Commit
•
4efef42
1
Parent(s):
3e79ed7
update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,100 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
datasets:
|
| 6 |
+
- common_voice
|
| 7 |
+
model-index:
|
| 8 |
+
- name: wav2vec2-large-xls-r-300m-as-g1
|
| 9 |
+
results: []
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 13 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 14 |
+
|
| 15 |
+
# wav2vec2-large-xls-r-300m-as-g1
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the common_voice dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 1.3327
|
| 20 |
+
- Wer: 0.5744
|
| 21 |
+
|
| 22 |
+
## Model description
|
| 23 |
+
|
| 24 |
+
More information needed
|
| 25 |
+
|
| 26 |
+
## Intended uses & limitations
|
| 27 |
+
|
| 28 |
+
More information needed
|
| 29 |
+
|
| 30 |
+
## Training and evaluation data
|
| 31 |
+
|
| 32 |
+
More information needed
|
| 33 |
+
|
| 34 |
+
## Training procedure
|
| 35 |
+
|
| 36 |
+
### Training hyperparameters
|
| 37 |
+
|
| 38 |
+
The following hyperparameters were used during training:
|
| 39 |
+
- learning_rate: 0.0003
|
| 40 |
+
- train_batch_size: 16
|
| 41 |
+
- eval_batch_size: 8
|
| 42 |
+
- seed: 42
|
| 43 |
+
- gradient_accumulation_steps: 2
|
| 44 |
+
- total_train_batch_size: 32
|
| 45 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
+
- lr_scheduler_type: linear
|
| 47 |
+
- lr_scheduler_warmup_steps: 1000
|
| 48 |
+
- num_epochs: 200
|
| 49 |
+
- mixed_precision_training: Native AMP
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 54 |
+
|:-------------:|:------:|:----:|:---------------:|:------:|
|
| 55 |
+
| 14.1958 | 5.26 | 100 | 7.1919 | 1.0 |
|
| 56 |
+
| 5.0035 | 10.51 | 200 | 3.9362 | 1.0 |
|
| 57 |
+
| 3.6193 | 15.77 | 300 | 3.4451 | 1.0 |
|
| 58 |
+
| 3.4852 | 21.05 | 400 | 3.3536 | 1.0 |
|
| 59 |
+
| 2.8489 | 26.31 | 500 | 1.6451 | 0.9100 |
|
| 60 |
+
| 0.9568 | 31.56 | 600 | 1.0514 | 0.7561 |
|
| 61 |
+
| 0.4865 | 36.82 | 700 | 1.0434 | 0.7184 |
|
| 62 |
+
| 0.322 | 42.1 | 800 | 1.0825 | 0.7210 |
|
| 63 |
+
| 0.2383 | 47.36 | 900 | 1.1304 | 0.6897 |
|
| 64 |
+
| 0.2136 | 52.62 | 1000 | 1.1150 | 0.6854 |
|
| 65 |
+
| 0.179 | 57.87 | 1100 | 1.2453 | 0.6875 |
|
| 66 |
+
| 0.1539 | 63.15 | 1200 | 1.2211 | 0.6704 |
|
| 67 |
+
| 0.1303 | 68.41 | 1300 | 1.2859 | 0.6747 |
|
| 68 |
+
| 0.1183 | 73.67 | 1400 | 1.2775 | 0.6721 |
|
| 69 |
+
| 0.0994 | 78.92 | 1500 | 1.2321 | 0.6404 |
|
| 70 |
+
| 0.0991 | 84.21 | 1600 | 1.2766 | 0.6524 |
|
| 71 |
+
| 0.0887 | 89.46 | 1700 | 1.3026 | 0.6344 |
|
| 72 |
+
| 0.0754 | 94.72 | 1800 | 1.3199 | 0.6704 |
|
| 73 |
+
| 0.0693 | 99.97 | 1900 | 1.3044 | 0.6361 |
|
| 74 |
+
| 0.0568 | 105.26 | 2000 | 1.3541 | 0.6254 |
|
| 75 |
+
| 0.0536 | 110.51 | 2100 | 1.3320 | 0.6249 |
|
| 76 |
+
| 0.0529 | 115.77 | 2200 | 1.3370 | 0.6271 |
|
| 77 |
+
| 0.048 | 121.05 | 2300 | 1.2757 | 0.6031 |
|
| 78 |
+
| 0.0419 | 126.31 | 2400 | 1.2661 | 0.6172 |
|
| 79 |
+
| 0.0349 | 131.56 | 2500 | 1.2897 | 0.6048 |
|
| 80 |
+
| 0.0309 | 136.82 | 2600 | 1.2688 | 0.5962 |
|
| 81 |
+
| 0.0278 | 142.1 | 2700 | 1.2885 | 0.5954 |
|
| 82 |
+
| 0.0254 | 147.36 | 2800 | 1.2988 | 0.5915 |
|
| 83 |
+
| 0.0223 | 152.62 | 2900 | 1.3153 | 0.5941 |
|
| 84 |
+
| 0.0216 | 157.87 | 3000 | 1.2936 | 0.5937 |
|
| 85 |
+
| 0.0186 | 163.15 | 3100 | 1.2906 | 0.5877 |
|
| 86 |
+
| 0.0156 | 168.41 | 3200 | 1.3476 | 0.5962 |
|
| 87 |
+
| 0.0158 | 173.67 | 3300 | 1.3363 | 0.5847 |
|
| 88 |
+
| 0.0142 | 178.92 | 3400 | 1.3367 | 0.5847 |
|
| 89 |
+
| 0.0153 | 184.21 | 3500 | 1.3105 | 0.5757 |
|
| 90 |
+
| 0.0119 | 189.46 | 3600 | 1.3255 | 0.5705 |
|
| 91 |
+
| 0.0115 | 194.72 | 3700 | 1.3340 | 0.5787 |
|
| 92 |
+
| 0.0103 | 199.97 | 3800 | 1.3327 | 0.5744 |
|
| 93 |
+
|
| 94 |
+
|
| 95 |
+
### Framework versions
|
| 96 |
+
|
| 97 |
+
- Transformers 4.16.2
|
| 98 |
+
- Pytorch 1.10.0+cu111
|
| 99 |
+
- Datasets 1.18.3
|
| 100 |
+
- Tokenizers 0.11.0
|