- type: mrr_at_5
value: 50.343
- type: ndcg_at_1
value: 42.508
- type: ndcg_at_10
value: 53.31399999999999
- type: ndcg_at_100
value: 57.245000000000005
- type: ndcg_at_1000
value: 58.794000000000004
- type: ndcg_at_3
value: 48.295
- type: ndcg_at_5
value: 50.415
- type: precision_at_1
value: 42.508
- type: precision_at_10
value: 8.458
- type: precision_at_100
value: 1.133
- type: precision_at_1000
value: 0.132
- type: precision_at_3
value: 21.191
- type: precision_at_5
value: 14.307
- type: recall_at_1
value: 37.239
- type: recall_at_10
value: 65.99000000000001
- type: recall_at_100
value: 82.99499999999999
- type: recall_at_1000
value: 94.128
- type: recall_at_3
value: 52.382
- type: recall_at_5
value: 57.648999999999994
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackGisRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 23.039
- type: map_at_10
value: 29.694
- type: map_at_100
value: 30.587999999999997
- type: map_at_1000
value: 30.692999999999998
- type: map_at_3
value: 27.708
- type: map_at_5
value: 28.774
- type: mrr_at_1
value: 24.633
- type: mrr_at_10
value: 31.478
- type: mrr_at_100
value: 32.299
- type: mrr_at_1000
value: 32.381
- type: mrr_at_3
value: 29.435
- type: mrr_at_5
value: 30.446
- type: ndcg_at_1
value: 24.633
- type: ndcg_at_10
value: 33.697
- type: ndcg_at_100
value: 38.080000000000005
- type: ndcg_at_1000
value: 40.812
- type: ndcg_at_3
value: 29.654000000000003
- type: ndcg_at_5
value: 31.474000000000004
- type: precision_at_1
value: 24.633
- type: precision_at_10
value: 5.0729999999999995
- type: precision_at_100
value: 0.753
- type: precision_at_1000
value: 0.10300000000000001
- type: precision_at_3
value: 12.279
- type: precision_at_5
value: 8.452
- type: recall_at_1
value: 23.039
- type: recall_at_10
value: 44.275999999999996
- type: recall_at_100
value: 64.4
- type: recall_at_1000
value: 85.135
- type: recall_at_3
value: 33.394
- type: recall_at_5
value: 37.687
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackMathematicaRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 13.594999999999999
- type: map_at_10
value: 19.933999999999997
- type: map_at_100
value: 20.966
- type: map_at_1000
value: 21.087
- type: map_at_3
value: 17.749000000000002
- type: map_at_5
value: 19.156000000000002
- type: mrr_at_1
value: 17.662
- type: mrr_at_10
value: 24.407
- type: mrr_at_100
value: 25.385
- type: mrr_at_1000
value: 25.465
- type: mrr_at_3
value: 22.056
- type: mrr_at_5
value: 23.630000000000003
- type: ndcg_at_1
value: 17.662
- type: ndcg_at_10
value: 24.391
- type: ndcg_at_100
value: 29.681
- type: ndcg_at_1000
value: 32.923
- type: ndcg_at_3
value: 20.271
- type: ndcg_at_5
value: 22.621
- type: precision_at_1
value: 17.662
- type: precision_at_10
value: 4.44
- type: precision_at_100
value: 0.8200000000000001
- type: precision_at_1000
value: 0.125
- type: precision_at_3
value: 9.577
- type: precision_at_5
value: 7.313
- type: recall_at_1
value: 13.594999999999999
- type: recall_at_10
value: 33.976
- type: recall_at_100
value: 57.43000000000001
- type: recall_at_1000
value: 80.958
- type: recall_at_3
value: 22.897000000000002
- type: recall_at_5
value: 28.714000000000002
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackPhysicsRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 26.683
- type: map_at_10
value: 35.068
- type: map_at_100
value: 36.311
- type: map_at_1000
value: 36.436
- type: map_at_3
value: 32.371
- type: map_at_5
value: 33.761
- type: mrr_at_1
value: 32.435
- type: mrr_at_10
value: 40.721000000000004
- type: mrr_at_100
value: 41.535
- type: mrr_at_1000
value: 41.593
- type: mrr_at_3
value: 38.401999999999994
- type: mrr_at_5
value: 39.567
- type: ndcg_at_1
value: 32.435
- type: ndcg_at_10
value: 40.538000000000004
- type: ndcg_at_100
value: 45.963
- type: ndcg_at_1000
value: 48.400999999999996
- type: ndcg_at_3
value: 36.048
- type: ndcg_at_5
value: 37.899
- type: precision_at_1
value: 32.435
- type: precision_at_10
value: 7.1129999999999995
- type: precision_at_100
value: 1.162
- type: precision_at_1000
value: 0.156
- type: precision_at_3
value: 16.683
- type: precision_at_5
value: 11.684
- type: recall_at_1
value: 26.683
- type: recall_at_10
value: 51.517
- type: recall_at_100
value: 74.553
- type: recall_at_1000
value: 90.649
- type: recall_at_3
value: 38.495000000000005
- type: recall_at_5
value: 43.495
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackProgrammersRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 24.186
- type: map_at_10
value: 31.972
- type: map_at_100
value: 33.117000000000004
- type: map_at_1000
value: 33.243
- type: map_at_3
value: 29.423
- type: map_at_5
value: 30.847
- type: mrr_at_1
value: 29.794999999999998
- type: mrr_at_10
value: 36.767
- type: mrr_at_100
value: 37.645
- type: mrr_at_1000
value: 37.716
- type: mrr_at_3
value: 34.513
- type: mrr_at_5
value: 35.791000000000004
- type: ndcg_at_1
value: 29.794999999999998
- type: ndcg_at_10
value: 36.786
- type: ndcg_at_100
value: 41.94
- type: ndcg_at_1000
value: 44.830999999999996
- type: ndcg_at_3
value: 32.504
- type: ndcg_at_5
value: 34.404
- type: precision_at_1
value: 29.794999999999998
- type: precision_at_10
value: 6.518
- type: precision_at_100
value: 1.0659999999999998
- type: precision_at_1000
value: 0.149
- type: precision_at_3
value: 15.296999999999999
- type: precision_at_5
value: 10.731
- type: recall_at_1
value: 24.186
- type: recall_at_10
value: 46.617
- type: recall_at_100
value: 68.75
- type: recall_at_1000
value: 88.864
- type: recall_at_3
value: 34.199
- type: recall_at_5
value: 39.462
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 24.22083333333333
- type: map_at_10
value: 31.606666666666662
- type: map_at_100
value: 32.6195
- type: map_at_1000
value: 32.739999999999995
- type: map_at_3
value: 29.37825
- type: map_at_5
value: 30.596083333333336
- type: mrr_at_1
value: 28.607916666666668
- type: mrr_at_10
value: 35.54591666666666
- type: mrr_at_100
value: 36.33683333333333
- type: mrr_at_1000
value: 36.40624999999999
- type: mrr_at_3
value: 33.526250000000005
- type: mrr_at_5
value: 34.6605
- type: ndcg_at_1
value: 28.607916666666668
- type: ndcg_at_10
value: 36.07966666666667
- type: ndcg_at_100
value: 40.73308333333333
- type: ndcg_at_1000
value: 43.40666666666666
- type: ndcg_at_3
value: 32.23525
- type: ndcg_at_5
value: 33.97083333333333
- type: precision_at_1
value: 28.607916666666668
- type: precision_at_10
value: 6.120333333333335
- type: precision_at_100
value: 0.9921666666666668
- type: precision_at_1000
value: 0.14091666666666666
- type: precision_at_3
value: 14.54975
- type: precision_at_5
value: 10.153166666666667
- type: recall_at_1
value: 24.22083333333333
- type: recall_at_10
value: 45.49183333333334
- type: recall_at_100
value: 66.28133333333332
- type: recall_at_1000
value: 85.16541666666667
- type: recall_at_3
value: 34.6485
- type: recall_at_5
value: 39.229749999999996
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackStatsRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 21.842
- type: map_at_10
value: 27.573999999999998
- type: map_at_100
value: 28.410999999999998
- type: map_at_1000
value: 28.502
- type: map_at_3
value: 25.921
- type: map_at_5
value: 26.888
- type: mrr_at_1
value: 24.08
- type: mrr_at_10
value: 29.915999999999997
- type: mrr_at_100
value: 30.669
- type: mrr_at_1000
value: 30.746000000000002
- type: mrr_at_3
value: 28.349000000000004
- type: mrr_at_5
value: 29.246
- type: ndcg_at_1
value: 24.08
- type: ndcg_at_10
value: 30.898999999999997
- type: ndcg_at_100
value: 35.272999999999996
- type: ndcg_at_1000
value: 37.679
- type: ndcg_at_3
value: 27.881
- type: ndcg_at_5
value: 29.432000000000002
- type: precision_at_1
value: 24.08
- type: precision_at_10
value: 4.678
- type: precision_at_100
value: 0.744
- type: precision_at_1000
value: 0.10300000000000001
- type: precision_at_3
value: 11.860999999999999
- type: precision_at_5
value: 8.16
- type: recall_at_1
value: 21.842
- type: recall_at_10
value: 38.66
- type: recall_at_100
value: 59.169000000000004
- type: recall_at_1000
value: 76.887
- type: recall_at_3
value: 30.532999999999998
- type: recall_at_5
value: 34.354
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackTexRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 17.145
- type: map_at_10
value: 22.729
- type: map_at_100
value: 23.574
- type: map_at_1000
value: 23.695
- type: map_at_3
value: 21.044
- type: map_at_5
value: 21.981
- type: mrr_at_1
value: 20.888
- type: mrr_at_10
value: 26.529000000000003
- type: mrr_at_100
value: 27.308
- type: mrr_at_1000
value: 27.389000000000003
- type: mrr_at_3
value: 24.868000000000002
- type: mrr_at_5
value: 25.825
- type: ndcg_at_1
value: 20.888
- type: ndcg_at_10
value: 26.457000000000004
- type: ndcg_at_100
value: 30.764000000000003
- type: ndcg_at_1000
value: 33.825
- type: ndcg_at_3
value: 23.483999999999998
- type: ndcg_at_5
value: 24.836
- type: precision_at_1
value: 20.888
- type: precision_at_10
value: 4.58
- type: precision_at_100
value: 0.784
- type: precision_at_1000
value: 0.121
- type: precision_at_3
value: 10.874
- type: precision_at_5
value: 7.639
- type: recall_at_1
value: 17.145
- type: recall_at_10
value: 33.938
- type: recall_at_100
value: 53.672
- type: recall_at_1000
value: 76.023
- type: recall_at_3
value: 25.363000000000003
- type: recall_at_5
value: 29.023
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackUnixRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 24.275
- type: map_at_10
value: 30.438
- type: map_at_100
value: 31.489
- type: map_at_1000
value: 31.601000000000003
- type: map_at_3
value: 28.647
- type: map_at_5
value: 29.660999999999998
- type: mrr_at_1
value: 28.077999999999996
- type: mrr_at_10
value: 34.098
- type: mrr_at_100
value: 35.025
- type: mrr_at_1000
value: 35.109
- type: mrr_at_3
value: 32.4
- type: mrr_at_5
value: 33.379999999999995
- type: ndcg_at_1
value: 28.077999999999996
- type: ndcg_at_10
value: 34.271
- type: ndcg_at_100
value: 39.352
- type: ndcg_at_1000
value: 42.199
- type: ndcg_at_3
value: 30.978
- type: ndcg_at_5
value: 32.498
- type: precision_at_1
value: 28.077999999999996
- type: precision_at_10
value: 5.345
- type: precision_at_100
value: 0.897
- type: precision_at_1000
value: 0.125
- type: precision_at_3
value: 13.526
- type: precision_at_5
value: 9.16
- type: recall_at_1
value: 24.275
- type: recall_at_10
value: 42.362
- type: recall_at_100
value: 64.461
- type: recall_at_1000
value: 84.981
- type: recall_at_3
value: 33.249
- type: recall_at_5
value: 37.214999999999996
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackWebmastersRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 22.358
- type: map_at_10
value: 30.062
- type: map_at_100
value: 31.189
- type: map_at_1000
value: 31.386999999999997
- type: map_at_3
value: 27.672
- type: map_at_5
value: 28.76
- type: mrr_at_1
value: 26.877000000000002
- type: mrr_at_10
value: 33.948
- type: mrr_at_100
value: 34.746
- type: mrr_at_1000
value: 34.816
- type: mrr_at_3
value: 31.884
- type: mrr_at_5
value: 33.001000000000005
- type: ndcg_at_1
value: 26.877000000000002
- type: ndcg_at_10
value: 34.977000000000004
- type: ndcg_at_100
value: 39.753
- type: ndcg_at_1000
value: 42.866
- type: ndcg_at_3
value: 30.956
- type: ndcg_at_5
value: 32.381
- type: precision_at_1
value: 26.877000000000002
- type: precision_at_10
value: 6.7
- type: precision_at_100
value: 1.287
- type: precision_at_1000
value: 0.215
- type: precision_at_3
value: 14.360999999999999
- type: precision_at_5
value: 10.119
- type: recall_at_1
value: 22.358
- type: recall_at_10
value: 44.183
- type: recall_at_100
value: 67.14
- type: recall_at_1000
value: 87.53999999999999
- type: recall_at_3
value: 32.79
- type: recall_at_5
value: 36.829
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackWordpressRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 19.198999999999998
- type: map_at_10
value: 25.229000000000003
- type: map_at_100
value: 26.003
- type: map_at_1000
value: 26.111
- type: map_at_3
value: 23.442
- type: map_at_5
value: 24.343
- type: mrr_at_1
value: 21.072
- type: mrr_at_10
value: 27.02
- type: mrr_at_100
value: 27.735
- type: mrr_at_1000
value: 27.815
- type: mrr_at_3
value: 25.416
- type: mrr_at_5
value: 26.173999999999996
- type: ndcg_at_1
value: 21.072
- type: ndcg_at_10
value: 28.862
- type: ndcg_at_100
value: 33.043
- type: ndcg_at_1000
value: 36.003
- type: ndcg_at_3
value: 25.35
- type: ndcg_at_5
value: 26.773000000000003
- type: precision_at_1
value: 21.072
- type: precision_at_10
value: 4.436
- type: precision_at_100
value: 0.713
- type: precision_at_1000
value: 0.106
- type: precision_at_3
value: 10.659
- type: precision_at_5
value: 7.32
- type: recall_at_1
value: 19.198999999999998
- type: recall_at_10
value: 38.376
- type: recall_at_100
value: 58.36900000000001
- type: recall_at_1000
value: 80.92099999999999
- type: recall_at_3
value: 28.715000000000003
- type: recall_at_5
value: 32.147
- task:
type: Retrieval
dataset:
type: climate-fever
name: MTEB ClimateFEVER
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 5.9319999999999995
- type: map_at_10
value: 10.483
- type: map_at_100
value: 11.97
- type: map_at_1000
value: 12.171999999999999
- type: map_at_3
value: 8.477
- type: map_at_5
value: 9.495000000000001
- type: mrr_at_1
value: 13.094
- type: mrr_at_10
value: 21.282
- type: mrr_at_100
value: 22.556
- type: mrr_at_1000
value: 22.628999999999998
- type: mrr_at_3
value: 18.218999999999998
- type: mrr_at_5
value: 19.900000000000002
- type: ndcg_at_1
value: 13.094
- type: ndcg_at_10
value: 15.811
- type: ndcg_at_100
value: 23.035
- type: ndcg_at_1000
value: 27.089999999999996
- type: ndcg_at_3
value: 11.905000000000001
- type: ndcg_at_5
value: 13.377
- type: precision_at_1
value: 13.094
- type: precision_at_10
value: 5.225
- type: precision_at_100
value: 1.2970000000000002
- type: precision_at_1000
value: 0.203
- type: precision_at_3
value: 8.86
- type: precision_at_5
value: 7.309
- type: recall_at_1
value: 5.9319999999999995
- type: recall_at_10
value: 20.305
- type: recall_at_100
value: 46.314
- type: recall_at_1000
value: 69.612
- type: recall_at_3
value: 11.21
- type: recall_at_5
value: 14.773
- task:
type: Retrieval
dataset:
type: dbpedia-entity
name: MTEB DBPedia
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 8.674
- type: map_at_10
value: 17.822
- type: map_at_100
value: 24.794
- type: map_at_1000
value: 26.214
- type: map_at_3
value: 12.690999999999999
- type: map_at_5
value: 15.033
- type: mrr_at_1
value: 61.75000000000001
- type: mrr_at_10
value: 71.58
- type: mrr_at_100
value: 71.923
- type: mrr_at_1000
value: 71.932
- type: mrr_at_3
value: 70.125
- type: mrr_at_5
value: 71.038
- type: ndcg_at_1
value: 51
- type: ndcg_at_10
value: 38.637
- type: ndcg_at_100
value: 42.398
- type: ndcg_at_1000
value: 48.962
- type: ndcg_at_3
value: 43.29
- type: ndcg_at_5
value: 40.763
- type: precision_at_1
value: 61.75000000000001
- type: precision_at_10
value: 30.125
- type: precision_at_100
value: 9.53
- type: precision_at_1000
value: 1.9619999999999997
- type: precision_at_3
value: 45.583
- type: precision_at_5
value: 38.95
- type: recall_at_1
value: 8.674
- type: recall_at_10
value: 23.122
- type: recall_at_100
value: 47.46
- type: recall_at_1000
value: 67.662
- type: recall_at_3
value: 13.946
- type: recall_at_5
value: 17.768
- task:
type: Classification
dataset:
type: mteb/emotion
name: MTEB EmotionClassification
config: default
split: test
revision: 4f58c6b202a23cf9a4da393831edf4f9183cad37
metrics:
- type: accuracy
value: 46.86000000000001
- type: f1
value: 41.343580452760776
- task:
type: Retrieval
dataset:
type: fever
name: MTEB FEVER
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 36.609
- type: map_at_10
value: 47.552
- type: map_at_100
value: 48.283
- type: map_at_1000
value: 48.321
- type: map_at_3
value: 44.869
- type: map_at_5
value: 46.509
- type: mrr_at_1
value: 39.214
- type: mrr_at_10
value: 50.434999999999995
- type: mrr_at_100
value: 51.122
- type: mrr_at_1000
value: 51.151
- type: mrr_at_3
value: 47.735
- type: mrr_at_5
value: 49.394
- type: ndcg_at_1
value: 39.214
- type: ndcg_at_10
value: 53.52400000000001
- type: ndcg_at_100
value: 56.997
- type: ndcg_at_1000
value: 57.975
- type: ndcg_at_3
value: 48.173
- type: ndcg_at_5
value: 51.05800000000001
- type: precision_at_1
value: 39.214
- type: precision_at_10
value: 7.573
- type: precision_at_100
value: 0.9440000000000001
- type: precision_at_1000
value: 0.104
- type: precision_at_3
value: 19.782
- type: precision_at_5
value: 13.453000000000001
- type: recall_at_1
value: 36.609
- type: recall_at_10
value: 69.247
- type: recall_at_100
value: 84.99600000000001
- type: recall_at_1000
value: 92.40899999999999
- type: recall_at_3
value: 54.856
- type: recall_at_5
value: 61.797000000000004
- task:
type: Retrieval
dataset:
type: fiqa
name: MTEB FiQA2018
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 16.466
- type: map_at_10
value: 27.060000000000002
- type: map_at_100
value: 28.511999999999997
- type: map_at_1000
value: 28.693
- type: map_at_3
value: 22.777
- type: map_at_5
value: 25.086000000000002
- type: mrr_at_1
value: 32.716
- type: mrr_at_10
value: 41.593999999999994
- type: mrr_at_100
value: 42.370000000000005
- type: mrr_at_1000
value: 42.419000000000004
- type: mrr_at_3
value: 38.143
- type: mrr_at_5
value: 40.288000000000004
- type: ndcg_at_1
value: 32.716
- type: ndcg_at_10
value: 34.795
- type: ndcg_at_100
value: 40.58
- type: ndcg_at_1000
value: 43.993
- type: ndcg_at_3
value: 29.573
- type: ndcg_at_5
value: 31.583
- type: precision_at_1
value: 32.716
- type: precision_at_10
value: 9.937999999999999
- type: precision_at_100
value: 1.585
- type: precision_at_1000
value: 0.22
- type: precision_at_3
value: 19.496
- type: precision_at_5
value: 15.247
- type: recall_at_1
value: 16.466
- type: recall_at_10
value: 42.886
- type: recall_at_100
value: 64.724
- type: recall_at_1000
value: 85.347
- type: recall_at_3
value: 26.765
- type: recall_at_5
value: 33.603
- task:
type: Retrieval
dataset:
type: hotpotqa
name: MTEB HotpotQA
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 33.025
- type: map_at_10
value: 47.343
- type: map_at_100
value: 48.207
- type: map_at_1000
value: 48.281
- type: map_at_3
value: 44.519
- type: map_at_5
value: 46.217000000000006
- type: mrr_at_1
value: 66.05
- type: mrr_at_10
value: 72.94699999999999
- type: mrr_at_100
value: 73.289
- type: mrr_at_1000
value: 73.30499999999999
- type: mrr_at_3
value: 71.686
- type: mrr_at_5
value: 72.491
- type: ndcg_at_1
value: 66.05
- type: ndcg_at_10
value: 56.338
- type: ndcg_at_100
value: 59.599999999999994
- type: ndcg_at_1000
value: 61.138000000000005
- type: ndcg_at_3
value: 52.034000000000006
- type: ndcg_at_5
value: 54.352000000000004
- type: precision_at_1
value: 66.05
- type: precision_at_10
value: 11.693000000000001
- type: precision_at_100
value: 1.425
- type: precision_at_1000
value: 0.163
- type: precision_at_3
value: 32.613
- type: precision_at_5
value: 21.401999999999997
- type: recall_at_1
value: 33.025
- type: recall_at_10
value: 58.467
- type: recall_at_100
value: 71.242
- type: recall_at_1000
value: 81.452
- type: recall_at_3
value: 48.92
- type: recall_at_5
value: 53.504
- task:
type: Classification
dataset:
type: mteb/imdb
name: MTEB ImdbClassification
config: default
split: test
revision: 3d86128a09e091d6018b6d26cad27f2739fc2db7
metrics:
- type: accuracy
value: 75.5492
- type: ap
value: 69.42911637216271
- type: f1
value: 75.39113704261024
- task:
type: Retrieval
dataset:
type: msmarco
name: MTEB MSMARCO
config: default
split: dev
revision: None
metrics:
- type: map_at_1
value: 23.173
- type: map_at_10
value: 35.453
- type: map_at_100
value: 36.573
- type: map_at_1000
value: 36.620999999999995
- type: map_at_3
value: 31.655
- type: map_at_5
value: 33.823
- type: mrr_at_1
value: 23.868000000000002
- type: mrr_at_10
value: 36.085
- type: mrr_at_100
value: 37.15
- type: mrr_at_1000
value: 37.193
- type: mrr_at_3
value: 32.376
- type: mrr_at_5
value: 34.501
- type: ndcg_at_1
value: 23.854
- type: ndcg_at_10
value: 42.33
- type: ndcg_at_100
value: 47.705999999999996
- type: ndcg_at_1000
value: 48.91
- type: ndcg_at_3
value: 34.604
- type: ndcg_at_5
value: 38.473
- type: precision_at_1
value: 23.854
- type: precision_at_10
value: 6.639
- type: precision_at_100
value: 0.932
- type: precision_at_1000
value: 0.104
- type: precision_at_3
value: 14.685
- type: precision_at_5
value: 10.782
- type: recall_at_1
value: 23.173
- type: recall_at_10
value: 63.441
- type: recall_at_100
value: 88.25
- type: recall_at_1000
value: 97.438
- type: recall_at_3
value: 42.434
- type: recall_at_5
value: 51.745
- task:
type: Classification
dataset:
type: mteb/mtop_domain
name: MTEB MTOPDomainClassification (en)
config: en
split: test
revision: d80d48c1eb48d3562165c59d59d0034df9fff0bf
metrics:
- type: accuracy
value: 92.05426356589147
- type: f1
value: 91.88068588063942
- task:
type: Classification
dataset:
type: mteb/mtop_intent
name: MTEB MTOPIntentClassification (en)
config: en
split: test
revision: ae001d0e6b1228650b7bd1c2c65fb50ad11a8aba
metrics:
- type: accuracy
value: 73.23985408116735
- type: f1
value: 55.858906745287506
- task:
type: Classification
dataset:
type: mteb/amazon_massive_intent
name: MTEB MassiveIntentClassification (en)
config: en
split: test
revision: 31efe3c427b0bae9c22cbb560b8f15491cc6bed7
metrics:
- type: accuracy
value: 72.21923335574984
- type: f1
value: 70.0174116204253
- task:
type: Classification
dataset:
type: mteb/amazon_massive_scenario
name: MTEB MassiveScenarioClassification (en)
config: en
split: test
revision: 7d571f92784cd94a019292a1f45445077d0ef634
metrics:
- type: accuracy
value: 75.77673167451245
- type: f1
value: 75.44811354778666
- task:
type: Clustering
dataset:
type: mteb/medrxiv-clustering-p2p
name: MTEB MedrxivClusteringP2P
config: default
split: test
revision: e7a26af6f3ae46b30dde8737f02c07b1505bcc73
metrics:
- type: v_measure
value: 31.340414710728737
- task:
type: Clustering
dataset:
type: mteb/medrxiv-clustering-s2s
name: MTEB MedrxivClusteringS2S
config: default
split: test
revision: 35191c8c0dca72d8ff3efcd72aa802307d469663
metrics:
- type: v_measure
value: 28.196676760061578
- task:
type: Reranking
dataset:
type: mteb/mind_small
name: MTEB MindSmallReranking
config: default
split: test
revision: 3bdac13927fdc888b903db93b2ffdbd90b295a69
metrics:
- type: map
value: 29.564149683482206
- type: mrr
value: 30.28995474250486
- task:
type: Retrieval
dataset:
type: nfcorpus
name: MTEB NFCorpus
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 5.93
- type: map_at_10
value: 12.828000000000001
- type: map_at_100
value: 15.501000000000001
- type: map_at_1000
value: 16.791
- type: map_at_3
value: 9.727
- type: map_at_5
value: 11.318999999999999
- type: mrr_at_1
value: 47.678
- type: mrr_at_10
value: 55.893
- type: mrr_at_100
value: 56.491
- type: mrr_at_1000
value: 56.53
- type: mrr_at_3
value: 54.386
- type: mrr_at_5
value: 55.516
- type: ndcg_at_1
value: 45.975
- type: ndcg_at_10
value: 33.928999999999995
- type: ndcg_at_100
value: 30.164
- type: ndcg_at_1000
value: 38.756
- type: ndcg_at_3
value: 41.077000000000005
- type: ndcg_at_5
value: 38.415
- type: precision_at_1
value: 47.678
- type: precision_at_10
value: 24.365000000000002
- type: precision_at_100
value: 7.344
- type: precision_at_1000
value: 1.994
- type: precision_at_3
value: 38.184000000000005
- type: precision_at_5
value: 33.003
- type: recall_at_1
value: 5.93
- type: recall_at_10
value: 16.239
- type: recall_at_100
value: 28.782999999999998
- type: recall_at_1000
value: 60.11
- type: recall_at_3
value: 10.700999999999999
- type: recall_at_5
value: 13.584
- task:
type: Retrieval
dataset:
type: nq
name: MTEB NQ
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 36.163000000000004
- type: map_at_10
value: 51.520999999999994
- type: map_at_100
value: 52.449
- type: map_at_1000
value: 52.473000000000006
- type: map_at_3
value: 47.666
- type: map_at_5
value: 50.043000000000006
- type: mrr_at_1
value: 40.266999999999996
- type: mrr_at_10
value: 54.074
- type: mrr_at_100
value: 54.722
- type: mrr_at_1000
value: 54.739000000000004
- type: mrr_at_3
value: 51.043000000000006
- type: mrr_at_5
value: 52.956
- type: ndcg_at_1
value: 40.238
- type: ndcg_at_10
value: 58.73199999999999
- type: ndcg_at_100
value: 62.470000000000006
- type: ndcg_at_1000
value: 63.083999999999996
- type: ndcg_at_3
value: 51.672
- type: ndcg_at_5
value: 55.564
- type: precision_at_1
value: 40.238
- type: precision_at_10
value: 9.279
- type: precision_at_100
value: 1.139
- type: precision_at_1000
value: 0.12
- type: precision_at_3
value: 23.078000000000003
- type: precision_at_5
value: 16.176
- type: recall_at_1
value: 36.163000000000004
- type: recall_at_10
value: 77.88199999999999
- type: recall_at_100
value: 93.83399999999999
- type: recall_at_1000
value: 98.465
- type: recall_at_3
value: 59.857000000000006
- type: recall_at_5
value: 68.73599999999999
- task:
type: Retrieval
dataset:
type: quora
name: MTEB QuoraRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 70.344
- type: map_at_10
value: 83.907
- type: map_at_100
value: 84.536
- type: map_at_1000
value: 84.557
- type: map_at_3
value: 80.984
- type: map_at_5
value: 82.844
- type: mrr_at_1
value: 81.02000000000001
- type: mrr_at_10
value: 87.158
- type: mrr_at_100
value: 87.268
- type: mrr_at_1000
value: 87.26899999999999
- type: mrr_at_3
value: 86.17
- type: mrr_at_5
value: 86.87
- type: ndcg_at_1
value: 81.02000000000001
- type: ndcg_at_10
value: 87.70700000000001
- type: ndcg_at_100
value: 89.004
- type: ndcg_at_1000
value: 89.139
- type: ndcg_at_3
value: 84.841
- type: ndcg_at_5
value: 86.455
- type: precision_at_1
value: 81.02000000000001
- type: precision_at_10
value: 13.248999999999999
- type: precision_at_100
value: 1.516
- type: precision_at_1000
value: 0.156
- type: precision_at_3
value: 36.963
- type: precision_at_5
value: 24.33
- type: recall_at_1
value: 70.344
- type: recall_at_10
value: 94.75099999999999
- type: recall_at_100
value: 99.30499999999999
- type: recall_at_1000
value: 99.928
- type: recall_at_3
value: 86.506
- type: recall_at_5
value: 91.083
- task:
type: Clustering
dataset:
type: mteb/reddit-clustering
name: MTEB RedditClustering
config: default
split: test
revision: 24640382cdbf8abc73003fb0fa6d111a705499eb
metrics:
- type: v_measure
value: 42.873718018378305
- task:
type: Clustering
dataset:
type: mteb/reddit-clustering-p2p
name: MTEB RedditClusteringP2P
config: default
split: test
revision: 282350215ef01743dc01b456c7f5241fa8937f16
metrics:
- type: v_measure
value: 56.39477366450528
- task:
type: Retrieval
dataset:
type: scidocs
name: MTEB SCIDOCS
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 3.868
- type: map_at_10
value: 9.611
- type: map_at_100
value: 11.087
- type: map_at_1000
value: 11.332
- type: map_at_3
value: 6.813
- type: map_at_5
value: 8.233
- type: mrr_at_1
value: 19
- type: mrr_at_10
value: 28.457
- type: mrr_at_100
value: 29.613
- type: mrr_at_1000
value: 29.695
- type: mrr_at_3
value: 25.55
- type: mrr_at_5
value: 27.29
- type: ndcg_at_1
value: 19
- type: ndcg_at_10
value: 16.419
- type: ndcg_at_100
value: 22.817999999999998
- type: ndcg_at_1000
value: 27.72
- type: ndcg_at_3
value: 15.379000000000001
- type: ndcg_at_5
value: 13.645
- type: precision_at_1
value: 19
- type: precision_at_10
value: 8.540000000000001
- type: precision_at_100
value: 1.7819999999999998
- type: precision_at_1000
value: 0.297
- type: precision_at_3
value: 14.267
- type: precision_at_5
value: 12.04
- type: recall_at_1
value: 3.868
- type: recall_at_10
value: 17.288
- type: recall_at_100
value: 36.144999999999996
- type: recall_at_1000
value: 60.199999999999996
- type: recall_at_3
value: 8.688
- type: recall_at_5
value: 12.198
- task:
type: STS
dataset:
type: mteb/sickr-sts
name: MTEB SICK-R
config: default
split: test
revision: a6ea5a8cab320b040a23452cc28066d9beae2cee
metrics:
- type: cos_sim_pearson
value: 83.96614722598582
- type: cos_sim_spearman
value: 78.9003023008781
- type: euclidean_pearson
value: 81.01829384436505
- type: euclidean_spearman
value: 78.93248416788914
- type: manhattan_pearson
value: 81.1665428926402
- type: manhattan_spearman
value: 78.93264116287453
- task:
type: STS
dataset:
type: mteb/sts12-sts
name: MTEB STS12
config: default
split: test
revision: a0d554a64d88156834ff5ae9920b964011b16384
metrics:
- type: cos_sim_pearson
value: 83.54613363895993
- type: cos_sim_spearman
value: 75.1883451602451
- type: euclidean_pearson
value: 79.70320886899894
- type: euclidean_spearman
value: 74.5917140136796
- type: manhattan_pearson
value: 79.82157067185999
- type: manhattan_spearman
value: 74.74185720594735
- task:
type: STS
dataset:
type: mteb/sts13-sts
name: MTEB STS13
config: default
split: test
revision: 7e90230a92c190f1bf69ae9002b8cea547a64cca
metrics:
- type: cos_sim_pearson
value: 81.30430156721782
- type: cos_sim_spearman
value: 81.79962989974364
- type: euclidean_pearson
value: 80.89058823224924
- type: euclidean_spearman
value: 81.35929372984597
- type: manhattan_pearson
value: 81.12204370487478
- type: manhattan_spearman
value: 81.6248963282232
- task:
type: STS
dataset:
type: mteb/sts14-sts
name: MTEB STS14
config: default
split: test
revision: 6031580fec1f6af667f0bd2da0a551cf4f0b2375
metrics:
- type: cos_sim_pearson
value: 81.13064504403134
- type: cos_sim_spearman
value: 78.48371403924872
- type: euclidean_pearson
value: 80.16794919665591
- type: euclidean_spearman
value: 78.29216082221699
- type: manhattan_pearson
value: 80.22308565207301
- type: manhattan_spearman
value: 78.37829229948022
- task:
type: STS
dataset:
type: mteb/sts15-sts
name: MTEB STS15
config: default
split: test
revision: ae752c7c21bf194d8b67fd573edf7ae58183cbe3
metrics:
- type: cos_sim_pearson
value: 86.52918899541099
- type: cos_sim_spearman
value: 87.49276894673142
- type: euclidean_pearson
value: 86.77440570164254
- type: euclidean_spearman
value: 87.5753295736756
- type: manhattan_pearson
value: 86.86098573892133
- type: manhattan_spearman
value: 87.65848591821947
- task:
type: STS
dataset:
type: mteb/sts16-sts
name: MTEB STS16
config: default
split: test
revision: 4d8694f8f0e0100860b497b999b3dbed754a0513
metrics:
- type: cos_sim_pearson
value: 82.86805307244882
- type: cos_sim_spearman
value: 84.58066253757511
- type: euclidean_pearson
value: 84.38377000876991
- type: euclidean_spearman
value: 85.1837278784528
- type: manhattan_pearson
value: 84.41903291363842
- type: manhattan_spearman
value: 85.19023736251052
- task:
type: STS
dataset:
type: mteb/sts17-crosslingual-sts
name: MTEB STS17 (en-en)
config: en-en
split: test
revision: af5e6fb845001ecf41f4c1e033ce921939a2a68d
metrics:
- type: cos_sim_pearson
value: 86.77218560282436
- type: cos_sim_spearman
value: 87.94243515296604
- type: euclidean_pearson
value: 88.22800939214864
- type: euclidean_spearman
value: 87.91106839439841
- type: manhattan_pearson
value: 88.17063269848741
- type: manhattan_spearman
value: 87.72751904126062
- task:
type: STS
dataset:
type: mteb/sts22-crosslingual-sts
name: MTEB STS22 (en)
config: en
split: test
revision: 6d1ba47164174a496b7fa5d3569dae26a6813b80
metrics:
- type: cos_sim_pearson
value: 60.40731554300387
- type: cos_sim_spearman
value: 63.76300532966479
- type: euclidean_pearson
value: 62.94727878229085
- type: euclidean_spearman
value: 63.678039531461216
- type: manhattan_pearson
value: 63.00661039863549
- type: manhattan_spearman
value: 63.6282591984376
- task:
type: STS
dataset:
type: mteb/stsbenchmark-sts
name: MTEB STSBenchmark
config: default
split: test
revision: b0fddb56ed78048fa8b90373c8a3cfc37b684831
metrics:
- type: cos_sim_pearson
value: 84.92731569745344
- type: cos_sim_spearman
value: 86.36336704300167
- type: euclidean_pearson
value: 86.09122224841195
- type: euclidean_spearman
value: 86.2116149319238
- type: manhattan_pearson
value: 86.07879456717032
- type: manhattan_spearman
value: 86.2022069635119
- task:
type: Reranking
dataset:
type: mteb/scidocs-reranking
name: MTEB SciDocsRR
config: default
split: test
revision: d3c5e1fc0b855ab6097bf1cda04dd73947d7caab
metrics:
- type: map
value: 79.75976311752326
- type: mrr
value: 94.15782837351466
- task:
type: Retrieval
dataset:
type: scifact
name: MTEB SciFact
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 51.193999999999996
- type: map_at_10
value: 61.224999999999994
- type: map_at_100
value: 62.031000000000006
- type: map_at_1000
value: 62.066
- type: map_at_3
value: 59.269000000000005
- type: map_at_5
value: 60.159
- type: mrr_at_1
value: 53.667
- type: mrr_at_10
value: 62.74999999999999
- type: mrr_at_100
value: 63.39399999999999
- type: mrr_at_1000
value: 63.425
- type: mrr_at_3
value: 61.389
- type: mrr_at_5
value: 61.989000000000004
- type: ndcg_at_1
value: 53.667
- type: ndcg_at_10
value: 65.596
- type: ndcg_at_100
value: 68.906
- type: ndcg_at_1000
value: 69.78999999999999
- type: ndcg_at_3
value: 62.261
- type: ndcg_at_5
value: 63.453
- type: precision_at_1
value: 53.667
- type: precision_at_10
value: 8.667
- type: precision_at_100
value: 1.04
- type: precision_at_1000
value: 0.11100000000000002
- type: precision_at_3
value: 24.556
- type: precision_at_5
value: 15.6
- type: recall_at_1
value: 51.193999999999996
- type: recall_at_10
value: 77.156
- type: recall_at_100
value: 91.43299999999999
- type: recall_at_1000
value: 98.333
- type: recall_at_3
value: 67.994
- type: recall_at_5
value: 71.14399999999999
- task:
type: PairClassification
dataset:
type: mteb/sprintduplicatequestions-pairclassification
name: MTEB SprintDuplicateQuestions
config: default
split: test
revision: d66bd1f72af766a5cc4b0ca5e00c162f89e8cc46
metrics:
- type: cos_sim_accuracy
value: 99.81485148514851
- type: cos_sim_ap
value: 95.28896513388551
- type: cos_sim_f1
value: 90.43478260869566
- type: cos_sim_precision
value: 92.56544502617801
- type: cos_sim_recall
value: 88.4
- type: dot_accuracy
value: 99.30594059405941
- type: dot_ap
value: 61.6432597455472
- type: dot_f1
value: 59.46481665014866
- type: dot_precision
value: 58.93909626719057
- type: dot_recall
value: 60
- type: euclidean_accuracy
value: 99.81980198019802
- type: euclidean_ap
value: 95.21411049527
- type: euclidean_f1
value: 91.06090373280944
- type: euclidean_precision
value: 89.47876447876449
- type: euclidean_recall
value: 92.7
- type: manhattan_accuracy
value: 99.81782178217821
- type: manhattan_ap
value: 95.32449994414968
- type: manhattan_f1
value: 90.86395233366436
- type: manhattan_precision
value: 90.23668639053254
- type: manhattan_recall
value: 91.5
- type: max_accuracy
value: 99.81980198019802
- type: max_ap
value: 95.32449994414968
- type: max_f1
value: 91.06090373280944
- task:
type: Clustering
dataset:
type: mteb/stackexchange-clustering
name: MTEB StackExchangeClustering
config: default
split: test
revision: 6cbc1f7b2bc0622f2e39d2c77fa502909748c259
metrics:
- type: v_measure
value: 59.08045614613064
- task:
type: Clustering
dataset:
type: mteb/stackexchange-clustering-p2p
name: MTEB StackExchangeClusteringP2P
config: default
split: test
revision: 815ca46b2622cec33ccafc3735d572c266efdb44
metrics:
- type: v_measure
value: 30.297802606804748
- task:
type: Reranking
dataset:
type: mteb/stackoverflowdupquestions-reranking
name: MTEB StackOverflowDupQuestions
config: default
split: test
revision: e185fbe320c72810689fc5848eb6114e1ef5ec69
metrics:
- type: map
value: 49.12801740706292
- type: mrr
value: 50.05592956879722
- task:
type: Summarization
dataset:
type: mteb/summeval
name: MTEB SummEval
config: default
split: test
revision: cda12ad7615edc362dbf25a00fdd61d3b1eaf93c
metrics:
- type: cos_sim_pearson
value: 31.523347880124497
- type: cos_sim_spearman
value: 31.388214436391014
- type: dot_pearson
value: 24.55403435439901
- type: dot_spearman
value: 23.50153210841191
- task:
type: Retrieval
dataset:
type: trec-covid
name: MTEB TRECCOVID
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 0.243
- type: map_at_10
value: 1.886
- type: map_at_100
value: 10.040000000000001
- type: map_at_1000
value: 23.768
- type: map_at_3
value: 0.674
- type: map_at_5
value: 1.079
- type: mrr_at_1
value: 88
- type: mrr_at_10
value: 93.667
- type: mrr_at_100
value: 93.667
- type: mrr_at_1000
value: 93.667
- type: mrr_at_3
value: 93.667
- type: mrr_at_5
value: 93.667
- type: ndcg_at_1
value: 83
- type: ndcg_at_10
value: 76.777
- type: ndcg_at_100
value: 55.153
- type: ndcg_at_1000
value: 47.912
- type: ndcg_at_3
value: 81.358
- type: ndcg_at_5
value: 80.74799999999999
- type: precision_at_1
value: 88
- type: precision_at_10
value: 80.80000000000001
- type: precision_at_100
value: 56.02
- type: precision_at_1000
value: 21.51
- type: precision_at_3
value: 86
- type: precision_at_5
value: 86
- type: recall_at_1
value: 0.243
- type: recall_at_10
value: 2.0869999999999997
- type: recall_at_100
value: 13.014000000000001
- type: recall_at_1000
value: 44.433
- type: recall_at_3
value: 0.6910000000000001
- type: recall_at_5
value: 1.1440000000000001
- task:
type: Retrieval
dataset:
type: webis-touche2020
name: MTEB Touche2020
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 3.066
- type: map_at_10
value: 10.615
- type: map_at_100
value: 16.463
- type: map_at_1000
value: 17.815
- type: map_at_3
value: 5.7860000000000005
- type: map_at_5
value: 7.353999999999999
- type: mrr_at_1
value: 38.775999999999996
- type: mrr_at_10
value: 53.846000000000004
- type: mrr_at_100
value: 54.37
- type: mrr_at_1000
value: 54.37
- type: mrr_at_3
value: 48.980000000000004
- type: mrr_at_5
value: 51.735
- type: ndcg_at_1
value: 34.694
- type: ndcg_at_10
value: 26.811
- type: ndcg_at_100
value: 37.342999999999996
- type: ndcg_at_1000
value: 47.964
- type: ndcg_at_3
value: 30.906
- type: ndcg_at_5
value: 27.77
- type: precision_at_1
value: 38.775999999999996
- type: precision_at_10
value: 23.878
- type: precision_at_100
value: 7.632999999999999
- type: precision_at_1000
value: 1.469
- type: precision_at_3
value: 31.973000000000003
- type: precision_at_5
value: 26.939
- type: recall_at_1
value: 3.066
- type: recall_at_10
value: 17.112
- type: recall_at_100
value: 47.723
- type: recall_at_1000
value: 79.50500000000001
- type: recall_at_3
value: 6.825
- type: recall_at_5
value: 9.584
- task:
type: Classification
dataset:
type: mteb/toxic_conversations_50k
name: MTEB ToxicConversationsClassification
config: default
split: test
revision: d7c0de2777da35d6aae2200a62c6e0e5af397c4c
metrics:
- type: accuracy
value: 72.76460000000002
- type: ap
value: 14.944240012137053
- type: f1
value: 55.89805777266571
- task:
type: Classification
dataset:
type: mteb/tweet_sentiment_extraction
name: MTEB TweetSentimentExtractionClassification
config: default
split: test
revision: d604517c81ca91fe16a244d1248fc021f9ecee7a
metrics:
- type: accuracy
value: 63.30503678551217
- type: f1
value: 63.57492701921179
- task:
type: Clustering
dataset:
type: mteb/twentynewsgroups-clustering
name: MTEB TwentyNewsgroupsClustering
config: default
split: test
revision: 6125ec4e24fa026cec8a478383ee943acfbd5449
metrics:
- type: v_measure
value: 37.51066495006874
- task:
type: PairClassification
dataset:
type: mteb/twittersemeval2015-pairclassification
name: MTEB TwitterSemEval2015
config: default
split: test
revision: 70970daeab8776df92f5ea462b6173c0b46fd2d1
metrics:
- type: cos_sim_accuracy
value: 86.07021517553794
- type: cos_sim_ap
value: 74.15520712370555
- type: cos_sim_f1
value: 68.64321608040201
- type: cos_sim_precision
value: 65.51558752997602
- type: cos_sim_recall
value: 72.0844327176781
- type: dot_accuracy
value: 80.23484532395541
- type: dot_ap
value: 54.298763810214176
- type: dot_f1
value: 53.22254659779924
- type: dot_precision
value: 46.32525410476936
- type: dot_recall
value: 62.532981530343015
- type: euclidean_accuracy
value: 86.04637301066937
- type: euclidean_ap
value: 73.85333854233123
- type: euclidean_f1
value: 68.77723660599845
- type: euclidean_precision
value: 66.87437686939182
- type: euclidean_recall
value: 70.79155672823218
- type: manhattan_accuracy
value: 85.98676759849795
- type: manhattan_ap
value: 73.56016090035973
- type: manhattan_f1
value: 68.48878539036647
- type: manhattan_precision
value: 63.9505607690547
- type: manhattan_recall
value: 73.7203166226913
- type: max_accuracy
value: 86.07021517553794
- type: max_ap
value: 74.15520712370555
- type: max_f1
value: 68.77723660599845
- task:
type: PairClassification
dataset:
type: mteb/twitterurlcorpus-pairclassification
name: MTEB TwitterURLCorpus
config: default
split: test
revision: 8b6510b0b1fa4e4c4f879467980e9be563ec1cdf
metrics:
- type: cos_sim_accuracy
value: 88.92769821865176
- type: cos_sim_ap
value: 85.78879502899773
- type: cos_sim_f1
value: 78.14414083990464
- type: cos_sim_precision
value: 74.61651607480563
- type: cos_sim_recall
value: 82.0218663381583
- type: dot_accuracy
value: 84.95750378390964
- type: dot_ap
value: 75.80219641857563
- type: dot_f1
value: 70.13966179585681
- type: dot_precision
value: 65.71140262361251
- type: dot_recall
value: 75.20788420080073
- type: euclidean_accuracy
value: 88.93546008460433
- type: euclidean_ap
value: 85.72056428301667
- type: euclidean_f1
value: 78.14387902598124
- type: euclidean_precision
value: 75.3376688344172
- type: euclidean_recall
value: 81.16723129042192
- type: manhattan_accuracy
value: 88.96262661543835
- type: manhattan_ap
value: 85.76605136314335
- type: manhattan_f1
value: 78.26696165191743
- type: manhattan_precision
value: 75.0990659496179
- type: manhattan_recall
value: 81.71388974437943
- type: max_accuracy
value: 88.96262661543835
- type: max_ap
value: 85.78879502899773
- type: max_f1
value: 78.26696165191743
language:
- en
license: mit
---
# E5-small
[Text Embeddings by Weakly-Supervised Contrastive Pre-training](https://arxiv.org/pdf/2212.03533.pdf).
Liang Wang, Nan Yang, Xiaolong Huang, Binxing Jiao, Linjun Yang, Daxin Jiang, Rangan Majumder, Furu Wei, arXiv 2022
This model has 12 layers and the embedding size is 384.
## Usage
Below is an example to encode queries and passages from the MS-MARCO passage ranking dataset.
```python
import torch.nn.functional as F
from torch import Tensor
from transformers import AutoTokenizer, AutoModel
def average_pool(last_hidden_states: Tensor,
attention_mask: Tensor) -> Tensor:
last_hidden = last_hidden_states.masked_fill(~attention_mask[..., None].bool(), 0.0)
return last_hidden.sum(dim=1) / attention_mask.sum(dim=1)[..., None]
# Each input text should start with "query: " or "passage: ".
# For tasks other than retrieval, you can simply use the "query: " prefix.
input_texts = ['query: how much protein should a female eat',
'query: summit define',
"passage: As a general guideline, the CDC's average requirement of protein for women ages 19 to 70 is 46 grams per day. But, as you can see from this chart, you'll need to increase that if you're expecting or training for a marathon. Check out the chart below to see how much protein you should be eating each day.",
"passage: Definition of summit for English Language Learners. : 1 the highest point of a mountain : the top of a mountain. : 2 the highest level. : 3 a meeting or series of meetings between the leaders of two or more governments."]
tokenizer = AutoTokenizer.from_pretrained('intfloat/e5-small')
model = AutoModel.from_pretrained('intfloat/e5-small')
# Tokenize the input texts
batch_dict = tokenizer(input_texts, max_length=512, padding=True, truncation=True, return_tensors='pt')
outputs = model(**batch_dict)
embeddings = average_pool(outputs.last_hidden_state, batch_dict['attention_mask'])
# (Optionally) normalize embeddings
embeddings = F.normalize(embeddings, p=2, dim=1)
scores = (embeddings[:2] @ embeddings[2:].T) * 100
print(scores.tolist())
```
## Training Details
Please refer to our paper at [https://arxiv.org/pdf/2212.03533.pdf](https://arxiv.org/pdf/2212.03533.pdf).
## Benchmark Evaluation
Check out [unilm/e5](https://github.com/microsoft/unilm/tree/master/e5) to reproduce evaluation results
on the [BEIR](https://arxiv.org/abs/2104.08663) and [MTEB benchmark](https://arxiv.org/abs/2210.07316).
## Citation
If you find our paper or models helpful, please consider cite as follows:
```
@article{wang2022text,
title={Text Embeddings by Weakly-Supervised Contrastive Pre-training},
author={Wang, Liang and Yang, Nan and Huang, Xiaolong and Jiao, Binxing and Yang, Linjun and Jiang, Daxin and Majumder, Rangan and Wei, Furu},
journal={arXiv preprint arXiv:2212.03533},
year={2022}
}
```
## Limitations
This model only works for English texts. Long texts will be truncated to at most 512 tokens. |
Akashpb13/Swahili_xlsr | [
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"sw",
"dataset:mozilla-foundation/common_voice_8_0",
"transformers",
"generated_from_trainer",
"hf-asr-leaderboard",
"model_for_talk",
"mozilla-foundation/common_voice_8_0",
"robust-speech-event",
"license:apache-2.0",
"model-index"
]
| automatic-speech-recognition | {
"architectures": [
"Wav2Vec2ForCTC"
],
"model_type": "wav2vec2",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 10 | null | ---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: my_awesome_model
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. -->
# my_awesome_model
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset.
## 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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
### Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1+cu113
- Tokenizers 0.13.2
|
Akashpb13/xlsr_hungarian_new | [
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"hu",
"dataset:mozilla-foundation/common_voice_8_0",
"transformers",
"generated_from_trainer",
"hf-asr-leaderboard",
"model_for_talk",
"mozilla-foundation/common_voice_8_0",
"robust-speech-event",
"license:apache-2.0",
"model-index"
]
| automatic-speech-recognition | {
"architectures": [
"Wav2Vec2ForCTC"
],
"model_type": "wav2vec2",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 7 | null | ---
license: creativeml-openrail-m
---
### June from [Obituary - A Grave Beginning](https://invidious.weblibre.org/watch?v=0l940bPkV1o) on [WD](https://huggingface.co/hakurei/waifu-diffusion) via Dreambooth
#### model by no3
This your waifu-diffusion v1.3 model fine-tuned june taught to waifu-diffusion v1.3 with Dreambooth.
It can be used by modifying the `instance_prompt`: **sks_june**
ou can also train your own concepts and upload them to the library by using [this notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/sd_dreambooth_training.ipynb).
And you can run your new concept via `diffusers`: [Colab Notebook for Inference](https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/sd_dreambooth_inference.ipynb), [Spaces with the Public Concepts loaded](https://huggingface.co/spaces/sd-dreambooth-library/stable-diffusion-dreambooth-concepts).
### note
If you want to to use in UI like [AUTOMATIC1111](https://github.com/AUTOMATIC1111/stable-diffusion-webui) or any UI that's uses .ckpt files just download ckpt file here for your convenience. **just click on "june-wd-1.3-beta2.ckpt"**
[june-wd-1.3-beta2.ckpt](https://huggingface.co/no3/june-wd-1.3-beta2/resolve/main/june-wd-1.3-beta2.ckpt)
If you have issues or questions feel free to visit the Community Tab and start discussion about it.
Here are images used for training this concept:
 |
AkshatSurolia/ConvNeXt-FaceMask-Finetuned | [
"pytorch",
"safetensors",
"convnext",
"image-classification",
"dataset:Face-Mask18K",
"transformers",
"license:apache-2.0",
"autotrain_compatible",
"has_space"
]
| image-classification | {
"architectures": [
"ConvNextForImageClassification"
],
"model_type": "convnext",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 56 | null | ---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
metrics:
- type: mean_reward
value: 226.55 +/- 49.07
name: mean_reward
verified: false
---
# **PPO** Agent playing **LunarLander-v2**
This is a trained model of a **PPO** agent playing **LunarLander-v2**
using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3).
## Usage (with Stable-baselines3)
TODO: Add your code
```python
from stable_baselines3 import ...
from huggingface_sb3 import load_from_hub
...
```
|
AkshatSurolia/DeiT-FaceMask-Finetuned | [
"pytorch",
"deit",
"image-classification",
"dataset:Face-Mask18K",
"transformers",
"license:apache-2.0",
"autotrain_compatible"
]
| image-classification | {
"architectures": [
"DeiTForImageClassification"
],
"model_type": "deit",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 46 | null | ---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: my_awesome_wnut_model
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. -->
# my_awesome_wnut_model
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset.
## 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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
### Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1+cu113
- Tokenizers 0.13.2
|
AkshayDev/BERT_Fine_Tuning | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language: en
license: apache-2.0
library_name: diffusers
tags: []
datasets: EmileEsmaili/sheet_music_ede2110
metrics: []
---
<!-- This model card has been generated automatically according to the information the training script had access to. You
should probably proofread and complete it, then remove this comment. -->
# ddpm-sheetmusic-colabVM
## Model description
This diffusion model is trained with the [🤗 Diffusers](https://github.com/huggingface/diffusers) library
on the `EmileEsmaili/sheet_music_ede2110` dataset.
## Intended uses & limitations
#### How to use
```python
# TODO: add an example code snippet for running this diffusion pipeline
```
#### Limitations and bias
[TODO: provide examples of latent issues and potential remediations]
## Training data
[TODO: describe the data used to train the model]
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0001
- train_batch_size: 16
- eval_batch_size: 16
- gradient_accumulation_steps: 1
- optimizer: AdamW with betas=(None, None), weight_decay=None and epsilon=None
- lr_scheduler: None
- lr_warmup_steps: 50
- ema_inv_gamma: None
- ema_inv_gamma: None
- ema_inv_gamma: None
- mixed_precision: no
### Training results
📈 [TensorBoard logs](https://huggingface.co/EmileEsmaili/ddpm-sheetmusic-colabVM/tensorboard?#scalars)
|
AkshaySg/GrammarCorrection | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
language:
- it
license: apache-2.0
tags:
- whisper-event
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
metrics:
- wer
model-index:
- name: Whisper Small Italian
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: mozilla-foundation/common_voice_11_0 it
type: mozilla-foundation/common_voice_11_0
config: it
split: test
args: it
metrics:
- name: Wer
type: wer
value: 9.26934935147778
---
<!-- 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. -->
# Whisper Small Italian
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the mozilla-foundation/common_voice_11_0 it dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2013
- Wer: 9.2693
## 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: 1e-05
- train_batch_size: 64
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 4000
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 0.2851 | 0.25 | 1000 | 0.2604 | 11.9744 |
| 0.1885 | 0.5 | 2000 | 0.2176 | 10.1358 |
| 0.1176 | 1.15 | 3000 | 0.2111 | 9.5664 |
| 0.1256 | 1.4 | 4000 | 0.2013 | 9.2693 |
### Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu117
- Datasets 2.7.1.dev0
- Tokenizers 0.13.2
|
Ale/Alen | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | null | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- wikisql
model-index:
- name: t5-small-finetuned-wikisql-with-cols
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. -->
# t5-small-finetuned-wikisql-with-cols
This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on the wikisql dataset using a (slightly modified) training script by [Manuel Romero](https://huggingface.co/mrm8488).
It achieves the following results on the evaluation set:
- Loss: 0.0282
- Rouge2 Precision: 0.9172
- Rouge2 Recall: 0.819
- Rouge2 Fmeasure: 0.8578
## 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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge2 Precision | Rouge2 Recall | Rouge2 Fmeasure |
|:-------------:|:-----:|:-----:|:---------------:|:----------------:|:-------------:|:---------------:|
| 0.0557 | 1.0 | 4049 | 0.0384 | 0.9004 | 0.8038 | 0.8417 |
| 0.0438 | 2.0 | 8098 | 0.0323 | 0.9101 | 0.8121 | 0.8507 |
| 0.0374 | 3.0 | 12147 | 0.0298 | 0.914 | 0.8162 | 0.8548 |
| 0.0353 | 4.0 | 16196 | 0.0286 | 0.9169 | 0.8189 | 0.8576 |
| 0.0343 | 5.0 | 20245 | 0.0282 | 0.9172 | 0.819 | 0.8578 |
### Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2
|
Aleenbo/Arcane | []
| null | {
"architectures": null,
"model_type": null,
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 0 | 2022-12-07T08:37:14Z | ---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
metrics:
- type: mean_reward
value: 190.04 +/- 65.99
name: mean_reward
verified: false
---
# **PPO** Agent playing **LunarLander-v2**
This is a trained model of a **PPO** agent playing **LunarLander-v2**
using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3).
## Usage (with Stable-baselines3)
TODO: Add your code
```python
from stable_baselines3 import ...
from huggingface_sb3 import load_from_hub
...
```
|
Aleksandar/bert-srb-ner-setimes | [
"pytorch",
"bert",
"token-classification",
"transformers",
"generated_from_trainer",
"autotrain_compatible"
]
| token-classification | {
"architectures": [
"BertForTokenClassification"
],
"model_type": "bert",
"task_specific_params": {
"conversational": {
"max_length": null
},
"summarization": {
"early_stopping": null,
"length_penalty": null,
"max_length": null,
"min_length": null,
"no_repeat_ngram_size": null,
"num_beams": null,
"prefix": null
},
"text-generation": {
"do_sample": null,
"max_length": null
},
"translation_en_to_de": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_fr": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
},
"translation_en_to_ro": {
"early_stopping": null,
"max_length": null,
"num_beams": null,
"prefix": null
}
}
} | 8 | null | ---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
metrics:
- type: mean_reward
value: -238.48 +/- 82.63
name: mean_reward
verified: false
---
# **PPO** Agent playing **LunarLander-v2**
This is a trained model of a **PPO** agent playing **LunarLander-v2**
using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3).
## Usage (with Stable-baselines3)
TODO: Add your code
```python
from stable_baselines3 import ...
from huggingface_sb3 import load_from_hub
...
```
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.