End of training

README.md

@@ -24,7 +24,7 @@ model-index:
     metrics:
     - name: Wer
       type: wer
-      value: 46.300985978395836
+      value: 16.45244089773603
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -34,9 +34,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [futureProofGlitch/whisper-small](https://huggingface.co/futureProofGlitch/whisper-small) on the Gigaspeech dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3434
-- Wer Ortho: 56.8717
-- Wer: 46.3010
+- Loss: 0.3078
+- Wer Ortho: 28.4362
+- Wer: 16.4524
 
 ## Model description
 
@@ -62,19 +62,15 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: constant_with_warmup
 - lr_scheduler_warmup_steps: 50
-- training_steps: 3000
+- training_steps: 1000
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer     |
 |:-------------:|:-----:|:----:|:---------------:|:---------:|:-------:|
-| 0.2268        | 0.5   | 500  | 0.3308          | 29.8287   | 18.2124 |
-| 0.2039        | 0.99  | 1000 | 0.3082          | 28.3139   | 16.3612 |
-| 0.1071        | 1.49  | 1500 | 0.3209          | 30.5425   | 18.9117 |
-| 0.1174        | 1.98  | 2000 | 0.3140          | 51.1370   | 40.1655 |
-| 0.0555        | 2.48  | 2500 | 0.3525          | 65.5197   | 53.9069 |
-| 0.0603        | 2.98  | 3000 | 0.3434          | 56.8717   | 46.3010 |
+| 0.2267        | 0.5   | 500  | 0.3309          | 29.5720   | 18.0966 |
+| 0.2035        | 0.99  | 1000 | 0.3078          | 28.4362   | 16.4524 |
 
 
 ### Framework versions