Update README.md

README.md

@@ -10,8 +10,11 @@ tags:
 - xlsr-fine-tuning-week
 datasets:
 - common_voice
+- ovm
+- pscr
+- vystadial2016
 model-index:
-- name: Czech comodoro Wav2Vec2 XLSR 300M CV8
+- name: Czech comodoro Wav2Vec2 XLSR 300M 250h data
   results:
   - task: 
       name: Automatic Speech Recognition 
@@ -23,25 +26,29 @@ model-index:
     metrics:
        - name: Test WER
          type: wer
-         value: 10.3
+         value: 10.0
        - name: Test CER
          type: cer 
          value: 2.6
 ---
-<!-- 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-300m-cs-cv8
+# Czech wav2vec2-xls-r-300m-cs-250
 
-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 8.0 dataset.
-It achieves the following results on the evaluation set while training:
-- Loss: 0.2327
-- Wer: 0.1608
-- Cer: 0.0376
+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 8.0 dataset as well as other datasets listed below.
+
+It achieves the following results on the evaluation set:
+- eval_loss: 0.1304
+- eval_wer: 0.1517
+- eval_cer: 0.0326
+- eval_runtime: 358.9895
+- eval_samples_per_second: 20.243
+- eval_steps_per_second: 2.532
+- epoch: 3.13
+- step: 31200
 
 The `eval.py` script results using a LM are:
-WER: 0.10281503199350225
-CER: 0.02622802241689026
+WER: 0.10053685691079459
+CER: 0.025859623842234124
 
 ## Model description
 
@@ -59,8 +66,8 @@ from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
 
 test_dataset = load_dataset("mozilla-foundation/common_voice_8_0", "cs", split="test[:2%]")
 
-processor = Wav2Vec2Processor.from_pretrained("comodoro/wav2vec2-xls-r-300m-cs-cv8")
-model = Wav2Vec2ForCTC.from_pretrained("comodoro/wav2vec2-xls-r-300m-cs-cv8")
+processor = Wav2Vec2Processor.from_pretrained("comodoro/wav2vec2-xls-r-300m-cs-250")
+model = Wav2Vec2ForCTC.from_pretrained("comodoro/wav2vec2-xls-r-300m-cs-250")
 
 resampler = torchaudio.transforms.Resample(48_000, 16_000)
 
@@ -87,83 +94,35 @@ print("Reference:", test_dataset[:2]["sentence"])
 
 The model can be evaluated using the attached `eval.py` script:
 ```
-python eval.py --model_id comodoro/wav2vec2-xls-r-300m-cs-cv8 --dataset mozilla-foundation/common-voice_8_0 --split test --config cs
+python eval.py --model_id comodoro/wav2vec2-xls-r-300m-cs-250 --dataset mozilla-foundation/common-voice_8_0 --split test --config cs
 ```
 
 ## Training and evaluation data
 
-The Common Voice 8.0 `train` and `validation` datasets were used for training
+The Common Voice 8.0 `train` and `validation` datasets were used for training, as well as the following datasets:
 
-## Training procedure
+- Šmídl, Luboš and Pražák, Aleš, 2013, OVM – Otázky Václava Moravce, LINDAT/CLARIAH-CZ digital library at the Institute of Formal and Applied Linguistics (ÚFAL), Faculty of Mathematics and Physics, Charles University, http://hdl.handle.net/11858/00-097C-0000-000D-EC98-3.
 
-### Training hyperparameters
+- Pražák, Aleš and Šmídl, Luboš, 2012, Czech Parliament Meetings, LINDAT/CLARIAH-CZ digital library at the Institute of Formal and Applied Linguistics (ÚFAL), Faculty of Mathematics and Physics, Charles University, http://hdl.handle.net/11858/00-097C-0000-0005-CF9C-4.
 
-The following hyperparameters were used during first stage of training:
+- Plátek, Ondřej; Dušek, Ondřej and Jurčíček, Filip, 2016, Vystadial 2016 – Czech data, LINDAT/CLARIAH-CZ digital library at the Institute of Formal and Applied Linguistics (ÚFAL), Faculty of Mathematics and Physics, Charles University, http://hdl.handle.net/11234/1-1740.
 
-- learning_rate: 7e-05
-- train_batch_size: 32
-- eval_batch_size: 8
-- seed: 42
-- gradient_accumulation_steps: 20
-- total_train_batch_size: 640
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 500
-- num_epochs: 150
-- mixed_precision_training: Native AMP
-
-The following hyperparameters were used during second stage of training:
+### Training hyperparameters
 
-- learning_rate: 0.001
-- train_batch_size: 32
+The following hyperparameters were used during training:
+- learning_rate: 1e-05
+- train_batch_size: 16
 - eval_batch_size: 8
 - seed: 42
-- gradient_accumulation_steps: 20
-- total_train_batch_size: 640
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 500
+- lr_scheduler_warmup_steps: 600
 - num_epochs: 50
 - mixed_precision_training: Native AMP
 
-### Training results
-
-| Training Loss | Epoch  | Step | Validation Loss | Wer    | Cer    |
-|:-------------:|:------:|:----:|:---------------:|:------:|:------:|
-| 7.2926        | 8.06   | 250  | 3.8497          | 1.0    | 1.0    |
-| 3.417         | 16.13  | 500  | 3.2852          | 1.0    | 0.9857 |
-| 2.0264        | 24.19  | 750  | 0.7099          | 0.7342 | 0.1768 |
-| 0.4018        | 32.25  | 1000 | 0.6188          | 0.6415 | 0.1551 |
-| 0.2444        | 40.32  | 1250 | 0.6632          | 0.6362 | 0.1600 |
-| 0.1882        | 48.38  | 1500 | 0.6070          | 0.5783 | 0.1388 |
-| 0.153         | 56.44  | 1750 | 0.6425          | 0.5720 | 0.1377 |
-| 0.1214        | 64.51  | 2000 | 0.6363          | 0.5546 | 0.1337 |
-| 0.1011        | 72.57  | 2250 | 0.6310          | 0.5222 | 0.1224 |
-| 0.0879        | 80.63  | 2500 | 0.6353          | 0.5258 | 0.1253 |
-| 0.0782        | 88.7   | 2750 | 0.6078          | 0.4904 | 0.1127 |
-| 0.0709        | 96.76  | 3000 | 0.6465          | 0.4960 | 0.1154 |
-| 0.0661        | 104.82 | 3250 | 0.6622          | 0.4945 | 0.1166 |
-| 0.0616        | 112.89 | 3500 | 0.6440          | 0.4786 | 0.1104 |
-| 0.0579        | 120.95 | 3750 | 0.6815          | 0.4887 | 0.1144 |
-| 0.0549        | 129.03 | 4000 | 0.6603          | 0.4780 | 0.1105 |
-| 0.0527        | 137.09 | 4250 | 0.6652          | 0.4749 | 0.1090 |
-| 0.0506        | 145.16 | 4500 | 0.6958          | 0.4846 | 0.1133 |
-
-Further fine-tuning with slightly different architecture and higher learning rate:
-
-| Training Loss | Epoch | Step | Validation Loss | Wer    | Cer    |
-|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
-| 0.576         | 8.06  | 250  | 0.2411          | 0.2340 | 0.0502 |
-| 0.2564        | 16.13 | 500  | 0.2305          | 0.2097 | 0.0492 |
-| 0.2018        | 24.19 | 750  | 0.2371          | 0.2059 | 0.0494 |
-| 0.1549        | 32.25 | 1000 | 0.2298          | 0.1844 | 0.0435 |
-| 0.1224        | 40.32 | 1250 | 0.2288          | 0.1725 | 0.0407 |
-| 0.1004        | 48.38 | 1500 | 0.2327          | 0.1608 | 0.0376 |
-
-
 ### Framework versions
 
-- Transformers 4.16.0.dev0
+- Transformers 4.16.2
 - Pytorch 1.10.1+cu102
-- Datasets 1.17.1.dev0
+- Datasets 1.18.3
 - Tokenizers 0.11.0