update model card README.md

README.md

@@ -0,0 +1,80 @@
+---
+license: apache-2.0
+tags:
+- generated_from_trainer
+datasets:
+- voxpopuli
+metrics:
+- wer
+model-index:
+- name: openai/whisper-small
+  results:
+  - task:
+      name: Automatic Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: voxpopuli
+      type: voxpopuli
+      config: hr
+      split: test
+      args: hr
+    metrics:
+    - name: Wer
+      type: wer
+      value: 25.43475821833277
+---
+
+<!-- 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. -->
+
+# openai/whisper-small
+
+This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the voxpopuli dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.6405
+- Wer: 25.4348
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 1e-05
+- train_batch_size: 64
+- eval_batch_size: 32
+- seed: 42
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- lr_scheduler_warmup_steps: 500
+- training_steps: 5000
+- mixed_precision_training: Native AMP
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss | Wer     |
+|:-------------:|:------:|:----:|:---------------:|:-------:|
+| 0.0028        | 24.01  | 1000 | 0.5303          | 26.1752 |
+| 0.0006        | 49.01  | 2000 | 0.5849          | 25.4123 |
+| 0.0003        | 74.01  | 3000 | 0.6141          | 25.6311 |
+| 0.0002        | 99.01  | 4000 | 0.6325          | 25.3731 |
+| 0.0002        | 124.01 | 5000 | 0.6405          | 25.4348 |
+
+
+### Framework versions
+
+- Transformers 4.26.0.dev0
+- Pytorch 1.13.0+cu117
+- Datasets 2.7.1.dev0
+- Tokenizers 0.13.2

run.log

@@ -363,3 +363,77 @@ xpu_backend=None,
 {'loss': 0.0003, 'learning_rate': 3.6133333333333336e-06, 'epoch': 84.0}
 {'loss': 0.0003, 'learning_rate': 3.5577777777777785e-06, 'epoch': 84.01}
 12/15/2022 06:41:03 - WARNING - datasets.download.streaming_download_manager - Got disconnected from remote data host. Retrying in 5sec [1/20]
+{'loss': 0.0003, 'learning_rate': 3.5022222222222225e-06, 'epoch': 85.0}
+{'loss': 0.0002, 'learning_rate': 3.446666666666667e-06, 'epoch': 86.0}
+{'loss': 0.0003, 'learning_rate': 3.391111111111111e-06, 'epoch': 86.01}
+{'loss': 0.0002, 'learning_rate': 3.335555555555556e-06, 'epoch': 87.0}
+{'loss': 0.0003, 'learning_rate': 3.2800000000000004e-06, 'epoch': 88.0}
+{'loss': 0.0002, 'learning_rate': 3.2244444444444444e-06, 'epoch': 88.01}
+{'loss': 0.0002, 'learning_rate': 3.1688888888888893e-06, 'epoch': 89.0}
+{'loss': 0.0002, 'learning_rate': 3.1133333333333337e-06, 'epoch': 89.01}
+{'loss': 0.0002, 'learning_rate': 3.0577777777777778e-06, 'epoch': 90.0}
+{'loss': 0.0002, 'learning_rate': 3.0022222222222227e-06, 'epoch': 91.0}
+{'loss': 0.0002, 'learning_rate': 2.946666666666667e-06, 'epoch': 91.01}
+{'loss': 0.0002, 'learning_rate': 2.891111111111111e-06, 'epoch': 92.0}
+{'loss': 0.0002, 'learning_rate': 2.835555555555556e-06, 'epoch': 93.0}
+{'loss': 0.0002, 'learning_rate': 2.7800000000000005e-06, 'epoch': 93.01}
+{'loss': 0.0002, 'learning_rate': 2.7244444444444445e-06, 'epoch': 94.0}
+{'loss': 0.0002, 'learning_rate': 2.6688888888888894e-06, 'epoch': 94.01}
+{'loss': 0.0002, 'learning_rate': 2.6133333333333334e-06, 'epoch': 95.0}
+{'loss': 0.0002, 'learning_rate': 2.557777777777778e-06, 'epoch': 96.0}
+{'loss': 0.0002, 'learning_rate': 2.5022222222222224e-06, 'epoch': 96.01}
+{'loss': 0.0002, 'learning_rate': 2.446666666666667e-06, 'epoch': 97.0}
+{'loss': 0.0002, 'learning_rate': 2.3911111111111113e-06, 'epoch': 98.0}
+{'loss': 0.0002, 'learning_rate': 2.3355555555555557e-06, 'epoch': 98.01}
+{'loss': 0.0002, 'learning_rate': 2.28e-06, 'epoch': 99.0}
+{'loss': 0.0002, 'learning_rate': 2.2244444444444447e-06, 'epoch': 99.01}
+{'eval_loss': 0.6324931383132935, 'eval_wer': 25.37305060024683, 'eval_runtime': 650.4597, 'eval_samples_per_second': 1.024, 'eval_steps_per_second': 0.032, 'epoch': 99.01}
+{'loss': 0.0002, 'learning_rate': 2.168888888888889e-06, 'epoch': 100.0}
+{'loss': 0.0002, 'learning_rate': 2.1133333333333336e-06, 'epoch': 101.0}
+{'loss': 0.0002, 'learning_rate': 2.057777777777778e-06, 'epoch': 101.01}
+{'loss': 0.0002, 'learning_rate': 2.0022222222222225e-06, 'epoch': 102.0}
+{'loss': 0.0002, 'learning_rate': 1.9466666666666665e-06, 'epoch': 103.0}
+{'loss': 0.0002, 'learning_rate': 1.8911111111111114e-06, 'epoch': 103.01}
+{'loss': 0.0002, 'learning_rate': 1.8355555555555557e-06, 'epoch': 104.0}
+{'loss': 0.0002, 'learning_rate': 1.7800000000000001e-06, 'epoch': 104.01}
+{'loss': 0.0002, 'learning_rate': 1.7244444444444448e-06, 'epoch': 105.0}
+{'loss': 0.0002, 'learning_rate': 1.668888888888889e-06, 'epoch': 106.0}
+{'loss': 0.0002, 'learning_rate': 1.6133333333333335e-06, 'epoch': 106.01}
+{'loss': 0.0002, 'learning_rate': 1.5577777777777777e-06, 'epoch': 107.0}
+{'loss': 0.0002, 'learning_rate': 1.5022222222222224e-06, 'epoch': 108.0}
+{'loss': 0.0002, 'learning_rate': 1.4466666666666669e-06, 'epoch': 108.01}
+{'loss': 0.0002, 'learning_rate': 1.3911111111111111e-06, 'epoch': 109.0}
+{'loss': 0.0002, 'learning_rate': 1.3355555555555558e-06, 'epoch': 109.01}
+{'loss': 0.0002, 'learning_rate': 1.28e-06, 'epoch': 110.0}
+{'loss': 0.0002, 'learning_rate': 1.2244444444444445e-06, 'epoch': 111.0}
+{'loss': 0.0002, 'learning_rate': 1.168888888888889e-06, 'epoch': 111.01}
+{'loss': 0.0002, 'learning_rate': 1.1133333333333334e-06, 'epoch': 112.0}
+{'loss': 0.0002, 'learning_rate': 1.0577777777777779e-06, 'epoch': 113.0}
+{'loss': 0.0002, 'learning_rate': 1.0022222222222223e-06, 'epoch': 113.01}
+{'loss': 0.0002, 'learning_rate': 9.466666666666667e-07, 'epoch': 114.0}
+{'loss': 0.0002, 'learning_rate': 8.911111111111112e-07, 'epoch': 114.01}
+{'loss': 0.0002, 'learning_rate': 8.355555555555556e-07, 'epoch': 115.0}
+{'loss': 0.0002, 'learning_rate': 7.8e-07, 'epoch': 116.0}
+{'loss': 0.0002, 'learning_rate': 7.244444444444446e-07, 'epoch': 116.01}
+{'loss': 0.0002, 'learning_rate': 6.68888888888889e-07, 'epoch': 117.0}
+{'loss': 0.0002, 'learning_rate': 6.133333333333333e-07, 'epoch': 118.0}
+{'loss': 0.0002, 'learning_rate': 5.577777777777779e-07, 'epoch': 118.01}
+{'loss': 0.0002, 'learning_rate': 5.022222222222222e-07, 'epoch': 119.0}
+{'loss': 0.0002, 'learning_rate': 4.466666666666667e-07, 'epoch': 119.01}
+{'loss': 0.0002, 'learning_rate': 3.9111111111111115e-07, 'epoch': 120.0}
+{'loss': 0.0002, 'learning_rate': 3.3555555555555556e-07, 'epoch': 121.0}
+{'loss': 0.0002, 'learning_rate': 2.8e-07, 'epoch': 121.01}
+{'loss': 0.0002, 'learning_rate': 2.2444444444444445e-07, 'epoch': 122.0}
+{'loss': 0.0002, 'learning_rate': 1.6888888888888888e-07, 'epoch': 123.0}
+{'loss': 0.0002, 'learning_rate': 1.1333333333333336e-07, 'epoch': 123.01}
+{'loss': 0.0002, 'learning_rate': 5.777777777777778e-08, 'epoch': 124.0}
+{'loss': 0.0002, 'learning_rate': 2.2222222222222225e-09, 'epoch': 124.01}
+{'eval_loss': 0.6405040621757507, 'eval_wer': 25.43475821833277, 'eval_runtime': 648.4677, 'eval_samples_per_second': 1.027, 'eval_steps_per_second': 0.032, 'epoch': 124.01}
+{'train_runtime': 39423.3866, 'train_samples_per_second': 8.117, 'train_steps_per_second': 0.127, 'train_loss': 0.04578730283072218, 'epoch': 124.01}
+12/15/2022 10:11:51 - WARNING - huggingface_hub.repository - Several commits (2) will be pushed upstream.
+12/15/2022 10:11:51 - WARNING - huggingface_hub.repository - The progress bars may be unreliable.
+12/15/2022 10:11:57 - WARNING - huggingface_hub.repository - remote: Scanning LFS files for validity, may be slow...        
+remote: LFS file scan complete.        
+To https://huggingface.co/mikr/whisper-small-hr-vox
+   ddcf8d0..1dba846  main -> main
+