End of training

README.md

@@ -3,40 +3,40 @@ library_name: transformers
 language:
 - ur
 license: apache-2.0
-base_model: GogetaBlueMUI/whisper-medium-ur
+base_model: openai/whisper-medium
 tags:
 - generated_from_trainer
 datasets:
-- fsicoli/common_voice_19_0
+- vfsicoli/common_voice_19_0
 metrics:
 - wer
 model-index:
-- name: Whisper V3 Medium Ur - Muhammad Abdullah on Common Voice 19
+- name: Whisper Medium Ur - Muhammad Abdullah on Common Voice 19
   results:
   - task:
       name: Automatic Speech Recognition
       type: automatic-speech-recognition
     dataset:
       name: Common Voice 19.0
-      type: fsicoli/common_voice_19_0
+      type: vfsicoli/common_voice_19_0
       config: ur
       split: test
       args: 'config: ur, split: test'
     metrics:
     - name: Wer
       type: wer
-      value: 28.60428249268761
+      value: 30.017639047044902
 ---
 
 <!-- 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 V3 Medium Ur - Muhammad Abdullah on Common Voice 19
+# Whisper Medium Ur - Muhammad Abdullah on Common Voice 19
 
-This model is a fine-tuned version of [GogetaBlueMUI/whisper-medium-ur](https://huggingface.co/GogetaBlueMUI/whisper-medium-ur) on the Common Voice 19.0 dataset.
+This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the Common Voice 19.0 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4501
-- Wer: 28.6043
+- Loss: 0.4195
+- Wer: 30.0176
 
 ## Model description
 
@@ -55,27 +55,27 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 1e-05
+- learning_rate: 1e-06
 - train_batch_size: 16
 - eval_batch_size: 8
 - seed: 42
 - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 60
-- training_steps: 500
+- lr_scheduler_warmup_steps: 80
+- training_steps: 600
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss | Wer     |
 |:-------------:|:------:|:----:|:---------------:|:-------:|
-| 0.1164        | 0.5568 | 250  | 0.4771          | 29.1714 |
-| 0.0553        | 1.1136 | 500  | 0.4501          | 28.6043 |
+| 0.3018        | 0.6682 | 300  | 0.4236          | 30.1561 |
+| 0.2544        | 1.3363 | 600  | 0.4195          | 30.0176 |
 
 
 ### Framework versions
 
-- Transformers 4.48.0
+- Transformers 4.48.1
 - Pytorch 2.5.1+cu121
 - Datasets 3.2.0
 - Tokenizers 0.21.0

generation_config.json

@@ -246,5 +246,5 @@
     "transcribe": 50359,
     "translate": 50358
   },
-  "transformers_version": "4.48.0"
+  "transformers_version": "4.48.1"
 }