Hugging Face's logo

luigimontaleone
/
asr_luigi_test

PEFT
TensorBoard
Safetensors
Italian
Generated from Trainer
Model card Files Metrics Community
luigimontaleone commited on
Commit
4fc7179
•
1 Parent(s): 13fad4a

End of training

Browse files
Files changed (5) hide show
  1. README.md +14 -9
  2. adapter_model.safetensors +1 -1
  3. runs/Jul30_18-22-25_gigio-desktop/events.out.tfevents.1722356549.gigio-desktop.29174.0 +3 -0
  4. runs/Jul30_18-23-24_gigio-desktop/events.out.tfevents.1722356608.gigio-desktop.29520.0 +3 -0
  5. training_args.bin +1 -1
README.md CHANGED
@@ -24,7 +24,7 @@ model-index:
24
  args: default
25
  metrics:
26
  - type: wer
27
- value: 100.0
28
  name: Wer
29
  ---
30
 
@@ -35,8 +35,8 @@ should probably proofread and complete it, then remove this comment. -->
35
 
36
  This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the b-brave/speech_disorders_voice dataset.
37
  It achieves the following results on the evaluation set:
38
- - Loss: 0.1865
39
- - Wer: 100.0
40
 
41
  ## Model description
42
 
@@ -64,18 +64,23 @@ The following hyperparameters were used during training:
64
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
65
  - lr_scheduler_type: linear
66
  - lr_scheduler_warmup_steps: 48
67
- - training_steps: 256
68
  - mixed_precision_training: Native AMP
69
 
70
  ### Training results
71
 
72
  | Training Loss | Epoch | Step | Validation Loss | Wer |
73
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
74
- | 1.423 | 0.375 | 48 | 0.3535 | 106.0166 |
75
- | 0.3121 | 0.75 | 96 | 0.2261 | 100.0 |
76
- | 0.2534 | 1.125 | 144 | 0.1995 | 100.0 |
77
- | 0.147 | 1.5 | 192 | 0.1973 | 99.7925 |
78
- | 0.1161 | 1.875 | 240 | 0.1865 | 100.0 |
 
 
 
 
 
79
 
80
 
81
  ### Framework versions
 
24
  args: default
25
  metrics:
26
  - type: wer
27
+ value: 420.2554744525547
28
  name: Wer
29
  ---
30
 
 
35
 
36
  This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the b-brave/speech_disorders_voice dataset.
37
  It achieves the following results on the evaluation set:
38
+ - Loss: 0.2259
39
+ - Wer: 420.2555
40
 
41
  ## Model description
42
 
 
64
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
65
  - lr_scheduler_type: linear
66
  - lr_scheduler_warmup_steps: 48
67
+ - num_epochs: 4
68
  - mixed_precision_training: Native AMP
69
 
70
  ### Training results
71
 
72
  | Training Loss | Epoch | Step | Validation Loss | Wer |
73
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
74
+ | 1.4274 | 0.375 | 48 | 0.4628 | 16.9708 |
75
+ | 0.3169 | 0.75 | 96 | 0.2702 | 56.7518 |
76
+ | 0.25 | 1.125 | 144 | 0.2318 | 25.7299 |
77
+ | 0.1453 | 1.5 | 192 | 0.2324 | 195.0730 |
78
+ | 0.1573 | 1.875 | 240 | 0.2431 | 98.7226 |
79
+ | 0.0813 | 2.25 | 288 | 0.2405 | 237.5912 |
80
+ | 0.0494 | 2.625 | 336 | 0.2150 | 279.9270 |
81
+ | 0.0677 | 3.0 | 384 | 0.2341 | 260.0365 |
82
+ | 0.0303 | 3.375 | 432 | 0.2274 | 372.4453 |
83
+ | 0.0241 | 3.75 | 480 | 0.2259 | 420.2555 |
84
 
85
 
86
  ### Framework versions
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5ea98cfa5c74ce2f8e2549afe7524f1935aca16d60c62d65a6f1407bc989f437
3
  size 62969640
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b06065a232ada2f1d91679fe339ae1b28a47c4e4aea846d3941a03d0c8ddecae
3
  size 62969640
runs/Jul30_18-22-25_gigio-desktop/events.out.tfevents.1722356549.gigio-desktop.29174.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:593b7b1f46a050df92534b2d86aa3752d0399044bdff511669a1764622700cc9
3
+ size 5795
runs/Jul30_18-23-24_gigio-desktop/events.out.tfevents.1722356608.gigio-desktop.29520.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a978e0b078c5321183369bfe6d87794885dc4d276bded30b71e2b61bd7546454
3
+ size 11419
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:97518e72887d592db266e32be8c538a0b955efc0037fda1bfe7fec5eed5e72c0
3
  size 5304
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d610ece67f421320b2942dfcc7584f8338a44acfeb34b19f2a33f03338b94c3
3
  size 5304