|
|
--- |
|
|
license: cc-by-nc-4.0 |
|
|
base_model: facebook/mms-1b-all |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
metrics: |
|
|
- wer |
|
|
model-index: |
|
|
- name: wav2vec2-large-mms-1b-kazakh-speech2ner-ksc_t-16b-8ep |
|
|
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. --> |
|
|
|
|
|
# wav2vec2-large-mms-1b-kazakh-speech2ner-ksc_t-16b-8ep |
|
|
|
|
|
This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the None dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 0.2356 |
|
|
- Wer: 0.3093 |
|
|
|
|
|
## 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: 16 |
|
|
- eval_batch_size: 8 |
|
|
- seed: 42 |
|
|
- distributed_type: multi-GPU |
|
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
|
- lr_scheduler_type: linear |
|
|
- lr_scheduler_warmup_steps: 1000 |
|
|
- num_epochs: 8 |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Wer | |
|
|
|:-------------:|:-----:|:-----:|:---------------:|:------:| |
|
|
| 5.1 | 0.22 | 2000 | 5.0320 | 1.0014 | |
|
|
| 3.8579 | 0.43 | 4000 | 3.8757 | 1.0001 | |
|
|
| 3.617 | 0.65 | 6000 | 3.6434 | 1.0001 | |
|
|
| 3.56 | 0.87 | 8000 | 3.5268 | 0.9999 | |
|
|
| 3.4363 | 1.09 | 10000 | 3.4524 | 0.9999 | |
|
|
| 3.3759 | 1.3 | 12000 | 3.4010 | 0.9999 | |
|
|
| 3.3024 | 1.52 | 14000 | 3.3323 | 1.0000 | |
|
|
| 2.721 | 1.74 | 16000 | 2.4051 | 1.0048 | |
|
|
| 0.3937 | 1.96 | 18000 | 0.3398 | 0.3681 | |
|
|
| 0.3456 | 2.17 | 20000 | 0.3013 | 0.3523 | |
|
|
| 0.3269 | 2.39 | 22000 | 0.2836 | 0.3410 | |
|
|
| 0.3275 | 2.61 | 24000 | 0.2738 | 0.3346 | |
|
|
| 0.3053 | 2.83 | 26000 | 0.2667 | 0.3294 | |
|
|
| 0.3041 | 3.04 | 28000 | 0.2619 | 0.3263 | |
|
|
| 0.3084 | 3.26 | 30000 | 0.2574 | 0.3230 | |
|
|
| 0.2915 | 3.48 | 32000 | 0.2547 | 0.3207 | |
|
|
| 0.2865 | 3.69 | 34000 | 0.2521 | 0.3187 | |
|
|
| 0.2814 | 3.91 | 36000 | 0.2500 | 0.3172 | |
|
|
| 0.2961 | 4.13 | 38000 | 0.2479 | 0.3159 | |
|
|
| 0.3037 | 4.35 | 40000 | 0.2464 | 0.3147 | |
|
|
| 0.3023 | 4.56 | 42000 | 0.2448 | 0.3148 | |
|
|
| 0.2977 | 4.78 | 44000 | 0.2433 | 0.3139 | |
|
|
| 0.2933 | 5.0 | 46000 | 0.2422 | 0.3133 | |
|
|
| 0.2838 | 5.22 | 48000 | 0.2413 | 0.3120 | |
|
|
| 0.2833 | 5.43 | 50000 | 0.2401 | 0.3123 | |
|
|
| 0.2774 | 5.65 | 52000 | 0.2398 | 0.3112 | |
|
|
| 0.2813 | 5.87 | 54000 | 0.2389 | 0.3111 | |
|
|
| 0.2779 | 6.08 | 56000 | 0.2380 | 0.3108 | |
|
|
| 0.2872 | 6.3 | 58000 | 0.2376 | 0.3106 | |
|
|
| 0.2758 | 6.52 | 60000 | 0.2372 | 0.3106 | |
|
|
| 0.275 | 6.74 | 62000 | 0.2369 | 0.3095 | |
|
|
| 0.2749 | 6.95 | 64000 | 0.2364 | 0.3100 | |
|
|
| 0.2828 | 7.17 | 66000 | 0.2362 | 0.3098 | |
|
|
| 0.2749 | 7.39 | 68000 | 0.2359 | 0.3093 | |
|
|
| 0.2775 | 7.61 | 70000 | 0.2358 | 0.3093 | |
|
|
| 0.2744 | 7.82 | 72000 | 0.2356 | 0.3093 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.33.0.dev0 |
|
|
- Pytorch 2.0.1+cu117 |
|
|
- Datasets 2.13.1 |
|
|
- Tokenizers 0.13.3 |
|
|
|
