End of training

README.md

@@ -6,37 +6,34 @@ base_model: openai/whisper-medium
 tags:
 - generated_from_trainer
 datasets:
-- parambharat/tamil_asr_corpus
+- Prajwal-143/ASR-Tamil-cleaned
 metrics:
 - wer
 model-index:
-- name: Whisper medium english - Logii33-Tamil
+- name: Whisper medium ta - Log-Tamil
   results:
   - task:
       name: Automatic Speech Recognition
       type: automatic-speech-recognition
     dataset:
       name: ' asr corpus'
-      type: parambharat/tamil_asr_corpus
-      config: default
-      split: test
-      args: default
+      type: Prajwal-143/ASR-Tamil-cleaned
     metrics:
     - name: Wer
       type: wer
-      value: 16.834654741982387
+      value: 10.598921515883243
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
 
-# Whisper medium english - Logii33-Tamil
+# Whisper medium ta - Log-Tamil
 
 This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the  asr corpus dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2995
-- Wer Ortho: 54.6978
-- Wer: 16.8347
+- Loss: 0.1653
+- Wer Ortho: 37.0213
+- Wer: 10.5989
 
 ## Model description
 
@@ -69,7 +66,7 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss | Wer Ortho | Wer     |
 |:-------------:|:------:|:----:|:---------------:|:---------:|:-------:|
-| 0.236         | 0.0088 | 500  | 0.2995          | 54.6978   | 16.8347 |
+| 0.1591        | 0.0143 | 500  | 0.1653          | 37.0213   | 10.5989 |
 
 
 ### Framework versions