File size: 4,381 Bytes
49d8476 0e62997 49d8476 0e62997 49d8476 0e62997 49d8476 0e62997 49d8476 0e62997 49d8476 0e62997 49d8476 0e62997 49d8476 |
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 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 |
---
language:
- eu
license: apache-2.0
base_model: openai/whisper-large-v3
tags:
- whisper-event
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_16_1
metrics:
- wer
model-index:
- name: Whisper Large-V3 Basque
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: mozilla-foundation/common_voice_16_1 eu
type: mozilla-foundation/common_voice_16_1
config: eu
split: test
args: eu
metrics:
- name: Wer
type: wer
value: 6.887994372362044
---
<!-- 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 Large-V3 Basque
This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the mozilla-foundation/common_voice_16_1 eu dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3688
- Wer: 6.8880
## 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: 32
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 40000
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:------:|:-----:|:---------------:|:------:|
| 0.0095 | 10.04 | 1000 | 0.2023 | 9.6803 |
| 0.0032 | 20.08 | 2000 | 0.2153 | 9.0521 |
| 0.0023 | 30.11 | 3000 | 0.2234 | 8.8645 |
| 0.0023 | 40.15 | 4000 | 0.2278 | 8.4366 |
| 0.0012 | 50.19 | 5000 | 0.2260 | 7.9911 |
| 0.0005 | 60.23 | 6000 | 0.2435 | 7.9060 |
| 0.0013 | 70.26 | 7000 | 0.2254 | 7.8484 |
| 0.0004 | 80.3 | 8000 | 0.2367 | 7.4830 |
| 0.0008 | 90.34 | 9000 | 0.2289 | 7.4420 |
| 0.0007 | 100.38 | 10000 | 0.2385 | 7.5319 |
| 0.001 | 110.41 | 11000 | 0.2293 | 7.6325 |
| 0.0001 | 120.45 | 12000 | 0.2473 | 7.1430 |
| 0.0001 | 130.49 | 13000 | 0.2488 | 7.1870 |
| 0.0004 | 140.53 | 14000 | 0.2398 | 7.1831 |
| 0.0 | 150.56 | 15000 | 0.2620 | 7.0590 |
| 0.0001 | 160.6 | 16000 | 0.2547 | 7.1967 |
| 0.0 | 170.64 | 17000 | 0.2768 | 7.0736 |
| 0.0 | 180.68 | 18000 | 0.2878 | 7.0004 |
| 0.0 | 190.72 | 19000 | 0.2962 | 6.9466 |
| 0.0013 | 200.75 | 20000 | 0.2354 | 7.6042 |
| 0.0 | 210.79 | 21000 | 0.2720 | 6.8948 |
| 0.0 | 220.83 | 22000 | 0.2865 | 6.8987 |
| 0.0 | 230.87 | 23000 | 0.2954 | 6.8890 |
| 0.0 | 240.9 | 24000 | 0.3031 | 6.8821 |
| 0.0 | 250.94 | 25000 | 0.3102 | 6.8772 |
| 0.0 | 260.98 | 26000 | 0.3166 | 6.8899 |
| 0.0 | 271.02 | 27000 | 0.3233 | 6.8919 |
| 0.0 | 281.05 | 28000 | 0.3248 | 6.8919 |
| 0.0 | 291.09 | 29000 | 0.3363 | 6.9026 |
| 0.0 | 301.13 | 30000 | 0.3419 | 6.9085 |
| 0.0 | 311.17 | 31000 | 0.3471 | 6.8851 |
| 0.0 | 321.2 | 32000 | 0.3526 | 6.8704 |
| 0.0 | 331.24 | 33000 | 0.3570 | 6.8831 |
| 0.0 | 341.28 | 34000 | 0.3614 | 6.8851 |
| 0.0 | 351.32 | 35000 | 0.3645 | 6.8782 |
| 0.0 | 361.36 | 36000 | 0.3663 | 6.8714 |
| 0.0 | 371.39 | 37000 | 0.3677 | 6.8675 |
| 0.0 | 381.43 | 38000 | 0.3681 | 6.8802 |
| 0.0 | 391.47 | 39000 | 0.3686 | 6.8880 |
| 0.0 | 401.51 | 40000 | 0.3688 | 6.8880 |
### Framework versions
- Transformers 4.37.2
- Pytorch 2.2.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
|