Hugging Face's logo

nomnoos37
/
stt-turbo-1227-v1.1-peft-eng-1k-1e-3-rank64-reduced

PEFT
TensorBoard
Safetensors
Korean
Generated from Trainer
Model card Files Metrics Community
nomnoos37 commited on
Commit
4162729
·
verified ·
1 Parent(s): 369f535

nomnoos37/stt-turbo-1226-v1.4

Browse files
Files changed (6) hide show
  1. README.md +162 -0
  2. adapter_config.json +35 -0
  3. adapter_model.safetensors +3 -0
  4. preprocessor_config.json +14 -0
  5. runs/Dec30_09-24-35_gglabs-a6000/events.out.tfevents.1735568675.gglabs-a6000 +3 -0
  6. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,162 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ language:
4
+ - ko
5
+ license: mit
6
+ base_model: openai/whisper-large-v3-turbo
7
+ tags:
8
+ - generated_from_trainer
9
+ model-index:
10
+ - name: Whisper Small ko
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # Whisper Small ko
18
+
19
+ This model is a fine-tuned version of [openai/whisper-large-v3-turbo](https://huggingface.co/openai/whisper-large-v3-turbo) on the custom dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 0.0681
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 0.001
41
+ - train_batch_size: 64
42
+ - eval_batch_size: 256
43
+ - seed: 42
44
+ - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
45
+ - lr_scheduler_type: linear
46
+ - lr_scheduler_warmup_steps: 200
47
+ - training_steps: 1000
48
+ - mixed_precision_training: Native AMP
49
+
50
+ ### Training results
51
+
52
+ | Training Loss | Epoch | Step | Validation Loss |
53
+ |:-------------:|:------:|:----:|:---------------:|
54
+ | 0.7511 | 0.0641 | 10 | 1.4828 |
55
+ | 0.6562 | 0.1282 | 20 | 1.2493 |
56
+ | 0.3357 | 0.1923 | 30 | 0.8826 |
57
+ | 0.1966 | 0.2564 | 40 | 0.7550 |
58
+ | 0.1566 | 0.3205 | 50 | 0.6223 |
59
+ | 0.1253 | 0.3846 | 60 | 0.5636 |
60
+ | 0.104 | 0.4487 | 70 | 0.5163 |
61
+ | 0.0979 | 0.5128 | 80 | 0.4967 |
62
+ | 0.0913 | 0.5769 | 90 | 0.4598 |
63
+ | 0.107 | 0.6410 | 100 | 0.4358 |
64
+ | 0.1104 | 0.7051 | 110 | 0.4337 |
65
+ | 0.0819 | 0.7692 | 120 | 0.4072 |
66
+ | 0.0892 | 0.8333 | 130 | 0.3902 |
67
+ | 0.0954 | 0.8974 | 140 | 0.3555 |
68
+ | 0.0789 | 0.9615 | 150 | 0.3516 |
69
+ | 0.0811 | 1.0256 | 160 | 0.3237 |
70
+ | 0.0822 | 1.0897 | 170 | 0.2995 |
71
+ | 0.0747 | 1.1538 | 180 | 0.3195 |
72
+ | 0.0838 | 1.2179 | 190 | 0.3436 |
73
+ | 0.0587 | 1.2821 | 200 | 0.3178 |
74
+ | 0.0675 | 1.3462 | 210 | 0.2981 |
75
+ | 0.062 | 1.4103 | 220 | 0.2745 |
76
+ | 0.0662 | 1.4744 | 230 | 0.2650 |
77
+ | 0.0639 | 1.5385 | 240 | 0.2802 |
78
+ | 0.0745 | 1.6026 | 250 | 0.3079 |
79
+ | 0.0632 | 1.6667 | 260 | 0.2631 |
80
+ | 0.0724 | 1.7308 | 270 | 0.2406 |
81
+ | 0.0728 | 1.7949 | 280 | 0.2496 |
82
+ | 0.0601 | 1.8590 | 290 | 0.2328 |
83
+ | 0.0653 | 1.9231 | 300 | 0.2230 |
84
+ | 0.0668 | 1.9872 | 310 | 0.2076 |
85
+ | 0.0606 | 2.0513 | 320 | 0.1849 |
86
+ | 0.0543 | 2.1154 | 330 | 0.1851 |
87
+ | 0.0493 | 2.1795 | 340 | 0.1697 |
88
+ | 0.047 | 2.2436 | 350 | 0.1630 |
89
+ | 0.0451 | 2.3077 | 360 | 0.1647 |
90
+ | 0.0403 | 2.3718 | 370 | 0.1634 |
91
+ | 0.0477 | 2.4359 | 380 | 0.1530 |
92
+ | 0.0491 | 2.5 | 390 | 0.1517 |
93
+ | 0.0444 | 2.5641 | 400 | 0.1464 |
94
+ | 0.0399 | 2.6282 | 410 | 0.1421 |
95
+ | 0.0622 | 2.6923 | 420 | 0.1690 |
96
+ | 0.0509 | 2.7564 | 430 | 0.1800 |
97
+ | 0.0447 | 2.8205 | 440 | 0.1773 |
98
+ | 0.0496 | 2.8846 | 450 | 0.1765 |
99
+ | 0.049 | 2.9487 | 460 | 0.1721 |
100
+ | 0.0388 | 3.0128 | 470 | 0.1666 |
101
+ | 0.0667 | 3.0769 | 480 | 0.1477 |
102
+ | 0.038 | 3.1410 | 490 | 0.1378 |
103
+ | 0.0363 | 3.2051 | 500 | 0.1376 |
104
+ | 0.0264 | 3.2692 | 510 | 0.1289 |
105
+ | 0.032 | 3.3333 | 520 | 0.1339 |
106
+ | 0.0324 | 3.3974 | 530 | 0.1226 |
107
+ | 0.034 | 3.4615 | 540 | 0.1196 |
108
+ | 0.0257 | 3.5256 | 550 | 0.1092 |
109
+ | 0.0355 | 3.5897 | 560 | 0.1054 |
110
+ | 0.0256 | 3.6538 | 570 | 0.1096 |
111
+ | 0.033 | 3.7179 | 580 | 0.1063 |
112
+ | 0.0337 | 3.7821 | 590 | 0.1121 |
113
+ | 0.0324 | 3.8462 | 600 | 0.1065 |
114
+ | 0.0327 | 3.9103 | 610 | 0.1075 |
115
+ | 0.0314 | 3.9744 | 620 | 0.1042 |
116
+ | 0.0407 | 4.0385 | 630 | 0.0929 |
117
+ | 0.0185 | 4.1026 | 640 | 0.1295 |
118
+ | 0.0215 | 4.1667 | 650 | 0.1258 |
119
+ | 0.0198 | 4.2308 | 660 | 0.1269 |
120
+ | 0.0204 | 4.2949 | 670 | 0.1243 |
121
+ | 0.019 | 4.3590 | 680 | 0.1252 |
122
+ | 0.0236 | 4.4231 | 690 | 0.1229 |
123
+ | 0.0263 | 4.4872 | 700 | 0.1196 |
124
+ | 0.0232 | 4.5513 | 710 | 0.1184 |
125
+ | 0.0189 | 4.6154 | 720 | 0.1171 |
126
+ | 0.0187 | 4.6795 | 730 | 0.1155 |
127
+ | 0.0199 | 4.7436 | 740 | 0.1114 |
128
+ | 0.026 | 4.8077 | 750 | 0.1105 |
129
+ | 0.0193 | 4.8718 | 760 | 0.1112 |
130
+ | 0.0174 | 4.9359 | 770 | 0.1116 |
131
+ | 0.0205 | 5.0 | 780 | 0.1117 |
132
+ | 0.0132 | 5.0641 | 790 | 0.1101 |
133
+ | 0.0125 | 5.1282 | 800 | 0.1100 |
134
+ | 0.0142 | 5.1923 | 810 | 0.1098 |
135
+ | 0.0555 | 5.2564 | 820 | 0.0805 |
136
+ | 0.0138 | 5.3205 | 830 | 0.0787 |
137
+ | 0.0193 | 5.3846 | 840 | 0.0748 |
138
+ | 0.0141 | 5.4487 | 850 | 0.0731 |
139
+ | 0.0158 | 5.5128 | 860 | 0.0724 |
140
+ | 0.0133 | 5.5769 | 870 | 0.0722 |
141
+ | 0.0166 | 5.6410 | 880 | 0.0716 |
142
+ | 0.0125 | 5.7051 | 890 | 0.0706 |
143
+ | 0.0133 | 5.7692 | 900 | 0.0701 |
144
+ | 0.0209 | 5.8333 | 910 | 0.0700 |
145
+ | 0.0144 | 5.8974 | 920 | 0.0696 |
146
+ | 0.0111 | 5.9615 | 930 | 0.0697 |
147
+ | 0.0109 | 6.0256 | 940 | 0.0691 |
148
+ | 0.0131 | 6.0897 | 950 | 0.0688 |
149
+ | 0.0174 | 6.1538 | 960 | 0.0685 |
150
+ | 0.0094 | 6.2179 | 970 | 0.0684 |
151
+ | 0.0128 | 6.2821 | 980 | 0.0682 |
152
+ | 0.011 | 6.3462 | 990 | 0.0681 |
153
+ | 0.0112 | 6.4103 | 1000 | 0.0681 |
154
+
155
+
156
+ ### Framework versions
157
+
158
+ - PEFT 0.14.0
159
+ - Transformers 4.47.1
160
+ - Pytorch 2.5.1+cu124
161
+ - Datasets 3.2.0
162
+ - Tokenizers 0.21.0
adapter_config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": {
4
+ "base_model_class": "WhisperForConditionalGeneration",
5
+ "parent_library": "transformers.models.whisper.modeling_whisper"
6
+ },
7
+ "base_model_name_or_path": "openai/whisper-large-v3-turbo",
8
+ "bias": "none",
9
+ "eva_config": null,
10
+ "exclude_modules": null,
11
+ "fan_in_fan_out": false,
12
+ "inference_mode": true,
13
+ "init_lora_weights": true,
14
+ "layer_replication": null,
15
+ "layers_pattern": null,
16
+ "layers_to_transform": null,
17
+ "loftq_config": {},
18
+ "lora_alpha": 64,
19
+ "lora_bias": false,
20
+ "lora_dropout": 0.05,
21
+ "megatron_config": null,
22
+ "megatron_core": "megatron.core",
23
+ "modules_to_save": null,
24
+ "peft_type": "LORA",
25
+ "r": 64,
26
+ "rank_pattern": {},
27
+ "revision": null,
28
+ "target_modules": [
29
+ "q_proj",
30
+ "v_proj"
31
+ ],
32
+ "task_type": null,
33
+ "use_dora": false,
34
+ "use_rslora": false
35
+ }
adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b66563ff4036fc09f27bdd41312caf5ea6a70089c3a858b822cc2f88e09366d2
3
+ size 52451624
preprocessor_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "chunk_length": 30,
3
+ "feature_extractor_type": "WhisperFeatureExtractor",
4
+ "feature_size": 128,
5
+ "hop_length": 160,
6
+ "n_fft": 400,
7
+ "n_samples": 480000,
8
+ "nb_max_frames": 3000,
9
+ "padding_side": "right",
10
+ "padding_value": 0.0,
11
+ "processor_class": "WhisperProcessor",
12
+ "return_attention_mask": false,
13
+ "sampling_rate": 16000
14
+ }
runs/Dec30_09-24-35_gglabs-a6000/events.out.tfevents.1735568675.gglabs-a6000 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20e7c9f04a2a7dcef5f00ca0a099e9f65e464631739feb63b6b1bfd235c3252d
3
+ size 54835
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:974d59dc71f0a091f4e6687f6460e4c0ad45c3c507c7b45f5f4ab063e8813e52
3
+ size 5560