End of training

README.md

@@ -18,8 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1142
-- Wer: 3.7226
+- Loss: 0.1362
+- Wer: 4.3516
 
 ## Model description
 
@@ -38,32 +38,34 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 1e-05
-- train_batch_size: 16
-- eval_batch_size: 8
+- learning_rate: 5e-06
+- train_batch_size: 8
+- eval_batch_size: 16
 - seed: 42
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 16
 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: linear
-- training_steps: 2000
+- lr_scheduler_warmup_steps: 50
+- training_steps: 3000
 - mixed_precision_training: Native AMP
 
 ### Training results
 
-| Training Loss | Epoch  | Step | Validation Loss | Wer    |
-|:-------------:|:------:|:----:|:---------------:|:------:|
-| 0.0667        | 0.3534 | 100  | 0.1267          | 4.8230 |
-| 0.0693        | 0.7067 | 200  | 0.1188          | 4.1087 |
-| 0.0614        | 1.0601 | 300  | 0.1133          | 3.7907 |
-| 0.0274        | 1.4134 | 400  | 0.1128          | 3.7503 |
-| 0.0365        | 1.7668 | 500  | 0.1113          | 3.8513 |
-| 0.014         | 2.1201 | 600  | 0.1106          | 3.7201 |
-| 0.0165        | 2.4735 | 700  | 0.1116          | 3.7352 |
-| 0.0137        | 2.8269 | 800  | 0.1096          | 3.6999 |
-| 0.007         | 3.1802 | 900  | 0.1114          | 3.6216 |
-| 0.007         | 3.5336 | 1000 | 0.1130          | 3.6317 |
-| 0.0077        | 3.8869 | 1100 | 0.1128          | 3.6645 |
-| 0.0056        | 4.2403 | 1200 | 0.1138          | 3.7302 |
-| 0.005         | 4.5936 | 1300 | 0.1142          | 3.7226 |
+| Training Loss | Epoch | Step | Validation Loss | Wer    |
+|:-------------:|:-----:|:----:|:---------------:|:------:|
+| 0.0146        | 0.4   | 100  | 0.1223          | 4.8708 |
+| 0.0159        | 0.8   | 200  | 0.1248          | 4.8306 |
+| 0.0104        | 1.2   | 300  | 0.1251          | 4.3803 |
+| 0.01          | 1.6   | 400  | 0.1259          | 4.3975 |
+| 0.0092        | 2.0   | 500  | 0.1263          | 4.4749 |
+| 0.0055        | 2.4   | 600  | 0.1301          | 4.3344 |
+| 0.0062        | 2.8   | 700  | 0.1303          | 4.4061 |
+| 0.0039        | 3.2   | 800  | 0.1324          | 4.5294 |
+| 0.0045        | 3.6   | 900  | 0.1337          | 4.3889 |
+| 0.0036        | 4.0   | 1000 | 0.1350          | 4.2626 |
+| 0.0033        | 4.4   | 1100 | 0.1358          | 4.3344 |
+| 0.0033        | 4.8   | 1200 | 0.1362          | 4.3516 |
 
 
 ### Framework versions

runs/Nov19_07-26-25_53be0e62afda/events.out.tfevents.1732001191.53be0e62afda.1551.2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9bb4f281604213476878de5057f8703193b4fe918e7e9e3ac54d255936f13e1b
-size 13202
+oid sha256:7757f3120fc815e9179ae6225a4d943ec4518c26e14319abafa423f4e847bdcb
+size 15036