File size: 4,203 Bytes
0bbdad1 01b2793 0bbdad1 01b2793 0bbdad1 01b2793 0bbdad1 01b2793 0bbdad1 c8b6183 0bbdad1 c8b6183 0bbdad1 c8b6183 0bbdad1 c8b6183 0bbdad1 |
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 |
---
language:
- ug
license: apache-2.0
tags:
- automatic-speech-recognition
- mozilla-foundation/common_voice_8_0
- generated_from_trainer
- robust-speech-event
- ug
datasets:
- mozilla-foundation/common_voice_8_0
model-index:
- name: XLS-R-300M Uyghur CV8
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 8
type: mozilla-foundation/common_voice_8_0
args: ug
metrics:
- name: Test WER
type: wer
value: 36.33
- name: Test CER
type: cer
value: 6.75
---
<!-- 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. -->
# XLS-R-300M Uyghur CV8
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_8_0 - UG dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2240
- Wer: 0.3693
## 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: 4e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2000
- num_epochs: 100.0
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:-----:|:---------------:|:------:|
| 4.1169 | 2.66 | 500 | 4.0146 | 1.0 |
| 3.2512 | 5.32 | 1000 | 3.2342 | 1.0 |
| 2.5435 | 7.97 | 1500 | 1.8155 | 1.0286 |
| 1.5575 | 10.64 | 2000 | 0.6346 | 0.7058 |
| 1.3979 | 13.3 | 2500 | 0.4885 | 0.6320 |
| 1.2874 | 15.95 | 3000 | 0.4271 | 0.6088 |
| 1.2383 | 18.61 | 3500 | 0.3889 | 0.5869 |
| 1.2054 | 21.28 | 4000 | 0.3609 | 0.5793 |
| 1.1866 | 23.93 | 4500 | 0.3450 | 0.5513 |
| 1.1332 | 26.59 | 5000 | 0.3214 | 0.5379 |
| 1.135 | 29.25 | 5500 | 0.3122 | 0.5384 |
| 1.0992 | 31.91 | 6000 | 0.2948 | 0.5078 |
| 1.0707 | 34.57 | 6500 | 0.2928 | 0.5128 |
| 1.0754 | 37.23 | 7000 | 0.2857 | 0.5017 |
| 1.0461 | 39.89 | 7500 | 0.2791 | 0.5099 |
| 1.0328 | 42.55 | 8000 | 0.2729 | 0.5120 |
| 1.0201 | 45.21 | 8500 | 0.2654 | 0.4720 |
| 1.0035 | 47.87 | 9000 | 0.2623 | 0.4659 |
| 1.0069 | 50.53 | 9500 | 0.2569 | 0.4593 |
| 0.9998 | 53.19 | 10000 | 0.2519 | 0.4405 |
| 0.9762 | 55.85 | 10500 | 0.2505 | 0.4588 |
| 0.9755 | 58.51 | 11000 | 0.2479 | 0.4564 |
| 0.9624 | 61.17 | 11500 | 0.2460 | 0.4298 |
| 0.9494 | 63.83 | 12000 | 0.2402 | 0.4182 |
| 0.948 | 66.49 | 12500 | 0.2412 | 0.4212 |
| 0.9312 | 69.15 | 13000 | 0.2352 | 0.3970 |
| 0.9172 | 71.81 | 13500 | 0.2357 | 0.3926 |
| 0.9101 | 74.47 | 14000 | 0.2305 | 0.3905 |
| 0.9177 | 77.13 | 14500 | 0.2307 | 0.3838 |
| 0.9083 | 79.78 | 15000 | 0.2313 | 0.3800 |
| 0.9068 | 82.45 | 15500 | 0.2275 | 0.3742 |
| 0.9087 | 85.11 | 16000 | 0.2283 | 0.3747 |
| 0.8838 | 87.76 | 16500 | 0.2286 | 0.3777 |
| 0.8868 | 90.42 | 17000 | 0.2269 | 0.3722 |
| 0.8895 | 93.08 | 17500 | 0.2246 | 0.3714 |
| 0.8926 | 95.74 | 18000 | 0.2241 | 0.3705 |
| 0.8856 | 98.4 | 18500 | 0.2242 | 0.3693 |
### Framework versions
- Transformers 4.16.0.dev0
- Pytorch 1.10.1+cu102
- Datasets 1.18.2.dev0
- Tokenizers 0.11.0
|