File size: 3,626 Bytes
f72b4a5 |
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 |
---
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
|