File size: 4,720 Bytes
c45140e ae4f787 c45140e ae4f787 c45140e ae4f787 c45140e 36b5d0b c45140e ae4f787 c45140e ae4f787 c45140e 36b5d0b c45140e 36b5d0b c45140e ae4f787 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 |
---
library_name: transformers
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: whisper-tiny-fr
results: []
---
<!-- 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-tiny-fr
This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6094
- Model Preparation Time: 0.0026
- Wer: 0.3301
- Cer: 0.1774
## 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: 0.0001
- train_batch_size: 8
- 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
- training_steps: 33000
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Model Preparation Time | Wer | Cer |
|:-------------:|:------:|:-----:|:---------------:|:----------------------:|:------:|:------:|
| 0.8933 | 0.0303 | 1000 | 1.3184 | 0.0026 | 0.5726 | 0.3162 |
| 0.7455 | 0.0606 | 2000 | 1.1738 | 0.0026 | 0.5533 | 0.2930 |
| 0.735 | 0.0909 | 3000 | 1.0462 | 0.0026 | 0.6032 | 0.3290 |
| 0.6426 | 0.1212 | 4000 | 1.0629 | 0.0026 | 0.4937 | 0.2473 |
| 0.6389 | 0.1515 | 5000 | 1.0132 | 0.0026 | 0.5671 | 0.3415 |
| 0.532 | 0.1818 | 6000 | 1.0194 | 0.0026 | 0.4639 | 0.2355 |
| 0.5341 | 0.2121 | 7000 | 1.0092 | 0.0026 | 0.4619 | 0.2435 |
| 0.4773 | 0.2424 | 8000 | 0.9672 | 0.0026 | 0.4666 | 0.2822 |
| 0.4932 | 0.2727 | 9000 | 0.9778 | 0.0026 | 0.4175 | 0.2187 |
| 0.479 | 0.3030 | 10000 | 0.9639 | 0.0026 | 0.4105 | 0.2169 |
| 0.4663 | 0.3333 | 11000 | 0.9689 | 0.0026 | 0.4236 | 0.2245 |
| 0.3647 | 0.3636 | 12000 | 1.0025 | 0.0026 | 0.4326 | 0.2297 |
| 0.451 | 0.3939 | 13000 | 0.8810 | 0.0026 | 0.4648 | 0.2591 |
| 0.4522 | 0.4242 | 14000 | 0.8283 | 0.0026 | 0.3869 | 0.1965 |
| 0.5064 | 0.4545 | 15000 | 0.8165 | 0.0026 | 0.3703 | 0.1898 |
| 0.4355 | 0.4848 | 16000 | 0.7857 | 0.0026 | 0.4367 | 0.2257 |
| 0.2953 | 0.5152 | 17000 | 0.8007 | 0.0026 | 0.3650 | 0.2020 |
| 0.4345 | 0.5455 | 18000 | 0.7823 | 0.0026 | 0.4544 | 0.2381 |
| 0.4117 | 0.5758 | 19000 | 0.7648 | 0.0026 | 0.3595 | 0.1823 |
| 0.4071 | 0.6061 | 20000 | 0.7475 | 0.0026 | 0.4121 | 0.2049 |
| 0.4371 | 0.6364 | 21000 | 0.7285 | 0.0026 | 0.3509 | 0.1842 |
| 0.34 | 0.6667 | 22000 | 0.7686 | 0.0026 | 0.3566 | 0.1860 |
| 0.335 | 0.6970 | 23000 | 0.7514 | 0.0026 | 0.3595 | 0.1846 |
| 0.2946 | 0.7273 | 24000 | 0.7928 | 0.0026 | 0.3742 | 0.2006 |
| 0.3916 | 0.7576 | 25000 | 0.6843 | 0.0026 | 0.3416 | 0.1747 |
| 0.3233 | 0.7879 | 26000 | 0.6478 | 0.0026 | 0.3178 | 0.1626 |
| 0.2981 | 0.8182 | 27000 | 0.6737 | 0.0026 | 0.3274 | 0.1669 |
| 0.2945 | 0.8485 | 28000 | 0.6512 | 0.0026 | 0.3302 | 0.1643 |
| 0.2956 | 0.8788 | 29000 | 0.6867 | 0.0026 | 0.3925 | 0.1991 |
| 0.2541 | 0.9091 | 30000 | 0.6333 | 0.0026 | 0.3186 | 0.1620 |
| 0.2475 | 0.9394 | 31000 | 0.7018 | 0.0026 | 0.3343 | 0.1722 |
| 0.2951 | 0.9697 | 32000 | 0.6527 | 0.0026 | 0.3168 | 0.1632 |
| 0.2879 | 1.0 | 33000 | 0.6440 | 0.0026 | 0.3102 | 0.1560 |
### Framework versions
- Transformers 4.49.0
- Pytorch 2.6.0+cu124
- Datasets 3.4.1
- Tokenizers 0.21.1
|