Hugging Face's logo

b1n1yam
/
shook-tiny

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Amharic
whisper
Generated from Trainer
Model card Files
xet
Metrics Community
b1n1yam commited on
Commit
145a4fe
·
verified ·
1 Parent(s): 26d6949

End of training

Browse files
Files changed (3) hide show
  1. README.md +60 -55
  2. model.safetensors +1 -1
  3. runs/Nov15_07-59-07_489f40e8c356/events.out.tfevents.1763193550.489f40e8c356.48.0 +2 -2
README.md CHANGED
@@ -9,19 +9,19 @@ tags:
9
  metrics:
10
  - wer
11
  model-index:
12
- - name: ' tiny Amharic - Biniyam Daniel'
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
- # tiny Amharic - Biniyam Daniel
20
 
21
  This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on an unknown dataset.
22
  It achieves the following results on the evaluation set:
23
- - Loss: 0.0883
24
- - Wer: 26.9112
25
 
26
  ## Model description
27
 
@@ -40,65 +40,70 @@ More information needed
40
  ### Training hyperparameters
41
 
42
  The following hyperparameters were used during training:
43
- - learning_rate: 5e-05
44
  - train_batch_size: 16
45
  - eval_batch_size: 8
46
  - seed: 42
47
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
48
- - lr_scheduler_type: linear
49
- - lr_scheduler_warmup_steps: 500
50
- - num_epochs: 3
51
  - mixed_precision_training: Native AMP
52
 
53
  ### Training results
54
 
55
- | Training Loss | Epoch | Step | Validation Loss | Wer |
56
- |:-------------:|:------:|:----:|:---------------:|:--------:|
57
- | 1.6022 | 0.0665 | 100 | 1.5464 | 127.9938 |
58
- | 1.4284 | 0.1330 | 200 | 1.4112 | 111.9973 |
59
- | 0.6919 | 0.1995 | 300 | 0.5602 | 92.5762 |
60
- | 0.3244 | 0.2660 | 400 | 0.3064 | 71.6748 |
61
- | 0.2487 | 0.3324 | 500 | 0.2587 | 63.6765 |
62
- | 0.2144 | 0.3989 | 600 | 0.2170 | 57.0482 |
63
- | 0.1957 | 0.4654 | 700 | 0.1978 | 53.9107 |
64
- | 0.1806 | 0.5319 | 800 | 0.1860 | 50.5966 |
65
- | 0.1636 | 0.5984 | 900 | 0.1717 | 48.1662 |
66
- | 0.1531 | 0.6649 | 1000 | 0.1626 | 47.1498 |
67
- | 0.1381 | 0.7314 | 1100 | 0.1530 | 44.8078 |
68
- | 0.1344 | 0.7979 | 1200 | 0.1476 | 42.8414 |
69
- | 0.1275 | 0.8644 | 1300 | 0.1420 | 42.8856 |
70
- | 0.1332 | 0.9309 | 1400 | 0.1372 | 39.9470 |
71
- | 0.1213 | 0.9973 | 1500 | 0.1364 | 41.8029 |
72
- | 0.097 | 1.0638 | 1600 | 0.1283 | 36.9642 |
73
- | 0.0942 | 1.1303 | 1700 | 0.1255 | 37.1852 |
74
- | 0.0902 | 1.1968 | 1800 | 0.1259 | 37.3398 |
75
- | 0.0976 | 1.2633 | 1900 | 0.1215 | 36.8979 |
76
- | 0.0914 | 1.3298 | 2000 | 0.1180 | 36.6991 |
77
- | 0.093 | 1.3963 | 2100 | 0.1158 | 36.5665 |
78
- | 0.0838 | 1.4628 | 2200 | 0.1116 | 33.6942 |
79
- | 0.0889 | 1.5293 | 2300 | 0.1100 | 34.5117 |
80
- | 0.084 | 1.5957 | 2400 | 0.1081 | 34.2024 |
81
- | 0.0931 | 1.6622 | 2500 | 0.1052 | 33.2302 |
82
- | 0.1029 | 1.7287 | 2600 | 0.1044 | 32.6779 |
83
- | 0.0934 | 1.7952 | 2700 | 0.1012 | 31.8604 |
84
- | 0.0816 | 1.8617 | 2800 | 0.0996 | 31.0871 |
85
- | 0.0851 | 1.9282 | 2900 | 0.0999 | 31.5510 |
86
- | 0.082 | 1.9947 | 3000 | 0.0982 | 30.9545 |
87
- | 0.0631 | 2.0612 | 3100 | 0.0983 | 30.2475 |
88
- | 0.0684 | 2.1277 | 3200 | 0.0981 | 30.3137 |
89
- | 0.0657 | 2.1941 | 3300 | 0.0999 | 29.6730 |
90
- | 0.0537 | 2.2606 | 3400 | 0.0973 | 30.4905 |
91
- | 0.0672 | 2.3271 | 3500 | 0.0962 | 29.1648 |
92
- | 0.0517 | 2.3936 | 3600 | 0.0960 | 28.3252 |
93
- | 0.0733 | 2.4601 | 3700 | 0.0934 | 27.9275 |
94
- | 0.0501 | 2.5266 | 3800 | 0.0919 | 28.0159 |
95
- | 0.0615 | 2.5931 | 3900 | 0.0909 | 28.0601 |
96
- | 0.0555 | 2.6596 | 4000 | 0.0906 | 27.4194 |
97
- | 0.0547 | 2.7261 | 4100 | 0.0891 | 27.0217 |
98
- | 0.0506 | 2.7926 | 4200 | 0.0886 | 27.2868 |
99
- | 0.0529 | 2.8590 | 4300 | 0.0885 | 27.0437 |
100
- | 0.0529 | 2.9255 | 4400 | 0.0887 | 26.9996 |
101
- | 0.0425 | 2.9920 | 4500 | 0.0883 | 26.9112 |
 
 
 
 
 
102
 
103
 
104
  ### Framework versions
 
9
  metrics:
10
  - wer
11
  model-index:
12
+ - name: ' tiny continued from check point 8e-6 - Biniyam Daniel'
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
+ # tiny continued from check point 8e-6 - Biniyam Daniel
20
 
21
  This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on an unknown dataset.
22
  It achieves the following results on the evaluation set:
23
+ - Loss: 0.0534
24
+ - Wer: 18.3821
25
 
26
  ## Model description
27
 
 
40
  ### Training hyperparameters
41
 
42
  The following hyperparameters were used during training:
43
+ - learning_rate: 8e-06
44
  - train_batch_size: 16
45
  - eval_batch_size: 8
46
  - seed: 42
47
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
48
+ - lr_scheduler_type: cosine
49
+ - lr_scheduler_warmup_ratio: 0.1
50
+ - training_steps: 5088
51
  - mixed_precision_training: Native AMP
52
 
53
  ### Training results
54
 
55
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
56
+ |:-------------:|:------:|:----:|:---------------:|:-------:|
57
+ | 0.0608 | 0.0337 | 100 | 0.0588 | 20.9205 |
58
+ | 0.0617 | 0.0675 | 200 | 0.0568 | 20.5858 |
59
+ | 0.0477 | 0.1012 | 300 | 0.0563 | 20.6137 |
60
+ | 0.0526 | 0.1350 | 400 | 0.0558 | 20.1953 |
61
+ | 0.0555 | 0.1687 | 500 | 0.0560 | 19.9721 |
62
+ | 0.0542 | 0.2025 | 600 | 0.0564 | 20.5858 |
63
+ | 0.0458 | 0.2362 | 700 | 0.0557 | 20.8368 |
64
+ | 0.0439 | 0.2700 | 800 | 0.0559 | 20.0837 |
65
+ | 0.0419 | 0.3037 | 900 | 0.0562 | 20.5021 |
66
+ | 0.0469 | 0.3375 | 1000 | 0.0556 | 19.6653 |
67
+ | 0.0457 | 0.3712 | 1100 | 0.0550 | 20.0 |
68
+ | 0.0465 | 0.4050 | 1200 | 0.0550 | 19.7768 |
69
+ | 0.0453 | 0.4387 | 1300 | 0.0552 | 19.3863 |
70
+ | 0.0425 | 0.4725 | 1400 | 0.0558 | 19.6095 |
71
+ | 0.0464 | 0.5062 | 1500 | 0.0547 | 19.6653 |
72
+ | 0.0396 | 0.5400 | 1600 | 0.0545 | 19.3863 |
73
+ | 0.043 | 0.5737 | 1700 | 0.0551 | 19.6653 |
74
+ | 0.0415 | 0.6075 | 1800 | 0.0550 | 19.3305 |
75
+ | 0.0396 | 0.6412 | 1900 | 0.0546 | 18.4937 |
76
+ | 0.0409 | 0.6750 | 2000 | 0.0542 | 18.7448 |
77
+ | 0.0418 | 0.7087 | 2100 | 0.0534 | 19.0237 |
78
+ | 0.0446 | 0.7425 | 2200 | 0.0538 | 19.1074 |
79
+ | 0.0364 | 0.7762 | 2300 | 0.0537 | 18.6053 |
80
+ | 0.0343 | 0.8100 | 2400 | 0.0537 | 18.4658 |
81
+ | 0.0437 | 0.8437 | 2500 | 0.0532 | 18.4100 |
82
+ | 0.0386 | 0.8775 | 2600 | 0.0530 | 18.9121 |
83
+ | 0.0426 | 0.9112 | 2700 | 0.0534 | 18.2706 |
84
+ | 0.0372 | 0.9450 | 2800 | 0.0536 | 18.6890 |
85
+ | 0.0325 | 0.9787 | 2900 | 0.0533 | 18.5495 |
86
+ | 0.03 | 1.0125 | 3000 | 0.0537 | 18.4100 |
87
+ | 0.0253 | 1.0462 | 3100 | 0.0545 | 18.5774 |
88
+ | 0.0316 | 1.0800 | 3200 | 0.0550 | 18.4658 |
89
+ | 0.0251 | 1.1137 | 3300 | 0.0556 | 18.7727 |
90
+ | 0.0261 | 1.1475 | 3400 | 0.0554 | 18.2427 |
91
+ | 0.0285 | 1.1812 | 3500 | 0.0551 | 18.4658 |
92
+ | 0.0234 | 1.2150 | 3600 | 0.0553 | 18.6890 |
93
+ | 0.0369 | 1.2487 | 3700 | 0.0549 | 18.3543 |
94
+ | 0.0248 | 1.2825 | 3800 | 0.0553 | 18.2985 |
95
+ | 0.0238 | 1.3162 | 3900 | 0.0551 | 18.2985 |
96
+ | 0.0278 | 1.3500 | 4000 | 0.0551 | 18.1311 |
97
+ | 0.0351 | 1.3837 | 4100 | 0.0544 | 18.4379 |
98
+ | 0.0459 | 1.4175 | 4200 | 0.0539 | 17.9916 |
99
+ | 0.0469 | 1.4512 | 4300 | 0.0537 | 18.3543 |
100
+ | 0.0384 | 1.4850 | 4400 | 0.0536 | 18.4658 |
101
+ | 0.0503 | 1.5187 | 4500 | 0.0536 | 18.4100 |
102
+ | 0.0369 | 1.5525 | 4600 | 0.0536 | 18.2985 |
103
+ | 0.0373 | 1.5862 | 4700 | 0.0535 | 18.4100 |
104
+ | 0.0376 | 1.6200 | 4800 | 0.0534 | 18.3821 |
105
+ | 0.0385 | 1.6537 | 4900 | 0.0534 | 18.3821 |
106
+ | 0.0362 | 1.6875 | 5000 | 0.0534 | 18.3821 |
107
 
108
 
109
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:685851086ff1f0f97300fb6c1c9ab0eb6695de36af3760906051a471ba3ef28b
3
  size 151061672
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4bb6c90097d377fda9a5d9dc9a6697e9ad7c0bc36be5e2a2f10c8b2ba2af27a4
3
  size 151061672
runs/Nov15_07-59-07_489f40e8c356/events.out.tfevents.1763193550.489f40e8c356.48.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b7100339ebed9016e1c786860abaaf2828ebb6c5d95f1670b4a13f4e7a611aa8
3
- size 63952
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bada86a4f0909299cd65f191d4efe5d257e4ec4b8da7fb4e0f56497fca69d532
3
+ size 64939