End of training

README.md

@@ -23,7 +23,7 @@ model-index:
     metrics:
     - name: Wer
       type: wer
-      value: 88.42434584755404
+      value: 65.37832712052841
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -33,8 +33,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the audiofolder dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.6423
-- Wer: 88.4243
+- Loss: 1.3225
+- Wer: 65.3783
 
 ## Model description
 
@@ -53,45 +53,28 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 5e-05
+- learning_rate: 1e-05
 - train_batch_size: 16
 - eval_batch_size: 8
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 4000
-- training_steps: 250
+- lr_scheduler_warmup_steps: 500
+- training_steps: 4000
 - mixed_precision_training: Native AMP
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | Wer      |
-|:-------------:|:-----:|:----:|:---------------:|:--------:|
-| No log        | 0.4   | 10   | 4.7002          | 139.7042 |
-| No log        | 0.8   | 20   | 4.6846          | 144.2548 |
-| 4.6631        | 1.2   | 30   | 4.5813          | 142.4630 |
-| 4.6631        | 1.6   | 40   | 4.4087          | 139.1354 |
-| 4.432         | 2.0   | 50   | 4.1992          | 146.7861 |
-| 4.432         | 2.4   | 60   | 3.9636          | 168.6860 |
-| 4.432         | 2.8   | 70   | 3.7453          | 165.1024 |
-| 3.9103        | 3.2   | 80   | 3.5625          | 156.6553 |
-| 3.9103        | 3.6   | 90   | 3.3555          | 148.3788 |
-| 3.4058        | 4.0   | 100  | 3.1750          | 166.2116 |
-| 3.4058        | 4.4   | 110  | 2.9982          | 163.3390 |
-| 3.4058        | 4.8   | 120  | 2.8273          | 156.1433 |
-| 2.9337        | 5.2   | 130  | 2.6632          | 171.3879 |
-| 2.9337        | 5.6   | 140  | 2.5056          | 156.8828 |
-| 2.4999        | 6.0   | 150  | 2.3768          | 132.7645 |
-| 2.4999        | 6.4   | 160  | 2.2670          | 117.0933 |
-| 2.4999        | 6.8   | 170  | 2.1677          | 109.2719 |
-| 2.1661        | 7.2   | 180  | 2.0791          | 111.1775 |
-| 2.1661        | 7.6   | 190  | 1.9999          | 99.6587  |
-| 1.9126        | 8.0   | 200  | 1.9264          | 100.7110 |
-| 1.9126        | 8.4   | 210  | 1.8584          | 99.9431  |
-| 1.9126        | 8.8   | 220  | 1.7963          | 102.2184 |
-| 1.7087        | 9.2   | 230  | 1.7424          | 100.3129 |
-| 1.7087        | 9.6   | 240  | 1.6920          | 102.0478 |
-| 1.5403        | 10.0  | 250  | 1.6423          | 88.4243  |
+| Training Loss | Epoch   | Step | Validation Loss | Wer     |
+|:-------------:|:-------:|:----:|:---------------:|:-------:|
+| 0.6326        | 7.5758  | 500  | 0.8675          | 71.1922 |
+| 0.1649        | 15.1515 | 1000 | 0.8976          | 65.7903 |
+| 0.0284        | 22.7273 | 1500 | 1.0661          | 66.8563 |
+| 0.0067        | 30.3030 | 2000 | 1.1925          | 66.7517 |
+| 0.0034        | 37.8788 | 2500 | 1.2540          | 68.6221 |
+| 0.0022        | 45.4545 | 3000 | 1.2904          | 67.0721 |
+| 0.0018        | 53.0303 | 3500 | 1.3139          | 67.5038 |
+| 0.0017        | 60.6061 | 4000 | 1.3225          | 65.3783 |
 
 
 ### Framework versions