modelId
string | author
string | last_modified
timestamp[us, tz=UTC] | downloads
int64 | likes
int64 | library_name
string | tags
list | pipeline_tag
string | createdAt
timestamp[us, tz=UTC] | card
string |
|---|---|---|---|---|---|---|---|---|---|
Vlasta/DNADebertaSentencepiece10k_continuation_continuation
|
Vlasta
| 2022-08-08T20:15:51Z
| 10
| 0
|
transformers
|
[
"transformers",
"pytorch",
"deberta",
"fill-mask",
"generated_from_trainer",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
fill-mask
| 2022-08-04T19:00:28Z
|
---
tags:
- generated_from_trainer
model-index:
- name: DNADebertaSentencepiece10k_continuation_continuation
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. -->
# DNADebertaSentencepiece10k_continuation_continuation
This model is a fine-tuned version of [Vlasta/DNADebertaSentencepiece10k_continuation](https://huggingface.co/Vlasta/DNADebertaSentencepiece10k_continuation) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 5.3056
## 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
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 15
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:------:|:---------------:|
| 5.4806 | 0.36 | 5000 | 5.4385 |
| 5.4848 | 0.72 | 10000 | 5.4333 |
| 5.4803 | 1.08 | 15000 | 5.4312 |
| 5.4759 | 1.45 | 20000 | 5.4223 |
| 5.4703 | 1.81 | 25000 | 5.4199 |
| 5.4626 | 2.17 | 30000 | 5.4147 |
| 5.4596 | 2.53 | 35000 | 5.4094 |
| 5.4534 | 2.89 | 40000 | 5.4014 |
| 5.4466 | 3.25 | 45000 | 5.4017 |
| 5.445 | 3.61 | 50000 | 5.3954 |
| 5.4446 | 3.97 | 55000 | 5.3916 |
| 5.4359 | 4.34 | 60000 | 5.3809 |
| 5.4327 | 4.7 | 65000 | 5.3846 |
| 5.4281 | 5.06 | 70000 | 5.3765 |
| 5.4207 | 5.42 | 75000 | 5.3744 |
| 5.4207 | 5.78 | 80000 | 5.3704 |
| 5.4167 | 6.14 | 85000 | 5.3685 |
| 5.41 | 6.5 | 90000 | 5.3641 |
| 5.4117 | 6.86 | 95000 | 5.3582 |
| 5.4075 | 7.23 | 100000 | 5.3568 |
| 5.4017 | 7.59 | 105000 | 5.3547 |
| 5.4006 | 7.95 | 110000 | 5.3494 |
| 5.3969 | 8.31 | 115000 | 5.3475 |
| 5.3935 | 8.67 | 120000 | 5.3453 |
| 5.3926 | 9.03 | 125000 | 5.3422 |
| 5.3895 | 9.39 | 130000 | 5.3351 |
| 5.3813 | 9.75 | 135000 | 5.3326 |
| 5.3841 | 10.12 | 140000 | 5.3340 |
| 5.3787 | 10.48 | 145000 | 5.3301 |
| 5.3781 | 10.84 | 150000 | 5.3280 |
| 5.3769 | 11.2 | 155000 | 5.3258 |
| 5.3733 | 11.56 | 160000 | 5.3198 |
| 5.3683 | 11.92 | 165000 | 5.3180 |
| 5.3682 | 12.28 | 170000 | 5.3181 |
| 5.3673 | 12.64 | 175000 | 5.3167 |
| 5.3623 | 13.01 | 180000 | 5.3116 |
| 5.3602 | 13.37 | 185000 | 5.3109 |
| 5.361 | 13.73 | 190000 | 5.3071 |
| 5.3573 | 14.09 | 195000 | 5.3078 |
| 5.3575 | 14.45 | 200000 | 5.3051 |
| 5.3544 | 14.81 | 205000 | 5.3038 |
### Framework versions
- Transformers 4.19.2
- Pytorch 1.11.0
- Datasets 2.2.2
- Tokenizers 0.12.1
|
ScottMueller/Cats_v_Dogs.ONNX
|
ScottMueller
| 2022-08-08T20:10:32Z
| 0
| 0
| null |
[
"onnx",
"license:mit",
"region:us"
] | null | 2022-08-08T19:54:50Z
|
---
license: mit
---
A simple single label classification model, ResNet18, to predict whether the provided image is a cat or a dog. The model was created in Fast.ai
and exported to ONNX using PyTorch's ONNX export capabilities.
The source dataset is the OXFORD-IIIT PET. Omkar M Parkhi, Andrea Vedaldi, Andrew Zisserman and C. V. Jawahar
We have created a 37 category pet dataset with roughly 200 images for each class.
The images have a large variations in scale, pose and lighting. All images havean
associated ground truth annotation of breed, head ROI, and pixel level trimap segmentation.
The ONNX model can be used in other frameworks like Elixir's Axon. An example of converting the ONNX model into Axon can be found at:
https://github.com/elixir-nx/axon/tree/main/notebooks/onnx_to_axon.livemd.
|
Vlasta/DNADebertaSentencepiece30k_continuation_continuation
|
Vlasta
| 2022-08-08T18:56:18Z
| 5
| 0
|
transformers
|
[
"transformers",
"pytorch",
"deberta",
"fill-mask",
"generated_from_trainer",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
fill-mask
| 2022-08-04T18:58:36Z
|
---
tags:
- generated_from_trainer
model-index:
- name: DNADebertaSentencepiece30k_continuation_continuation
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. -->
# DNADebertaSentencepiece30k_continuation_continuation
This model is a fine-tuned version of [Vlasta/DNADebertaSentencepiece30k_continuation](https://huggingface.co/Vlasta/DNADebertaSentencepiece30k_continuation) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 5.9867
## 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
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 15
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:------:|:---------------:|
| 6.1786 | 0.41 | 5000 | 6.1475 |
| 6.1856 | 0.81 | 10000 | 6.1490 |
| 6.1769 | 1.22 | 15000 | 6.1370 |
| 6.1714 | 1.62 | 20000 | 6.1330 |
| 6.1633 | 2.03 | 25000 | 6.1221 |
| 6.1548 | 2.44 | 30000 | 6.1180 |
| 6.1495 | 2.84 | 35000 | 6.1141 |
| 6.1453 | 3.25 | 40000 | 6.1026 |
| 6.1362 | 3.66 | 45000 | 6.0984 |
| 6.1325 | 4.06 | 50000 | 6.0961 |
| 6.1227 | 4.47 | 55000 | 6.0874 |
| 6.1215 | 4.87 | 60000 | 6.0806 |
| 6.1149 | 5.28 | 65000 | 6.0779 |
| 6.1099 | 5.69 | 70000 | 6.0701 |
| 6.104 | 6.09 | 75000 | 6.0633 |
| 6.0963 | 6.5 | 80000 | 6.0628 |
| 6.095 | 6.91 | 85000 | 6.0572 |
| 6.0858 | 7.31 | 90000 | 6.0525 |
| 6.0895 | 7.72 | 95000 | 6.0430 |
| 6.0804 | 8.12 | 100000 | 6.0437 |
| 6.0767 | 8.53 | 105000 | 6.0371 |
| 6.0748 | 8.94 | 110000 | 6.0312 |
| 6.0702 | 9.34 | 115000 | 6.0293 |
| 6.0668 | 9.75 | 120000 | 6.0242 |
| 6.0615 | 10.16 | 125000 | 6.0213 |
| 6.0568 | 10.56 | 130000 | 6.0183 |
| 6.0552 | 10.97 | 135000 | 6.0125 |
| 6.0496 | 11.37 | 140000 | 6.0087 |
| 6.0493 | 11.78 | 145000 | 6.0084 |
| 6.0466 | 12.19 | 150000 | 6.0060 |
| 6.042 | 12.59 | 155000 | 6.0008 |
| 6.0375 | 13.0 | 160000 | 5.9986 |
| 6.0345 | 13.41 | 165000 | 5.9940 |
| 6.0336 | 13.81 | 170000 | 5.9905 |
| 6.0334 | 14.22 | 175000 | 5.9891 |
| 6.0313 | 14.62 | 180000 | 5.9887 |
### Framework versions
- Transformers 4.19.2
- Pytorch 1.11.0
- Datasets 2.2.2
- Tokenizers 0.12.1
|
keljai/ppo-LunarLander-v2
|
keljai
| 2022-08-08T17:14:03Z
| 0
| 0
|
stable-baselines3
|
[
"stable-baselines3",
"LunarLander-v2",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-08T17:13:37Z
|
---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- metrics:
- type: mean_reward
value: 227.24 +/- 21.38
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
---
# **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
...
```
|
sofiaoliveira/q-Taxi-v3
|
sofiaoliveira
| 2022-08-08T16:12:21Z
| 0
| 0
| null |
[
"Taxi-v3",
"q-learning",
"reinforcement-learning",
"custom-implementation",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-08T14:35:35Z
|
---
tags:
- Taxi-v3
- q-learning
- reinforcement-learning
- custom-implementation
model-index:
- name: q-Taxi-v3
results:
- metrics:
- type: mean_reward
value: 7.56 +/- 2.71
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: Taxi-v3
type: Taxi-v3
---
# **Q-Learning** Agent playing **Taxi-v3**
This is a trained model of a **Q-Learning** agent playing **Taxi-v3** .
## Usage
```python
model = load_from_hub(repo_id="sofiaoliveira/q-Taxi-v3", filename="q-learning.pkl")
# Don't forget to check if you need to add additional attributes (is_slippery=False etc)
env = gym.make(model["env_id"])
evaluate_agent(env, model["max_steps"], model["n_eval_episodes"], model["qtable"], model["eval_seed"])
```
|
abdulmatinomotoso/multi_news_article_title_12000_2
|
abdulmatinomotoso
| 2022-08-08T15:54:55Z
| 11
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"pegasus",
"text2text-generation",
"generated_from_trainer",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text2text-generation
| 2022-08-08T12:42:58Z
|
---
tags:
- generated_from_trainer
model-index:
- name: multi_news_article_title_12000_2
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. -->
# multi_news_article_title_12000_2
This model is a fine-tuned version of [google/pegasus-multi_news](https://huggingface.co/google/pegasus-multi_news) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1917
## 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: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 16
- 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: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 0.3887 | 0.65 | 500 | 0.2781 |
| 0.2484 | 1.31 | 1000 | 0.2000 |
| 0.2314 | 1.96 | 1500 | 0.1917 |
### Framework versions
- Transformers 4.18.0
- Pytorch 1.11.0
- Datasets 2.1.0
- Tokenizers 0.12.1
|
lauer/distilbert-base-uncased-finetuned-emotion
|
lauer
| 2022-08-08T15:47:05Z
| 4
| 1
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_trainer",
"dataset:emotion",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text-classification
| 2022-08-08T15:14:30Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- emotion
metrics:
- accuracy
- f1
model-index:
- name: distilbert-base-uncased-finetuned-emotion
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: emotion
type: emotion
config: default
split: train
args: default
metrics:
- name: Accuracy
type: accuracy
value: 0.9305
- name: F1
type: f1
value: 0.930580220497549
---
<!-- 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 the emotion dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2133
- Accuracy: 0.9305
- F1: 0.9306
## 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 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.825 | 1.0 | 250 | 0.2957 | 0.909 | 0.9069 |
| 0.2399 | 2.0 | 500 | 0.2133 | 0.9305 | 0.9306 |
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
Mozart-coder/DNA_bert_3-finetuned
|
Mozart-coder
| 2022-08-08T15:45:59Z
| 161
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"bert",
"fill-mask",
"generated_from_trainer",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
fill-mask
| 2022-08-08T15:29:41Z
|
---
tags:
- generated_from_trainer
model-index:
- name: DNA_bert_3-finetuned
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. -->
# DNA_bert_3-finetuned
This model is a fine-tuned version of [armheb/DNA_bert_3](https://huggingface.co/armheb/DNA_bert_3) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5788
## 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: 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: 10
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 0.8244 | 1.0 | 62 | 0.6044 |
| 0.5987 | 2.0 | 124 | 0.5933 |
| 0.5915 | 3.0 | 186 | 0.5856 |
| 0.585 | 4.0 | 248 | 0.5844 |
| 0.5817 | 5.0 | 310 | 0.5818 |
| 0.5791 | 6.0 | 372 | 0.5809 |
| 0.5801 | 7.0 | 434 | 0.5807 |
| 0.5768 | 8.0 | 496 | 0.5796 |
| 0.5741 | 9.0 | 558 | 0.5790 |
| 0.574 | 10.0 | 620 | 0.5788 |
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
MLRS/mBERTu
|
MLRS
| 2022-08-08T15:44:11Z
| 34
| 3
|
transformers
|
[
"transformers",
"pytorch",
"bert",
"fill-mask",
"mt",
"dataset:MLRS/korpus_malti",
"license:cc-by-nc-sa-4.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
fill-mask
| 2022-04-14T13:54:27Z
|
---
language:
- mt
datasets:
- MLRS/korpus_malti
model-index:
- name: mBERTu
results:
- task:
type: dependency-parsing
name: Dependency Parsing
dataset:
type: universal_dependencies
args: mt_mudt
name: Maltese Universal Dependencies Treebank (MUDT)
metrics:
- type: uas
value: 92.10
name: Unlabelled Attachment Score
- type: las
value: 87.87
name: Labelled Attachment Score
- task:
type: part-of-speech-tagging
name: Part-of-Speech Tagging
dataset:
type: mlrs_pos
name: MLRS POS dataset
metrics:
- type: accuracy
value: 98.66
name: UPOS Accuracy
args: upos
- type: accuracy
value: 98.58
name: XPOS Accuracy
args: xpos
- task:
type: named-entity-recognition
name: Named Entity Recognition
dataset:
type: wikiann
name: WikiAnn (Maltese)
args: mt
metrics:
- type: f1
args: span
value: 86.60
name: Span-based F1
- task:
type: sentiment-analysis
name: Sentiment Analysis
dataset:
type: mt-sentiment-analysis
name: Maltese Sentiment Analysis Dataset
metrics:
- type: f1
args: macro
value: 76.79
name: Macro-averaged F1
license: cc-by-nc-sa-4.0
widget:
- text: "Malta huwa pajjiż fl-[MASK]."
---
# mBERTu
A Maltese multilingual model pre-trained on the Korpus Malti v4.0 using multilingual BERT as the initial checkpoint.
## License
This work is licensed under a
[Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License][cc-by-nc-sa].
Permissions beyond the scope of this license may be available at [https://mlrs.research.um.edu.mt/](https://mlrs.research.um.edu.mt/).
[![CC BY-NC-SA 4.0][cc-by-nc-sa-image]][cc-by-nc-sa]
[cc-by-nc-sa]: http://creativecommons.org/licenses/by-nc-sa/4.0/
[cc-by-nc-sa-image]: https://licensebuttons.net/l/by-nc-sa/4.0/88x31.png
## Citation
This work was first presented in [Pre-training Data Quality and Quantity for a Low-Resource Language: New Corpus and BERT Models for Maltese](https://aclanthology.org/2022.deeplo-1.10/).
Cite it as follows:
```bibtex
@inproceedings{BERTu,
title = "Pre-training Data Quality and Quantity for a Low-Resource Language: New Corpus and {BERT} Models for {M}altese",
author = "Micallef, Kurt and
Gatt, Albert and
Tanti, Marc and
van der Plas, Lonneke and
Borg, Claudia",
booktitle = "Proceedings of the Third Workshop on Deep Learning for Low-Resource Natural Language Processing",
month = jul,
year = "2022",
address = "Hybrid",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.deeplo-1.10",
doi = "10.18653/v1/2022.deeplo-1.10",
pages = "90--101",
}
```
|
wenkai-li/finetuned-marktextepoch-n500
|
wenkai-li
| 2022-08-08T15:03:23Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"roberta",
"fill-mask",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
fill-mask
| 2022-08-08T04:03:26Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: finetuned-marktextepoch-n500
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-marktextepoch-n500
This model is a fine-tuned version of [leokai/finetuned-marktextepoch-n200](https://huggingface.co/leokai/finetuned-marktextepoch-n200) on the None dataset.
It achieves the following results on the evaluation set:
- eval_loss: 2.4281
- eval_runtime: 11.4175
- eval_samples_per_second: 279.571
- eval_steps_per_second: 34.946
- epoch: 218.0
- step: 350108
## 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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 300
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
neskue/ppo-LunarLander-v2
|
neskue
| 2022-08-08T14:57:11Z
| 0
| 0
|
stable-baselines3
|
[
"stable-baselines3",
"LunarLander-v2",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-08T14:56:33Z
|
---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- metrics:
- type: mean_reward
value: 27.28 +/- 144.51
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
---
# **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
...
```
|
sofiaoliveira/q-FrozenLake-v1-8x8-noSlippery
|
sofiaoliveira
| 2022-08-08T14:08:49Z
| 0
| 0
| null |
[
"FrozenLake-v1-8x8-no_slippery",
"q-learning",
"reinforcement-learning",
"custom-implementation",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-08T14:08:41Z
|
---
tags:
- FrozenLake-v1-8x8-no_slippery
- q-learning
- reinforcement-learning
- custom-implementation
model-index:
- name: q-FrozenLake-v1-8x8-noSlippery
results:
- metrics:
- type: mean_reward
value: 1.00 +/- 0.00
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: FrozenLake-v1-8x8-no_slippery
type: FrozenLake-v1-8x8-no_slippery
---
# **Q-Learning** Agent playing **FrozenLake-v1**
This is a trained model of a **Q-Learning** agent playing **FrozenLake-v1** .
## Usage
```python
model = load_from_hub(repo_id="sofiaoliveira/q-FrozenLake-v1-8x8-noSlippery", filename="q-learning.pkl")
# Don't forget to check if you need to add additional attributes (is_slippery=False etc)
env = gym.make(model["env_id"])
evaluate_agent(env, model["max_steps"], model["n_eval_episodes"], model["qtable"], model["eval_seed"])
```
|
nvidia/segformer-b0-finetuned-cityscapes-1024-1024
|
nvidia
| 2022-08-08T13:43:30Z
| 5,316
| 7
|
transformers
|
[
"transformers",
"pytorch",
"tf",
"segformer",
"vision",
"image-segmentation",
"dataset:cityscapes",
"arxiv:2105.15203",
"license:other",
"endpoints_compatible",
"region:us"
] |
image-segmentation
| 2022-03-02T23:29:05Z
|
---
license: other
tags:
- vision
- image-segmentation
datasets:
- cityscapes
widget:
- src: https://cdn-media.huggingface.co/Inference-API/Sample-results-on-the-Cityscapes-dataset-The-above-images-show-how-our-method-can-handle.png
example_title: Road
---
# SegFormer (b0-sized) model fine-tuned on CityScapes
SegFormer model fine-tuned on CityScapes at resolution 1024x1024. It was introduced in the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Xie et al. and first released in [this repository](https://github.com/NVlabs/SegFormer).
Disclaimer: The team releasing SegFormer did not write a model card for this model so this model card has been written by the Hugging Face team.
## Model description
SegFormer consists of a hierarchical Transformer encoder and a lightweight all-MLP decode head to achieve great results on semantic segmentation benchmarks such as ADE20K and Cityscapes. The hierarchical Transformer is first pre-trained on ImageNet-1k, after which a decode head is added and fine-tuned altogether on a downstream dataset.
## Intended uses & limitations
You can use the raw model for semantic segmentation. See the [model hub](https://huggingface.co/models?other=segformer) to look for fine-tuned versions on a task that interests you.
### How to use
Here is how to use this model to classify an image of the COCO 2017 dataset into one of the 1,000 ImageNet classes:
```python
from transformers import SegformerFeatureExtractor, SegformerForSemanticSegmentation
from PIL import Image
import requests
feature_extractor = SegformerFeatureExtractor.from_pretrained("nvidia/segformer-b0-finetuned-cityscapes-1024-1024")
model = SegformerForSemanticSegmentation.from_pretrained("nvidia/segformer-b0-finetuned-cityscapes-1024-1024")
url = "http://images.cocodataset.org/val2017/000000039769.jpg"
image = Image.open(requests.get(url, stream=True).raw)
inputs = feature_extractor(images=image, return_tensors="pt")
outputs = model(**inputs)
logits = outputs.logits # shape (batch_size, num_labels, height/4, width/4)
```
For more code examples, we refer to the [documentation](https://huggingface.co/transformers/model_doc/segformer.html#).
### License
The license for this model can be found [here](https://github.com/NVlabs/SegFormer/blob/master/LICENSE).
### BibTeX entry and citation info
```bibtex
@article{DBLP:journals/corr/abs-2105-15203,
author = {Enze Xie and
Wenhai Wang and
Zhiding Yu and
Anima Anandkumar and
Jose M. Alvarez and
Ping Luo},
title = {SegFormer: Simple and Efficient Design for Semantic Segmentation with
Transformers},
journal = {CoRR},
volume = {abs/2105.15203},
year = {2021},
url = {https://arxiv.org/abs/2105.15203},
eprinttype = {arXiv},
eprint = {2105.15203},
timestamp = {Wed, 02 Jun 2021 11:46:42 +0200},
biburl = {https://dblp.org/rec/journals/corr/abs-2105-15203.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
```
|
dminiotas05/distilbert-base-uncased-finetuned-ft1500_norm500_aug1
|
dminiotas05
| 2022-08-08T13:27:41Z
| 105
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text-classification
| 2022-08-08T11:37:51Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: distilbert-base-uncased-finetuned-ft1500_norm500_aug1
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-ft1500_norm500_aug1
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: 2.9086
- Mse: 3.6357
- Mae: 1.0762
- R2: 0.2894
- Accuracy: 0.5170
## 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: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Mse | Mae | R2 | Accuracy |
|:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:------:|:--------:|
| 1.5856 | 1.0 | 5847 | 3.3101 | 4.1376 | 1.1447 | 0.1913 | 0.4965 |
| 0.442 | 2.0 | 11694 | 2.7448 | 3.4311 | 1.0934 | 0.3294 | 0.4523 |
| 0.2703 | 3.0 | 17541 | 2.9300 | 3.6625 | 1.0907 | 0.2841 | 0.4933 |
| 0.1699 | 4.0 | 23388 | 2.7979 | 3.4973 | 1.0808 | 0.3164 | 0.4805 |
| 0.1168 | 5.0 | 29235 | 2.9086 | 3.6357 | 1.0762 | 0.2894 | 0.5170 |
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
Lvxue/distilled-mt5-small-0.5
|
Lvxue
| 2022-08-08T10:41:51Z
| 8
| 0
|
transformers
|
[
"transformers",
"pytorch",
"mt5",
"text2text-generation",
"generated_from_trainer",
"en",
"ro",
"dataset:wmt16",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text2text-generation
| 2022-08-08T08:12:07Z
|
---
language:
- en
- ro
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- wmt16
metrics:
- bleu
model-index:
- name: distilled-mt5-small-0.5
results:
- task:
name: Translation
type: translation
dataset:
name: wmt16 ro-en
type: wmt16
args: ro-en
metrics:
- name: Bleu
type: bleu
value: 1.2575
---
<!-- 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. -->
# distilled-mt5-small-0.5
This model is a distilled version of [Lvxue/finetuned-mt5-base](https://huggingface.co/Lvxue/finetuned-mt5-base) on [google/mt5-small](https://huggingface.co/google/mt5-small) on the wmt16 ro-en dataset.
It achieves the following results on the evaluation set:
- Loss: 3.7455
- Bleu: 1.2575
- Gen Len: 94.3597
## 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: 4
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5.0
### Training results
### Framework versions
- Transformers 4.20.1
- Pytorch 1.12.0+cu102
- Datasets 2.3.2
- Tokenizers 0.12.1
|
mohammadhadiarabi/ddpm-butterflies-128
|
mohammadhadiarabi
| 2022-08-08T10:35:57Z
| 2
| 0
|
diffusers
|
[
"diffusers",
"tensorboard",
"en",
"dataset:huggan/smithsonian_butterflies_subset",
"license:apache-2.0",
"diffusers:DDPMPipeline",
"region:us"
] | null | 2022-08-08T09:22:10Z
|
---
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
# 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: 500
- ema_inv_gamma: None
- ema_inv_gamma: None
- ema_inv_gamma: None
- mixed_precision: fp16
### Training results
📈 [TensorBoard logs](https://huggingface.co/mohammadhadiarabi/ddpm-butterflies-128/tensorboard?#scalars)
|
osanseviero/distilroberta-base-sentence-transformer
|
osanseviero
| 2022-08-08T09:33:52Z
| 1
| 0
|
sentence-transformers
|
[
"sentence-transformers",
"pytorch",
"roberta",
"feature-extraction",
"sentence-similarity",
"transformers",
"dataset:embedding-data/QQP_triplets",
"autotrain_compatible",
"text-embeddings-inference",
"endpoints_compatible",
"region:us"
] |
sentence-similarity
| 2022-08-08T09:33:42Z
|
---
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- feature-extraction
- sentence-similarity
- transformers
datasets:
- embedding-data/QQP_triplets
---
# osanseviero/distilroberta-base-sentence-transformer
This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.
<!--- Describe your model here -->
## Usage (Sentence-Transformers)
Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
```
pip install -U sentence-transformers
```
Then you can use the model like this:
```python
from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]
model = SentenceTransformer('osanseviero/distilroberta-base-sentence-transformer')
embeddings = model.encode(sentences)
print(embeddings)
```
## Usage (HuggingFace Transformers)
Without [sentence-transformers](https://www.SBERT.net), you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.
```python
from transformers import AutoTokenizer, AutoModel
import torch
#Mean Pooling - Take attention mask into account for correct averaging
def mean_pooling(model_output, attention_mask):
token_embeddings = model_output[0] #First element of model_output contains all token embeddings
input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
# Sentences we want sentence embeddings for
sentences = ['This is an example sentence', 'Each sentence is converted']
# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained('osanseviero/distilroberta-base-sentence-transformer')
model = AutoModel.from_pretrained('osanseviero/distilroberta-base-sentence-transformer')
# Tokenize sentences
encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
# Compute token embeddings
with torch.no_grad():
model_output = model(**encoded_input)
# Perform pooling. In this case, mean pooling.
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
print("Sentence embeddings:")
print(sentence_embeddings)
```
## Evaluation Results
<!--- Describe how your model was evaluated -->
For an automated evaluation of this model, see the *Sentence Embeddings Benchmark*: [https://seb.sbert.net](https://seb.sbert.net?model_name=osanseviero/distilroberta-base-sentence-transformer)
## Training
The model was trained with the parameters:
**DataLoader**:
`torch.utils.data.dataloader.DataLoader` of length 63 with parameters:
```
{'batch_size': 16, 'sampler': 'torch.utils.data.sampler.RandomSampler', 'batch_sampler': 'torch.utils.data.sampler.BatchSampler'}
```
**Loss**:
`sentence_transformers.losses.TripletLoss.TripletLoss` with parameters:
```
{'distance_metric': 'TripletDistanceMetric.EUCLIDEAN', 'triplet_margin': 5}
```
Parameters of the fit()-Method:
```
{
"epochs": 10,
"evaluation_steps": 0,
"evaluator": "NoneType",
"max_grad_norm": 1,
"optimizer_class": "<class 'torch.optim.adamw.AdamW'>",
"optimizer_params": {
"lr": 2e-05
},
"scheduler": "WarmupLinear",
"steps_per_epoch": null,
"warmup_steps": 63,
"weight_decay": 0.01
}
```
## Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: RobertaModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
)
```
## Citing & Authors
<!--- Describe where people can find more information -->
|
Lvxue/distilled-mt5-small-0.9
|
Lvxue
| 2022-08-08T09:27:38Z
| 6
| 0
|
transformers
|
[
"transformers",
"pytorch",
"mt5",
"text2text-generation",
"generated_from_trainer",
"en",
"ro",
"dataset:wmt16",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text2text-generation
| 2022-08-08T08:18:23Z
|
---
language:
- en
- ro
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- wmt16
metrics:
- bleu
model-index:
- name: distilled-mt5-small-0.9
results:
- task:
name: Translation
type: translation
dataset:
name: wmt16 ro-en
type: wmt16
args: ro-en
metrics:
- name: Bleu
type: bleu
value: 2.6938
---
<!-- 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. -->
# distilled-mt5-small-0.9
This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on the wmt16 ro-en dataset.
It achieves the following results on the evaluation set:
- Loss: 3.4137
- Bleu: 2.6938
- Gen Len: 69.7484
## 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: 4
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5.0
### Training results
### Framework versions
- Transformers 4.20.1
- Pytorch 1.12.0+cu102
- Datasets 2.3.2
- Tokenizers 0.12.1
|
Mahmoud7/q-Taxi-v3
|
Mahmoud7
| 2022-08-08T09:21:53Z
| 0
| 0
| null |
[
"Taxi-v3",
"q-learning",
"reinforcement-learning",
"custom-implementation",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-08T09:21:45Z
|
---
tags:
- Taxi-v3
- q-learning
- reinforcement-learning
- custom-implementation
model-index:
- name: q-Taxi-v3
results:
- metrics:
- type: mean_reward
value: 7.56 +/- 2.71
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: Taxi-v3
type: Taxi-v3
---
# **Q-Learning** Agent playing **Taxi-v3**
This is a trained model of a **Q-Learning** agent playing **Taxi-v3** .
## Usage
```python
model = load_from_hub(repo_id="Mahmoud7/q-Taxi-v3", filename="q-learning.pkl")
# Don't forget to check if you need to add additional attributes (is_slippery=False etc)
env = gym.make(model["env_id"])
evaluate_agent(env, model["max_steps"], model["n_eval_episodes"], model["qtable"], model["eval_seed"])
```
|
eliwill/distilgpt2-discursive-krishna
|
eliwill
| 2022-08-08T07:56:32Z
| 4
| 0
|
transformers
|
[
"transformers",
"tf",
"tensorboard",
"gpt2",
"text-generation",
"generated_from_keras_callback",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text-generation
| 2022-08-08T07:49:44Z
|
---
license: apache-2.0
tags:
- generated_from_keras_callback
model-index:
- name: eliwill/distilgpt2-discursive-krishna
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. -->
# eliwill/distilgpt2-discursive-krishna
This model is a fine-tuned version of [distilgpt2](https://huggingface.co/distilgpt2) on an unknown dataset.
It achieves the following results on the evaluation set:
- Train Loss: 3.2503
- Validation Loss: 3.1371
- 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: {'name': 'AdamWeightDecay', 'learning_rate': 2e-05, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False, 'weight_decay_rate': 0.01}
- training_precision: float32
### Training results
| Train Loss | Validation Loss | Epoch |
|:----------:|:---------------:|:-----:|
| 3.2503 | 3.1371 | 0 |
### Framework versions
- Transformers 4.21.1
- TensorFlow 2.8.2
- Datasets 2.4.0
- Tokenizers 0.12.1
|
202015004/Spoof_detection
|
202015004
| 2022-08-08T07:48:41Z
| 37
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] |
automatic-speech-recognition
| 2022-08-05T09:32:36Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: Spoof_detection
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. -->
# Spoof_detection
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.7448
- Wer: 0.1090
## 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: 4
- 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: 1000
- num_epochs: 30
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:-----:|:---------------:|:------:|
| 95.9046 | 0.66 | 500 | 992.2993 | 0.6180 |
| 14.0322 | 1.33 | 1000 | 1.8873 | 0.1090 |
| 1.8659 | 1.99 | 1500 | 1.7827 | 0.1090 |
| 1.851 | 2.65 | 2000 | 1.8489 | 0.1090 |
| 1.8218 | 3.32 | 2500 | 1.8943 | 0.1090 |
| 1.8108 | 3.98 | 3000 | 1.9250 | 0.1090 |
| 1.8228 | 4.64 | 3500 | 1.7555 | 0.1090 |
| 1.832 | 5.31 | 4000 | 1.7837 | 0.1090 |
| 1.8403 | 5.97 | 4500 | 1.6644 | 0.1090 |
| 1.8292 | 6.63 | 5000 | 1.6906 | 0.1090 |
| 1.8223 | 7.29 | 5500 | 1.6966 | 0.1090 |
| 1.8007 | 7.96 | 6000 | 1.6951 | 0.1090 |
| 1.7986 | 8.62 | 6500 | 1.7436 | 0.1090 |
| 1.7933 | 9.28 | 7000 | 1.8169 | 0.1090 |
| 1.7861 | 9.95 | 7500 | 1.7209 | 0.1090 |
| 1.7843 | 10.61 | 8000 | 1.9379 | 0.1090 |
| 1.7743 | 11.27 | 8500 | 1.9834 | 0.1090 |
| 1.7721 | 11.94 | 9000 | 1.9279 | 0.1090 |
| 1.7719 | 12.6 | 9500 | 1.8187 | 0.1090 |
| 1.7616 | 13.26 | 10000 | 1.7804 | 0.1090 |
| 1.7638 | 13.93 | 10500 | 1.7884 | 0.1090 |
| 1.7651 | 14.59 | 11000 | 1.7476 | 0.1090 |
| 1.7603 | 15.25 | 11500 | 1.7570 | 0.1090 |
| 1.7543 | 15.92 | 12000 | 1.7356 | 0.1090 |
| 1.7556 | 16.58 | 12500 | 1.7140 | 0.1090 |
| 1.751 | 17.24 | 13000 | 1.7453 | 0.1090 |
| 1.75 | 17.9 | 13500 | 1.7648 | 0.1090 |
| 1.7492 | 18.57 | 14000 | 1.7338 | 0.1090 |
| 1.7484 | 19.23 | 14500 | 1.7093 | 0.1090 |
| 1.7461 | 19.89 | 15000 | 1.7393 | 0.1090 |
| 1.7429 | 20.56 | 15500 | 1.7605 | 0.1090 |
| 1.7446 | 21.22 | 16000 | 1.7782 | 0.1090 |
| 1.7435 | 21.88 | 16500 | 1.6749 | 0.1090 |
| 1.7392 | 22.55 | 17000 | 1.7468 | 0.1090 |
| 1.741 | 23.21 | 17500 | 1.7406 | 0.1090 |
| 1.7394 | 23.87 | 18000 | 1.7787 | 0.1090 |
| 1.739 | 24.54 | 18500 | 1.7969 | 0.1090 |
| 1.7341 | 25.2 | 19000 | 1.7490 | 0.1090 |
| 1.7371 | 25.86 | 19500 | 1.7783 | 0.1090 |
| 1.735 | 26.53 | 20000 | 1.7540 | 0.1090 |
| 1.7353 | 27.19 | 20500 | 1.7735 | 0.1090 |
| 1.7331 | 27.85 | 21000 | 1.7188 | 0.1090 |
| 1.7308 | 28.51 | 21500 | 1.7349 | 0.1090 |
| 1.7341 | 29.18 | 22000 | 1.7531 | 0.1090 |
| 1.7305 | 29.84 | 22500 | 1.7448 | 0.1090 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu102
- Datasets 1.16.1
- Tokenizers 0.12.1
|
amartyobanerjee/mt5-small-finetuned-amazon-en-es
|
amartyobanerjee
| 2022-08-08T07:43:46Z
| 8
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"mt5",
"text2text-generation",
"summarization",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
summarization
| 2022-08-08T06:18:06Z
|
---
license: apache-2.0
tags:
- summarization
- generated_from_trainer
metrics:
- rouge
model-index:
- name: mt5-small-finetuned-amazon-en-es
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. -->
# mt5-small-finetuned-amazon-en-es
This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 3.0294
- Rouge1: 16.497
- Rouge2: 8.0618
- Rougel: 16.2979
- Rougelsum: 16.1465
## 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: 5.6e-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
- num_epochs: 8
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:------:|:-------:|:---------:|
| 6.5928 | 1.0 | 1209 | 3.3005 | 14.7843 | 6.5518 | 14.2805 | 14.2951 |
| 3.9024 | 2.0 | 2418 | 3.1399 | 16.8202 | 8.6739 | 16.1194 | 16.0844 |
| 3.5806 | 3.0 | 3627 | 3.0869 | 18.1223 | 9.3051 | 17.7533 | 17.7254 |
| 3.4201 | 4.0 | 4836 | 3.0590 | 17.654 | 9.0154 | 17.1853 | 17.1769 |
| 3.3202 | 5.0 | 6045 | 3.0598 | 17.612 | 8.6707 | 17.4662 | 17.2963 |
| 3.2436 | 6.0 | 7254 | 3.0409 | 16.7938 | 8.3054 | 16.6141 | 16.4853 |
| 3.2079 | 7.0 | 8463 | 3.0332 | 16.7246 | 8.2362 | 16.5065 | 16.3611 |
| 3.1801 | 8.0 | 9672 | 3.0294 | 16.497 | 8.0618 | 16.2979 | 16.1465 |
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
naveenkb/ppo-LunarLander-v2
|
naveenkb
| 2022-08-08T07:35:10Z
| 0
| 0
|
stable-baselines3
|
[
"stable-baselines3",
"LunarLander-v2",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-08T07:34:35Z
|
---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- metrics:
- type: mean_reward
value: 193.11 +/- 17.14
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
---
# **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
...
```
|
wenkai-li/finetuned-marktextepoch-n200
|
wenkai-li
| 2022-08-08T03:37:00Z
| 161
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"roberta",
"fill-mask",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
fill-mask
| 2022-08-07T17:29:25Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: finetuned-marktextepoch-n200
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-marktextepoch-n200
This model is a fine-tuned version of [distilroberta-base](https://huggingface.co/distilroberta-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 2.0880
## 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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 200
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:------:|:---------------:|
| 2.5742 | 1.0 | 1606 | 2.4071 |
| 2.4441 | 2.0 | 3212 | 2.2715 |
| 2.3699 | 3.0 | 4818 | 2.2896 |
| 2.3074 | 4.0 | 6424 | 2.2295 |
| 2.2667 | 5.0 | 8030 | 2.2147 |
| 2.2376 | 6.0 | 9636 | 2.1886 |
| 2.2161 | 7.0 | 11242 | 2.1816 |
| 2.1611 | 8.0 | 12848 | 2.1690 |
| 2.1418 | 9.0 | 14454 | 2.1541 |
| 2.1198 | 10.0 | 16060 | 2.1355 |
| 2.1033 | 11.0 | 17666 | 2.1132 |
| 2.0738 | 12.0 | 19272 | 2.1441 |
| 2.0581 | 13.0 | 20878 | 2.1068 |
| 2.0555 | 14.0 | 22484 | 2.1035 |
| 2.0375 | 15.0 | 24090 | 2.1000 |
| 2.0071 | 16.0 | 25696 | 2.1084 |
| 1.9942 | 17.0 | 27302 | 2.0711 |
| 1.9554 | 18.0 | 28908 | 2.0978 |
| 1.9469 | 19.0 | 30514 | 2.0705 |
| 1.9414 | 20.0 | 32120 | 2.0597 |
| 1.9331 | 21.0 | 33726 | 2.0782 |
| 1.9132 | 22.0 | 35332 | 2.0622 |
| 1.9003 | 23.0 | 36938 | 2.0426 |
| 1.9019 | 24.0 | 38544 | 2.0562 |
| 1.8733 | 25.0 | 40150 | 2.0419 |
| 1.8556 | 26.0 | 41756 | 2.0572 |
| 1.8399 | 27.0 | 43362 | 2.0453 |
| 1.8332 | 28.0 | 44968 | 2.0517 |
| 1.8296 | 29.0 | 46574 | 2.0580 |
| 1.788 | 30.0 | 48180 | 2.0454 |
| 1.7944 | 31.0 | 49786 | 2.0193 |
| 1.7716 | 32.0 | 51392 | 2.0595 |
| 1.7799 | 33.0 | 52998 | 2.0379 |
| 1.7633 | 34.0 | 54604 | 2.0392 |
| 1.7477 | 35.0 | 56210 | 2.0122 |
| 1.7407 | 36.0 | 57816 | 2.0293 |
| 1.7163 | 37.0 | 59422 | 2.0339 |
| 1.72 | 38.0 | 61028 | 1.9987 |
| 1.729 | 39.0 | 62634 | 2.0135 |
| 1.709 | 40.0 | 64240 | 2.0455 |
| 1.7019 | 41.0 | 65846 | 2.0206 |
| 1.6958 | 42.0 | 67452 | 2.0408 |
| 1.6789 | 43.0 | 69058 | 2.0470 |
| 1.6907 | 44.0 | 70664 | 2.0280 |
| 1.6531 | 45.0 | 72270 | 2.0514 |
| 1.6563 | 46.0 | 73876 | 2.0428 |
| 1.6364 | 47.0 | 75482 | 2.0305 |
| 1.6534 | 48.0 | 77088 | 2.0200 |
| 1.6312 | 49.0 | 78694 | 2.0444 |
| 1.6092 | 50.0 | 80300 | 2.0154 |
| 1.5998 | 51.0 | 81906 | 2.0249 |
| 1.5808 | 52.0 | 83512 | 2.0235 |
| 1.5945 | 53.0 | 85118 | 2.0286 |
| 1.6004 | 54.0 | 86724 | 2.0288 |
| 1.5802 | 55.0 | 88330 | 2.0346 |
| 1.5665 | 56.0 | 89936 | 2.0120 |
| 1.5723 | 57.0 | 91542 | 2.0257 |
| 1.5553 | 58.0 | 93148 | 2.0146 |
| 1.5445 | 59.0 | 94754 | 2.0333 |
| 1.5669 | 60.0 | 96360 | 2.0325 |
| 1.5318 | 61.0 | 97966 | 2.0250 |
| 1.5117 | 62.0 | 99572 | 2.0343 |
| 1.5248 | 63.0 | 101178 | 2.0183 |
| 1.5149 | 64.0 | 102784 | 2.0422 |
| 1.5087 | 65.0 | 104390 | 2.0236 |
| 1.5087 | 66.0 | 105996 | 2.0275 |
| 1.4938 | 67.0 | 107602 | 2.0384 |
| 1.5008 | 68.0 | 109208 | 2.0167 |
| 1.4871 | 69.0 | 110814 | 2.0456 |
| 1.4931 | 70.0 | 112420 | 2.0083 |
| 1.467 | 71.0 | 114026 | 2.0313 |
| 1.4519 | 72.0 | 115632 | 2.0254 |
| 1.448 | 73.0 | 117238 | 2.0289 |
| 1.4475 | 74.0 | 118844 | 2.0051 |
| 1.4522 | 75.0 | 120450 | 2.0378 |
| 1.4508 | 76.0 | 122056 | 2.0612 |
| 1.4428 | 77.0 | 123662 | 2.0479 |
| 1.4496 | 78.0 | 125268 | 2.0082 |
| 1.4305 | 79.0 | 126874 | 2.0376 |
| 1.4072 | 80.0 | 128480 | 2.0294 |
| 1.4148 | 81.0 | 130086 | 2.0565 |
| 1.4078 | 82.0 | 131692 | 2.0309 |
| 1.3931 | 83.0 | 133298 | 2.0371 |
| 1.4038 | 84.0 | 134904 | 2.0318 |
| 1.3893 | 85.0 | 136510 | 2.0413 |
| 1.3862 | 86.0 | 138116 | 2.0503 |
| 1.3782 | 87.0 | 139722 | 2.0182 |
| 1.3757 | 88.0 | 141328 | 2.0253 |
| 1.3879 | 89.0 | 142934 | 2.0357 |
| 1.3768 | 90.0 | 144540 | 2.0405 |
| 1.3494 | 91.0 | 146146 | 2.0495 |
| 1.3492 | 92.0 | 147752 | 2.0586 |
| 1.353 | 93.0 | 149358 | 2.0779 |
| 1.3397 | 94.0 | 150964 | 2.0564 |
| 1.3486 | 95.0 | 152570 | 2.0459 |
| 1.3262 | 96.0 | 154176 | 2.0692 |
| 1.349 | 97.0 | 155782 | 2.0765 |
| 1.3228 | 98.0 | 157388 | 2.0443 |
| 1.3291 | 99.0 | 158994 | 2.0617 |
| 1.3211 | 100.0 | 160600 | 2.0552 |
| 1.3344 | 101.0 | 162206 | 2.0626 |
| 1.307 | 102.0 | 163812 | 2.0492 |
| 1.2968 | 103.0 | 165418 | 2.0461 |
| 1.2919 | 104.0 | 167024 | 2.0725 |
| 1.3004 | 105.0 | 168630 | 2.0424 |
| 1.303 | 106.0 | 170236 | 2.0484 |
| 1.2847 | 107.0 | 171842 | 2.0083 |
| 1.2861 | 108.0 | 173448 | 2.0491 |
| 1.2763 | 109.0 | 175054 | 2.0505 |
| 1.2852 | 110.0 | 176660 | 2.0691 |
| 1.2611 | 111.0 | 178266 | 2.0711 |
| 1.2739 | 112.0 | 179872 | 2.0730 |
| 1.278 | 113.0 | 181478 | 2.0551 |
| 1.2581 | 114.0 | 183084 | 2.0554 |
| 1.2532 | 115.0 | 184690 | 2.0513 |
| 1.2322 | 116.0 | 186296 | 2.0292 |
| 1.2774 | 117.0 | 187902 | 2.0409 |
| 1.242 | 118.0 | 189508 | 2.0517 |
| 1.2476 | 119.0 | 191114 | 2.0612 |
| 1.2314 | 120.0 | 192720 | 2.0795 |
| 1.2379 | 121.0 | 194326 | 2.0679 |
| 1.2291 | 122.0 | 195932 | 2.0472 |
| 1.2515 | 123.0 | 197538 | 2.0829 |
| 1.2467 | 124.0 | 199144 | 2.0662 |
| 1.2437 | 125.0 | 200750 | 2.0962 |
| 1.2373 | 126.0 | 202356 | 2.0692 |
| 1.2099 | 127.0 | 203962 | 2.0688 |
| 1.1911 | 128.0 | 205568 | 2.0803 |
| 1.2311 | 129.0 | 207174 | 2.0765 |
| 1.2095 | 130.0 | 208780 | 2.0697 |
| 1.2093 | 131.0 | 210386 | 2.0507 |
| 1.2065 | 132.0 | 211992 | 2.0658 |
| 1.1964 | 133.0 | 213598 | 2.0542 |
| 1.2085 | 134.0 | 215204 | 2.0722 |
| 1.1871 | 135.0 | 216810 | 2.0806 |
| 1.1863 | 136.0 | 218416 | 2.0691 |
| 1.1763 | 137.0 | 220022 | 2.0869 |
| 1.1816 | 138.0 | 221628 | 2.0780 |
| 1.1854 | 139.0 | 223234 | 2.0462 |
| 1.1902 | 140.0 | 224840 | 2.0880 |
| 1.1762 | 141.0 | 226446 | 2.0682 |
| 1.1551 | 142.0 | 228052 | 2.0837 |
| 1.171 | 143.0 | 229658 | 2.1028 |
| 1.1571 | 144.0 | 231264 | 2.0726 |
| 1.1627 | 145.0 | 232870 | 2.0863 |
| 1.1537 | 146.0 | 234476 | 2.0857 |
| 1.1695 | 147.0 | 236082 | 2.0620 |
| 1.1477 | 148.0 | 237688 | 2.0817 |
| 1.1592 | 149.0 | 239294 | 2.0705 |
| 1.1478 | 150.0 | 240900 | 2.0841 |
| 1.1398 | 151.0 | 242506 | 2.0886 |
| 1.144 | 152.0 | 244112 | 2.0673 |
| 1.1646 | 153.0 | 245718 | 2.0620 |
| 1.12 | 154.0 | 247324 | 2.0821 |
| 1.1419 | 155.0 | 248930 | 2.0632 |
| 1.1436 | 156.0 | 250536 | 2.0817 |
| 1.1365 | 157.0 | 252142 | 2.0663 |
| 1.1318 | 158.0 | 253748 | 2.0796 |
| 1.1219 | 159.0 | 255354 | 2.0825 |
| 1.1306 | 160.0 | 256960 | 2.0837 |
| 1.1295 | 161.0 | 258566 | 2.0564 |
| 1.1261 | 162.0 | 260172 | 2.0722 |
| 1.1273 | 163.0 | 261778 | 2.1058 |
| 1.1143 | 164.0 | 263384 | 2.0963 |
| 1.1276 | 165.0 | 264990 | 2.0948 |
| 1.1238 | 166.0 | 266596 | 2.0695 |
| 1.1222 | 167.0 | 268202 | 2.0801 |
| 1.1145 | 168.0 | 269808 | 2.0768 |
| 1.1093 | 169.0 | 271414 | 2.0664 |
| 1.1141 | 170.0 | 273020 | 2.0903 |
| 1.0936 | 171.0 | 274626 | 2.1012 |
| 1.1048 | 172.0 | 276232 | 2.1033 |
| 1.0991 | 173.0 | 277838 | 2.0761 |
| 1.1164 | 174.0 | 279444 | 2.0689 |
| 1.0935 | 175.0 | 281050 | 2.0754 |
| 1.1032 | 176.0 | 282656 | 2.0810 |
| 1.1124 | 177.0 | 284262 | 2.0790 |
| 1.1107 | 178.0 | 285868 | 2.0762 |
| 1.085 | 179.0 | 287474 | 2.0697 |
| 1.093 | 180.0 | 289080 | 2.0856 |
| 1.1034 | 181.0 | 290686 | 2.0734 |
| 1.0983 | 182.0 | 292292 | 2.0837 |
| 1.0972 | 183.0 | 293898 | 2.1063 |
| 1.0909 | 184.0 | 295504 | 2.0873 |
| 1.0805 | 185.0 | 297110 | 2.0888 |
| 1.0893 | 186.0 | 298716 | 2.0498 |
| 1.096 | 187.0 | 300322 | 2.0906 |
| 1.0781 | 188.0 | 301928 | 2.0905 |
| 1.0981 | 189.0 | 303534 | 2.0767 |
| 1.093 | 190.0 | 305140 | 2.0695 |
| 1.0814 | 191.0 | 306746 | 2.0763 |
| 1.0862 | 192.0 | 308352 | 2.0890 |
| 1.0833 | 193.0 | 309958 | 2.1026 |
| 1.0806 | 194.0 | 311564 | 2.0978 |
| 1.0834 | 195.0 | 313170 | 2.1004 |
| 1.0807 | 196.0 | 314776 | 2.0953 |
| 1.0827 | 197.0 | 316382 | 2.1129 |
| 1.0826 | 198.0 | 317988 | 2.1069 |
| 1.0796 | 199.0 | 319594 | 2.0867 |
| 1.0881 | 200.0 | 321200 | 2.0880 |
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
jjjjjjjjjj/Reinforce-PixelCopterV1
|
jjjjjjjjjj
| 2022-08-08T03:32:35Z
| 0
| 0
| null |
[
"Pixelcopter-PLE-v0",
"reinforce",
"reinforcement-learning",
"custom-implementation",
"deep-rl-class",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-08T03:22:44Z
|
---
tags:
- Pixelcopter-PLE-v0
- reinforce
- reinforcement-learning
- custom-implementation
- deep-rl-class
model-index:
- name: Reinforce-PixelCopterV1
results:
- metrics:
- type: mean_reward
value: 18.50 +/- 12.47
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: Pixelcopter-PLE-v0
type: Pixelcopter-PLE-v0
---
# **Reinforce** Agent playing **Pixelcopter-PLE-v0**
This is a trained model of a **Reinforce** agent playing **Pixelcopter-PLE-v0** .
To learn to use this model and train yours check Unit 5 of the Deep Reinforcement Learning Class: https://github.com/huggingface/deep-rl-class/tree/main/unit5
|
abdulmatinomotoso/multi_news_article_title_25000_1
|
abdulmatinomotoso
| 2022-08-08T02:32:22Z
| 10
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"pegasus",
"text2text-generation",
"generated_from_trainer",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text2text-generation
| 2022-08-07T23:09:49Z
|
---
tags:
- generated_from_trainer
model-index:
- name: multi_news_article_title_25000_1
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. -->
# multi_news_article_title_25000_1
This model is a fine-tuned version of [google/pegasus-multi_news](https://huggingface.co/google/pegasus-multi_news) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1973
## 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: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 16
- 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: 1
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 0.3431 | 0.32 | 500 | 0.2731 |
| 0.2171 | 0.64 | 1000 | 0.2056 |
| 0.2211 | 0.96 | 1500 | 0.1973 |
### Framework versions
- Transformers 4.18.0
- Pytorch 1.11.0
- Datasets 2.1.0
- Tokenizers 0.12.1
|
ariesutiono/scibert-lm-v2-finetuned-20
|
ariesutiono
| 2022-08-08T02:13:38Z
| 5
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"bert",
"fill-mask",
"generated_from_trainer",
"dataset:conll2003",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
fill-mask
| 2022-08-08T01:28:54Z
|
---
tags:
- generated_from_trainer
datasets:
- conll2003
model-index:
- name: scibert-lm-v2-finetuned-20
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. -->
# scibert-lm-v2-finetuned-20
This model is a fine-tuned version of [allenai/scibert_scivocab_cased](https://huggingface.co/allenai/scibert_scivocab_cased) on the conll2003 dataset.
It achieves the following results on the evaluation set:
- Loss: 15.7952
## 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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 0.0211 | 1.0 | 878 | 15.1971 |
| 0.0001 | 2.0 | 1756 | 16.8774 |
| 0.0001 | 3.0 | 2634 | 15.7952 |
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
jjjjjjjjjj/testpyramidsrnd
|
jjjjjjjjjj
| 2022-08-08T02:10:40Z
| 2
| 0
|
ml-agents
|
[
"ml-agents",
"tensorboard",
"onnx",
"unity-ml-agents",
"deep-reinforcement-learning",
"reinforcement-learning",
"ML-Agents-Pyramids",
"region:us"
] |
reinforcement-learning
| 2022-08-08T02:07:17Z
|
---
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: jjjjjjjjjj/testpyramidsrnd
3. Step 2: Select your *.nn /*.onnx file
4. Click on Watch the agent play 👀
|
yuewu/toc_titler
|
yuewu
| 2022-08-08T01:55:41Z
| 23
| 0
|
transformers
|
[
"transformers",
"pytorch",
"vision-encoder-decoder",
"image-text-to-text",
"image-to-text",
"image-captioning",
"license:mit",
"endpoints_compatible",
"region:us"
] |
image-to-text
| 2022-08-08T01:11:08Z
|
---
license: mit
tags:
- image-to-text
- image-captioning
---
A model that inputs chemistry journal article table of contents (ToC) images and generates appropriate titles. Trained on all JACS ToCs and titles.
|
ultra-coder54732/comment-detection-prop-16
|
ultra-coder54732
| 2022-08-08T00:29:14Z
| 7
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"bert",
"text-classification",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text-classification
| 2022-08-07T05:37:42Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: comment-detection-prop-16
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. -->
# comment-detection-prop-16
This model is a fine-tuned version of [bert-base-cased](https://huggingface.co/bert-base-cased) 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: 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
- num_epochs: 20
### Training results
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
srcocotero/bert-qa-es
|
srcocotero
| 2022-08-07T21:19:06Z
| 12
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"bert",
"question-answering",
"generated_from_trainer",
"dataset:squad_es",
"endpoints_compatible",
"region:us"
] |
question-answering
| 2022-08-07T18:16:10Z
|
---
tags:
- generated_from_trainer
datasets:
- squad_es
model-index:
- name: bert-qa-es
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-qa-es
This model is a fine-tuned version of [dccuchile/bert-base-spanish-wwm-uncased](https://huggingface.co/dccuchile/bert-base-spanish-wwm-uncased) on the squad_es 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: 3e-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
- num_epochs: 2.0
### Training results
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
paola-md/RELEXset-MLM
|
paola-md
| 2022-08-07T20:07:52Z
| 3
| 0
|
transformers
|
[
"transformers",
"pytorch",
"roberta",
"fill-mask",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
fill-mask
| 2022-08-07T13:14:24Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: recipe-distilroberta-Is
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-distilroberta-Is
This model is a fine-tuned version of [distilroberta-base](https://huggingface.co/distilroberta-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 4.7427
## 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
- num_epochs: 25
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:-----:|:---------------:|
| 19.6191 | 1.0 | 2135 | 10.5217 |
| 8.6838 | 2.0 | 4270 | 7.3017 |
| 6.884 | 3.0 | 6405 | 6.4445 |
| 6.2953 | 4.0 | 8540 | 6.0610 |
| 6.0205 | 5.0 | 10675 | 5.9047 |
| 5.851 | 6.0 | 12810 | 5.7790 |
| 5.7464 | 7.0 | 14945 | 5.7164 |
| 5.6684 | 8.0 | 17080 | 5.6415 |
| 5.6138 | 9.0 | 19215 | 5.5671 |
| 5.5638 | 10.0 | 21350 | 5.5360 |
| 5.5288 | 11.0 | 23485 | 5.5069 |
| 5.4968 | 12.0 | 25620 | 5.4968 |
| 5.4696 | 13.0 | 27755 | 5.4539 |
| 5.4468 | 14.0 | 29890 | 5.4416 |
| 5.4177 | 15.0 | 32025 | 5.3722 |
| 5.3717 | 16.0 | 34160 | 5.3226 |
| 5.317 | 17.0 | 36295 | 5.2197 |
| 5.2367 | 18.0 | 38430 | 5.0888 |
| 5.1543 | 19.0 | 40565 | 4.9954 |
| 5.0919 | 20.0 | 42700 | 4.9306 |
| 5.038 | 21.0 | 44835 | 4.8657 |
| 4.9983 | 22.0 | 46970 | 4.8137 |
| 4.9639 | 23.0 | 49105 | 4.7704 |
| 4.9426 | 24.0 | 51240 | 4.7486 |
| 4.9312 | 25.0 | 53375 | 4.7427 |
### Framework versions
- Transformers 4.19.0.dev0
- Pytorch 1.9.0+cu111
- Datasets 2.4.0
- Tokenizers 0.12.1
|
theicfire/ppo-LunarLander-v2
|
theicfire
| 2022-08-07T19:48:16Z
| 1
| 0
|
stable-baselines3
|
[
"stable-baselines3",
"LunarLander-v2",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-07T19:47:53Z
|
---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- metrics:
- type: mean_reward
value: 245.21 +/- 40.58
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
---
# **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
...
```
|
ycchen/TrOCR-base-ver021-v1
|
ycchen
| 2022-08-07T19:23:45Z
| 45
| 0
|
transformers
|
[
"transformers",
"pytorch",
"vision-encoder-decoder",
"image-text-to-text",
"endpoints_compatible",
"region:us"
] |
image-text-to-text
| 2022-08-07T18:49:55Z
|
### How to use
Here is how to use this model in PyTorch:
```python
from transformers import TrOCRProcessor, VisionEncoderDecoderModel
from PIL import Image
import requests
# load image from the IAM database (actually this model is meant to be used on printed text)
url = 'https://fki.tic.heia-fr.ch/static/img/a01-122-02-00.jpg'
image = Image.open(requests.get(url, stream=True).raw).convert("RGB")
processor = TrOCRProcessor.from_pretrained('ycchen/TrOCR-base-ver021-v1')
model = VisionEncoderDecoderModel.from_pretrained('ycchen/TrOCR-base-ver021-v1')
pixel_values = processor(images=image, return_tensors="pt").pixel_values
generated_ids = model.generate(pixel_values)
generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
```
|
cataluna84/xlm-roberta-base-finetuned-panx-en
|
cataluna84
| 2022-08-07T18:15:27Z
| 8
| 0
|
transformers
|
[
"transformers",
"pytorch",
"xlm-roberta",
"token-classification",
"generated_from_trainer",
"dataset:xtreme",
"license:mit",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
token-classification
| 2022-08-07T17:56:51Z
|
---
license: mit
tags:
- generated_from_trainer
datasets:
- xtreme
metrics:
- f1
model-index:
- name: xlm-roberta-base-finetuned-panx-en
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: xtreme
type: xtreme
args: PAN-X.en
metrics:
- name: F1
type: f1
value: 0.6886160714285715
---
<!-- 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. -->
# xlm-roberta-base-finetuned-panx-en
This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the xtreme dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4043
- F1: 0.6886
## 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: 24
- eval_batch_size: 24
- 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 | F1 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 1.1347 | 1.0 | 50 | 0.5771 | 0.4880 |
| 0.5066 | 2.0 | 100 | 0.4209 | 0.6582 |
| 0.3631 | 3.0 | 150 | 0.4043 | 0.6886 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.12.0+cu113
- Datasets 1.16.1
- Tokenizers 0.10.3
|
ultra-coder54732/stance-detection-prop-16
|
ultra-coder54732
| 2022-08-07T17:32:24Z
| 5
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"bert",
"text-classification",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text-classification
| 2022-08-07T06:06:49Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: stance-detection-prop-16
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. -->
# stance-detection-prop-16
This model is a fine-tuned version of [bert-base-cased](https://huggingface.co/bert-base-cased) 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: 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
- num_epochs: 20
### Training results
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
ai4bharat/IndicBART
|
ai4bharat
| 2022-08-07T17:12:33Z
| 1,663
| 29
|
transformers
|
[
"transformers",
"pytorch",
"mbart",
"text2text-generation",
"multilingual",
"nlp",
"indicnlp",
"as",
"bn",
"gu",
"hi",
"kn",
"ml",
"mr",
"or",
"pa",
"ta",
"te",
"arxiv:2109.02903",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text2text-generation
| 2022-03-02T23:29:05Z
|
---
language:
- as
- bn
- gu
- hi
- kn
- ml
- mr
- or
- pa
- ta
- te
tags:
- multilingual
- nlp
- indicnlp
---
IndicBART is a multilingual, sequence-to-sequence pre-trained model focusing on Indic languages and English. It currently supports 11 Indian languages and is based on the mBART architecture. You can use IndicBART model to build natural language generation applications for Indian languages by finetuning the model with supervised training data for tasks like machine translation, summarization, question generation, etc. Some salient features of the IndicBART are:
<ul>
<li >Supported languages: Assamese, Bengali, Gujarati, Hindi, Marathi, Odiya, Punjabi, Kannada, Malayalam, Tamil, Telugu and English. Not all of these languages are supported by mBART50 and mT5. </li>
<li >The model is much smaller than the mBART and mT5(-base) models, so less computationally expensive for finetuning and decoding. </li>
<li> Trained on large Indic language corpora (452 million sentences and 9 billion tokens) which also includes Indian English content. </li>
<li> All languages, except English, have been represented in Devanagari script to encourage transfer learning among the related languages. </li>
</ul>
You can read more about IndicBART in this <a href="https://arxiv.org/abs/2109.02903">paper</a>.
For detailed documentation, look here: https://github.com/AI4Bharat/indic-bart/ and https://indicnlp.ai4bharat.org/indic-bart/
# Pre-training corpus
We used the <a href="https://indicnlp.ai4bharat.org/corpora/">IndicCorp</a> data spanning 12 languages with 452 million sentences (9 billion tokens). The model was trained using the text-infilling objective used in mBART.
# Usage:
```
from transformers import MBartForConditionalGeneration, AutoModelForSeq2SeqLM
from transformers import AlbertTokenizer, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("ai4bharat/IndicBART", do_lower_case=False, use_fast=False, keep_accents=True)
# Or use tokenizer = AlbertTokenizer.from_pretrained("ai4bharat/IndicBART", do_lower_case=False, use_fast=False, keep_accents=True)
model = AutoModelForSeq2SeqLM.from_pretrained("ai4bharat/IndicBART")
# Or use model = MBartForConditionalGeneration.from_pretrained("ai4bharat/IndicBART")
# Some initial mapping
bos_id = tokenizer._convert_token_to_id_with_added_voc("<s>")
eos_id = tokenizer._convert_token_to_id_with_added_voc("</s>")
pad_id = tokenizer._convert_token_to_id_with_added_voc("<pad>")
# To get lang_id use any of ['<2as>', '<2bn>', '<2en>', '<2gu>', '<2hi>', '<2kn>', '<2ml>', '<2mr>', '<2or>', '<2pa>', '<2ta>', '<2te>']
# First tokenize the input and outputs. The format below is how IndicBART was trained so the input should be "Sentence </s> <2xx>" where xx is the language code. Similarly, the output should be "<2yy> Sentence </s>".
inp = tokenizer("I am a boy </s> <2en>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids # tensor([[ 466, 1981, 80, 25573, 64001, 64004]])
out = tokenizer("<2hi> मैं एक लड़का हूँ </s>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids # tensor([[64006, 942, 43, 32720, 8384, 64001]])
# Note that if you use any language other than Hindi or Marathi, you should convert its script to Devanagari using the Indic NLP Library.
model_outputs=model(input_ids=inp, decoder_input_ids=out[:,0:-1], labels=out[:,1:])
# For loss
model_outputs.loss ## This is not label smoothed.
# For logits
model_outputs.logits
# For generation. Pardon the messiness. Note the decoder_start_token_id.
model.eval() # Set dropouts to zero
model_output=model.generate(inp, use_cache=True, num_beams=4, max_length=20, min_length=1, early_stopping=True, pad_token_id=pad_id, bos_token_id=bos_id, eos_token_id=eos_id, decoder_start_token_id=tokenizer._convert_token_to_id_with_added_voc("<2en>"))
# Decode to get output strings
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
print(decoded_output) # I am a boy
# Note that if your output language is not Hindi or Marathi, you should convert its script from Devanagari to the desired language using the Indic NLP Library.
# What if we mask?
inp = tokenizer("I am [MASK] </s> <2en>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids
model_output=model.generate(inp, use_cache=True, num_beams=4, max_length=20, min_length=1, early_stopping=True, pad_token_id=pad_id, bos_token_id=bos_id, eos_token_id=eos_id, decoder_start_token_id=tokenizer._convert_token_to_id_with_added_voc("<2en>"))
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
print(decoded_output) # I am happy
inp = tokenizer("मैं [MASK] हूँ </s> <2hi>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids
model_output=model.generate(inp, use_cache=True, num_beams=4, max_length=20, min_length=1, early_stopping=True, pad_token_id=pad_id, bos_token_id=bos_id, eos_token_id=eos_id, decoder_start_token_id=tokenizer._convert_token_to_id_with_added_voc("<2en>"))
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
print(decoded_output) # मैं जानता हूँ
inp = tokenizer("मला [MASK] पाहिजे </s> <2mr>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids
model_output=model.generate(inp, use_cache=True, num_beams=4, max_length=20, min_length=1, early_stopping=True, pad_token_id=pad_id, bos_token_id=bos_id, eos_token_id=eos_id, decoder_start_token_id=tokenizer._convert_token_to_id_with_added_voc("<2en>"))
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
print(decoded_output) # मला ओळखलं पाहिजे
```
# Notes:
1. This is compatible with the latest version of transformers but was developed with version 4.3.2 so consider using 4.3.2 if possible.
2. While I have only shown how to get logits and loss and how to generate outputs, you can do pretty much everything the MBartForConditionalGeneration class can do as in https://huggingface.co/docs/transformers/model_doc/mbart#transformers.MBartForConditionalGeneration
3. Note that the tokenizer I have used is based on sentencepiece and not BPE. Therefore, I used the AlbertTokenizer class and not the MBartTokenizer class.
4. If you wish to use any language written in a non-Devanagari script (except English), then you should first convert it to Devanagari using the <a href="https://github.com/anoopkunchukuttan/indic_nlp_library">Indic NLP Library</a>. After you get the output, you should convert it back into the original script.
# Fine-tuning on a downstream task
1. If you wish to fine-tune this model, then you can do so using the <a href="https://github.com/prajdabre/yanmtt">YANMTT</a> toolkit, following the instructions <a href="https://github.com/AI4Bharat/indic-bart ">here</a>.
2. (Untested) Alternatively, you may use the official huggingface scripts for <a href="https://github.com/huggingface/transformers/tree/master/examples/pytorch/translation">translation</a> and <a href="https://github.com/huggingface/transformers/tree/master/examples/pytorch/summarization">summarization</a>.
# Contributors
<ul>
<li> Raj Dabre </li>
<li> Himani Shrotriya </li>
<li> Anoop Kunchukuttan </li>
<li> Ratish Puduppully </li>
<li> Mitesh M. Khapra </li>
<li> Pratyush Kumar </li>
</ul>
# Paper
If you use IndicBART, please cite the following paper:
```
@misc{dabre2021indicbart,
title={IndicBART: A Pre-trained Model for Natural Language Generation of Indic Languages},
author={Raj Dabre and Himani Shrotriya and Anoop Kunchukuttan and Ratish Puduppully and Mitesh M. Khapra and Pratyush Kumar},
year={2021},
eprint={2109.02903},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
# License
The model is available under the MIT License.
|
abcp4/a2c-CartPole-v1
|
abcp4
| 2022-08-07T16:42:18Z
| 1
| 0
|
stable-baselines3
|
[
"stable-baselines3",
"CartPole-v1",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-07T02:07:19Z
|
---
library_name: stable-baselines3
tags:
- CartPole-v1
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: A2C
results:
- metrics:
- type: mean_reward
value: 16.90 +/- 6.92
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: CartPole-v1
type: CartPole-v1
---
# **A2C** Agent playing **CartPole-v1**
This is a trained model of a **A2C** agent playing **CartPole-v1**
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
...
```
|
Yuri/xlm-roberta-base-finetuned-panx-de-fr
|
Yuri
| 2022-08-07T16:26:58Z
| 3
| 0
|
transformers
|
[
"transformers",
"pytorch",
"xlm-roberta",
"token-classification",
"generated_from_trainer",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
token-classification
| 2022-08-07T16:01:49Z
|
---
license: mit
tags:
- generated_from_trainer
metrics:
- f1
model-index:
- name: xlm-roberta-base-finetuned-panx-de-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. -->
# xlm-roberta-base-finetuned-panx-de-fr
This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1608
- F1: 0.8593
## 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: 24
- eval_batch_size: 24
- 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 | F1 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 0.2888 | 1.0 | 715 | 0.1779 | 0.8233 |
| 0.1437 | 2.0 | 1430 | 0.1570 | 0.8497 |
| 0.0931 | 3.0 | 2145 | 0.1608 | 0.8593 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.12.0+cu113
- Datasets 1.16.1
- Tokenizers 0.10.3
|
huggingtweets/apesahoy-dril_gpt2-nigella_lawson
|
huggingtweets
| 2022-08-07T16:24:31Z
| 3
| 0
|
transformers
|
[
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
text-generation
| 2022-08-07T16:22:50Z
|
---
language: en
thumbnail: http://www.huggingtweets.com/apesahoy-dril_gpt2-nigella_lawson/1659889467093/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1196519479364268034/5QpniWSP_400x400.jpg')">
</div>
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1493693969825423365/pqVtK9q0_400x400.jpg')">
</div>
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1514451221054173189/BWP3wqQj_400x400.jpg')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI CYBORG 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">Humongous Ape MP & Nigella Lawson & wint but Al</div>
<div style="text-align: center; font-size: 14px;">@apesahoy-dril_gpt2-nigella_lawson</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from Humongous Ape MP & Nigella Lawson & wint but Al.
| Data | Humongous Ape MP | Nigella Lawson | wint but Al |
| --- | --- | --- | --- |
| Tweets downloaded | 3245 | 3250 | 3229 |
| Retweets | 199 | 60 | 47 |
| Short tweets | 612 | 667 | 57 |
| Tweets kept | 2434 | 2523 | 3125 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/2nku0vp7/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @apesahoy-dril_gpt2-nigella_lawson's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/1gjr59ua) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/1gjr59ua/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/apesahoy-dril_gpt2-nigella_lawson')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
Yuri/xlm-roberta-base-finetuned-panx-de
|
Yuri
| 2022-08-07T15:54:38Z
| 6
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"xlm-roberta",
"token-classification",
"generated_from_trainer",
"dataset:xtreme",
"license:mit",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
token-classification
| 2022-08-07T13:23:41Z
|
---
license: mit
tags:
- generated_from_trainer
datasets:
- xtreme
metrics:
- f1
model-index:
- name: xlm-roberta-base-finetuned-panx-de
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: xtreme
type: xtreme
args: PAN-X.de
metrics:
- name: F1
type: f1
value: 0.8648740833380706
---
<!-- 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. -->
# xlm-roberta-base-finetuned-panx-de
This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the xtreme dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1365
- F1: 0.8649
## 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: 24
- eval_batch_size: 24
- 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 | F1 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 0.2553 | 1.0 | 525 | 0.1575 | 0.8279 |
| 0.1284 | 2.0 | 1050 | 0.1386 | 0.8463 |
| 0.0813 | 3.0 | 1575 | 0.1365 | 0.8649 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.12.0+cu113
- Datasets 1.16.1
- Tokenizers 0.10.3
|
huggingtweets/donalds28__-dril-kommmipakk
|
huggingtweets
| 2022-08-07T15:45:18Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
text-generation
| 2022-08-07T15:45:12Z
|
---
language: en
thumbnail: https://github.com/borisdayma/huggingtweets/blob/master/img/logo.png?raw=true
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1520796357149315073/VpjToHNe_400x400.jpg')">
</div>
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1241482050370224135/TAQSqgng_400x400.jpg')">
</div>
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1510917391533830145/XW-zSFDJ_400x400.jpg')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI CYBORG 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">owen & Ⓐju goblin 🦉 & wint</div>
<div style="text-align: center; font-size: 14px;">@donalds28__-dril-kommmipakk</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from owen & Ⓐju goblin 🦉 & wint.
| Data | owen | Ⓐju goblin 🦉 | wint |
| --- | --- | --- | --- |
| Tweets downloaded | 2359 | 2701 | 3224 |
| Retweets | 94 | 609 | 497 |
| Short tweets | 725 | 744 | 287 |
| Tweets kept | 1540 | 1348 | 2440 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/27r6w6f8/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @donalds28__-dril-kommmipakk's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/3pen7vd2) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/3pen7vd2/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/donalds28__-dril-kommmipakk')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
huggingtweets/apesahoy-chai_ste-punishedvirgo
|
huggingtweets
| 2022-08-07T14:48:33Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
text-generation
| 2022-08-07T14:47:43Z
|
---
language: en
thumbnail: http://www.huggingtweets.com/apesahoy-chai_ste-punishedvirgo/1659883708573/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1479595267800322048/Aqqb82wz_400x400.png')">
</div>
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1196519479364268034/5QpniWSP_400x400.jpg')">
</div>
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1513525852906180612/yOaU5tCT_400x400.jpg')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI CYBORG 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">ste 🍊 & Humongous Ape MP & radler inspector 🍋🍺🇵🇸🇺🇦</div>
<div style="text-align: center; font-size: 14px;">@apesahoy-chai_ste-punishedvirgo</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from ste 🍊 & Humongous Ape MP & radler inspector 🍋🍺🇵🇸🇺🇦.
| Data | ste 🍊 | Humongous Ape MP | radler inspector 🍋🍺🇵🇸🇺🇦 |
| --- | --- | --- | --- |
| Tweets downloaded | 3192 | 3245 | 3235 |
| Retweets | 293 | 199 | 657 |
| Short tweets | 494 | 612 | 675 |
| Tweets kept | 2405 | 2434 | 1903 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/eaah6yyc/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @apesahoy-chai_ste-punishedvirgo's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/deqhdcw0) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/deqhdcw0/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/apesahoy-chai_ste-punishedvirgo')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
dbmdz/detectron2-model
|
dbmdz
| 2022-08-07T14:35:23Z
| 0
| 2
| null |
[
"license:mit",
"region:us"
] | null | 2022-08-07T14:25:43Z
|
---
license: mit
---
# Detectron2 model
This repository hosts our trained Detectron2 model, that can detect segments from digitized books.
The following classes are supported:
- Illustration
- Illumination
The model is based on `faster_rcnn_R_50_FPN_3x` and was fine-tuned on own and manually annotated segments from digitized books.
|
jonahank/KlimaBERT
|
jonahank
| 2022-08-07T14:34:37Z
| 11
| 4
|
transformers
|
[
"transformers",
"pytorch",
"bert",
"text-classification",
"climate change",
"climate-classifier",
"political quotes",
"klimabert",
"da",
"arxiv:1810.04805",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text-classification
| 2022-06-01T08:21:00Z
|
---
language:
- da
tags:
- climate change
- climate-classifier
- political quotes
- klimabert
---
# Identifying and Analysing political quotes from the Danish Parliament related to climate change using NLP
**KlimaBERT**, a sequence-classifier fine-tuned to predict whether political quotes are climate-related. When predicting the positive class 1, "climate-related", the model achieves a F1-score of 0.97, Precision of 0.97, and Recall of 0.97. The negative class, 0, is defined as "non-climate-related".
KlimaBERT is fine-tuned using the pre-trained DaBERT-uncased model, on a training set of 1.000 manually labelled data-points. The training set contains both political quotes and summaries of bills from the [Danish Parliament](https://www.ft.dk/).
The model is created to identify political quotes related to climate change, and performs best on official texts from the Danish Parliament.
### Fine-tuning
To fine-tune a model similar to KlimaBERT, follow the [fine-tuning notebooks](https://github.com/jonahank/Vote-Prediction-Model/tree/main/climate_classifier)
### References
BERT:
Devlin, J., M.-W. Chang, K. Lee, and K. Toutanova (2018). Bert: Pre-training of deep
bidirectional transformers for language understanding.
https://arxiv.org/abs/1810.04805
DaBERT:
Certainly (2021). Certainly has trained the most advanced danish bert model to date.
https://www.certainly.io/blog/danish-bert-model/.
### Acknowledgements
The resources are created through the work of my Master's thesis, so I would like to thank my supervisors [Leon Derczynski](https://www.derczynski.com/itu/) and [Vedran Sekara](https://vedransekara.github.io/) for the great support throughout the project! And a HUGE thanks to [Gustav Gyrst](https://github.com/Gyrst) for great sparring and co-development of the tools you find in this repo.
### Contact
For any further help, questions, comments etc. feel free to contact the author Jonathan Kristensen on [LinedIn](https://www.linkedin.com/in/jonathan-kristensen-444a96104) or by creating a "discussion" on this model's page.
|
VietAI/gpt-neo-1.3B-vietnamese-news
|
VietAI
| 2022-08-07T14:32:07Z
| 1,341
| 28
|
transformers
|
[
"transformers",
"pytorch",
"gpt_neo",
"text-generation",
"causal-lm",
"gpt",
"vi",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text-generation
| 2022-03-02T23:29:05Z
|
---
language:
- vi
tags:
- pytorch
- causal-lm
- gpt
---
# GPT-Neo 1.3B on Vietnamese News
Details will be available soon.
For more information, please contact [email protected] (Dương) / [email protected] (Thành) / [email protected] (Bình).
### How to use
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("VietAI/gpt-neo-1.3B-vietnamese-news")
model = AutoModelForCausalLM.from_pretrained("VietAI/gpt-neo-1.3B-vietnamese-news", low_cpu_mem_usage=True)
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model.to(device)
prompt = "Tiềm năng của trí tuệ nhân tạo" # your input sentence
input_ids = tokenizer(prompt, return_tensors="pt")['input_ids'].to(device)
gen_tokens = model.generate(
input_ids,
max_length=max_length,
do_sample=True,
temperature=0.9,
top_k=20,
)
gen_text = tokenizer.batch_decode(gen_tokens)[0]
print(gen_text)
```
|
VietAI/gpt-j-6B-vietnamese-news
|
VietAI
| 2022-08-07T14:31:36Z
| 111
| 12
|
transformers
|
[
"transformers",
"pytorch",
"gptj",
"text-generation",
"causal-lm",
"vi",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text-generation
| 2022-03-02T23:29:05Z
|
---
language:
- vi
tags:
- pytorch
- causal-lm
- text-generation
---
# GPT-J 6B on Vietnamese News
Details will be available soon.
For more information, please contact [email protected] (Dương) / [email protected] (Thành) / [email protected] (Bình).
### How to use
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("VietAI/gpt-j-6B-vietnamese-news")
model = AutoModelForCausalLM.from_pretrained("VietAI/gpt-j-6B-vietnamese-news", low_cpu_mem_usage=True)
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model.to(device)
prompt = "Tiềm năng của trí tuệ nhân tạo" # your input sentence
input_ids = tokenizer(prompt, return_tensors="pt")['input_ids'].to(device)
gen_tokens = model.generate(
input_ids,
max_length=max_length,
do_sample=True,
temperature=0.9,
top_k=20,
)
gen_text = tokenizer.batch_decode(gen_tokens)[0]
print(gen_text)
```
|
mohammedbriman/t5-small-finetuned-tf-xsum
|
mohammedbriman
| 2022-08-07T13:52:08Z
| 4
| 0
|
transformers
|
[
"transformers",
"tf",
"tensorboard",
"t5",
"text2text-generation",
"generated_from_keras_callback",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text2text-generation
| 2022-07-16T14:46:07Z
|
---
tags:
- generated_from_keras_callback
model-index:
- name: t5-small-finetuned-tf-xsum
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. -->
# t5-small-finetuned-tf-xsum
This model was trained from scratch on xsum dataset.
It achieves the following results on the evaluation set:
- Train Loss: 2.3494
- Validation Loss: 2.1933
- Train Rouge1: 32.0241
- Train Rouge2: 10.1025
- Train Rougel: 25.8834
- Train Rougelsum: 25.9662
- Train Gen Len: 18.69
- Epoch: 8
## 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': 2e-05, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False, 'weight_decay_rate': 0.01}
- training_precision: float32
### Training results
| Train Loss | Validation Loss | Train Rouge1 | Train Rouge2 | Train Rougel | Train Rougelsum | Train Gen Len | Epoch |
| :--------: | :-------------: | :----------: | :----------: | :----------: | :-------------: | :-----------: | :---: |
| 2.7197 | 2.4028 | 29.6376 | 8.8596 | 22.8598 | 22.8401 | 18.82 | 1 |
| 2.5822 | 2.3407 | 30.6849 | 9.3100 | 23.8971 | 23.9096 | 18.745 | 2 |
| 2.5174 | 2.2979 | 32.3706 | 11.5463 | 26.4253 | 26.3525 | 18.75 | 3 |
| 2.4711 | 2.2703 | 32.2768 | 11.0460 | 26.2472 | 26.1540 | 18.825 | 4 |
| 2.4305 | 2.2432 | 29.3935 | 8.3337 | 22.2859 | 22.3557 | 18.65 | 5 |
| 2.3994 | 2.2237 | 31.0993 | 8.7932 | 23.6971 | 23.7702 | 18.815 | 6 |
| 2.3732 | 2.2071 | 31.4819 | 10.0677 | 25.1846 | 25.2829 | 18.675 | 7 |
| 2.3494 | 2.1933 | 32.0241 | 10.1025 | 25.8834 | 25.9662 | 18.69 | 8 |
### Framework versions
- Transformers 4.21.1
- TensorFlow 2.8.2
- Datasets 2.4.0
- Tokenizers 0.12.1
|
sunilkumardash9/finetuning-sentiment-model-3000-samples
|
sunilkumardash9
| 2022-08-07T12:16:22Z
| 8
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_trainer",
"dataset:imdb",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text-classification
| 2022-08-07T12:04:41Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- imdb
metrics:
- accuracy
- f1
model-index:
- name: finetuning-sentiment-model-3000-samples
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: 0.8666666666666667
- name: F1
type: f1
value: 0.8684210526315789
---
<!-- 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-3000-samples
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: 0.3213
- Accuracy: 0.8667
- F1: 0.8684
## 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 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
yhavinga/t5-small-24L-ccmatrix-multi
|
yhavinga
| 2022-08-07T12:07:16Z
| 8
| 2
|
transformers
|
[
"transformers",
"pytorch",
"jax",
"tensorboard",
"t5",
"text2text-generation",
"translation",
"seq2seq",
"nl",
"en",
"dataset:yhavinga/mc4_nl_cleaned",
"dataset:yhavinga/ccmatrix",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
translation
| 2022-04-27T09:46:23Z
|
---
language:
- nl
- en
datasets:
- yhavinga/mc4_nl_cleaned
- yhavinga/ccmatrix
tags:
- t5
- translation
- seq2seq
pipeline_tag: translation
widget:
- text: "It is a painful and tragic spectacle that rises before me: I have drawn back the curtain from the rottenness of man. This word, in my mouth, is at least free from one suspicion: that it involves a moral accusation against humanity."
- text: "Young Wehling was hunched in his chair, his head in his hand. He was so rumpled, so still and colorless as to be virtually invisible. His camouflage was perfect, since the waiting room had a disorderly and demoralized air, too. Chairs and ashtrays had been moved away from the walls. The floor was paved with spattered dropcloths."
license: apache-2.0
---
# t5-small-24L-ccmatrix-multi
A [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) model finetuned for Dutch to English and English to Dutch translation on the CCMatrix dataset.
Evaluation metrics of this model are listed in the **Translation models** section below.
You can use this model directly with a pipeline for text translation:
```python
model_name = "yhavinga/t5-small-24L-ccmatrix-multi"
from transformers import AutoTokenizer
from transformers import AutoModelForSeq2SeqLM
from transformers import pipeline
import torch
device_num = 0 if torch.cuda.is_available() else -1
device = "cpu" if device_num < 0 else f"cuda:{device_num}"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name).to(device)
params = {"max_length": 128, "num_beams": 4, "early_stopping": True}
en_to_nl = pipeline("translation_en_to_nl", tokenizer=tokenizer, model=model, device=device_num)
print(en_to_nl("""Young Wehling was hunched in his chair, his head in his hand. He was so rumpled, so still and colorless as to be virtually invisible.""",
**params)[0]['translation_text'])
nl_to_en = pipeline("translation_nl_to_en", tokenizer=tokenizer, model=model, device=device_num)
print(nl_to_en("""De jonge Wehling zat gebogen in zijn stoel, zijn hoofd in zijn hand. Hij was zo stoffig, zo stil en kleurloos dat hij vrijwel onzichtbaar was.""",
**params)[0]['translation_text'])
```
This **t5 eff** model has **249M** parameters.
It was pre-trained with masked language modeling (denoise token span corruption) objective on the dataset
`mc4_nl_cleaned` config `large_en_nl` for **1** epoch(s) and a duration of **4d10h**,
with a sequence length of **512**, batch size **128** and **851852** total steps (**56B** tokens).
Pre-training evaluation loss and accuracy are **1,18** and **0,74**.
Refer to the evaluation section below for a comparison of the pre-trained models on summarization and translation.
## Tokenizer
The model uses a cased SentencePiece tokenizer configured with the `Nmt, NFKC, Replace multi-space to single-space` normalizers
and has 32003 tokens.
It was trained on Dutch and English with scripts from the Huggingface Transformers [Flax examples](https://github.com/huggingface/transformers/tree/master/examples/flax/language-modeling).
See [./raw/main/tokenizer.json](tokenizer.json) for details.
## Dataset(s)
All models listed below are pre-trained on
[cleaned Dutch mC4](https://huggingface.co/datasets/yhavinga/mc4_nl_cleaned),
which is the original mC4, except
* Documents that contained words from a selection of the Dutch and English [List of Dirty Naught Obscene and Otherwise Bad Words](https://github.com/LDNOOBW/List-of-Dirty-Naughty-Obscene-and-Otherwise-Bad-Words) are removed
* Sentences with less than 3 words are removed
* Sentences with a word of more than 1000 characters are removed
* Documents with less than 5 sentences are removed
* Documents with "javascript", "lorum ipsum", "terms of use", "privacy policy", "cookie policy", "uses cookies",
"use of cookies", "use cookies", "elementen ontbreken", "deze printversie" are removed.
The Dutch and English models are pre-trained on a 50/50% mix of Dutch mC4 and English C4.
The translation models are fine-tuned on [CCMatrix](https://huggingface.co/datasets/yhavinga/ccmatrix).
## Dutch T5 Models
Three types of [Dutch T5 models have been trained (blog)](https://huggingface.co/spaces/yhavinga/pre-training-dutch-t5-models).
`t5-base-dutch` is the only model with an original T5 config.
The other model types t5-v1.1 and t5-eff have `gated-relu` instead of `relu` as activation function,
and trained with a drop-out of `0.0` unless training would diverge (`t5-v1.1-large-dutch-cased`).
The T5-eff models are models that differ in their number of layers. The table will list
the several dimensions of these models. Not all t5-eff models are efficient, the best example being the inefficient
`t5-xl-4L-dutch-english-cased`.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1.1-large-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-large-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-xl-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-xl-8l-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) |
|:------------------|:----------------|:-----------------------------|:---------------------------|:----------------------------|:-----------------------------------|:----------------------------------------|:-----------------------------|:-------------------------------|:----------------------------------|:-----------------------------------|:--------------------------------------|
| *type* | t5 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5 eff | t5 eff | t5 eff | t5 eff | t5 eff |
| *d_model* | 768 | 768 | 768 | 1024 | 768 | 768 | 512 | 2048 | 768 | 1024 | 1024 |
| *d_ff* | 3072 | 2048 | 2048 | 2816 | 2048 | 2048 | 1920 | 5120 | 2560 | 16384 | 4096 |
| *num_heads* | 12 | 12 | 12 | 16 | 12 | 12 | 8 | 32 | 12 | 32 | 16 |
| *d_kv* | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 128 | 64 |
| *num_layers* | 12 | 12 | 12 | 24 | 12 | 12 | 24 | 4 | 36 | 8 | 8 |
| *num parameters* | 223M | 248M | 248M | 783M | 248M | 248M | 250M | 585M | 729M | 1241M | 335M |
| *feed_forward_proj* | relu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu |
| *dropout* | 0.1 | 0.0 | 0.0 | 0.1 | 0.0 | 0.0 | 0.0 | 0.1 | 0.0 | 0.0 | 0.0 |
| *dataset* | mc4_nl_cleaned | mc4_nl_cleaned full | mc4_nl_cleaned full | mc4_nl_cleaned | mc4_nl_cleaned small_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl |
| *tr. seq len* | 512 | 1024 | 1024 | 512 | 512 | 1024 | 512 | 512 | 512 | 512 | 512 |
| *batch size* | 128 | 64 | 64 | 64 | 128 | 64 | 128 | 512 | 512 | 64 | 128 |
| *total steps* | 527500 | 1014525 | 1210154 | 1120k/2427498 | 2839630 | 1520k/3397024 | 851852 | 212963 | 212963 | 538k/1703705 | 851850 |
| *epochs* | 1 | 2 | 2 | 2 | 10 | 4 | 1 | 1 | 1 | 1 | 1 |
| *duration* | 2d9h | 5d5h | 6d6h | 8d13h | 11d18h | 9d1h | 4d10h | 6d1h | 17d15h | 4d 19h | 3d 23h |
| *optimizer* | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor |
| *lr* | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.009 | 0.005 | 0.005 |
| *warmup* | 10000.0 | 10000.0 | 10000.0 | 10000.0 | 10000.0 | 5000.0 | 20000.0 | 2500.0 | 1000.0 | 1500.0 | 1500.0 |
| *eval loss* | 1,38 | 1,20 | 0,96 | 1,07 | 1,11 | 1,13 | 1,18 | 1,27 | 1,05 | 1,3019 | 1,15 |
| *eval acc* | 0,70 | 0,73 | 0,78 | 0,76 | 0,75 | 0,74 | 0,74 | 0,72 | 0,76 | 0,71 | 0,74 |
## Evaluation
Most models from the list above have been fine-tuned for summarization and translation.
The figure below shows the evaluation scores, where the x-axis shows the translation Bleu score (higher is better)
and y-axis the summarization Rouge1 translation score (higher is better).
Point size is proportional to the model size. Models with faster inference speed are green, slower inference speed is
plotted as bleu.
Evaluation was run on fine-tuned models trained with the following settings:
| | Summarization | Translation |
|---------------:|------------------|-------------------|
| Dataset | CNN Dailymail NL | CCMatrix en -> nl |
| #train samples | 50K | 50K |
| Optimizer | Adam | Adam |
| learning rate | 0.001 | 0.0005 |
| source length | 1024 | 128 |
| target length | 142 | 128 |
|label smoothing | 0.05 | 0.1 |
| #eval samples | 1000 | 1000 |
Note that the amount of training data is limited to a fraction of the total dataset sizes, therefore the scores
below can only be used to compare the 'transfer-learning' strength. The fine-tuned checkpoints for this evaluation
are not saved, since they were trained for comparison of pre-trained models only.
The numbers for summarization are the Rouge scores on 1000 documents from the test split.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) | mt5-base |
|:------------------------|----------------:|-----------------------------:|---------------------------:|-----------------------------------:|----------------------------------------:|-----------------------------:|-------------------------------:|----------------------------------:|--------------------------------------:|-----------:|
| *rouge1* | 33.38 | 33.97 | 34.39 | 33.38 | 34.97 | 34.38 | 30.35 | **35.04** | 34.04 | 33.25 |
| *rouge2* | 13.32 | 13.85 | 13.98 | 13.47 | 14.01 | 13.89 | 11.57 | **14.23** | 13.76 | 12.74 |
| *rougeL* | 24.22 | 24.72 | 25.1 | 24.34 | 24.99 | **25.25** | 22.69 | 25.05 | 24.75 | 23.5 |
| *rougeLsum* | 30.23 | 30.9 | 31.44 | 30.51 | 32.01 | 31.38 | 27.5 | **32.12** | 31.12 | 30.15 |
| *samples_per_second* | 3.18 | 3.02 | 2.99 | 3.22 | 2.97 | 1.57 | 2.8 | 0.61 | **3.27** | 1.22 |
The models below have been evaluated for English to Dutch translation.
Note that the first four models are pre-trained on Dutch only. That they still perform adequate is probably because
the translation direction is English to Dutch.
The numbers reported are the Bleu scores on 1000 documents from the test split.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1.1-large-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-large-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) | mt5-base |
|:-------------------------------|----------------:|-----------------------------:|---------------------------:|----------------------------:|-----------------------------------:|----------------------------------------:|-----------------------------:|-------------------------------:|----------------------------------:|--------------------------------------:|-----------:|
| *precision_ng1* | 74.17 | 78.09 | 77.08 | 72.12 | 77.19 | 78.76 | 78.59 | 77.3 | **79.75** | 78.88 | 73.47 |
| *precision_ng2* | 52.42 | 57.52 | 55.31 | 48.7 | 55.39 | 58.01 | 57.83 | 55.27 | **59.89** | 58.27 | 50.12 |
| *precision_ng3* | 39.55 | 45.2 | 42.54 | 35.54 | 42.25 | 45.13 | 45.02 | 42.06 | **47.4** | 45.95 | 36.59 |
| *precision_ng4* | 30.23 | 36.04 | 33.26 | 26.27 | 32.74 | 35.72 | 35.41 | 32.61 | **38.1** | 36.91 | 27.26 |
| *bp* | 0.99 | 0.98 | 0.97 | 0.98 | 0.98 | 0.98 | 0.98 | 0.97 | 0.98 | 0.98 | 0.98 |
| *score* | 45.88 | 51.21 | 48.31 | 41.59 | 48.17 | 51.31 | 50.82 | 47.83 | **53** | 51.79 | 42.74 |
| *samples_per_second* | **45.19** | 45.05 | 38.67 | 10.12 | 42.19 | 42.61 | 12.85 | 33.74 | 9.07 | 37.86 | 9.03 |
## Translation models
The models `t5-small-24L-dutch-english` and `t5-base-36L-dutch-english` have been fine-tuned for both language
directions on the first 25M samples from CCMatrix, giving a total of 50M training samples.
Evaluation is performed on out-of-sample CCMatrix and also on Tatoeba and Opus Books.
The `_bp` columns list the *brevity penalty*. The `avg_bleu` score is the bleu score
averaged over all three evaluation datasets. The best scores displayed in bold for both translation directions.
| | [t5-base-36L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-base-36L-ccmatrix-multi) | [t5-base-36L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-base-36L-ccmatrix-multi) | [t5-small-24L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-small-24L-ccmatrix-multi) | [t5-small-24L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-small-24L-ccmatrix-multi) |
|:-----------------------|:-----------------------------|:-----------------------------|:------------------------------|:------------------------------|
| *source_lang* | en | nl | en | nl |
| *target_lang* | nl | en | nl | en |
| *source_prefix* | translate English to Dutch: | translate Dutch to English: | translate English to Dutch: | translate Dutch to English: |
| *ccmatrix_bleu* | **56.8** | 62.8 | 57.4 | **63.1** |
| *tatoeba_bleu* | **46.6** | **52.8** | 46.4 | 51.7 |
| *opus_books_bleu* | **13.5** | **24.9** | 12.9 | 23.4 |
| *ccmatrix_bp* | 0.95 | 0.96 | 0.95 | 0.96 |
| *tatoeba_bp* | 0.97 | 0.94 | 0.98 | 0.94 |
| *opus_books_bp* | 0.8 | 0.94 | 0.77 | 0.89 |
| *avg_bleu* | **38.96** | **46.86** | 38.92 | 46.06 |
| *max_source_length* | 128 | 128 | 128 | 128 |
| *max_target_length* | 128 | 128 | 128 | 128 |
| *adam_beta1* | 0.9 | 0.9 | 0.9 | 0.9 |
| *adam_beta2* | 0.997 | 0.997 | 0.997 | 0.997 |
| *weight_decay* | 0.05 | 0.05 | 0.002 | 0.002 |
| *lr* | 5e-05 | 5e-05 | 0.0005 | 0.0005 |
| *label_smoothing_factor* | 0.15 | 0.15 | 0.1 | 0.1 |
| *train_batch_size* | 128 | 128 | 128 | 128 |
| *warmup_steps* | 2000 | 2000 | 2000 | 2000 |
| *total steps* | 390625 | 390625 | 390625 | 390625 |
| *duration* | 4d 5h | 4d 5h | 3d 2h | 3d 2h |
| *num parameters* | 729M | 729M | 250M | 250M |
## Acknowledgements
This project would not have been possible without compute generously provided by Google through the
[TPU Research Cloud](https://sites.research.google/trc/). The HuggingFace 🤗 ecosystem was instrumental in all parts
of the training. Weights & Biases made it possible to keep track of many training sessions
and orchestrate hyper-parameter sweeps with insightful visualizations.
The following repositories where helpful in setting up the TPU-VM,
and getting an idea what sensible hyper-parameters are for training gpt2 from scratch:
* [Gsarti's Pretrain and Fine-tune a T5 model with Flax on GCP](https://github.com/gsarti/t5-flax-gcp)
* [Flax/Jax Community week t5-base-dutch](https://huggingface.co/flax-community/t5-base-dutch)
Created by [Yeb Havinga](https://www.linkedin.com/in/yeb-havinga-86530825/)
|
yhavinga/t5-base-36L-ccmatrix-multi
|
yhavinga
| 2022-08-07T12:07:12Z
| 13
| 1
|
transformers
|
[
"transformers",
"pytorch",
"jax",
"tensorboard",
"t5",
"text2text-generation",
"translation",
"seq2seq",
"nl",
"en",
"dataset:yhavinga/mc4_nl_cleaned",
"dataset:yhavinga/ccmatrix",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
translation
| 2022-04-22T07:56:31Z
|
---
language:
- nl
- en
datasets:
- yhavinga/mc4_nl_cleaned
- yhavinga/ccmatrix
tags:
- t5
- translation
- seq2seq
pipeline_tag: translation
widget:
- text: "It is a painful and tragic spectacle that rises before me: I have drawn back the curtain from the rottenness of man. This word, in my mouth, is at least free from one suspicion: that it involves a moral accusation against humanity."
- text: "Young Wehling was hunched in his chair, his head in his hand. He was so rumpled, so still and colorless as to be virtually invisible. His camouflage was perfect, since the waiting room had a disorderly and demoralized air, too. Chairs and ashtrays had been moved away from the walls. The floor was paved with spattered dropcloths."
license: apache-2.0
---
# t5-base-36L-ccmatrix-multi
A [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) model finetuned for Dutch to English and English to Dutch translation on the CCMatrix dataset.
Evaluation metrics of this model are listed in the **Translation models** section below.
You can use this model directly with a pipeline for text translation:
```python
model_name = "yhavinga/t5-base-36L-ccmatrix-multi"
from transformers import AutoTokenizer
from transformers import AutoModelForSeq2SeqLM
from transformers import pipeline
import torch
device_num = 0 if torch.cuda.is_available() else -1
device = "cpu" if device_num < 0 else f"cuda:{device_num}"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name).to(device)
params = {"max_length": 128, "num_beams": 4, "early_stopping": True}
en_to_nl = pipeline("translation_en_to_nl", tokenizer=tokenizer, model=model, device=device_num)
print(en_to_nl("""Young Wehling was hunched in his chair, his head in his hand. He was so rumpled, so still and colorless as to be virtually invisible.""",
**params)[0]['translation_text'])
nl_to_en = pipeline("translation_nl_to_en", tokenizer=tokenizer, model=model, device=device_num)
print(nl_to_en("""De jonge Wehling zat gebogen in zijn stoel, zijn hoofd in zijn hand. Hij was zo stoffig, zo stil en kleurloos dat hij vrijwel onzichtbaar was.""",
**params)[0]['translation_text'])
```
This **t5 eff** model has **728M** parameters.
It was pre-trained with masked language modeling (denoise token span corruption) objective on the dataset
`mc4_nl_cleaned` config `large_en_nl` for **1** epoch(s) and a duration of **17d15h**,
with a sequence length of **512**, batch size **512** and **212963** total steps (**56B** tokens).
Pre-training evaluation loss and accuracy are **1,05** and **0,76**.
Refer to the evaluation section below for a comparison of the pre-trained models on summarization and translation.
## Tokenizer
The model uses a cased SentencePiece tokenizer configured with the `Nmt, NFKC, Replace multi-space to single-space` normalizers
and has 32003 tokens.
It was trained on Dutch and English with scripts from the Huggingface Transformers [Flax examples](https://github.com/huggingface/transformers/tree/master/examples/flax/language-modeling).
See [./raw/main/tokenizer.json](tokenizer.json) for details.
## Dataset(s)
All models listed below are pre-trained on
[cleaned Dutch mC4](https://huggingface.co/datasets/yhavinga/mc4_nl_cleaned),
which is the original mC4, except
* Documents that contained words from a selection of the Dutch and English [List of Dirty Naught Obscene and Otherwise Bad Words](https://github.com/LDNOOBW/List-of-Dirty-Naughty-Obscene-and-Otherwise-Bad-Words) are removed
* Sentences with less than 3 words are removed
* Sentences with a word of more than 1000 characters are removed
* Documents with less than 5 sentences are removed
* Documents with "javascript", "lorum ipsum", "terms of use", "privacy policy", "cookie policy", "uses cookies",
"use of cookies", "use cookies", "elementen ontbreken", "deze printversie" are removed.
The Dutch and English models are pre-trained on a 50/50% mix of Dutch mC4 and English C4.
The translation models are fine-tuned on [CCMatrix](https://huggingface.co/datasets/yhavinga/ccmatrix).
## Dutch T5 Models
Three types of [Dutch T5 models have been trained (blog)](https://huggingface.co/spaces/yhavinga/pre-training-dutch-t5-models).
`t5-base-dutch` is the only model with an original T5 config.
The other model types t5-v1.1 and t5-eff have `gated-relu` instead of `relu` as activation function,
and trained with a drop-out of `0.0` unless training would diverge (`t5-v1.1-large-dutch-cased`).
The T5-eff models are models that differ in their number of layers. The table will list
the several dimensions of these models. Not all t5-eff models are efficient, the best example being the inefficient
`t5-xl-4L-dutch-english-cased`.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1.1-large-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-large-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-xl-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-xl-8l-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) |
|:------------------|:----------------|:-----------------------------|:---------------------------|:----------------------------|:-----------------------------------|:----------------------------------------|:-----------------------------|:-------------------------------|:----------------------------------|:-----------------------------------|:--------------------------------------|
| *type* | t5 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5 eff | t5 eff | t5 eff | t5 eff | t5 eff |
| *d_model* | 768 | 768 | 768 | 1024 | 768 | 768 | 512 | 2048 | 768 | 1024 | 1024 |
| *d_ff* | 3072 | 2048 | 2048 | 2816 | 2048 | 2048 | 1920 | 5120 | 2560 | 16384 | 4096 |
| *num_heads* | 12 | 12 | 12 | 16 | 12 | 12 | 8 | 32 | 12 | 32 | 16 |
| *d_kv* | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 128 | 64 |
| *num_layers* | 12 | 12 | 12 | 24 | 12 | 12 | 24 | 4 | 36 | 8 | 8 |
| *num parameters* | 223M | 248M | 248M | 783M | 248M | 248M | 250M | 585M | 729M | 1241M | 335M |
| *feed_forward_proj* | relu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu |
| *dropout* | 0.1 | 0.0 | 0.0 | 0.1 | 0.0 | 0.0 | 0.0 | 0.1 | 0.0 | 0.0 | 0.0 |
| *dataset* | mc4_nl_cleaned | mc4_nl_cleaned full | mc4_nl_cleaned full | mc4_nl_cleaned | mc4_nl_cleaned small_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl |
| *tr. seq len* | 512 | 1024 | 1024 | 512 | 512 | 1024 | 512 | 512 | 512 | 512 | 512 |
| *batch size* | 128 | 64 | 64 | 64 | 128 | 64 | 128 | 512 | 512 | 64 | 128 |
| *total steps* | 527500 | 1014525 | 1210154 | 1120k/2427498 | 2839630 | 1520k/3397024 | 851852 | 212963 | 212963 | 538k/1703705 | 851850 |
| *epochs* | 1 | 2 | 2 | 2 | 10 | 4 | 1 | 1 | 1 | 1 | 1 |
| *duration* | 2d9h | 5d5h | 6d6h | 8d13h | 11d18h | 9d1h | 4d10h | 6d1h | 17d15h | 4d 19h | 3d 23h |
| *optimizer* | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor |
| *lr* | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.009 | 0.005 | 0.005 |
| *warmup* | 10000.0 | 10000.0 | 10000.0 | 10000.0 | 10000.0 | 5000.0 | 20000.0 | 2500.0 | 1000.0 | 1500.0 | 1500.0 |
| *eval loss* | 1,38 | 1,20 | 0,96 | 1,07 | 1,11 | 1,13 | 1,18 | 1,27 | 1,05 | 1,3019 | 1,15 |
| *eval acc* | 0,70 | 0,73 | 0,78 | 0,76 | 0,75 | 0,74 | 0,74 | 0,72 | 0,76 | 0,71 | 0,74 |
## Evaluation
Most models from the list above have been fine-tuned for summarization and translation.
The figure below shows the evaluation scores, where the x-axis shows the translation Bleu score (higher is better)
and y-axis the summarization Rouge1 translation score (higher is better).
Point size is proportional to the model size. Models with faster inference speed are green, slower inference speed is
plotted as bleu.
Evaluation was run on fine-tuned models trained with the following settings:
| | Summarization | Translation |
|---------------:|------------------|-------------------|
| Dataset | CNN Dailymail NL | CCMatrix en -> nl |
| #train samples | 50K | 50K |
| Optimizer | Adam | Adam |
| learning rate | 0.001 | 0.0005 |
| source length | 1024 | 128 |
| target length | 142 | 128 |
|label smoothing | 0.05 | 0.1 |
| #eval samples | 1000 | 1000 |
Note that the amount of training data is limited to a fraction of the total dataset sizes, therefore the scores
below can only be used to compare the 'transfer-learning' strength. The fine-tuned checkpoints for this evaluation
are not saved, since they were trained for comparison of pre-trained models only.
The numbers for summarization are the Rouge scores on 1000 documents from the test split.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) | mt5-base |
|:------------------------|----------------:|-----------------------------:|---------------------------:|-----------------------------------:|----------------------------------------:|-----------------------------:|-------------------------------:|----------------------------------:|--------------------------------------:|-----------:|
| *rouge1* | 33.38 | 33.97 | 34.39 | 33.38 | 34.97 | 34.38 | 30.35 | **35.04** | 34.04 | 33.25 |
| *rouge2* | 13.32 | 13.85 | 13.98 | 13.47 | 14.01 | 13.89 | 11.57 | **14.23** | 13.76 | 12.74 |
| *rougeL* | 24.22 | 24.72 | 25.1 | 24.34 | 24.99 | **25.25** | 22.69 | 25.05 | 24.75 | 23.5 |
| *rougeLsum* | 30.23 | 30.9 | 31.44 | 30.51 | 32.01 | 31.38 | 27.5 | **32.12** | 31.12 | 30.15 |
| *samples_per_second* | 3.18 | 3.02 | 2.99 | 3.22 | 2.97 | 1.57 | 2.8 | 0.61 | **3.27** | 1.22 |
The models below have been evaluated for English to Dutch translation.
Note that the first four models are pre-trained on Dutch only. That they still perform adequate is probably because
the translation direction is English to Dutch.
The numbers reported are the Bleu scores on 1000 documents from the test split.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1.1-large-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-large-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) | mt5-base |
|:-------------------------------|----------------:|-----------------------------:|---------------------------:|----------------------------:|-----------------------------------:|----------------------------------------:|-----------------------------:|-------------------------------:|----------------------------------:|--------------------------------------:|-----------:|
| *precision_ng1* | 74.17 | 78.09 | 77.08 | 72.12 | 77.19 | 78.76 | 78.59 | 77.3 | **79.75** | 78.88 | 73.47 |
| *precision_ng2* | 52.42 | 57.52 | 55.31 | 48.7 | 55.39 | 58.01 | 57.83 | 55.27 | **59.89** | 58.27 | 50.12 |
| *precision_ng3* | 39.55 | 45.2 | 42.54 | 35.54 | 42.25 | 45.13 | 45.02 | 42.06 | **47.4** | 45.95 | 36.59 |
| *precision_ng4* | 30.23 | 36.04 | 33.26 | 26.27 | 32.74 | 35.72 | 35.41 | 32.61 | **38.1** | 36.91 | 27.26 |
| *bp* | 0.99 | 0.98 | 0.97 | 0.98 | 0.98 | 0.98 | 0.98 | 0.97 | 0.98 | 0.98 | 0.98 |
| *score* | 45.88 | 51.21 | 48.31 | 41.59 | 48.17 | 51.31 | 50.82 | 47.83 | **53** | 51.79 | 42.74 |
| *samples_per_second* | **45.19** | 45.05 | 38.67 | 10.12 | 42.19 | 42.61 | 12.85 | 33.74 | 9.07 | 37.86 | 9.03 |
## Translation models
The models `t5-small-24L-dutch-english` and `t5-base-36L-dutch-english` have been fine-tuned for both language
directions on the first 25M samples from CCMatrix, giving a total of 50M training samples.
Evaluation is performed on out-of-sample CCMatrix and also on Tatoeba and Opus Books.
The `_bp` columns list the *brevity penalty*. The `avg_bleu` score is the bleu score
averaged over all three evaluation datasets. The best scores displayed in bold for both translation directions.
| | [t5-base-36L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-base-36L-ccmatrix-multi) | [t5-base-36L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-base-36L-ccmatrix-multi) | [t5-small-24L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-small-24L-ccmatrix-multi) | [t5-small-24L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-small-24L-ccmatrix-multi) |
|:-----------------------|:-----------------------------|:-----------------------------|:------------------------------|:------------------------------|
| *source_lang* | en | nl | en | nl |
| *target_lang* | nl | en | nl | en |
| *source_prefix* | translate English to Dutch: | translate Dutch to English: | translate English to Dutch: | translate Dutch to English: |
| *ccmatrix_bleu* | **56.8** | 62.8 | 57.4 | **63.1** |
| *tatoeba_bleu* | **46.6** | **52.8** | 46.4 | 51.7 |
| *opus_books_bleu* | **13.5** | **24.9** | 12.9 | 23.4 |
| *ccmatrix_bp* | 0.95 | 0.96 | 0.95 | 0.96 |
| *tatoeba_bp* | 0.97 | 0.94 | 0.98 | 0.94 |
| *opus_books_bp* | 0.8 | 0.94 | 0.77 | 0.89 |
| *avg_bleu* | **38.96** | **46.86** | 38.92 | 46.06 |
| *max_source_length* | 128 | 128 | 128 | 128 |
| *max_target_length* | 128 | 128 | 128 | 128 |
| *adam_beta1* | 0.9 | 0.9 | 0.9 | 0.9 |
| *adam_beta2* | 0.997 | 0.997 | 0.997 | 0.997 |
| *weight_decay* | 0.05 | 0.05 | 0.002 | 0.002 |
| *lr* | 5e-05 | 5e-05 | 0.0005 | 0.0005 |
| *label_smoothing_factor* | 0.15 | 0.15 | 0.1 | 0.1 |
| *train_batch_size* | 128 | 128 | 128 | 128 |
| *warmup_steps* | 2000 | 2000 | 2000 | 2000 |
| *total steps* | 390625 | 390625 | 390625 | 390625 |
| *duration* | 4d 5h | 4d 5h | 3d 2h | 3d 2h |
| *num parameters* | 729M | 729M | 250M | 250M |
## Acknowledgements
This project would not have been possible without compute generously provided by Google through the
[TPU Research Cloud](https://sites.research.google/trc/). The HuggingFace 🤗 ecosystem was instrumental in all parts
of the training. Weights & Biases made it possible to keep track of many training sessions
and orchestrate hyper-parameter sweeps with insightful visualizations.
The following repositories where helpful in setting up the TPU-VM,
and getting an idea what sensible hyper-parameters are for training gpt2 from scratch:
* [Gsarti's Pretrain and Fine-tune a T5 model with Flax on GCP](https://github.com/gsarti/t5-flax-gcp)
* [Flax/Jax Community week t5-base-dutch](https://huggingface.co/flax-community/t5-base-dutch)
Created by [Yeb Havinga](https://www.linkedin.com/in/yeb-havinga-86530825/)
|
yhavinga/t5-eff-large-8l-dutch-english-cased
|
yhavinga
| 2022-08-07T12:07:07Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"seq2seq",
"nl",
"en",
"dataset:yhavinga/mc4_nl_cleaned",
"arxiv:1910.10683",
"arxiv:2109.10686",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"region:us"
] |
text2text-generation
| 2022-05-06T07:48:22Z
|
---
language:
- nl
- en
datasets:
- yhavinga/mc4_nl_cleaned
tags:
- t5
- seq2seq
inference: false
license: apache-2.0
---
# t5-eff-large-8l-dutch-english-cased
A [T5](https://ai.googleblog.com/2020/02/exploring-transfer-learning-with-t5.html) sequence to sequence model
pre-trained from scratch on [cleaned Dutch 🇳🇱🇧🇪 mC4 and cleaned English 🇬🇧 C4](https://huggingface.co/datasets/yhavinga/mc4_nl_cleaned).
This **t5 eff** model has **334M** parameters.
It was pre-trained with masked language modeling (denoise token span corruption) objective on the dataset
`mc4_nl_cleaned` config `large_en_nl` for **1** epoch(s) and a duration of **3d 23h**,
with a sequence length of **512**, batch size **128** and **851850** total steps (**56B** tokens).
Pre-training evaluation loss and accuracy are **1,15** and **0,74**.
Refer to the evaluation section below for a comparison of the pre-trained models on summarization and translation.
* Pre-trained T5 models need to be finetuned before they can be used for downstream tasks, therefore the inference widget on the right has been turned off.
* For a demo of the Dutch CNN summarization models, head over to the Hugging Face Spaces for
the **[Netherformer 📰](https://huggingface.co/spaces/flax-community/netherformer)** example application!
Please refer to the original T5 papers and Scale Efficiently papers for more information about the T5 architecture
and configs, though it must be noted that this model (t5-eff-large-8l-dutch-english-cased) is unrelated to these projects and not an 'official' checkpoint.
* **[Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/pdf/1910.10683.pdf)** by *Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, Peter J. Liu*.
* **[Scale Efficiently: Insights from Pre-training and Fine-tuning Transformers](https://arxiv.org/abs/2109.10686)** by *Yi Tay, Mostafa Dehghani, Jinfeng Rao, William Fedus, Samira Abnar, Hyung Won Chung, Sharan Narang, Dani Yogatama, Ashish Vaswani, Donald Metzler*.
## Tokenizer
The model uses a cased SentencePiece tokenizer configured with the `Nmt, NFKC, Replace multi-space to single-space` normalizers
and has 32003 tokens.
It was trained on Dutch and English with scripts from the Huggingface Transformers [Flax examples](https://github.com/huggingface/transformers/tree/master/examples/flax/language-modeling).
See [./raw/main/tokenizer.json](tokenizer.json) for details.
## Dataset(s)
All models listed below are pre-trained on
[cleaned Dutch mC4](https://huggingface.co/datasets/yhavinga/mc4_nl_cleaned),
which is the original mC4, except
* Documents that contained words from a selection of the Dutch and English [List of Dirty Naught Obscene and Otherwise Bad Words](https://github.com/LDNOOBW/List-of-Dirty-Naughty-Obscene-and-Otherwise-Bad-Words) are removed
* Sentences with less than 3 words are removed
* Sentences with a word of more than 1000 characters are removed
* Documents with less than 5 sentences are removed
* Documents with "javascript", "lorum ipsum", "terms of use", "privacy policy", "cookie policy", "uses cookies",
"use of cookies", "use cookies", "elementen ontbreken", "deze printversie" are removed.
The Dutch and English models are pre-trained on a 50/50% mix of Dutch mC4 and English C4.
The translation models are fine-tuned on [CCMatrix](https://huggingface.co/datasets/yhavinga/ccmatrix).
## Dutch T5 Models
Three types of [Dutch T5 models have been trained (blog)](https://huggingface.co/spaces/yhavinga/pre-training-dutch-t5-models).
`t5-base-dutch` is the only model with an original T5 config.
The other model types t5-v1.1 and t5-eff have `gated-relu` instead of `relu` as activation function,
and trained with a drop-out of `0.0` unless training would diverge (`t5-v1.1-large-dutch-cased`).
The T5-eff models are models that differ in their number of layers. The table will list
the several dimensions of these models. Not all t5-eff models are efficient, the best example being the inefficient
`t5-xl-4L-dutch-english-cased`.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1.1-large-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-large-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-xl-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-xl-8l-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) |
|:------------------|:----------------|:-----------------------------|:---------------------------|:----------------------------|:-----------------------------------|:----------------------------------------|:-----------------------------|:-------------------------------|:----------------------------------|:-----------------------------------|:--------------------------------------|
| *type* | t5 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5 eff | t5 eff | t5 eff | t5 eff | t5 eff |
| *d_model* | 768 | 768 | 768 | 1024 | 768 | 768 | 512 | 2048 | 768 | 1024 | 1024 |
| *d_ff* | 3072 | 2048 | 2048 | 2816 | 2048 | 2048 | 1920 | 5120 | 2560 | 16384 | 4096 |
| *num_heads* | 12 | 12 | 12 | 16 | 12 | 12 | 8 | 32 | 12 | 32 | 16 |
| *d_kv* | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 128 | 64 |
| *num_layers* | 12 | 12 | 12 | 24 | 12 | 12 | 24 | 4 | 36 | 8 | 8 |
| *num parameters* | 223M | 248M | 248M | 783M | 248M | 248M | 250M | 585M | 729M | 1241M | 335M |
| *feed_forward_proj* | relu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu |
| *dropout* | 0.1 | 0.0 | 0.0 | 0.1 | 0.0 | 0.0 | 0.0 | 0.1 | 0.0 | 0.0 | 0.0 |
| *dataset* | mc4_nl_cleaned | mc4_nl_cleaned full | mc4_nl_cleaned full | mc4_nl_cleaned | mc4_nl_cleaned small_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl |
| *tr. seq len* | 512 | 1024 | 1024 | 512 | 512 | 1024 | 512 | 512 | 512 | 512 | 512 |
| *batch size* | 128 | 64 | 64 | 64 | 128 | 64 | 128 | 512 | 512 | 64 | 128 |
| *total steps* | 527500 | 1014525 | 1210154 | 1120k/2427498 | 2839630 | 1520k/3397024 | 851852 | 212963 | 212963 | 538k/1703705 | 851850 |
| *epochs* | 1 | 2 | 2 | 2 | 10 | 4 | 1 | 1 | 1 | 1 | 1 |
| *duration* | 2d9h | 5d5h | 6d6h | 8d13h | 11d18h | 9d1h | 4d10h | 6d1h | 17d15h | 4d 19h | 3d 23h |
| *optimizer* | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor |
| *lr* | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.009 | 0.005 | 0.005 |
| *warmup* | 10000.0 | 10000.0 | 10000.0 | 10000.0 | 10000.0 | 5000.0 | 20000.0 | 2500.0 | 1000.0 | 1500.0 | 1500.0 |
| *eval loss* | 1,38 | 1,20 | 0,96 | 1,07 | 1,11 | 1,13 | 1,18 | 1,27 | 1,05 | 1,3019 | 1,15 |
| *eval acc* | 0,70 | 0,73 | 0,78 | 0,76 | 0,75 | 0,74 | 0,74 | 0,72 | 0,76 | 0,71 | 0,74 |
## Evaluation
Most models from the list above have been fine-tuned for summarization and translation.
The figure below shows the evaluation scores, where the x-axis shows the translation Bleu score (higher is better)
and y-axis the summarization Rouge1 translation score (higher is better).
Point size is proportional to the model size. Models with faster inference speed are green, slower inference speed is
plotted as bleu.
Evaluation was run on fine-tuned models trained with the following settings:
| | Summarization | Translation |
|---------------:|------------------|-------------------|
| Dataset | CNN Dailymail NL | CCMatrix en -> nl |
| #train samples | 50K | 50K |
| Optimizer | Adam | Adam |
| learning rate | 0.001 | 0.0005 |
| source length | 1024 | 128 |
| target length | 142 | 128 |
|label smoothing | 0.05 | 0.1 |
| #eval samples | 1000 | 1000 |
Note that the amount of training data is limited to a fraction of the total dataset sizes, therefore the scores
below can only be used to compare the 'transfer-learning' strength. The fine-tuned checkpoints for this evaluation
are not saved, since they were trained for comparison of pre-trained models only.
The numbers for summarization are the Rouge scores on 1000 documents from the test split.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) | mt5-base |
|:------------------------|----------------:|-----------------------------:|---------------------------:|-----------------------------------:|----------------------------------------:|-----------------------------:|-------------------------------:|----------------------------------:|--------------------------------------:|-----------:|
| *rouge1* | 33.38 | 33.97 | 34.39 | 33.38 | 34.97 | 34.38 | 30.35 | **35.04** | 34.04 | 33.25 |
| *rouge2* | 13.32 | 13.85 | 13.98 | 13.47 | 14.01 | 13.89 | 11.57 | **14.23** | 13.76 | 12.74 |
| *rougeL* | 24.22 | 24.72 | 25.1 | 24.34 | 24.99 | **25.25** | 22.69 | 25.05 | 24.75 | 23.5 |
| *rougeLsum* | 30.23 | 30.9 | 31.44 | 30.51 | 32.01 | 31.38 | 27.5 | **32.12** | 31.12 | 30.15 |
| *samples_per_second* | 3.18 | 3.02 | 2.99 | 3.22 | 2.97 | 1.57 | 2.8 | 0.61 | **3.27** | 1.22 |
The models below have been evaluated for English to Dutch translation.
Note that the first four models are pre-trained on Dutch only. That they still perform adequate is probably because
the translation direction is English to Dutch.
The numbers reported are the Bleu scores on 1000 documents from the test split.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1.1-large-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-large-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) | mt5-base |
|:-------------------------------|----------------:|-----------------------------:|---------------------------:|----------------------------:|-----------------------------------:|----------------------------------------:|-----------------------------:|-------------------------------:|----------------------------------:|--------------------------------------:|-----------:|
| *precision_ng1* | 74.17 | 78.09 | 77.08 | 72.12 | 77.19 | 78.76 | 78.59 | 77.3 | **79.75** | 78.88 | 73.47 |
| *precision_ng2* | 52.42 | 57.52 | 55.31 | 48.7 | 55.39 | 58.01 | 57.83 | 55.27 | **59.89** | 58.27 | 50.12 |
| *precision_ng3* | 39.55 | 45.2 | 42.54 | 35.54 | 42.25 | 45.13 | 45.02 | 42.06 | **47.4** | 45.95 | 36.59 |
| *precision_ng4* | 30.23 | 36.04 | 33.26 | 26.27 | 32.74 | 35.72 | 35.41 | 32.61 | **38.1** | 36.91 | 27.26 |
| *bp* | 0.99 | 0.98 | 0.97 | 0.98 | 0.98 | 0.98 | 0.98 | 0.97 | 0.98 | 0.98 | 0.98 |
| *score* | 45.88 | 51.21 | 48.31 | 41.59 | 48.17 | 51.31 | 50.82 | 47.83 | **53** | 51.79 | 42.74 |
| *samples_per_second* | **45.19** | 45.05 | 38.67 | 10.12 | 42.19 | 42.61 | 12.85 | 33.74 | 9.07 | 37.86 | 9.03 |
## Translation models
The models `t5-small-24L-dutch-english` and `t5-base-36L-dutch-english` have been fine-tuned for both language
directions on the first 25M samples from CCMatrix, giving a total of 50M training samples.
Evaluation is performed on out-of-sample CCMatrix and also on Tatoeba and Opus Books.
The `_bp` columns list the *brevity penalty*. The `avg_bleu` score is the bleu score
averaged over all three evaluation datasets. The best scores displayed in bold for both translation directions.
| | [t5-base-36L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-base-36L-ccmatrix-multi) | [t5-base-36L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-base-36L-ccmatrix-multi) | [t5-small-24L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-small-24L-ccmatrix-multi) | [t5-small-24L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-small-24L-ccmatrix-multi) |
|:-----------------------|:-----------------------------|:-----------------------------|:------------------------------|:------------------------------|
| *source_lang* | en | nl | en | nl |
| *target_lang* | nl | en | nl | en |
| *source_prefix* | translate English to Dutch: | translate Dutch to English: | translate English to Dutch: | translate Dutch to English: |
| *ccmatrix_bleu* | **56.8** | 62.8 | 57.4 | **63.1** |
| *tatoeba_bleu* | **46.6** | **52.8** | 46.4 | 51.7 |
| *opus_books_bleu* | **13.5** | **24.9** | 12.9 | 23.4 |
| *ccmatrix_bp* | 0.95 | 0.96 | 0.95 | 0.96 |
| *tatoeba_bp* | 0.97 | 0.94 | 0.98 | 0.94 |
| *opus_books_bp* | 0.8 | 0.94 | 0.77 | 0.89 |
| *avg_bleu* | **38.96** | **46.86** | 38.92 | 46.06 |
| *max_source_length* | 128 | 128 | 128 | 128 |
| *max_target_length* | 128 | 128 | 128 | 128 |
| *adam_beta1* | 0.9 | 0.9 | 0.9 | 0.9 |
| *adam_beta2* | 0.997 | 0.997 | 0.997 | 0.997 |
| *weight_decay* | 0.05 | 0.05 | 0.002 | 0.002 |
| *lr* | 5e-05 | 5e-05 | 0.0005 | 0.0005 |
| *label_smoothing_factor* | 0.15 | 0.15 | 0.1 | 0.1 |
| *train_batch_size* | 128 | 128 | 128 | 128 |
| *warmup_steps* | 2000 | 2000 | 2000 | 2000 |
| *total steps* | 390625 | 390625 | 390625 | 390625 |
| *duration* | 4d 5h | 4d 5h | 3d 2h | 3d 2h |
| *num parameters* | 729M | 729M | 250M | 250M |
## Acknowledgements
This project would not have been possible without compute generously provided by Google through the
[TPU Research Cloud](https://sites.research.google/trc/). The HuggingFace 🤗 ecosystem was instrumental in all parts
of the training. Weights & Biases made it possible to keep track of many training sessions
and orchestrate hyper-parameter sweeps with insightful visualizations.
The following repositories where helpful in setting up the TPU-VM,
and getting an idea what sensible hyper-parameters are for training gpt2 from scratch:
* [Gsarti's Pretrain and Fine-tune a T5 model with Flax on GCP](https://github.com/gsarti/t5-flax-gcp)
* [Flax/Jax Community week t5-base-dutch](https://huggingface.co/flax-community/t5-base-dutch)
Created by [Yeb Havinga](https://www.linkedin.com/in/yeb-havinga-86530825/)
|
yhavinga/t5-base-36L-dutch-english-cased
|
yhavinga
| 2022-08-07T12:06:50Z
| 6
| 0
|
transformers
|
[
"transformers",
"jax",
"t5",
"text2text-generation",
"seq2seq",
"nl",
"en",
"dataset:yhavinga/mc4_nl_cleaned",
"arxiv:1910.10683",
"arxiv:2109.10686",
"license:apache-2.0",
"autotrain_compatible",
"region:us"
] |
text2text-generation
| 2022-03-07T08:59:11Z
|
---
language:
- nl
- en
datasets:
- yhavinga/mc4_nl_cleaned
tags:
- t5
- seq2seq
inference: false
license: apache-2.0
---
# t5-base-36L-dutch-english-cased
A [T5](https://ai.googleblog.com/2020/02/exploring-transfer-learning-with-t5.html) sequence to sequence model
pre-trained from scratch on [cleaned Dutch 🇳🇱🇧🇪 mC4 and cleaned English 🇬🇧 C4](https://huggingface.co/datasets/yhavinga/mc4_nl_cleaned).
This **t5 eff** model has **728M** parameters.
It was pre-trained with masked language modeling (denoise token span corruption) objective on the dataset
`mc4_nl_cleaned` config `large_en_nl` for **1** epoch(s) and a duration of **17d15h**,
with a sequence length of **512**, batch size **512** and **212963** total steps (**56B** tokens).
Pre-training evaluation loss and accuracy are **1,05** and **0,76**.
Refer to the evaluation section below for a comparison of the pre-trained models on summarization and translation.
* Pre-trained T5 models need to be finetuned before they can be used for downstream tasks, therefore the inference widget on the right has been turned off.
* For a demo of the Dutch CNN summarization models, head over to the Hugging Face Spaces for
the **[Netherformer 📰](https://huggingface.co/spaces/flax-community/netherformer)** example application!
Please refer to the original T5 papers and Scale Efficiently papers for more information about the T5 architecture
and configs, though it must be noted that this model (t5-base-36L-dutch-english-cased) is unrelated to these projects and not an 'official' checkpoint.
* **[Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/pdf/1910.10683.pdf)** by *Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, Peter J. Liu*.
* **[Scale Efficiently: Insights from Pre-training and Fine-tuning Transformers](https://arxiv.org/abs/2109.10686)** by *Yi Tay, Mostafa Dehghani, Jinfeng Rao, William Fedus, Samira Abnar, Hyung Won Chung, Sharan Narang, Dani Yogatama, Ashish Vaswani, Donald Metzler*.
## Tokenizer
The model uses a cased SentencePiece tokenizer configured with the `Nmt, NFKC, Replace multi-space to single-space` normalizers
and has 32003 tokens.
It was trained on Dutch and English with scripts from the Huggingface Transformers [Flax examples](https://github.com/huggingface/transformers/tree/master/examples/flax/language-modeling).
See [./raw/main/tokenizer.json](tokenizer.json) for details.
## Dataset(s)
All models listed below are pre-trained on
[cleaned Dutch mC4](https://huggingface.co/datasets/yhavinga/mc4_nl_cleaned),
which is the original mC4, except
* Documents that contained words from a selection of the Dutch and English [List of Dirty Naught Obscene and Otherwise Bad Words](https://github.com/LDNOOBW/List-of-Dirty-Naughty-Obscene-and-Otherwise-Bad-Words) are removed
* Sentences with less than 3 words are removed
* Sentences with a word of more than 1000 characters are removed
* Documents with less than 5 sentences are removed
* Documents with "javascript", "lorum ipsum", "terms of use", "privacy policy", "cookie policy", "uses cookies",
"use of cookies", "use cookies", "elementen ontbreken", "deze printversie" are removed.
The Dutch and English models are pre-trained on a 50/50% mix of Dutch mC4 and English C4.
The translation models are fine-tuned on [CCMatrix](https://huggingface.co/datasets/yhavinga/ccmatrix).
## Dutch T5 Models
Three types of [Dutch T5 models have been trained (blog)](https://huggingface.co/spaces/yhavinga/pre-training-dutch-t5-models).
`t5-base-dutch` is the only model with an original T5 config.
The other model types t5-v1.1 and t5-eff have `gated-relu` instead of `relu` as activation function,
and trained with a drop-out of `0.0` unless training would diverge (`t5-v1.1-large-dutch-cased`).
The T5-eff models are models that differ in their number of layers. The table will list
the several dimensions of these models. Not all t5-eff models are efficient, the best example being the inefficient
`t5-xl-4L-dutch-english-cased`.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1.1-large-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-large-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-xl-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-xl-8l-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) |
|:------------------|:----------------|:-----------------------------|:---------------------------|:----------------------------|:-----------------------------------|:----------------------------------------|:-----------------------------|:-------------------------------|:----------------------------------|:-----------------------------------|:--------------------------------------|
| *type* | t5 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5 eff | t5 eff | t5 eff | t5 eff | t5 eff |
| *d_model* | 768 | 768 | 768 | 1024 | 768 | 768 | 512 | 2048 | 768 | 1024 | 1024 |
| *d_ff* | 3072 | 2048 | 2048 | 2816 | 2048 | 2048 | 1920 | 5120 | 2560 | 16384 | 4096 |
| *num_heads* | 12 | 12 | 12 | 16 | 12 | 12 | 8 | 32 | 12 | 32 | 16 |
| *d_kv* | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 128 | 64 |
| *num_layers* | 12 | 12 | 12 | 24 | 12 | 12 | 24 | 4 | 36 | 8 | 8 |
| *num parameters* | 223M | 248M | 248M | 783M | 248M | 248M | 250M | 585M | 729M | 1241M | 335M |
| *feed_forward_proj* | relu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu |
| *dropout* | 0.1 | 0.0 | 0.0 | 0.1 | 0.0 | 0.0 | 0.0 | 0.1 | 0.0 | 0.0 | 0.0 |
| *dataset* | mc4_nl_cleaned | mc4_nl_cleaned full | mc4_nl_cleaned full | mc4_nl_cleaned | mc4_nl_cleaned small_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl |
| *tr. seq len* | 512 | 1024 | 1024 | 512 | 512 | 1024 | 512 | 512 | 512 | 512 | 512 |
| *batch size* | 128 | 64 | 64 | 64 | 128 | 64 | 128 | 512 | 512 | 64 | 128 |
| *total steps* | 527500 | 1014525 | 1210154 | 1120k/2427498 | 2839630 | 1520k/3397024 | 851852 | 212963 | 212963 | 538k/1703705 | 851850 |
| *epochs* | 1 | 2 | 2 | 2 | 10 | 4 | 1 | 1 | 1 | 1 | 1 |
| *duration* | 2d9h | 5d5h | 6d6h | 8d13h | 11d18h | 9d1h | 4d10h | 6d1h | 17d15h | 4d 19h | 3d 23h |
| *optimizer* | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor |
| *lr* | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.009 | 0.005 | 0.005 |
| *warmup* | 10000.0 | 10000.0 | 10000.0 | 10000.0 | 10000.0 | 5000.0 | 20000.0 | 2500.0 | 1000.0 | 1500.0 | 1500.0 |
| *eval loss* | 1,38 | 1,20 | 0,96 | 1,07 | 1,11 | 1,13 | 1,18 | 1,27 | 1,05 | 1,3019 | 1,15 |
| *eval acc* | 0,70 | 0,73 | 0,78 | 0,76 | 0,75 | 0,74 | 0,74 | 0,72 | 0,76 | 0,71 | 0,74 |
## Evaluation
Most models from the list above have been fine-tuned for summarization and translation.
The figure below shows the evaluation scores, where the x-axis shows the translation Bleu score (higher is better)
and y-axis the summarization Rouge1 translation score (higher is better).
Point size is proportional to the model size. Models with faster inference speed are green, slower inference speed is
plotted as bleu.
Evaluation was run on fine-tuned models trained with the following settings:
| | Summarization | Translation |
|---------------:|------------------|-------------------|
| Dataset | CNN Dailymail NL | CCMatrix en -> nl |
| #train samples | 50K | 50K |
| Optimizer | Adam | Adam |
| learning rate | 0.001 | 0.0005 |
| source length | 1024 | 128 |
| target length | 142 | 128 |
|label smoothing | 0.05 | 0.1 |
| #eval samples | 1000 | 1000 |
Note that the amount of training data is limited to a fraction of the total dataset sizes, therefore the scores
below can only be used to compare the 'transfer-learning' strength. The fine-tuned checkpoints for this evaluation
are not saved, since they were trained for comparison of pre-trained models only.
The numbers for summarization are the Rouge scores on 1000 documents from the test split.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) | mt5-base |
|:------------------------|----------------:|-----------------------------:|---------------------------:|-----------------------------------:|----------------------------------------:|-----------------------------:|-------------------------------:|----------------------------------:|--------------------------------------:|-----------:|
| *rouge1* | 33.38 | 33.97 | 34.39 | 33.38 | 34.97 | 34.38 | 30.35 | **35.04** | 34.04 | 33.25 |
| *rouge2* | 13.32 | 13.85 | 13.98 | 13.47 | 14.01 | 13.89 | 11.57 | **14.23** | 13.76 | 12.74 |
| *rougeL* | 24.22 | 24.72 | 25.1 | 24.34 | 24.99 | **25.25** | 22.69 | 25.05 | 24.75 | 23.5 |
| *rougeLsum* | 30.23 | 30.9 | 31.44 | 30.51 | 32.01 | 31.38 | 27.5 | **32.12** | 31.12 | 30.15 |
| *samples_per_second* | 3.18 | 3.02 | 2.99 | 3.22 | 2.97 | 1.57 | 2.8 | 0.61 | **3.27** | 1.22 |
The models below have been evaluated for English to Dutch translation.
Note that the first four models are pre-trained on Dutch only. That they still perform adequate is probably because
the translation direction is English to Dutch.
The numbers reported are the Bleu scores on 1000 documents from the test split.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1.1-large-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-large-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) | mt5-base |
|:-------------------------------|----------------:|-----------------------------:|---------------------------:|----------------------------:|-----------------------------------:|----------------------------------------:|-----------------------------:|-------------------------------:|----------------------------------:|--------------------------------------:|-----------:|
| *precision_ng1* | 74.17 | 78.09 | 77.08 | 72.12 | 77.19 | 78.76 | 78.59 | 77.3 | **79.75** | 78.88 | 73.47 |
| *precision_ng2* | 52.42 | 57.52 | 55.31 | 48.7 | 55.39 | 58.01 | 57.83 | 55.27 | **59.89** | 58.27 | 50.12 |
| *precision_ng3* | 39.55 | 45.2 | 42.54 | 35.54 | 42.25 | 45.13 | 45.02 | 42.06 | **47.4** | 45.95 | 36.59 |
| *precision_ng4* | 30.23 | 36.04 | 33.26 | 26.27 | 32.74 | 35.72 | 35.41 | 32.61 | **38.1** | 36.91 | 27.26 |
| *bp* | 0.99 | 0.98 | 0.97 | 0.98 | 0.98 | 0.98 | 0.98 | 0.97 | 0.98 | 0.98 | 0.98 |
| *score* | 45.88 | 51.21 | 48.31 | 41.59 | 48.17 | 51.31 | 50.82 | 47.83 | **53** | 51.79 | 42.74 |
| *samples_per_second* | **45.19** | 45.05 | 38.67 | 10.12 | 42.19 | 42.61 | 12.85 | 33.74 | 9.07 | 37.86 | 9.03 |
## Translation models
The models `t5-small-24L-dutch-english` and `t5-base-36L-dutch-english` have been fine-tuned for both language
directions on the first 25M samples from CCMatrix, giving a total of 50M training samples.
Evaluation is performed on out-of-sample CCMatrix and also on Tatoeba and Opus Books.
The `_bp` columns list the *brevity penalty*. The `avg_bleu` score is the bleu score
averaged over all three evaluation datasets. The best scores displayed in bold for both translation directions.
| | [t5-base-36L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-base-36L-ccmatrix-multi) | [t5-base-36L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-base-36L-ccmatrix-multi) | [t5-small-24L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-small-24L-ccmatrix-multi) | [t5-small-24L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-small-24L-ccmatrix-multi) |
|:-----------------------|:-----------------------------|:-----------------------------|:------------------------------|:------------------------------|
| *source_lang* | en | nl | en | nl |
| *target_lang* | nl | en | nl | en |
| *source_prefix* | translate English to Dutch: | translate Dutch to English: | translate English to Dutch: | translate Dutch to English: |
| *ccmatrix_bleu* | **56.8** | 62.8 | 57.4 | **63.1** |
| *tatoeba_bleu* | **46.6** | **52.8** | 46.4 | 51.7 |
| *opus_books_bleu* | **13.5** | **24.9** | 12.9 | 23.4 |
| *ccmatrix_bp* | 0.95 | 0.96 | 0.95 | 0.96 |
| *tatoeba_bp* | 0.97 | 0.94 | 0.98 | 0.94 |
| *opus_books_bp* | 0.8 | 0.94 | 0.77 | 0.89 |
| *avg_bleu* | **38.96** | **46.86** | 38.92 | 46.06 |
| *max_source_length* | 128 | 128 | 128 | 128 |
| *max_target_length* | 128 | 128 | 128 | 128 |
| *adam_beta1* | 0.9 | 0.9 | 0.9 | 0.9 |
| *adam_beta2* | 0.997 | 0.997 | 0.997 | 0.997 |
| *weight_decay* | 0.05 | 0.05 | 0.002 | 0.002 |
| *lr* | 5e-05 | 5e-05 | 0.0005 | 0.0005 |
| *label_smoothing_factor* | 0.15 | 0.15 | 0.1 | 0.1 |
| *train_batch_size* | 128 | 128 | 128 | 128 |
| *warmup_steps* | 2000 | 2000 | 2000 | 2000 |
| *total steps* | 390625 | 390625 | 390625 | 390625 |
| *duration* | 4d 5h | 4d 5h | 3d 2h | 3d 2h |
| *num parameters* | 729M | 729M | 250M | 250M |
## Acknowledgements
This project would not have been possible without compute generously provided by Google through the
[TPU Research Cloud](https://sites.research.google/trc/). The HuggingFace 🤗 ecosystem was instrumental in all parts
of the training. Weights & Biases made it possible to keep track of many training sessions
and orchestrate hyper-parameter sweeps with insightful visualizations.
The following repositories where helpful in setting up the TPU-VM,
and getting an idea what sensible hyper-parameters are for training gpt2 from scratch:
* [Gsarti's Pretrain and Fine-tune a T5 model with Flax on GCP](https://github.com/gsarti/t5-flax-gcp)
* [Flax/Jax Community week t5-base-dutch](https://huggingface.co/flax-community/t5-base-dutch)
Created by [Yeb Havinga](https://www.linkedin.com/in/yeb-havinga-86530825/)
|
yhavinga/t5-small-24L-dutch-english
|
yhavinga
| 2022-08-07T12:06:40Z
| 22
| 1
|
transformers
|
[
"transformers",
"jax",
"t5",
"text2text-generation",
"seq2seq",
"nl",
"en",
"dataset:yhavinga/mc4_nl_cleaned",
"arxiv:1910.10683",
"arxiv:2109.10686",
"license:apache-2.0",
"autotrain_compatible",
"region:us"
] |
text2text-generation
| 2022-03-09T20:39:13Z
|
---
language:
- nl
- en
datasets:
- yhavinga/mc4_nl_cleaned
tags:
- t5
- seq2seq
inference: false
license: apache-2.0
---
# t5-small-24L-dutch-english
A [T5](https://ai.googleblog.com/2020/02/exploring-transfer-learning-with-t5.html) sequence to sequence model
pre-trained from scratch on [cleaned Dutch 🇳🇱🇧🇪 mC4 and cleaned English 🇬🇧 C4](https://huggingface.co/datasets/yhavinga/mc4_nl_cleaned).
This **t5 eff** model has **249M** parameters.
It was pre-trained with masked language modeling (denoise token span corruption) objective on the dataset
`mc4_nl_cleaned` config `large_en_nl` for **1** epoch(s) and a duration of **4d10h**,
with a sequence length of **512**, batch size **128** and **851852** total steps (**56B** tokens).
Pre-training evaluation loss and accuracy are **1,18** and **0,74**.
Refer to the evaluation section below for a comparison of the pre-trained models on summarization and translation.
* Pre-trained T5 models need to be finetuned before they can be used for downstream tasks, therefore the inference widget on the right has been turned off.
* For a demo of the Dutch CNN summarization models, head over to the Hugging Face Spaces for
the **[Netherformer 📰](https://huggingface.co/spaces/flax-community/netherformer)** example application!
Please refer to the original T5 papers and Scale Efficiently papers for more information about the T5 architecture
and configs, though it must be noted that this model (t5-small-24L-dutch-english) is unrelated to these projects and not an 'official' checkpoint.
* **[Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/pdf/1910.10683.pdf)** by *Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, Peter J. Liu*.
* **[Scale Efficiently: Insights from Pre-training and Fine-tuning Transformers](https://arxiv.org/abs/2109.10686)** by *Yi Tay, Mostafa Dehghani, Jinfeng Rao, William Fedus, Samira Abnar, Hyung Won Chung, Sharan Narang, Dani Yogatama, Ashish Vaswani, Donald Metzler*.
## Tokenizer
The model uses a cased SentencePiece tokenizer configured with the `Nmt, NFKC, Replace multi-space to single-space` normalizers
and has 32003 tokens.
It was trained on Dutch and English with scripts from the Huggingface Transformers [Flax examples](https://github.com/huggingface/transformers/tree/master/examples/flax/language-modeling).
See [./raw/main/tokenizer.json](tokenizer.json) for details.
## Dataset(s)
All models listed below are pre-trained on
[cleaned Dutch mC4](https://huggingface.co/datasets/yhavinga/mc4_nl_cleaned),
which is the original mC4, except
* Documents that contained words from a selection of the Dutch and English [List of Dirty Naught Obscene and Otherwise Bad Words](https://github.com/LDNOOBW/List-of-Dirty-Naughty-Obscene-and-Otherwise-Bad-Words) are removed
* Sentences with less than 3 words are removed
* Sentences with a word of more than 1000 characters are removed
* Documents with less than 5 sentences are removed
* Documents with "javascript", "lorum ipsum", "terms of use", "privacy policy", "cookie policy", "uses cookies",
"use of cookies", "use cookies", "elementen ontbreken", "deze printversie" are removed.
The Dutch and English models are pre-trained on a 50/50% mix of Dutch mC4 and English C4.
The translation models are fine-tuned on [CCMatrix](https://huggingface.co/datasets/yhavinga/ccmatrix).
## Dutch T5 Models
Three types of [Dutch T5 models have been trained (blog)](https://huggingface.co/spaces/yhavinga/pre-training-dutch-t5-models).
`t5-base-dutch` is the only model with an original T5 config.
The other model types t5-v1.1 and t5-eff have `gated-relu` instead of `relu` as activation function,
and trained with a drop-out of `0.0` unless training would diverge (`t5-v1.1-large-dutch-cased`).
The T5-eff models are models that differ in their number of layers. The table will list
the several dimensions of these models. Not all t5-eff models are efficient, the best example being the inefficient
`t5-xl-4L-dutch-english-cased`.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1.1-large-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-large-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-xl-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-xl-8l-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) |
|:------------------|:----------------|:-----------------------------|:---------------------------|:----------------------------|:-----------------------------------|:----------------------------------------|:-----------------------------|:-------------------------------|:----------------------------------|:-----------------------------------|:--------------------------------------|
| *type* | t5 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5-v1.1 | t5 eff | t5 eff | t5 eff | t5 eff | t5 eff |
| *d_model* | 768 | 768 | 768 | 1024 | 768 | 768 | 512 | 2048 | 768 | 1024 | 1024 |
| *d_ff* | 3072 | 2048 | 2048 | 2816 | 2048 | 2048 | 1920 | 5120 | 2560 | 16384 | 4096 |
| *num_heads* | 12 | 12 | 12 | 16 | 12 | 12 | 8 | 32 | 12 | 32 | 16 |
| *d_kv* | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 64 | 128 | 64 |
| *num_layers* | 12 | 12 | 12 | 24 | 12 | 12 | 24 | 4 | 36 | 8 | 8 |
| *num parameters* | 223M | 248M | 248M | 783M | 248M | 248M | 250M | 585M | 729M | 1241M | 335M |
| *feed_forward_proj* | relu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu | gated-gelu |
| *dropout* | 0.1 | 0.0 | 0.0 | 0.1 | 0.0 | 0.0 | 0.0 | 0.1 | 0.0 | 0.0 | 0.0 |
| *dataset* | mc4_nl_cleaned | mc4_nl_cleaned full | mc4_nl_cleaned full | mc4_nl_cleaned | mc4_nl_cleaned small_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl | mc4_nl_cleaned large_en_nl |
| *tr. seq len* | 512 | 1024 | 1024 | 512 | 512 | 1024 | 512 | 512 | 512 | 512 | 512 |
| *batch size* | 128 | 64 | 64 | 64 | 128 | 64 | 128 | 512 | 512 | 64 | 128 |
| *total steps* | 527500 | 1014525 | 1210154 | 1120k/2427498 | 2839630 | 1520k/3397024 | 851852 | 212963 | 212963 | 538k/1703705 | 851850 |
| *epochs* | 1 | 2 | 2 | 2 | 10 | 4 | 1 | 1 | 1 | 1 | 1 |
| *duration* | 2d9h | 5d5h | 6d6h | 8d13h | 11d18h | 9d1h | 4d10h | 6d1h | 17d15h | 4d 19h | 3d 23h |
| *optimizer* | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor | adafactor |
| *lr* | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.005 | 0.009 | 0.005 | 0.005 |
| *warmup* | 10000.0 | 10000.0 | 10000.0 | 10000.0 | 10000.0 | 5000.0 | 20000.0 | 2500.0 | 1000.0 | 1500.0 | 1500.0 |
| *eval loss* | 1,38 | 1,20 | 0,96 | 1,07 | 1,11 | 1,13 | 1,18 | 1,27 | 1,05 | 1,3019 | 1,15 |
| *eval acc* | 0,70 | 0,73 | 0,78 | 0,76 | 0,75 | 0,74 | 0,74 | 0,72 | 0,76 | 0,71 | 0,74 |
## Evaluation
Most models from the list above have been fine-tuned for summarization and translation.
The figure below shows the evaluation scores, where the x-axis shows the translation Bleu score (higher is better)
and y-axis the summarization Rouge1 translation score (higher is better).
Point size is proportional to the model size. Models with faster inference speed are green, slower inference speed is
plotted as bleu.
Evaluation was run on fine-tuned models trained with the following settings:
| | Summarization | Translation |
|---------------:|------------------|-------------------|
| Dataset | CNN Dailymail NL | CCMatrix en -> nl |
| #train samples | 50K | 50K |
| Optimizer | Adam | Adam |
| learning rate | 0.001 | 0.0005 |
| source length | 1024 | 128 |
| target length | 142 | 128 |
|label smoothing | 0.05 | 0.1 |
| #eval samples | 1000 | 1000 |
Note that the amount of training data is limited to a fraction of the total dataset sizes, therefore the scores
below can only be used to compare the 'transfer-learning' strength. The fine-tuned checkpoints for this evaluation
are not saved, since they were trained for comparison of pre-trained models only.
The numbers for summarization are the Rouge scores on 1000 documents from the test split.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) | mt5-base |
|:------------------------|----------------:|-----------------------------:|---------------------------:|-----------------------------------:|----------------------------------------:|-----------------------------:|-------------------------------:|----------------------------------:|--------------------------------------:|-----------:|
| *rouge1* | 33.38 | 33.97 | 34.39 | 33.38 | 34.97 | 34.38 | 30.35 | **35.04** | 34.04 | 33.25 |
| *rouge2* | 13.32 | 13.85 | 13.98 | 13.47 | 14.01 | 13.89 | 11.57 | **14.23** | 13.76 | 12.74 |
| *rougeL* | 24.22 | 24.72 | 25.1 | 24.34 | 24.99 | **25.25** | 22.69 | 25.05 | 24.75 | 23.5 |
| *rougeLsum* | 30.23 | 30.9 | 31.44 | 30.51 | 32.01 | 31.38 | 27.5 | **32.12** | 31.12 | 30.15 |
| *samples_per_second* | 3.18 | 3.02 | 2.99 | 3.22 | 2.97 | 1.57 | 2.8 | 0.61 | **3.27** | 1.22 |
The models below have been evaluated for English to Dutch translation.
Note that the first four models are pre-trained on Dutch only. That they still perform adequate is probably because
the translation direction is English to Dutch.
The numbers reported are the Bleu scores on 1000 documents from the test split.
| | [t5-base-dutch](https://huggingface.co/yhavinga/t5-base-dutch) | [t5-v1.1-base-dutch-uncased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-uncased) | [t5-v1.1-base-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-base-dutch-cased) | [t5-v1.1-large-dutch-cased](https://huggingface.co/yhavinga/t5-v1.1-large-dutch-cased) | [t5-v1_1-base-dutch-english-cased](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased) | [t5-v1_1-base-dutch-english-cased-1024](https://huggingface.co/yhavinga/t5-v1_1-base-dutch-english-cased-1024) | [t5-small-24L-dutch-english](https://huggingface.co/yhavinga/t5-small-24L-dutch-english) | [t5-xl-4L-dutch-english-cased](https://huggingface.co/yhavinga/t5-xl-4L-dutch-english-cased) | [t5-base-36L-dutch-english-cased](https://huggingface.co/yhavinga/t5-base-36L-dutch-english-cased) | [t5-eff-large-8l-dutch-english-cased](https://huggingface.co/yhavinga/t5-eff-large-8l-dutch-english-cased) | mt5-base |
|:-------------------------------|----------------:|-----------------------------:|---------------------------:|----------------------------:|-----------------------------------:|----------------------------------------:|-----------------------------:|-------------------------------:|----------------------------------:|--------------------------------------:|-----------:|
| *precision_ng1* | 74.17 | 78.09 | 77.08 | 72.12 | 77.19 | 78.76 | 78.59 | 77.3 | **79.75** | 78.88 | 73.47 |
| *precision_ng2* | 52.42 | 57.52 | 55.31 | 48.7 | 55.39 | 58.01 | 57.83 | 55.27 | **59.89** | 58.27 | 50.12 |
| *precision_ng3* | 39.55 | 45.2 | 42.54 | 35.54 | 42.25 | 45.13 | 45.02 | 42.06 | **47.4** | 45.95 | 36.59 |
| *precision_ng4* | 30.23 | 36.04 | 33.26 | 26.27 | 32.74 | 35.72 | 35.41 | 32.61 | **38.1** | 36.91 | 27.26 |
| *bp* | 0.99 | 0.98 | 0.97 | 0.98 | 0.98 | 0.98 | 0.98 | 0.97 | 0.98 | 0.98 | 0.98 |
| *score* | 45.88 | 51.21 | 48.31 | 41.59 | 48.17 | 51.31 | 50.82 | 47.83 | **53** | 51.79 | 42.74 |
| *samples_per_second* | **45.19** | 45.05 | 38.67 | 10.12 | 42.19 | 42.61 | 12.85 | 33.74 | 9.07 | 37.86 | 9.03 |
## Translation models
The models `t5-small-24L-dutch-english` and `t5-base-36L-dutch-english` have been fine-tuned for both language
directions on the first 25M samples from CCMatrix, giving a total of 50M training samples.
Evaluation is performed on out-of-sample CCMatrix and also on Tatoeba and Opus Books.
The `_bp` columns list the *brevity penalty*. The `avg_bleu` score is the bleu score
averaged over all three evaluation datasets. The best scores displayed in bold for both translation directions.
| | [t5-base-36L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-base-36L-ccmatrix-multi) | [t5-base-36L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-base-36L-ccmatrix-multi) | [t5-small-24L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-small-24L-ccmatrix-multi) | [t5-small-24L-ccmatrix-multi](https://huggingface.co/yhavinga/t5-small-24L-ccmatrix-multi) |
|:-----------------------|:-----------------------------|:-----------------------------|:------------------------------|:------------------------------|
| *source_lang* | en | nl | en | nl |
| *target_lang* | nl | en | nl | en |
| *source_prefix* | translate English to Dutch: | translate Dutch to English: | translate English to Dutch: | translate Dutch to English: |
| *ccmatrix_bleu* | **56.8** | 62.8 | 57.4 | **63.1** |
| *tatoeba_bleu* | **46.6** | **52.8** | 46.4 | 51.7 |
| *opus_books_bleu* | **13.5** | **24.9** | 12.9 | 23.4 |
| *ccmatrix_bp* | 0.95 | 0.96 | 0.95 | 0.96 |
| *tatoeba_bp* | 0.97 | 0.94 | 0.98 | 0.94 |
| *opus_books_bp* | 0.8 | 0.94 | 0.77 | 0.89 |
| *avg_bleu* | **38.96** | **46.86** | 38.92 | 46.06 |
| *max_source_length* | 128 | 128 | 128 | 128 |
| *max_target_length* | 128 | 128 | 128 | 128 |
| *adam_beta1* | 0.9 | 0.9 | 0.9 | 0.9 |
| *adam_beta2* | 0.997 | 0.997 | 0.997 | 0.997 |
| *weight_decay* | 0.05 | 0.05 | 0.002 | 0.002 |
| *lr* | 5e-05 | 5e-05 | 0.0005 | 0.0005 |
| *label_smoothing_factor* | 0.15 | 0.15 | 0.1 | 0.1 |
| *train_batch_size* | 128 | 128 | 128 | 128 |
| *warmup_steps* | 2000 | 2000 | 2000 | 2000 |
| *total steps* | 390625 | 390625 | 390625 | 390625 |
| *duration* | 4d 5h | 4d 5h | 3d 2h | 3d 2h |
| *num parameters* | 729M | 729M | 250M | 250M |
## Acknowledgements
This project would not have been possible without compute generously provided by Google through the
[TPU Research Cloud](https://sites.research.google/trc/). The HuggingFace 🤗 ecosystem was instrumental in all parts
of the training. Weights & Biases made it possible to keep track of many training sessions
and orchestrate hyper-parameter sweeps with insightful visualizations.
The following repositories where helpful in setting up the TPU-VM,
and getting an idea what sensible hyper-parameters are for training gpt2 from scratch:
* [Gsarti's Pretrain and Fine-tune a T5 model with Flax on GCP](https://github.com/gsarti/t5-flax-gcp)
* [Flax/Jax Community week t5-base-dutch](https://huggingface.co/flax-community/t5-base-dutch)
Created by [Yeb Havinga](https://www.linkedin.com/in/yeb-havinga-86530825/)
|
Mahmoud7/ppo-LunarLander-v2
|
Mahmoud7
| 2022-08-07T08:49:54Z
| 3
| 0
|
stable-baselines3
|
[
"stable-baselines3",
"LunarLander-v2",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-07T08:49:23Z
|
---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- metrics:
- type: mean_reward
value: 216.14 +/- 19.57
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
---
# **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
...
```
|
osanseviero/llama_or_alpaca
|
osanseviero
| 2022-08-07T08:23:43Z
| 0
| 0
|
fastai
|
[
"fastai",
"region:us"
] | null | 2022-08-07T08:23:33Z
|
---
tags:
- fastai
---
# Amazing!
🥳 Congratulations on hosting your fastai model on the Hugging Face Hub!
# Some next steps
1. Fill out this model card with more information (see the template below and the [documentation here](https://huggingface.co/docs/hub/model-repos))!
2. Create a demo in Gradio or Streamlit using 🤗 Spaces ([documentation here](https://huggingface.co/docs/hub/spaces)).
3. Join the fastai community on the [Fastai Discord](https://discord.com/invite/YKrxeNn)!
Greetings fellow fastlearner 🤝! Don't forget to delete this content from your model card.
---
# Model card
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
|
osanseviero/is_it_a_llama
|
osanseviero
| 2022-08-07T07:02:11Z
| 0
| 0
|
fastai
|
[
"fastai",
"region:us"
] | null | 2022-08-07T07:02:02Z
|
---
tags:
- fastai
---
# Amazing!
🥳 Congratulations on hosting your fastai model on the Hugging Face Hub!
# Some next steps
1. Fill out this model card with more information (see the template below and the [documentation here](https://huggingface.co/docs/hub/model-repos))!
2. Create a demo in Gradio or Streamlit using 🤗 Spaces ([documentation here](https://huggingface.co/docs/hub/spaces)).
3. Join the fastai community on the [Fastai Discord](https://discord.com/invite/YKrxeNn)!
Greetings fellow fastlearner 🤝! Don't forget to delete this content from your model card.
---
# Model card
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
|
pritoms/opt-350m-opty-350m-lectures
|
pritoms
| 2022-08-07T06:08:51Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"opt",
"text-generation",
"generated_from_trainer",
"license:other",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
text-generation
| 2022-08-07T05:52:41Z
|
---
license: other
tags:
- generated_from_trainer
model-index:
- name: opt-350m-opty-350m-lectures
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. -->
# opt-350m-opty-350m-lectures
This model is a fine-tuned version of [facebook/opt-350m](https://huggingface.co/facebook/opt-350m) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 2.3830
## 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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3.0
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| No log | 1.0 | 5 | 2.7828 |
| No log | 2.0 | 10 | 2.4889 |
| No log | 3.0 | 15 | 2.3830 |
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
Mozart-coder/DNA_BigBird_3
|
Mozart-coder
| 2022-08-07T04:37:56Z
| 5
| 1
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"big_bird",
"fill-mask",
"generated_from_trainer",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
fill-mask
| 2022-08-07T03:30:19Z
|
---
tags:
- generated_from_trainer
model-index:
- name: dna_3-Pretrained
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. -->
# dna_3-Pretrained
This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6003
## 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: 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: 10
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 0.6057 | 1.0 | 62 | 0.6032 |
| 0.5983 | 2.0 | 124 | 0.6002 |
| 0.5934 | 3.0 | 186 | 0.5992 |
| 0.5853 | 4.0 | 248 | 0.5979 |
| 0.5799 | 5.0 | 310 | 0.5982 |
| 0.5769 | 6.0 | 372 | 0.5999 |
| 0.5754 | 7.0 | 434 | 0.5996 |
| 0.5713 | 8.0 | 496 | 0.6000 |
| 0.5687 | 9.0 | 558 | 0.5988 |
| 0.5666 | 10.0 | 620 | 0.6003 |
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
JTH/twitter_classification
|
JTH
| 2022-08-07T04:28:59Z
| 4
| 1
|
transformers
|
[
"transformers",
"pytorch",
"distilbert",
"text-classification",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text-classification
| 2022-08-06T15:46:21Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: twitter_classification
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. -->
# twitter_classification
Label mapping:
```
{
'Business': 2,
'Civil Society': 4,
'Government': 5,
'Individual': 0,
'News & Info': 1,
'Research & Academia': 3
}
```
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: 5
### Training results
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
sherwinseah/Fine-tuned-T5-for-MCQGenerator
|
sherwinseah
| 2022-08-07T04:18:04Z
| 10
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"t5",
"text2text-generation",
"generated_from_trainer",
"dataset:squad_modified_for_t5_qg",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
text2text-generation
| 2022-08-07T03:07:33Z
|
---
tags:
- generated_from_trainer
datasets:
- squad_modified_for_t5_qg
model-index:
- name: Fine-tuned-T5-for-MCQGenerator
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. -->
# Fine-tuned-T5-for-MCQGenerator
This model is a fine-tuned version of [ramsrigouthamg/t5_squad_v1](https://huggingface.co/ramsrigouthamg/t5_squad_v1) on the squad_modified_for_t5_qg 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: 0.0001
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 7
### Training results
### Framework versions
- Transformers 4.21.1
- Pytorch 1.11.0+cpu
- Datasets 2.4.0
- Tokenizers 0.12.1
|
voice/wav2vec2-large-xlsr-common1000asli-demo-colab-dd
|
voice
| 2022-08-07T03:39:10Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"dataset:common_voice",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] |
automatic-speech-recognition
| 2022-08-02T05:56:51Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- common_voice
model-index:
- name: wav2vec2-large-xlsr-common1000asli-demo-colab-dd
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-xlsr-common1000asli-demo-colab-dd
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the common_voice dataset.
It achieves the following results on the evaluation set:
- Loss: 1.0671
- Wer: 0.5268
## 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: 128
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 1000
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:------:|:-----:|:---------------:|:------:|
| 10.4469 | 10.53 | 400 | 2.9884 | 0.9788 |
| 1.6358 | 21.05 | 800 | 0.5968 | 0.6578 |
| 0.3404 | 31.58 | 1200 | 0.5348 | 0.5757 |
| 0.2218 | 42.11 | 1600 | 0.5945 | 0.5659 |
| 0.1676 | 52.63 | 2000 | 0.6318 | 0.5753 |
| 0.1442 | 63.16 | 2400 | 0.6395 | 0.5559 |
| 0.1237 | 73.68 | 2800 | 0.6981 | 0.5522 |
| 0.1132 | 84.21 | 3200 | 0.7174 | 0.5428 |
| 0.1056 | 94.74 | 3600 | 0.7120 | 0.5531 |
| 0.0942 | 105.26 | 4000 | 0.7865 | 0.5529 |
| 0.0967 | 115.79 | 4400 | 0.7796 | 0.5546 |
| 0.0854 | 126.32 | 4800 | 0.7392 | 0.5507 |
| 0.0816 | 136.84 | 5200 | 0.8173 | 0.5575 |
| 0.0748 | 147.37 | 5600 | 0.8164 | 0.5550 |
| 0.0691 | 157.89 | 6000 | 0.8061 | 0.5444 |
| 0.0654 | 168.42 | 6400 | 0.8098 | 0.5524 |
| 0.0627 | 178.95 | 6800 | 0.8527 | 0.5655 |
| 0.0653 | 189.47 | 7200 | 0.8210 | 0.5412 |
| 0.0639 | 200.0 | 7600 | 0.8619 | 0.5602 |
| 0.0614 | 210.53 | 8000 | 0.8453 | 0.5606 |
| 0.0647 | 221.05 | 8400 | 0.8248 | 0.5564 |
| 0.0611 | 231.58 | 8800 | 0.8323 | 0.5637 |
| 0.0606 | 242.11 | 9200 | 0.8754 | 0.5654 |
| 0.0587 | 252.63 | 9600 | 0.8684 | 0.5528 |
| 0.0524 | 263.16 | 10000 | 0.8798 | 0.5556 |
| 0.0499 | 273.68 | 10400 | 0.8593 | 0.5553 |
| 0.0466 | 284.21 | 10800 | 0.9079 | 0.5520 |
| 0.0505 | 294.74 | 11200 | 0.8680 | 0.5607 |
| 0.0517 | 305.26 | 11600 | 0.8709 | 0.5557 |
| 0.0522 | 315.79 | 12000 | 0.8687 | 0.5570 |
| 0.0453 | 326.32 | 12400 | 0.8585 | 0.5614 |
| 0.047 | 336.84 | 12800 | 0.9249 | 0.5581 |
| 0.0431 | 347.37 | 13200 | 0.8934 | 0.5543 |
| 0.0454 | 357.89 | 13600 | 0.8837 | 0.5583 |
| 0.0472 | 368.42 | 14000 | 0.9070 | 0.5565 |
| 0.0431 | 378.95 | 14400 | 0.9202 | 0.5526 |
| 0.0404 | 389.47 | 14800 | 0.9234 | 0.5543 |
| 0.0386 | 400.0 | 15200 | 0.9056 | 0.5549 |
| 0.0372 | 410.53 | 15600 | 0.9901 | 0.5493 |
| 0.0376 | 421.05 | 16000 | 0.9109 | 0.5460 |
| 0.0365 | 431.58 | 16400 | 0.9313 | 0.5487 |
| 0.0347 | 442.11 | 16800 | 0.9027 | 0.5496 |
| 0.0361 | 452.63 | 17200 | 0.9614 | 0.5457 |
| 0.0323 | 463.16 | 17600 | 0.9782 | 0.5558 |
| 0.0325 | 473.68 | 18000 | 0.9549 | 0.5481 |
| 0.032 | 484.21 | 18400 | 0.9781 | 0.5431 |
| 0.0289 | 494.74 | 18800 | 0.9840 | 0.5463 |
| 0.0292 | 505.26 | 19200 | 0.9397 | 0.5357 |
| 0.0276 | 515.79 | 19600 | 0.9228 | 0.5467 |
| 0.0283 | 526.32 | 20000 | 0.9683 | 0.5394 |
| 0.0281 | 536.84 | 20400 | 0.9783 | 0.5479 |
| 0.026 | 547.37 | 20800 | 0.9663 | 0.5472 |
| 0.0288 | 557.89 | 21200 | 0.9424 | 0.5426 |
| 0.0275 | 568.42 | 21600 | 0.9788 | 0.5435 |
| 0.0264 | 578.95 | 22000 | 0.9703 | 0.5473 |
| 0.0259 | 589.47 | 22400 | 0.9994 | 0.5446 |
| 0.0243 | 600.0 | 22800 | 0.9637 | 0.5590 |
| 0.0251 | 610.53 | 23200 | 0.9577 | 0.5457 |
| 0.0222 | 621.05 | 23600 | 0.9780 | 0.5419 |
| 0.0227 | 631.58 | 24000 | 0.9582 | 0.5417 |
| 0.0222 | 642.11 | 24400 | 0.9847 | 0.5432 |
| 0.0214 | 652.63 | 24800 | 1.0171 | 0.5449 |
| 0.022 | 663.16 | 25200 | 0.9819 | 0.5430 |
| 0.0202 | 673.68 | 25600 | 0.9737 | 0.5413 |
| 0.0187 | 684.21 | 26000 | 0.9977 | 0.5440 |
| 0.0213 | 694.74 | 26400 | 0.9919 | 0.5464 |
| 0.0197 | 705.26 | 26800 | 0.9769 | 0.5357 |
| 0.0183 | 715.79 | 27200 | 0.9964 | 0.5377 |
| 0.0187 | 726.32 | 27600 | 0.9973 | 0.5341 |
| 0.0191 | 736.84 | 28000 | 0.9970 | 0.5399 |
| 0.0183 | 747.37 | 28400 | 1.0179 | 0.5371 |
| 0.0176 | 757.89 | 28800 | 1.0020 | 0.5440 |
| 0.018 | 768.42 | 29200 | 0.9992 | 0.5394 |
| 0.0157 | 778.95 | 29600 | 1.0502 | 0.5397 |
| 0.0165 | 789.47 | 30000 | 1.0463 | 0.5397 |
| 0.0147 | 800.0 | 30400 | 1.0363 | 0.5430 |
| 0.0153 | 810.53 | 30800 | 0.9890 | 0.5407 |
| 0.0145 | 821.05 | 31200 | 1.0139 | 0.5369 |
| 0.0143 | 831.58 | 31600 | 1.0260 | 0.5346 |
| 0.0141 | 842.11 | 32000 | 1.0277 | 0.5361 |
| 0.0139 | 852.63 | 32400 | 1.0639 | 0.5335 |
| 0.0132 | 863.16 | 32800 | 1.0661 | 0.5314 |
| 0.013 | 873.68 | 33200 | 1.0537 | 0.5335 |
| 0.0126 | 884.21 | 33600 | 1.0433 | 0.5347 |
| 0.0121 | 894.74 | 34000 | 1.0275 | 0.5326 |
| 0.0128 | 905.26 | 34400 | 1.0405 | 0.5327 |
| 0.0112 | 915.79 | 34800 | 1.0626 | 0.5296 |
| 0.0115 | 926.32 | 35200 | 1.0583 | 0.5284 |
| 0.0109 | 936.84 | 35600 | 1.0494 | 0.5287 |
| 0.0113 | 947.37 | 36000 | 1.0655 | 0.5294 |
| 0.0104 | 957.89 | 36400 | 1.0723 | 0.5269 |
| 0.0108 | 968.42 | 36800 | 1.0680 | 0.5267 |
| 0.0104 | 978.95 | 37200 | 1.0707 | 0.5261 |
| 0.0108 | 989.47 | 37600 | 1.0649 | 0.5268 |
| 0.0103 | 1000.0 | 38000 | 1.0671 | 0.5268 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu102
- Datasets 1.13.3
- Tokenizers 0.10.3
|
arinze/t5_finetuned_xsum_hr
|
arinze
| 2022-08-07T02:11:35Z
| 17
| 0
|
transformers
|
[
"transformers",
"pytorch",
"t5",
"text2text-generation",
"generated_from_trainer",
"dataset:xsum",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
text2text-generation
| 2022-08-05T14:50:04Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- xsum
metrics:
- rouge
model-index:
- name: t5_finetuned_xsum_hr
results:
- task:
name: Sequence-to-sequence Language Modeling
type: text2text-generation
dataset:
name: xsum
type: xsum
config: default
split: train
args: default
metrics:
- name: Rouge1
type: rouge
value: 29.1239
---
<!-- 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_finetuned_xsum_hr
This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on the xsum dataset.
It achieves the following results on the evaluation set:
- Loss: 2.4224
- Rouge1: 29.1239
- Rouge2: 8.3165
- Rougel: 22.9946
- Rougelsum: 22.9996
- Gen Len: 18.819
## 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
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|:-------------:|:-----:|:-----:|:---------------:|:-------:|:------:|:-------:|:---------:|:-------:|
| 2.6839 | 1.0 | 12753 | 2.4474 | 28.721 | 8.0397 | 22.6568 | 22.6527 | 18.8324 |
| 2.6425 | 2.0 | 25506 | 2.4224 | 29.1239 | 8.3165 | 22.9946 | 22.9996 | 18.819 |
### Framework versions
- Transformers 4.21.0
- Pytorch 1.12.0
- Datasets 2.4.0
- Tokenizers 0.12.1
|
huggingtweets/carterhiggins
|
huggingtweets
| 2022-08-07T01:19:20Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
text-generation
| 2022-07-07T01:12:44Z
|
---
language: en
thumbnail: http://www.huggingtweets.com/carterhiggins/1659835083112/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1296229510510030849/0dyqAcul_400x400.jpg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">Carter Higgins</div>
<div style="text-align: center; font-size: 14px;">@carterhiggins</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from Carter Higgins.
| Data | Carter Higgins |
| --- | --- |
| Tweets downloaded | 3246 |
| Retweets | 539 |
| Short tweets | 568 |
| Tweets kept | 2139 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/3jjhg6sk/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @carterhiggins's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/14ee5p5o) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/14ee5p5o/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/carterhiggins')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
AdamAbate1/wav2vec2-base-finetuned-ks
|
AdamAbate1
| 2022-08-07T00:53:49Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"wav2vec2",
"audio-classification",
"generated_from_trainer",
"dataset:superb",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] |
audio-classification
| 2022-08-06T11:29:56Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- superb
metrics:
- accuracy
model-index:
- name: wav2vec2-base-finetuned-ks
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-base-finetuned-ks
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the superb dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0993
- Accuracy: 0.9812
## 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
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 0.7367 | 1.0 | 399 | 0.6341 | 0.8819 |
| 0.3087 | 2.0 | 798 | 0.1900 | 0.9771 |
| 0.1979 | 3.0 | 1197 | 0.1232 | 0.9800 |
| 0.171 | 4.0 | 1596 | 0.1057 | 0.9794 |
| 0.1253 | 5.0 | 1995 | 0.0993 | 0.9812 |
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0
- Datasets 2.4.0
- Tokenizers 0.12.1
|
huggingartists/gspd
|
huggingartists
| 2022-08-06T23:30:23Z
| 5
| 0
|
transformers
|
[
"transformers",
"pytorch",
"jax",
"gpt2",
"text-generation",
"huggingartists",
"lyrics",
"lm-head",
"causal-lm",
"en",
"dataset:huggingartists/gspd",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
text-generation
| 2022-03-02T23:29:05Z
|
---
language: en
datasets:
- huggingartists/gspd
tags:
- huggingartists
- lyrics
- lm-head
- causal-lm
widget:
- text: "I am"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:DISPLAY_1; margin-left: auto; margin-right: auto; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://images.genius.com/9409ae2b38424a74b42cb1e4bb66b83a.1000x1000x1.jpg')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 HuggingArtists Model 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">GSPD</div>
<a href="https://genius.com/artists/gspd">
<div style="text-align: center; font-size: 14px;">@gspd</div>
</a>
</div>
I was made with [huggingartists](https://github.com/AlekseyKorshuk/huggingartists).
Create your own bot based on your favorite artist with [the demo](https://colab.research.google.com/github/AlekseyKorshuk/huggingartists/blob/master/huggingartists-demo.ipynb)!
## How does it work?
To understand how the model was developed, check the [W&B report](https://wandb.ai/huggingartists/huggingartists/reportlist).
## Training data
The model was trained on lyrics from GSPD.
Dataset is available [here](https://huggingface.co/datasets/huggingartists/gspd).
And can be used with:
```python
from datasets import load_dataset
dataset = load_dataset("huggingartists/gspd")
```
[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/3jof0sex/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on GSPD's lyrics.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/2nxhrny4) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/2nxhrny4/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingartists/gspd')
generator("I am", num_return_sequences=5)
```
Or with Transformers library:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead
tokenizer = AutoTokenizer.from_pretrained("huggingartists/gspd")
model = AutoModelWithLMHead.from_pretrained("huggingartists/gspd")
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Aleksey Korshuk*
[](https://github.com/AlekseyKorshuk)
[](https://twitter.com/intent/follow?screen_name=alekseykorshuk)
[](https://t.me/joinchat/_CQ04KjcJ-4yZTky)
For more details, visit the project repository.
[](https://github.com/AlekseyKorshuk/huggingartists)
|
huggingartists/morgenshtern
|
huggingartists
| 2022-08-06T23:27:49Z
| 11
| 0
|
transformers
|
[
"transformers",
"pytorch",
"jax",
"gpt2",
"text-generation",
"huggingartists",
"lyrics",
"lm-head",
"causal-lm",
"en",
"dataset:huggingartists/morgenshtern",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
text-generation
| 2022-03-02T23:29:05Z
|
---
language: en
datasets:
- huggingartists/morgenshtern
tags:
- huggingartists
- lyrics
- lm-head
- causal-lm
widget:
- text: "I am"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:DISPLAY_1; margin-left: auto; margin-right: auto; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://images.genius.com/cdfb190640789439daae426c799e5e32.1000x1000x1.jpg')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 HuggingArtists Model 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">MORGENSHTERN</div>
<a href="https://genius.com/artists/morgenshtern">
<div style="text-align: center; font-size: 14px;">@morgenshtern</div>
</a>
</div>
I was made with [huggingartists](https://github.com/AlekseyKorshuk/huggingartists).
Create your own bot based on your favorite artist with [the demo](https://colab.research.google.com/github/AlekseyKorshuk/huggingartists/blob/master/huggingartists-demo.ipynb)!
## How does it work?
To understand how the model was developed, check the [W&B report](https://wandb.ai/huggingartists/huggingartists/reportlist).
## Training data
The model was trained on lyrics from MORGENSHTERN.
Dataset is available [here](https://huggingface.co/datasets/huggingartists/morgenshtern).
And can be used with:
```python
from datasets import load_dataset
dataset = load_dataset("huggingartists/morgenshtern")
```
[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/lmrnk6sz/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on MORGENSHTERN's lyrics.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/1m2jynlh) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/1m2jynlh/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingartists/morgenshtern')
generator("I am", num_return_sequences=5)
```
Or with Transformers library:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead
tokenizer = AutoTokenizer.from_pretrained("huggingartists/morgenshtern")
model = AutoModelWithLMHead.from_pretrained("huggingartists/morgenshtern")
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Aleksey Korshuk*
[](https://github.com/AlekseyKorshuk)
[](https://twitter.com/intent/follow?screen_name=alekseykorshuk)
[](https://t.me/joinchat/_CQ04KjcJ-4yZTky)
For more details, visit the project repository.
[](https://github.com/AlekseyKorshuk/huggingartists)
|
theojolliffe/bart-paraphrase-v2-e1
|
theojolliffe
| 2022-08-06T22:09:16Z
| 7
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"bart",
"text2text-generation",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text2text-generation
| 2022-08-06T21:11:36Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: bart-paraphrase-v2-e1
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. -->
# bart-paraphrase-v2-e1
This model is a fine-tuned version of [eugenesiow/bart-paraphrase](https://huggingface.co/eugenesiow/bart-paraphrase) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1536
- Rouge1: 72.6807
- Rouge2: 68.1957
- Rougel: 71.1787
- Rougelsum: 71.3376
- Gen Len: 19.5698
## 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: 2
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:-------:|
| 0.0682 | 1.0 | 7093 | 0.1536 | 72.6807 | 68.1957 | 71.1787 | 71.3376 | 19.5698 |
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
huggingtweets/chai_ste
|
huggingtweets
| 2022-08-06T21:51:04Z
| 3
| 0
|
transformers
|
[
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
text-generation
| 2022-08-06T20:38:06Z
|
---
language: en
thumbnail: http://www.huggingtweets.com/chai_ste/1659822641053/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1479595267800322048/Aqqb82wz_400x400.png')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">ste 🍊</div>
<div style="text-align: center; font-size: 14px;">@chai_ste</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from ste 🍊.
| Data | ste 🍊 |
| --- | --- |
| Tweets downloaded | 3192 |
| Retweets | 293 |
| Short tweets | 497 |
| Tweets kept | 2402 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/1pklrw4p/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @chai_ste's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/ef8xv0wo) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/ef8xv0wo/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/chai_ste')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
abcp4/a2c-HalfCheetahBulletEnv-v0
|
abcp4
| 2022-08-06T20:44:56Z
| 3
| 0
|
stable-baselines3
|
[
"stable-baselines3",
"HalfCheetahBulletEnv-v0",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-06T20:10:39Z
|
---
library_name: stable-baselines3
tags:
- HalfCheetahBulletEnv-v0
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: A2C
results:
- metrics:
- type: mean_reward
value: -891.37 +/- 122.88
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: HalfCheetahBulletEnv-v0
type: HalfCheetahBulletEnv-v0
---
# **A2C** Agent playing **HalfCheetahBulletEnv-v0**
This is a trained model of a **A2C** agent playing **HalfCheetahBulletEnv-v0**
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
...
```
|
bassemessam/wav2vec2-large-xls-r-300m-arabic-saudi-colab
|
bassemessam
| 2022-08-06T20:37:36Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] |
automatic-speech-recognition
| 2022-08-06T19:09:00Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: wav2vec2-large-xls-r-300m-arabic-saudi-colab
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-xls-r-300m-arabic-saudi-colab
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) 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: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- 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: 5
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.21.1
- Pytorch 1.10.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
Yousef7/U
|
Yousef7
| 2022-08-06T19:50:32Z
| 0
| 0
| null |
[
"region:us"
] | null | 2022-08-06T19:43:17Z
|
pip install rubika
pip install colorama
pip install rubpy
|
jackoyoungblood/q-FrozenLake-v1-4x4-noSlippery
|
jackoyoungblood
| 2022-08-06T19:13:27Z
| 0
| 0
| null |
[
"FrozenLake-v1-4x4-no_slippery",
"q-learning",
"reinforcement-learning",
"custom-implementation",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-06T19:13:22Z
|
---
tags:
- FrozenLake-v1-4x4-no_slippery
- q-learning
- reinforcement-learning
- custom-implementation
model-index:
- name: q-FrozenLake-v1-4x4-noSlippery
results:
- metrics:
- type: mean_reward
value: 1.00 +/- 0.00
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: FrozenLake-v1-4x4-no_slippery
type: FrozenLake-v1-4x4-no_slippery
---
# **Q-Learning** Agent playing **FrozenLake-v1**
This is a trained model of a **Q-Learning** agent playing **FrozenLake-v1** .
## Usage
```python
model = load_from_hub(repo_id="/q-FrozenLake-v1-4x4-noSlippery", filename="q-learning.pkl")
# Don't forget to check if you need to add additional attributes (is_slippery=False etc)
env = gym.make(model["env_id"])
evaluate_agent(env, model["max_steps"], model["n_eval_episodes"], model["qtable"], model["eval_seed"])
```
|
JTH/results
|
JTH
| 2022-08-06T15:36:15Z
| 10
| 0
|
transformers
|
[
"transformers",
"pytorch",
"distilbert",
"text-classification",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text-classification
| 2022-07-29T16:43:15Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: results
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. -->
# results
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: 5
### Training results
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
Einmalumdiewelt/DistilBART_CNN_GNAD
|
Einmalumdiewelt
| 2022-08-06T15:16:21Z
| 5
| 0
|
transformers
|
[
"transformers",
"pytorch",
"bart",
"text2text-generation",
"generated_from_trainer",
"de",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text2text-generation
| 2022-08-05T19:47:30Z
|
---
language:
- de
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: DistilBART_CNN_GNAD
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. -->
# DistilBART_CNN_GNAD
This model is a fine-tuned version of [Einmalumdiewelt/DistilBART_CNN_GNAD](https://huggingface.co/Einmalumdiewelt/DistilBART_CNN_GNAD) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 2.8723
- Rouge1: 27.4368
- Rouge2: 8.159
- Rougel: 18.1359
- Rougelsum: 23.1339
- Gen Len: 91.5847
## 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
- num_epochs: 10.0
### Training results
### Framework versions
- Transformers 4.22.0.dev0
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
L-macc/autotrain-Biomedical_sc_summ-1217846142
|
L-macc
| 2022-08-06T14:50:47Z
| 15
| 0
|
transformers
|
[
"transformers",
"pytorch",
"longt5",
"text2text-generation",
"autotrain",
"summarization",
"unk",
"dataset:L-macc/autotrain-data-Biomedical_sc_summ",
"model-index",
"co2_eq_emissions",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
summarization
| 2022-08-04T07:45:06Z
|
---
tags:
- autotrain
- summarization
language:
- unk
widget:
- text: "I love AutoTrain \U0001F917"
datasets:
- L-macc/autotrain-data-Biomedical_sc_summ
co2_eq_emissions:
emissions: 16.211223325053414
model-index:
- name: L-macc/autotrain-Biomedical_sc_summ-1217846142
results:
- task:
type: summarization
name: Summarization
dataset:
name: Blaise-g/SumPubmed
type: Blaise-g/SumPubmed
config: Blaise-g--SumPubmed
split: test
metrics:
- name: ROUGE-1
type: rouge
value: 38.2324
verified: true
- name: ROUGE-2
type: rouge
value: 12.2914
verified: true
- name: ROUGE-L
type: rouge
value: 21.4512
verified: true
- name: ROUGE-LSUM
type: rouge
value: 34.3031
verified: true
- name: loss
type: loss
value: 2.3273825645446777
verified: true
- name: gen_len
type: gen_len
value: 130.3934
verified: true
---
# Model Trained Using AutoTrain
- Problem type: Summarization
- Model ID: 1217846142
- CO2 Emissions (in grams): 16.2112
## Validation Metrics
- Loss: 2.159
- Rouge1: 40.236
- Rouge2: 12.161
- RougeL: 23.255
- RougeLsum: 35.138
- Gen Len: 121.504
## Usage
You can use cURL to access this model:
```
$ curl -X POST -H "Authorization: Bearer YOUR_HUGGINGFACE_API_KEY" -H "Content-Type: application/json" -d '{"inputs": "I love AutoTrain"}' https://api-inference.huggingface.co/L-macc/autotrain-Biomedical_sc_summ-1217846142
```
|
qiaoyi/Comment_Summarization4DesignTutor
|
qiaoyi
| 2022-08-06T14:34:39Z
| 14
| 0
|
transformers
|
[
"transformers",
"pytorch",
"jax",
"rust",
"t5",
"text2text-generation",
"summarization",
"translation",
"en",
"fr",
"ro",
"de",
"dataset:c4",
"arxiv:1805.12471",
"arxiv:1708.00055",
"arxiv:1704.05426",
"arxiv:1606.05250",
"arxiv:1808.09121",
"arxiv:1810.12885",
"arxiv:1905.10044",
"arxiv:1910.10683",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
translation
| 2022-08-06T14:06:42Z
|
---
language:
- en
- fr
- ro
- de
datasets:
- c4
tags:
- summarization
- translation
license: apache-2.0
---
[Google's T5](https://ai.googleblog.com/2020/02/exploring-transfer-learning-with-t5.html)
## PreTraining
The model was pre-trained on a on a **multi-task mixture of unsupervised (1.) and supervised tasks (2.)**.
Thereby, the following datasets were being used for (1.) and (2.):
1. **Datasets used for Unsupervised denoising objective**:
- [C4](https://huggingface.co/datasets/c4)
- [Wiki-DPR](https://huggingface.co/datasets/wiki_dpr)
2. **Datasets used for Supervised text-to-text language modeling objective**
- Sentence acceptability judgment
- CoLA [Warstadt et al., 2018](https://arxiv.org/abs/1805.12471)
- Sentiment analysis
- SST-2 [Socher et al., 2013](https://nlp.stanford.edu/~socherr/EMNLP2013_RNTN.pdf)
- Paraphrasing/sentence similarity
- MRPC [Dolan and Brockett, 2005](https://aclanthology.org/I05-5002)
- STS-B [Ceret al., 2017](https://arxiv.org/abs/1708.00055)
- QQP [Iyer et al., 2017](https://quoradata.quora.com/First-Quora-Dataset-Release-Question-Pairs)
- Natural language inference
- MNLI [Williams et al., 2017](https://arxiv.org/abs/1704.05426)
- QNLI [Rajpurkar et al.,2016](https://arxiv.org/abs/1606.05250)
- RTE [Dagan et al., 2005](https://link.springer.com/chapter/10.1007/11736790_9)
- CB [De Marneff et al., 2019](https://semanticsarchive.net/Archive/Tg3ZGI2M/Marneffe.pdf)
- Sentence completion
- COPA [Roemmele et al., 2011](https://www.researchgate.net/publication/221251392_Choice_of_Plausible_Alternatives_An_Evaluation_of_Commonsense_Causal_Reasoning)
- Word sense disambiguation
- WIC [Pilehvar and Camacho-Collados, 2018](https://arxiv.org/abs/1808.09121)
- Question answering
- MultiRC [Khashabi et al., 2018](https://aclanthology.org/N18-1023)
- ReCoRD [Zhang et al., 2018](https://arxiv.org/abs/1810.12885)
- BoolQ [Clark et al., 2019](https://arxiv.org/abs/1905.10044)
## All T5 checkpoints
Other Community Checkpoints: [here](https://huggingface.co/models?search=t5)
## Paper
For more information, please take a look at the original paper.
Paper: [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/pdf/1910.10683.pdf)
Authors: *Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, Peter J. Liu*
**Abstract**
Transfer learning, where a model is first pre-trained on a data-rich task before being fine-tuned on a downstream task, has emerged as a powerful technique in natural language processing (NLP). The effectiveness of transfer learning has given rise to a diversity of approaches, methodology, and practice. In this paper, we explore the landscape of transfer learning techniques for NLP by introducing a unified framework that converts every language problem into a text-to-text format. Our systematic study compares pre-training objectives, architectures, unlabeled datasets, transfer approaches, and other factors on dozens of language understanding tasks. By combining the insights from our exploration with scale and our new “Colossal Clean Crawled Corpus”, we achieve state-of-the-art results on many benchmarks covering summarization, question answering, text classification, and more. To facilitate future work on transfer learning for NLP, we release our dataset, pre-trained models, and code.
|
paola-md/recipe-reg
|
paola-md
| 2022-08-06T11:51:52Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"roberta",
"text-classification",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text-classification
| 2022-08-05T12:31:23Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: recipe-reg
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-reg
This model is a fine-tuned version of [distilroberta-base](https://huggingface.co/distilroberta-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 3.3717
- Rmse: 1.8362
- Mse: 3.3717
- Mae: 1.6145
## 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: 15
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rmse | Mse | Mae |
|:-------------:|:-----:|:------:|:---------------:|:------:|:------:|:------:|
| 3.3247 | 1.0 | 12809 | 3.3717 | 1.8362 | 3.3717 | 1.6145 |
| 3.3238 | 2.0 | 25618 | 3.3722 | 1.8363 | 3.3722 | 1.6145 |
| 3.3217 | 3.0 | 38427 | 3.3718 | 1.8362 | 3.3718 | 1.6145 |
| 3.3215 | 4.0 | 51236 | 3.3754 | 1.8372 | 3.3754 | 1.6145 |
| 3.3203 | 5.0 | 64045 | 3.3721 | 1.8363 | 3.3721 | 1.6145 |
| 3.3199 | 6.0 | 76854 | 3.3731 | 1.8366 | 3.3731 | 1.6145 |
| 3.319 | 7.0 | 89663 | 3.3731 | 1.8366 | 3.3731 | 1.6145 |
| 3.3188 | 8.0 | 102472 | 3.3717 | 1.8362 | 3.3717 | 1.6145 |
### Framework versions
- Transformers 4.19.0.dev0
- Pytorch 1.9.0+cu111
- Datasets 2.4.0
- Tokenizers 0.12.1
|
nvidia/mit-b2
|
nvidia
| 2022-08-06T10:26:08Z
| 20,071
| 4
|
transformers
|
[
"transformers",
"pytorch",
"tf",
"segformer",
"image-classification",
"vision",
"dataset:imagenet_1k",
"arxiv:2105.15203",
"license:other",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
image-classification
| 2022-03-02T23:29:05Z
|
---
license: other
tags:
- vision
datasets:
- imagenet_1k
widget:
- src: https://huggingface.co/datasets/hf-internal-testing/fixtures_ade20k/resolve/main/ADE_val_00000001.jpg
example_title: House
- src: https://huggingface.co/datasets/hf-internal-testing/fixtures_ade20k/resolve/main/ADE_val_00000002.jpg
example_title: Castle
---
# SegFormer (b2-sized) encoder pre-trained-only
SegFormer encoder fine-tuned on Imagenet-1k. It was introduced in the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Xie et al. and first released in [this repository](https://github.com/NVlabs/SegFormer).
Disclaimer: The team releasing SegFormer did not write a model card for this model so this model card has been written by the Hugging Face team.
## Model description
SegFormer consists of a hierarchical Transformer encoder and a lightweight all-MLP decode head to achieve great results on semantic segmentation benchmarks such as ADE20K and Cityscapes. The hierarchical Transformer is first pre-trained on ImageNet-1k, after which a decode head is added and fine-tuned altogether on a downstream dataset.
This repository only contains the pre-trained hierarchical Transformer, hence it can be used for fine-tuning purposes.
## Intended uses & limitations
You can use the model for fine-tuning of semantic segmentation. See the [model hub](https://huggingface.co/models?other=segformer) to look for fine-tuned versions on a task that interests you.
### How to use
Here is how to use this model to classify an image of the COCO 2017 dataset into one of the 1,000 ImageNet classes:
```python
from transformers import SegformerFeatureExtractor, SegformerForImageClassification
from PIL import Image
import requests
url = "http://images.cocodataset.org/val2017/000000039769.jpg"
image = Image.open(requests.get(url, stream=True).raw)
feature_extractor = SegformerFeatureExtractor.from_pretrained("nvidia/mit-b2")
model = SegformerForImageClassification.from_pretrained("nvidia/mit-b2")
inputs = feature_extractor(images=image, return_tensors="pt")
outputs = model(**inputs)
logits = outputs.logits
# model predicts one of the 1000 ImageNet classes
predicted_class_idx = logits.argmax(-1).item()
print("Predicted class:", model.config.id2label[predicted_class_idx])
```
For more code examples, we refer to the [documentation](https://huggingface.co/transformers/model_doc/segformer.html#).
### License
The license for this model can be found [here](https://github.com/NVlabs/SegFormer/blob/master/LICENSE).
### BibTeX entry and citation info
```bibtex
@article{DBLP:journals/corr/abs-2105-15203,
author = {Enze Xie and
Wenhai Wang and
Zhiding Yu and
Anima Anandkumar and
Jose M. Alvarez and
Ping Luo},
title = {SegFormer: Simple and Efficient Design for Semantic Segmentation with
Transformers},
journal = {CoRR},
volume = {abs/2105.15203},
year = {2021},
url = {https://arxiv.org/abs/2105.15203},
eprinttype = {arXiv},
eprint = {2105.15203},
timestamp = {Wed, 02 Jun 2021 11:46:42 +0200},
biburl = {https://dblp.org/rec/journals/corr/abs-2105-15203.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
```
|
nvidia/segformer-b5-finetuned-ade-640-640
|
nvidia
| 2022-08-06T10:25:55Z
| 210,546
| 39
|
transformers
|
[
"transformers",
"pytorch",
"tf",
"segformer",
"vision",
"image-segmentation",
"dataset:scene_parse_150",
"arxiv:2105.15203",
"license:other",
"endpoints_compatible",
"region:us"
] |
image-segmentation
| 2022-03-02T23:29:05Z
|
---
license: other
tags:
- vision
- image-segmentation
datasets:
- scene_parse_150
widget:
- src: https://huggingface.co/datasets/hf-internal-testing/fixtures_ade20k/resolve/main/ADE_val_00000001.jpg
example_title: House
- src: https://huggingface.co/datasets/hf-internal-testing/fixtures_ade20k/resolve/main/ADE_val_00000002.jpg
example_title: Castle
---
# SegFormer (b5-sized) model fine-tuned on ADE20k
SegFormer model fine-tuned on ADE20k at resolution 640x640. It was introduced in the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Xie et al. and first released in [this repository](https://github.com/NVlabs/SegFormer).
Disclaimer: The team releasing SegFormer did not write a model card for this model so this model card has been written by the Hugging Face team.
## Model description
SegFormer consists of a hierarchical Transformer encoder and a lightweight all-MLP decode head to achieve great results on semantic segmentation benchmarks such as ADE20K and Cityscapes. The hierarchical Transformer is first pre-trained on ImageNet-1k, after which a decode head is added and fine-tuned altogether on a downstream dataset.
## Intended uses & limitations
You can use the raw model for semantic segmentation. See the [model hub](https://huggingface.co/models?other=segformer) to look for fine-tuned versions on a task that interests you.
### How to use
Here is how to use this model to classify an image of the COCO 2017 dataset into one of the 1,000 ImageNet classes:
```python
from transformers import SegformerFeatureExtractor, SegformerForSemanticSegmentation
from PIL import Image
import requests
feature_extractor = SegformerFeatureExtractor.from_pretrained("nvidia/segformer-b5-finetuned-ade-512-512")
model = SegformerForSemanticSegmentation.from_pretrained("nvidia/segformer-b5-finetuned-ade-512-512")
url = "http://images.cocodataset.org/val2017/000000039769.jpg"
image = Image.open(requests.get(url, stream=True).raw)
inputs = feature_extractor(images=image, return_tensors="pt")
outputs = model(**inputs)
logits = outputs.logits # shape (batch_size, num_labels, height/4, width/4)
```
For more code examples, we refer to the [documentation](https://huggingface.co/transformers/model_doc/segformer.html#).
### License
The license for this model can be found [here](https://github.com/NVlabs/SegFormer/blob/master/LICENSE).
### BibTeX entry and citation info
```bibtex
@article{DBLP:journals/corr/abs-2105-15203,
author = {Enze Xie and
Wenhai Wang and
Zhiding Yu and
Anima Anandkumar and
Jose M. Alvarez and
Ping Luo},
title = {SegFormer: Simple and Efficient Design for Semantic Segmentation with
Transformers},
journal = {CoRR},
volume = {abs/2105.15203},
year = {2021},
url = {https://arxiv.org/abs/2105.15203},
eprinttype = {arXiv},
eprint = {2105.15203},
timestamp = {Wed, 02 Jun 2021 11:46:42 +0200},
biburl = {https://dblp.org/rec/journals/corr/abs-2105-15203.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
```
|
nvidia/mit-b1
|
nvidia
| 2022-08-06T10:25:12Z
| 38,555
| 1
|
transformers
|
[
"transformers",
"pytorch",
"tf",
"segformer",
"image-classification",
"vision",
"dataset:imagenet_1k",
"arxiv:2105.15203",
"license:other",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
image-classification
| 2022-03-02T23:29:05Z
|
---
license: other
tags:
- vision
datasets:
- imagenet_1k
widget:
- src: https://huggingface.co/datasets/hf-internal-testing/fixtures_ade20k/resolve/main/ADE_val_00000001.jpg
example_title: House
- src: https://huggingface.co/datasets/hf-internal-testing/fixtures_ade20k/resolve/main/ADE_val_00000002.jpg
example_title: Castle
---
# SegFormer (b1-sized) encoder pre-trained-only
SegFormer encoder fine-tuned on Imagenet-1k. It was introduced in the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Xie et al. and first released in [this repository](https://github.com/NVlabs/SegFormer).
Disclaimer: The team releasing SegFormer did not write a model card for this model so this model card has been written by the Hugging Face team.
## Model description
SegFormer consists of a hierarchical Transformer encoder and a lightweight all-MLP decode head to achieve great results on semantic segmentation benchmarks such as ADE20K and Cityscapes. The hierarchical Transformer is first pre-trained on ImageNet-1k, after which a decode head is added and fine-tuned altogether on a downstream dataset.
This repository only contains the pre-trained hierarchical Transformer, hence it can be used for fine-tuning purposes.
## Intended uses & limitations
You can use the model for fine-tuning of semantic segmentation. See the [model hub](https://huggingface.co/models?other=segformer) to look for fine-tuned versions on a task that interests you.
### How to use
Here is how to use this model to classify an image of the COCO 2017 dataset into one of the 1,000 ImageNet classes:
```python
from transformers import SegformerFeatureExtractor, SegformerForImageClassification
from PIL import Image
import requests
url = "http://images.cocodataset.org/val2017/000000039769.jpg"
image = Image.open(requests.get(url, stream=True).raw)
feature_extractor = SegformerFeatureExtractor.from_pretrained("nvidia/mit-b1")
model = SegformerForImageClassification.from_pretrained("nvidia/mit-b1")
inputs = feature_extractor(images=image, return_tensors="pt")
outputs = model(**inputs)
logits = outputs.logits
# model predicts one of the 1000 ImageNet classes
predicted_class_idx = logits.argmax(-1).item()
print("Predicted class:", model.config.id2label[predicted_class_idx])
```
For more code examples, we refer to the [documentation](https://huggingface.co/transformers/model_doc/segformer.html#).
### License
The license for this model can be found [here](https://github.com/NVlabs/SegFormer/blob/master/LICENSE).
### BibTeX entry and citation info
```bibtex
@article{DBLP:journals/corr/abs-2105-15203,
author = {Enze Xie and
Wenhai Wang and
Zhiding Yu and
Anima Anandkumar and
Jose M. Alvarez and
Ping Luo},
title = {SegFormer: Simple and Efficient Design for Semantic Segmentation with
Transformers},
journal = {CoRR},
volume = {abs/2105.15203},
year = {2021},
url = {https://arxiv.org/abs/2105.15203},
eprinttype = {arXiv},
eprint = {2105.15203},
timestamp = {Wed, 02 Jun 2021 11:46:42 +0200},
biburl = {https://dblp.org/rec/journals/corr/abs-2105-15203.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
```
|
maan909/unisumm
|
maan909
| 2022-08-06T09:14:26Z
| 17
| 0
|
transformers
|
[
"transformers",
"jax",
"rust",
"bart",
"text2text-generation",
"summarization",
"en",
"dataset:cnn_dailymail",
"dataset:xsum",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
summarization
| 2022-08-01T06:32:51Z
|
---
language: en
tags:
- summarization
license: apache-2.0
datasets:
- cnn_dailymail
- xsum
thumbnail: https://huggingface.co/front/thumbnails/distilbart_medium.png
---
### Usage
This checkpoint should be loaded into `BartForConditionalGeneration.from_pretrained`. See the [BART docs](https://huggingface.co/transformers/model_doc/bart.html?#transformers.BartForConditionalGeneration) for more information.
### Metrics for DistilBART models
| Model Name | MM Params | Inference Time (MS) | Speedup | Rouge 2 | Rouge-L |
|:---------------------------|------------:|----------------------:|----------:|----------:|----------:|
| distilbart-xsum-12-1 | 222 | 90 | 2.54 | 18.31 | 33.37 |
| distilbart-xsum-6-6 | 230 | 132 | 1.73 | 20.92 | 35.73 |
| distilbart-xsum-12-3 | 255 | 106 | 2.16 | 21.37 | 36.39 |
| distilbart-xsum-9-6 | 268 | 136 | 1.68 | 21.72 | 36.61 |
| bart-large-xsum (baseline) | 406 | 229 | 1 | 21.85 | 36.50 |
| distilbart-xsum-12-6 | 306 | 137 | 1.68 | 22.12 | 36.99 |
| bart-large-cnn (baseline) | 406 | 381 | 1 | 21.06 | 30.63 |
| distilbart-12-3-cnn | 255 | 214 | 1.78 | 20.57 | 30.00 |
| distilbart-12-6-cnn | 306 | 307 | 1.24 | 21.26 | 30.59 |
| distilbart-6-6-cnn | 230 | 182 | 2.09 | 20.17 | 29.70 |
|
trtd56/ppo-CartPole
|
trtd56
| 2022-08-05T23:42:39Z
| 0
| 0
| null |
[
"tensorboard",
"CartPole-v1",
"ppo",
"deep-reinforcement-learning",
"reinforcement-learning",
"custom-implementation",
"deep-rl-class",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-05T23:42:30Z
|
---
tags:
- CartPole-v1
- ppo
- deep-reinforcement-learning
- reinforcement-learning
- custom-implementation
- deep-rl-class
model-index:
- name: PPO
results:
- metrics:
- type: mean_reward
value: 166.60 +/- 82.10
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: CartPole-v1
type: CartPole-v1
---
# PPO Agent Playing CartPole-v1
This is a trained model of a PPO agent playing CartPole-v1.
To learn to code your own PPO agent and train it Unit 8 of the Deep Reinforcement Learning Class: https://github.com/huggingface/deep-rl-class/tree/main/unit8
# Hyperparameters
```python
{'exp_name': 'ppo'
'seed': 1
'torch_deterministic': True
'cuda': True
'track': False
'wandb_project_name': 'cleanRL'
'wandb_entity': None
'capture_video': False
'env_id': 'CartPole-v1'
'total_timesteps': 50000
'learning_rate': 0.00025
'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': 'trtd56/ppo-CartPole'
'batch_size': 512
'minibatch_size': 128}
```
|
nikhilnayak/lyrics.ai
|
nikhilnayak
| 2022-08-05T23:02:19Z
| 2
| 0
|
transformers
|
[
"transformers",
"pytorch",
"endpoints_compatible",
"region:us"
] | null | 2022-08-05T18:55:44Z
|
---
widget:
- text: "Jens Peter Hansen kommer fra Danmark"
- pipeline-tag: text-generation
---
|
huggingtweets/chipflake
|
huggingtweets
| 2022-08-05T22:38:19Z
| 3
| 0
|
transformers
|
[
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] |
text-generation
| 2022-08-05T22:37:32Z
|
---
language: en
thumbnail: http://www.huggingtweets.com/chipflake/1659739094566/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1338527116644806663/XkhjylPj_400x400.jpg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">the one, singular chip</div>
<div style="text-align: center; font-size: 14px;">@chipflake</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from the one, singular chip.
| Data | the one, singular chip |
| --- | --- |
| Tweets downloaded | 1214 |
| Retweets | 80 |
| Short tweets | 220 |
| Tweets kept | 914 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/38uu3y9r/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @chipflake's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/40p3p4l2) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/40p3p4l2/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/chipflake')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
Hazam/distilbert-base-uncased-finetuned-imdb
|
Hazam
| 2022-08-05T22:00:52Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"fill-mask",
"generated_from_trainer",
"dataset:imdb",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
fill-mask
| 2022-08-05T21:02:02Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- imdb
model-index:
- name: distilbert-base-uncased-finetuned-imdb
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-imdb
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: 2.4721
## 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: 3.0
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 2.7086 | 1.0 | 157 | 2.4898 |
| 2.5796 | 2.0 | 314 | 2.4230 |
| 2.5269 | 3.0 | 471 | 2.4354 |
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
1-800-BAD-CODE/morsecode_en_quartznet_10x5
|
1-800-BAD-CODE
| 2022-08-05T21:55:46Z
| 9
| 1
|
nemo
|
[
"nemo",
"region:us"
] | null | 2022-08-05T21:31:47Z
|
## Model Overview
This model is a Morse Code recognition model. It was trained with the package at https://github.com/1-800-BAD-CODE/MorseCodeToolkit.
This model accepts as input audio signals sampled at 8khz containing Morse code. The model produces the English transcription of the Morse code signal.
For inference, only the base NeMo package needs to be installed because this is just an ASR model trained to decode Morse code signals rather than speech signals.
## How to Use this Model
With NeMo is installed, this model can be used to run inference on Morse code audio files.
### Automatically instantiate the model
```python
import nemo.collections.asr as nemo_asr
asr_model = nemo_asr.models.ASRModel.from_pretrained("1-800-BAD-CODE/morsecode_en_quartznet_10x5")
```
### Transcribing using Python
First, let's download an example Morse code audio file from Wikipedia:
```
wget https://upload.wikimedia.org/wikipedia/commons/0/04/Wikipedia-Morse.ogg
```
Then simply do:
```
asr_model.transcribe(['Wikipedia-Morse.ogg'])
['WELCOME TO WIKIPEDIA, THE FREE ENCYCLOPEDIA THAT ANYONE CAN EDIT.']
```
## Limitations
This model was trained on synthetic Morse code data generated by https://github.com/1-800-BAD-CODE/MorseCodeToolkit. Any Morse code generated with parameters outside of the range of the parameters used to generate the training data will not be well recognized by the model.
|
skr1125/xlm-roberta-base-finetuned-panx-all
|
skr1125
| 2022-08-05T21:43:23Z
| 5
| 0
|
transformers
|
[
"transformers",
"pytorch",
"xlm-roberta",
"token-classification",
"generated_from_trainer",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
token-classification
| 2022-08-05T21:20:17Z
|
---
license: mit
tags:
- generated_from_trainer
metrics:
- f1
model-index:
- name: xlm-roberta-base-finetuned-panx-all
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. -->
# xlm-roberta-base-finetuned-panx-all
This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1781
- F1: 0.8538
## 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: 24
- eval_batch_size: 24
- 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 | F1 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 0.2986 | 1.0 | 835 | 0.1929 | 0.8055 |
| 0.1547 | 2.0 | 1670 | 0.1804 | 0.8380 |
| 0.1003 | 3.0 | 2505 | 0.1781 | 0.8538 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.12.0+cu113
- Datasets 1.16.1
- Tokenizers 0.10.3
|
skr1125/xlm-roberta-base-finetuned-panx-en
|
skr1125
| 2022-08-05T21:18:38Z
| 5
| 0
|
transformers
|
[
"transformers",
"pytorch",
"xlm-roberta",
"token-classification",
"generated_from_trainer",
"dataset:xtreme",
"license:mit",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
token-classification
| 2022-08-05T21:03:08Z
|
---
license: mit
tags:
- generated_from_trainer
datasets:
- xtreme
metrics:
- f1
model-index:
- name: xlm-roberta-base-finetuned-panx-en
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: xtreme
type: xtreme
args: PAN-X.en
metrics:
- name: F1
type: f1
value: 0.7032474804031354
---
<!-- 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. -->
# xlm-roberta-base-finetuned-panx-en
This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the xtreme dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3932
- F1: 0.7032
## 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: 24
- eval_batch_size: 24
- 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 | F1 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 1.1504 | 1.0 | 50 | 0.5992 | 0.4786 |
| 0.5147 | 2.0 | 100 | 0.4307 | 0.6468 |
| 0.3717 | 3.0 | 150 | 0.3932 | 0.7032 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.12.0+cu113
- Datasets 1.16.1
- Tokenizers 0.10.3
|
abdulmatinomotoso/multi_news_article_title_5000
|
abdulmatinomotoso
| 2022-08-05T21:05:21Z
| 10
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"pegasus",
"text2text-generation",
"generated_from_trainer",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
text2text-generation
| 2022-08-05T19:01:50Z
|
---
tags:
- generated_from_trainer
model-index:
- name: multi_news_article_title_5000
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. -->
# multi_news_article_title_5000
This model is a fine-tuned version of [google/pegasus-multi_news](https://huggingface.co/google/pegasus-multi_news) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2547
## 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: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 16
- 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: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 0.3601 | 1.52 | 500 | 0.2547 |
### Framework versions
- Transformers 4.18.0
- Pytorch 1.11.0
- Datasets 2.1.0
- Tokenizers 0.12.1
|
skr1125/xlm-roberta-base-finetuned-panx-fr
|
skr1125
| 2022-08-05T20:47:17Z
| 3
| 0
|
transformers
|
[
"transformers",
"pytorch",
"xlm-roberta",
"token-classification",
"generated_from_trainer",
"dataset:xtreme",
"license:mit",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
token-classification
| 2022-08-05T20:30:21Z
|
---
license: mit
tags:
- generated_from_trainer
datasets:
- xtreme
metrics:
- f1
model-index:
- name: xlm-roberta-base-finetuned-panx-fr
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: xtreme
type: xtreme
args: PAN-X.fr
metrics:
- name: F1
type: f1
value: 0.8330673475901692
---
<!-- 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. -->
# xlm-roberta-base-finetuned-panx-fr
This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the xtreme dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2851
- F1: 0.8331
## 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: 24
- eval_batch_size: 24
- 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 | F1 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 0.5989 | 1.0 | 191 | 0.3200 | 0.7991 |
| 0.2617 | 2.0 | 382 | 0.2897 | 0.8236 |
| 0.1672 | 3.0 | 573 | 0.2851 | 0.8331 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.12.0+cu113
- Datasets 1.16.1
- Tokenizers 0.10.3
|
bennyeatspants/Cat-vs-Dog
|
bennyeatspants
| 2022-08-05T20:03:42Z
| 0
| 2
|
fastai
|
[
"fastai",
"region:us"
] | null | 2022-08-05T18:21:41Z
|
---
tags:
- fastai
---
# Amazing!
🥳 Congratulations on hosting your fastai model on the Hugging Face Hub!
# Some next steps
1. Fill out this model card with more information (see the template below and the [documentation here](https://huggingface.co/docs/hub/model-repos))!
2. Create a demo in Gradio or Streamlit using 🤗 Spaces ([documentation here](https://huggingface.co/docs/hub/spaces)).
3. Join the fastai community on the [Fastai Discord](https://discord.com/invite/YKrxeNn)!
Greetings fellow fastlearner 🤝! Don't forget to delete this content from your model card.
---
# Model card
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
|
jackoyoungblood/ppo-LunarLander-v2c
|
jackoyoungblood
| 2022-08-05T19:46:18Z
| 3
| 0
|
stable-baselines3
|
[
"stable-baselines3",
"LunarLander-v2",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-07-29T23:03:33Z
|
---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- metrics:
- type: mean_reward
value: 267.50 +/- 18.00
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
---
# **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
...
```
|
embedding-data/distilroberta-base-sentence-transformer
|
embedding-data
| 2022-08-05T19:12:47Z
| 4,095
| 1
|
sentence-transformers
|
[
"sentence-transformers",
"pytorch",
"roberta",
"feature-extraction",
"sentence-similarity",
"transformers",
"dataset:embedding-data/QQP_triplets",
"autotrain_compatible",
"text-embeddings-inference",
"endpoints_compatible",
"region:us"
] |
sentence-similarity
| 2022-08-05T18:36:25Z
|
---
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- feature-extraction
- sentence-similarity
- transformers
datasets:
- embedding-data/QQP_triplets
---
# embedding-data/distilroberta-base-sentence-transformer
This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.
<!--- Describe your model here -->
## Usage (Sentence-Transformers)
Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
```
pip install -U sentence-transformers
```
Then you can use the model like this:
```python
from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]
model = SentenceTransformer('embedding-data/distilroberta-base-sentence-transformer')
embeddings = model.encode(sentences)
print(embeddings)
```
## Usage (HuggingFace Transformers)
Without [sentence-transformers](https://www.SBERT.net), you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.
```python
from transformers import AutoTokenizer, AutoModel
import torch
#Mean Pooling - Take attention mask into account for correct averaging
def mean_pooling(model_output, attention_mask):
token_embeddings = model_output[0] #First element of model_output contains all token embeddings
input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
# Sentences we want sentence embeddings for
sentences = ['This is an example sentence', 'Each sentence is converted']
# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained('embedding-data/distilroberta-base-sentence-transformer')
model = AutoModel.from_pretrained('embedding-data/distilroberta-base-sentence-transformer')
# Tokenize sentences
encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
# Compute token embeddings
with torch.no_grad():
model_output = model(**encoded_input)
# Perform pooling. In this case, mean pooling.
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
print("Sentence embeddings:")
print(sentence_embeddings)
```
## Evaluation Results
<!--- Describe how your model was evaluated -->
For an automated evaluation of this model, see the *Sentence Embeddings Benchmark*: [https://seb.sbert.net](https://seb.sbert.net?model_name=embedding-data/distilroberta-base-sentence-transformer)
## Training
The model was trained with the parameters:
**DataLoader**:
`torch.utils.data.dataloader.DataLoader` of length 7 with parameters:
```
{'batch_size': 16, 'sampler': 'torch.utils.data.sampler.RandomSampler', 'batch_sampler': 'torch.utils.data.sampler.BatchSampler'}
```
**Loss**:
`sentence_transformers.losses.TripletLoss.TripletLoss` with parameters:
```
{'distance_metric': 'TripletDistanceMetric.EUCLIDEAN', 'triplet_margin': 5}
```
Parameters of the fit()-Method:
```
{
"epochs": 10,
"evaluation_steps": 0,
"evaluator": "NoneType",
"max_grad_norm": 1,
"optimizer_class": "<class 'torch.optim.adamw.AdamW'>",
"optimizer_params": {
"lr": 2e-05
},
"scheduler": "WarmupLinear",
"steps_per_epoch": null,
"warmup_steps": 7,
"weight_decay": 0.01
}
```
## Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: RobertaModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
)
```
## Citing & Authors
<!--- Describe where people can find more information -->
|
skr1125/xlm-roberta-base-finetuned-panx-de
|
skr1125
| 2022-08-05T17:50:14Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"xlm-roberta",
"token-classification",
"generated_from_trainer",
"dataset:xtreme",
"license:mit",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
token-classification
| 2022-08-02T01:50:37Z
|
---
license: mit
tags:
- generated_from_trainer
datasets:
- xtreme
metrics:
- f1
model-index:
- name: xlm-roberta-base-finetuned-panx-de
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: xtreme
type: xtreme
args: PAN-X.de
metrics:
- name: F1
type: f1
value: 0.863677639046538
---
<!-- 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. -->
# xlm-roberta-base-finetuned-panx-de
This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the xtreme dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1343
- F1: 0.8637
## 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: 24
- eval_batch_size: 24
- 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 | F1 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 0.2578 | 1.0 | 525 | 0.1562 | 0.8273 |
| 0.1297 | 2.0 | 1050 | 0.1330 | 0.8474 |
| 0.0809 | 3.0 | 1575 | 0.1343 | 0.8637 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.12.0+cu113
- Datasets 1.16.1
- Tokenizers 0.10.3
|
Evelyn18/roberta-base-bne-ROBERTaBECAS
|
Evelyn18
| 2022-08-05T16:38:34Z
| 4
| 0
|
transformers
|
[
"transformers",
"pytorch",
"tensorboard",
"roberta",
"question-answering",
"generated_from_trainer",
"dataset:becasv2",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] |
question-answering
| 2022-08-05T15:49:11Z
|
---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- becasv2
model-index:
- name: roberta-base-bne-ROBERTaBECAS
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. -->
# roberta-base-bne-ROBERTaBECAS
This model is a fine-tuned version of [BSC-TeMU/roberta-base-bne](https://huggingface.co/BSC-TeMU/roberta-base-bne) on the becasv2 dataset.
It achieves the following results on the evaluation set:
- Loss: 2.5760
## 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: 11
- eval_batch_size: 11
- 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 |
|:-------------:|:-----:|:----:|:---------------:|
| No log | 1.0 | 6 | 4.3366 |
| No log | 2.0 | 12 | 3.1395 |
| No log | 3.0 | 18 | 2.6092 |
| No log | 4.0 | 24 | 2.5084 |
| No log | 5.0 | 30 | 2.5760 |
### Framework versions
- Transformers 4.21.1
- Pytorch 1.9.0+cu111
- Datasets 2.4.0
- Tokenizers 0.12.1
|
roberta-sgariglia/ppo-LunarLander-v2
|
roberta-sgariglia
| 2022-08-05T16:04:03Z
| 7
| 1
|
stable-baselines3
|
[
"stable-baselines3",
"LunarLander-v2",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-05T16:03:39Z
|
---
library_name: stable-baselines3
tags:
- LunarLander-v2
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: PPO
results:
- metrics:
- type: mean_reward
value: 230.59 +/- 38.03
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: LunarLander-v2
type: LunarLander-v2
---
# **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
...
```
|
mrm8488/dqn-BeamRiderNoFrameskip-v4
|
mrm8488
| 2022-08-05T14:05:47Z
| 11
| 1
|
stable-baselines3
|
[
"stable-baselines3",
"BeamRiderNoFrameskip-v4",
"deep-reinforcement-learning",
"reinforcement-learning",
"model-index",
"region:us"
] |
reinforcement-learning
| 2022-08-05T14:03:10Z
|
---
library_name: stable-baselines3
tags:
- BeamRiderNoFrameskip-v4
- deep-reinforcement-learning
- reinforcement-learning
- stable-baselines3
model-index:
- name: DQN
results:
- metrics:
- type: mean_reward
value: 2519.60 +/- 1330.45
name: mean_reward
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: BeamRiderNoFrameskip-v4
type: BeamRiderNoFrameskip-v4
---
# **DQN** Agent playing **BeamRiderNoFrameskip-v4**
This is a trained model of a **DQN** agent playing **BeamRiderNoFrameskip-v4**
using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3)
and the [RL Zoo](https://github.com/DLR-RM/rl-baselines3-zoo).
The RL Zoo is a training framework for Stable Baselines3
reinforcement learning agents,
with hyperparameter optimization and pre-trained agents included.
## Usage (with SB3 RL Zoo)
RL Zoo: https://github.com/DLR-RM/rl-baselines3-zoo<br/>
SB3: https://github.com/DLR-RM/stable-baselines3<br/>
SB3 Contrib: https://github.com/Stable-Baselines-Team/stable-baselines3-contrib
```
# Download model and save it into the logs/ folder
python -m utils.load_from_hub --algo dqn --env BeamRiderNoFrameskip-v4 -orga mrm8488 -f logs/
python enjoy.py --algo dqn --env BeamRiderNoFrameskip-v4 -f logs/
```
## Training (with the RL Zoo)
```
python train.py --algo dqn --env BeamRiderNoFrameskip-v4 -f logs/
# Upload the model and generate video (when possible)
python -m utils.push_to_hub --algo dqn --env BeamRiderNoFrameskip-v4 -f logs/ -orga mrm8488
```
## Hyperparameters
```python
OrderedDict([('batch_size', 1024),
('buffer_size', 100000),
('env_wrapper',
['stable_baselines3.common.atari_wrappers.AtariWrapper']),
('exploration_final_eps', 0.01),
('exploration_fraction', 0.1),
('frame_stack', 4),
('gradient_steps', 1),
('learning_rate', 0.0001),
('learning_starts', 100000),
('n_timesteps', 600000.0),
('optimize_memory_usage', False),
('policy', 'CnnPolicy'),
('target_update_interval', 1000),
('train_freq', 4),
('normalize', False)])
```
|
AlexN/xls-r-300m-pt
|
AlexN
| 2022-08-05T11:24:32Z
| 36
| 1
|
transformers
|
[
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"robust-speech-event",
"mozilla-foundation/common_voice_8_0",
"generated_from_trainer",
"hf-asr-leaderboard",
"pt",
"dataset:mozilla-foundation/common_voice_8_0",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us"
] |
automatic-speech-recognition
| 2022-03-02T23:29:04Z
|
---
language:
- pt
license: apache-2.0
tags:
- automatic-speech-recognition
- robust-speech-event
- mozilla-foundation/common_voice_8_0
- generated_from_trainer
- hf-asr-leaderboard
datasets:
- mozilla-foundation/common_voice_8_0
model-index:
- name: xls-r-300m-pt
results:
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 8.0 pt
type: mozilla-foundation/common_voice_8_0
args: pt
metrics:
- name: Test WER
type: wer
value: 19.361
- name: Test CER
type: cer
value: 5.533
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: Robust Speech Event - Dev Data
type: speech-recognition-community-v2/dev_data
args: fr
metrics:
- name: Validation WER
type: wer
value: 47.812
- name: Validation CER
type: cer
value: 18.805
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 8.0
type: mozilla-foundation/common_voice_8_0
args: pt
metrics:
- name: Test WER
type: wer
value: 19.36
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Robust Speech Event - Dev Data
type: speech-recognition-community-v2/dev_data
args: pt
metrics:
- name: Test WER
type: wer
value: 48.01
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Robust Speech Event - Test Data
type: speech-recognition-community-v2/eval_data
args: pt
metrics:
- name: Test WER
type: wer
value: 49.21
---
<!-- 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. -->
#
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_8_0 - PT dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2290
- Wer: 0.2382
## 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: 0.0002
- 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_steps: 1500
- num_epochs: 15.0
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:-----:|:---------------:|:------:|
| 3.0952 | 0.64 | 500 | 3.0982 | 1.0 |
| 1.7975 | 1.29 | 1000 | 0.7887 | 0.5651 |
| 1.4138 | 1.93 | 1500 | 0.5238 | 0.4389 |
| 1.344 | 2.57 | 2000 | 0.4775 | 0.4318 |
| 1.2737 | 3.21 | 2500 | 0.4648 | 0.4075 |
| 1.2554 | 3.86 | 3000 | 0.4069 | 0.3678 |
| 1.1996 | 4.5 | 3500 | 0.3914 | 0.3668 |
| 1.1427 | 5.14 | 4000 | 0.3694 | 0.3572 |
| 1.1372 | 5.78 | 4500 | 0.3568 | 0.3501 |
| 1.0831 | 6.43 | 5000 | 0.3331 | 0.3253 |
| 1.1074 | 7.07 | 5500 | 0.3332 | 0.3352 |
| 1.0536 | 7.71 | 6000 | 0.3131 | 0.3152 |
| 1.0248 | 8.35 | 6500 | 0.3024 | 0.3023 |
| 1.0075 | 9.0 | 7000 | 0.2948 | 0.3028 |
| 0.979 | 9.64 | 7500 | 0.2796 | 0.2853 |
| 0.9594 | 10.28 | 8000 | 0.2719 | 0.2789 |
| 0.9172 | 10.93 | 8500 | 0.2620 | 0.2695 |
| 0.9047 | 11.57 | 9000 | 0.2537 | 0.2596 |
| 0.8777 | 12.21 | 9500 | 0.2438 | 0.2525 |
| 0.8629 | 12.85 | 10000 | 0.2409 | 0.2493 |
| 0.8575 | 13.5 | 10500 | 0.2366 | 0.2440 |
| 0.8361 | 14.14 | 11000 | 0.2317 | 0.2385 |
| 0.8126 | 14.78 | 11500 | 0.2290 | 0.2382 |
### Framework versions
- Transformers 4.17.0.dev0
- Pytorch 1.10.2+cu102
- Datasets 1.18.2.dev0
- Tokenizers 0.11.0
|
keepitreal/mini-phoelectra-v1
|
keepitreal
| 2022-08-05T10:47:11Z
| 5
| 0
|
transformers
|
[
"transformers",
"pytorch",
"electra",
"fill-mask",
"generated_from_trainer",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] |
fill-mask
| 2022-08-05T09:19:28Z
|
---
tags:
- generated_from_trainer
model-index:
- name: mini-phoelectra-v1
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. -->
# mini-phoelectra-v1
This model is a fine-tuned version of [](https://huggingface.co/) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 6.4503
## 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: 64
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
### Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
|
Subsets and Splits
Filtered Qwen2.5 Distill Models
Identifies specific configurations of models by filtering cards that contain 'distill', 'qwen2.5', '7b' while excluding certain base models and incorrect model ID patterns, uncovering unique model variants.
Filtered Model Cards Count
Finds the count of entries with specific card details that include 'distill', 'qwen2.5', '7b' but exclude certain base models, revealing valuable insights about the dataset's content distribution.
Filtered Distill Qwen 7B Models
Filters for specific card entries containing 'distill', 'qwen', and '7b', excluding certain strings and patterns, to identify relevant model configurations.
Filtered Qwen-7b Model Cards
The query performs a detailed filtering based on specific keywords and excludes certain entries, which could be useful for identifying a specific subset of cards but does not provide deeper insights or trends.
Filtered Qwen 7B Model Cards
The query filters for specific terms related to "distilled" or "distill", "qwen", and "7b" in the 'card' column but excludes certain base models, providing a limited set of entries for further inspection.
Qwen 7B Distilled Models
The query provides a basic filtering of records to find specific card names that include keywords related to distilled Qwen 7b models, excluding a particular base model, which gives limited insight but helps in focusing on relevant entries.
Qwen 7B Distilled Model Cards
The query filters data based on specific keywords in the modelId and card fields, providing limited insight primarily useful for locating specific entries rather than revealing broad patterns or trends.
Qwen 7B Distilled Models
Finds all entries containing the terms 'distilled', 'qwen', and '7b' in a case-insensitive manner, providing a filtered set of records but without deeper analysis.
Distilled Qwen 7B Models
The query filters for specific model IDs containing 'distilled', 'qwen', and '7b', providing a basic retrieval of relevant entries but without deeper analysis or insight.
Filtered Model Cards with Distill Qwen2.
Filters and retrieves records containing specific keywords in the card description while excluding certain phrases, providing a basic count of relevant entries.
Filtered Model Cards with Distill Qwen 7
The query filters specific variations of card descriptions containing 'distill', 'qwen', and '7b' while excluding a particular base model, providing limited but specific data retrieval.
Distill Qwen 7B Model Cards
The query filters and retrieves rows where the 'card' column contains specific keywords ('distill', 'qwen', and '7b'), providing a basic filter result that can help in identifying specific entries.