FCameCode
/

whisper-tiny-it-8

Automatic Speech Recognition

hf-asr-leaderboard

Generated from Trainer

Inference Endpoints

Model card Files Files and versions Metrics Training metrics Community

FCameCode commited on Dec 9, 2022

Commit

1f03fb1

•

1 Parent(s): db70a31

Update README.md

Files changed (1) hide show

README.md +4 -4

README.md CHANGED Viewed

@@ -29,8 +29,8 @@ model-index:
 # Whisper Tiny it 8
 This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the Common Voice 11.0 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.999731
-- Wer: 58.361065
 ## Model description
@@ -62,12 +62,12 @@ 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: 4000
 - mixed_precision_training: Native AMP
 ### Training results
 | Training Loss | Epoch | Step | Validation Loss | Wer     |
 |:-------------:|:-----:|:----:|:---------------:|:-------:|
-| 0.444900      | 3.82  | 4000 | 0.999731        |58.361065|
 ### Framework versions
 - Transformers 4.26.0.dev0
 - Pytorch 1.12.1+cu113

 # Whisper Tiny it 8
 This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the Common Voice 11.0 dataset.
 It achieves the following results on the evaluation set:
+- Loss: 1.011502
+- Wer: 56.905158
 ## Model description
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 500
+- training_steps: 3000
 - mixed_precision_training: Native AMP
 ### Training results
 | Training Loss | Epoch | Step | Validation Loss | Wer     |
 |:-------------:|:-----:|:----:|:---------------:|:-------:|
+| 0.525800      | 3.82  | 3000 | 1.011502        |56.905158|
 ### Framework versions
 - Transformers 4.26.0.dev0
 - Pytorch 1.12.1+cu113