mkadiyala commited on
Commit
558b0a8
·
verified ·
1 Parent(s): 53c3efb

End of training

Browse files
Files changed (5) hide show
  1. README.md +84 -0
  2. config.json +96 -0
  3. model.safetensors +3 -0
  4. preprocessor_config.json +9 -0
  5. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,84 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: apache-2.0
4
+ base_model: ntu-spml/distilhubert
5
+ tags:
6
+ - generated_from_trainer
7
+ datasets:
8
+ - marsyas/gtzan
9
+ metrics:
10
+ - accuracy
11
+ model-index:
12
+ - name: distilhubert-finetuned-gtzan
13
+ results:
14
+ - task:
15
+ name: Audio Classification
16
+ type: audio-classification
17
+ dataset:
18
+ name: GTZAN
19
+ type: marsyas/gtzan
20
+ config: all
21
+ split: train
22
+ args: all
23
+ metrics:
24
+ - name: Accuracy
25
+ type: accuracy
26
+ value: 0.87
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
+ # distilhubert-finetuned-gtzan
33
+
34
+ This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on the GTZAN dataset.
35
+ It achieves the following results on the evaluation set:
36
+ - Loss: 1.2205
37
+ - Accuracy: 0.87
38
+
39
+ ## Model description
40
+
41
+ More information needed
42
+
43
+ ## Intended uses & limitations
44
+
45
+ More information needed
46
+
47
+ ## Training and evaluation data
48
+
49
+ More information needed
50
+
51
+ ## Training procedure
52
+
53
+ ### Training hyperparameters
54
+
55
+ The following hyperparameters were used during training:
56
+ - learning_rate: 1e-05
57
+ - train_batch_size: 1
58
+ - eval_batch_size: 1
59
+ - seed: 42
60
+ - gradient_accumulation_steps: 4
61
+ - total_train_batch_size: 4
62
+ - optimizer: Use OptimizerNames.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: 5
66
+ - mixed_precision_training: Native AMP
67
+
68
+ ### Training results
69
+
70
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
71
+ |:-------------:|:------:|:----:|:---------------:|:--------:|
72
+ | 0.366 | 1.0 | 225 | 1.2412 | 0.84 |
73
+ | 0.0 | 2.0 | 450 | 1.3556 | 0.85 |
74
+ | 0.0 | 3.0 | 675 | 1.2284 | 0.86 |
75
+ | 0.0 | 4.0 | 900 | 1.2393 | 0.87 |
76
+ | 0.4097 | 4.9789 | 1120 | 1.2205 | 0.87 |
77
+
78
+
79
+ ### Framework versions
80
+
81
+ - Transformers 4.47.1
82
+ - Pytorch 2.5.1+cu124
83
+ - Datasets 3.2.0
84
+ - Tokenizers 0.21.0
config.json ADDED
@@ -0,0 +1,96 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "./distilhubert-finetuned-gtzan",
3
+ "activation_dropout": 0.1,
4
+ "apply_spec_augment": false,
5
+ "architectures": [
6
+ "HubertForSequenceClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "bos_token_id": 1,
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
+ "hidden_dropout": 0.1,
50
+ "hidden_size": 768,
51
+ "id2label": {
52
+ "0": "blues",
53
+ "1": "classical",
54
+ "2": "country",
55
+ "3": "disco",
56
+ "4": "hiphop",
57
+ "5": "jazz",
58
+ "6": "metal",
59
+ "7": "pop",
60
+ "8": "reggae",
61
+ "9": "rock"
62
+ },
63
+ "initializer_range": 0.02,
64
+ "intermediate_size": 3072,
65
+ "label2id": {
66
+ "blues": "0",
67
+ "classical": "1",
68
+ "country": "2",
69
+ "disco": "3",
70
+ "hiphop": "4",
71
+ "jazz": "5",
72
+ "metal": "6",
73
+ "pop": "7",
74
+ "reggae": "8",
75
+ "rock": "9"
76
+ },
77
+ "layer_norm_eps": 1e-05,
78
+ "layerdrop": 0.0,
79
+ "mask_feature_length": 10,
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
+ "num_conv_pos_embedding_groups": 16,
88
+ "num_conv_pos_embeddings": 128,
89
+ "num_feat_extract_layers": 7,
90
+ "num_hidden_layers": 2,
91
+ "pad_token_id": 0,
92
+ "torch_dtype": "float32",
93
+ "transformers_version": "4.47.1",
94
+ "use_weighted_layer_sum": false,
95
+ "vocab_size": 32
96
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f3a20ca4237821fff488ca4bb131a9b6c414d4d247cc4f6df5c86fcc71164856
3
+ size 94771728
preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0,
7
+ "return_attention_mask": true,
8
+ "sampling_rate": 16000
9
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:13698094d2cb6a05987b135b0ca97702189134f66a3cf3636f489f5f33a7b178
3
+ size 5368