modelId
string | author
string | last_modified
timestamp[us, tz=UTC] | downloads
int64 | likes
int64 | library_name
string | tags
sequence | pipeline_tag
string | createdAt
timestamp[us, tz=UTC] | card
string |
|---|---|---|---|---|---|---|---|---|---|
huggingtweets/clortown | huggingtweets | 2022-04-02T04:51:29Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-04-02T02:36:56Z | ---
language: en
thumbnail: http://www.huggingtweets.com/clortown/1648875085007/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/1488574779351187458/RlIQNUFG_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">yeosang elf agenda</div>
<div style="text-align: center; font-size: 14px;">@clortown</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 yeosang elf agenda.
| Data | yeosang elf agenda |
| --- | --- |
| Tweets downloaded | 3140 |
| Retweets | 538 |
| Short tweets | 463 |
| Tweets kept | 2139 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/3cupnlna/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 @clortown's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/uii743r9) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/uii743r9/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/clortown')
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)
|
nikhil6041/wav2vec2-commonvoice-hindi | nikhil6041 | 2022-04-02T04:48:26Z | 5 | 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 | 2022-03-31T04:27:46Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- common_voice
model-index:
- name: wav2vec2-commonvoice-hindi
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-commonvoice-hindi
This model is a fine-tuned version of [theainerd/Wav2Vec2-large-xlsr-hindi](https://huggingface.co/theainerd/Wav2Vec2-large-xlsr-hindi) on the common_voice dataset.
It achieves the following results on the evaluation set:
- Loss: 0.9825
- Wer: 0.6763
## 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: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 30
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| No log | 20.0 | 100 | 0.8801 | 0.6754 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu111
- Datasets 1.18.3
- Tokenizers 0.10.3
|
JustAdvanceTechonology/medical_research_dataset_marian-finetuned-kde4-fr-to-en | JustAdvanceTechonology | 2022-04-02T00:07:29Z | 4 | 0 | transformers | [
"transformers",
"tf",
"marian",
"text2text-generation",
"generated_from_keras_callback",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-31T10:16:30Z | ---
license: apache-2.0
tags:
- generated_from_keras_callback
model-index:
- name: JustAdvanceTechonology/medical_research_dataset_marian-finetuned-kde4-fr-to-en
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. -->
# JustAdvanceTechonology/medical_research_dataset_marian-finetuned-kde4-fr-to-en
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-en-fr](https://huggingface.co/Helsinki-NLP/opus-mt-en-fr) on an unknown dataset.
It achieves the following results on the evaluation set:
- Train Loss: 0.6429
- Validation Loss: 0.8071
- Epoch: 2
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 5e-05, 'decay_steps': 17733, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}
- training_precision: mixed_float16
### Training results
| Train Loss | Validation Loss | Epoch |
|:----------:|:---------------:|:-----:|
| 0.6423 | 0.8071 | 0 |
| 0.6424 | 0.8071 | 1 |
| 0.6429 | 0.8071 | 2 |
### Framework versions
- Transformers 4.16.2
- TensorFlow 2.5.0
- Datasets 2.0.0
- Tokenizers 0.10.1
|
DrishtiSharma/poem-gen-spanish-t5-small-d2 | DrishtiSharma | 2022-04-01T22:38:26Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"t5",
"text2text-generation",
"generated_from_trainer",
"license:mit",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-04-01T17:08:12Z | ---
license: mit
tags:
- generated_from_trainer
model-index:
- name: poem-gen-spanish-t5-small-d2
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. -->
# poem-gen-spanish-t5-small-d2
This model is a fine-tuned version of [flax-community/spanish-t5-small](https://huggingface.co/flax-community/spanish-t5-small) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 2.9027
## 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: 6
- eval_batch_size: 6
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:------:|:---------------:|
| 3.223 | 0.73 | 30000 | 3.1479 |
| 3.0109 | 1.46 | 60000 | 3.0544 |
| 2.8649 | 2.19 | 90000 | 2.9730 |
| 2.7603 | 2.93 | 120000 | 2.9301 |
| 2.6343 | 3.66 | 150000 | 2.9188 |
| 2.5094 | 4.39 | 180000 | 2.9064 |
| 2.391 | 5.12 | 210000 | 2.9073 |
| 2.3592 | 5.85 | 240000 | 2.9022 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
lgris/wav2vec2-large-xlsr-open-brazilian-portuguese-v2 | lgris | 2022-04-01T20:35:26Z | 858 | 18 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"audio",
"speech",
"pt",
"portuguese-speech-corpus",
"PyTorch",
"hf-asr-leaderboard",
"dataset:common_voice",
"dataset:mls",
"dataset:cetuc",
"dataset:lapsbm",
"dataset:voxforge",
"arxiv:2012.03411",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-02T23:29:05Z | ---
language: pt
datasets:
- common_voice
- mls
- cetuc
- lapsbm
- voxforge
metrics:
- wer
tags:
- audio
- speech
- wav2vec2
- pt
- portuguese-speech-corpus
- automatic-speech-recognition
- speech
- PyTorch
- hf-asr-leaderboard
model-index:
- name: wav2vec2-large-xlsr-open-brazilian-portuguese-v2
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice
type: common_voice
args: pt
metrics:
- name: Test WER
type: wer
value: 10.69
license: apache-2.0
---
# Wav2vec 2.0 With Open Brazilian Portuguese Datasets v2
This a the demonstration of a fine-tuned Wav2vec model for Brazilian Portuguese using the following datasets:
- [CETUC](http://www02.smt.ufrj.br/~igor.quintanilha/alcaim.tar.gz): contains approximately 145 hours of Brazilian Portuguese speech distributed among 50 male and 50 female speakers, each pronouncing approximately 1,000 phonetically balanced sentences selected from the [CETEN-Folha](https://www.linguateca.pt/cetenfolha/) corpus.
- [Multilingual Librispeech (MLS)](https://arxiv.org/abs/2012.03411): a massive dataset available in many languages. The MLS is based on audiobook recordings in public domain like [LibriVox](https://librivox.org/). The dataset contains a total of 6k hours of transcribed data in many languages. The set in Portuguese [used in this work](http://www.openslr.org/94/) (mostly Brazilian variant) has approximately 284 hours of speech, obtained from 55 audiobooks read by 62 speakers.
- [VoxForge](http://www.voxforge.org/): is a project with the goal to build open datasets for acoustic models. The corpus contains approximately 100 speakers and 4,130 utterances of Brazilian Portuguese, with sample rates varying from 16kHz to 44.1kHz.
- [Common Voice 6.1](https://commonvoice.mozilla.org/pt): is a project proposed by Mozilla Foundation with the goal to create a wide open dataset in different languages to train ASR models. In this project, volunteers donate and validate speech using the [oficial site](https://commonvoice.mozilla.org/pt). The set in Portuguese (mostly Brazilian variant) used in this work is the 6.1 version (pt_63h_2020-12-11) that contains about 50 validated hours and 1,120 unique speakers.
- [Lapsbm](https://github.com/falabrasil/gitlab-resources): "Falabrasil - UFPA" is a dataset used by the Fala Brasil group to benchmark ASR systems in Brazilian Portuguese. Contains 35 speakers (10 females), each one pronouncing 20 unique sentences, totalling 700 utterances in Brazilian Portuguese. The audios were recorded in 22.05 kHz without environment control.
These datasets were combined to build a larger Brazilian Portuguese dataset. All data was used for training except Common Voice dev/test sets, that were used for validation/test respectively.
The original model was fine-tuned using [fairseq](https://github.com/pytorch/fairseq). This notebook uses a converted version of the original one.
__NOTE: The common voice test reports 10% of WER, however, this model was trained using all the validated instances of Common Voice, except the instances of the test set. This means that some speakers of the train set can be present on the test set.__
## Imports and dependencies
```python
%%capture
!pip install datasets
!pip install jiwer
!pip install torchaudio
!pip install transformers
!pip install soundfile
```
```python
import torchaudio
from datasets import load_dataset, load_metric
from transformers import (
Wav2Vec2ForCTC,
Wav2Vec2Processor,
)
import torch
import re
import sys
```
## Preparation
```python
chars_to_ignore_regex = '[\,\?\.\!\;\:\"]' # noqa: W605
wer = load_metric("wer")
device = "cuda"
```
```python
model_name = 'lgris/wav2vec2-large-xlsr-open-brazilian-portuguese-v2'
model = Wav2Vec2ForCTC.from_pretrained(model_name).to(device)
processor = Wav2Vec2Processor.from_pretrained(model_name)
```
```python
def map_to_pred(batch):
features = processor(batch["speech"], sampling_rate=batch["sampling_rate"][0], padding=True, return_tensors="pt")
input_values = features.input_values.to(device)
attention_mask = features.attention_mask.to(device)
with torch.no_grad():
logits = model(input_values, attention_mask=attention_mask).logits
pred_ids = torch.argmax(logits, dim=-1)
batch["predicted"] = processor.batch_decode(pred_ids)
batch["predicted"] = [pred.lower() for pred in batch["predicted"]]
batch["target"] = batch["sentence"]
return batch
```
## Tests
### Test against Common Voice (In-domain)
```python
dataset = load_dataset("common_voice", "pt", split="test", data_dir="./cv-corpus-6.1-2020-12-11")
resampler = torchaudio.transforms.Resample(orig_freq=48_000, new_freq=16_000)
def map_to_array(batch):
speech, _ = torchaudio.load(batch["path"])
batch["speech"] = resampler.forward(speech.squeeze(0)).numpy()
batch["sampling_rate"] = resampler.new_freq
batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower().replace("’", "'")
return batch
```
```python
ds = dataset.map(map_to_array)
result = ds.map(map_to_pred, batched=True, batch_size=1, remove_columns=list(ds.features.keys()))
print(wer.compute(predictions=result["predicted"], references=result["target"]))
for pred, target in zip(result["predicted"][:10], result["target"][:10]):
print(pred, "|", target)
```
**Result**: 10.69%
### Test against [TEDx](http://www.openslr.org/100/) (Out-of-domain)
```python
!gdown --id 1HJEnvthaGYwcV_whHEywgH2daIN4bQna
!tar -xf tedx.tar.gz
```
```python
dataset = load_dataset('csv', data_files={'test': 'test.csv'})['test']
def map_to_array(batch):
speech, _ = torchaudio.load(batch["path"])
batch["speech"] = speech.squeeze(0).numpy()
batch["sampling_rate"] = resampler.new_freq
batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower().replace("’", "'")
return batch
```
```python
ds = dataset.map(map_to_array)
result = ds.map(map_to_pred, batched=True, batch_size=1, remove_columns=list(ds.features.keys()))
print(wer.compute(predictions=result["predicted"], references=result["target"]))
for pred, target in zip(result["predicted"][:10], result["target"][:10]):
print(pred, "|", target)
```
**Result**: 34.53% |
lgris/bp500-base10k_voxpopuli | lgris | 2022-04-01T20:34:35Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"audio",
"speech",
"pt",
"portuguese-speech-corpus",
"PyTorch",
"hf-asr-leaderboard",
"dataset:common_voice",
"dataset:mls",
"dataset:cetuc",
"dataset:lapsbm",
"dataset:voxforge",
"dataset:tedx",
"dataset:sid",
"arxiv:2012.03411",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-02T23:29:05Z | ---
language: pt
datasets:
- common_voice
- mls
- cetuc
- lapsbm
- voxforge
- tedx
- sid
metrics:
- wer
tags:
- audio
- speech
- wav2vec2
- pt
- portuguese-speech-corpus
- automatic-speech-recognition
- speech
- PyTorch
- hf-asr-leaderboard
model-index:
- name: bp500-base10k_voxpopuli
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice
type: common_voice
args: pt
metrics:
- name: Test WER
type: wer
value: 24.9
license: apache-2.0
---
# bp500-base10k_voxpopuli: Wav2vec 2.0 with Brazilian Portuguese (BP) Dataset
This is a the demonstration of a fine-tuned Wav2vec model for Brazilian Portuguese using the following datasets:
- [CETUC](http://www02.smt.ufrj.br/~igor.quintanilha/alcaim.tar.gz): contains approximately 145 hours of Brazilian Portuguese speech distributed among 50 male and 50 female speakers, each pronouncing approximately 1,000 phonetically balanced sentences selected from the [CETEN-Folha](https://www.linguateca.pt/cetenfolha/) corpus.
- [Common Voice 7.0](https://commonvoice.mozilla.org/pt): is a project proposed by Mozilla Foundation with the goal to create a wide open dataset in different languages. In this project, volunteers donate and validate speech using the [oficial site](https://commonvoice.mozilla.org/pt).
- [Lapsbm](https://github.com/falabrasil/gitlab-resources): "Falabrasil - UFPA" is a dataset used by the Fala Brasil group to benchmark ASR systems in Brazilian Portuguese. Contains 35 speakers (10 females), each one pronouncing 20 unique sentences, totalling 700 utterances in Brazilian Portuguese. The audios were recorded in 22.05 kHz without environment control.
- [Multilingual Librispeech (MLS)](https://arxiv.org/abs/2012.03411): a massive dataset available in many languages. The MLS is based on audiobook recordings in public domain like [LibriVox](https://librivox.org/). The dataset contains a total of 6k hours of transcribed data in many languages. The set in Portuguese [used in this work](http://www.openslr.org/94/) (mostly Brazilian variant) has approximately 284 hours of speech, obtained from 55 audiobooks read by 62 speakers.
- [Multilingual TEDx](http://www.openslr.org/100): a collection of audio recordings from TEDx talks in 8 source languages. The Portuguese set (mostly Brazilian Portuguese variant) contains 164 hours of transcribed speech.
- [Sidney](https://igormq.github.io/datasets/) (SID): contains 5,777 utterances recorded by 72 speakers (20 women) from 17 to 59 years old with fields such as place of birth, age, gender, education, and occupation;
- [VoxForge](http://www.voxforge.org/): is a project with the goal to build open datasets for acoustic models. The corpus contains approximately 100 speakers and 4,130 utterances of Brazilian Portuguese, with sample rates varying from 16kHz to 44.1kHz.
These datasets were combined to build a larger Brazilian Portuguese dataset. All data was used for training except Common Voice dev/test sets, that were used for validation/test respectively. We also made test sets for all the gathered datasets.
| Dataset | Train | Valid | Test |
|--------------------------------|-------:|------:|------:|
| CETUC | 94.0h | -- | 5.4h |
| Common Voice | 37.8h | 8.9h | 9.5h |
| LaPS BM | 0.8h | -- | 0.1h |
| MLS | 161.0h | -- | 3.7h |
| Multilingual TEDx (Portuguese) | 148.9h | -- | 1.8h |
| SID | 7.2h | -- | 1.0h |
| VoxForge | 3.9h | -- | 0.1h |
| Total | 453.6h | 8.9h | 21.6h |
The original model was fine-tuned using [fairseq](https://github.com/pytorch/fairseq). This notebook uses a converted version of the original one. The link to the original fairseq model is available [here](https://drive.google.com/file/d/19kkENi8uvczmw9OLSdqnjvKqBE53cl_W/view?usp=sharing).
#### Summary
| | CETUC | CV | LaPS | MLS | SID | TEDx | VF | AVG |
|----------------------|---------------|----------------|----------------|----------------|----------------|----------------|----------------|----------------|
| bp\_500-base10k_voxpopuli (demonstration below) | 0.120 | 0.249 | 0.039 | 0.227 | 0.169 | 0.349 | 0.116 | 0.181 |
| bp\_500-base10k_voxpopuli + 4-gram (demonstration below) | 0.074 | 0.174 | 0.032 | 0.182 | 0.181 | 0.349 | 0.111 | 0.157 |
#### Transcription examples
| Text | Transcription |
|------------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------------|
|suco de uva e água misturam bem|suco **deúva** e água **misturão** bem|
|culpa do dinheiro|**cupa** do dinheiro|
|eu amo shooters call of duty é o meu favorito|eu **omo** **shúters cofedete** é meu favorito|
|você pode explicar por que isso acontece|você pode explicar *por* que isso **ontece**|
|no futuro você desejará ter começado a investir hoje|no futuro você desejará **a** ter começado a investir hoje|
## Demonstration
```python
MODEL_NAME = "lgris/bp500-base10k_voxpopuli"
```
### Imports and dependencies
```python
%%capture
!pip install torch==1.8.2+cu111 torchvision==0.9.2+cu111 torchaudio===0.8.2 -f https://download.pytorch.org/whl/lts/1.8/torch_lts.html
!pip install datasets
!pip install jiwer
!pip install transformers
!pip install soundfile
!pip install pyctcdecode
!pip install https://github.com/kpu/kenlm/archive/master.zip
```
```python
import jiwer
import torchaudio
from datasets import load_dataset, load_metric
from transformers import (
Wav2Vec2ForCTC,
Wav2Vec2Processor,
)
from pyctcdecode import build_ctcdecoder
import torch
import re
import sys
```
### Helpers
```python
chars_to_ignore_regex = '[\,\?\.\!\;\:\"]' # noqa: W605
def map_to_array(batch):
speech, _ = torchaudio.load(batch["path"])
batch["speech"] = speech.squeeze(0).numpy()
batch["sampling_rate"] = 16_000
batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower().replace("’", "'")
batch["target"] = batch["sentence"]
return batch
```
```python
def calc_metrics(truths, hypos):
wers = []
mers = []
wils = []
for t, h in zip(truths, hypos):
try:
wers.append(jiwer.wer(t, h))
mers.append(jiwer.mer(t, h))
wils.append(jiwer.wil(t, h))
except: # Empty string?
pass
wer = sum(wers)/len(wers)
mer = sum(mers)/len(mers)
wil = sum(wils)/len(wils)
return wer, mer, wil
```
```python
def load_data(dataset):
data_files = {'test': f'{dataset}/test.csv'}
dataset = load_dataset('csv', data_files=data_files)["test"]
return dataset.map(map_to_array)
```
### Model
```python
class STT:
def __init__(self,
model_name,
device='cuda' if torch.cuda.is_available() else 'cpu',
lm=None):
self.model_name = model_name
self.model = Wav2Vec2ForCTC.from_pretrained(model_name).to(device)
self.processor = Wav2Vec2Processor.from_pretrained(model_name)
self.vocab_dict = self.processor.tokenizer.get_vocab()
self.sorted_dict = {
k.lower(): v for k, v in sorted(self.vocab_dict.items(),
key=lambda item: item[1])
}
self.device = device
self.lm = lm
if self.lm:
self.lm_decoder = build_ctcdecoder(
list(self.sorted_dict.keys()),
self.lm
)
def batch_predict(self, batch):
features = self.processor(batch["speech"],
sampling_rate=batch["sampling_rate"][0],
padding=True,
return_tensors="pt")
input_values = features.input_values.to(self.device)
with torch.no_grad():
logits = self.model(input_values).logits
if self.lm:
logits = logits.cpu().numpy()
batch["predicted"] = []
for sample_logits in logits:
batch["predicted"].append(self.lm_decoder.decode(sample_logits))
else:
pred_ids = torch.argmax(logits, dim=-1)
batch["predicted"] = self.processor.batch_decode(pred_ids)
return batch
```
### Download datasets
```python
%%capture
!gdown --id 1HFECzIizf-bmkQRLiQD0QVqcGtOG5upI
!mkdir bp_dataset
!unzip bp_dataset -d bp_dataset/
```
```python
%cd bp_dataset
```
/content/bp_dataset
### Tests
```python
stt = STT(MODEL_NAME)
```
#### CETUC
```python
ds = load_data('cetuc_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("CETUC WER:", wer)
```
CETUC WER: 0.12096759949218888
#### Common Voice
```python
ds = load_data('commonvoice_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("CV WER:", wer)
```
CV WER: 0.24977003159495725
#### LaPS
```python
ds = load_data('lapsbm_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("Laps WER:", wer)
```
Laps WER: 0.039769570707070705
#### MLS
```python
ds = load_data('mls_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("MLS WER:", wer)
```
MLS WER: 0.2269637077788063
#### SID
```python
ds = load_data('sid_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("Sid WER:", wer)
```
Sid WER: 0.1691680138494731
#### TEDx
```python
ds = load_data('tedx_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("TEDx WER:", wer)
```
TEDx WER: 0.34908555859018014
#### VoxForge
```python
ds = load_data('voxforge_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("VoxForge WER:", wer)
```
VoxForge WER: 0.11649350649350651
### Tests with LM
```python
!rm -rf ~/.cache
!gdown --id 1GJIKseP5ZkTbllQVgOL98R4yYAcIySFP # trained with wikipedia
stt = STT(MODEL_NAME, lm='pt-BR-wiki.word.4-gram.arpa')
# !gdown --id 1dLFldy7eguPtyJj5OAlI4Emnx0BpFywg # trained with bp
# stt = STT(MODEL_NAME, lm='pt-BR.word.4-gram.arpa')
```
### Cetuc
```python
ds = load_data('cetuc_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("CETUC WER:", wer)
```
CETUC WER: 0.07499558425787961
#### Common Voice
```python
ds = load_data('commonvoice_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("CV WER:", wer)
```
CV WER: 0.17442648452610307
#### LaPS
```python
ds = load_data('lapsbm_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("Laps WER:", wer)
```
Laps WER: 0.032774621212121206
#### MLS
```python
ds = load_data('mls_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("MLS WER:", wer)
```
MLS WER: 0.18213620321569274
#### SID
```python
ds = load_data('sid_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("Sid WER:", wer)
```
Sid WER: 0.18102544972868206
#### TEDx
```python
ds = load_data('tedx_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("TEDx WER:", wer)
```
TEDx WER: 0.3491402028105601
#### VoxForge
```python
ds = load_data('voxforge_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("VoxForge WER:", wer)
```
VoxForge WER: 0.11189529220779222
|
lgris/bp500-xlsr | lgris | 2022-04-01T20:33:47Z | 15 | 1 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"audio",
"speech",
"pt",
"portuguese-speech-corpus",
"PyTorch",
"hf-asr-leaderboard",
"dataset:common_voice",
"dataset:mls",
"dataset:cetuc",
"dataset:lapsbm",
"dataset:voxforge",
"dataset:tedx",
"dataset:sid",
"arxiv:2012.03411",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-02T23:29:05Z | ---
language: pt
datasets:
- common_voice
- mls
- cetuc
- lapsbm
- voxforge
- tedx
- sid
metrics:
- wer
tags:
- audio
- speech
- wav2vec2
- pt
- portuguese-speech-corpus
- automatic-speech-recognition
- speech
- PyTorch
- hf-asr-leaderboard
model-index:
- name: bp400-xlsr
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice
type: common_voice
args: pt
metrics:
- name: Test WER
type: wer
value: 13.6
license: apache-2.0
---
# bp500-xlsr: Wav2vec 2.0 with Brazilian Portuguese (BP) Dataset
This is a the demonstration of a fine-tuned Wav2vec model for Brazilian Portuguese using the following datasets:
- [CETUC](http://www02.smt.ufrj.br/~igor.quintanilha/alcaim.tar.gz): contains approximately 145 hours of Brazilian Portuguese speech distributed among 50 male and 50 female speakers, each pronouncing approximately 1,000 phonetically balanced sentences selected from the [CETEN-Folha](https://www.linguateca.pt/cetenfolha/) corpus;
- [Common Voice 7.0](https://commonvoice.mozilla.org/pt): is a project proposed by Mozilla Foundation with the goal to create a wide open dataset in different languages. In this project, volunteers donate and validate speech using the [oficial site](https://commonvoice.mozilla.org/pt);
- [Lapsbm](https://github.com/falabrasil/gitlab-resources): "Falabrasil - UFPA" is a dataset used by the Fala Brasil group to benchmark ASR systems in Brazilian Portuguese. Contains 35 speakers (10 females), each one pronouncing 20 unique sentences, totalling 700 utterances in Brazilian Portuguese. The audios were recorded in 22.05 kHz without environment control;
- [Multilingual Librispeech (MLS)](https://arxiv.org/abs/2012.03411): a massive dataset available in many languages. The MLS is based on audiobook recordings in public domain like [LibriVox](https://librivox.org/). The dataset contains a total of 6k hours of transcribed data in many languages. The set in Portuguese [used in this work](http://www.openslr.org/94/) (mostly Brazilian variant) has approximately 284 hours of speech, obtained from 55 audiobooks read by 62 speakers;
- [VoxForge](http://www.voxforge.org/): is a project with the goal to build open datasets for acoustic models. The corpus contains approximately 100 speakers and 4,130 utterances of Brazilian Portuguese, with sample rates varying from 16kHz to 44.1kHz.
These datasets were combined to build a larger Brazilian Portuguese dataset. All data was used for training except Common Voice dev/test sets, that were used for validation/test respectively. We also made test sets for all the gathered datasets.
| Dataset | Train | Valid | Test |
|--------------------------------|-------:|------:|------:|
| CETUC | 93.9h | -- | 5.4h |
| Common Voice | 37.6h | 8.9h | 9.5h |
| LaPS BM | 0.8h | -- | 0.1h |
| MLS | 161.0h | -- | 3.7h |
| Multilingual TEDx (Portuguese) | 144.2h | -- | 1.8h |
| SID | 5.0h | -- | 1.0h |
| VoxForge | 2.8h | -- | 0.1h |
| Total | 437.2h | 8.9h | 21.6h |
The original model was fine-tuned using [fairseq](https://github.com/pytorch/fairseq). This notebook uses a converted version of the original one. The link to the original fairseq model is available [here](https://drive.google.com/file/d/1J8aR1ltDLQFe-dVrGuyxoRm2uyJjCWgf/view?usp=sharing).
#### Summary
| | CETUC | CV | LaPS | MLS | SID | TEDx | VF | AVG |
|----------------------|---------------|----------------|----------------|----------------|----------------|----------------|----------------|----------------|
| bp\_500 (demonstration below) | 0.051 | 0.136 | 0.032 | 0.118 | 0.095 | 0.248 | 0.082 | 0.108 |
| bp\_500 + 4-gram (demonstration below) | 0.032 | 0.097 | 0.022 | 0.114 | 0.125 | 0.246 | 0.065 | 0.100 |
#### Transcription examples
| Text | Transcription |
|------------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------------|
|não há um departamento de mediadores independente das federações e das agremiações|não há um **dearamento** de mediadores independente das federações e das **agrebiações**|
|mas que bodega|**masque** bodega|
|a cortina abriu o show começou|a cortina abriu o **chô** começou|
|por sorte havia uma passadeira|**busote avinhoa** **passadeiro**|
|estou maravilhada está tudo pronto|**stou** estou maravilhada está tudo pronto|
## Demonstration
```python
MODEL_NAME = "lgris/bp500-xlsr"
```
### Imports and dependencies
```python
%%capture
!pip install torch==1.8.2+cu111 torchvision==0.9.2+cu111 torchaudio===0.8.2 -f https://download.pytorch.org/whl/lts/1.8/torch_lts.html
!pip install datasets
!pip install jiwer
!pip install transformers
!pip install soundfile
!pip install pyctcdecode
!pip install https://github.com/kpu/kenlm/archive/master.zip
```
```python
import jiwer
import torchaudio
from datasets import load_dataset, load_metric
from transformers import (
Wav2Vec2ForCTC,
Wav2Vec2Processor,
)
from pyctcdecode import build_ctcdecoder
import torch
import re
import sys
```
### Helpers
```python
chars_to_ignore_regex = '[\,\?\.\!\;\:\"]' # noqa: W605
def map_to_array(batch):
speech, _ = torchaudio.load(batch["path"])
batch["speech"] = speech.squeeze(0).numpy()
batch["sampling_rate"] = 16_000
batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower().replace("’", "'")
batch["target"] = batch["sentence"]
return batch
```
```python
def calc_metrics(truths, hypos):
wers = []
mers = []
wils = []
for t, h in zip(truths, hypos):
try:
wers.append(jiwer.wer(t, h))
mers.append(jiwer.mer(t, h))
wils.append(jiwer.wil(t, h))
except: # Empty string?
pass
wer = sum(wers)/len(wers)
mer = sum(mers)/len(mers)
wil = sum(wils)/len(wils)
return wer, mer, wil
```
```python
def load_data(dataset):
data_files = {'test': f'{dataset}/test.csv'}
dataset = load_dataset('csv', data_files=data_files)["test"]
return dataset.map(map_to_array)
```
### Model
```python
class STT:
def __init__(self,
model_name,
device='cuda' if torch.cuda.is_available() else 'cpu',
lm=None):
self.model_name = model_name
self.model = Wav2Vec2ForCTC.from_pretrained(model_name).to(device)
self.processor = Wav2Vec2Processor.from_pretrained(model_name)
self.vocab_dict = self.processor.tokenizer.get_vocab()
self.sorted_dict = {
k.lower(): v for k, v in sorted(self.vocab_dict.items(),
key=lambda item: item[1])
}
self.device = device
self.lm = lm
if self.lm:
self.lm_decoder = build_ctcdecoder(
list(self.sorted_dict.keys()),
self.lm
)
def batch_predict(self, batch):
features = self.processor(batch["speech"],
sampling_rate=batch["sampling_rate"][0],
padding=True,
return_tensors="pt")
input_values = features.input_values.to(self.device)
attention_mask = features.attention_mask.to(self.device)
with torch.no_grad():
logits = self.model(input_values, attention_mask=attention_mask).logits
if self.lm:
logits = logits.cpu().numpy()
batch["predicted"] = []
for sample_logits in logits:
batch["predicted"].append(self.lm_decoder.decode(sample_logits))
else:
pred_ids = torch.argmax(logits, dim=-1)
batch["predicted"] = self.processor.batch_decode(pred_ids)
return batch
```
### Download datasets
```python
%%capture
!gdown --id 1HFECzIizf-bmkQRLiQD0QVqcGtOG5upI
!mkdir bp_dataset
!unzip bp_dataset -d bp_dataset/
```
```python
%cd bp_dataset
```
/content/bp_dataset
### Tests
```python
stt = STT(MODEL_NAME)
```
#### CETUC
```python
ds = load_data('cetuc_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("CETUC WER:", wer)
```
CETUC WER: 0.05159097808687998
#### Common Voice
```python
ds = load_data('commonvoice_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("CV WER:", wer)
```
CV WER: 0.13659981509705973
#### LaPS
```python
ds = load_data('lapsbm_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("Laps WER:", wer)
```
Laps WER: 0.03196969696969697
#### MLS
```python
ds = load_data('mls_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("MLS WER:", wer)
```
MLS WER: 0.1178481066463896
#### SID
```python
ds = load_data('sid_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("Sid WER:", wer)
```
Sid WER: 0.09544588416964224
#### TEDx
```python
ds = load_data('tedx_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("TEDx WER:", wer)
```
TEDx WER: 0.24868046340420813
#### VoxForge
```python
ds = load_data('voxforge_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("VoxForge WER:", wer)
```
VoxForge WER: 0.08246076839826841
### Tests with LM
```python
!rm -rf ~/.cache
!gdown --id 1GJIKseP5ZkTbllQVgOL98R4yYAcIySFP # trained with wikipedia
stt = STT(MODEL_NAME, lm='pt-BR-wiki.word.4-gram.arpa')
# !gdown --id 1dLFldy7eguPtyJj5OAlI4Emnx0BpFywg # trained with bp
# stt = STT(MODEL_NAME, lm='pt-BR.word.4-gram.arpa')
```
### Cetuc
```python
ds = load_data('cetuc_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("CETUC WER:", wer)
```
CETUC WER: 0.03222801788375573
#### Common Voice
```python
ds = load_data('commonvoice_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("CV WER:", wer)
```
CV WER: 0.09713866021093655
#### LaPS
```python
ds = load_data('lapsbm_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("Laps WER:", wer)
```
Laps WER: 0.022310606060606065
#### MLS
```python
ds = load_data('mls_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("MLS WER:", wer)
```
MLS WER: 0.11408590958696524
#### SID
```python
ds = load_data('sid_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("Sid WER:", wer)
```
Sid WER: 0.12502797252979136
#### TEDx
```python
ds = load_data('tedx_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("TEDx WER:", wer)
```
TEDx WER: 0.24603179403904793
#### VoxForge
```python
ds = load_data('voxforge_dataset')
result = ds.map(stt.batch_predict, batched=True, batch_size=8)
wer, mer, wil = calc_metrics(result["sentence"], result["predicted"])
print("VoxForge WER:", wer)
```
VoxForge WER: 0.06542207792207791
|
lgris/wav2vec2-large-xlsr-open-brazilian-portuguese | lgris | 2022-04-01T20:32:58Z | 268 | 9 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"audio",
"speech",
"pt",
"portuguese-speech-corpus",
"PyTorch",
"hf-asr-leaderboard",
"dataset:common_voice",
"dataset:mls",
"dataset:cetuc",
"dataset:lapsbm",
"dataset:voxforge",
"arxiv:2012.03411",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-02T23:29:05Z | ---
language: pt
datasets:
- common_voice
- mls
- cetuc
- lapsbm
- voxforge
metrics:
- wer
tags:
- audio
- speech
- wav2vec2
- pt
- portuguese-speech-corpus
- automatic-speech-recognition
- speech
- PyTorch
- hf-asr-leaderboard
license: apache-2.0
model-index:
- name: Lucas Gris XLSR Wav2Vec2 Large 53 Brazilian Portuguese
results:
- task:
name: Speech Recognition
type: automatic-speech-recognition
metrics:
- name: Test WER
type: wer
value: 12.905054857823264%
---
# Wav2vec 2.0 With Open Brazilian Portuguese Datasets
This a the demonstration of a fine-tuned Wav2vec model for Brazilian Portuguese using the following datasets:
- [CETUC](http://www02.smt.ufrj.br/~igor.quintanilha/alcaim.tar.gz): contains approximately 145 hours of Brazilian Portuguese speech distributed among 50 male and 50 female speakers, each pronouncing approximately 1,000 phonetically balanced sentences selected from the [CETEN-Folha](https://www.linguateca.pt/cetenfolha/) corpus.
- [Multilingual Librispeech (MLS)](https://arxiv.org/abs/2012.03411): a massive dataset available in many languages. The MLS is based on audiobook recordings in public domain like [LibriVox](https://librivox.org/). The dataset contains a total of 6k hours of transcribed data in many languages. The set in Portuguese [used in this work](http://www.openslr.org/94/) (mostly Brazilian variant) has approximately 284 hours of speech, obtained from 55 audiobooks read by 62 speakers.
- [VoxForge](http://www.voxforge.org/): is a project with the goal to build open datasets for acoustic models. The corpus contains approximately 100 speakers and 4,130 utterances of Brazilian Portuguese, with sample rates varying from 16kHz to 44.1kHz.
- [Common Voice 6.1](https://commonvoice.mozilla.org/pt) (_only train_): is a project proposed by Mozilla Foundation with the goal to create a wide open dataset in different languages to train ASR models. In this project, volunteers donate and validate speech using the [oficial site](https://commonvoice.mozilla.org/pt). The set in Portuguese (mostly Brazilian variant) used in this work is the 6.1 version (pt_63h_2020-12-11) that contains about 50 validated hours and 1,120 unique speakers.
- [Lapsbm](https://github.com/falabrasil/gitlab-resources): "Falabrasil - UFPA" is a dataset used by the Fala Brasil group to benchmark ASR systems in Brazilian Portuguese. Contains 35 speakers (10 females), each one pronouncing 20 unique sentences, totalling 700 utterances in Brazilian Portuguese. The audios were recorded in 22.05 kHz without environment control.
These datasets were combined to build a larger Brazilian Portuguese dataset. All data was used for training except Common Voice dev/test sets, that were used for validation/test respectively.
The original model was fine-tuned using [fairseq](https://github.com/pytorch/fairseq). This notebook uses a converted version of the original one. The link to the original fairseq model is available [here](https://drive.google.com/drive/folders/1XTKIUB4kp3oYOavwH97wq8IPFsxP5sNz?usp=sharing).
This model was trained in 80k updates.
#### Datasets in number of instances and number of frames
The following image shows the overall distribution of the dataset:
#### Transcription examples
| Text | Transcription |
|------------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------------|
| É comum os usuários confundirem software livre com software livre | É comum os __usuares__ __confunder em__ __softwerlivr__ com __softwerlivre__ |
| Ele fez tanto ghostwriting que ele começa a se sentir como um fantasma também | Ele fez tanto __golstraitn__ que ele __começou__ a se sentir como um fantasma também |
| Arnold apresentou um gráfico mostrando quantas cegonhas ele havia contado nos últimos dez anos | Arnold apresentou um gráfico mostrando quantas __segonhas__ ele havia contado nos últimos dez anos |
| Mais cedo ou mais tarde eles descobrirão como ler esses hieróglifos | Mais __sedo__ ou mais tarde eles descobriram como __de__ esses __ierogrôficos__ |
| Viver juntos compartilhar objetivos e ter um bom relacionamento | __E ver__ juntos __signafica__ viver juntos ou __fartlhar__ objetivos ter um bom __relacionamentoo__ |
| Da mesma forma uma patente pode impedir que concorrentes desenvolvam produtos similares | Da mesma forma uma patente pode impedir que concorrentes __desenvolva__ produtos similares |
| Duas mulheres e uma menina levantam com troféus | Duas mulheres e uma menina levantam com __trofés__ |
| Esse acrobata de circo deve ter um sistema vestibular bem treinado pensou o espectador | Esse acrobata de __cirko__ deve ter um sistema vestibular __bemtreinado__ pensou o espectador |
| Durante a exposição o tribunal pode fazer quaisquer perguntas ou esclarecimentos que considere apropriados | Durante a exposição o tribunal pode fazer quaisquer perguntas ou esclarecimentos que considere __apropriado__ |
## Imports and dependencies
```python
%%capture
!pip install datasets
!pip install jiwer
!pip install torchaudio
!pip install transformers
!pip install soundfile
```
```python
import torchaudio
from datasets import load_dataset, load_metric
from transformers import (
Wav2Vec2ForCTC,
Wav2Vec2Processor,
)
import torch
import re
import sys
```
## Preparation
```python
chars_to_ignore_regex = '[\,\?\.\!\;\:\"]' # noqa: W605
wer = load_metric("wer")
device = "cuda"
```
```python
model_name = 'lgris/wav2vec2-large-xlsr-open-brazilian-portuguese'
model = Wav2Vec2ForCTC.from_pretrained(model_name).to(device)
processor = Wav2Vec2Processor.from_pretrained(model_name)
```
```python
def map_to_pred(batch):
features = processor(batch["speech"], sampling_rate=batch["sampling_rate"][0], padding=True, return_tensors="pt")
input_values = features.input_values.to(device)
attention_mask = features.attention_mask.to(device)
with torch.no_grad():
logits = model(input_values, attention_mask=attention_mask).logits
pred_ids = torch.argmax(logits, dim=-1)
batch["predicted"] = processor.batch_decode(pred_ids)
batch["predicted"] = [pred.lower() for pred in batch["predicted"]]
batch["target"] = batch["sentence"]
return batch
```
## Tests
### Test against Common Voice (In-domain)
```python
dataset = load_dataset("common_voice", "pt", split="test", data_dir="./cv-corpus-6.1-2020-12-11")
resampler = torchaudio.transforms.Resample(orig_freq=48_000, new_freq=16_000)
def map_to_array(batch):
speech, _ = torchaudio.load(batch["path"])
batch["speech"] = resampler.forward(speech.squeeze(0)).numpy()
batch["sampling_rate"] = resampler.new_freq
batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower().replace("’", "'")
return batch
```
```python
ds = dataset.map(map_to_array)
result = ds.map(map_to_pred, batched=True, batch_size=1, remove_columns=list(ds.features.keys()))
print(wer.compute(predictions=result["predicted"], references=result["target"]))
for pred, target in zip(result["predicted"][:10], result["target"][:10]):
print(pred, "|", target)
```
0.12905054857823264
nem o varanin os altros influmindo os de teterno um bombederster | nem o radar nem os outros instrumentos detectaram o bombardeiro stealth
pedir dinheiro é emprestado das pessoas do aldeia | pedir dinheiro emprestado às pessoas da aldeia
oito | oito
teno calcos | trancá-los
realizaram a investigação para resolver o problema | realizar uma investigação para resolver o problema
iotube ainda é a melhor plataforma de vídeos | o youtube ainda é a melhor plataforma de vídeos
menina e menino beijando nas sombras | menina e menino beijando nas sombras
eu sou o senhor | eu sou o senhor
duas metcas sentam-se para baixo randes jornais | duas mulheres que sentam-se para baixo lendo jornais
eu originalmente esperava | eu originalmente esperava
**Result**: 12.90%
### Test against [TEDx](http://www.openslr.org/100/) (Out-of-domain)
```python
!gdown --id 1HJEnvthaGYwcV_whHEywgH2daIN4bQna
!tar -xf tedx.tar.gz
```
```python
dataset = load_dataset('csv', data_files={'test': 'tedx/test.csv'})['test']
def map_to_array(batch):
speech, _ = torchaudio.load(batch["path"])
batch["speech"] = speech.squeeze(0).numpy()
batch["sampling_rate"] = resampler.new_freq
batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower().replace("’", "'")
return batch
```
```python
ds = dataset.map(map_to_array)
result = ds.map(map_to_pred, batched=True, batch_size=1, remove_columns=list(ds.features.keys()))
print(wer.compute(predictions=result["predicted"], references=result["target"]))
for pred, target in zip(result["predicted"][:10], result["target"][:10]):
print(pred, "|", target)
```
0.35215851987208774
com isso a gente vê que essa rede de pactuação de de deparcerias nos remete a um raciocínio lógico que ao que a gente crê que é a prevenção | com isso a gente vê que essa rede de pactuação de parcerias nos remete a um raciocínio lógico que é o que a gente crê que é a prevenção
ente vai para o resultado | e aí a gente vai pro resultado
curiosidade hé o que eu descobri desde que comecei a fazer pesquisa lá no ensino médio | e a curiosidade é algo que descobri desde que comecei a fazer pesquisa lá no ensino médio
val des quemesho | há vários caminhos
que é uma opcissão por comer soldado | que é uma obsessão por comer saudável
isso é tão é forte algoltão universal que existem dados que mostram que setenta e cinco por cento das reuniões são dominadas pela voz masculina | e isso é tão forte é algo tão universal que existem dados que mostram que das reuniões são dominadas pela voz masculina
não era exatamente isso não estávamos deveto | e não era exatamente isso que nós estávamos a ver
durante meci do médio ofiz pesquisa estudei numa escola que chamam a fundação liberate ficava relativamente próximo daqui | durante o ensino médio eu fiz pesquisa estudei numa escola que se chama fundação liberato que fica relativamente próxima daqui
oito anos atrás eu fui apresentado por uma doença que até então eu não conhecia e que é bem provável que a maior parte de nós todos aqui não conheçamos | oito anos atrás fui apresentado para uma doença que até então eu não conhecia e que é bem provável que a maior parte de nós todos aqui não conheçamos
o terceiro é o museu do ripiopeco | o terceiro é o museu do hip hop
**Result**: 35.21% |
anwarvic/distilbert-base-uncased-for-fakenews | anwarvic | 2022-04-01T19:12:49Z | 6 | 0 | transformers | [
"transformers",
"pytorch",
"distilbert",
"text-classification",
"exbert",
"en",
"dataset:bookcorpus",
"dataset:wikipedia",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-29T21:56:17Z | ---
language: en
tags:
- exbert
license: apache-2.0
datasets:
- bookcorpus
- wikipedia
---
# DistilBERT (uncased) for FaceNews Classification
This model is a classification model built by fine-tuning
[DistilBERT base model](https://huggingface.co/distilbert-base-uncased).
This model was trained using
[fake-and-real-news-dataset](https://www.kaggle.com/clmentbisaillon/fake-and-real-news-dataset)
for five epochs.
> **NOTE:**
This model is just a POC (proof-of-concept) for a fellowship I was applying for.
## Intended uses & limitations
Note that this model is primarily aimed at classifying an article to either
"Fake" or "Real".
### How to use
Check this [notebook](https://www.kaggle.com/code/mohamedanwarvic/fakenewsclassifier-fatima-fellowship) on Kaggle. |
juaner/distilbert-base-uncased-finetuned-cola | juaner | 2022-04-01T18:20:42Z | 5 | 0 | transformers | [
"transformers",
"tf",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_keras_callback",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-04-01T17:59:52Z | ---
license: apache-2.0
tags:
- generated_from_keras_callback
model-index:
- name: juaner/distilbert-base-uncased-finetuned-cola
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. -->
# juaner/distilbert-base-uncased-finetuned-cola
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:
- Train Loss: 0.1909
- Validation Loss: 0.5553
- Train Matthews Correlation: 0.5279
- Epoch: 2
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- optimizer: {'name': 'Adam', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 2670, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False}
- training_precision: float32
### Training results
| Train Loss | Validation Loss | Train Matthews Correlation | Epoch |
|:----------:|:---------------:|:--------------------------:|:-----:|
| 0.5191 | 0.4491 | 0.4718 | 0 |
| 0.3270 | 0.4571 | 0.5196 | 1 |
| 0.1909 | 0.5553 | 0.5279 | 2 |
### Framework versions
- Transformers 4.16.2
- TensorFlow 2.8.0
- Datasets 1.18.3
- Tokenizers 0.11.0
|
vicl/canine-c-finetuned-cola | vicl | 2022-04-01T17:38:35Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"canine",
"text-classification",
"generated_from_trainer",
"dataset:glue",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-04-01T17:13:12Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- matthews_correlation
model-index:
- name: canine-c-finetuned-cola
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: glue
type: glue
args: cola
metrics:
- name: Matthews Correlation
type: matthews_correlation
value: 0.0990441507705203
---
<!-- 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. -->
# canine-c-finetuned-cola
This model is a fine-tuned version of [google/canine-c](https://huggingface.co/google/canine-c) on the glue dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6246
- Matthews Correlation: 0.0990
## 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
| Training Loss | Epoch | Step | Validation Loss | Matthews Correlation |
|:-------------:|:-----:|:----:|:---------------:|:--------------------:|
| 0.6142 | 1.0 | 535 | 0.6268 | 0.0 |
| 0.607 | 2.0 | 1070 | 0.6234 | 0.0 |
| 0.6104 | 3.0 | 1605 | 0.6226 | 0.0 |
| 0.5725 | 4.0 | 2140 | 0.6246 | 0.0990 |
| 0.5426 | 5.0 | 2675 | 0.6866 | 0.0495 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
bitsanlp/distilbert-base-uncased-distilbert-fakenews-detection | bitsanlp | 2022-04-01T17:17:55Z | 7 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-04-01T16:12:00Z | ---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
model-index:
- name: distilbert-base-uncased-distilbert-fakenews-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. -->
# distilbert-base-uncased-distilbert-fakenews-detection
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0000
- Accuracy: 1.0
- F1: 1.0
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---:|
| 0.0125 | 1.0 | 978 | 0.0000 | 1.0 | 1.0 |
| 0.0 | 2.0 | 1956 | 0.0000 | 1.0 | 1.0 |
| 0.0 | 3.0 | 2934 | 0.0000 | 1.0 | 1.0 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.2+cu102
- Datasets 2.0.0
- Tokenizers 0.11.6
|
ahmedzaky91/Fatima-Fake_news_calssifier | ahmedzaky91 | 2022-04-01T16:54:24Z | 0 | 0 | null | [
"region:us"
] | null | 2022-04-01T00:00:39Z | ## This model is a fine-tuned version of distilbert-base-uncased-finetuned-sst-2-english on Fake and real dataset on kaggle
## The following hyperparameters were used during training:
learning_rate: 5e-05
train_batch_size: 8
num_epochs: 2
|
vicl/canine-c-finetuned-mrpc | vicl | 2022-04-01T16:33:28Z | 4 | 1 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"canine",
"text-classification",
"generated_from_trainer",
"dataset:glue",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-04-01T16:05:44Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- accuracy
- f1
model-index:
- name: canine-c-finetuned-mrpc
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: glue
type: glue
args: mrpc
metrics:
- name: Accuracy
type: accuracy
value: 0.8627450980392157
- name: F1
type: f1
value: 0.9014084507042254
---
<!-- 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. -->
# canine-c-finetuned-mrpc
This model is a fine-tuned version of [google/canine-c](https://huggingface.co/google/canine-c) on the glue dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4066
- Accuracy: 0.8627
- F1: 0.9014
## 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
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| No log | 1.0 | 230 | 0.5014 | 0.7696 | 0.8479 |
| No log | 2.0 | 460 | 0.4755 | 0.7892 | 0.8622 |
| 0.5096 | 3.0 | 690 | 0.3645 | 0.8431 | 0.8869 |
| 0.5096 | 4.0 | 920 | 0.4066 | 0.8627 | 0.9014 |
| 0.2619 | 5.0 | 1150 | 0.4551 | 0.8431 | 0.8877 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
avialfont/ner-dummy-model | avialfont | 2022-04-01T14:59:22Z | 5 | 0 | transformers | [
"transformers",
"tf",
"bert",
"token-classification",
"generated_from_keras_callback",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | token-classification | 2022-04-01T10:59:27Z | ---
license: apache-2.0
tags:
- generated_from_keras_callback
model-index:
- name: ner-dummy-model
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. -->
# ner-dummy-model
This model is a fine-tuned version of [bert-base-cased](https://huggingface.co/bert-base-cased) on an unknown dataset.
It achieves the following results on the evaluation set:
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 2631, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}
- training_precision: float32
### Training results
### Framework versions
- Transformers 4.16.2
- TensorFlow 2.8.0
- Datasets 1.18.3
- Tokenizers 0.11.6
|
notexist/ttt | notexist | 2022-04-01T13:16:50Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-04-01T12:45:30Z | ---
license: apache-2.0
---
|
bmichele/poetry-generation-nextline-mbart-ws-fi-single | bmichele | 2022-04-01T11:51:32Z | 0 | 0 | null | [
"pytorch",
"region:us"
] | null | 2022-04-01T11:35:07Z | # poetry-generation-nextline-mbart-ws-fi-single
* `nextline`: generates a poem line from previous line(s)
* `mbart`: base model is [facebook/mbart-large-cc25](https://huggingface.co/facebook/mbart-large-cc25)
* `ws`: trained on Wikisource data
* `fi`: Finnish language
* `single`: uses only last poem line as input for generation |
blacktree/distilbert-base-uncased-finetuned-cola | blacktree | 2022-04-01T09:00:33Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_trainer",
"dataset:glue",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-31T15:48:48Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- glue
metrics:
- matthews_correlation
model-index:
- name: distilbert-base-uncased-finetuned-cola
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: glue
type: glue
args: cola
metrics:
- name: Matthews Correlation
type: matthews_correlation
value: 0.5285676961321106
---
<!-- 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-cola
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the glue dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4883
- Matthews Correlation: 0.5286
## 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
| Training Loss | Epoch | Step | Validation Loss | Matthews Correlation |
|:-------------:|:-----:|:----:|:---------------:|:--------------------:|
| 0.5269 | 1.0 | 535 | 0.5197 | 0.4187 |
| 0.3477 | 2.0 | 1070 | 0.4883 | 0.5286 |
| 0.2333 | 3.0 | 1605 | 0.6530 | 0.5079 |
| 0.17 | 4.0 | 2140 | 0.7567 | 0.5272 |
| 0.1271 | 5.0 | 2675 | 0.8887 | 0.5259 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.12.0
|
yy642/bert-base-uncased-finetuned-mnli-rte-wnli-10 | yy642 | 2022-04-01T06:04:00Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"bert",
"text-classification",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-31T23:51:06Z | ---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: bert-base-uncased-finetuned-mnli-rte-wnli-10
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# bert-base-uncased-finetuned-mnli-rte-wnli-10
This model is a fine-tuned version of [yy642/bert-base-uncased-finetuned-mnli-rte-wnli-5](https://huggingface.co/yy642/bert-base-uncased-finetuned-mnli-rte-wnli-5) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5876
- Accuracy: 0.9206
## 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
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
| 0.0641 | 1.0 | 16558 | 0.4528 | 0.9138 |
| 0.0479 | 2.0 | 33116 | 0.5116 | 0.9153 |
| 0.0363 | 3.0 | 49674 | 0.5660 | 0.9138 |
| 0.0244 | 4.0 | 66232 | 0.5876 | 0.9206 |
| 0.0145 | 5.0 | 82790 | 0.6156 | 0.9192 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.11.0a0+17540c5
- Datasets 2.0.0
- Tokenizers 0.11.6
|
Yaxin/xlm-roberta-base-amazon-en-es-fr-mlm | Yaxin | 2022-04-01T05:28:33Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"xlm-roberta",
"fill-mask",
"generated_from_trainer",
"dataset:Yaxin/amazon_reviews_multi",
"license:mit",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | fill-mask | 2022-03-31T14:56:00Z | ---
license: mit
tags:
- generated_from_trainer
datasets:
- Yaxin/amazon_reviews_multi
metrics:
- accuracy
model-index:
- name: xlm-roberta-base-amazon-en-es-fr-mlm
results:
- task:
name: Masked Language Modeling
type: fill-mask
dataset:
name: Yaxin/amazon_reviews_multi
type: Yaxin/amazon_reviews_multi
metrics:
- name: Accuracy
type: accuracy
value: 0.6951035447140035
---
<!-- 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-amazon-en-es-fr-mlm
This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the Yaxin/amazon_reviews_multi dataset.
It achieves the following results on the evaluation set:
- Loss: 1.3936
- Accuracy: 0.6951
## 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: 6
- eval_batch_size: 6
- 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 4.18.0.dev0
- Pytorch 1.11.0
- Datasets 2.0.0
- Tokenizers 0.11.6
|
dchung117/distilbert-base-uncased-finetuned-squad-d5716d28 | dchung117 | 2022-04-01T02:02:28Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"distilbert",
"fill-mask",
"question-answering",
"en",
"dataset:squad",
"arxiv:1910.01108",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | question-answering | 2022-04-01T01:51:41Z | ---
language:
- en
thumbnail: https://github.com/karanchahal/distiller/blob/master/distiller.jpg
tags:
- question-answering
license: apache-2.0
datasets:
- squad
metrics:
- squad
---
# DistilBERT with a second step of distillation
## Model description
This model replicates the "DistilBERT (D)" model from Table 2 of the [DistilBERT paper](https://arxiv.org/pdf/1910.01108.pdf). In this approach, a DistilBERT student is fine-tuned on SQuAD v1.1, but with a BERT model (also fine-tuned on SQuAD v1.1) acting as a teacher for a second step of task-specific distillation.
In this version, the following pre-trained models were used:
* Student: `distilbert-base-uncased`
* Teacher: `lewtun/bert-base-uncased-finetuned-squad-v1`
## Training data
This model was trained on the SQuAD v1.1 dataset which can be obtained from the `datasets` library as follows:
```python
from datasets import load_dataset
squad = load_dataset('squad')
```
## Training procedure
## Eval results
| | Exact Match | F1 |
|------------------|-------------|------|
| DistilBERT paper | 79.1 | 86.9 |
| Ours | 78.4 | 86.5 |
The scores were calculated using the `squad` metric from `datasets`.
### BibTeX entry and citation info
```bibtex
@misc{sanh2020distilbert,
title={DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter},
author={Victor Sanh and Lysandre Debut and Julien Chaumond and Thomas Wolf},
year={2020},
eprint={1910.01108},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
``` |
arjundd/vortex-release | arjundd | 2022-03-31T21:54:43Z | 0 | 0 | null | [
"mri",
"reconstruction",
"artifact correction",
"en",
"arxiv:2111.02549",
"license:apache-2.0",
"region:us"
] | null | 2022-03-02T23:29:05Z | ---
language: en
license: apache-2.0
tags:
- mri
- reconstruction
- artifact correction
---
# VORTEX
<div align="center">
<img src="https://drive.google.com/uc?export=view&id=1q0jAm6Kg5ZhRg3h0w0ZbtIgcRF3_-Vgb" alt="Vortex Schematic" width="700px" />
</div>
> **VORTEX: Physics-Driven Data Augmentations for Consistency Training for Robust Accelerated MRI Reconstruction**\
> Arjun Desai, Beliz Gunel, Batu Ozturkler, Harris Beg, Shreyas Vasanawala, Brian Hargreaves, Christopher Ré, John Pauly, Akshay Chaudhari\
> https://arxiv.org/abs/2111.02549
This repository contains the artifacts for the VORTEX paper. To use our code
and artifacts in your research, please use the [Meddlr](https://github.com/ad12/meddlr) package.
|
anisdismail/celebA-orientation-detection | anisdismail | 2022-03-31T21:51:37Z | 0 | 2 | null | [
"image-classification",
"pytorch",
"en",
"dataset:nielsr/CelebA-faces",
"license:cc-by-nc-4.0",
"model-index",
"region:us"
] | image-classification | 2022-03-31T19:48:26Z | ---
language:
- en
license: cc-by-nc-4.0
tags:
- image-classification
- pytorch
datasets:
- nielsr/CelebA-faces
model-index:
- name: celebA_orientation_detection_model
results:
- task:
type: image_classification # Required. Example: automatic-speech-recognition
name: Image Classification # Optional. Example: Speech Recognition
dataset:
type: nielsr/CelebA-faces
name: CelebA-faces
metrics:
- type: f1score # Required. Example: wer
value: 0.97 # Required. Example: 20.90
name: Val F1 Score # Optional. Example: Test WER
---
## Detecting the Orientation of CelebA pictures using Deep Learning
This model has been trained on a modified version of the CelebA-faces dataset, which was made from flipping 20,000 images upside down and keeping 20,000 images intact.<br>
The model relies on Resnet-18 as a backbone and is connected to one output node to classify whether the images are flipped upside down (1) or not (0). |
arjundd/noise2recon-release | arjundd | 2022-03-31T21:50:44Z | 0 | 1 | null | [
"mri",
"reconstruction",
"denoising",
"en",
"arxiv:2110.00075",
"license:apache-2.0",
"region:us"
] | null | 2022-03-02T23:29:05Z | ---
language: en
license: apache-2.0
tags:
- mri
- reconstruction
- denoising
---
# Noise2Recon
> **Noise2Recon: A Semi-Supervised Framework for Joint MRI Reconstruction and Denoising**\
> Arjun Desai, Batu Ozturkler, Christopher Sandino, Shreyas Vasanawala, Brian Hargreaves, Christopher Ré, John Pauly, Akshay Chaudhari\
> https://arxiv.org/abs/2110.00075
This repository contains the artifacts for the Noise2Recon paper. To use our code
and artifacts in your research, please use the [Meddlr](https://github.com/ad12/meddlr) package.
|
magitz/distilbert-base-uncased-finetuned-emotion | magitz | 2022-03-31T20:48:43Z | 8 | 0 | 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 | 2022-03-31T20:41:54Z | ---
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
args: default
metrics:
- name: Accuracy
type: accuracy
value: 0.9265
- name: F1
type: f1
value: 0.9267965474109292
---
<!-- 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: 0.2235
- Accuracy: 0.9265
- F1: 0.9268
## 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 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.8101 | 1.0 | 250 | 0.3177 | 0.9045 | 0.9010 |
| 0.2472 | 2.0 | 500 | 0.2235 | 0.9265 | 0.9268 |
### Framework versions
- Transformers 4.16.2
- Pytorch 1.8.1
- Datasets 1.18.3
- Tokenizers 0.11.0
|
arampacha/gpt-neo-therapist-small | arampacha | 2022-03-31T20:34:26Z | 17 | 1 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"onnx",
"gpt_neo",
"text-generation",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-30T08:40:54Z | ---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: gpt-neo-therapist-small
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. -->
# gpt-neo-therapist-small
This model is a fine-tuned version of [EleutherAI/gpt-neo-125M](https://huggingface.co/EleutherAI/gpt-neo-125M) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 4.6731
- Rouge1: 39.5028
- Rouge2: 6.43
- Rougel: 24.0091
- Rougelsum: 35.4481
- Gen Len: 204.1329
## 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: 24
- gradient_accumulation_steps: 64
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.98) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:------:|:-------:|:---------:|:--------:|
| 9.9955 | 0.97 | 7 | 6.8195 | 18.6047 | 1.0194 | 14.8565 | 17.9774 | 212.0983 |
| 6.9729 | 1.97 | 14 | 5.6783 | 26.3789 | 3.0779 | 18.5195 | 24.8592 | 203.0925 |
| 5.2614 | 2.97 | 21 | 5.0506 | 34.9428 | 4.921 | 21.9741 | 32.1122 | 206.2775 |
| 5.0599 | 3.97 | 28 | 4.7372 | 38.5235 | 6.2251 | 23.5923 | 34.5633 | 204.2428 |
| 4.5479 | 4.97 | 35 | 4.6731 | 39.5028 | 6.43 | 24.0091 | 35.4481 | 204.1329 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
novarac23/distilbert-base-uncased-finetuned-emotion | novarac23 | 2022-03-31T19:39:15Z | 6 | 0 | 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 | 2022-03-31T19:05:57Z | ---
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
args: default
metrics:
- name: Accuracy
type: accuracy
value: 0.925
- name: F1
type: f1
value: 0.9251919899321654
---
<!-- 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: 0.2234
- Accuracy: 0.925
- F1: 0.9252
## 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 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.8213 | 1.0 | 250 | 0.3210 | 0.9025 | 0.8989 |
| 0.2463 | 2.0 | 500 | 0.2234 | 0.925 | 0.9252 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu111
- Datasets 1.16.1
- Tokenizers 0.10.3
|
Tahsin-Mayeesha/distilbert-finetuned-fakenews | Tahsin-Mayeesha | 2022-03-31T17:11:42Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-31T15:58:31Z | ---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
model-index:
- name: distilbert-finetuned-fakenews
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-finetuned-fakenews
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0049
- Accuracy: 0.9995
- F1: 0.9995
## 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
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.0392 | 1.0 | 500 | 0.0059 | 0.999 | 0.999 |
| 0.002 | 2.0 | 1000 | 0.0047 | 0.9995 | 0.9995 |
| 0.0001 | 3.0 | 1500 | 0.0047 | 0.9995 | 0.9995 |
| 0.0001 | 4.0 | 2000 | 0.0049 | 0.9995 | 0.9995 |
| 0.0 | 5.0 | 2500 | 0.0049 | 0.9995 | 0.9995 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.12.0
|
eren23/pneumonia-bielefeld-dl-course | eren23 | 2022-03-31T15:55:27Z | 61 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"vit",
"image-classification",
"huggingpics",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | image-classification | 2022-03-27T12:17:21Z | ---
tags:
- image-classification
- pytorch
- huggingpics
metrics:
- accuracy
model-index:
- name: pneumonia-bielefeld-dl-course
results:
- task:
name: Image Classification
type: image-classification
metrics:
- name: Accuracy
type: accuracy
value: 0.8456632494926453
---
# pneumonia-bielefeld-dl-course
This registry contains the model for making pneumonia predictions and was prepared for
Bielefeld University Deep Learning course homework.
The code used for this implementation mostly comes from here: https://github.com/nateraw/huggingpics it was a ready pipeline for model fine-tuning with huggingface and PyTorch Lightning for another dataset.
|
huggingtweets/youtube | huggingtweets | 2022-03-31T14:06:33Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-31T14:05:50Z | ---
language: en
thumbnail: http://www.huggingtweets.com/youtube/1648735587597/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/1427292844612595720/RC1YSvuT_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">YouTube</div>
<div style="text-align: center; font-size: 14px;">@youtube</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 YouTube.
| Data | YouTube |
| --- | --- |
| Tweets downloaded | 3250 |
| Retweets | 23 |
| Short tweets | 104 |
| Tweets kept | 3123 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/2dx34obn/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 @youtube's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/p527w5q3) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/p527w5q3/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/youtube')
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)
|
Edresson/wav2vec2-large-xlsr-coraa-portuguese | Edresson | 2022-03-31T13:28:43Z | 632 | 15 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"audio",
"speech",
"pt",
"portuguese-speech-corpus",
"hf-asr-leaderboard",
"PyTorch",
"dataset:CORAA",
"arxiv:2110.15731",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-02T23:29:04Z | ---
language: pt
datasets:
- CORAA
metrics:
- wer
tags:
- audio
- speech
- wav2vec2
- pt
- portuguese-speech-corpus
- automatic-speech-recognition
- hf-asr-leaderboard
- speech
- PyTorch
license: apache-2.0
model-index:
- name: Edresson Casanova XLSR Wav2Vec2 Large 53 Portuguese
results:
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: CORAA
type: CORAA
args: pt
metrics:
- name: Test CORAA WER
type: wer
value: 25.26
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 7
type: mozilla-foundation/common_voice_7_0
args: pt
metrics:
- name: Test WER on Common Voice 7
type: wer
value: 20.08
---
# Wav2vec 2.0 trained with CORAA Portuguese Dataset
This a the demonstration of a fine-tuned Wav2vec model for Portuguese using the following [CORAA dataset](https://github.com/nilc-nlp/CORAA)
# Use this model
```python
from transformers import AutoTokenizer, Wav2Vec2ForCTC
tokenizer = AutoTokenizer.from_pretrained("Edresson/wav2vec2-large-xlsr-coraa-portuguese")
model = Wav2Vec2ForCTC.from_pretrained("Edresson/wav2vec2-large-xlsr-coraa-portuguese")
```
# Results
For the results check the [CORAA article](https://arxiv.org/abs/2110.15731)
# Example test with Common Voice Dataset
```python
dataset = load_dataset("common_voice", "pt", split="test", data_dir="./cv-corpus-6.1-2020-12-11")
resampler = torchaudio.transforms.Resample(orig_freq=48_000, new_freq=16_000)
def map_to_array(batch):
speech, _ = torchaudio.load(batch["path"])
batch["speech"] = resampler.forward(speech.squeeze(0)).numpy()
batch["sampling_rate"] = resampler.new_freq
batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower().replace("’", "'")
return batch
```
```python
ds = dataset.map(map_to_array)
result = ds.map(map_to_pred, batched=True, batch_size=1, remove_columns=list(ds.features.keys()))
print(wer.compute(predictions=result["predicted"], references=result["target"]))
```
|
scasutt/wav2vec2-base_toy_train_data_slow_10pct | scasutt | 2022-03-31T13:12:54Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-27T02:28:24Z | ---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: wav2vec2-base_toy_train_data_slow_10pct
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_toy_train_data_slow_10pct
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.3248
- Wer: 0.7175
## 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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 20
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 3.0663 | 2.1 | 500 | 3.0725 | 0.9982 |
| 1.1679 | 4.2 | 1000 | 1.3620 | 0.8889 |
| 0.6789 | 6.3 | 1500 | 1.2182 | 0.8160 |
| 0.5764 | 8.4 | 2000 | 1.2469 | 0.7667 |
| 0.4603 | 10.5 | 2500 | 1.2851 | 0.7533 |
| 0.4085 | 12.6 | 3000 | 1.2351 | 0.7401 |
| 0.3583 | 14.7 | 3500 | 1.2455 | 0.7367 |
| 0.3158 | 16.81 | 4000 | 1.3663 | 0.7261 |
| 0.2817 | 18.91 | 4500 | 1.3248 | 0.7175 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.11.0+cu102
- Datasets 2.0.0
- Tokenizers 0.11.6
|
mustapha/flipped-image-ViT | mustapha | 2022-03-31T12:30:19Z | 61 | 2 | transformers | [
"transformers",
"pytorch",
"vit",
"image-classification",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | image-classification | 2022-03-30T21:57:42Z | Hello world,
This model have been created in the context of ` Fatima Fellowship Programme`. The model was trained on the Cifar10 dataset with a googd final accuracy of arround 98%.
This model determines wether an image is flipped of not. |
Khalsuu/2nd-wav2vec2-l-xls-r-300m-turkish-test | Khalsuu | 2022-03-31T12:09:32Z | 5 | 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 | 2022-03-31T08:45:25Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- common_voice
model-index:
- name: 2nd-wav2vec2-l-xls-r-300m-turkish-test
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. -->
# 2nd-wav2vec2-l-xls-r-300m-turkish-test
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the common_voice dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6019
- Wer: 0.4444
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.001
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 20
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 3.0522 | 3.67 | 400 | 0.7773 | 0.7296 |
| 0.5369 | 7.34 | 800 | 0.6282 | 0.5888 |
| 0.276 | 11.01 | 1200 | 0.5998 | 0.5330 |
| 0.1725 | 14.68 | 1600 | 0.5859 | 0.4908 |
| 0.1177 | 18.35 | 2000 | 0.6019 | 0.4444 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu111
- Datasets 1.18.3
- Tokenizers 0.10.3
|
Neulvo/bert-finetuned-squad | Neulvo | 2022-03-31T12:08:42Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"bert",
"question-answering",
"generated_from_trainer",
"dataset:squad",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | question-answering | 2022-03-31T10:54:31Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- squad
model-index:
- name: bert-finetuned-squad
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# bert-finetuned-squad
This model is a fine-tuned version of [bert-base-cased](https://huggingface.co/bert-base-cased) on the squad dataset.
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 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
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.16.2
- Pytorch 1.10.2
- Datasets 1.18.3
- Tokenizers 0.11.0
|
YiTian/wav2vec2-common_voice-tr-demo | YiTian | 2022-03-31T11:40:04Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"automatic-speech-recognition",
"common_voice",
"generated_from_trainer",
"tr",
"dataset:common_voice",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-31T09:39:08Z | ---
language:
- tr
license: apache-2.0
tags:
- automatic-speech-recognition
- common_voice
- generated_from_trainer
datasets:
- common_voice
model-index:
- name: wav2vec2-common_voice-tr-demo
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-common_voice-tr-demo
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 - TR dataset.
It achieves the following results on the evaluation set:
- Loss: 2.9841
- Wer: 0.9999
## 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: 64
- 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: 15.0
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| No log | 7.14 | 100 | 3.6689 | 1.0 |
| No log | 14.29 | 200 | 3.0280 | 0.9999 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.9.0
- Datasets 1.18.0
- Tokenizers 0.11.6
|
frtna/jwt300_mt-Italian-to-Spanish_transformers | frtna | 2022-03-31T11:18:09Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"t5",
"text2text-generation",
"generated_from_trainer",
"dataset:new_dataset",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-29T09:49:05Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- new_dataset
metrics:
- sacrebleu
model-index:
- name: jwt300_mt-Italian-to-Spanish_transformers
results:
- task:
name: Sequence-to-sequence Language Modeling
type: text2text-generation
dataset:
name: new_dataset
type: new_dataset
args: jwt300_mt
metrics:
- name: Sacrebleu
type: sacrebleu
value: 0.9057
---
<!-- 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. -->
# jwt300_mt-Italian-to-Spanish_transformers
This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on the new_dataset dataset.
It achieves the following results on the evaluation set:
- Loss: 2.4425
- Sacrebleu: 0.9057
- Gen Len: 18.1276
## 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: 1
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Sacrebleu | Gen Len |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:-------:|
| 2.7545 | 1.0 | 2229 | 2.4425 | 0.9057 | 18.1276 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.11.0
- Datasets 2.0.0
- Tokenizers 0.11.6
|
scasutt/wav2vec2-base_toy_train_data_random_low_pass | scasutt | 2022-03-31T10:42:02Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-31T08:21:35Z | ---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: wav2vec2-base_toy_train_data_random_low_pass
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_toy_train_data_random_low_pass
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.3227
- Wer: 0.7288
## 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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 20
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 3.0795 | 2.1 | 500 | 3.2227 | 0.9982 |
| 1.21 | 4.2 | 1000 | 1.3713 | 0.8879 |
| 0.742 | 6.3 | 1500 | 1.2660 | 0.8296 |
| 0.5877 | 8.4 | 2000 | 1.2921 | 0.7794 |
| 0.4823 | 10.5 | 2500 | 1.2899 | 0.7565 |
| 0.4036 | 12.6 | 3000 | 1.3486 | 0.7494 |
| 0.391 | 14.7 | 3500 | 1.2701 | 0.7466 |
| 0.3426 | 16.81 | 4000 | 1.3570 | 0.7279 |
| 0.3015 | 18.91 | 4500 | 1.3227 | 0.7288 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.11.0+cu102
- Datasets 2.0.0
- Tokenizers 0.11.6
|
nikhil6041/wav2vec2-commonvoice-tamil | nikhil6041 | 2022-03-31T09:24:01Z | 18 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"dataset:common_voice",
"license:mit",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-31T04:00:23Z | ---
license: mit
tags:
- generated_from_trainer
datasets:
- common_voice
model-index:
- name: wav2vec2-commonvoice-tamil
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-commonvoice-tamil
This model is a fine-tuned version of [Harveenchadha/vakyansh-wav2vec2-tamil-tam-250](https://huggingface.co/Harveenchadha/vakyansh-wav2vec2-tamil-tam-250) on the common_voice dataset.
It achieves the following results on the evaluation set:
- Loss: 3.3415
- Wer: 1.0
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0003
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 400
- num_epochs: 30
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:---:|
| 5.384 | 1.69 | 200 | 3.3400 | 1.0 |
| 3.3085 | 3.39 | 400 | 3.3609 | 1.0 |
| 3.3008 | 5.08 | 600 | 3.3331 | 1.0 |
| 3.2852 | 6.78 | 800 | 3.3492 | 1.0 |
| 3.2908 | 8.47 | 1000 | 3.3318 | 1.0 |
| 3.2865 | 10.17 | 1200 | 3.3501 | 1.0 |
| 3.2826 | 11.86 | 1400 | 3.3403 | 1.0 |
| 3.2875 | 13.56 | 1600 | 3.3335 | 1.0 |
| 3.2899 | 15.25 | 1800 | 3.3311 | 1.0 |
| 3.2755 | 16.95 | 2000 | 3.3617 | 1.0 |
| 3.2877 | 18.64 | 2200 | 3.3317 | 1.0 |
| 3.2854 | 20.34 | 2400 | 3.3560 | 1.0 |
| 3.2878 | 22.03 | 2600 | 3.3332 | 1.0 |
| 3.2766 | 23.73 | 2800 | 3.3317 | 1.0 |
| 3.2943 | 25.42 | 3000 | 3.3737 | 1.0 |
| 3.2845 | 27.12 | 3200 | 3.3347 | 1.0 |
| 3.2765 | 28.81 | 3400 | 3.3415 | 1.0 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu111
- Datasets 1.18.3
- Tokenizers 0.10.3
|
emiyasstar/ch-w2v-conformer | emiyasstar | 2022-03-31T08:48:13Z | 0 | 2 | null | [
"region:us"
] | null | 2022-03-29T15:44:56Z | The ch-w2v-conformer model uses following datasets to pretrain:
ISML datasets (6 languages,70k hours): internal dataset contains 40k hours Chinese, Cantonese, Tibetan, Inner Mongolian, Inner Kazakh, Uighur.
Babel datasets (17 languages, 2k hours): Assamese, Bengali, Cantonese, Cebuano, Georgian, Haitian, Kazakh, Kurmanji, Lao, Pashto, Swahili, Tagalog, Tamil, Tok, Turkish, Vietnamese, Zulu
After pretraining, we build ASR system based on CTC-Attention structure. In very low resource task, we find that if too many initialization network structures are constructed in the upper layer of pre-training conformer encoder, the migration performance of the pre-training model will be destroyed, so we only build a single-layer transformer decoder for joint training.
pretrained model link:
## constrained-plus Task Performance
* Languages: Cantonese,mongolian,kazakh
* config: conf/train_conformer_large_10h.yaml
* Feature info: using mfcc feature, with dither 1.0, without cmvn
* Training info: lr 0.001, batch size 10, 4 gpus on V100, acc_grad 1, 80 epochs
* Decoding info: ctc_weight 0.5, average_num 35
dev set results trained only with 10 hours training set
## w2v-Conformer
| decoding_method | Cantonese(CER) | mongolian(WER) |
|:-------------------:|:----:|:----:|
| ctc_greedy_search | 31.46 | 53.64 |
| ctc_prefix_search | 31.47 | 53.50 |
| attention_rescoring | 31.45 | 52.96 |
## Conformer （train from scartch）
| decoding_method | Cantonese(CER) | mongolian(WER) |
|:-------------------:|----:|:----:|
| ctc_greedy_search | 61.43 | 89.38 |
| ctc_prefix_search | 61.37 | 89.53|
| attention_rescoring | 60.61 | 89.60| |
thaind/layoutlmv2-jaen-gemai | thaind | 2022-03-31T08:13:42Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"layoutlmv2",
"token-classification",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | token-classification | 2022-03-31T07:38:07Z | This is model fine tune from layoutlmv2 model for japanese and english language
|
snehatyagi/wav2vec2_test | snehatyagi | 2022-03-31T07:21:45Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-26T09:11:57Z | ---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: wav2vec2_test
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_test
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 91.1661
- Wer: 0.5714
## 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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- num_epochs: 1000
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:------:|:----:|:---------------:|:------:|
| 11.9459 | 100.0 | 100 | 46.9901 | 1.0 |
| 3.2175 | 200.0 | 200 | 73.0950 | 1.0 |
| 1.8117 | 300.0 | 300 | 78.4884 | 0.6735 |
| 1.3694 | 400.0 | 400 | 84.0168 | 0.6327 |
| 1.1392 | 500.0 | 500 | 85.2083 | 0.5918 |
| 0.979 | 600.0 | 600 | 88.9109 | 0.5918 |
| 0.8917 | 700.0 | 700 | 89.0310 | 0.5918 |
| 0.8265 | 800.0 | 800 | 90.0659 | 0.6122 |
| 0.769 | 900.0 | 900 | 91.8476 | 0.5714 |
| 0.7389 | 1000.0 | 1000 | 91.1661 | 0.5714 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.2
- Datasets 1.18.3
- Tokenizers 0.11.6
|
unjustify/autotrain-commonsence-689620825 | unjustify | 2022-03-31T06:38:08Z | 7 | 0 | transformers | [
"transformers",
"pytorch",
"distilbert",
"text-classification",
"autotrain",
"en",
"dataset:unjustify/autotrain-data-commonsence",
"co2_eq_emissions",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-31T06:18:51Z | ---
tags: autotrain
language: en
widget:
- text: "I love AutoTrain 🤗"
datasets:
- unjustify/autotrain-data-commonsence
co2_eq_emissions: 20.656741915705204
---
# Model Trained Using AutoTrain
- Problem type: Binary Classification
- Model ID: 689620825
- CO2 Emissions (in grams): 20.656741915705204
## Validation Metrics
- Loss: 0.7315372824668884
- Accuracy: 0.6354949675117849
- Precision: 0.63792194092827
- Recall: 0.6191451241361658
- AUC: 0.6912165223485615
- F1: 0.6283932978308872
## Usage
You can use cURL to access this model:
```
$ curl -X POST -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"inputs": "I love AutoTrain"}' https://api-inference.huggingface.co/models/unjustify/autotrain-commonsence-689620825
```
Or Python API:
```
from transformers import AutoModelForSequenceClassification, AutoTokenizer
model = AutoModelForSequenceClassification.from_pretrained("unjustify/autotrain-commonsence-689620825", use_auth_token=True)
tokenizer = AutoTokenizer.from_pretrained("unjustify/autotrain-commonsence-689620825", use_auth_token=True)
inputs = tokenizer("I love AutoTrain", return_tensors="pt")
outputs = model(**inputs)
``` |
lazyturtl/roomclassifier | lazyturtl | 2022-03-31T01:09:57Z | 2,692 | 16 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"vit",
"image-classification",
"huggingpics",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | image-classification | 2022-03-31T01:09:48Z | ---
tags:
- image-classification
- pytorch
- huggingpics
metrics:
- accuracy
model-index:
- name: roomclassifier
results:
- task:
name: Image Classification
type: image-classification
metrics:
- name: Accuracy
type: accuracy
value: 0.9402984976768494
---
# roomclassifier
Autogenerated by HuggingPics🤗🖼️
Create your own image classifier for **anything** by running [the demo on Google Colab](https://colab.research.google.com/github/nateraw/huggingpics/blob/main/HuggingPics.ipynb).
Report any issues with the demo at the [github repo](https://github.com/nateraw/huggingpics).
## Example Images
#### Bathroom
#### Bedroom
#### DinningRoom
#### Kitchen
#### Laundry room
#### Livingroom
 |
michiyasunaga/BioLinkBERT-large | michiyasunaga | 2022-03-31T00:54:57Z | 4,470 | 33 | transformers | [
"transformers",
"pytorch",
"bert",
"feature-extraction",
"exbert",
"linkbert",
"biolinkbert",
"fill-mask",
"question-answering",
"text-classification",
"token-classification",
"en",
"dataset:pubmed",
"arxiv:2203.15827",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-08T06:20:38Z | ---
license: apache-2.0
language: en
datasets:
- pubmed
tags:
- bert
- exbert
- linkbert
- biolinkbert
- feature-extraction
- fill-mask
- question-answering
- text-classification
- token-classification
widget:
- text: "Sunitinib is a tyrosine kinase inhibitor"
---
## BioLinkBERT-large
BioLinkBERT-large model pretrained on [PubMed](https://pubmed.ncbi.nlm.nih.gov/) abstracts along with citation link information. It is introduced in the paper [LinkBERT: Pretraining Language Models with Document Links (ACL 2022)](https://arxiv.org/abs/2203.15827). The code and data are available in [this repository](https://github.com/michiyasunaga/LinkBERT).
This model achieves state-of-the-art performance on several biomedical NLP benchmarks such as [BLURB](https://microsoft.github.io/BLURB/) and [MedQA-USMLE](https://github.com/jind11/MedQA).
## Model description
LinkBERT is a transformer encoder (BERT-like) model pretrained on a large corpus of documents. It is an improvement of BERT that newly captures **document links** such as hyperlinks and citation links to include knowledge that spans across multiple documents. Specifically, it was pretrained by feeding linked documents into the same language model context, besides a single document.
LinkBERT can be used as a drop-in replacement for BERT. It achieves better performance for general language understanding tasks (e.g. text classification), and is also particularly effective for **knowledge-intensive** tasks (e.g. question answering) and **cross-document** tasks (e.g. reading comprehension, document retrieval).
## Intended uses & limitations
The model can be used by fine-tuning on a downstream task, such as question answering, sequence classification, and token classification.
You can also use the raw model for feature extraction (i.e. obtaining embeddings for input text).
### How to use
To use the model to get the features of a given text in PyTorch:
```python
from transformers import AutoTokenizer, AutoModel
tokenizer = AutoTokenizer.from_pretrained('michiyasunaga/BioLinkBERT-large')
model = AutoModel.from_pretrained('michiyasunaga/BioLinkBERT-large')
inputs = tokenizer("Sunitinib is a tyrosine kinase inhibitor", return_tensors="pt")
outputs = model(**inputs)
last_hidden_states = outputs.last_hidden_state
```
For fine-tuning, you can use [this repository](https://github.com/michiyasunaga/LinkBERT) or follow any other BERT fine-tuning codebases.
## Evaluation results
When fine-tuned on downstream tasks, LinkBERT achieves the following results.
**Biomedical benchmarks ([BLURB](https://microsoft.github.io/BLURB/), [MedQA](https://github.com/jind11/MedQA), [MMLU](https://github.com/hendrycks/test), etc.):** BioLinkBERT attains new state-of-the-art.
| | BLURB score | PubMedQA | BioASQ | MedQA-USMLE |
| ---------------------- | -------- | -------- | ------- | -------- |
| PubmedBERT-base | 81.10 | 55.8 | 87.5 | 38.1 |
| **BioLinkBERT-base** | **83.39** | **70.2** | **91.4** | **40.0** |
| **BioLinkBERT-large** | **84.30** | **72.2** | **94.8** | **44.6** |
| | MMLU-professional medicine |
| ---------------------- | -------- |
| GPT-3 (175 params) | 38.7 |
| UnifiedQA (11B params) | 43.2 |
| **BioLinkBERT-large (340M params)** | **50.7** |
## Citation
If you find LinkBERT useful in your project, please cite the following:
```bibtex
@InProceedings{yasunaga2022linkbert,
author = {Michihiro Yasunaga and Jure Leskovec and Percy Liang},
title = {LinkBERT: Pretraining Language Models with Document Links},
year = {2022},
booktitle = {Association for Computational Linguistics (ACL)},
}
```
|
michiyasunaga/BioLinkBERT-base | michiyasunaga | 2022-03-31T00:51:21Z | 6,225 | 36 | transformers | [
"transformers",
"pytorch",
"bert",
"feature-extraction",
"exbert",
"linkbert",
"biolinkbert",
"fill-mask",
"question-answering",
"text-classification",
"token-classification",
"en",
"dataset:pubmed",
"arxiv:2203.15827",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-08T07:22:12Z | ---
license: apache-2.0
language: en
datasets:
- pubmed
tags:
- bert
- exbert
- linkbert
- biolinkbert
- feature-extraction
- fill-mask
- question-answering
- text-classification
- token-classification
widget:
- text: "Sunitinib is a tyrosine kinase inhibitor"
---
## BioLinkBERT-base
BioLinkBERT-base model pretrained on [PubMed](https://pubmed.ncbi.nlm.nih.gov/) abstracts along with citation link information. It is introduced in the paper [LinkBERT: Pretraining Language Models with Document Links (ACL 2022)](https://arxiv.org/abs/2203.15827). The code and data are available in [this repository](https://github.com/michiyasunaga/LinkBERT).
This model achieves state-of-the-art performance on several biomedical NLP benchmarks such as [BLURB](https://microsoft.github.io/BLURB/) and [MedQA-USMLE](https://github.com/jind11/MedQA).
## Model description
LinkBERT is a transformer encoder (BERT-like) model pretrained on a large corpus of documents. It is an improvement of BERT that newly captures **document links** such as hyperlinks and citation links to include knowledge that spans across multiple documents. Specifically, it was pretrained by feeding linked documents into the same language model context, besides a single document.
LinkBERT can be used as a drop-in replacement for BERT. It achieves better performance for general language understanding tasks (e.g. text classification), and is also particularly effective for **knowledge-intensive** tasks (e.g. question answering) and **cross-document** tasks (e.g. reading comprehension, document retrieval).
## Intended uses & limitations
The model can be used by fine-tuning on a downstream task, such as question answering, sequence classification, and token classification.
You can also use the raw model for feature extraction (i.e. obtaining embeddings for input text).
### How to use
To use the model to get the features of a given text in PyTorch:
```python
from transformers import AutoTokenizer, AutoModel
tokenizer = AutoTokenizer.from_pretrained('michiyasunaga/BioLinkBERT-base')
model = AutoModel.from_pretrained('michiyasunaga/BioLinkBERT-base')
inputs = tokenizer("Sunitinib is a tyrosine kinase inhibitor", return_tensors="pt")
outputs = model(**inputs)
last_hidden_states = outputs.last_hidden_state
```
For fine-tuning, you can use [this repository](https://github.com/michiyasunaga/LinkBERT) or follow any other BERT fine-tuning codebases.
## Evaluation results
When fine-tuned on downstream tasks, LinkBERT achieves the following results.
**Biomedical benchmarks ([BLURB](https://microsoft.github.io/BLURB/), [MedQA](https://github.com/jind11/MedQA), [MMLU](https://github.com/hendrycks/test), etc.):** BioLinkBERT attains new state-of-the-art.
| | BLURB score | PubMedQA | BioASQ | MedQA-USMLE |
| ---------------------- | -------- | -------- | ------- | -------- |
| PubmedBERT-base | 81.10 | 55.8 | 87.5 | 38.1 |
| **BioLinkBERT-base** | **83.39** | **70.2** | **91.4** | **40.0** |
| **BioLinkBERT-large** | **84.30** | **72.2** | **94.8** | **44.6** |
| | MMLU-professional medicine |
| ---------------------- | -------- |
| GPT-3 (175 params) | 38.7 |
| UnifiedQA (11B params) | 43.2 |
| **BioLinkBERT-large (340M params)** | **50.7** |
## Citation
If you find LinkBERT useful in your project, please cite the following:
```bibtex
@InProceedings{yasunaga2022linkbert,
author = {Michihiro Yasunaga and Jure Leskovec and Percy Liang},
title = {LinkBERT: Pretraining Language Models with Document Links},
year = {2022},
booktitle = {Association for Computational Linguistics (ACL)},
}
```
|
GleamEyeBeast/ascend_with_english | GleamEyeBeast | 2022-03-30T23:35:00Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"dataset:timit_asr",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-30T22:09:15Z | ---
tags:
- generated_from_trainer
datasets:
- timit_asr
model-index:
- name: ascend_with_english
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. -->
# ascend_with_english
This model is a fine-tuned version of [GleamEyeBeast/ascend](https://huggingface.co/GleamEyeBeast/ascend) on the timit_asr dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3049
- Wer: 0.2251
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| No log | 1.0 | 289 | 0.3524 | 0.3016 |
| 0.4246 | 2.0 | 578 | 0.3132 | 0.2607 |
| 0.4246 | 3.0 | 867 | 0.3044 | 0.2373 |
| 0.2008 | 4.0 | 1156 | 0.3075 | 0.2302 |
| 0.2008 | 5.0 | 1445 | 0.3049 | 0.2251 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
yinde/fatimah_fake_news_bert | yinde | 2022-03-30T22:41:12Z | 16 | 1 | transformers | [
"transformers",
"pytorch",
"distilbert",
"text-classification",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-30T20:54:21Z | ---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: fatimah_fake_news_bert
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. -->
# fatimah_fake_news_bert
This model is a fine-tuned version of [distilbert-base-uncased-finetuned-sst-2-english](https://huggingface.co/distilbert-base-uncased-finetuned-sst-2-english) on [Fake and real dataset on kaggle ]([distilbert-base-uncased-finetuned-sst-2-english](https://huggingface.co/distilbert-base-uncased-finetuned-sst-2-english))
It achieves the following results on the evaluation set:
- Loss: 0.0010
- Accuracy: 0.9998
## 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: 10
- eval_batch_size: 20
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- num_epochs: 1
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 0.3298 | 0.06 | 200 | 0.0094 | 0.9987 |
| 0.0087 | 0.11 | 400 | 0.0091 | 0.9988 |
| 0.0126 | 0.17 | 600 | 0.0132 | 0.9965 |
| 0.0081 | 0.22 | 800 | 0.0100 | 0.9987 |
| 0.0132 | 0.28 | 1000 | 0.0086 | 0.9990 |
| 0.0131 | 0.33 | 1200 | 0.0070 | 0.9986 |
| 0.0086 | 0.39 | 1400 | 0.0079 | 0.9990 |
| 0.0041 | 0.45 | 1600 | 0.0057 | 0.9991 |
| 0.0069 | 0.5 | 1800 | 0.0083 | 0.9989 |
| 0.0052 | 0.56 | 2000 | 0.0043 | 0.9993 |
| 0.0 | 0.61 | 2200 | 0.0047 | 0.9993 |
| 0.003 | 0.67 | 2400 | 0.0052 | 0.9994 |
| 0.0126 | 0.72 | 2600 | 0.0028 | 0.9997 |
| 0.0047 | 0.78 | 2800 | 0.0018 | 0.9996 |
| 0.0 | 0.84 | 3000 | 0.0027 | 0.9996 |
| 0.0001 | 0.89 | 3200 | 0.0029 | 0.9996 |
| 0.0079 | 0.95 | 3400 | 0.0010 | 0.9998 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
UBC-NLP/MARBERTv2 | UBC-NLP | 2022-03-30T21:52:31Z | 3,124 | 8 | transformers | [
"transformers",
"pytorch",
"tf",
"bert",
"fill-mask",
"Arabic BERT",
"MSA",
"Twitter",
"Masked Langauge Model",
"ar",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | fill-mask | 2022-03-02T23:29:05Z | ---
language:
- ar
tags:
- Arabic BERT
- MSA
- Twitter
- Masked Langauge Model
widget:
- text: "اللغة العربية هي لغة [MASK]."
---
<img src="https://raw.githubusercontent.com/UBC-NLP/marbert/main/ARBERT_MARBERT.jpg" alt="drawing" width="30%" height="30%" align="right"/>
**MARBERTv2** is one of three models described in our **ACL 2021 paper** **["ARBERT & MARBERT: Deep Bidirectional Transformers for Arabic"](https://aclanthology.org/2021.acl-long.551.pdf)**.
We find that results with ARBERT and MARBERT on QA are not competitive, a clear discrepancy from what we have observed thus far on other tasksWe hypothesize this is because the two models are pre-trained with a sequence length of only 128, which does not allow them to sufficiently capture both a question and its likely answer within the same sequence window during the pre-training.
To rectify this, we further pre-train the stronger model, MARBERT, on the same MSA data as ARBERT in addition to AraNews dataset but with a bigger sequence length of 512 tokens for 40 epochs. We call this
further pre-trained model **MARBERTv2**, noting it has **29B tokens**. MARBERTv2 acquires best performance on all but one test set, where XLM-RLarge marginally outperforms us (only in F1).
For more information, please visit our own GitHub [repo](https://github.com/UBC-NLP/marbert).
# BibTex
If you use our models (ARBERT, MARBERT, or MARBERTv2) for your scientific publication, or if you find the resources in this repository useful, please cite our paper as follows (to be updated):
```bibtex
@inproceedings{abdul-mageed-etal-2021-arbert,
title = "{ARBERT} {\&} {MARBERT}: Deep Bidirectional Transformers for {A}rabic",
author = "Abdul-Mageed, Muhammad and
Elmadany, AbdelRahim and
Nagoudi, El Moatez Billah",
booktitle = "Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)",
month = aug,
year = "2021",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.acl-long.551",
doi = "10.18653/v1/2021.acl-long.551",
pages = "7088--7105",
abstract = "Pre-trained language models (LMs) are currently integral to many natural language processing systems. Although multilingual LMs were also introduced to serve many languages, these have limitations such as being costly at inference time and the size and diversity of non-English data involved in their pre-training. We remedy these issues for a collection of diverse Arabic varieties by introducing two powerful deep bidirectional transformer-based models, ARBERT and MARBERT. To evaluate our models, we also introduce ARLUE, a new benchmark for multi-dialectal Arabic language understanding evaluation. ARLUE is built using 42 datasets targeting six different task clusters, allowing us to offer a series of standardized experiments under rich conditions. When fine-tuned on ARLUE, our models collectively achieve new state-of-the-art results across the majority of tasks (37 out of 48 classification tasks, on the 42 datasets). Our best model acquires the highest ARLUE score (77.40) across all six task clusters, outperforming all other models including XLM-R Large ( 3.4x larger size). Our models are publicly available at https://github.com/UBC-NLP/marbert and ARLUE will be released through the same repository.",
}
```
## Acknowledgments
We gratefully acknowledge support from the Natural Sciences and Engineering Research Council of Canada, the Social Sciences and Humanities Research Council of Canada, Canadian Foundation for Innovation, [ComputeCanada](www.computecanada.ca) and [UBC ARC-Sockeye](https://doi.org/10.14288/SOCKEYE). We also thank the [Google TensorFlow Research Cloud (TFRC)](https://www.tensorflow.org/tfrc) program for providing us with free TPU access.
|
mrm8488/biomedtra-small-es | mrm8488 | 2022-03-30T21:07:50Z | 3 | 2 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"electra",
"pretraining",
"Spanish",
"Electra",
"Bio",
"Medical",
"es",
"dataset:cowese",
"arxiv:1406.2661",
"endpoints_compatible",
"region:us"
] | null | 2022-03-02T23:29:05Z | ---
language: es
tags:
- Spanish
- Electra
- Bio
- Medical
datasets:
- cowese
---
## 🦠 BIOMEDtra 🏥
**BIOMEDtra** (small) is an Electra like model (discriminator in this case) trained on [Spanish Biomedical Crawled Corpus](https://zenodo.org/record/5510033#.Yhdk1ZHMLJx).
As mentioned in the original [paper](https://openreview.net/pdf?id=r1xMH1BtvB):
**ELECTRA** is a new method for self-supervised language representation learning. It can be used to pre-train transformer networks using relatively little compute. ELECTRA models are trained to distinguish "real" input tokens vs "fake" input tokens generated by another neural network, similar to the discriminator of a [GAN](https://arxiv.org/pdf/1406.2661.pdf). At small scale, ELECTRA achieves strong results even when trained on a single GPU. At large scale, ELECTRA achieves state-of-the-art results on the [SQuAD 2.0](https://rajpurkar.github.io/SQuAD-explorer/) dataset.
For a detailed description and experimental results, please refer the paper [ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators](https://openreview.net/pdf?id=r1xMH1BtvB).
## Training details
The model was trained using the Electra base code for 3 days on 1 GPU (Tesla V100 16GB).
## Dataset details
The largest Spanish biomedical and heath corpus to date gathered from a massive Spanish health domain crawler over more than 3,000 URLs were downloaded and preprocessed. The collected data have been preprocessed to produce the **CoWeSe** (Corpus Web Salud Español) resource, a large-scale and high-quality corpus intended for biomedical and health NLP in Spanish.
## Model details ⚙
|Param| # Value|
|-----|--------|
|Layers| 12 |
|Hidden | 256 |
|Params| 14M |
## Evaluation metrics (for discriminator) 🧾
|Metric | # Score |
|-------|---------|
|Accuracy| 0.9561|
|Precision| 0.808|
|Recall | 0.531 |
|AUC | 0.949|
## Benchmarks 🔨
WIP 🚧
## How to use the discriminator in `transformers`
```py
from transformers import ElectraForPreTraining, ElectraTokenizerFast
import torch
discriminator = ElectraForPreTraining.from_pretrained("mrm8488/biomedtra-small-es")
tokenizer = ElectraTokenizerFast.from_pretrained("mrm8488/biomedtra-small-es")
sentence = "Los españoles tienden a sufir déficit de vitamina c"
fake_sentence = "Los españoles tienden a déficit sufrir de vitamina c"
fake_tokens = tokenizer.tokenize(fake_sentence)
fake_inputs = tokenizer.encode(fake_sentence, return_tensors="pt")
discriminator_outputs = discriminator(fake_inputs)
predictions = torch.round((torch.sign(discriminator_outputs[0]) + 1) / 2)
[print("%7s" % token, end="") for token in fake_tokens]
[print("%7s" % prediction, end="") for prediction in predictions.tolist()]
```
## Acknowledgments
TBA
## Citation
If you want to cite this model you can use this:
```bibtex
@misc{mromero2022biomedtra,
title={Spanish BioMedical Electra (small)},
author={Romero, Manuel},
publisher={Hugging Face},
journal={Hugging Face Hub},
howpublished={\url{https://huggingface.co/mrm8488/biomedtra-small-es},
year={2022}
}
```
> Created by [Manuel Romero/@mrm8488](https://twitter.com/mrm8488)
> Made with <span style="color: #e25555;">♥</span> in Spain |
vlsb/autotrain-security-text-classification-albert-688320769 | vlsb | 2022-03-30T20:59:32Z | 15 | 2 | transformers | [
"transformers",
"pytorch",
"albert",
"text-classification",
"autotrain",
"unk",
"dataset:vlsb/autotrain-data-security-text-classification-albert",
"co2_eq_emissions",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-30T20:55:59Z | ---
tags: autotrain
language: unk
widget:
- text: "I love AutoTrain 🤗"
datasets:
- vlsb/autotrain-data-security-text-classification-albert
co2_eq_emissions: 3.670416179055797
---
# Model Trained Using AutoTrain
- Problem type: Binary Classification
- Model ID: 688320769
- CO2 Emissions (in grams): 3.670416179055797
## Validation Metrics
- Loss: 0.3046899139881134
- Accuracy: 0.8826530612244898
- Precision: 0.9181818181818182
- Recall: 0.8782608695652174
- AUC: 0.9423510466988727
- F1: 0.8977777777777778
## Usage
You can use cURL to access this model:
```
$ curl -X POST -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"inputs": "I love AutoTrain"}' https://api-inference.huggingface.co/models/vlsb/autotrain-security-text-classification-albert-688320769
```
Or Python API:
```
from transformers import AutoModelForSequenceClassification, AutoTokenizer
model = AutoModelForSequenceClassification.from_pretrained("vlsb/autotrain-security-text-classification-albert-688320769", use_auth_token=True)
tokenizer = AutoTokenizer.from_pretrained("vlsb/autotrain-security-text-classification-albert-688320769", use_auth_token=True)
inputs = tokenizer("I love AutoTrain", return_tensors="pt")
outputs = model(**inputs)
``` |
mrm8488/longformer-base-4096-spanish | mrm8488 | 2022-03-30T20:36:36Z | 49 | 16 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"roberta",
"fill-mask",
"Long documents",
"longformer",
"bertin",
"spanish",
"es",
"dataset:spanish_large_corpus",
"arxiv:2004.05150",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | fill-mask | 2022-03-02T23:29:05Z | ---
language:
- es
license: mit
widget:
- text: "Manuel Romero ha creado con el equipo de BERTIN un modelo que procesa documentos <mask> largos."
tags:
- Long documents
- longformer
- bertin
- spanish
datasets:
- spanish_large_corpus
---
# longformer-base-4096-spanish
## [Longformer](https://arxiv.org/abs/2004.05150) is a Transformer model for long documents.
`longformer-base-4096` is a BERT-like model started from the RoBERTa checkpoint (**BERTIN** in this case) and pre-trained for *MLM* on long documents (from BETO's `all_wikis`). It supports sequences of length up to 4,096!
**Longformer** uses a combination of a sliding window (*local*) attention and *global* attention. Global attention is user-configured based on the task to allow the model to learn task-specific representations.
This model was made following the research done by [Iz Beltagy and Matthew E. Peters and Arman Cohan](https://arxiv.org/abs/2004.05150).
## Citation
If you want to cite this model you can use this:
```bibtex
@misc{mromero2022longformer-base-4096-spanish,
title={Spanish LongFormer by Manuel Romero},
author={Romero, Manuel},
publisher={Hugging Face},
journal={Hugging Face Hub},
howpublished={\url{https://huggingface.co/mrm8488/longformer-base-4096-spanish}},
year={2022}
}
``` |
sc2qa/msmarco_qa_classifier | sc2qa | 2022-03-30T18:33:34Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"roberta",
"text-classification",
"arxiv:2109.04689",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-02T23:29:05Z | For details, please refer to the following links.
Github repo: https://github.com/amazon-research/SC2QA-DRIL
Paper: [Generating Self-Contained and Summary-Centric Question Answer Pairs via Differentiable Reward Imitation Learning](https://arxiv.org/pdf/2109.04689.pdf) |
SAGAR4REAL/wav2vec2hindiasr | SAGAR4REAL | 2022-03-30T17:32:46Z | 5 | 1 | 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 | 2022-03-30T14:51:24Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- common_voice
model-index:
- name: wav2vec2hindiasr
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. -->
# wav2vec2hindiasr
This model is a fine-tuned version of [theainerd/Wav2Vec2-large-xlsr-hindi](https://huggingface.co/theainerd/Wav2Vec2-large-xlsr-hindi) on the common_voice 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: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 30
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu111
- Datasets 1.18.3
- Tokenizers 0.10.3
|
hoangbinhmta99/wav2vec-demo | hoangbinhmta99 | 2022-03-30T17:18:48Z | 9 | 2 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-02T23:29:05Z | Convert from model .pt to transformer
Link: https://huggingface.co/tommy19970714/wav2vec2-base-960h
Bash:
```bash
pip install transformers[sentencepiece]
pip install fairseq -U
git clone https://github.com/huggingface/transformers.git
cp transformers/src/transformers/models/wav2vec2/convert_wav2vec2_original_pytorch_checkpoint_to_pytorch.py .
wget https://dl.fbaipublicfiles.com/fairseq/wav2vec/wav2vec_small.pt -O ./wav2vec_small.pt
mkdir dict
wget https://dl.fbaipublicfiles.com/fairseq/wav2vec/dict.ltr.txt
mkdir outputs
python convert_wav2vec2_original_pytorch_checkpoint_to_pytorch.py
--pytorch_dump_folder_path ./outputs --checkpoint_path ./finetuned/wav2vec_small.pt
--dict_path ./dict/dict.ltr.txt --not_finetuned
```
# install and upload model
```
curl -s https://packagecloud.io/install/repositories/github/git-lfs/script.deb.sh | sudo bash
git lfs install
sudo apt-get install git-lfs
git lfs install
git clone https://huggingface.co/hoangbinhmta99/wav2vec-demo
ls
cd wav2vec-demo/
git status
git add .
git commit -m "First model version"
git config --global user.email [yourname]
git config --global user.name [yourpass]
git commit -m "First model version"
git push
```
|
scasutt/wav2vec2-base_toy_train_data_random_high_pass | scasutt | 2022-03-30T16:37:23Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-30T13:17:36Z | ---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: wav2vec2-base_toy_train_data_random_high_pass
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_toy_train_data_random_high_pass
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.2841
- Wer: 0.7222
## 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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 20
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 3.061 | 2.1 | 500 | 3.0551 | 1.0 |
| 1.1294 | 4.2 | 1000 | 1.3102 | 0.8777 |
| 0.7051 | 6.3 | 1500 | 1.2081 | 0.8092 |
| 0.5421 | 8.4 | 2000 | 1.2280 | 0.7684 |
| 0.448 | 10.5 | 2500 | 1.2459 | 0.7506 |
| 0.3777 | 12.6 | 3000 | 1.3533 | 0.7631 |
| 0.3611 | 14.7 | 3500 | 1.2058 | 0.7291 |
| 0.3177 | 16.81 | 4000 | 1.3168 | 0.7185 |
| 0.279 | 18.91 | 4500 | 1.2841 | 0.7222 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.11.0+cu102
- Datasets 2.0.0
- Tokenizers 0.11.6
|
facebook/stylenerf-ffhq-config-basic | facebook | 2022-03-30T14:59:16Z | 0 | 2 | null | [
"license:cc-by-nc-4.0",
"region:us"
] | null | 2022-03-20T23:34:44Z | ---
license: cc-by-nc-4.0
---
## StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image Synthesis
**Abstract:** *We propose StyleNeRF, a 3D-aware generative model for photo-realistic high-resolution image synthesis with high multi-view consistency, which can be trained on unstructured 2D images. Existing approaches either cannot synthesize high-resolution images with fine details or yield noticeable 3D-inconsistent artifacts. In addition, many of them lack control over style attributes and explicit 3D camera poses. StyleNeRF integrates the neural radiance field (NeRF) into a style-based generator to tackle the aforementioned challenges, i.e., improving rendering efficiency and 3D consistency for high-resolution image generation. We perform volume rendering only to produce a low-resolution feature map and progressively apply upsampling in 2D to address the first issue. To mitigate the inconsistencies caused by 2D upsampling, we propose multiple designs, including a better upsampler and a new regularization loss. With these designs, StyleNeRF can synthesize high-resolution images at interactive rates while preserving 3D consistency at high quality. StyleNeRF also enables control of camera poses and different levels of styles, which can generalize to unseen views. It also supports challenging tasks, including zoom-in and-out, style mixing, inversion, and semantic editing.*
## Model description
This is a pre-trained StyleNeRF checkpoint at a resolution of 512^2 based on the basic configuration used in the original paper.
## How to use
Please check the official opensource code at [here](https://github.com/facebookresearch/StyleNeRF). |
manu/lilt-camembert-base | manu | 2022-03-30T14:49:30Z | 5 | 1 | transformers | [
"transformers",
"pytorch",
"liltrobertalike",
"fill-mask",
"token-classification",
"fr",
"dataset:iit-cdip",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | token-classification | 2022-03-28T13:16:58Z | ---
language:
- fr
tags:
- token-classification
- fill-mask
license: mit
datasets:
- iit-cdip
---
This model is the combined camembert-base model, with the pretrained lilt checkpoint from the paper "LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding".
Original repository: https://github.com/jpWang/LiLT
To use it, it is necessary to fork the modeling and configuration files from the original repository, and load the pretrained model from the corresponding classes (LiLTRobertaLikeConfig, LiLTRobertaLikeForRelationExtraction, LiLTRobertaLikeForTokenClassification, LiLTRobertaLikeModel).
They can also be preloaded with the AutoConfig/model factories as such:
```python
from transformers import AutoModelForTokenClassification, AutoConfig
from path_to_custom_classes import (
LiLTRobertaLikeConfig,
LiLTRobertaLikeForRelationExtraction,
LiLTRobertaLikeForTokenClassification,
LiLTRobertaLikeModel
)
def patch_transformers():
AutoConfig.register("liltrobertalike", LiLTRobertaLikeConfig)
AutoModel.register(LiLTRobertaLikeConfig, LiLTRobertaLikeModel)
AutoModelForTokenClassification.register(LiLTRobertaLikeConfig, LiLTRobertaLikeForTokenClassification)
# etc...
```
To load the model, it is then possible to use:
```python
# patch_transformers() must have been executed beforehand
tokenizer = AutoTokenizer.from_pretrained("camembert-base")
model = AutoModel.from_pretrained("manu/lilt-camembert-base")
model = AutoModelForTokenClassification.from_pretrained("manu/lilt-camembert-base") # to be fine-tuned on a token classification task
``` |
GioReg/ita1 | GioReg | 2022-03-30T14:42:06Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"bert",
"text-classification",
"generated_from_trainer",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-28T20:17:13Z | ---
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
model-index:
- name: ita1
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. -->
# ita1
This model is a fine-tuned version of [m-polignano-uniba/bert_uncased_L-12_H-768_A-12_italian_alb3rt0](https://huggingface.co/m-polignano-uniba/bert_uncased_L-12_H-768_A-12_italian_alb3rt0) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5892
- Accuracy: 0.776
- F1: 0.5912
## 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 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
abdusah/aradia-ctc-v1 | abdusah | 2022-03-30T13:48:41Z | 23 | 0 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"abdusahmbzuai/arabic_speech_massive_300hrs",
"generated_from_trainer",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-23T10:58:05Z | ---
tags:
- automatic-speech-recognition
- abdusahmbzuai/arabic_speech_massive_300hrs
- generated_from_trainer
model-index:
- name: aradia-ctc-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. -->
# aradia-ctc-v1
This model is a fine-tuned version of [/l/users/abdulwahab.sahyoun/aradia/aradia-ctc-v1](https://huggingface.co//l/users/abdulwahab.sahyoun/aradia/aradia-ctc-v1) on the ABDUSAHMBZUAI/ARABIC_SPEECH_MASSIVE_300HRS - NA dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7171
- Wer: 0.3336
## 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: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 20.0
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| No log | 0.22 | 100 | 5.1889 | 1.0 |
| No log | 0.43 | 200 | 3.1129 | 1.0 |
| No log | 0.65 | 300 | 3.0503 | 1.0 |
| No log | 0.87 | 400 | 3.0279 | 1.0 |
| 6.2756 | 1.09 | 500 | 2.9965 | 1.0 |
| 6.2756 | 1.3 | 600 | 2.3618 | 0.9993 |
| 6.2756 | 1.52 | 700 | 1.2715 | 0.8758 |
| 6.2756 | 1.74 | 800 | 0.9971 | 0.7156 |
| 6.2756 | 1.96 | 900 | 0.8927 | 0.6382 |
| 1.712 | 2.17 | 1000 | 0.8252 | 0.5926 |
| 1.712 | 2.39 | 1100 | 0.7794 | 0.5434 |
| 1.712 | 2.61 | 1200 | 0.7557 | 0.5092 |
| 1.712 | 2.83 | 1300 | 0.7347 | 0.5203 |
| 1.712 | 3.04 | 1400 | 0.7189 | 0.4929 |
| 0.9305 | 3.26 | 1500 | 0.6820 | 0.4595 |
| 0.9305 | 3.48 | 1600 | 0.6792 | 0.4504 |
| 0.9305 | 3.69 | 1700 | 0.6596 | 0.4442 |
| 0.9305 | 3.91 | 1800 | 0.6756 | 0.4432 |
| 0.9305 | 4.13 | 1900 | 0.6663 | 0.4392 |
| 0.737 | 4.35 | 2000 | 0.6479 | 0.4372 |
| 0.737 | 4.56 | 2100 | 0.6353 | 0.4203 |
| 0.737 | 4.78 | 2200 | 0.6251 | 0.4088 |
| 0.737 | 5.0 | 2300 | 0.6209 | 0.4177 |
| 0.737 | 5.22 | 2400 | 0.6639 | 0.4094 |
| 0.6247 | 5.43 | 2500 | 0.6408 | 0.3970 |
| 0.6247 | 5.65 | 2600 | 0.6373 | 0.3932 |
| 0.6247 | 5.87 | 2700 | 0.6411 | 0.3928 |
| 0.6247 | 6.09 | 2800 | 0.6378 | 0.3897 |
| 0.6247 | 6.3 | 2900 | 0.6396 | 0.3929 |
| 0.5443 | 6.52 | 3000 | 0.6544 | 0.3864 |
| 0.5443 | 6.74 | 3100 | 0.6218 | 0.3786 |
| 0.5443 | 6.96 | 3200 | 0.6200 | 0.3784 |
| 0.5443 | 7.17 | 3300 | 0.6157 | 0.3791 |
| 0.5443 | 7.39 | 3400 | 0.6317 | 0.3798 |
| 0.4845 | 7.61 | 3500 | 0.6540 | 0.3771 |
| 0.4845 | 7.83 | 3600 | 0.6436 | 0.3670 |
| 0.4845 | 8.04 | 3700 | 0.6335 | 0.3695 |
| 0.4845 | 8.26 | 3800 | 0.6579 | 0.3610 |
| 0.4845 | 8.48 | 3900 | 0.6170 | 0.3613 |
| 0.4279 | 8.69 | 4000 | 0.6523 | 0.3617 |
| 0.4279 | 8.91 | 4100 | 0.6349 | 0.3577 |
| 0.4279 | 9.13 | 4200 | 0.6344 | 0.3673 |
| 0.4279 | 9.35 | 4300 | 0.6215 | 0.3641 |
| 0.4279 | 9.56 | 4400 | 0.6513 | 0.3608 |
| 0.3825 | 9.78 | 4500 | 0.6386 | 0.3605 |
| 0.3825 | 10.0 | 4600 | 0.6724 | 0.3549 |
| 0.3825 | 10.22 | 4700 | 0.6776 | 0.3602 |
| 0.3825 | 10.43 | 4800 | 0.6739 | 0.3544 |
| 0.3825 | 10.65 | 4900 | 0.6688 | 0.3557 |
| 0.3477 | 10.87 | 5000 | 0.6674 | 0.3564 |
| 0.3477 | 11.09 | 5100 | 0.6786 | 0.3476 |
| 0.3477 | 11.3 | 5200 | 0.6818 | 0.3478 |
| 0.3477 | 11.52 | 5300 | 0.6874 | 0.3470 |
| 0.3477 | 11.74 | 5400 | 0.6993 | 0.3424 |
| 0.3101 | 11.96 | 5500 | 0.6950 | 0.3404 |
| 0.3101 | 12.17 | 5600 | 0.6872 | 0.3406 |
| 0.3101 | 12.39 | 5700 | 0.6846 | 0.3424 |
| 0.3101 | 12.61 | 5800 | 0.7051 | 0.3405 |
| 0.3101 | 12.83 | 5900 | 0.7051 | 0.3378 |
| 0.2859 | 13.04 | 6000 | 0.6955 | 0.3403 |
| 0.2859 | 13.26 | 6100 | 0.7115 | 0.3390 |
| 0.2859 | 13.48 | 6200 | 0.7074 | 0.3384 |
| 0.2859 | 13.69 | 6300 | 0.7002 | 0.3376 |
| 0.2859 | 13.91 | 6400 | 0.7171 | 0.3360 |
| 0.2714 | 14.13 | 6500 | 0.7193 | 0.3341 |
| 0.2714 | 14.35 | 6600 | 0.7132 | 0.3347 |
| 0.2714 | 14.56 | 6700 | 0.7184 | 0.3353 |
| 0.2714 | 14.78 | 6800 | 0.7171 | 0.3331 |
### Framework versions
- Transformers 4.18.0.dev0
- Pytorch 1.10.2+cu113
- Datasets 1.18.4
- Tokenizers 0.11.6
|
javilonso/classificationEsp3_Attraction | javilonso | 2022-03-30T12:09:19Z | 5 | 0 | transformers | [
"transformers",
"tf",
"gpt2",
"text-classification",
"generated_from_keras_callback",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-30T11:07:40Z | ---
license: apache-2.0
tags:
- generated_from_keras_callback
model-index:
- name: javilonso/classificationEsp3_Attraction
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. -->
# javilonso/classificationEsp3_Attraction
This model is a fine-tuned version of [PlanTL-GOB-ES/gpt2-base-bne](https://huggingface.co/PlanTL-GOB-ES/gpt2-base-bne) on an unknown dataset.
It achieves the following results on the evaluation set:
- Train Loss: 0.0055
- Validation Loss: 0.0515
- Epoch: 2
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 17958, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}
- training_precision: mixed_float16
### Training results
| Train Loss | Validation Loss | Epoch |
|:----------:|:---------------:|:-----:|
| 0.0964 | 0.0662 | 0 |
| 0.0265 | 0.0500 | 1 |
| 0.0055 | 0.0515 | 2 |
### Framework versions
- Transformers 4.17.0
- TensorFlow 2.6.0
- Datasets 2.0.0
- Tokenizers 0.11.6
|
joe5campbell/Horovod_Tweet_Sentiment_1K_4eps | joe5campbell | 2022-03-30T11:38:32Z | 5 | 0 | transformers | [
"transformers",
"tf",
"bert",
"text-classification",
"generated_from_keras_callback",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-24T12:35:50Z | ---
license: apache-2.0
tags:
- generated_from_keras_callback
model-index:
- name: Horovod_Tweet_Sentiment_1K_4eps
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. -->
# Horovod_Tweet_Sentiment_1K_4eps
This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Train Loss: 0.6803332
- Train Accuracy: 0.57187504
- Validation Loss: 0.6883397
- Validation Accuracy: 0.54375
- Epoch: 3
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- optimizer: {'name': 'Adam', 'clipnorm': 1.0, 'learning_rate': 0.0003, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False}
- training_precision: float32
### Training results
| Train Loss | Train Accuracy | Validation Loss | Validation Accuracy | Epoch |
|:----------:|:--------------:|:---------------:|:-------------------:|:-----:|
| 0.70931095 | 0.5078125 | 0.81717503 | 0.528125 | 0 |
| 0.77384466 | 0.5296875 | 0.68696874 | 0.51875 | 1 |
| 0.68944424 | 0.53125 | 0.6837756 | 0.53125 | 2 |
| 0.6803332 | 0.57187504 | 0.6883397 | 0.54375 | 3 |
### Framework versions
- Transformers 4.17.0
- TensorFlow 2.6.0
- Tokenizers 0.11.6
|
jeniakim/hedgehog | jeniakim | 2022-03-30T09:27:38Z | 51 | 2 | transformers | [
"transformers",
"pytorch",
"bert",
"token-classification",
"en",
"license:mit",
"autotrain_compatible",
"region:us"
] | token-classification | 2022-03-02T23:29:05Z | ---
language: en
license: mit
inference: false
---
🦔 HEDGEhog 🦔: BERT-based multi-class uncertainty cues recognition
====================================================================
# Description
A fine-tuned multi-class classification model that detects four different types of uncertainty cues (a.k.a hedges) on a token level.
# Uncertainty types
label | type | description | example
---| ---| ---| ---
E | Epistemic | The proposition is possible, but its truth-value cannot be decided at the moment. | She **may** be already asleep.
I | Investigation | The proposition is in the process of having its truth-value determined. | She **examined** the role of NF-kappaB in protein activation.
D | Doxatic | The proposition expresses beliefs and hypotheses, which may be known as true or false by others. | She **believes** that the Earth is flat.
N | Condition | The proposition is true or false based on the truth-value of another proposition. | **If** she gets the job, she will move to Utrecht.
C | *certain* | *n/a* | *n/a*
# Intended uses and limitations
- The model was fine-tuned with the [Simple Transformers](https://simpletransformers.ai/) library. This library is based on Transformers but the model cannot be used directly with Transformers `pipeline` and classes; doing so would generate incorrect outputs. For this reason, the API on this page is disabled.
# How to use
To generate predictions with the model, use the [Simple Transformers](https://simpletransformers.ai/) library:
```
from simpletransformers.ner import NERModel
model = NERModel(
'bert',
'jeniakim/hedgehog',
use_cuda=False,
labels=["C", "D", "E", "I", "N"],
)
example = "As much as I definitely enjoy solitude, I wouldn't mind perhaps spending little time with you (Björk)"
predictions, raw_outputs = model.predict([example])
```
The predictions look like this:
```
[[{'As': 'C'},
{'much': 'C'},
{'as': 'C'},
{'I': 'C'},
{'definitely': 'C'},
{'enjoy': 'C'},
{'solitude,': 'C'},
{'I': 'C'},
{"wouldn't": 'C'},
{'mind': 'C'},
{'perhaps': 'E'},
{'spending': 'C'},
{'little': 'C'},
{'time': 'C'},
{'with': 'C'},
{'you': 'C'},
{'(Björk)': 'C'}]]
```
In other words, the token 'perhaps' is recognized as an **epistemic uncertainty cue** and all the other tokens are not uncertainty cues.
# Training Data
HEDGEhog is trained and evaluated on the [Szeged Uncertainty Corpus](https://rgai.inf.u-szeged.hu/node/160) (Szarvas et al. 2012<sup>1</sup>). The original sentence-level XML version of this dataset is available [here](https://rgai.inf.u-szeged.hu/node/160).
The token-level version that was used for the training can be downloaded from [here](https://1drv.ms/u/s!AvPkt_QxBozXk7BiazucDqZkVxLo6g?e=IisuM6) in a form of pickled pandas DataFrame's. You can download either the split sets (```train.pkl``` 137MB, ```test.pkl``` 17MB, ```dev.pkl``` 17MB) or the full dataset (```szeged_fixed.pkl``` 172MB). Each row in the df contains a token, its features (these are not relevant for HEDGEhog; they were used to train the baseline CRF model, see [here](https://github.com/vanboefer/uncertainty_crf)), its sentence ID, and its label.
# Training Procedure
The following training parameters were used:
- Optimizer: AdamW
- Learning rate: 4e-5
- Num train epochs: 1
- Train batch size: 16
# Evaluation Results
class | precision | recall | F1-score | support
---|---|---|---|---
Epistemic | 0.90 | 0.85 | 0.88 | 624
Doxatic | 0.88 | 0.92 | 0.90 | 142
Investigation | 0.83 | 0.86 | 0.84 | 111
Condition | 0.85 | 0.87 | 0.86 | 86
Certain | 1.00 | 1.00 | 1.00 | 104,751
**macro average** | **0.89** | **0.90** | **0.89** | 105,714
# References
<sup>1</sup> Szarvas, G., Vincze, V., Farkas, R., Móra, G., & Gurevych, I. (2012). Cross-genre and cross-domain detection of semantic uncertainty. *Computational Linguistics, 38*(2), 335-367.
|
Peltarion/xlm-roberta-longformer-base-4096 | Peltarion | 2022-03-30T09:23:58Z | 75 | 8 | transformers | [
"transformers",
"pytorch",
"xlm-roberta",
"fill-mask",
"longformer",
"multilingual",
"dataset:wikitext",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | fill-mask | 2022-03-02T23:29:04Z | ---
tags:
- longformer
language: multilingual
license: apache-2.0
datasets:
- wikitext
---
## XLM-R Longformer Model
XLM-R Longformer is a XLM-R model, that has been extended to allow sequence lengths up to 4096 tokens, instead of the regular 512. The model was pre-trained from the XLM-RoBERTa checkpoint using the Longformer [pre-training scheme](https://github.com/allenai/longformer/blob/master/scripts/convert_model_to_long.ipynb) on the English WikiText-103 corpus.
The reason for this was to investigate methods for creating efficient Transformers for low-resource languages, such as Swedish, without the need to pre-train them on long-context datasets in each respecitve language. The trained model came as a result of a master thesis project at [Peltarion](https://peltarion.com/) and was fine-tuned on multilingual quesion-answering tasks, with code available [here](https://github.com/MarkusSagen/Master-Thesis-Multilingual-Longformer#xlm-r).
Since both XLM-R model and Longformer models are large models, it it recommended to run the models with NVIDIA Apex (16bit precision), large GPU and several gradient accumulation steps.
## How to Use
The model can be used as expected to fine-tune on a downstream task.
For instance for QA.
```python
import torch
from transformers import AutoModel, AutoTokenizer
MAX_SEQUENCE_LENGTH = 4096
MODEL_NAME_OR_PATH = "markussagen/xlm-roberta-longformer-base-4096"
tokenizer = AutoTokenizer.from_pretrained(
MODEL_NAME_OR_PATH,
max_length=MAX_SEQUENCE_LENGTH,
padding="max_length",
truncation=True,
)
model = AutoModelForQuestionAnswering.from_pretrained(
MODEL_NAME_OR_PATH,
max_length=MAX_SEQUENCE_LENGTH,
)
```
## Training Procedure
The model have been trained on the WikiText-103 corpus, using a **48GB** GPU with the following training script and parameters. The model was pre-trained for 6000 iterations and took ~5 days. See the full [training script](https://github.com/MarkusSagen/Master-Thesis-Multilingual-Longformer/blob/main/scripts/finetune_qa_models.py) and [Github repo](https://github.com/MarkusSagen/Master-Thesis-Multilingual-Longformer) for more information
```sh
wget https://s3.amazonaws.com/research.metamind.io/wikitext/wikitext-103-raw-v1.zip
unzip wikitext-103-raw-v1.zip
export DATA_DIR=./wikitext-103-raw
scripts/run_long_lm.py \
--model_name_or_path xlm-roberta-base \
--model_name xlm-roberta-to-longformer \
--output_dir ./output \
--logging_dir ./logs \
--val_file_path $DATA_DIR/wiki.valid.raw \
--train_file_path $DATA_DIR/wiki.train.raw \
--seed 42 \
--max_pos 4096 \
--adam_epsilon 1e-8 \
--warmup_steps 500 \
--learning_rate 3e-5 \
--weight_decay 0.01 \
--max_steps 6000 \
--evaluate_during_training \
--logging_steps 50 \
--eval_steps 50 \
--save_steps 6000 \
--max_grad_norm 1.0 \
--per_device_eval_batch_size 2 \
--per_device_train_batch_size 1 \
--gradient_accumulation_steps 64 \
--overwrite_output_dir \
--fp16 \
--do_train \
--do_eval
```
|
Aureliano/electra-if | Aureliano | 2022-03-30T09:07:27Z | 6 | 0 | transformers | [
"transformers",
"pytorch",
"tf",
"electra",
"feature-extraction",
"en",
"arxiv:1406.2661",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | feature-extraction | 2022-03-11T15:40:21Z | ---
language: en
license: apache-2.0
---
## ELECTRA for IF
**ELECTRA** is a method for self-supervised language representation learning. They are trained to distinguish "real" input tokens vs "fake" input tokens generated by another neural network, similar to the discriminator of a [GAN](https://arxiv.org/pdf/1406.2661.pdf).
For a detailed description and experimental results, please refer to the original paper [ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators](https://openreview.net/pdf?id=r1xMH1BtvB).
This repository contains a small ELECTRA discriminator finetuned on a corpus of interactive fiction commands labelled with the WordNet synset offset of the verb in the sentence. The original dataset has been collected from the list of action in the walkthroughs for the game included in the [Jericho](https://github.com/microsoft/jericho) framework and manually annotated. For more information visit https://github.com/aporporato/electra and https://github.com/aporporato/jericho-corpora.
## How to use the discriminator in `transformers`
(Heavily based on: https://github.com/huggingface/notebooks/blob/master/examples/text_classification-tf.ipynb)
```python
import math
import numpy as np
import tensorflow as tf
from datasets import load_metric, Dataset, DatasetDict
from transformers import TFAutoModelForSequenceClassification, AutoTokenizer, DataCollatorWithPadding, create_optimizer
from transformers.keras_callbacks import KerasMetricCallback
# This example shows how this model can be used:
# you should finetune the model of your specific corpus if commands, bigger than this
dict_train = {
"idx": ["0", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15", "16", "17", "18",
"19", "20"],
"sentence": ["e", "get pen", "drop book", "x paper", "i", "south", "get paper", "drop the pen", "x book",
"inventory", "n", "get the book", "drop paper", "look at Pen", "inv", "g", "s", "get sandwich",
"drop sandwich", "x sandwich", "agin"],
"label": ["travel.v.01", "take.v.04", "drop.v.01", "examine.v.02", "inventory.v.01", "travel.v.01", "take.v.04",
"drop.v.01", "examine.v.02", "inventory.v.01", "travel.v.01", "take.v.04", "drop.v.01", "examine.v.02",
"inventory.v.01", "repeat.v.01", "travel.v.01", "take.v.04", "drop.v.01", "examine.v.02", "repeat.v.01"]
}
dict_val = {
"idx": ["0", "1", "2", "3", "4", "5"],
"sentence": ["w", "get shield", "drop sword", "x spikes", "i", "repeat"],
"label": ["travel.v.01", "take.v.04", "drop.v.01", "examine.v.02", "inventory.v.01", "repeat.v.01"]
}
raw_train_dataset = Dataset.from_dict(dict_train)
raw_val_dataset = Dataset.from_dict(dict_val)
raw_dataset = DatasetDict()
raw_dataset["train"] = raw_train_dataset
raw_dataset["val"] = raw_val_dataset
raw_dataset = raw_dataset.class_encode_column("label")
print(raw_dataset)
print(raw_dataset["train"].features)
print(raw_dataset["val"].features)
print(raw_dataset["train"][1])
label2id = {}
id2label = {}
for i, l in enumerate(raw_dataset["train"].features["label"].names):
label2id[l] = i
id2label[i] = l
discriminator = TFAutoModelForSequenceClassification.from_pretrained("Aureliano/electra-if",
label2id=label2id,
id2label=id2label)
tokenizer = AutoTokenizer.from_pretrained("Aureliano/electra-if")
tokenize_function = lambda example: tokenizer(example["sentence"], truncation=True)
pre_tokenizer_columns = set(raw_dataset["train"].features)
encoded_dataset = raw_dataset.map(tokenize_function, batched=True)
tokenizer_columns = list(set(encoded_dataset["train"].features) - pre_tokenizer_columns)
data_collator = DataCollatorWithPadding(tokenizer=tokenizer, return_tensors="tf")
batch_size = len(encoded_dataset["train"])
tf_train_dataset = encoded_dataset["train"].to_tf_dataset(
columns=tokenizer_columns,
label_cols=["labels"],
shuffle=True,
batch_size=batch_size,
collate_fn=data_collator
)
tf_validation_dataset = encoded_dataset["val"].to_tf_dataset(
columns=tokenizer_columns,
label_cols=["labels"],
shuffle=False,
batch_size=batch_size,
collate_fn=data_collator
)
loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True)
num_epochs = 25
batches_per_epoch = math.ceil(len(encoded_dataset["train"]) / batch_size)
total_train_steps = int(batches_per_epoch * num_epochs)
optimizer, schedule = create_optimizer(
init_lr=5e-5, num_warmup_steps=total_train_steps // 5, num_train_steps=total_train_steps
)
metric = load_metric("accuracy")
def compute_metrics(eval_predictions):
logits, labels = eval_predictions
predictions = np.argmax(logits, axis=-1)
return metric.compute(predictions=predictions, references=labels)
metric_callback = KerasMetricCallback(metric_fn=compute_metrics, eval_dataset=tf_validation_dataset)
callbacks = [metric_callback]
discriminator.compile(optimizer=optimizer, loss=loss, metrics=["sparse_categorical_accuracy"])
discriminator.fit(
tf_train_dataset,
epochs=num_epochs,
validation_data=tf_validation_dataset,
callbacks=callbacks
)
print("Evaluate on test data")
results = discriminator.evaluate(tf_validation_dataset)
print("test loss, test acc:", results)
text = "i"
encoded_input = tokenizer(text, return_tensors='tf')
output = discriminator(encoded_input)
prediction = tf.nn.softmax(output["logits"][0], -1)
label = id2label[tf.math.argmax(prediction).numpy()]
print("\n", text, ":", label,
"\n") # ideally 'inventory.v.01' (-> "make or include in an itemized record or report"), but probably only with a better finetuning dataset
text = "get lamp"
encoded_input = tokenizer(text, return_tensors='tf')
output = discriminator(encoded_input)
prediction = tf.nn.softmax(output["logits"][0], -1)
label = id2label[tf.math.argmax(prediction).numpy()]
print("\n", text, ":", label,
"\n") # ideally 'take.v.04' (-> "get into one's hands, take physically"), but probably only with a better finetuning dataset
text = "w"
encoded_input = tokenizer(text, return_tensors='tf')
output = discriminator(encoded_input)
prediction = tf.nn.softmax(output["logits"][0], -1)
label = id2label[tf.math.argmax(prediction).numpy()]
print("\n", text, ":", label,
"\n") # ideally 'travel.v.01' (-> "change location; move, travel, or proceed, also metaphorically"), but probably only with a better finetuning dataset
```
|
javilonso/classificationPolEsp1 | javilonso | 2022-03-30T09:02:50Z | 3 | 0 | transformers | [
"transformers",
"tf",
"bert",
"text-classification",
"generated_from_keras_callback",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-30T07:49:20Z | ---
license: mit
tags:
- generated_from_keras_callback
model-index:
- name: javilonso/classificationPolEsp1
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. -->
# javilonso/classificationPolEsp1
This model is a fine-tuned version of [nlptown/bert-base-multilingual-uncased-sentiment](https://huggingface.co/nlptown/bert-base-multilingual-uncased-sentiment) on an unknown dataset.
It achieves the following results on the evaluation set:
- Train Loss: 0.3728
- Validation Loss: 0.6217
- Epoch: 2
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 17958, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}
- training_precision: mixed_float16
### Training results
| Train Loss | Validation Loss | Epoch |
|:----------:|:---------------:|:-----:|
| 0.6282 | 0.6017 | 0 |
| 0.5129 | 0.6177 | 1 |
| 0.3728 | 0.6217 | 2 |
### Framework versions
- Transformers 4.17.0
- TensorFlow 2.6.0
- Datasets 2.0.0
- Tokenizers 0.11.6
|
neibla/distilbert-base-uncased-finetuned-emotion | neibla | 2022-03-30T08:56:26Z | 9 | 0 | 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 | 2022-03-30T08:22:55Z | ---
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
args: default
metrics:
- name: Accuracy
type: accuracy
value: 0.9255
- name: F1
type: f1
value: 0.9254917237562972
---
<!-- 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: 0.2187
- Accuracy: 0.9255
- F1: 0.9255
## 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 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.855 | 1.0 | 250 | 0.3211 | 0.905 | 0.9017 |
| 0.2561 | 2.0 | 500 | 0.2187 | 0.9255 | 0.9255 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu111
- Datasets 1.16.1
- Tokenizers 0.10.3
|
loulou/distilbert-base-uncased-finetuned-emotion | loulou | 2022-03-30T04:57:58Z | 3 | 0 | 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 | 2022-03-22T04:55:48Z | ---
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
args: default
metrics:
- name: Accuracy
type: accuracy
value: 0.922
- name: F1
type: f1
value: 0.9221931901873676
---
<!-- 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: 0.2285
- Accuracy: 0.922
- F1: 0.9222
## 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 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.8366 | 1.0 | 250 | 0.3212 | 0.9025 | 0.8990 |
| 0.2588 | 2.0 | 500 | 0.2285 | 0.922 | 0.9222 |
### Framework versions
- Transformers 4.18.0.dev0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
scasutt/wav2vec2-large-xlsr-53_toy_train_data_masked_audio | scasutt | 2022-03-30T03:35:01Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-29T11:30:40Z | ---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: wav2vec2-large-xlsr-53_toy_train_data_masked_audio
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-53_toy_train_data_masked_audio
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6445
- Wer: 0.4938
## 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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 20
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 3.3761 | 1.05 | 250 | 3.4022 | 0.9954 |
| 3.0858 | 2.1 | 500 | 3.4684 | 0.9954 |
| 2.6302 | 3.15 | 750 | 1.7989 | 0.9865 |
| 1.1292 | 4.2 | 1000 | 0.8558 | 0.7355 |
| 0.8371 | 5.25 | 1250 | 0.7319 | 0.6621 |
| 0.5992 | 6.3 | 1500 | 0.6848 | 0.6147 |
| 0.5189 | 7.35 | 1750 | 0.6522 | 0.5742 |
| 0.454 | 8.4 | 2000 | 0.6601 | 0.5531 |
| 0.3896 | 9.45 | 2250 | 0.6138 | 0.5439 |
| 0.3678 | 10.5 | 2500 | 0.6436 | 0.5320 |
| 0.3232 | 11.55 | 2750 | 0.5920 | 0.5174 |
| 0.2926 | 12.6 | 3000 | 0.6615 | 0.5107 |
| 0.3041 | 13.65 | 3250 | 0.6311 | 0.5015 |
| 0.2882 | 14.7 | 3500 | 0.6182 | 0.5004 |
| 0.2868 | 15.75 | 3750 | 0.6266 | 0.4943 |
| 0.2508 | 16.81 | 4000 | 0.6587 | 0.4965 |
| 0.2563 | 17.86 | 4250 | 0.6634 | 0.4939 |
| 0.2213 | 18.91 | 4500 | 0.6441 | 0.4925 |
| 0.2255 | 19.96 | 4750 | 0.6445 | 0.4938 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.11.0+cu102
- Datasets 2.0.0
- Tokenizers 0.11.6
|
cammiemw/bert-marco-hdct | cammiemw | 2022-03-30T01:21:38Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"bert",
"text-classification",
"license:cc-by-nc-4.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-30T01:09:55Z | ---
license: cc-by-nc-4.0
---
|
DrishtiSharma/poem-gen-spanish-t5-small-v7 | DrishtiSharma | 2022-03-30T00:34:41Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"t5",
"text2text-generation",
"generated_from_trainer",
"license:mit",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-29T19:14:40Z | ---
license: mit
tags:
- generated_from_trainer
model-index:
- name: poem-gen-spanish-t5-small-v7
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. -->
# poem-gen-spanish-t5-small-v7
This model is a fine-tuned version of [hackathon-pln-es/poem-gen-spanish-t5-small](https://huggingface.co/hackathon-pln-es/poem-gen-spanish-t5-small) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 2.9201
## 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: 0.000333
- train_batch_size: 6
- eval_batch_size: 6
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:------:|:---------------:|
| 3.1716 | 0.73 | 30000 | 3.1114 |
| 2.9666 | 1.46 | 60000 | 3.0271 |
| 2.8292 | 2.19 | 90000 | 2.9531 |
| 2.7264 | 2.93 | 120000 | 2.9126 |
| 2.6057 | 3.66 | 150000 | 2.9175 |
| 2.4876 | 4.39 | 180000 | 2.9077 |
| 2.3791 | 5.12 | 210000 | 2.9240 |
| 2.3515 | 5.85 | 240000 | 2.9169 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
DrishtiSharma/poem-gen-spanish-t5-small-v6 | DrishtiSharma | 2022-03-29T23:45:09Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"t5",
"text2text-generation",
"generated_from_trainer",
"license:mit",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-29T18:58:46Z | ---
license: mit
tags:
- generated_from_trainer
model-index:
- name: poem-gen-spanish-t5-small-v6
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. -->
# poem-gen-spanish-t5-small-v6
This model is a fine-tuned version of [hackathon-pln-es/poem-gen-spanish-t5-small](https://huggingface.co/hackathon-pln-es/poem-gen-spanish-t5-small) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 2.8831
## 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: 7.5e-05
- train_batch_size: 6
- eval_batch_size: 6
- 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 |
|:-------------:|:-----:|:------:|:---------------:|
| 2.8551 | 0.73 | 30000 | 2.9296 |
| 2.6961 | 1.46 | 60000 | 2.9005 |
| 2.5756 | 2.19 | 90000 | 2.8786 |
| 2.5095 | 2.93 | 120000 | 2.8621 |
| 2.4061 | 3.66 | 150000 | 2.8830 |
| 2.3161 | 4.39 | 180000 | 2.8865 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
efederici/sentence-it5-base | efederici | 2022-03-29T23:09:01Z | 35 | 4 | sentence-transformers | [
"sentence-transformers",
"pytorch",
"t5",
"feature-extraction",
"sentence-similarity",
"transformers",
"it",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | sentence-similarity | 2022-03-29T19:57:59Z | ---
pipeline_tag: sentence-similarity
language:
- it
tags:
- sentence-transformers
- feature-extraction
- sentence-similarity
- transformers
---
# sentence-IT5-base
This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 512 dimensional dense vector space and can be used for tasks like clustering or semantic search. It is a T5 ([IT5](https://huggingface.co/gsarti/it5-base)) base model. It is trained on a dataset made from question/context pairs ([squad-it](https://github.com/crux82/squad-it)), tags/news-article pairs, headline/text pairs ([change-it](https://huggingface.co/datasets/gsarti/change_it)) and on [stsb](https://huggingface.co/datasets/stsb_multi_mt/viewer/it/train).
## 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 = ["Questo è un esempio di frase", "Questo è un ulteriore esempio"]
model = SentenceTransformer('efederici/sentence-IT5-base')
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 = ["Questo è un esempio di frase", "Questo è un ulteriore esempio"]
# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained('efederici/sentence-IT5-base')
model = AutoModel.from_pretrained('efederici/sentence-IT5-base')
# 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)
```
## Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': None, 'do_lower_case': False}) with Transformer model: T5EncoderModel
(1): Pooling({'word_embedding_dimension': 512, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
)
```
|
espnet/bur_openslr80_hubert | espnet | 2022-03-29T22:19:50Z | 0 | 0 | null | [
"region:us"
] | null | 2022-03-28T22:04:54Z | <!-- Generated by scripts/utils/show_asr_result.sh -->
# RESULTS
## Environments
- date: `Mon Mar 21 22:59:35 UTC 2022`
- python version: `3.9.7 (default, Sep 16 2021, 13:09:58) [GCC 7.5.0]`
- espnet version: `espnet 0.10.7a1`
- pytorch version: `pytorch 1.10.1`
- Git hash: `7ae4efd81778436a98b822483e8123adba6aa430`
- Commit date: `Tue Mar 15 20:11:18 2022 -0400`
## asr_train_asr_hubert_transformer_adam_specaug_raw_bpe150
### WER
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|---|---|---|---|---|---|---|---|---|
|decode_asr_batch_size1_lm_lm_train_lm_bpe150_valid.loss.ave_asr_model_valid.acc.best/bur_test|480|4227|39.1|50.4|10.5|6.1|67.0|99.8|
### CER
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|---|---|---|---|---|---|---|---|---|
|decode_asr_batch_size1_lm_lm_train_lm_bpe150_valid.loss.ave_asr_model_valid.acc.best/bur_test|480|33345|82.2|7.6|10.1|3.6|21.4|99.8|
### TER
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|---|---|---|---|---|---|---|---|---|
|decode_asr_batch_size1_lm_lm_train_lm_bpe150_valid.loss.ave_asr_model_valid.acc.best/bur_test|480|18237|70.7|17.7|11.6|2.5|31.8|99.8|
|
Chikashi/t5-small-finetuned-cnndm_3epoch | Chikashi | 2022-03-29T19:28:09Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"t5",
"text2text-generation",
"generated_from_trainer",
"dataset:cnn_dailymail",
"license:apache-2.0",
"model-index",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-29T00:14:31Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- cnn_dailymail
metrics:
- rouge
model-index:
- name: t5-small-finetuned-cnndm_3epoch
results:
- task:
name: Sequence-to-sequence Language Modeling
type: text2text-generation
dataset:
name: cnn_dailymail
type: cnn_dailymail
args: 3.0.0
metrics:
- name: Rouge1
type: rouge
value: 24.5435
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# t5-small-finetuned-cnndm_3epoch
This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on the cnn_dailymail dataset.
It achieves the following results on the evaluation set:
- Loss: 1.6622
- Rouge1: 24.5435
- Rouge2: 11.7919
- Rougel: 20.2929
- Rougelsum: 23.1661
- Gen Len: 18.9996
## 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
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|:-------------:|:-----:|:------:|:---------------:|:-------:|:-------:|:-------:|:---------:|:-------:|
| 1.9113 | 0.14 | 5000 | 1.7162 | 24.4374 | 11.6932 | 20.1741 | 23.0427 | 18.9997 |
| 1.8772 | 0.28 | 10000 | 1.7008 | 24.3715 | 11.6699 | 20.1387 | 22.9772 | 18.9997 |
| 1.8609 | 0.42 | 15000 | 1.6911 | 24.4174 | 11.6986 | 20.1756 | 23.0205 | 18.9997 |
| 1.8564 | 0.56 | 20000 | 1.6871 | 24.4374 | 11.6801 | 20.1663 | 23.0366 | 18.9995 |
| 1.8495 | 0.7 | 25000 | 1.6796 | 24.4019 | 11.6901 | 20.177 | 23.034 | 18.999 |
| 1.8448 | 0.84 | 30000 | 1.6787 | 24.4813 | 11.7227 | 20.1985 | 23.0847 | 18.999 |
| 1.8427 | 0.98 | 35000 | 1.6762 | 24.4905 | 11.7591 | 20.2548 | 23.1006 | 18.9993 |
| 1.8341 | 1.11 | 40000 | 1.6747 | 24.4743 | 11.7124 | 20.1782 | 23.0726 | 18.9996 |
| 1.822 | 1.25 | 45000 | 1.6753 | 24.4797 | 11.7292 | 20.2319 | 23.0816 | 18.9993 |
| 1.8262 | 1.39 | 50000 | 1.6713 | 24.4865 | 11.7079 | 20.2214 | 23.0919 | 18.9986 |
| 1.8281 | 1.53 | 55000 | 1.6702 | 24.5095 | 11.7364 | 20.2534 | 23.1264 | 18.9991 |
| 1.8228 | 1.67 | 60000 | 1.6678 | 24.5153 | 11.7595 | 20.2544 | 23.1138 | 18.9993 |
| 1.824 | 1.81 | 65000 | 1.6662 | 24.5324 | 11.7804 | 20.2671 | 23.1498 | 18.9997 |
| 1.8265 | 1.95 | 70000 | 1.6648 | 24.5795 | 11.7917 | 20.2935 | 23.1855 | 18.9992 |
| 1.8179 | 2.09 | 75000 | 1.6658 | 24.5426 | 11.804 | 20.2861 | 23.1586 | 18.9996 |
| 1.8147 | 2.23 | 80000 | 1.6646 | 24.5429 | 11.7914 | 20.2889 | 23.1542 | 18.9993 |
| 1.8026 | 2.37 | 85000 | 1.6632 | 24.5451 | 11.8045 | 20.2781 | 23.1555 | 18.9996 |
| 1.8141 | 2.51 | 90000 | 1.6643 | 24.5078 | 11.7781 | 20.2631 | 23.121 | 18.9996 |
| 1.8124 | 2.65 | 95000 | 1.6628 | 24.5728 | 11.7958 | 20.2875 | 23.178 | 18.9996 |
| 1.8098 | 2.79 | 100000 | 1.6635 | 24.5534 | 11.7998 | 20.2979 | 23.169 | 18.9996 |
| 1.8153 | 2.93 | 105000 | 1.6622 | 24.5435 | 11.7919 | 20.2929 | 23.1661 | 18.9996 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
GleamEyeBeast/ascend | GleamEyeBeast | 2022-03-29T16:49:48Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-29T01:37:59Z | ---
tags:
- generated_from_trainer
model-index:
- name: ascend
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. -->
# ascend
This model is a fine-tuned version of [GleamEyeBeast/ascend](https://huggingface.co/GleamEyeBeast/ascend) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.3718
- Wer: 0.6412
- Cer: 0.2428
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|
| 0.5769 | 1.0 | 688 | 1.1864 | 0.7716 | 0.3159 |
| 0.5215 | 2.0 | 1376 | 1.1613 | 0.7504 | 0.2965 |
| 0.4188 | 3.0 | 2064 | 1.1644 | 0.7389 | 0.2950 |
| 0.3695 | 4.0 | 2752 | 1.1937 | 0.7184 | 0.2815 |
| 0.3404 | 5.0 | 3440 | 1.1947 | 0.7083 | 0.2719 |
| 0.2885 | 6.0 | 4128 | 1.2314 | 0.7108 | 0.2685 |
| 0.2727 | 7.0 | 4816 | 1.2243 | 0.6850 | 0.2616 |
| 0.2417 | 8.0 | 5504 | 1.2506 | 0.6767 | 0.2608 |
| 0.2207 | 9.0 | 6192 | 1.2804 | 0.6922 | 0.2595 |
| 0.2195 | 10.0 | 6880 | 1.2582 | 0.6818 | 0.2575 |
| 0.1896 | 11.0 | 7568 | 1.3101 | 0.6814 | 0.2545 |
| 0.1961 | 12.0 | 8256 | 1.2793 | 0.6706 | 0.2526 |
| 0.1752 | 13.0 | 8944 | 1.2643 | 0.6584 | 0.2509 |
| 0.1638 | 14.0 | 9632 | 1.3152 | 0.6588 | 0.2482 |
| 0.1522 | 15.0 | 10320 | 1.3098 | 0.6433 | 0.2439 |
| 0.1351 | 16.0 | 11008 | 1.3253 | 0.6537 | 0.2447 |
| 0.1266 | 17.0 | 11696 | 1.3394 | 0.6365 | 0.2418 |
| 0.1289 | 18.0 | 12384 | 1.3718 | 0.6412 | 0.2443 |
| 0.1204 | 19.0 | 13072 | 1.3708 | 0.6433 | 0.2433 |
| 0.1189 | 20.0 | 13760 | 1.3718 | 0.6412 | 0.2428 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
tbosse/bert-base-german-cased-finetuned-subj_v1 | tbosse | 2022-03-29T15:59:49Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"bert",
"token-classification",
"generated_from_trainer",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | token-classification | 2022-03-29T14:22:30Z | ---
license: mit
tags:
- generated_from_trainer
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: bert-base-german-cased-finetuned-subj_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. -->
# bert-base-german-cased-finetuned-subj_v1
This model is a fine-tuned version of [bert-base-german-cased](https://huggingface.co/bert-base-german-cased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1594
- Precision: 0.1875
- Recall: 0.0077
- F1: 0.0147
- Accuracy: 0.9508
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
| No log | 1.0 | 136 | 0.1591 | 1.0 | 0.0051 | 0.0102 | 0.9523 |
| No log | 2.0 | 272 | 0.1571 | 0.375 | 0.0077 | 0.015 | 0.9518 |
| No log | 3.0 | 408 | 0.1594 | 0.1875 | 0.0077 | 0.0147 | 0.9508 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
sayef/fsner-bert-base-uncased | sayef | 2022-03-29T14:20:35Z | 9 | 6 | transformers | [
"transformers",
"pytorch",
"bert",
"feature-extraction",
"arxiv:2008.10570",
"text-embeddings-inference",
"endpoints_compatible",
"region:us"
] | feature-extraction | 2022-03-02T23:29:05Z | # FSNER
Implemented by [sayef](https://huggingface.co/sayef).
# Overview
The FSNER model was proposed in [Example-Based Named Entity Recognition](https://arxiv.org/abs/2008.10570) by Morteza
Ziyadi, Yuting Sun, Abhishek Goswami, Jade Huang, Weizhu Chen. To identify entity spans in a new domain, it uses a
train-free few-shot learning approach inspired by question-answering.
## Abstract
> We present a novel approach to named entity recognition (NER) in the presence of scarce data that we call example-based NER. Our train-free few-shot learning approach takes inspiration from question-answering to identify entity spans in a new and unseen domain. In comparison with the current state-of-the-art, the proposed method performs significantly better, especially when using a low number of support examples.
## Model Training Details
| identifier | epochs | datasets |
| ---------- |:------:|:-----------------------------------------------------------------------------------------------:|
| [sayef/fsner-bert-base-uncased](https://huggingface.co/sayef/fsner-bert-base-uncased) | 25 | ontonotes5, conll2003, wnut2017, mit_movie_trivia, mit_restaurant and fin (Alvarado et al.). |
## Installation and Example Usage
You can use the FSNER model in 3 ways:
1. Install directly from PyPI: `pip install fsner` and import the model as shown in the code example below
or
2. Install from source: `python install .` and import the model as shown in the code example below
or
3. Clone [repo](https://github.com/sayef/fsner) and add absolute path of `fsner/src` directory to your PYTHONPATH and
import the model as shown in the code example below
```python
import json
from fsner import FSNERModel, FSNERTokenizerUtils, pretty_embed
query_texts = [
"Does Luke's serve lunch?",
"Chang does not speak Taiwanese very well.",
"I like Berlin."
]
# Each list in supports are the examples of one entity type
# Wrap entities around with [E] and [/E] in the examples.
# Each sentence should have only one pair of [E] ... [/E]
support_texts = {
"Restaurant": [
"What time does [E] Subway [/E] open for breakfast?",
"Is there a [E] China Garden [/E] restaurant in newark?",
"Does [E] Le Cirque [/E] have valet parking?",
"Is there a [E] McDonalds [/E] on main street?",
"Does [E] Mike's Diner [/E] offer huge portions and outdoor dining?"
],
"Language": [
"Although I understood no [E] French [/E] in those days , I was prepared to spend the whole day with Chien - chien .",
"like what the hell 's that called in [E] English [/E] ? I have to register to be here like since I 'm a foreigner .",
"So , I 'm also working on an [E] English [/E] degree because that 's my real interest .",
"Al - Jazeera TV station , established in November 1996 in Qatar , is an [E] Arabic - language [/E] news TV station broadcasting global news and reports nonstop around the clock .",
"They think it 's far better for their children to be here improving their [E] English [/E] than sitting at home in front of a TV . \"",
"The only solution seemed to be to have her learn [E] French [/E] .",
"I have to read sixty pages of [E] Russian [/E] today ."
]
}
device = 'cpu'
tokenizer = FSNERTokenizerUtils("sayef/fsner-bert-base-uncased")
queries = tokenizer.tokenize(query_texts).to(device)
supports = tokenizer.tokenize(list(support_texts.values())).to(device)
model = FSNERModel("sayef/fsner-bert-base-uncased")
model.to(device)
p_starts, p_ends = model.predict(queries, supports)
# One can prepare supports once and reuse multiple times with different queries
# ------------------------------------------------------------------------------
# start_token_embeddings, end_token_embeddings = model.prepare_supports(supports)
# p_starts, p_ends = model.predict(queries, start_token_embeddings=start_token_embeddings,
# end_token_embeddings=end_token_embeddings)
output = tokenizer.extract_entity_from_scores(query_texts, queries, p_starts, p_ends,
entity_keys=list(support_texts.keys()), thresh=0.50)
print(json.dumps(output, indent=2))
# install displacy for pretty embed
pretty_embed(query_texts, output, list(support_texts.keys()))
```
<!DOCTYPE html>
<html lang="en">
<head>
<title>displaCy</title>
</head>
<body style="font-size: 16px; font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Helvetica, Arial, sans-serif, 'Apple Color Emoji', 'Segoe UI Emoji', 'Segoe UI Symbol'; padding: 4rem 2rem; direction: ltr">
<figure style="margin-bottom: 6rem">
<div class="entities" style="line-height: 2.5; direction: ltr">
<div class="entities" style="line-height: 2.5; direction: ltr">Does
<mark class="entity" style="background: #7aecec; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;">
Luke's
<span style="font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem">Restaurant</span>
</mark>
serve lunch?</div>
<div class="entities" style="line-height: 2.5; direction: ltr">Chang does not speak
<mark class="entity" style="background: #bfeeb7; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;">
Taiwanese
<span style="font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem">Language</span>
</mark>
very well.</div>
<div class="entities" style="line-height: 2.5; direction: ltr">I like Berlin.</div>
</div>
</figure>
</body>
</html>
## Datasets preparation
1. We need to convert dataset into the following format. Let's say we have a dataset file train.json like following.
2. Each list in supports are the examples of one entity type
3. Wrap entities around with [E] and [/E] in the examples.
4. Each example should have only one pair of [E] ... [/E].
```json
{
"CARDINAL_NUMBER": [
"Washington , cloudy , [E] 2 [/E] to 6 degrees .",
"New Dehli , sunny , [E] 6 [/E] to 19 degrees .",
"Well this is number [E] two [/E] .",
"....."
],
"LANGUAGE": [
"They do n't have the Quicken [E] Dutch [/E] version ?",
"they learned a lot of [E] German [/E] .",
"and then [E] Dutch [/E] it 's Mifrau",
"...."
],
"MONEY": [
"Per capita personal income ranged from $ [E] 11,116 [/E] in Mississippi to $ 23,059 in Connecticut ... .",
"The trade surplus was [E] 582 million US dollars [/E] .",
"It settled with a loss of 4.95 cents at $ [E] 1.3210 [/E] a pound .",
"...."
]
}
```
2. Converted ontonotes5 dataset can be found here:
1. [train](https://gist.githubusercontent.com/sayef/46deaf7e6c6e1410b430ddc8aff9c557/raw/ea7ae2ae933bfc9c0daac1aa52a9dc093d5b36f4/ontonotes5.train.json)
2. [dev](https://gist.githubusercontent.com/sayef/46deaf7e6c6e1410b430ddc8aff9c557/raw/ea7ae2ae933bfc9c0daac1aa52a9dc093d5b36f4/ontonotes5.dev.json)
3. Then trainer script can be used to train/evaluate your fsner model.
```bash
fsner trainer --pretrained-model bert-base-uncased --mode train --train-data train.json --val-data val.json \
--train-batch-size 6 --val-batch-size 6 --n-examples-per-entity 10 --neg-example-batch-ratio 1/3 --max-epochs 25 --device gpu \
--gpus -1 --strategy ddp
``` |
Rishav-hub/xlm-roberta-base-finetuned-panx-de | Rishav-hub | 2022-03-29T11:05:37Z | 3 | 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 | 2022-03-29T10:26:12Z | ---
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: 0.8591260810195721
---
<!-- 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: 0.1352
- F1: 0.8591
## 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 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 0.257 | 1.0 | 525 | 0.1512 | 0.8302 |
| 0.1305 | 2.0 | 1050 | 0.1401 | 0.8447 |
| 0.0817 | 3.0 | 1575 | 0.1352 | 0.8591 |
### Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu111
- Datasets 1.16.1
- Tokenizers 0.10.3
|
beston91/gpt2-xl_ft_logits_5k_experiment | beston91 | 2022-03-29T10:27:12Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"gpt2",
"text-generation",
"generated_from_trainer",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-29T03:13:26Z | ---
tags:
- generated_from_trainer
model-index:
- name: gpt2-xl_ft_logits_5k_experiment
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. -->
# gpt2-xl_ft_logits_5k_experiment
This model is a fine-tuned version of [gpt2-xl](https://huggingface.co/gpt2-xl) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 6.8601
## 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
- gradient_accumulation_steps: 64
- total_train_batch_size: 512
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100.0
- num_epochs: 4
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| No log | 0.9 | 7 | 6.1556 |
| No log | 1.9 | 14 | 6.3365 |
| No log | 2.9 | 21 | 6.5909 |
| No log | 3.9 | 28 | 6.8601 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
### Perplexity
Score: 17.589759826660156 |
KeithHorgan/TweetClimateAnalysis | KeithHorgan | 2022-03-29T10:01:24Z | 4 | 1 | transformers | [
"transformers",
"pytorch",
"roberta",
"text-classification",
"autotrain",
"unk",
"dataset:KeithHorgan98/autotrain-data-TweetClimateAnalysis",
"co2_eq_emissions",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-29T10:16:42Z | ---
tags: autotrain
language: unk
widget:
- text: "Climate Change is a hoax"
- text: "It is freezing, where is global warming"
datasets:
- KeithHorgan98/autotrain-data-TweetClimateAnalysis
co2_eq_emissions: 133.19491276284793
---
# Model Trained Using AutoTrain
- Problem type: Multi-class Classification
- Model ID: 678720226
- CO2 Emissions (in grams): 133.19491276284793
## Validation Metrics
- Loss: 0.4864234924316406
- Accuracy: 0.865424430641822
- Macro F1: 0.7665472174344069
- Micro F1: 0.8654244306418221
- Weighted F1: 0.8586375445115083
- Macro Precision: 0.8281449061702826
- Micro Precision: 0.865424430641822
- Weighted Precision: 0.8619727477790186
- Macro Recall: 0.736576343905098
- Micro Recall: 0.865424430641822
- Weighted Recall: 0.865424430641822
## Usage
You can use cURL to access this model:
```
$ curl -X POST -H "Authorization: Bearer YOUR_API_KEY" -H "Content-Type: application/json" -d '{"inputs": "I love AutoTrain"}' https://api-inference.huggingface.co/models/KeithHorgan98/autotrain-TweetClimateAnalysis-678720226
```
Or Python API:
```
from transformers import AutoModelForSequenceClassification, AutoTokenizer
model = AutoModelForSequenceClassification.from_pretrained("KeithHorgan98/autotrain-TweetClimateAnalysis-678720226", use_auth_token=True)
tokenizer = AutoTokenizer.from_pretrained("KeithHorgan98/autotrain-TweetClimateAnalysis-678720226", use_auth_token=True)
inputs = tokenizer("I love AutoTrain", return_tensors="pt")
outputs = model(**inputs)
``` |
ai4bharat/MultiIndicWikiBioSS | ai4bharat | 2022-03-29T09:22:47Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"mbart",
"text2text-generation",
"wikibio",
"multilingual",
"nlp",
"indicnlp",
"as",
"bn",
"hi",
"kn",
"ml",
"or",
"pa",
"ta",
"te",
"dataset:ai4bharat/IndicWikiBio",
"arxiv:2203.05437",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-16T11:36:23Z | ---
tags:
- wikibio
- multilingual
- nlp
- indicnlp
datasets:
- ai4bharat/IndicWikiBio
language:
- as
- bn
- hi
- kn
- ml
- or
- pa
- ta
- te
licenses:
- cc-by-nc-4.0
widget:
- text: <TAG> name </TAG> राम नरेश पांडेय <TAG> office </TAG> विधायक - 205 - कुशीनगर विधान सभा निर्वाचन क्षेत्र , उत्तर प्रदेश <TAG> term </TAG> 1967 से 1968 <TAG> nationality </TAG> भारतीय </s> <2hi>
---
# MultiIndicWikiBioSS
MultiIndicWikiBioSS is a multilingual, sequence-to-sequence pre-trained model, a [IndicBARTSS](https://huggingface.co/ai4bharat/IndicBARTSS) checkpoint fine-tuned on the 9 languages of [IndicWikiBio](https://huggingface.co/datasets/ai4bharat/IndicWikiBio) dataset. For fine-tuning details,
see the [paper](https://arxiv.org/abs/2203.05437). You can use MultiIndicWikiBioSS to build biography generation applications for Indian languages by fine-tuning the model with supervised training data. Some salient features of the MultiIndicWikiBioSS are:
<ul>
<li >Supported languages: Assamese, Bengali, Hindi, Oriya, Punjabi, Kannada, Malayalam, Tamil, and Telugu. 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> Fine-tuned on an Indic language corpora (34,653 examples). </li>
<li> Unlike ai4bharat/MultiIndicWikiBioUnified, each language is written in its own script, so you do not need to perform any script mapping to/from Devanagari. </li>
</ul>
You can read more about MultiIndicWikiBioSS in this <a href="https://arxiv.org/abs/2203.05437">paper</a>.
## Using this model in `transformers`
```
from transformers import MBartForConditionalGeneration, AutoModelForSeq2SeqLM
from transformers import AlbertTokenizer, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("ai4bharat/MultiIndicWikiBioSS", do_lower_case=False, use_fast=False, keep_accents=True)
# Or use tokenizer = AlbertTokenizer.from_pretrained("ai4bharat/MultiIndicWikiBioSS", do_lower_case=False, use_fast=False, keep_accents=True)
model = AutoModelForSeq2SeqLM.from_pretrained("ai4bharat/MultiIndicWikiBioSS")
# Or use model = MBartForConditionalGeneration.from_pretrained("ai4bharat/MultiIndicWikiBioSS")
# 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>', '<2hi>', '<2kn>', '<2ml>', '<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("<TAG> name </TAG> भीखा लाल <TAG> office </TAG> विधायक - 318 - हसनगंज विधान सभा निर्वाचन क्षेत्र , उत्तर प्रदेश <TAG> term </TAG> 1957 से 1962 <TAG> nationality </TAG> भारतीय</s><2hi>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids
out = tokenizer("<2hi> भीखा लाल ,भारत के उत्तर प्रदेश की दूसरी विधानसभा सभा में विधायक रहे। </s>", add_special_tokens=False, return_tensors="pt", padding=True).input_ids
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,no_repeat_ngram_size=3,encoder_no_repeat_ngram_size=3, 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("<2hi>"))
# Decode to get output strings
decoded_output=tokenizer.decode(model_output[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
print(decoded_output) # __भीखा लाल ,भारत के उत्तर प्रदेश की दूसरी विधानसभा सभा में विधायक रहे।
```
## Benchmarks
Scores on the `IndicWikiBio` test sets are as follows:
Language | RougeL
---------|----------------------------
as | 56.50
bn | 56.58
hi | 67.34
kn | 39.37
ml | 38.42
or | 70.71
pa | 52.78
ta | 51.11
te | 51.72
## Citation
If you use this model, please cite the following paper:
```
@inproceedings{Kumar2022IndicNLGSM,
title={IndicNLG Suite: Multilingual Datasets for Diverse NLG Tasks in Indic Languages},
author={Aman Kumar and Himani Shrotriya and Prachi Sahu and Raj Dabre and Ratish Puduppully and Anoop Kunchukuttan and Amogh Mishra and Mitesh M. Khapra and Pratyush Kumar},
year={2022},
url = "https://arxiv.org/abs/2203.05437"
}
```
# License
The model is available under the MIT License. |
Davlan/m2m100_418M-eng-yor-mt | Davlan | 2022-03-29T09:21:53Z | 820 | 1 | transformers | [
"transformers",
"pytorch",
"m2m_100",
"text2text-generation",
"arxiv:2103.08647",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z | Hugging Face's logo
---
language:
- yo
- en
datasets:
- JW300 + [Menyo-20k](https://huggingface.co/datasets/menyo20k_mt)
---
# m2m100_418M-eng-yor-mt
## Model description
**m2m100_418M-eng-yor-mt** is a **machine translation** model from English language to Yorùbá language based on a fine-tuned facebook/m2m100_418M model. It establishes a **strong baseline** for automatically translating texts from English to Yorùbá.
Specifically, this model is a *facebook/m2m100_418M* model that was fine-tuned on JW300 Yorùbá corpus and [Menyo-20k](https://huggingface.co/datasets/menyo20k_mt).
#### Limitations and bias
This model is limited by its training dataset. This may not generalize well for all use cases in different domains.
## Training data
This model was fine-tuned on JW300 corpus and [Menyo-20k](https://huggingface.co/datasets/menyo20k_mt) dataset
## Training procedure
This model was trained on NVIDIA V100 GPU
## Eval results on Test set (BLEU score)
Fine-tuning m2m100_418M achieves **13.39 BLEU** on [Menyo-20k test set](https://arxiv.org/abs/2103.08647) while mt5-base achieves 9.82
### BibTeX entry and citation info
By David Adelani
```
```
|
Davlan/m2m100_418M-yor-eng-mt | Davlan | 2022-03-29T09:21:03Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"m2m_100",
"text2text-generation",
"arxiv:2103.08647",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z | Hugging Face's logo
---
language:
- yo
- en
datasets:
- JW300 + [Menyo-20k](https://huggingface.co/datasets/menyo20k_mt)
---
# m2m100_418M-eng-yor-mt
## Model description
**m2m100_418M-yor-eng-mt** is a **machine translation** model from Yorùbá language to English language based on a fine-tuned facebook/m2m100_418M model. It establishes a **strong baseline** for automatically translating texts from Yorùbá to English.
Specifically, this model is a *facebook/m2m100_418M* model that was fine-tuned on JW300 Yorùbá corpus and [Menyo-20k](https://huggingface.co/datasets/menyo20k_mt).
#### Limitations and bias
This model is limited by its training dataset. This may not generalize well for all use cases in different domains.
## Training data
This model was fine-tuned on JW300 corpus and [Menyo-20k](https://huggingface.co/datasets/menyo20k_mt) dataset
## Training procedure
This model was trained on NVIDIA V100 GPU
## Eval results on Test set (BLEU score)
Fine-tuning m2m100_418M achieves **16.76 BLEU** on [Menyo-20k test set](https://arxiv.org/abs/2103.08647) while mt5-base achieves 15.57
### BibTeX entry and citation info
By David Adelani
```
```
|
PereLluis13/wav2vec2-xls-r-1b-ca | PereLluis13 | 2022-03-29T08:44:49Z | 17 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"automatic-speech-recognition",
"collectivat/tv3_parla",
"generated_from_trainer",
"hf-asr-leaderboard",
"mozilla-foundation/common_voice_8_0",
"projecte-aina/parlament_parla",
"robust-speech-event",
"ca",
"dataset:mozilla-foundation/common_voice_8_0",
"dataset:collectivat/tv3_parla",
"dataset:projecte-aina/parlament_parla",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-02T23:29:04Z | ---
language:
- ca
license: apache-2.0
tags:
- automatic-speech-recognition
- collectivat/tv3_parla
- generated_from_trainer
- hf-asr-leaderboard
- mozilla-foundation/common_voice_8_0
- projecte-aina/parlament_parla
- robust-speech-event
datasets:
- mozilla-foundation/common_voice_8_0
- collectivat/tv3_parla
- projecte-aina/parlament_parla
model-index:
- name: wav2vec2-xls-r-1b-ca
results:
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: mozilla-foundation/common_voice_8_0 ca
type: mozilla-foundation/common_voice_8_0
args: ca
metrics:
- name: Test WER
type: wer
value: 11.030639657300516
- name: Test CER
type: cer
value: 2.8405630530040634
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: projecte-aina/parlament_parla ca
type: projecte-aina/parlament_parla
args: clean
metrics:
- name: Test WER
type: wer
value: 6.483115660665961
- name: Test CER
type: cer
value: 2.0212863746191828
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: collectivat/tv3_parla ca
type: collectivat/tv3_parla
args: ca
metrics:
- name: Test WER
type: wer
value: 17.917773414943988
- name: Test CER
type: cer
value: 8.872589572206396
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: Robust Speech Event - Catalan Dev Data
type: speech-recognition-community-v2/dev_data
args: ca
metrics:
- name: Test WER
type: wer
value: 27.126683954209097
- name: Test CER
type: cer
value: 14.213308815078726
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Robust Speech Event - Test Data
type: speech-recognition-community-v2/eval_data
args: ca
metrics:
- name: Test WER
type: wer
value: 18.7
---
<!-- 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-xls-r-1b-ca
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 - CA, the [tv3_parla](https://huggingface.co/datasets/collectivat/tv3_parla) and [parlament_parla](https://huggingface.co/datasets/projecte-aina/parlament_parla) datasets.
## Model description
Please check the original [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) Model card. This is just a finetuned version of that model.
## Intended uses & limitations
As any model trained on crowdsourced data, this model can show the biases and particularities of the data and model used to train this model. Moreover, since this is a speech recognition model, it may underperform for some lower-resourced dialects for the catalan language.
## Training and evaluation data
## Training procedure
The data is preprocessed to remove characters not on the catalan alphabet. Moreover, numbers are verbalized using code provided by [@ccoreilly](https://github.com/ccoreilly), which can be found on the text/ folder or [here](https://github.com/CollectivaT-dev/catotron-cpu/blob/master/text/numbers_ca.py).
### Training results
Check the Tensorboard tab to check the training profile and evaluation results along training. The model was evaluated on the test splits for each of the datasets used during training.
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2000
- num_epochs: 10.0
- mixed_precision_training: Native AMP
### Framework versions
- Transformers 4.17.0.dev0
- Pytorch 1.10.2+cu102
- Datasets 1.18.3
- Tokenizers 0.11.0
# Thanks
Want to thank both [@ccoreilly](https://github.com/ccoreilly) and [@gullabi](https://github.com/gullabi) who have contributed with their own resources and knowledge into making this model possible. |
johnowhitaker/sketchy_unet_rn34 | johnowhitaker | 2022-03-29T08:02:43Z | 0 | 0 | null | [
"license:cc-by-4.0",
"region:us"
] | null | 2022-03-29T07:57:40Z | ---
license: cc-by-4.0
---
This is the exported model for a small project I' working on, to test integration with spaces.
It is a fastai model and needs some custom code to work.
For now please ignore :) |
gayanin/t5-small-med-term-conditional-masking-0 | gayanin | 2022-03-29T03:19:04Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"t5",
"text2text-generation",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-28T22:04:47Z | ---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: t5-small-med-term-conditional-masking-0
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# t5-small-med-term-conditional-masking-0
This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6688
- Rouge2 Precision: 0.694
- Rouge2 Recall: 0.4781
- Rouge2 Fmeasure: 0.5479
## 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: 10
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge2 Precision | Rouge2 Recall | Rouge2 Fmeasure |
|:-------------:|:-----:|:------:|:---------------:|:----------------:|:-------------:|:---------------:|
| 0.9525 | 1.0 | 13915 | 0.8148 | 0.6657 | 0.4581 | 0.5252 |
| 0.8541 | 2.0 | 27830 | 0.7562 | 0.6779 | 0.4694 | 0.5371 |
| 0.8183 | 3.0 | 41745 | 0.7268 | 0.6827 | 0.4722 | 0.5405 |
| 0.8033 | 4.0 | 55660 | 0.7074 | 0.6861 | 0.4729 | 0.5419 |
| 0.7727 | 5.0 | 69575 | 0.6934 | 0.6872 | 0.4726 | 0.5419 |
| 0.7704 | 6.0 | 83490 | 0.6832 | 0.6901 | 0.4742 | 0.544 |
| 0.7485 | 7.0 | 97405 | 0.6771 | 0.6926 | 0.4772 | 0.5469 |
| 0.7528 | 8.0 | 111320 | 0.6722 | 0.6934 | 0.4782 | 0.5478 |
| 0.7535 | 9.0 | 125235 | 0.6696 | 0.6944 | 0.4782 | 0.5481 |
| 0.7444 | 10.0 | 139150 | 0.6688 | 0.694 | 0.4781 | 0.5479 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
DrishtiSharma/wav2vec2-base-finetuned-sentiment-mesd-v9 | DrishtiSharma | 2022-03-29T00:52:52Z | 5 | 2 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"audio-classification",
"generated_from_trainer",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | audio-classification | 2022-03-29T00:13:34Z | ---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: wav2vec2-base-finetuned-sentiment-mesd-v9
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-sentiment-mesd-v9
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3500
- Accuracy: 0.9154
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0001
- train_batch_size: 64
- eval_batch_size: 40
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.01
- num_epochs: 100
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| No log | 0.86 | 3 | 1.7825 | 0.1846 |
| 1.9553 | 1.86 | 6 | 1.7212 | 0.4308 |
| 1.9553 | 2.86 | 9 | 1.6164 | 0.3769 |
| 2.002 | 3.86 | 12 | 1.4904 | 0.3769 |
| 1.6191 | 4.86 | 15 | 1.4426 | 0.4385 |
| 1.6191 | 5.86 | 18 | 1.3516 | 0.5231 |
| 1.6209 | 6.86 | 21 | 1.2176 | 0.5538 |
| 1.6209 | 7.86 | 24 | 1.1683 | 0.5692 |
| 1.371 | 8.86 | 27 | 1.0885 | 0.5923 |
| 1.1568 | 9.86 | 30 | 1.0152 | 0.6385 |
| 1.1568 | 10.86 | 33 | 0.9289 | 0.6385 |
| 1.1023 | 11.86 | 36 | 0.9141 | 0.6308 |
| 1.1023 | 12.86 | 39 | 0.8526 | 0.6462 |
| 0.9448 | 13.86 | 42 | 0.8420 | 0.6769 |
| 0.7972 | 14.86 | 45 | 0.7976 | 0.6692 |
| 0.7972 | 15.86 | 48 | 0.8192 | 0.7308 |
| 0.7793 | 16.86 | 51 | 0.7108 | 0.7615 |
| 0.7793 | 17.86 | 54 | 0.6712 | 0.7769 |
| 0.6468 | 18.86 | 57 | 0.6684 | 0.7923 |
| 0.5083 | 19.86 | 60 | 0.6922 | 0.7385 |
| 0.5083 | 20.86 | 63 | 0.6148 | 0.7923 |
| 0.4988 | 21.86 | 66 | 0.5846 | 0.7923 |
| 0.4988 | 22.86 | 69 | 0.6050 | 0.8154 |
| 0.4123 | 23.86 | 72 | 0.5506 | 0.7846 |
| 0.3511 | 24.86 | 75 | 0.6095 | 0.7846 |
| 0.3511 | 25.86 | 78 | 0.5916 | 0.8154 |
| 0.3268 | 26.86 | 81 | 0.5912 | 0.8077 |
| 0.3268 | 27.86 | 84 | 0.5142 | 0.8538 |
| 0.3036 | 28.86 | 87 | 0.5492 | 0.8077 |
| 0.3066 | 29.86 | 90 | 0.6007 | 0.8231 |
| 0.3066 | 30.86 | 93 | 0.5748 | 0.8231 |
| 0.2538 | 31.86 | 96 | 0.6027 | 0.7692 |
| 0.2538 | 32.86 | 99 | 0.6979 | 0.7462 |
| 0.2281 | 33.86 | 102 | 0.7002 | 0.7615 |
| 0.2183 | 34.86 | 105 | 0.6650 | 0.7769 |
| 0.2183 | 35.86 | 108 | 0.5192 | 0.8462 |
| 0.2202 | 36.86 | 111 | 0.5389 | 0.8308 |
| 0.2202 | 37.86 | 114 | 0.5050 | 0.8385 |
| 0.1906 | 38.86 | 117 | 0.5722 | 0.7769 |
| 0.154 | 39.86 | 120 | 0.5239 | 0.8308 |
| 0.154 | 40.86 | 123 | 0.4448 | 0.8615 |
| 0.1474 | 41.86 | 126 | 0.4623 | 0.8615 |
| 0.1474 | 42.86 | 129 | 0.4282 | 0.8615 |
| 0.1345 | 43.86 | 132 | 0.5087 | 0.8615 |
| 0.1567 | 44.86 | 135 | 0.4859 | 0.8385 |
| 0.1567 | 45.86 | 138 | 0.6603 | 0.8077 |
| 0.1731 | 46.86 | 141 | 0.5379 | 0.8385 |
| 0.1731 | 47.86 | 144 | 0.8666 | 0.7538 |
| 0.1606 | 48.86 | 147 | 0.7518 | 0.8 |
| 0.1484 | 49.86 | 150 | 0.5986 | 0.8385 |
| 0.1484 | 50.86 | 153 | 0.6368 | 0.8231 |
| 0.2256 | 51.86 | 156 | 0.4639 | 0.8692 |
| 0.2256 | 52.86 | 159 | 0.5533 | 0.8462 |
| 0.1178 | 53.86 | 162 | 0.5038 | 0.8615 |
| 0.0815 | 54.86 | 165 | 0.5052 | 0.8692 |
| 0.0815 | 55.86 | 168 | 0.4337 | 0.8846 |
| 0.0998 | 56.86 | 171 | 0.4422 | 0.8769 |
| 0.0998 | 57.86 | 174 | 0.4317 | 0.8692 |
| 0.0855 | 58.86 | 177 | 0.4025 | 0.8923 |
| 0.0962 | 59.86 | 180 | 0.4605 | 0.8769 |
| 0.0962 | 60.86 | 183 | 0.4356 | 0.8769 |
| 0.0763 | 61.86 | 186 | 0.4614 | 0.8769 |
| 0.0763 | 62.86 | 189 | 0.4382 | 0.8846 |
| 0.0902 | 63.86 | 192 | 0.4701 | 0.8692 |
| 0.0654 | 64.86 | 195 | 0.4922 | 0.8692 |
| 0.0654 | 65.86 | 198 | 0.5413 | 0.8538 |
| 0.0651 | 66.86 | 201 | 0.5759 | 0.8615 |
| 0.0651 | 67.86 | 204 | 0.4238 | 0.9 |
| 0.0822 | 68.86 | 207 | 0.3500 | 0.9154 |
| 0.0625 | 69.86 | 210 | 0.3878 | 0.8923 |
| 0.0625 | 70.86 | 213 | 0.4952 | 0.8615 |
| 0.0548 | 71.86 | 216 | 0.4544 | 0.8615 |
| 0.0548 | 72.86 | 219 | 0.5497 | 0.8769 |
| 0.054 | 73.86 | 222 | 0.4434 | 0.8846 |
| 0.0543 | 74.86 | 225 | 0.4732 | 0.8769 |
| 0.0543 | 75.86 | 228 | 0.4425 | 0.8923 |
| 0.0881 | 76.86 | 231 | 0.4788 | 0.8769 |
| 0.0881 | 77.86 | 234 | 0.5448 | 0.8769 |
| 0.061 | 78.86 | 237 | 0.4221 | 0.9077 |
| 0.0567 | 79.86 | 240 | 0.4404 | 0.8769 |
| 0.0567 | 80.86 | 243 | 0.4099 | 0.9 |
| 0.052 | 81.86 | 246 | 0.5259 | 0.8769 |
| 0.052 | 82.86 | 249 | 0.5874 | 0.8692 |
| 0.0444 | 83.86 | 252 | 0.5555 | 0.8846 |
| 0.0332 | 84.86 | 255 | 0.5156 | 0.8615 |
| 0.0332 | 85.86 | 258 | 0.4564 | 0.8615 |
| 0.0449 | 86.86 | 261 | 0.4826 | 0.8692 |
| 0.0449 | 87.86 | 264 | 0.4726 | 0.8615 |
| 0.0385 | 88.86 | 267 | 0.4206 | 0.8846 |
| 0.0356 | 89.86 | 270 | 0.4050 | 0.8769 |
| 0.0356 | 90.86 | 273 | 0.4161 | 0.8923 |
| 0.0391 | 91.86 | 276 | 0.4100 | 0.9077 |
| 0.0391 | 92.86 | 279 | 0.4047 | 0.9 |
| 0.0249 | 93.86 | 282 | 0.4044 | 0.9 |
| 0.0399 | 94.86 | 285 | 0.3968 | 0.8846 |
| 0.0399 | 95.86 | 288 | 0.3802 | 0.9 |
| 0.031 | 96.86 | 291 | 0.3689 | 0.9 |
| 0.031 | 97.86 | 294 | 0.3616 | 0.9077 |
| 0.036 | 98.86 | 297 | 0.3584 | 0.9077 |
| 0.0386 | 99.86 | 300 | 0.3574 | 0.9077 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
i-was-neo-first/hubert-large-ami-shard-experiment-colab | i-was-neo-first | 2022-03-29T00:39:37Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"hubert",
"automatic-speech-recognition",
"generated_from_trainer",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-20T02:10:11Z | ---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: hubert-large-ami-shard-experiment-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. -->
# hubert-large-ami-shard-experiment-colab
This model is a fine-tuned version of [facebook/hubert-large-ls960-ft](https://huggingface.co/facebook/hubert-large-ls960-ft) on the None dataset.
It achieves the following results on the evaluation set:
- eval_loss: nan
- eval_wer: 1.0
- eval_runtime: 6.0682
- eval_samples_per_second: 16.479
- eval_steps_per_second: 2.142
- epoch: 1.02
- step: 1000
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0001
- train_batch_size: 1
- 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
### Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu111
- Datasets 1.18.3
- Tokenizers 0.10.3
|
sanchit-gandhi/wav2vec2-2-bart-large-cnn | sanchit-gandhi | 2022-03-29T00:24:41Z | 25 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"speech-encoder-decoder",
"automatic-speech-recognition",
"generated_from_trainer",
"dataset:librispeech_asr",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-22T16:26:40Z | ---
tags:
- generated_from_trainer
datasets:
- librispeech_asr
model-index:
- name: ''
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. -->
#
This model was trained from scratch on the librispeech_asr dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3524
- Wer: 0.1042
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- gradient_accumulation_steps: 16
- total_train_batch_size: 256
- total_eval_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: 10.0
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 2.7605 | 4.5 | 500 | 2.6299 | 1.4451 |
| 0.1177 | 9.01 | 1000 | 0.3524 | 0.1042 |
### Framework versions
- Transformers 4.17.0.dev0
- Pytorch 1.10.2+cu113
- Datasets 1.18.3
- Tokenizers 0.11.0
|
frtna/ted_mt-Spanish-to-Italian | frtna | 2022-03-28T22:04:21Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"marian",
"text2text-generation",
"generated_from_trainer",
"dataset:new_dataset",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:05Z | ---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- new_dataset
model-index:
- name: ted_mt-Spanish-to-Italian
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. -->
# ted_mt-Spanish-to-Italian
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-es-it](https://huggingface.co/Helsinki-NLP/opus-mt-es-it) on the new_dataset 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: 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: 1
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Sacrebleu | Gen Len |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:-------:|
| No log | 1.0 | 46 | 1.4873 | 29.6133 | 26.9081 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.11.0
- Datasets 2.0.0
- Tokenizers 0.11.6
|
jorge-henao/spanish-t5-small-disco-poetry | jorge-henao | 2022-03-28T21:26:45Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"t5",
"text2text-generation",
"generated_from_trainer",
"license:mit",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-28T18:15:25Z | ---
license: mit
tags:
- generated_from_trainer
model-index:
- name: spanish-t5-small-disco-poetry
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. -->
# spanish-t5-small-disco-poetry
This model is a fine-tuned version of [flax-community/spanish-t5-small](https://huggingface.co/flax-community/spanish-t5-small) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0477
## 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: 6
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 0.1417 | 1.0 | 1284 | 0.0577 |
| 0.0902 | 2.0 | 2568 | 0.0516 |
| 0.0803 | 3.0 | 3852 | 0.0494 |
| 0.0733 | 4.0 | 5136 | 0.0488 |
| 0.0683 | 5.0 | 6420 | 0.0480 |
| 0.067 | 6.0 | 7704 | 0.0477 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
hf-test/xls-r-300m-sv | hf-test | 2022-03-28T20:07:57Z | 28 | 3 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"hf-asr-leaderboard",
"hello",
"model_for_talk",
"mozilla-foundation/common_voice_7_0",
"robust-speech-event",
"sv",
"dataset:mozilla-foundation/common_voice_7_0",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-02T23:29:05Z | ---
language:
- sv-SE
license: apache-2.0
tags:
- automatic-speech-recognition
- generated_from_trainer
- hf-asr-leaderboard
- hello
- model_for_talk
- mozilla-foundation/common_voice_7_0
- robust-speech-event
- sv
datasets:
- mozilla-foundation/common_voice_7_0
model-index:
- name: XLS-R-300M - Swedish
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 7
type: mozilla-foundation/common_voice_7_0
args: sv-SE
metrics:
- name: Test WER
type: wer
value: 16.98
- name: Test CER
type: cer
value: 5.66
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Robust Speech Event - Dev Data
type: speech-recognition-community-v2/dev_data
args: sv
metrics:
- name: Test WER
type: wer
value: 27.01
- name: Test CER
type: cer
value: 13.14
---
<!-- 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. -->
# XLS-R-300m-SV
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_7_0 - SV-SE dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3171
- Wer: 0.2468
## 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: 7.5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2000
- num_epochs: 50.0
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:-----:|:---------------:|:------:|
| 3.3349 | 1.45 | 500 | 3.2858 | 1.0 |
| 2.9298 | 2.91 | 1000 | 2.9225 | 1.0000 |
| 2.0839 | 4.36 | 1500 | 1.1546 | 0.8295 |
| 1.7093 | 5.81 | 2000 | 0.6827 | 0.5701 |
| 1.5855 | 7.27 | 2500 | 0.5597 | 0.4947 |
| 1.4831 | 8.72 | 3000 | 0.4923 | 0.4527 |
| 1.4416 | 10.17 | 3500 | 0.4670 | 0.4270 |
| 1.3848 | 11.63 | 4000 | 0.4341 | 0.3980 |
| 1.3749 | 13.08 | 4500 | 0.4203 | 0.4011 |
| 1.3311 | 14.53 | 5000 | 0.4310 | 0.3961 |
| 1.317 | 15.99 | 5500 | 0.3898 | 0.4322 |
| 1.2799 | 17.44 | 6000 | 0.3806 | 0.3572 |
| 1.2771 | 18.89 | 6500 | 0.3828 | 0.3427 |
| 1.2451 | 20.35 | 7000 | 0.3702 | 0.3359 |
| 1.2182 | 21.8 | 7500 | 0.3685 | 0.3270 |
| 1.2152 | 23.26 | 8000 | 0.3650 | 0.3308 |
| 1.1837 | 24.71 | 8500 | 0.3568 | 0.3187 |
| 1.1721 | 26.16 | 9000 | 0.3659 | 0.3249 |
| 1.1764 | 27.61 | 9500 | 0.3547 | 0.3145 |
| 1.1606 | 29.07 | 10000 | 0.3514 | 0.3104 |
| 1.1431 | 30.52 | 10500 | 0.3469 | 0.3062 |
| 1.1047 | 31.97 | 11000 | 0.3313 | 0.2979 |
| 1.1315 | 33.43 | 11500 | 0.3298 | 0.2992 |
| 1.1022 | 34.88 | 12000 | 0.3296 | 0.2973 |
| 1.0935 | 36.34 | 12500 | 0.3278 | 0.2926 |
| 1.0676 | 37.79 | 13000 | 0.3208 | 0.2868 |
| 1.0571 | 39.24 | 13500 | 0.3322 | 0.2885 |
| 1.0536 | 40.7 | 14000 | 0.3245 | 0.2831 |
| 1.0525 | 42.15 | 14500 | 0.3285 | 0.2826 |
| 1.0464 | 43.6 | 15000 | 0.3223 | 0.2796 |
| 1.0415 | 45.06 | 15500 | 0.3166 | 0.2774 |
| 1.0356 | 46.51 | 16000 | 0.3177 | 0.2746 |
| 1.04 | 47.96 | 16500 | 0.3150 | 0.2735 |
| 1.0209 | 49.42 | 17000 | 0.3175 | 0.2731 |
### Framework versions
- Transformers 4.16.0.dev0
- Pytorch 1.10.0+cu102
- Datasets 1.17.1.dev0
- Tokenizers 0.10.3
#### Evaluation Commands
1. To evaluate on `mozilla-foundation/common_voice_7_0` with split `test`
```bash
python eval.py --model_id hf-test/xls-r-300m-sv --dataset mozilla-foundation/common_voice_7_0 --config sv-SE --split test
```
2. To evaluate on `speech-recognition-community-v2/dev_data`
```bash
python eval.py --model_id hf-test/xls-r-300m-sv --dataset speech-recognition-community-v2/dev_data --config sv --split validation --chunk_length_s 5.0 --stride_length_s 1.0
```
### Inference With LM
```python
import torch
from datasets import load_dataset
from transformers import AutoModelForCTC, AutoProcessor
import torchaudio.functional as F
model_id = "hf-test/xls-r-300m-sv"
sample_iter = iter(load_dataset("mozilla-foundation/common_voice_7_0", "sv-SE", split="test", streaming=True, use_auth_token=True))
sample = next(sample_iter)
resampled_audio = F.resample(torch.tensor(sample["audio"]["array"]), 48_000, 16_000).numpy()
model = AutoModelForCTC.from_pretrained(model_id)
processor = AutoProcessor.from_pretrained(model_id)
input_values = processor(resampled_audio, return_tensors="pt").input_values
with torch.no_grad():
logits = model(input_values).logits
transcription = processor.batch_decode(logits.numpy()).text
# => "jag lämnade grovjobbet åt honom"
```
### Eval results on Common Voice 7 "test" (WER):
| Without LM | With LM (run `./eval.py`) |
|---|---|
| 24.68 | 16.98 |
|
Symbermine/rare-puppers | Symbermine | 2022-03-28T19:38:23Z | 57 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"vit",
"image-classification",
"huggingpics",
"model-index",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | image-classification | 2022-03-28T19:38:13Z | ---
tags:
- image-classification
- pytorch
- huggingpics
metrics:
- accuracy
model-index:
- name: rare-puppers
results:
- task:
name: Image Classification
type: image-classification
metrics:
- name: Accuracy
type: accuracy
value: 0.9285714030265808
---
# rare-puppers
Autogenerated by HuggingPics🤗🖼️
Create your own image classifier for **anything** by running [the demo on Google Colab](https://colab.research.google.com/github/nateraw/huggingpics/blob/main/HuggingPics.ipynb).
Report any issues with the demo at the [github repo](https://github.com/nateraw/huggingpics).
## Example Images
#### Husky siberiano
#### cocker spaniel
#### galgo
#### labrador
#### pastor aleman
 |
joniponi/distilbert-base-uncased-finetuned-emotion | joniponi | 2022-03-28T19:06:11Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"distilbert",
"text-classification",
"generated_from_trainer",
"license:apache-2.0",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text-classification | 2022-03-28T15:57:55Z | ---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
model-index:
- name: distilbert-base-uncased-finetuned-emotion
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-emotion
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.8357
- Accuracy: 0.6309
- F1: 0.6469
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
| 0.9559 | 1.0 | 78 | 0.8585 | 0.6223 | 0.6363 |
| 0.7998 | 2.0 | 156 | 0.8472 | 0.6202 | 0.6354 |
| 0.7207 | 3.0 | 234 | 0.8357 | 0.6309 | 0.6469 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
DrishtiSharma/wav2vec2-base-finetuned-sentiment-mesd-v2 | DrishtiSharma | 2022-03-28T19:04:20Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"audio-classification",
"generated_from_trainer",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | audio-classification | 2022-03-28T17:20:20Z | ---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: wav2vec2-base-finetuned-sentiment-mesd-v2
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# wav2vec2-base-finetuned-sentiment-mesd-v2
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.7213
- Accuracy: 0.3923
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1.25e-05
- train_batch_size: 64
- eval_batch_size: 40
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 20
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| No log | 0.86 | 3 | 1.7961 | 0.1462 |
| 1.9685 | 1.86 | 6 | 1.7932 | 0.1692 |
| 1.9685 | 2.86 | 9 | 1.7891 | 0.2 |
| 2.1386 | 3.86 | 12 | 1.7820 | 0.2923 |
| 1.9492 | 4.86 | 15 | 1.7750 | 0.2923 |
| 1.9492 | 5.86 | 18 | 1.7684 | 0.2846 |
| 2.1143 | 6.86 | 21 | 1.7624 | 0.3231 |
| 2.1143 | 7.86 | 24 | 1.7561 | 0.3308 |
| 2.0945 | 8.86 | 27 | 1.7500 | 0.3462 |
| 1.9121 | 9.86 | 30 | 1.7443 | 0.3385 |
| 1.9121 | 10.86 | 33 | 1.7386 | 0.3231 |
| 2.0682 | 11.86 | 36 | 1.7328 | 0.3231 |
| 2.0682 | 12.86 | 39 | 1.7272 | 0.3769 |
| 2.0527 | 13.86 | 42 | 1.7213 | 0.3923 |
| 1.8705 | 14.86 | 45 | 1.7154 | 0.3846 |
| 1.8705 | 15.86 | 48 | 1.7112 | 0.3846 |
| 2.0263 | 16.86 | 51 | 1.7082 | 0.3769 |
| 2.0263 | 17.86 | 54 | 1.7044 | 0.3846 |
| 2.0136 | 18.86 | 57 | 1.7021 | 0.3846 |
| 1.8429 | 19.86 | 60 | 1.7013 | 0.3846 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
DrishtiSharma/xls-r-es-test-lm-finetuned-sentiment-mesd | DrishtiSharma | 2022-03-28T19:03:37Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"tensorboard",
"wav2vec2",
"audio-classification",
"generated_from_trainer",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | audio-classification | 2022-03-28T14:54:48Z | ---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: xls-r-es-test-lm-finetuned-sentiment-mesd
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. -->
# xls-r-es-test-lm-finetuned-sentiment-mesd
This model is a fine-tuned version of [glob-asr/xls-r-es-test-lm](https://huggingface.co/glob-asr/xls-r-es-test-lm) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.7851
- Accuracy: 0.2385
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1.25e-05
- train_batch_size: 64
- eval_batch_size: 40
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 20
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| No log | 0.86 | 3 | 1.7876 | 0.1923 |
| 1.9709 | 1.86 | 6 | 1.7869 | 0.2 |
| 1.9709 | 2.86 | 9 | 1.7859 | 0.2308 |
| 2.146 | 3.86 | 12 | 1.7851 | 0.2385 |
| 1.9622 | 4.86 | 15 | 1.7842 | 0.1923 |
| 1.9622 | 5.86 | 18 | 1.7834 | 0.1769 |
| 2.137 | 6.86 | 21 | 1.7823 | 0.1923 |
| 2.137 | 7.86 | 24 | 1.7812 | 0.1923 |
| 2.1297 | 8.86 | 27 | 1.7800 | 0.1846 |
| 1.9502 | 9.86 | 30 | 1.7787 | 0.1846 |
| 1.9502 | 10.86 | 33 | 1.7772 | 0.1846 |
| 2.1234 | 11.86 | 36 | 1.7760 | 0.1846 |
| 2.1234 | 12.86 | 39 | 1.7748 | 0.1846 |
| 2.1186 | 13.86 | 42 | 1.7736 | 0.1846 |
| 1.9401 | 14.86 | 45 | 1.7725 | 0.1846 |
| 1.9401 | 15.86 | 48 | 1.7715 | 0.1923 |
| 2.112 | 16.86 | 51 | 1.7706 | 0.1923 |
| 2.112 | 17.86 | 54 | 1.7701 | 0.1923 |
| 2.1094 | 18.86 | 57 | 1.7697 | 0.2 |
| 1.934 | 19.86 | 60 | 1.7696 | 0.2 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6
|
scasutt/wav2vec2-large-xlsr-53_toy_train_data_fast_10pct | scasutt | 2022-03-28T18:53:54Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"generated_from_trainer",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-28T12:30:15Z | ---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: wav2vec2-large-xlsr-53_toy_train_data_fast_10pct
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-53_toy_train_data_fast_10pct
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6983
- Wer: 0.5026
## 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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 20
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 3.3619 | 1.05 | 250 | 3.4334 | 1.0 |
| 3.0818 | 2.1 | 500 | 3.4914 | 1.0 |
| 2.3245 | 3.15 | 750 | 1.6483 | 0.9486 |
| 1.0233 | 4.2 | 1000 | 0.8817 | 0.7400 |
| 0.7522 | 5.25 | 1250 | 0.7374 | 0.6529 |
| 0.5343 | 6.3 | 1500 | 0.6972 | 0.6068 |
| 0.4452 | 7.35 | 1750 | 0.6757 | 0.5740 |
| 0.4275 | 8.4 | 2000 | 0.6789 | 0.5551 |
| 0.3688 | 9.45 | 2250 | 0.6468 | 0.5394 |
| 0.3363 | 10.5 | 2500 | 0.6798 | 0.5358 |
| 0.3036 | 11.55 | 2750 | 0.6439 | 0.5265 |
| 0.3173 | 12.6 | 3000 | 0.6898 | 0.5196 |
| 0.2985 | 13.65 | 3250 | 0.6791 | 0.5169 |
| 0.288 | 14.7 | 3500 | 0.6442 | 0.5090 |
| 0.2673 | 15.75 | 3750 | 0.6984 | 0.5119 |
| 0.2575 | 16.81 | 4000 | 0.7146 | 0.5084 |
| 0.239 | 17.86 | 4250 | 0.6847 | 0.5040 |
| 0.2266 | 18.91 | 4500 | 0.6900 | 0.5028 |
| 0.22 | 19.96 | 4750 | 0.6983 | 0.5026 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.11.0+cu102
- Datasets 2.0.0
- Tokenizers 0.11.6
|
aapot/wav2vec2-large-xlsr-53-finnish | aapot | 2022-03-28T17:56:36Z | 9 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"wav2vec2",
"automatic-speech-recognition",
"audio",
"speech",
"xlsr-fine-tuning-week",
"fi",
"dataset:common_voice",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-02T23:29:05Z | ---
language: fi
datasets:
- common_voice
metrics:
- wer
tags:
- audio
- automatic-speech-recognition
- speech
- xlsr-fine-tuning-week
license: apache-2.0
model-index:
- name: XLSR Wav2Vec2 Finnish by Aapo Tanskanen
results:
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice fi
type: common_voice
args: fi
metrics:
- name: Test WER
type: wer
value: 32.378771
---
# NOTE: this is an old model and should not be used anymore!! There are a lot better newer models available at our orgnization hub: [Finnish-NLP/wav2vec2-xlsr-1b-finnish-lm-v2](https://huggingface.co/Finnish-NLP/wav2vec2-xlsr-1b-finnish-lm-v2) and [Finnish-NLP/wav2vec2-xlsr-300m-finnish-lm](https://huggingface.co/Finnish-NLP/wav2vec2-xlsr-300m-finnish-lm)
# Wav2Vec2-Large-XLSR-53-Finnish
Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on Finnish using the [Common Voice](https://huggingface.co/datasets/common_voice), [CSS10 Finnish](https://www.kaggle.com/bryanpark/finnish-single-speaker-speech-dataset) and [Finnish parliament session 2](https://b2share.eudat.eu/records/4df422d631544ce682d6af1d4714b2d4) datasets.
When using this model, make sure that your speech input is sampled at 16kHz.
## Usage
The model can be used directly (without a language model) as follows:
```python
import librosa
import torch
import torchaudio
from datasets import load_dataset
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
test_dataset = load_dataset("common_voice", "fi", split="test[:2%]")
processor = Wav2Vec2Processor.from_pretrained("aapot/wav2vec2-large-xlsr-53-finnish")
model = Wav2Vec2ForCTC.from_pretrained("aapot/wav2vec2-large-xlsr-53-finnish")
resampler = lambda sr, y: librosa.resample(y.numpy().squeeze(), sr, 16_000)
# Preprocessing the datasets.
# We need to read the audio files as arrays
def speech_file_to_array_fn(batch):
speech_array, sampling_rate = torchaudio.load(batch["path"])
batch["speech"] = resampler(sampling_rate, speech_array).squeeze()
return batch
test_dataset = test_dataset.map(speech_file_to_array_fn)
inputs = processor(test_dataset["speech"][:2], sampling_rate=16_000, return_tensors="pt", padding=True)
with torch.no_grad():
logits = model(inputs.input_values, attention_mask=inputs.attention_mask).logits
predicted_ids = torch.argmax(logits, dim=-1)
print("Prediction:", processor.batch_decode(predicted_ids))
print("Reference:", test_dataset["sentence"][:2])
```
## Evaluation
The model can be evaluated as follows on the Finnish test data of Common Voice.
```python
import librosa
import torch
import torchaudio
from datasets import load_dataset, load_metric
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
import re
test_dataset = load_dataset("common_voice", "fi", split="test")
wer = load_metric("wer")
processor = Wav2Vec2Processor.from_pretrained("aapot/wav2vec2-large-xlsr-53-finnish")
model = Wav2Vec2ForCTC.from_pretrained("aapot/wav2vec2-large-xlsr-53-finnish")
model.to("cuda")
chars_to_ignore_regex = '[\,\?\.\!\-\;\:\"\“\%\‘\”\�\'\...\…\–\é]'
resampler = lambda sr, y: librosa.resample(y.numpy().squeeze(), sr, 16_000)
# Preprocessing the datasets.
# We need to read the audio files as arrays
def speech_file_to_array_fn(batch):
batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower()
speech_array, sampling_rate = torchaudio.load(batch["path"])
batch["speech"] = resampler(sampling_rate, speech_array).squeeze()
return batch
test_dataset = test_dataset.map(speech_file_to_array_fn)
# Preprocessing the datasets.
# We need to read the audio files as arrays
def evaluate(batch):
inputs = processor(batch["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
with torch.no_grad():
logits = model(inputs.input_values.to("cuda"), attention_mask=inputs.attention_mask.to("cuda")).logits
pred_ids = torch.argmax(logits, dim=-1)
batch["pred_strings"] = processor.batch_decode(pred_ids)
return batch
result = test_dataset.map(evaluate, batched=True, batch_size=8)
print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"], references=result["sentence"])))
```
**Test Result**: 32.378771 %
## Training
The Common Voice `train`, `validation` and `other` datasets were used for training as well as `CSS10 Finnish` and `Finnish parliament session 2` datasets.
The script used for training can be found from [Google Colab](https://colab.research.google.com/drive/1vnEGC9BnNRmVyIHj-0UsVulh_cUYSGWA?usp=sharing) |
aapot/wav2vec2-xlsr-300m-finnish-lm | aapot | 2022-03-28T17:22:08Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"wav2vec2",
"automatic-speech-recognition",
"fi",
"finnish",
"generated_from_trainer",
"hf-asr-leaderboard",
"robust-speech-event",
"dataset:mozilla-foundation/common_voice_7_0",
"arxiv:2111.09296",
"license:apache-2.0",
"model-index",
"endpoints_compatible",
"region:us"
] | automatic-speech-recognition | 2022-03-02T23:29:05Z | ---
license: apache-2.0
language: fi
metrics:
- wer
- cer
tags:
- automatic-speech-recognition
- fi
- finnish
- generated_from_trainer
- hf-asr-leaderboard
- robust-speech-event
datasets:
- mozilla-foundation/common_voice_7_0
model-index:
- name: wav2vec2-xlsr-300m-finnish-lm
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 7
type: mozilla-foundation/common_voice_7_0
args: fi
metrics:
- name: Test WER
type: wer
value: 8.16
- name: Test CER
type: cer
value: 1.97
---
# Wav2Vec2 XLS-R for Finnish ASR
This acoustic model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) for Finnish ASR. The model has been fine-tuned with 275.6 hours of Finnish transcribed speech data. Wav2Vec2 XLS-R was introduced in
[this paper](https://arxiv.org/abs/2111.09296) and first released at [this page](https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#wav2vec-20).
This repository also includes Finnish KenLM language model used in the decoding phase with the acoustic model.
**Note**: this model is exactly the same as the [Finnish-NLP/wav2vec2-xlsr-300m-finnish-lm](https://huggingface.co/Finnish-NLP/wav2vec2-xlsr-300m-finnish-lm) model so this model has just been copied/moved to the `Finnish-NLP` Hugging Face organization.
## Model description
Wav2Vec2 XLS-R is Facebook AI's large-scale multilingual pretrained model for speech. It is pretrained on 436k hours of unlabeled speech, including VoxPopuli, MLS, CommonVoice, BABEL, and VoxLingua107. It uses the wav2vec 2.0 objective, in 128 languages.
You can read more about the pretrained model from [this blog](https://ai.facebook.com/blog/xls-r-self-supervised-speech-processing-for-128-languages) and [this paper](https://arxiv.org/abs/2111.09296).
This model is fine-tuned version of the pretrained model (300 million parameter variant) for Finnish ASR.
## Intended uses & limitations
You can use this model for Finnish ASR (speech-to-text) task.
### How to use
Check the [run-finnish-asr-models.ipynb](https://huggingface.co/aapot/wav2vec2-xlsr-300m-finnish-lm/blob/main/run-finnish-asr-models.ipynb) notebook in this repository for an detailed example on how to use this model.
### Limitations and bias
This model was fine-tuned with audio samples which maximum length was 20 seconds so this model most likely works the best for quite short audios of similar length. However, you can try this model with a lot longer audios too and see how it works. If you encounter out of memory errors with very long audio files you can use the audio chunking method introduced in [this blog post](https://huggingface.co/blog/asr-chunking).
A vast majority of the data used for fine-tuning was from the Finnish Parliament dataset so this model may not generalize so well to very different domains like common daily spoken Finnish with dialects etc. In addition, audios of the datasets tend to be adult male dominated so this model may not work as well for speeches of children and women, for example.
The Finnish KenLM language model used in the decoding phase has been trained with text data from the audio transcriptions and from a subset of Finnish Wikipedia. Thus, the decoder's language model may not generalize to very different language, for example to spoken daily language with dialects (because especially the Wikipedia contains mostly formal Finnish language). It may be beneficial to train your own KenLM language model for your domain language and use that in the decoding.
## Training data
This model was fine-tuned with 275.6 hours of Finnish transcribed speech data from following datasets:
| Dataset | Hours | % of total hours |
|:------------------------------------------------------------------------------------------------------------------------------ |:--------:|:----------------:|
| [Common Voice 7.0 Finnish train + evaluation + other splits](https://huggingface.co/datasets/mozilla-foundation/common_voice_7_0) | 9.70 h | 3.52 % |
| [Finnish parliament session 2](https://b2share.eudat.eu/records/4df422d631544ce682d6af1d4714b2d4) | 0.24 h | 0.09 % |
| [VoxPopuli Finnish](https://github.com/facebookresearch/voxpopuli) | 21.97 h | 7.97 % |
| [CSS10 Finnish](https://github.com/kyubyong/css10) | 10.32 h | 3.74 % |
| [Aalto Finnish Parliament ASR Corpus](http://urn.fi/urn:nbn:fi:lb-2021051903) | 228.00 h | 82.73 % |
| [Finnish Broadcast Corpus](http://urn.fi/urn:nbn:fi:lb-2016042502) | 5.37 h | 1.95 % |
Datasets were filtered to include maximum length of 20 seconds long audio samples.
## Training procedure
This model was trained during [Robust Speech Challenge Event](https://discuss.huggingface.co/t/open-to-the-community-robust-speech-recognition-challenge/13614) organized by Hugging Face. Training was done on a Tesla V100 GPU, sponsored by OVHcloud.
Training script was provided by Hugging Face and it is available [here](https://github.com/huggingface/transformers/blob/main/examples/research_projects/robust-speech-event/run_speech_recognition_ctc_bnb.py). We only modified its data loading for our custom datasets.
For the KenLM language model training, we followed the [blog post tutorial](https://huggingface.co/blog/wav2vec2-with-ngram) provided by Hugging Face. Training data for the 5-gram KenLM were text transcriptions of the audio training data and 100k random samples of cleaned [Finnish Wikipedia](https://huggingface.co/datasets/wikipedia) (August 2021) dataset.
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-04
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: [8-bit Adam](https://github.com/facebookresearch/bitsandbytes) with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 10
- mixed_precision_training: Native AMP
The pretrained `facebook/wav2vec2-xls-r-300m` model was initialized with following hyperparameters:
- attention_dropout: 0.094
- hidden_dropout: 0.047
- feat_proj_dropout: 0.04
- mask_time_prob: 0.082
- layerdrop: 0.041
- activation_dropout: 0.055
- ctc_loss_reduction: "mean"
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:-----:|:---------------:|:------:|
| 0.973 | 0.17 | 500 | 0.5750 | 0.6844 |
| 0.713 | 0.34 | 1000 | 0.3356 | 0.4518 |
| 0.6563 | 0.5 | 1500 | 0.3007 | 0.4039 |
| 0.642 | 0.67 | 2000 | 0.2619 | 0.3674 |
| 0.6203 | 0.84 | 2500 | 0.2488 | 0.3558 |
| 0.6016 | 1.01 | 3000 | 0.2795 | 0.3835 |
| 0.5423 | 1.17 | 3500 | 0.2652 | 0.3310 |
| 0.5639 | 1.34 | 4000 | 0.2479 | 0.3462 |
| 0.586 | 1.51 | 4500 | 0.2409 | 0.3295 |
| 0.5169 | 1.68 | 5000 | 0.2728 | 0.3352 |
| 0.5176 | 1.84 | 5500 | 0.2254 | 0.3149 |
| 0.4983 | 2.01 | 6000 | 0.2169 | 0.3009 |
| 0.4982 | 2.18 | 6500 | 0.2215 | 0.3079 |
| 0.4898 | 2.35 | 7000 | 0.2174 | 0.3023 |
| 0.4922 | 2.51 | 7500 | 0.2217 | 0.3081 |
| 0.5025 | 2.68 | 8000 | 0.2002 | 0.2710 |
| 0.4745 | 2.85 | 8500 | 0.1935 | 0.2783 |
| 0.4377 | 3.02 | 9000 | 0.1859 | 0.2742 |
| 0.4511 | 3.18 | 9500 | 0.2038 | 0.2786 |
| 0.4411 | 3.35 | 10000 | 0.1863 | 0.2651 |
| 0.4501 | 3.52 | 10500 | 0.1948 | 0.2605 |
| 0.4557 | 3.69 | 11000 | 0.1872 | 0.2695 |
| 0.4493 | 3.85 | 11500 | 0.1888 | 0.2632 |
| 0.4047 | 4.02 | 12000 | 0.1818 | 0.2559 |
| 0.4319 | 4.19 | 12500 | 0.1896 | 0.2648 |
| 0.4162 | 4.36 | 13000 | 0.1953 | 0.2595 |
| 0.4046 | 4.52 | 13500 | 0.1864 | 0.2606 |
| 0.4195 | 4.69 | 14000 | 0.1843 | 0.2467 |
| 0.4146 | 4.86 | 14500 | 0.1686 | 0.2450 |
| 0.378 | 5.03 | 15000 | 0.1731 | 0.2401 |
| 0.3792 | 5.19 | 15500 | 0.1676 | 0.2325 |
| 0.3855 | 5.36 | 16000 | 0.1740 | 0.2326 |
| 0.4029 | 5.53 | 16500 | 0.1674 | 0.2345 |
| 0.386 | 5.7 | 17000 | 0.1735 | 0.2280 |
| 0.3811 | 5.86 | 17500 | 0.1692 | 0.2258 |
| 0.3607 | 6.03 | 18000 | 0.1797 | 0.2279 |
| 0.3604 | 6.2 | 18500 | 0.1651 | 0.2206 |
| 0.3362 | 6.37 | 19000 | 0.1627 | 0.2199 |
| 0.3611 | 6.53 | 19500 | 0.1652 | 0.2172 |
| 0.3671 | 6.7 | 20000 | 0.1564 | 0.2140 |
| 0.3769 | 6.87 | 20500 | 0.1525 | 0.2101 |
| 0.3539 | 7.04 | 21000 | 0.1639 | 0.2096 |
| 0.3225 | 7.21 | 21500 | 0.1611 | 0.2087 |
| 0.3323 | 7.37 | 22000 | 0.1633 | 0.2008 |
| 0.3327 | 7.54 | 22500 | 0.1692 | 0.1975 |
| 0.3456 | 7.71 | 23000 | 0.1555 | 0.1991 |
| 0.3058 | 7.88 | 23500 | 0.1590 | 0.1959 |
| 0.3034 | 8.04 | 24000 | 0.1531 | 0.1973 |
| 0.2925 | 8.21 | 24500 | 0.1583 | 0.1978 |
| 0.2967 | 8.38 | 25000 | 0.1546 | 0.1906 |
| 0.2974 | 8.55 | 25500 | 0.1540 | 0.1869 |
| 0.3131 | 8.71 | 26000 | 0.1534 | 0.1850 |
| 0.3306 | 8.88 | 26500 | 0.1482 | 0.1844 |
| 0.2842 | 9.05 | 27000 | 0.1490 | 0.1854 |
| 0.2879 | 9.22 | 27500 | 0.1463 | 0.1799 |
| 0.27 | 9.38 | 28000 | 0.1454 | 0.1798 |
| 0.2874 | 9.55 | 28500 | 0.1504 | 0.1787 |
| 0.2757 | 9.72 | 29000 | 0.1512 | 0.1784 |
| 0.3017 | 9.89 | 29500 | 0.1484 | 0.1800 |
### Framework versions
- Transformers 4.17.0.dev0
- Pytorch 1.10.2+cu102
- Datasets 1.18.3
- Tokenizers 0.11.0
## Evaluation results
Evaluation was done with the [Common Voice 7.0 Finnish test split](https://huggingface.co/datasets/mozilla-foundation/common_voice_7_0).
To evaluate this model, run the `eval.py` script in this repository:
```bash
python3 eval.py --model_id aapot/wav2vec2-xlsr-300m-finnish-lm --dataset mozilla-foundation/common_voice_7_0 --config fi --split test
```
This model (the third row of the table) achieves the following WER (Word Error Rate) and CER (Character Error Rate) results compared to our other models:
| | WER (with LM) | WER (without LM) | CER (with LM) | CER (without LM) |
|-----------------------------------------|---------------|------------------|---------------|------------------|
|aapot/wav2vec2-xlsr-1b-finnish-lm-v2 |**4.09** |**9.73** |**0.88** |**1.65** |
|aapot/wav2vec2-xlsr-1b-finnish-lm |5.65 |13.11 |1.20 |2.23 |
|aapot/wav2vec2-xlsr-300m-finnish-lm |8.16 |17.92 |1.97 |3.36 |
## Team Members
- Aapo Tanskanen, [Hugging Face profile](https://huggingface.co/aapot), [LinkedIn profile](https://www.linkedin.com/in/aapotanskanen/)
- Rasmus Toivanen, [Hugging Face profile](https://huggingface.co/RASMUS), [LinkedIn profile](https://www.linkedin.com/in/rasmustoivanen/)
Feel free to contact us for more details 🤗 |
Subsets and Splits