Hugging Face's logo

asserr
/
speecht5-tunis_finalll

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
speecht5
Generated from Trainer
Model card Files
xet
Metrics Community
asserr commited on
Commit
7b15550
·
verified ·
1 Parent(s): fb6a8eb

Model save

Browse files
Files changed (5) hide show
  1. README.md +23 -18
  2. final_model/model.safetensors +1 -1
  3. final_model/training_args.bin +1 -1
  4. model.safetensors +1 -1
  5. runs/Jan24_23-42-43_b212ad9c366f/events.out.tfevents.1769303596.b212ad9c366f.38.4 +3 -0
README.md CHANGED
@@ -18,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
18
 
19
  This model is a fine-tuned version of [microsoft/speecht5_asr](https://huggingface.co/microsoft/speecht5_asr) on the None dataset.
20
  It achieves the following results on the evaluation set:
21
- - Loss: 0.1951
22
- - Wer Ortho: 63.3333
23
- - Wer: 62.4672
24
 
25
  ## Model description
26
 
@@ -39,7 +39,7 @@ More information needed
39
  ### Training hyperparameters
40
 
41
  The following hyperparameters were used during training:
42
- - learning_rate: 1e-05
43
  - train_batch_size: 2
44
  - eval_batch_size: 8
45
  - seed: 42
@@ -47,24 +47,29 @@ The following hyperparameters were used during training:
47
  - total_train_batch_size: 16
48
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
49
  - lr_scheduler_type: cosine
50
- - lr_scheduler_warmup_steps: 100
51
- - training_steps: 1000
52
  - mixed_precision_training: Native AMP
53
 
54
  ### Training results
55
 
56
- | Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer |
57
- |:-------------:|:------:|:----:|:---------------:|:---------:|:--------:|
58
- | 1.7381 | 0.3731 | 100 | 1.4437 | 371.3889 | 213.9108 |
59
- | 0.8437 | 0.7463 | 200 | 0.5686 | 80.5556 | 81.6273 |
60
- | 0.4461 | 1.1194 | 300 | 0.3668 | 76.1111 | 77.4278 |
61
- | 0.3753 | 1.4925 | 400 | 0.2760 | 72.7778 | 74.0157 |
62
- | 0.3416 | 1.8657 | 500 | 0.2392 | 80.8333 | 84.7769 |
63
- | 0.2656 | 2.2388 | 600 | 0.2138 | 67.7778 | 67.9790 |
64
- | 0.2706 | 2.6119 | 700 | 0.2085 | 74.7222 | 77.1654 |
65
- | 0.2509 | 2.9851 | 800 | 0.1995 | 63.0556 | 62.2047 |
66
- | 0.2314 | 3.3582 | 900 | 0.1949 | 62.5 | 61.6798 |
67
- | 0.2806 | 3.7313 | 1000 | 0.1951 | 63.3333 | 62.4672 |
 
 
 
 
 
68
 
69
 
70
  ### Framework versions
 
18
 
19
  This model is a fine-tuned version of [microsoft/speecht5_asr](https://huggingface.co/microsoft/speecht5_asr) on the None dataset.
20
  It achieves the following results on the evaluation set:
21
+ - Loss: 0.2003
22
+ - Wer Ortho: 62.9526
23
+ - Wer: 59.7855
24
 
25
  ## Model description
26
 
 
39
  ### Training hyperparameters
40
 
41
  The following hyperparameters were used during training:
42
+ - learning_rate: 5e-06
43
  - train_batch_size: 2
44
  - eval_batch_size: 8
45
  - seed: 42
 
47
  - total_train_batch_size: 16
48
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
49
  - lr_scheduler_type: cosine
50
+ - lr_scheduler_warmup_steps: 50
51
+ - training_steps: 1500
52
  - mixed_precision_training: Native AMP
53
 
54
  ### Training results
55
 
56
+ | Training Loss | Epoch | Step | Validation Loss | Wer | Wer Ortho |
57
+ |:-------------:|:------:|:----:|:---------------:|:--------:|:---------:|
58
+ | 1.7381 | 0.3731 | 100 | 1.4437 | 213.9108 | 371.3889 |
59
+ | 0.8437 | 0.7463 | 200 | 0.5686 | 81.6273 | 80.5556 |
60
+ | 0.4461 | 1.1194 | 300 | 0.3668 | 77.4278 | 76.1111 |
61
+ | 0.3753 | 1.4925 | 400 | 0.2760 | 74.0157 | 72.7778 |
62
+ | 0.3416 | 1.8657 | 500 | 0.2392 | 84.7769 | 80.8333 |
63
+ | 0.2656 | 2.2388 | 600 | 0.2138 | 67.9790 | 67.7778 |
64
+ | 0.2706 | 2.6119 | 700 | 0.2085 | 77.1654 | 74.7222 |
65
+ | 0.2509 | 2.9851 | 800 | 0.1995 | 62.2047 | 63.0556 |
66
+ | 0.2314 | 3.3582 | 900 | 0.1949 | 61.6798 | 62.5 |
67
+ | 0.2806 | 3.7313 | 1000 | 0.1951 | 62.4672 | 63.3333 |
68
+ | 0.2254 | 4.1045 | 1100 | 0.1912 | 68.6111 | 69.2913 |
69
+ | 0.2674 | 4.4776 | 1200 | 0.1863 | 68.6111 | 69.8163 |
70
+ | 0.301 | 4.8507 | 1300 | 0.1862 | 67.5 | 67.9790 |
71
+ | 0.2354 | 5.2239 | 1400 | 0.1850 | 61.1111 | 59.8425 |
72
+ | 0.2349 | 5.5970 | 1500 | 0.1851 | 67.2222 | 67.7165 |
73
 
74
 
75
  ### Framework versions
final_model/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3dc61cb1446fcbb6002384682ffdd95b1f127033e00738521480a2388356e233
3
  size 604711248
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e97f0f2105b52dd2fb6ccb67620876f0077649b8cfc24a74306dada9954c2c30
3
  size 604711248
final_model/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e1870232b40870feeeb9e828e86f936d1b7f61afd2c72c99e07956df5e914e39
3
  size 5432
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e0a2743349e0bef163b7e3dc33ca78b1ee4ab8fa27db594b0364ec3235b2d068
3
  size 5432
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:545df90da3bca7fa4d96468a34d09d50aa7365598f7822a78be7c1faf444f325
3
  size 604711248
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e97f0f2105b52dd2fb6ccb67620876f0077649b8cfc24a74306dada9954c2c30
3
  size 604711248
runs/Jan24_23-42-43_b212ad9c366f/events.out.tfevents.1769303596.b212ad9c366f.38.4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c66d60c3cfc5c2e316b508605b82f4b21680db3cc1aef03f4ca396150a3d7e8c
3
+ size 459