sewd-classifier-aug-ref

fydhfzh/sewd-classifier-aug-ref

sewd-classifier-aug-ref is a audio classification model from fydhfzh. Use it for the audio classification task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as apache-2.0.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

0% of all-time downloads

All-time downloads

526

Public

Parameters

24.2M

582 MB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors97 MB · 100%

At a glance

Task: Audio Classification
Library: transformers
License: apache-2.0
Model type: sew-d
Access: Public
Created: Jul 12, 2024
Updated: Jul 12, 2024
SHA: bd190f1e

Base models

asapp/sew-d-tiny-100k-ft-ls100h

Model card

From the Hugging Face model README

sewd-classifier-aug-ref

This model is a fine-tuned version of asapp/sew-d-tiny-100k-ft-ls100h on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 1.2210
Accuracy: 0.6402
Precision: 0.6291
Recall: 0.6402
F1: 0.6137
Binary: 0.7478

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: 32
eval_batch_size: 32
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 128
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Binary
No log	0.13	50	4.3839	0.0337	0.0029	0.0337	0.0052	0.2004
No log	0.27	100	4.1179	0.0687	0.0353	0.0687	0.0281	0.3295
No log	0.4	150	3.8682	0.0889	0.0335	0.0889	0.0338	0.3523
No log	0.54	200	3.6594	0.0997	0.0311	0.0997	0.0366	0.3635
No log	0.67	250	3.5084	0.1280	0.0534	0.1280	0.0494	0.3838
No log	0.81	300	3.3946	0.1469	0.0476	0.1469	0.0628	0.3964
No log	0.94	350	3.2570	0.1604	0.0699	0.1604	0.0830	0.4082
3.9151	1.08	400	3.1540	0.1806	0.0844	0.1806	0.0987	0.4224
3.9151	1.21	450	3.0449	0.1846	0.1000	0.1846	0.1005	0.4260
3.9151	1.35	500	2.9543	0.2237	0.1403	0.2237	0.1376	0.4534
3.9151	1.48	550	2.8691	0.2507	0.1621	0.2507	0.1606	0.4706
3.9151	1.62	600	2.7812	0.2493	0.1520	0.2493	0.1592	0.4718
3.9151	1.75	650	2.6598	0.2871	0.1855	0.2871	0.1856	0.4981
3.9151	1.89	700	2.6099	0.2951	0.2123	0.2951	0.2019	0.5047
3.1406	2.02	750	2.5039	0.3235	0.2106	0.3235	0.2230	0.5236
3.1406	2.16	800	2.4359	0.3383	0.2454	0.3383	0.2501	0.5358
3.1406	2.29	850	2.3869	0.3154	0.2329	0.3154	0.2324	0.5179
3.1406	2.43	900	2.3144	0.3612	0.2937	0.3612	0.2798	0.5513
3.1406	2.56	950	2.2470	0.3720	0.3122	0.3720	0.2908	0.5584
3.1406	2.7	1000	2.1944	0.3774	0.3099	0.3774	0.2992	0.5632
3.1406	2.83	1050	2.1421	0.4030	0.3250	0.4030	0.3226	0.5819
3.1406	2.97	1100	2.0630	0.4137	0.3442	0.4137	0.3336	0.5899
2.6974	3.1	1150	2.0115	0.4245	0.3679	0.4245	0.3510	0.5974
2.6974	3.24	1200	1.9716	0.4434	0.3964	0.4434	0.3729	0.6093
2.6974	3.37	1250	1.9255	0.4488	0.3972	0.4488	0.3883	0.6150
2.6974	3.51	1300	1.8715	0.4623	0.4112	0.4623	0.3969	0.6228
2.6974	3.64	1350	1.8223	0.4825	0.4534	0.4825	0.4222	0.6369
2.6974	3.78	1400	1.7951	0.5013	0.4728	0.5013	0.4500	0.6511
2.6974	3.91	1450	1.7427	0.5270	0.4855	0.5270	0.4804	0.6686
2.3963	4.05	1500	1.7319	0.5	0.4618	0.5	0.4452	0.6493
2.3963	4.18	1550	1.7098	0.4960	0.4588	0.4960	0.4454	0.6473
2.3963	4.32	1600	1.6518	0.5310	0.5051	0.5310	0.4855	0.6709
2.3963	4.45	1650	1.6535	0.5067	0.4838	0.5067	0.4552	0.6539
2.3963	4.59	1700	1.6011	0.5539	0.5106	0.5539	0.5061	0.6865
2.3963	4.72	1750	1.5894	0.5404	0.4940	0.5404	0.4923	0.6767
2.3963	4.86	1800	1.5580	0.5660	0.5371	0.5660	0.5285	0.6964
2.3963	4.99	1850	1.5375	0.5431	0.5032	0.5431	0.4968	0.6803
2.1926	5.12	1900	1.5166	0.5620	0.5237	0.5620	0.5193	0.6941
2.1926	5.26	1950	1.5168	0.5526	0.5198	0.5526	0.5085	0.6860
2.1926	5.39	2000	1.4773	0.5836	0.5615	0.5836	0.5455	0.7073
2.1926	5.53	2050	1.4488	0.5782	0.5564	0.5782	0.5396	0.7054
2.1926	5.66	2100	1.4335	0.5916	0.5691	0.5916	0.5560	0.7143
2.1926	5.8	2150	1.4078	0.5957	0.5782	0.5957	0.5641	0.7177
2.1926	5.93	2200	1.4092	0.5863	0.5691	0.5863	0.5506	0.7105
2.0446	6.07	2250	1.3942	0.5755	0.5405	0.5755	0.5334	0.7026
2.0446	6.2	2300	1.3828	0.5930	0.5776	0.5930	0.5613	0.7148
2.0446	6.34	2350	1.3625	0.6065	0.5886	0.6065	0.5688	0.7247
2.0446	6.47	2400	1.3444	0.6119	0.6008	0.6119	0.5754	0.7284
2.0446	6.61	2450	1.3088	0.6267	0.6134	0.6267	0.5914	0.7388
2.0446	6.74	2500	1.3183	0.6038	0.5869	0.6038	0.5729	0.7228
2.0446	6.88	2550	1.3000	0.6173	0.5886	0.6173	0.5810	0.7322
1.9441	7.01	2600	1.2930	0.6213	0.6048	0.6213	0.5900	0.7341
1.9441	7.15	2650	1.2757	0.6226	0.6097	0.6226	0.5959	0.7361
1.9441	7.28	2700	1.2787	0.6226	0.6091	0.6226	0.5963	0.7361
1.9441	7.42	2750	1.2566	0.6240	0.6204	0.6240	0.5983	0.7375
1.9441	7.55	2800	1.2549	0.6253	0.6055	0.6253	0.5970	0.7380
1.9441	7.69	2850	1.2396	0.6240	0.6255	0.6240	0.5954	0.7371
1.9441	7.82	2900	1.2400	0.6388	0.6361	0.6388	0.6128	0.7478
1.9441	7.96	2950	1.2369	0.6294	0.6188	0.6294	0.5996	0.7407
1.8636	8.09	3000	1.2235	0.6375	0.6363	0.6375	0.6151	0.7460
1.8636	8.23	3050	1.2178	0.6456	0.6415	0.6456	0.6196	0.7520
1.8636	8.36	3100	1.2093	0.6402	0.6346	0.6402	0.6149	0.7482
1.8636	8.5	3150	1.2210	0.6402	0.6291	0.6402	0.6137	0.7478

Framework versions

Transformers 4.38.2
Pytorch 2.3.0
Datasets 2.19.1
Tokenizers 0.15.1

Task: Audio Classification
Library: transformers
Type: sew-d
License: apache-2.0
Created: Jul 12, 2024
Updated: Jul 12, 2024