Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
Marathi
whisper
hf-asr-leaderboard
Generated from Trainer
Eval Results (legacy)
# Load model directly
from transformers import AutoProcessor, AutoModelForSpeechSeq2Seq
processor = AutoProcessor.from_pretrained("var2093/whisper-tiny-mr")
model = AutoModelForSpeechSeq2Seq.from_pretrained("var2093/whisper-tiny-mr")Quick Links
Whisper Tiny Mr - varun
This model is a fine-tuned version of openai/whisper-tiny on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:
- Loss: 1.1336
- Wer: 418.5997
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 5
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- training_steps: 100
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 1.5808 | 0.06 | 50 | 1.5279 | 339.9007 |
| 1.0984 | 0.13 | 100 | 1.1336 | 418.5997 |
Framework versions
- Transformers 4.28.0.dev0
- Pytorch 1.13.1+cu116
- Datasets 2.10.1
- Tokenizers 0.13.2
- Downloads last month
- 4
Evaluation results
- Wer on Common Voice 11.0test set self-reported418.600
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="var2093/whisper-tiny-mr")