Hugging Face's logo

JacobLinCool
/
wft-test-model-merged

Automatic Speech Recognition
Transformers
Safetensors
English
whisper
wft
audio
speech
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
JacobLinCool commited on
Commit
3b11400
·
verified ·
1 Parent(s): 2a764ea

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +22 -13
README.md CHANGED
@@ -26,7 +26,7 @@ model-index:
26
  type: hf-internal-testing/librispeech_asr_dummy
27
  metrics:
28
  - type: wer
29
- value: 59.01234567901235
30
  name: Wer
31
  ---
32
 
@@ -37,9 +37,12 @@ should probably proofread and complete it, then remove this comment. -->
37
 
38
  This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the hf-internal-testing/librispeech_asr_dummy dataset.
39
  It achieves the following results on the evaluation set:
40
- - Loss: 1.0778
41
- - Wer: 59.0123
42
- - Cer: 23.9022
 
 
 
43
 
44
  ## Model description
45
 
@@ -60,21 +63,27 @@ More information needed
60
  The following hyperparameters were used during training:
61
  - learning_rate: 0.0005
62
  - train_batch_size: 4
63
- - eval_batch_size: 4
64
  - seed: 42
65
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
66
  - lr_scheduler_type: linear
67
- - num_epochs: 5
 
68
 
69
  ### Training results
70
 
71
- | Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
72
- |:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|
73
- | 1.8837 | 1.0 | 8 | 1.5660 | 67.6543 | 28.1575 |
74
- | 0.8916 | 2.0 | 16 | 1.3180 | 58.0247 | 23.3590 |
75
- | 0.3949 | 3.0 | 24 | 1.1546 | 60.7407 | 25.2603 |
76
- | 0.2231 | 4.0 | 32 | 1.0957 | 59.7531 | 24.7623 |
77
- | 0.17 | 5.0 | 40 | 1.0778 | 59.0123 | 23.9022 |
 
 
 
 
 
78
 
79
 
80
  ### Framework versions
 
26
  type: hf-internal-testing/librispeech_asr_dummy
27
  metrics:
28
  - type: wer
29
+ value: 4.724409448818897
30
  name: Wer
31
  ---
32
 
 
37
 
38
  This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the hf-internal-testing/librispeech_asr_dummy dataset.
39
  It achieves the following results on the evaluation set:
40
+ - Loss: 0.1218
41
+ - Wer: 4.7244
42
+ - Cer: 100.7102
43
+ - Decode Time: 0.5595
44
+ - Wer Time: 0.0060
45
+ - Cer Time: 0.0027
46
 
47
  ## Model description
48
 
 
63
  The following hyperparameters were used during training:
64
  - learning_rate: 0.0005
65
  - train_batch_size: 4
66
+ - eval_batch_size: 8
67
  - seed: 42
68
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
69
  - lr_scheduler_type: linear
70
+ - lr_scheduler_warmup_steps: 50
71
+ - training_steps: 100
72
 
73
  ### Training results
74
 
75
+ | Training Loss | Epoch | Step | Validation Loss | Wer | Cer | Decode Time | Wer Time | Cer Time |
76
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|:--------:|:-----------:|:--------:|:--------:|
77
+ | 2.4038 | 0.1 | 10 | 1.9871 | 299.6063 | 116.5483 | 0.5542 | 0.0112 | 0.0038 |
78
+ | 1.2202 | 1.01 | 20 | 1.1610 | 172.0472 | 91.6903 | 0.5178 | 0.0055 | 0.0028 |
79
+ | 0.8765 | 1.11 | 30 | 0.8079 | 31.8898 | 56.1080 | 0.4854 | 0.0051 | 0.0024 |
80
+ | 0.4415 | 2.02 | 40 | 0.6279 | 25.9843 | 82.5284 | 0.5218 | 0.0060 | 0.0028 |
81
+ | 0.4307 | 2.12 | 50 | 0.4509 | 16.9291 | 98.2955 | 0.5377 | 0.0056 | 0.0030 |
82
+ | 0.2363 | 3.03 | 60 | 0.2952 | 9.4488 | 102.9119 | 0.5310 | 0.0048 | 0.0027 |
83
+ | 0.2245 | 3.13 | 70 | 0.2046 | 7.4803 | 96.3778 | 0.5429 | 0.0056 | 0.0029 |
84
+ | 0.1053 | 4.04 | 80 | 0.1700 | 5.5118 | 96.4489 | 0.5306 | 0.0067 | 0.0028 |
85
+ | 0.0731 | 4.14 | 90 | 0.1320 | 4.7244 | 97.1591 | 0.5509 | 0.0050 | 0.0026 |
86
+ | 0.0782 | 5.05 | 100 | 0.1218 | 4.7244 | 100.7102 | 0.5595 | 0.0060 | 0.0027 |
87
 
88
 
89
  ### Framework versions