End of training

README.md

@@ -20,8 +20,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the AI4Bharat-svarah dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3790
-- Wer: 22.8606
+- Loss: 0.1469
+- Wer: 23.6355
 
 ## Model description
 
@@ -49,7 +49,7 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 500
-- training_steps: 1500
+- training_steps: 2000
 - mixed_precision_training: Native AMP
 
 ### Training results
@@ -57,6 +57,7 @@ The following hyperparameters were used during training:
 | Training Loss | Epoch  | Step | Validation Loss | Wer     |
 |:-------------:|:------:|:----:|:---------------:|:-------:|
 | 0.1893        | 2.6702 | 1000 | 0.3790          | 22.8606 |
+| 0.0709        | 5.3458 | 2000 | 0.1469          | 23.6355 |
 
 
 ### Framework versions

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1e41893893f3673b22ccb09355e1a193f5efdb47615328ccfada2872eafa32e1
+oid sha256:ba45c5495a8744a7fb7e626076d59c5daa9d9e9799de00432b30106c9b70f39d
 size 290403936