Hugging Face's logo

Aysha630
/
whisper-large-v3-MH-fine-tuned

PEFT
TensorBoard
Safetensors
Generated from Trainer
Model card Files Metrics Community
Aysha630 commited on
Commit
37d84f5
•
1 Parent(s): 0f25b88

End of training

Browse files
Files changed (5) hide show
  1. README.md +23 -28
  2. adapter_config.json +2 -2
  3. adapter_model.safetensors +1 -1
  4. runs/May19_11-28-54_d247403fb5a3/events.out.tfevents.1716118142.d247403fb5a3.34.0 +3 -0
  5. training_args.bin +1 -1
README.md CHANGED
@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
16
 
17
  This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the None dataset.
18
  It achieves the following results on the evaluation set:
19
- - Loss: 0.6328
20
 
21
  ## Model description
22
 
@@ -41,39 +41,34 @@ The following hyperparameters were used during training:
41
  - seed: 42
42
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
43
  - lr_scheduler_type: linear
44
- - lr_scheduler_warmup_steps: 18
45
- - num_epochs: 25
46
  - mixed_precision_training: Native AMP
47
 
48
  ### Training results
49
 
50
  | Training Loss | Epoch | Step | Validation Loss |
51
  |:-------------:|:-----:|:----:|:---------------:|
52
- | 0.983 | 1.0 | 7 | 0.6322 |
53
- | 0.6898 | 2.0 | 14 | 0.5551 |
54
- | 0.3441 | 3.0 | 21 | 0.5087 |
55
- | 0.2457 | 4.0 | 28 | 0.4763 |
56
- | 0.1372 | 5.0 | 35 | 0.4975 |
57
- | 0.109 | 6.0 | 42 | 0.4832 |
58
- | 0.0663 | 7.0 | 49 | 0.4969 |
59
- | 0.0452 | 8.0 | 56 | 0.5020 |
60
- | 0.0322 | 9.0 | 63 | 0.5067 |
61
- | 0.0218 | 10.0 | 70 | 0.5277 |
62
- | 0.0147 | 11.0 | 77 | 0.5655 |
63
- | 0.0157 | 12.0 | 84 | 0.5682 |
64
- | 0.0214 | 13.0 | 91 | 0.6323 |
65
- | 0.0091 | 14.0 | 98 | 0.6149 |
66
- | 0.0105 | 15.0 | 105 | 0.5824 |
67
- | 0.0111 | 16.0 | 112 | 0.5962 |
68
- | 0.011 | 17.0 | 119 | 0.5991 |
69
- | 0.0078 | 18.0 | 126 | 0.6121 |
70
- | 0.0065 | 19.0 | 133 | 0.6199 |
71
- | 0.0054 | 20.0 | 140 | 0.6285 |
72
- | 0.0041 | 21.0 | 147 | 0.6269 |
73
- | 0.0051 | 22.0 | 154 | 0.6299 |
74
- | 0.003 | 23.0 | 161 | 0.6244 |
75
- | 0.0025 | 24.0 | 168 | 0.6320 |
76
- | 0.0035 | 25.0 | 175 | 0.6328 |
77
 
78
 
79
  ### Framework versions
 
16
 
17
  This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the None dataset.
18
  It achieves the following results on the evaluation set:
19
+ - Loss: 0.7223
20
 
21
  ## Model description
22
 
 
41
  - seed: 42
42
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
43
  - lr_scheduler_type: linear
44
+ - lr_scheduler_warmup_steps: 14
45
+ - num_epochs: 20
46
  - mixed_precision_training: Native AMP
47
 
48
  ### Training results
49
 
50
  | Training Loss | Epoch | Step | Validation Loss |
51
  |:-------------:|:-----:|:----:|:---------------:|
52
+ | 0.8453 | 1.0 | 7 | 0.6698 |
53
+ | 0.5918 | 2.0 | 14 | 0.6178 |
54
+ | 0.3676 | 3.0 | 21 | 0.6309 |
55
+ | 0.2071 | 4.0 | 28 | 0.5503 |
56
+ | 0.0815 | 5.0 | 35 | 0.5673 |
57
+ | 0.0716 | 6.0 | 42 | 0.6529 |
58
+ | 0.0498 | 7.0 | 49 | 0.6263 |
59
+ | 0.0385 | 8.0 | 56 | 0.6218 |
60
+ | 0.0279 | 9.0 | 63 | 0.6489 |
61
+ | 0.0223 | 10.0 | 70 | 0.6558 |
62
+ | 0.0156 | 11.0 | 77 | 0.7044 |
63
+ | 0.012 | 12.0 | 84 | 0.6791 |
64
+ | 0.0101 | 13.0 | 91 | 0.7385 |
65
+ | 0.3867 | 14.0 | 98 | 0.7011 |
66
+ | 0.0131 | 15.0 | 105 | 0.7007 |
67
+ | 0.0117 | 16.0 | 112 | 0.7196 |
68
+ | 0.0107 | 17.0 | 119 | 0.7192 |
69
+ | 0.0075 | 18.0 | 126 | 0.7183 |
70
+ | 0.0081 | 19.0 | 133 | 0.7164 |
71
+ | 0.0067 | 20.0 | 140 | 0.7223 |
 
 
 
 
 
72
 
73
 
74
  ### Framework versions
adapter_config.json CHANGED
@@ -23,8 +23,8 @@
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
- "q_proj",
27
- "v_proj"
28
  ],
29
  "task_type": null,
30
  "use_dora": false,
 
23
  "rank_pattern": {},
24
  "revision": null,
25
  "target_modules": [
26
+ "v_proj",
27
+ "q_proj"
28
  ],
29
  "task_type": null,
30
  "use_dora": false,
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c40e0bb5892b210b63c739223e07c7d14ed6a8e9df8cb6b18a0bbdbde1e3160b
3
  size 62969640
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c76b905ec786e43258477324849ee297a86f6d9cc571373a1a4fe84ec3484c0d
3
  size 62969640
runs/May19_11-28-54_d247403fb5a3/events.out.tfevents.1716118142.d247403fb5a3.34.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:075475ee62c3f913bf7ac907041203ab8def5a561766ce728dcd0bb51049ae6c
3
+ size 21245
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2c5150d9179768eccaf7272af9e0aad4f9f7068863c3143fc20a5b295f6f04e5
3
  size 5176
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f16e7be59e49ea3773d8b2f96f0fb85ee6d72a92d39aeb29c24f49f80e2c1491
3
  size 5176