Hugging Face's logo

MathildeB3
/
52Hz-small-LORA-fr

PEFT
TensorBoard
Safetensors
Transformers
French
lora
Model card Files
xet
Metrics Community
MathildeB3 commited on
Commit
d3d3ac1
·
verified ·
1 Parent(s): a086abe

Model save

Browse files
Files changed (2) hide show
  1. README.md +37 -24
  2. adapter_model.safetensors +1 -1
README.md CHANGED
@@ -1,7 +1,5 @@
1
  ---
2
  library_name: peft
3
- language:
4
- - fr
5
  license: apache-2.0
6
  base_model: openai/whisper-small
7
  tags:
@@ -11,19 +9,19 @@ tags:
11
  metrics:
12
  - wer
13
  model-index:
14
- - name: 52Hz Small Fr - IMT Atlantique X 52 Hertz
15
  results: []
16
  ---
17
 
18
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
19
  should probably proofread and complete it, then remove this comment. -->
20
 
21
- # 52Hz Small Fr - IMT Atlantique X 52 Hertz
22
 
23
- This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Premier dataset organisé de 52 Hertz dataset.
24
  It achieves the following results on the evaluation set:
25
- - Loss: 0.3461
26
- - Wer: 29.8507
27
 
28
  ## Model description
29
 
@@ -42,7 +40,7 @@ More information needed
42
  ### Training hyperparameters
43
 
44
  The following hyperparameters were used during training:
45
- - learning_rate: 2e-05
46
  - train_batch_size: 4
47
  - eval_batch_size: 8
48
  - seed: 42
@@ -51,28 +49,43 @@ The following hyperparameters were used during training:
51
  - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
52
  - lr_scheduler_type: cosine
53
  - lr_scheduler_warmup_ratio: 0.1
54
- - num_epochs: 15
55
  - mixed_precision_training: Native AMP
56
 
57
  ### Training results
58
 
59
  | Training Loss | Epoch | Step | Validation Loss | Wer |
60
  |:-------------:|:-----:|:----:|:---------------:|:-------:|
61
- | 2.9259 | 1.0 | 21 | 1.1730 | 70.5224 |
62
- | 1.8966 | 2.0 | 42 | 0.9073 | 47.3881 |
63
- | 1.3732 | 3.0 | 63 | 0.7192 | 41.0448 |
64
- | 0.9515 | 4.0 | 84 | 0.5480 | 38.0597 |
65
- | 0.8467 | 5.0 | 105 | 0.4615 | 33.2090 |
66
- | 0.6615 | 6.0 | 126 | 0.4187 | 32.6493 |
67
- | 0.5408 | 7.0 | 147 | 0.3862 | 31.1567 |
68
- | 0.4857 | 8.0 | 168 | 0.3715 | 30.0373 |
69
- | 0.5076 | 9.0 | 189 | 0.3638 | 30.0373 |
70
- | 0.4837 | 10.0 | 210 | 0.3582 | 30.0373 |
71
- | 0.4008 | 11.0 | 231 | 0.3522 | 29.6642 |
72
- | 0.4278 | 12.0 | 252 | 0.3485 | 29.4776 |
73
- | 0.4311 | 13.0 | 273 | 0.3465 | 29.8507 |
74
- | 0.4184 | 14.0 | 294 | 0.3462 | 29.8507 |
75
- | 0.4092 | 15.0 | 315 | 0.3461 | 29.8507 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
76
 
77
 
78
  ### Framework versions
 
1
  ---
2
  library_name: peft
 
 
3
  license: apache-2.0
4
  base_model: openai/whisper-small
5
  tags:
 
9
  metrics:
10
  - wer
11
  model-index:
12
+ - name: 52Hz-small-LORA-fr
13
  results: []
14
  ---
15
 
16
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
17
  should probably proofread and complete it, then remove this comment. -->
18
 
19
+ # 52Hz-small-LORA-fr
20
 
21
+ This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the None dataset.
22
  It achieves the following results on the evaluation set:
23
+ - Loss: 0.3311
24
+ - Wer: 19.5896
25
 
26
  ## Model description
27
 
 
40
  ### Training hyperparameters
41
 
42
  The following hyperparameters were used during training:
43
+ - learning_rate: 8e-05
44
  - train_batch_size: 4
45
  - eval_batch_size: 8
46
  - seed: 42
 
49
  - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
50
  - lr_scheduler_type: cosine
51
  - lr_scheduler_warmup_ratio: 0.1
52
+ - num_epochs: 30
53
  - mixed_precision_training: Native AMP
54
 
55
  ### Training results
56
 
57
  | Training Loss | Epoch | Step | Validation Loss | Wer |
58
  |:-------------:|:-----:|:----:|:---------------:|:-------:|
59
+ | 2.6698 | 1.0 | 21 | 1.0703 | 65.6716 |
60
+ | 1.4505 | 2.0 | 42 | 0.7071 | 40.1119 |
61
+ | 0.8063 | 3.0 | 63 | 0.4246 | 33.9552 |
62
+ | 0.5185 | 4.0 | 84 | 0.3435 | 30.5970 |
63
+ | 0.3976 | 5.0 | 105 | 0.3134 | 28.5448 |
64
+ | 0.2784 | 6.0 | 126 | 0.3100 | 28.1716 |
65
+ | 0.2058 | 7.0 | 147 | 0.2878 | 23.6940 |
66
+ | 0.1687 | 8.0 | 168 | 0.3124 | 25.3731 |
67
+ | 0.1543 | 9.0 | 189 | 0.3016 | 26.6791 |
68
+ | 0.1509 | 10.0 | 210 | 0.2871 | 22.5746 |
69
+ | 0.0932 | 11.0 | 231 | 0.2906 | 24.4403 |
70
+ | 0.0881 | 12.0 | 252 | 0.3020 | 22.0149 |
71
+ | 0.0774 | 13.0 | 273 | 0.2969 | 20.3358 |
72
+ | 0.0807 | 14.0 | 294 | 0.2988 | 21.6418 |
73
+ | 0.0558 | 15.0 | 315 | 0.3083 | 19.5896 |
74
+ | 0.0487 | 16.0 | 336 | 0.3145 | 18.4701 |
75
+ | 0.0489 | 17.0 | 357 | 0.3200 | 19.4030 |
76
+ | 0.0376 | 18.0 | 378 | 0.3349 | 21.8284 |
77
+ | 0.0417 | 19.0 | 399 | 0.3251 | 22.9478 |
78
+ | 0.0288 | 20.0 | 420 | 0.3154 | 18.2836 |
79
+ | 0.0311 | 21.0 | 441 | 0.3236 | 20.3358 |
80
+ | 0.0306 | 22.0 | 462 | 0.3177 | 19.5896 |
81
+ | 0.0229 | 23.0 | 483 | 0.3253 | 19.7761 |
82
+ | 0.0231 | 24.0 | 504 | 0.3298 | 19.9627 |
83
+ | 0.0198 | 25.0 | 525 | 0.3316 | 19.7761 |
84
+ | 0.0212 | 26.0 | 546 | 0.3322 | 19.7761 |
85
+ | 0.0208 | 27.0 | 567 | 0.3309 | 19.5896 |
86
+ | 0.0183 | 28.0 | 588 | 0.3317 | 19.5896 |
87
+ | 0.0254 | 29.0 | 609 | 0.3312 | 19.5896 |
88
+ | 0.0229 | 30.0 | 630 | 0.3311 | 19.5896 |
89
 
90
 
91
  ### Framework versions
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:de6e8398fe779d2cdc492c26a2170e41130beaf2947c5cc05bcd780f2c275237
3
  size 103862440
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c62aa3b3be255327f19e73ec6e7d4f50f7f97255184d76e9d663d81dfc12a568
3
  size 103862440