End of training

README.md

@@ -21,7 +21,7 @@ model-index:
     metrics:
     - name: Wer
       type: wer
-      value: 1.0114942528735633
+      value: 0.6237000547345375
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -31,8 +31,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [](https://huggingface.co/) on the common_voice_17_0 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 2.3875
-- Wer: 1.0115
+- Loss: 0.2006
+- Wer: 0.6237
 
 ## Model description
 
@@ -51,7 +51,7 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 0.0005
+- learning_rate: 3e-05
 - train_batch_size: 16
 - eval_batch_size: 8
 - seed: 42
@@ -67,15 +67,15 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch   | Step  | Validation Loss | Wer    |
 |:-------------:|:-------:|:-----:|:---------------:|:------:|
-| 0.5699        | 2.1942  | 2000  | 0.5029          | 0.9499 |
-| 0.4929        | 4.3884  | 4000  | 0.4351          | 0.9349 |
-| 0.4568        | 6.5826  | 6000  | 0.3928          | 0.9080 |
-| 0.4274        | 8.7767  | 8000  | 0.3524          | 0.8848 |
-| 0.389         | 10.9709 | 10000 | 0.3127          | 0.8331 |
-| 0.6396        | 13.1651 | 12000 | 0.5697          | 0.9324 |
-| 2.5933        | 15.3593 | 14000 | 2.3935          | 1.0109 |
-| 2.5858        | 17.5535 | 16000 | 2.3892          | 1.0120 |
-| 2.5724        | 19.7477 | 18000 | 2.3875          | 1.0115 |
+| 0.3837        | 2.1942  | 2000  | 0.3241          | 0.8196 |
+| 0.3176        | 4.3884  | 4000  | 0.2855          | 0.7830 |
+| 0.2886        | 6.5826  | 6000  | 0.2620          | 0.7499 |
+| 0.2659        | 8.7767  | 8000  | 0.2431          | 0.7154 |
+| 0.2464        | 10.9709 | 10000 | 0.2285          | 0.6877 |
+| 0.2252        | 13.1651 | 12000 | 0.2163          | 0.6552 |
+| 0.2132        | 15.3593 | 14000 | 0.2087          | 0.6461 |
+| 0.2083        | 17.5535 | 16000 | 0.2032          | 0.6286 |
+| 0.2034        | 19.7477 | 18000 | 0.2006          | 0.6237 |
 
 
 ### Framework versions

generation_config.json

@@ -1,7 +1,6 @@
 {
   "_from_model_config": true,
   "bos_token_id": 50256,
-  "decoder_start_token_id": 50256,
   "eos_token_id": 50256,
   "transformers_version": "4.45.2"
 }