Upload 5 files
Browse files- README.md +47 -27
- config.json +15 -62
- preprocessor_config.json +8 -4
- pytorch_model.bin +2 -2
- training_args.bin +2 -2
README.md
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
---
|
2 |
-
license:
|
3 |
tags:
|
4 |
- generated_from_trainer
|
5 |
datasets:
|
@@ -7,19 +7,32 @@ datasets:
|
|
7 |
metrics:
|
8 |
- accuracy
|
9 |
model-index:
|
10 |
-
- name:
|
11 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
12 |
---
|
13 |
|
14 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
15 |
should probably proofread and complete it, then remove this comment. -->
|
16 |
|
17 |
-
#
|
18 |
|
19 |
-
This model is a fine-tuned version of [
|
20 |
It achieves the following results on the evaluation set:
|
21 |
-
- Loss: 0.
|
22 |
-
- Accuracy: 0.
|
23 |
|
24 |
## Model description
|
25 |
|
@@ -39,38 +52,45 @@ More information needed
|
|
39 |
|
40 |
The following hyperparameters were used during training:
|
41 |
- learning_rate: 5e-05
|
42 |
-
- train_batch_size:
|
43 |
-
- eval_batch_size:
|
44 |
- seed: 42
|
|
|
|
|
45 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
46 |
- lr_scheduler_type: linear
|
47 |
- lr_scheduler_warmup_ratio: 0.1
|
48 |
-
- num_epochs:
|
49 |
|
50 |
### Training results
|
51 |
|
52 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
53 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
54 |
-
|
|
55 |
-
|
|
56 |
-
|
|
57 |
-
| 0.
|
58 |
-
| 0.
|
59 |
-
| 0.
|
60 |
-
| 0.
|
61 |
-
| 0.
|
62 |
-
| 0.
|
63 |
-
| 0.
|
64 |
-
| 0.
|
65 |
-
| 0.
|
66 |
-
| 0.
|
67 |
-
| 0.
|
68 |
-
| 0.
|
|
|
|
|
|
|
|
|
|
|
69 |
|
70 |
|
71 |
### Framework versions
|
72 |
|
73 |
-
- Transformers 4.
|
74 |
- Pytorch 2.0.1+cu118
|
75 |
-
- Datasets 2.
|
76 |
- Tokenizers 0.13.3
|
|
|
1 |
---
|
2 |
+
license: bsd-3-clause
|
3 |
tags:
|
4 |
- generated_from_trainer
|
5 |
datasets:
|
|
|
7 |
metrics:
|
8 |
- accuracy
|
9 |
model-index:
|
10 |
+
- name: ast-finetuned-audioset-10-10-0.4593-finetuned-gtzan
|
11 |
+
results:
|
12 |
+
- task:
|
13 |
+
name: Audio Classification
|
14 |
+
type: audio-classification
|
15 |
+
dataset:
|
16 |
+
name: GTZAN
|
17 |
+
type: marsyas/gtzan
|
18 |
+
config: all
|
19 |
+
split: train
|
20 |
+
args: all
|
21 |
+
metrics:
|
22 |
+
- name: Accuracy
|
23 |
+
type: accuracy
|
24 |
+
value: 0.9
|
25 |
---
|
26 |
|
27 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
28 |
should probably proofread and complete it, then remove this comment. -->
|
29 |
|
30 |
+
# ast-finetuned-audioset-10-10-0.4593-finetuned-gtzan
|
31 |
|
32 |
+
This model is a fine-tuned version of [MIT/ast-finetuned-audioset-10-10-0.4593](https://huggingface.co/MIT/ast-finetuned-audioset-10-10-0.4593) on the GTZAN dataset.
|
33 |
It achieves the following results on the evaluation set:
|
34 |
+
- Loss: 0.4717
|
35 |
+
- Accuracy: 0.9
|
36 |
|
37 |
## Model description
|
38 |
|
|
|
52 |
|
53 |
The following hyperparameters were used during training:
|
54 |
- learning_rate: 5e-05
|
55 |
+
- train_batch_size: 4
|
56 |
+
- eval_batch_size: 4
|
57 |
- seed: 42
|
58 |
+
- gradient_accumulation_steps: 4
|
59 |
+
- total_train_batch_size: 16
|
60 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
61 |
- lr_scheduler_type: linear
|
62 |
- lr_scheduler_warmup_ratio: 0.1
|
63 |
+
- num_epochs: 20
|
64 |
|
65 |
### Training results
|
66 |
|
67 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
68 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
69 |
+
| 0.7581 | 1.0 | 56 | 0.7029 | 0.78 |
|
70 |
+
| 0.3942 | 1.99 | 112 | 0.4646 | 0.86 |
|
71 |
+
| 0.3298 | 2.99 | 168 | 0.3861 | 0.88 |
|
72 |
+
| 0.1227 | 4.0 | 225 | 0.4702 | 0.86 |
|
73 |
+
| 0.0774 | 5.0 | 281 | 0.4492 | 0.9 |
|
74 |
+
| 0.0039 | 5.99 | 337 | 0.4607 | 0.9 |
|
75 |
+
| 0.0014 | 6.99 | 393 | 0.5022 | 0.9 |
|
76 |
+
| 0.0022 | 8.0 | 450 | 0.4711 | 0.9 |
|
77 |
+
| 0.0193 | 9.0 | 506 | 0.5226 | 0.86 |
|
78 |
+
| 0.0004 | 9.99 | 562 | 0.6055 | 0.82 |
|
79 |
+
| 0.0003 | 10.99 | 618 | 0.4793 | 0.89 |
|
80 |
+
| 0.0002 | 12.0 | 675 | 0.5052 | 0.9 |
|
81 |
+
| 0.0002 | 13.0 | 731 | 0.4652 | 0.89 |
|
82 |
+
| 0.0001 | 13.99 | 787 | 0.4617 | 0.9 |
|
83 |
+
| 0.0001 | 14.99 | 843 | 0.4653 | 0.9 |
|
84 |
+
| 0.0001 | 16.0 | 900 | 0.4635 | 0.91 |
|
85 |
+
| 0.0001 | 17.0 | 956 | 0.4693 | 0.9 |
|
86 |
+
| 0.0001 | 17.99 | 1012 | 0.4697 | 0.9 |
|
87 |
+
| 0.0001 | 18.99 | 1068 | 0.4715 | 0.9 |
|
88 |
+
| 0.0025 | 19.91 | 1120 | 0.4717 | 0.9 |
|
89 |
|
90 |
|
91 |
### Framework versions
|
92 |
|
93 |
+
- Transformers 4.31.0.dev0
|
94 |
- Pytorch 2.0.1+cu118
|
95 |
+
- Datasets 2.13.1
|
96 |
- Tokenizers 0.13.3
|
config.json
CHANGED
@@ -1,52 +1,12 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "
|
3 |
-
"activation_dropout": 0.1,
|
4 |
-
"apply_spec_augment": false,
|
5 |
"architectures": [
|
6 |
-
"
|
7 |
],
|
8 |
-
"
|
9 |
-
"
|
10 |
-
"classifier_proj_size": 256,
|
11 |
-
"conv_bias": false,
|
12 |
-
"conv_dim": [
|
13 |
-
512,
|
14 |
-
512,
|
15 |
-
512,
|
16 |
-
512,
|
17 |
-
512,
|
18 |
-
512,
|
19 |
-
512
|
20 |
-
],
|
21 |
-
"conv_kernel": [
|
22 |
-
10,
|
23 |
-
3,
|
24 |
-
3,
|
25 |
-
3,
|
26 |
-
3,
|
27 |
-
2,
|
28 |
-
2
|
29 |
-
],
|
30 |
-
"conv_stride": [
|
31 |
-
5,
|
32 |
-
2,
|
33 |
-
2,
|
34 |
-
2,
|
35 |
-
2,
|
36 |
-
2,
|
37 |
-
2
|
38 |
-
],
|
39 |
-
"ctc_loss_reduction": "sum",
|
40 |
-
"ctc_zero_infinity": false,
|
41 |
-
"do_stable_layer_norm": false,
|
42 |
-
"eos_token_id": 2,
|
43 |
-
"feat_extract_activation": "gelu",
|
44 |
-
"feat_extract_norm": "group",
|
45 |
-
"feat_proj_dropout": 0.0,
|
46 |
-
"feat_proj_layer_norm": false,
|
47 |
-
"final_dropout": 0.0,
|
48 |
"hidden_act": "gelu",
|
49 |
-
"
|
50 |
"hidden_size": 768,
|
51 |
"id2label": {
|
52 |
"0": "blues",
|
@@ -74,23 +34,16 @@
|
|
74 |
"reggae": "8",
|
75 |
"rock": "9"
|
76 |
},
|
77 |
-
"layer_norm_eps": 1e-
|
78 |
-
"
|
79 |
-
"
|
80 |
-
"mask_feature_min_masks": 0,
|
81 |
-
"mask_feature_prob": 0.0,
|
82 |
-
"mask_time_length": 10,
|
83 |
-
"mask_time_min_masks": 2,
|
84 |
-
"mask_time_prob": 0.05,
|
85 |
-
"model_type": "hubert",
|
86 |
"num_attention_heads": 12,
|
87 |
-
"
|
88 |
-
"
|
89 |
-
"
|
90 |
-
"
|
91 |
-
"
|
|
|
92 |
"torch_dtype": "float32",
|
93 |
-
"transformers_version": "4.
|
94 |
-
"use_weighted_layer_sum": false,
|
95 |
-
"vocab_size": 32
|
96 |
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "MIT/ast-finetuned-audioset-10-10-0.4593",
|
|
|
|
|
3 |
"architectures": [
|
4 |
+
"ASTForAudioClassification"
|
5 |
],
|
6 |
+
"attention_probs_dropout_prob": 0.0,
|
7 |
+
"frequency_stride": 10,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
"hidden_act": "gelu",
|
9 |
+
"hidden_dropout_prob": 0.0,
|
10 |
"hidden_size": 768,
|
11 |
"id2label": {
|
12 |
"0": "blues",
|
|
|
34 |
"reggae": "8",
|
35 |
"rock": "9"
|
36 |
},
|
37 |
+
"layer_norm_eps": 1e-12,
|
38 |
+
"max_length": 1024,
|
39 |
+
"model_type": "audio-spectrogram-transformer",
|
|
|
|
|
|
|
|
|
|
|
|
|
40 |
"num_attention_heads": 12,
|
41 |
+
"num_hidden_layers": 12,
|
42 |
+
"num_mel_bins": 128,
|
43 |
+
"patch_size": 16,
|
44 |
+
"problem_type": "single_label_classification",
|
45 |
+
"qkv_bias": true,
|
46 |
+
"time_stride": 10,
|
47 |
"torch_dtype": "float32",
|
48 |
+
"transformers_version": "4.31.0.dev0"
|
|
|
|
|
49 |
}
|
preprocessor_config.json
CHANGED
@@ -1,9 +1,13 @@
|
|
1 |
{
|
2 |
"do_normalize": true,
|
3 |
-
"feature_extractor_type": "
|
4 |
"feature_size": 1,
|
|
|
|
|
|
|
5 |
"padding_side": "right",
|
6 |
-
"padding_value": 0,
|
7 |
-
"return_attention_mask":
|
8 |
-
"sampling_rate": 16000
|
|
|
9 |
}
|
|
|
1 |
{
|
2 |
"do_normalize": true,
|
3 |
+
"feature_extractor_type": "ASTFeatureExtractor",
|
4 |
"feature_size": 1,
|
5 |
+
"max_length": 1024,
|
6 |
+
"mean": -4.2677393,
|
7 |
+
"num_mel_bins": 128,
|
8 |
"padding_side": "right",
|
9 |
+
"padding_value": 0.0,
|
10 |
+
"return_attention_mask": false,
|
11 |
+
"sampling_rate": 16000,
|
12 |
+
"std": 4.5689974
|
13 |
}
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:aaddab9baccf21fdd4b71c5748750967fccfbd7145c9f067636b0798b06c99d8
|
3 |
+
size 344860025
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:820d83430bf5ca24a07288b6880ffe6c418e766a8e4e37faf5d2821727ea8e2f
|
3 |
+
size 4027
|