update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
tags:
|
| 3 |
+
- generated_from_trainer
|
| 4 |
+
datasets:
|
| 5 |
+
- common_voice
|
| 6 |
+
model-index:
|
| 7 |
+
- name: wav2vec2_base_10k_8khz_pt_cv7_2
|
| 8 |
+
results: []
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 12 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 13 |
+
|
| 14 |
+
# wav2vec2_base_10k_8khz_pt_cv7_2
|
| 15 |
+
|
| 16 |
+
This model is a fine-tuned version of [lgris/seasr_2022_base_10k_8khz_pt](https://huggingface.co/lgris/seasr_2022_base_10k_8khz_pt) on the common_voice dataset.
|
| 17 |
+
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 76.3426
|
| 19 |
+
- Wer: 0.1979
|
| 20 |
+
|
| 21 |
+
## Model description
|
| 22 |
+
|
| 23 |
+
More information needed
|
| 24 |
+
|
| 25 |
+
## Intended uses & limitations
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Training and evaluation data
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Training procedure
|
| 34 |
+
|
| 35 |
+
### Training hyperparameters
|
| 36 |
+
|
| 37 |
+
The following hyperparameters were used during training:
|
| 38 |
+
- learning_rate: 1e-05
|
| 39 |
+
- train_batch_size: 8
|
| 40 |
+
- eval_batch_size: 8
|
| 41 |
+
- seed: 42
|
| 42 |
+
- gradient_accumulation_steps: 2
|
| 43 |
+
- total_train_batch_size: 16
|
| 44 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
+
- lr_scheduler_type: linear
|
| 46 |
+
- lr_scheduler_warmup_steps: 100
|
| 47 |
+
- training_steps: 10000
|
| 48 |
+
|
| 49 |
+
### Training results
|
| 50 |
+
|
| 51 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 52 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|
|
| 53 |
+
| 189.1362 | 0.65 | 500 | 80.6347 | 0.2139 |
|
| 54 |
+
| 174.2587 | 1.3 | 1000 | 80.2062 | 0.2116 |
|
| 55 |
+
| 164.676 | 1.95 | 1500 | 78.2161 | 0.2073 |
|
| 56 |
+
| 176.5856 | 2.6 | 2000 | 78.8920 | 0.2074 |
|
| 57 |
+
| 164.3583 | 3.25 | 2500 | 77.2865 | 0.2066 |
|
| 58 |
+
| 161.414 | 3.9 | 3000 | 77.8888 | 0.2048 |
|
| 59 |
+
| 158.283 | 4.55 | 3500 | 77.3472 | 0.2033 |
|
| 60 |
+
| 159.2265 | 5.19 | 4000 | 79.0953 | 0.2036 |
|
| 61 |
+
| 156.3967 | 5.84 | 4500 | 76.6855 | 0.2029 |
|
| 62 |
+
| 154.2743 | 6.49 | 5000 | 77.7785 | 0.2015 |
|
| 63 |
+
| 156.6497 | 7.14 | 5500 | 77.1220 | 0.2033 |
|
| 64 |
+
| 157.3038 | 7.79 | 6000 | 76.2926 | 0.2027 |
|
| 65 |
+
| 162.8151 | 8.44 | 6500 | 76.7602 | 0.2013 |
|
| 66 |
+
| 151.8613 | 9.09 | 7000 | 77.4777 | 0.2011 |
|
| 67 |
+
| 153.0225 | 9.74 | 7500 | 76.5206 | 0.2001 |
|
| 68 |
+
| 157.52 | 10.39 | 8000 | 76.1061 | 0.2006 |
|
| 69 |
+
| 145.0592 | 11.04 | 8500 | 76.7855 | 0.1992 |
|
| 70 |
+
| 150.0066 | 11.69 | 9000 | 76.0058 | 0.1988 |
|
| 71 |
+
| 146.8128 | 12.34 | 9500 | 76.2853 | 0.1987 |
|
| 72 |
+
| 146.9148 | 12.99 | 10000 | 76.3426 | 0.1979 |
|
| 73 |
+
|
| 74 |
+
|
| 75 |
+
### Framework versions
|
| 76 |
+
|
| 77 |
+
- Transformers 4.16.2
|
| 78 |
+
- Pytorch 1.10.0+cu111
|
| 79 |
+
- Datasets 1.18.3
|
| 80 |
+
- Tokenizers 0.11.0
|