Downloads · 30 days
12
2% of all-time downloads
Whisper_Small_Odia_CV_11.0_5k_steps
Ranjit/Whisper_Small_Odia_CV_11.0_5k_steps
Whisper_Small_Odia_CV_11.0_5k_steps is a automatic speech recognition model from Ranjit. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as afl-3.0.
should probably proofread and complete it, then remove this comment. --
All-time downloads
558
Public
Repo size
1.9 GB
Likes
1
Public
Repo makeup
Click a slice to open those files.
.bin967 MB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- License
- afl-3.0
- Model type
- whisper
- Access
- Public
- Created
- Mar 29, 2023
- Updated
- May 31, 2023
- SHA
- b741ea6c
Spaces
Model card
From the Hugging Face model README
<!-- 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_Small_Odia_CV_11.0_5k_steps
This model is a fine-tuned version of Ranjit/Whisper_Small_Odia_10k_steps on the mozilla-foundation/common_voice_11_0 or dataset. It achieves the following results on the evaluation set:
- Loss: 0.4827
- Wer: 23.4979
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.98) and epsilon=1e-06
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 200
- training_steps: 5000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0018 | 50.0 | 1000 | 0.3315 | 24.0903 |
| 0.0 | 100.0 | 2000 | 0.4098 | 23.7236 |
| 0.0 | 150.0 | 3000 | 0.4827 | 23.4979 |
| 0.0 | 200.0 | 4000 | 0.4914 | 23.8928 |
| 0.0 | 250.0 | 5000 | 0.4953 | 23.7800 |
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- whisper
- License
- afl-3.0
- Languages
- or
- Created
- Mar 29, 2023
- Updated
- May 31, 2023