Hugging Face's logo

UniversalAlgorithmic
/
wav2vec2-base-ft-keyword-spotting

Audio Classification
Transformers
TensorBoard
Safetensors
wav2vec2
Generated from Trainer
Eval Results (legacy)
Model card Files
xet
Metrics Community
Zheng Li commited on
Commit
cde85ba
·
verified ·
1 Parent(s): 56191df

Model save

Browse files
Files changed (3) hide show
  1. README.md +14 -17
  2. model.safetensors +1 -1
  3. runs/May15_02-50-35_cs-Precision-7960-Tower/events.out.tfevents.1747291839.cs-Precision-7960-Tower.145708.0 +2 -2
README.md CHANGED
@@ -3,7 +3,6 @@ library_name: transformers
3
  license: apache-2.0
4
  base_model: facebook/wav2vec2-base
5
  tags:
6
- - audio-classification
7
  - generated_from_trainer
8
  datasets:
9
  - superb
@@ -34,7 +33,7 @@ should probably proofread and complete it, then remove this comment. -->
34
 
35
  This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the superb dataset.
36
  It achieves the following results on the evaluation set:
37
- - Loss: 0.1243
38
  - Accuracy: 0.9812
39
 
40
  ## Model description
@@ -55,31 +54,29 @@ More information needed
55
 
56
  The following hyperparameters were used during training:
57
  - learning_rate: 3e-05
58
- - train_batch_size: 128
59
  - eval_batch_size: 32
60
  - seed: 0
61
  - gradient_accumulation_steps: 4
62
- - total_train_batch_size: 512
63
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
64
  - lr_scheduler_type: linear
65
  - lr_scheduler_warmup_ratio: 0.1
66
- - num_epochs: 10.0
67
  - mixed_precision_training: Native AMP
68
 
69
  ### Training results
70
 
71
- | Training Loss | Epoch | Step | Validation Loss | Accuracy |
72
- |:-------------:|:-----:|:----:|:---------------:|:--------:|
73
- | 1.8178 | 1.0 | 100 | 1.7386 | 0.6209 |
74
- | 1.3138 | 2.0 | 200 | 1.1779 | 0.6511 |
75
- | 0.9632 | 3.0 | 300 | 0.8326 | 0.8679 |
76
- | 0.499 | 4.0 | 400 | 0.3697 | 0.9725 |
77
- | 0.3191 | 5.0 | 500 | 0.2240 | 0.9760 |
78
- | 0.242 | 6.0 | 600 | 0.1709 | 0.9793 |
79
- | 0.2144 | 7.0 | 700 | 0.1460 | 0.9806 |
80
- | 0.1977 | 8.0 | 800 | 0.1344 | 0.9800 |
81
- | 0.1742 | 9.0 | 900 | 0.1282 | 0.9797 |
82
- | 0.1867 | 10.0 | 1000 | 0.1243 | 0.9812 |
83
 
84
 
85
  ### Framework versions
 
3
  license: apache-2.0
4
  base_model: facebook/wav2vec2-base
5
  tags:
 
6
  - generated_from_trainer
7
  datasets:
8
  - superb
 
33
 
34
  This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the superb dataset.
35
  It achieves the following results on the evaluation set:
36
+ - Loss: 0.0896
37
  - Accuracy: 0.9812
38
 
39
  ## Model description
 
54
 
55
  The following hyperparameters were used during training:
56
  - learning_rate: 3e-05
57
+ - train_batch_size: 48
58
  - eval_batch_size: 32
59
  - seed: 0
60
  - gradient_accumulation_steps: 4
61
+ - total_train_batch_size: 192
62
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
63
  - lr_scheduler_type: linear
64
  - lr_scheduler_warmup_ratio: 0.1
65
+ - num_epochs: 8.0
66
  - mixed_precision_training: Native AMP
67
 
68
  ### Training results
69
 
70
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
71
+ |:-------------:|:------:|:----:|:---------------:|:--------:|
72
+ | 1.3624 | 1.0 | 267 | 1.1959 | 0.6546 |
73
+ | 0.3854 | 2.0 | 534 | 0.2675 | 0.9734 |
74
+ | 0.2473 | 3.0 | 801 | 0.1461 | 0.9768 |
75
+ | 0.1997 | 4.0 | 1068 | 0.1088 | 0.9804 |
76
+ | 0.1723 | 5.0 | 1335 | 0.0954 | 0.9826 |
77
+ | 0.1442 | 6.0 | 1602 | 0.0927 | 0.9813 |
78
+ | 0.1397 | 7.0 | 1869 | 0.0892 | 0.9812 |
79
+ | 0.1368 | 7.9728 | 2128 | 0.0896 | 0.9812 |
 
 
80
 
81
 
82
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:35d82637ef5fb6e79ddad46224e9089ddb69ec6072b5909195be196ba6d6737e
3
  size 385400352
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d6e4f1c85d883f3e41ebfab4cd7752ab2e6d6b968b847795be22e1e0662657a3
3
  size 385400352
runs/May15_02-50-35_cs-Precision-7960-Tower/events.out.tfevents.1747291839.cs-Precision-7960-Tower.145708.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f9fc01b1e46f7782baad490bdd081471fcd69fde8ef1c845604aea4d1e57ea90
3
- size 54276
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e2609c766efe5f43d9a53c21fa56ab879064fa3337296506169f1a2b478142a4
3
+ size 54953