metadata

license: apache-2.0
base_model: ntu-spml/distilhubert
tags:
  - generated_from_trainer
datasets:
  - marsyas/gtzan
metrics:
  - accuracy
model-index:
  - name: distilhubert-finetuned-gtzan
    results:
      - task:
          name: Audio Classification
          type: audio-classification
        dataset:
          name: GTZAN
          type: marsyas/gtzan
          config: all
          split: train
          args: all
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.775

distilhubert-finetuned-gtzan

This model is a fine-tuned version of ntu-spml/distilhubert on the GTZAN dataset. It achieves the following results on the evaluation set:

Loss: 1.2754
Accuracy: 0.775

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 3e-05
train_batch_size: 12
eval_batch_size: 12
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 50
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.2856	1.0	67	2.2801	0.19
2.1936	2.0	134	2.1829	0.335
1.9496	3.0	201	1.9189	0.5
1.6727	4.0	268	1.6280	0.595
1.5444	5.0	335	1.4530	0.635
1.0974	6.0	402	1.2269	0.67
1.0647	7.0	469	1.0802	0.72
0.8521	8.0	536	0.9819	0.72
0.7618	9.0	603	0.9660	0.74
0.5022	10.0	670	0.8664	0.75
0.4576	11.0	737	0.8972	0.7
0.2801	12.0	804	0.8073	0.76
0.2404	13.0	871	0.7892	0.765
0.1493	14.0	938	0.8512	0.74
0.0945	15.0	1005	0.8876	0.74
0.049	16.0	1072	0.9735	0.72
0.0311	17.0	1139	0.9881	0.76
0.0225	18.0	1206	1.0965	0.735
0.0164	19.0	1273	1.0578	0.76
0.0124	20.0	1340	1.0298	0.75
0.0109	21.0	1407	1.0762	0.745
0.0085	22.0	1474	1.1168	0.75
0.0071	23.0	1541	1.1697	0.73
0.0063	24.0	1608	1.1204	0.765
0.0054	25.0	1675	1.1270	0.765
0.005	26.0	1742	1.1315	0.76
0.0521	27.0	1809	1.1868	0.755
0.004	28.0	1876	1.1645	0.77
0.0468	29.0	1943	1.1515	0.775
0.0036	30.0	2010	1.1655	0.775
0.0595	31.0	2077	1.2069	0.76
0.003	32.0	2144	1.2012	0.77
0.0029	33.0	2211	1.2369	0.755
0.0027	34.0	2278	1.2397	0.765
0.0026	35.0	2345	1.2581	0.765
0.029	36.0	2412	1.2226	0.76
0.0024	37.0	2479	1.1833	0.775
0.0023	38.0	2546	1.2723	0.765
0.0023	39.0	2613	1.2575	0.77
0.0284	40.0	2680	1.2945	0.76
0.002	41.0	2747	1.2345	0.765
0.0203	42.0	2814	1.2607	0.77
0.002	43.0	2881	1.2945	0.765
0.0019	44.0	2948	1.2487	0.77
0.0018	45.0	3015	1.2626	0.78
0.0018	46.0	3082	1.2692	0.77
0.0017	47.0	3149	1.2783	0.77
0.0018	48.0	3216	1.2813	0.775
0.0017	49.0	3283	1.2861	0.775
0.0275	50.0	3350	1.2754	0.775

Framework versions

Transformers 4.36.2
Pytorch 2.1.2+cu121
Datasets 2.16.1
Tokenizers 0.15.0