Hugging Face's logo

kurosekurose
/
wav2vec2-base-EMOPIA

Transformers
PyTorch
Safetensors
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
kurosekurose commited on
Commit
1f5b79b
•
1 Parent(s): 896cb83

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +18 -23
README.md CHANGED
@@ -16,8 +16,8 @@ should probably proofread and complete it, then remove this comment. -->
16
 
17
  This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
18
  It achieves the following results on the evaluation set:
19
- - Loss: 1.0626
20
- - Accuracy: 0.6800
21
 
22
  ## Model description
23
 
@@ -45,33 +45,28 @@ The following hyperparameters were used during training:
45
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
  - lr_scheduler_type: linear
47
  - lr_scheduler_warmup_steps: 300
48
- - num_epochs: 20
49
  - mixed_precision_training: Native AMP
50
 
51
  ### Training results
52
 
53
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
54
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
55
- | 1.4068 | 0.98 | 41 | 1.3707 | 0.3000 |
56
- | 1.3568 | 1.98 | 82 | 1.3093 | 0.3200 |
57
- | 1.2503 | 2.98 | 123 | 1.1783 | 0.4600 |
58
- | 1.1088 | 3.98 | 164 | 1.1606 | 0.4200 |
59
- | 0.9996 | 4.98 | 205 | 1.0271 | 0.5400 |
60
- | 0.924 | 5.98 | 246 | 1.0093 | 0.5800 |
61
- | 0.8449 | 6.98 | 287 | 0.9613 | 0.5800 |
62
- | 0.7559 | 7.98 | 328 | 0.9370 | 0.6000 |
63
- | 0.7744 | 8.98 | 369 | 1.0094 | 0.5600 |
64
- | 0.6745 | 9.98 | 410 | 0.9141 | 0.6400 |
65
- | 0.6245 | 10.98 | 451 | 1.1952 | 0.5 |
66
- | 0.5174 | 11.98 | 492 | 0.9249 | 0.6600 |
67
- | 0.5091 | 12.98 | 533 | 0.9293 | 0.7200 |
68
- | 0.391 | 13.98 | 574 | 0.9356 | 0.6600 |
69
- | 0.3689 | 14.98 | 615 | 1.0245 | 0.6200 |
70
- | 0.2816 | 15.98 | 656 | 1.0421 | 0.6600 |
71
- | 0.3003 | 16.98 | 697 | 1.0427 | 0.6400 |
72
- | 0.2495 | 17.98 | 738 | 1.0957 | 0.6600 |
73
- | 0.2269 | 18.98 | 779 | 1.0539 | 0.6600 |
74
- | 0.2055 | 19.98 | 820 | 1.0626 | 0.6800 |
75
 
76
 
77
  ### Framework versions
 
16
 
17
  This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
18
  It achieves the following results on the evaluation set:
19
+ - Loss: 0.9346
20
+ - Accuracy: 0.7000
21
 
22
  ## Model description
23
 
 
45
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
  - lr_scheduler_type: linear
47
  - lr_scheduler_warmup_steps: 300
48
+ - num_epochs: 15
49
  - mixed_precision_training: Native AMP
50
 
51
  ### Training results
52
 
53
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
54
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
55
+ | 1.3867 | 0.99 | 58 | 1.3518 | 0.3286 |
56
+ | 1.2842 | 1.99 | 116 | 1.1872 | 0.4571 |
57
+ | 1.0725 | 2.99 | 174 | 1.0476 | 0.5571 |
58
+ | 0.9343 | 3.99 | 232 | 0.9677 | 0.5714 |
59
+ | 0.8053 | 4.99 | 290 | 0.9525 | 0.6143 |
60
+ | 0.7895 | 5.99 | 348 | 0.7735 | 0.6857 |
61
+ | 0.6867 | 6.99 | 406 | 0.8556 | 0.6429 |
62
+ | 0.6218 | 7.99 | 464 | 0.8454 | 0.6714 |
63
+ | 0.558 | 8.99 | 522 | 0.8405 | 0.6571 |
64
+ | 0.5033 | 9.99 | 580 | 1.0190 | 0.6286 |
65
+ | 0.4403 | 10.99 | 638 | 0.8249 | 0.7000 |
66
+ | 0.3995 | 11.99 | 696 | 0.8997 | 0.7143 |
67
+ | 0.3534 | 12.99 | 754 | 0.9177 | 0.7000 |
68
+ | 0.3023 | 13.99 | 812 | 0.9544 | 0.6571 |
69
+ | 0.2752 | 14.99 | 870 | 0.9346 | 0.7000 |
 
 
 
 
 
70
 
71
 
72
  ### Framework versions