Hugging Face's logo

8688chris
/
Helldivers2ASR_V4

Safetensors
wav2vec2
Generated from Trainer
Model card Files
xet
 Community
8688chris commited on
Commit
6b8918d
·
verified ·
1 Parent(s): 5d30e79

Model save

Browse files
Files changed (4) hide show
  1. README.md +63 -63
  2. config.json +4 -4
  3. model.safetensors +1 -1
  4. training_args.bin +1 -1
README.md CHANGED
@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
17
 
18
  This model is a fine-tuned version of [facebook/wav2vec2-large-960h](https://huggingface.co/facebook/wav2vec2-large-960h) on an unknown dataset.
19
  It achieves the following results on the evaluation set:
20
- - Loss: 32.5406
21
- - Wer: 0.0515
22
 
23
  ## Model description
24
 
@@ -42,7 +42,7 @@ The following hyperparameters were used during training:
42
  - eval_batch_size: 16
43
  - seed: 42
44
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
45
- - lr_scheduler_type: linear
46
  - lr_scheduler_warmup_ratio: 0.1
47
  - num_epochs: 60
48
  - mixed_precision_training: Native AMP
@@ -51,66 +51,66 @@ The following hyperparameters were used during training:
51
 
52
  | Training Loss | Epoch | Step | Validation Loss | Wer |
53
  |:-------------:|:-----:|:-----:|:---------------:|:------:|
54
- | 1153.3836 | 1.0 | 488 | 850.3044 | 0.6449 |
55
- | 928.434 | 2.0 | 976 | 691.6172 | 0.5685 |
56
- | 811.1838 | 3.0 | 1464 | 598.8566 | 0.4987 |
57
- | 717.7333 | 4.0 | 1952 | 552.0085 | 0.4545 |
58
- | 647.3931 | 5.0 | 2440 | 463.2961 | 0.3979 |
59
- | 586.5474 | 6.0 | 2928 | 417.2820 | 0.3674 |
60
- | 533.4568 | 7.0 | 3416 | 398.9988 | 0.3179 |
61
- | 488.8762 | 8.0 | 3904 | 369.4180 | 0.3068 |
62
- | 444.0211 | 9.0 | 4392 | 301.7917 | 0.2615 |
63
- | 413.0845 | 10.0 | 4880 | 251.3396 | 0.2400 |
64
- | 372.2671 | 11.0 | 5368 | 249.1012 | 0.2238 |
65
- | 347.6574 | 12.0 | 5856 | 230.7993 | 0.1970 |
66
- | 314.716 | 13.0 | 6344 | 217.9398 | 0.1959 |
67
- | 298.8911 | 14.0 | 6832 | 189.8561 | 0.1857 |
68
- | 283.848 | 15.0 | 7320 | 191.2605 | 0.1741 |
69
- | 258.3081 | 16.0 | 7808 | 143.3509 | 0.1580 |
70
- | 243.6687 | 17.0 | 8296 | 145.5848 | 0.1499 |
71
- | 233.2926 | 18.0 | 8784 | 140.6835 | 0.1495 |
72
- | 211.4206 | 19.0 | 9272 | 124.2010 | 0.1386 |
73
- | 208.7865 | 20.0 | 9760 | 118.0089 | 0.1269 |
74
- | 195.4273 | 21.0 | 10248 | 117.8243 | 0.1262 |
75
- | 184.6244 | 22.0 | 10736 | 97.7215 | 0.1164 |
76
- | 171.8441 | 23.0 | 11224 | 111.4605 | 0.1131 |
77
- | 163.1848 | 24.0 | 11712 | 89.1710 | 0.1134 |
78
- | 153.9752 | 25.0 | 12200 | 87.3499 | 0.1047 |
79
- | 147.2633 | 26.0 | 12688 | 87.4700 | 0.1082 |
80
- | 142.8953 | 27.0 | 13176 | 83.6597 | 0.1020 |
81
- | 136.2411 | 28.0 | 13664 | 82.2492 | 0.0981 |
82
- | 125.4659 | 29.0 | 14152 | 80.2814 | 0.0996 |
83
- | 123.5863 | 30.0 | 14640 | 72.3761 | 0.0878 |
84
- | 123.4564 | 31.0 | 15128 | 74.1428 | 0.0898 |
85
- | 119.1795 | 32.0 | 15616 | 60.4706 | 0.0805 |
86
- | 106.3762 | 33.0 | 16104 | 70.6227 | 0.0834 |
87
- | 108.1521 | 34.0 | 16592 | 59.9063 | 0.0785 |
88
- | 102.576 | 35.0 | 17080 | 56.8914 | 0.0781 |
89
- | 94.8957 | 36.0 | 17568 | 58.9398 | 0.0785 |
90
- | 96.225 | 37.0 | 18056 | 55.8423 | 0.0777 |
91
- | 91.3152 | 38.0 | 18544 | 55.6489 | 0.0747 |
92
- | 94.8994 | 39.0 | 19032 | 55.0808 | 0.0701 |
93
- | 86.3052 | 40.0 | 19520 | 51.2320 | 0.0716 |
94
- | 81.8472 | 41.0 | 20008 | 47.2280 | 0.0641 |
95
- | 84.5362 | 42.0 | 20496 | 48.6431 | 0.0654 |
96
- | 80.2845 | 43.0 | 20984 | 44.2606 | 0.0663 |
97
- | 79.6424 | 44.0 | 21472 | 46.5519 | 0.0612 |
98
- | 78.9052 | 45.0 | 21960 | 47.1750 | 0.0654 |
99
- | 75.1781 | 46.0 | 22448 | 41.0039 | 0.0630 |
100
- | 66.398 | 47.0 | 22936 | 41.1136 | 0.0599 |
101
- | 68.9468 | 48.0 | 23424 | 39.4498 | 0.0577 |
102
- | 69.0842 | 49.0 | 23912 | 39.5141 | 0.0590 |
103
- | 63.1563 | 50.0 | 24400 | 38.1891 | 0.0572 |
104
- | 66.0264 | 51.0 | 24888 | 36.2524 | 0.0606 |
105
- | 64.2084 | 52.0 | 25376 | 39.5151 | 0.0579 |
106
- | 63.0642 | 53.0 | 25864 | 36.7211 | 0.0579 |
107
- | 61.9061 | 54.0 | 26352 | 37.9883 | 0.0564 |
108
- | 60.9176 | 55.0 | 26840 | 35.0138 | 0.0528 |
109
- | 56.4366 | 56.0 | 27328 | 33.8095 | 0.0519 |
110
- | 60.1974 | 57.0 | 27816 | 32.5853 | 0.0492 |
111
- | 57.5505 | 58.0 | 28304 | 34.2817 | 0.0512 |
112
- | 58.9594 | 59.0 | 28792 | 32.7452 | 0.0519 |
113
- | 58.9188 | 60.0 | 29280 | 32.5406 | 0.0515 |
114
 
115
 
116
  ### Framework versions
 
17
 
18
  This model is a fine-tuned version of [facebook/wav2vec2-large-960h](https://huggingface.co/facebook/wav2vec2-large-960h) on an unknown dataset.
19
  It achieves the following results on the evaluation set:
20
+ - Loss: 22.0317
21
+ - Wer: 0.0531
22
 
23
  ## Model description
24
 
 
42
  - eval_batch_size: 16
43
  - seed: 42
44
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
45
+ - lr_scheduler_type: constant
46
  - lr_scheduler_warmup_ratio: 0.1
47
  - num_epochs: 60
48
  - mixed_precision_training: Native AMP
 
51
 
52
  | Training Loss | Epoch | Step | Validation Loss | Wer |
53
  |:-------------:|:-----:|:-----:|:---------------:|:------:|
54
+ | 898.4883 | 1.0 | 505 | 613.9966 | 0.5030 |
55
+ | 700.9432 | 2.0 | 1010 | 474.9732 | 0.4287 |
56
+ | 616.9108 | 3.0 | 1515 | 412.2095 | 0.3757 |
57
+ | 546.7158 | 4.0 | 2020 | 376.4230 | 0.3477 |
58
+ | 495.3992 | 5.0 | 2525 | 327.5403 | 0.3052 |
59
+ | 457.4469 | 6.0 | 3030 | 277.4165 | 0.2692 |
60
+ | 419.2608 | 7.0 | 3535 | 264.2311 | 0.2557 |
61
+ | 387.4688 | 8.0 | 4040 | 260.7827 | 0.2309 |
62
+ | 358.4347 | 9.0 | 4545 | 213.9615 | 0.2112 |
63
+ | 334.2494 | 10.0 | 5050 | 198.8723 | 0.1903 |
64
+ | 309.9395 | 11.0 | 5555 | 164.4527 | 0.1853 |
65
+ | 288.1617 | 12.0 | 6060 | 164.5457 | 0.1783 |
66
+ | 265.1981 | 13.0 | 6565 | 151.1678 | 0.1663 |
67
+ | 247.7314 | 14.0 | 7070 | 142.4500 | 0.1665 |
68
+ | 234.6288 | 15.0 | 7575 | 119.6140 | 0.1476 |
69
+ | 217.8525 | 16.0 | 8080 | 118.9178 | 0.1399 |
70
+ | 212.3975 | 17.0 | 8585 | 109.2142 | 0.1278 |
71
+ | 197.9598 | 18.0 | 9090 | 98.2246 | 0.1245 |
72
+ | 181.3615 | 19.0 | 9595 | 98.2521 | 0.1250 |
73
+ | 177.8549 | 20.0 | 10100 | 83.1861 | 0.1130 |
74
+ | 170.1361 | 21.0 | 10605 | 83.6687 | 0.1139 |
75
+ | 152.4907 | 22.0 | 11110 | 72.1740 | 0.0974 |
76
+ | 149.9199 | 23.0 | 11615 | 76.4454 | 0.0990 |
77
+ | 146.3055 | 24.0 | 12120 | 66.7212 | 0.0949 |
78
+ | 140.3444 | 25.0 | 12625 | 65.7427 | 0.0927 |
79
+ | 131.9466 | 26.0 | 13130 | 62.4010 | 0.0875 |
80
+ | 131.0567 | 27.0 | 13635 | 63.4794 | 0.0921 |
81
+ | 120.657 | 28.0 | 14140 | 58.2065 | 0.0824 |
82
+ | 118.117 | 29.0 | 14645 | 57.8397 | 0.0762 |
83
+ | 112.2041 | 30.0 | 15150 | 48.3423 | 0.0791 |
84
+ | 112.5141 | 31.0 | 15655 | 51.5712 | 0.0825 |
85
+ | 103.9626 | 32.0 | 16160 | 42.7415 | 0.0661 |
86
+ | 100.5436 | 33.0 | 16665 | 45.6458 | 0.0704 |
87
+ | 101.6996 | 34.0 | 17170 | 41.2959 | 0.0647 |
88
+ | 94.8516 | 35.0 | 17675 | 41.7417 | 0.0694 |
89
+ | 96.989 | 36.0 | 18180 | 41.1608 | 0.0647 |
90
+ | 92.9812 | 37.0 | 18685 | 44.2804 | 0.0738 |
91
+ | 88.2919 | 38.0 | 19190 | 41.8735 | 0.0654 |
92
+ | 84.4574 | 39.0 | 19695 | 41.3105 | 0.0615 |
93
+ | 82.7767 | 40.0 | 20200 | 39.5144 | 0.0651 |
94
+ | 80.9119 | 41.0 | 20705 | 31.5262 | 0.0545 |
95
+ | 80.6782 | 42.0 | 21210 | 34.7946 | 0.0603 |
96
+ | 76.5013 | 43.0 | 21715 | 31.3224 | 0.0594 |
97
+ | 76.2344 | 44.0 | 22220 | 35.6723 | 0.0625 |
98
+ | 76.0117 | 45.0 | 22725 | 32.7431 | 0.0688 |
99
+ | 75.4117 | 46.0 | 23230 | 34.8090 | 0.0617 |
100
+ | 74.5649 | 47.0 | 23735 | 35.5215 | 0.0726 |
101
+ | 70.9009 | 48.0 | 24240 | 30.4729 | 0.0654 |
102
+ | 70.0265 | 49.0 | 24745 | 34.5584 | 0.0719 |
103
+ | 70.4782 | 50.0 | 25250 | 38.5341 | 0.0699 |
104
+ | 66.4355 | 51.0 | 25755 | 31.3699 | 0.0700 |
105
+ | 68.6737 | 52.0 | 26260 | 37.2897 | 0.0695 |
106
+ | 64.2738 | 53.0 | 26765 | 33.8824 | 0.0699 |
107
+ | 62.5659 | 54.0 | 27270 | 29.2762 | 0.0632 |
108
+ | 62.469 | 55.0 | 27775 | 32.5962 | 0.0692 |
109
+ | 60.5259 | 56.0 | 28280 | 28.6182 | 0.0656 |
110
+ | 61.5349 | 57.0 | 28785 | 27.0813 | 0.0617 |
111
+ | 59.6488 | 58.0 | 29290 | 23.4508 | 0.0565 |
112
+ | 59.56 | 59.0 | 29795 | 29.5856 | 0.0575 |
113
+ | 58.864 | 60.0 | 30300 | 22.0317 | 0.0531 |
114
 
115
 
116
  ### Framework versions
config.json CHANGED
@@ -9,7 +9,7 @@
9
  "architectures": [
10
  "Wav2Vec2ForCTC"
11
  ],
12
- "attention_dropout": 0.1,
13
  "bos_token_id": 1,
14
  "classifier_proj_size": 256,
15
  "codevector_dim": 256,
@@ -50,11 +50,11 @@
50
  "feat_extract_activation": "gelu",
51
  "feat_extract_dropout": 0.0,
52
  "feat_extract_norm": "group",
53
- "feat_proj_dropout": 0.15,
54
  "feat_quantizer_dropout": 0.0,
55
- "final_dropout": 0.1,
56
  "hidden_act": "gelu",
57
- "hidden_dropout": 0.1,
58
  "hidden_dropout_prob": 0.1,
59
  "hidden_size": 1024,
60
  "initializer_range": 0.02,
 
9
  "architectures": [
10
  "Wav2Vec2ForCTC"
11
  ],
12
+ "attention_dropout": 0.15,
13
  "bos_token_id": 1,
14
  "classifier_proj_size": 256,
15
  "codevector_dim": 256,
 
50
  "feat_extract_activation": "gelu",
51
  "feat_extract_dropout": 0.0,
52
  "feat_extract_norm": "group",
53
+ "feat_proj_dropout": 0.2,
54
  "feat_quantizer_dropout": 0.0,
55
+ "final_dropout": 0.15,
56
  "hidden_act": "gelu",
57
+ "hidden_dropout": 0.15,
58
  "hidden_dropout_prob": 0.1,
59
  "hidden_size": 1024,
60
  "initializer_range": 0.02,
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2697b097013f682f8cd1e0f08264f66eca8babc158c67656d80e4295dbb0f059
3
  size 1261897368
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e186bfc7b9a89f2b578b81e13de26edaaf81350f8208394fedc55303ef463591
3
  size 1261897368
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c3933599f5e4a5a9973599931182d82005e10cb49b72302ce0b90417f54bbfe1
3
  size 5240
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6ece9d55fd867ab888f57a748a5372a385c13395916b233f23479cc1fc0bc414
3
  size 5240