eai6 commited on
Commit
0525178
1 Parent(s): ada3e0d

End of training

Browse files
README.md CHANGED
@@ -4,39 +4,37 @@ language:
4
  license: apache-2.0
5
  base_model: openai/whisper-tiny.en
6
  tags:
7
- - nyansapo_ai-asr-leaderboard
8
  - generated_from_trainer
9
  datasets:
10
- - NyansapoAI/azure-dataset
11
  metrics:
12
  - wer
13
  model-index:
14
- - name: whisper-tiny.en
15
  results:
16
  - task:
17
  name: Automatic Speech Recognition
18
  type: automatic-speech-recognition
19
  dataset:
20
- name: Azure-dataset
21
- type: NyansapoAI/azure-dataset
22
- config: default
23
- split: test
24
- args: 'split: test'
25
  metrics:
26
  - name: Wer
27
  type: wer
28
- value: 1.8181818181818181
29
  ---
30
 
31
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
32
  should probably proofread and complete it, then remove this comment. -->
33
 
34
- # whisper-tiny.en
35
 
36
- This model is a fine-tuned version of [openai/whisper-tiny.en](https://huggingface.co/openai/whisper-tiny.en) on the Azure-dataset dataset.
37
  It achieves the following results on the evaluation set:
38
- - Loss: 0.0219
39
- - Wer: 1.8182
40
 
41
  ## Model description
42
 
@@ -61,20 +59,22 @@ The following hyperparameters were used during training:
61
  - seed: 42
62
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
63
  - lr_scheduler_type: linear
64
- - lr_scheduler_warmup_steps: 500
65
- - training_steps: 2000
66
 
67
  ### Training results
68
 
69
- | Training Loss | Epoch | Step | Validation Loss | Wer |
70
- |:-------------:|:-----:|:----:|:---------------:|:------:|
71
- | 0.0756 | 6.21 | 1000 | 0.0262 | 2.1212 |
72
- | 0.045 | 12.42 | 2000 | 0.0219 | 1.8182 |
 
 
73
 
74
 
75
  ### Framework versions
76
 
77
- - Transformers 4.33.0.dev0
78
- - Pytorch 2.0.1
79
- - Datasets 2.14.4
80
- - Tokenizers 0.13.3
 
4
  license: apache-2.0
5
  base_model: openai/whisper-tiny.en
6
  tags:
7
+ - hf-asr-leaderboard
8
  - generated_from_trainer
9
  datasets:
10
+ - eai6/bungoma_training
11
  metrics:
12
  - wer
13
  model-index:
14
+ - name: Whisper tiny
15
  results:
16
  - task:
17
  name: Automatic Speech Recognition
18
  type: automatic-speech-recognition
19
  dataset:
20
+ name: Nyansapo AI Dataset
21
+ type: eai6/bungoma_training
22
+ args: 'config: en, split: test'
 
 
23
  metrics:
24
  - name: Wer
25
  type: wer
26
+ value: 152.94117647058823
27
  ---
28
 
29
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
30
  should probably proofread and complete it, then remove this comment. -->
31
 
32
+ # Whisper tiny
33
 
34
+ This model is a fine-tuned version of [openai/whisper-tiny.en](https://huggingface.co/openai/whisper-tiny.en) on the Nyansapo AI Dataset dataset.
35
  It achieves the following results on the evaluation set:
36
+ - Loss: 0.4537
37
+ - Wer: 152.9412
38
 
39
  ## Model description
40
 
 
59
  - seed: 42
60
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
61
  - lr_scheduler_type: linear
62
+ - lr_scheduler_warmup_steps: 250
63
+ - training_steps: 1000
64
 
65
  ### Training results
66
 
67
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
68
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|
69
+ | 0.3783 | 13.89 | 250 | 0.6768 | 80.6228 |
70
+ | 0.0434 | 27.78 | 500 | 0.4529 | 92.3875 |
71
+ | 0.0395 | 41.67 | 750 | 0.4625 | 73.3564 |
72
+ | 0.036 | 55.56 | 1000 | 0.4537 | 152.9412 |
73
 
74
 
75
  ### Framework versions
76
 
77
+ - Transformers 4.39.1
78
+ - Pytorch 2.2.1
79
+ - Datasets 2.18.0
80
+ - Tokenizers 0.15.2
generation_config.json CHANGED
@@ -47,10 +47,11 @@
47
  ]
48
  ],
49
  "is_multilingual": false,
50
- "max_initial_timestamp_index": 1,
51
  "max_length": 448,
52
  "no_timestamps_token_id": 50362,
53
  "pad_token_id": 50256,
 
54
  "return_timestamps": false,
55
  "suppress_tokens": [
56
  1,
@@ -144,5 +145,5 @@
144
  50360,
145
  50361
146
  ],
147
- "transformers_version": "4.33.0.dev0"
148
  }
 
47
  ]
48
  ],
49
  "is_multilingual": false,
50
+ "max_initial_timestamp_index": 50,
51
  "max_length": 448,
52
  "no_timestamps_token_id": 50362,
53
  "pad_token_id": 50256,
54
+ "prev_sot_token_id": 50360,
55
  "return_timestamps": false,
56
  "suppress_tokens": [
57
  1,
 
145
  50360,
146
  50361
147
  ],
148
+ "transformers_version": "4.39.1"
149
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:748b0b0fd88ea7e4683de92551b4a7bd1119736b8e16e599822c312ff9a8660c
3
  size 151060136
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c17d6963510b98c8e70f94f7343b27cf1a3520b58d9e8c9f89e9ce731367ddd9
3
  size 151060136
runs/Mar23_16-21-54_Edwards-MacBook-Pro-2.local/events.out.tfevents.1711225316.Edwards-MacBook-Pro-2.local.80784.1 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3d0c63887fe04b14e1bd7955e022e7a6e6aeef524399b15b2bbe44708b041889
3
- size 15131
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b701c8bc95e9deb59834fddb2a696eb0bea40b98acd311cf312526aaa4190f8
3
+ size 15485