Model save

README.md

@@ -4,24 +4,11 @@ license: apache-2.0
 base_model: razhan/whisper-base-hawrami
 tags:
 - generated_from_trainer
-datasets:
-- razhan/DOLMA-speech
 metrics:
 - wer
 model-index:
 - name: whisper-base-hawrami-transcription
-  results:
-  - task:
-      name: Automatic Speech Recognition
-      type: automatic-speech-recognition
-    dataset:
-      name: razhan/DOLMA-speech hawrami
-      type: razhan/DOLMA-speech
-      args: hawrami
-    metrics:
-    - name: Wer
-      type: wer
-      value: 2.3790849673202614
+  results: []
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -29,10 +16,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 # whisper-base-hawrami-transcription
 
-This model is a fine-tuned version of [razhan/whisper-base-hawrami](https://huggingface.co/razhan/whisper-base-hawrami) on the razhan/DOLMA-speech hawrami dataset.
+This model is a fine-tuned version of [razhan/whisper-base-hawrami](https://huggingface.co/razhan/whisper-base-hawrami) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 3.5885
-- Wer: 2.3791
+- Loss: 0.2612
+- Wer: 0.4013
+- Cer: 0.0856
 
 ## Model description
 
@@ -52,7 +40,7 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 1e-05
-- train_batch_size: 256
+- train_batch_size: 128
 - eval_batch_size: 128
 - seed: 42
 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
@@ -63,12 +51,12 @@ The following hyperparameters were used during training:
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | Wer    |
-|:-------------:|:-----:|:----:|:---------------:|:------:|
-| No log        | 1.0   | 1    | 3.5885          | 2.3791 |
-| No log        | 2.0   | 2    | 3.5885          | 2.3791 |
-| No log        | 3.0   | 3    | 3.5885          | 2.3791 |
-| No log        | 4.0   | 4    | 3.5885          | 2.3791 |
+| Training Loss | Epoch | Step | Validation Loss | Wer    | Cer    |
+|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
+| 0.6798        | 1.0   | 80   | 0.5513          | 0.6855 | 0.1788 |
+| 0.3095        | 2.0   | 160  | 0.2984          | 0.4486 | 0.0972 |
+| 0.2673        | 3.0   | 240  | 0.2676          | 0.4143 | 0.0882 |
+| 0.2428        | 4.0   | 320  | 0.2612          | 0.4013 | 0.0856 |
 
 
 ### Framework versions

generation_config.json

@@ -239,7 +239,7 @@
     50361,
     50362
   ],
-  "task": "translate",
+  "task": "transcribe",
   "task_to_id": {
     "transcribe": 50359,
     "translate": 50358