Hugging Face's logo

miosipof
/
whisper_medium_BB_and_EC_v2.2

PEFT
Safetensors
Italian
Generated from Trainer
Eval Results
Model card Files Community
miosipof commited on
Commit
7eeb259
·
verified ·
1 Parent(s): 5742463

End of training

Browse files
Files changed (3) hide show
  1. README.md +9 -10
  2. adapter_model.safetensors +1 -1
  3. training_args.bin +1 -1
README.md CHANGED
@@ -23,7 +23,7 @@ model-index:
23
  args: default
24
  metrics:
25
  - type: wer
26
- value: 36.80297397769517
27
  name: Wer
28
  ---
29
 
@@ -34,8 +34,8 @@ should probably proofread and complete it, then remove this comment. -->
34
 
35
  This model is a fine-tuned version of [b-brave/asr_double_training_15-10-2024_merged](https://huggingface.co/b-brave/asr_double_training_15-10-2024_merged) on the ASR_BB_and_EC dataset.
36
  It achieves the following results on the evaluation set:
37
- - Loss: 0.4435
38
- - Wer: 36.8030
39
 
40
  ## Model description
41
 
@@ -54,7 +54,7 @@ More information needed
54
  ### Training hyperparameters
55
 
56
  The following hyperparameters were used during training:
57
- - learning_rate: 3e-06
58
  - train_batch_size: 16
59
  - eval_batch_size: 16
60
  - seed: 42
@@ -62,18 +62,17 @@ The following hyperparameters were used during training:
62
  - total_train_batch_size: 32
63
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
64
  - lr_scheduler_type: constant
65
- - lr_scheduler_warmup_steps: 50
66
- - num_epochs: 4
67
  - mixed_precision_training: Native AMP
68
 
69
  ### Training results
70
 
71
  | Training Loss | Epoch | Step | Validation Loss | Wer |
72
  |:-------------:|:------:|:----:|:---------------:|:-------:|
73
- | 1.332 | 0.9852 | 100 | 0.4732 | 35.5638 |
74
- | 1.241 | 1.9704 | 200 | 0.4624 | 34.9442 |
75
- | 1.17 | 2.9557 | 300 | 0.4524 | 36.0595 |
76
- | 1.1215 | 3.9409 | 400 | 0.4435 | 36.8030 |
77
 
78
 
79
  ### Framework versions
 
23
  args: default
24
  metrics:
25
  - type: wer
26
+ value: 35.5638166047088
27
  name: Wer
28
  ---
29
 
 
34
 
35
  This model is a fine-tuned version of [b-brave/asr_double_training_15-10-2024_merged](https://huggingface.co/b-brave/asr_double_training_15-10-2024_merged) on the ASR_BB_and_EC dataset.
36
  It achieves the following results on the evaluation set:
37
+ - Loss: 0.4733
38
+ - Wer: 35.5638
39
 
40
  ## Model description
41
 
 
54
  ### Training hyperparameters
55
 
56
  The following hyperparameters were used during training:
57
+ - learning_rate: 1e-06
58
  - train_batch_size: 16
59
  - eval_batch_size: 16
60
  - seed: 42
 
62
  - total_train_batch_size: 32
63
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
64
  - lr_scheduler_type: constant
65
+ - lr_scheduler_warmup_steps: 10
66
+ - num_epochs: 3
67
  - mixed_precision_training: Native AMP
68
 
69
  ### Training results
70
 
71
  | Training Loss | Epoch | Step | Validation Loss | Wer |
72
  |:-------------:|:------:|:----:|:---------------:|:-------:|
73
+ | 1.3876 | 0.9852 | 100 | 0.4835 | 36.3073 |
74
+ | 1.3282 | 1.9704 | 200 | 0.4776 | 36.1834 |
75
+ | 1.2853 | 2.9557 | 300 | 0.4733 | 35.5638 |
 
76
 
77
 
78
  ### Framework versions
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ac188f6e9543f0aa612a9005cd4ca5d8ae755877d80f7f6bd04fc453752555f9
3
  size 37789960
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d3d2b85482420bc00dab401f4661faa90b66725977c43830b2169e80148da91
3
  size 37789960
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:667717a6276888986d0ad1b3ea893bf6af9957459fd9e9d8e7e666ac7758a1d5
3
  size 5368
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d24895e77f4add2d62f40cdbd9f7b0f54486223e759c30eb1e5d1bdaaf651cc8
3
  size 5368