update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,87 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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-tr
|
| 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-tr
|
| 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: 0.2891
|
| 20 |
+
- Wer: 0.4741
|
| 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: 500
|
| 48 |
+
- num_epochs: 10
|
| 49 |
+
- mixed_precision_training: Native AMP
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 54 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|
|
| 55 |
+
| 5.4933 | 0.39 | 400 | 1.0543 | 0.9316 |
|
| 56 |
+
| 0.7039 | 0.78 | 800 | 0.6927 | 0.7702 |
|
| 57 |
+
| 0.4768 | 1.17 | 1200 | 0.4779 | 0.6774 |
|
| 58 |
+
| 0.4004 | 1.57 | 1600 | 0.4462 | 0.6450 |
|
| 59 |
+
| 0.3739 | 1.96 | 2000 | 0.4287 | 0.6296 |
|
| 60 |
+
| 0.317 | 2.35 | 2400 | 0.4395 | 0.6248 |
|
| 61 |
+
| 0.3027 | 2.74 | 2800 | 0.4052 | 0.6027 |
|
| 62 |
+
| 0.2633 | 3.13 | 3200 | 0.4026 | 0.5938 |
|
| 63 |
+
| 0.245 | 3.52 | 3600 | 0.3814 | 0.5902 |
|
| 64 |
+
| 0.2415 | 3.91 | 4000 | 0.3691 | 0.5708 |
|
| 65 |
+
| 0.2193 | 4.31 | 4400 | 0.3626 | 0.5623 |
|
| 66 |
+
| 0.2057 | 4.7 | 4800 | 0.3591 | 0.5551 |
|
| 67 |
+
| 0.1874 | 5.09 | 5200 | 0.3670 | 0.5512 |
|
| 68 |
+
| 0.1782 | 5.48 | 5600 | 0.3483 | 0.5406 |
|
| 69 |
+
| 0.1706 | 5.87 | 6000 | 0.3392 | 0.5338 |
|
| 70 |
+
| 0.153 | 6.26 | 6400 | 0.3189 | 0.5207 |
|
| 71 |
+
| 0.1493 | 6.65 | 6800 | 0.3185 | 0.5164 |
|
| 72 |
+
| 0.1381 | 7.05 | 7200 | 0.3199 | 0.5185 |
|
| 73 |
+
| 0.1244 | 7.44 | 7600 | 0.3082 | 0.4993 |
|
| 74 |
+
| 0.1182 | 7.83 | 8000 | 0.3122 | 0.4998 |
|
| 75 |
+
| 0.1136 | 8.22 | 8400 | 0.3003 | 0.4936 |
|
| 76 |
+
| 0.1047 | 8.61 | 8800 | 0.2945 | 0.4858 |
|
| 77 |
+
| 0.0986 | 9.0 | 9200 | 0.2827 | 0.4809 |
|
| 78 |
+
| 0.0925 | 9.39 | 9600 | 0.2894 | 0.4786 |
|
| 79 |
+
| 0.0885 | 9.78 | 10000 | 0.2891 | 0.4741 |
|
| 80 |
+
|
| 81 |
+
|
| 82 |
+
### Framework versions
|
| 83 |
+
|
| 84 |
+
- Transformers 4.18.0
|
| 85 |
+
- Pytorch 1.12.1+cu116
|
| 86 |
+
- Datasets 2.1.0
|
| 87 |
+
- Tokenizers 0.12.1
|