Commit
·
9f94b88
1
Parent(s):
5526bc6
update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,106 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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-ar
|
| 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-ar
|
| 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: 0.4819
|
| 20 |
+
- Wer: 0.4244
|
| 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: 3e-05
|
| 40 |
+
- train_batch_size: 32
|
| 41 |
+
- eval_batch_size: 4
|
| 42 |
+
- seed: 42
|
| 43 |
+
- gradient_accumulation_steps: 2
|
| 44 |
+
- total_train_batch_size: 64
|
| 45 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
+
- lr_scheduler_type: linear
|
| 47 |
+
- lr_scheduler_warmup_steps: 400
|
| 48 |
+
- num_epochs: 30
|
| 49 |
+
- mixed_precision_training: Native AMP
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 54 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|
|
| 55 |
+
| 11.0435 | 0.67 | 400 | 4.3104 | 1.0 |
|
| 56 |
+
| 3.4451 | 1.34 | 800 | 3.1566 | 1.0 |
|
| 57 |
+
| 3.1399 | 2.01 | 1200 | 3.0532 | 0.9990 |
|
| 58 |
+
| 2.8538 | 2.68 | 1600 | 1.6994 | 0.9238 |
|
| 59 |
+
| 1.7195 | 3.35 | 2000 | 0.8867 | 0.6727 |
|
| 60 |
+
| 1.326 | 4.02 | 2400 | 0.6603 | 0.5834 |
|
| 61 |
+
| 1.1561 | 4.69 | 2800 | 0.5809 | 0.5479 |
|
| 62 |
+
| 1.0764 | 5.36 | 3200 | 0.5943 | 0.5495 |
|
| 63 |
+
| 1.0144 | 6.03 | 3600 | 0.5344 | 0.5251 |
|
| 64 |
+
| 0.965 | 6.7 | 4000 | 0.4844 | 0.4936 |
|
| 65 |
+
| 0.927 | 7.37 | 4400 | 0.5048 | 0.5019 |
|
| 66 |
+
| 0.8985 | 8.04 | 4800 | 0.5809 | 0.5267 |
|
| 67 |
+
| 0.8684 | 8.71 | 5200 | 0.4740 | 0.4753 |
|
| 68 |
+
| 0.8581 | 9.38 | 5600 | 0.4813 | 0.4834 |
|
| 69 |
+
| 0.8334 | 10.05 | 6000 | 0.4515 | 0.4545 |
|
| 70 |
+
| 0.8134 | 10.72 | 6400 | 0.4370 | 0.4543 |
|
| 71 |
+
| 0.8002 | 11.39 | 6800 | 0.4225 | 0.4384 |
|
| 72 |
+
| 0.7884 | 12.06 | 7200 | 0.4593 | 0.4565 |
|
| 73 |
+
| 0.7675 | 12.73 | 7600 | 0.4752 | 0.4680 |
|
| 74 |
+
| 0.7607 | 13.4 | 8000 | 0.4950 | 0.4771 |
|
| 75 |
+
| 0.7475 | 14.07 | 8400 | 0.4373 | 0.4391 |
|
| 76 |
+
| 0.7397 | 14.74 | 8800 | 0.4506 | 0.4541 |
|
| 77 |
+
| 0.7289 | 15.41 | 9200 | 0.4840 | 0.4691 |
|
| 78 |
+
| 0.722 | 16.08 | 9600 | 0.4701 | 0.4571 |
|
| 79 |
+
| 0.7067 | 16.75 | 10000 | 0.4561 | 0.4461 |
|
| 80 |
+
| 0.7033 | 17.42 | 10400 | 0.4384 | 0.4347 |
|
| 81 |
+
| 0.6915 | 18.09 | 10800 | 0.4424 | 0.4290 |
|
| 82 |
+
| 0.6854 | 18.76 | 11200 | 0.4635 | 0.4360 |
|
| 83 |
+
| 0.6813 | 19.43 | 11600 | 0.4280 | 0.4147 |
|
| 84 |
+
| 0.6776 | 20.1 | 12000 | 0.4610 | 0.4344 |
|
| 85 |
+
| 0.67 | 20.77 | 12400 | 0.4540 | 0.4367 |
|
| 86 |
+
| 0.6653 | 21.44 | 12800 | 0.4509 | 0.4234 |
|
| 87 |
+
| 0.6609 | 22.11 | 13200 | 0.4874 | 0.4444 |
|
| 88 |
+
| 0.6541 | 22.78 | 13600 | 0.4542 | 0.4230 |
|
| 89 |
+
| 0.6528 | 23.45 | 14000 | 0.4732 | 0.4373 |
|
| 90 |
+
| 0.6463 | 24.12 | 14400 | 0.4483 | 0.4188 |
|
| 91 |
+
| 0.6399 | 24.79 | 14800 | 0.4731 | 0.4341 |
|
| 92 |
+
| 0.6353 | 25.46 | 15200 | 0.5031 | 0.4412 |
|
| 93 |
+
| 0.6358 | 26.13 | 15600 | 0.4986 | 0.4397 |
|
| 94 |
+
| 0.6317 | 26.8 | 16000 | 0.5000 | 0.4360 |
|
| 95 |
+
| 0.6262 | 27.47 | 16400 | 0.4958 | 0.4318 |
|
| 96 |
+
| 0.6317 | 28.14 | 16800 | 0.4738 | 0.4234 |
|
| 97 |
+
| 0.6205 | 28.81 | 17200 | 0.4853 | 0.4262 |
|
| 98 |
+
| 0.6205 | 29.48 | 17600 | 0.4819 | 0.4244 |
|
| 99 |
+
|
| 100 |
+
|
| 101 |
+
### Framework versions
|
| 102 |
+
|
| 103 |
+
- Transformers 4.17.0.dev0
|
| 104 |
+
- Pytorch 1.10.2+cu102
|
| 105 |
+
- Datasets 1.18.3
|
| 106 |
+
- Tokenizers 0.11.0
|