wav2vec2-tiny-demo

This model is a fine-tuned version of yenpolin/wav2vec2-tiny on the DNA_R9.4.1 - NA dataset. It achieves the following results on the evaluation set:

• Loss: 1.3174
• Mean Acc: 41.4052
• Median Acc: 54.7558

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: 0.0003
• train_batch_size: 400
• eval_batch_size: 800
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100.0

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Acc	Median Acc
No log	1.0	250	1.2708	0.0	0.0
1.2586	2.0	500	1.2401	0.0	0.0
1.2586	3.0	750	1.2212	0.0	0.0
1.1999	4.0	1000	1.1985	0.0026	0.0
1.1999	5.0	1250	1.1824	0.0122	0.0
1.1635	6.0	1500	1.1716	0.0301	0.0
1.1635	7.0	1750	1.1562	0.0296	0.0
1.1361	8.0	2000	1.1507	0.0088	0.0
1.1361	9.0	2250	1.1393	0.0649	0.0
1.1142	10.0	2500	1.1311	0.0150	0.0
1.1142	11.0	2750	1.0849	0.0534	0.0
1.0648	12.0	3000	1.0644	0.0912	0.0
1.0648	13.0	3250	1.0415	0.1348	0.0
1.0185	14.0	3500	1.0390	0.0443	0.0
1.0185	15.0	3750	1.0214	0.1120	0.0
0.9951	16.0	4000	1.0152	0.1647	0.0
0.9951	17.0	4250	1.0193	0.1194	0.0
0.9813	18.0	4500	1.0092	0.1182	0.0
0.9813	19.0	4750	1.0165	0.0672	0.0
0.9625	20.0	5000	0.9430	9.0082	0.0
0.9625	21.0	5250	0.9301	13.2452	0.0
0.8958	22.0	5500	0.9261	8.9878	0.0
0.8958	23.0	5750	0.9174	15.3641	0.0
0.8756	24.0	6000	0.9116	14.9011	0.0
0.8756	25.0	6250	0.9130	12.2681	0.0
0.8607	26.0	6500	0.9114	15.2874	0.0
0.8607	27.0	6750	0.9105	24.1205	0.0
0.8482	28.0	7000	0.9083	18.6607	0.0
0.8482	29.0	7250	0.9196	17.7246	0.0
0.8359	30.0	7500	0.9149	19.7733	0.0
0.8359	31.0	7750	0.9133	18.6297	0.0
0.8232	32.0	8000	0.9479	12.2703	0.0
0.8232	33.0	8250	0.9303	19.9043	0.0
0.8092	34.0	8500	0.9300	22.3510	0.0
0.8092	35.0	8750	0.9295	27.8118	0.0
0.7951	36.0	9000	0.9439	23.2963	0.0
0.7951	37.0	9250	0.9620	20.9074	0.0
0.7803	38.0	9500	0.9571	28.7828	52.3026
0.7803	39.0	9750	0.9814	25.2679	0.0
0.7669	40.0	10000	0.9787	31.2820	53.3865
0.7669	41.0	10250	0.9765	28.6333	52.2034
0.7529	42.0	10500	1.0038	27.5154	0.0
0.7529	43.0	10750	1.0338	28.4940	52.1368
0.7411	44.0	11000	1.0279	28.7206	52.2782
0.7411	45.0	11250	1.0177	29.0583	52.4116
0.7299	46.0	11500	1.0148	32.9230	53.6
0.7299	47.0	11750	1.0400	33.5174	53.8136
0.7198	48.0	12000	1.0478	30.9309	53.2544
0.7198	49.0	12250	1.0538	29.2445	52.5692
0.7109	50.0	12500	1.0524	27.2312	0.0
0.7109	51.0	12750	1.0862	33.3504	53.7234
0.7036	52.0	13000	1.0744	31.8779	53.3679
0.7036	53.0	13250	1.0628	28.3728	51.7123
0.6963	54.0	13500	1.0587	30.9822	53.125
0.6963	55.0	13750	1.0834	33.2130	53.5714
0.69	56.0	14000	1.1077	36.3846	54.2857
0.69	57.0	14250	1.1150	32.3586	53.4954
0.6855	58.0	14500	1.1352	36.9014	54.4061
0.6855	59.0	14750	1.1557	33.8871	53.8462
0.6811	60.0	15000	1.1315	33.5959	53.7736
0.6811	61.0	15250	1.0957	32.9908	53.5714
0.6768	62.0	15500	1.1236	32.7653	53.6122
0.6768	63.0	15750	1.1153	34.8283	53.9432
0.6722	64.0	16000	1.1300	35.0110	53.9683
0.6722	65.0	16250	1.1826	35.9229	54.1463
0.6682	66.0	16500	1.1534	38.5511	54.4850
0.6682	67.0	16750	1.1636	35.8289	54.0984
0.6653	68.0	17000	1.1404	34.6543	53.8462
0.6653	69.0	17250	1.1473	36.2493	54.1209
0.6624	70.0	17500	1.1532	39.5140	54.5775
0.6624	71.0	17750	1.1715	36.2354	54.1254
0.6597	72.0	18000	1.1875	35.5023	54.0856
0.6597	73.0	18250	1.1643	34.5047	53.9024
0.657	74.0	18500	1.1894	38.7561	54.4928
0.657	75.0	18750	1.2082	38.2157	54.4601
0.6543	76.0	19000	1.1843	34.2857	53.8835
0.6543	77.0	19250	1.1689	38.4264	54.4355
0.652	78.0	19500	1.2085	37.7495	54.3750
0.652	79.0	19750	1.1985	39.1444	54.5190
0.6497	80.0	20000	1.2331	40.0190	54.6125
0.6497	81.0	20250	1.2403	39.0511	54.5205
0.6476	82.0	20500	1.1910	37.2936	54.2453
0.6476	83.0	20750	1.2035	41.2572	54.7244
0.6457	84.0	21000	1.2124	38.7565	54.4892
0.6457	85.0	21250	1.2327	39.1286	54.4959
0.6437	86.0	21500	1.2252	39.8797	54.5775
0.6437	87.0	21750	1.2346	38.1128	54.3860
0.642	88.0	22000	1.2441	40.4358	54.6512
0.642	89.0	22250	1.2488	39.9912	54.5977
0.6403	90.0	22500	1.2483	39.9116	54.6053
0.6403	91.0	22750	1.2674	40.9458	54.7009
0.6387	92.0	23000	1.2694	40.4423	54.6358
0.6387	93.0	23250	1.2717	40.7397	54.6763
0.6371	94.0	23500	1.2820	40.5719	54.6392
0.6371	95.0	23750	1.2883	40.3534	54.6667
0.6358	96.0	24000	1.3039	40.6046	54.6667
0.6358	97.0	24250	1.3068	41.4525	54.7739
0.6347	98.0	24500	1.3132	41.4247	54.7541
0.6347	99.0	24750	1.3124	41.2563	54.7445
0.6339	100.0	25000	1.3174	41.4052	54.7558

Framework versions

• Transformers 4.26.0
• Pytorch 1.12.1
• Datasets 2.9.0
• Tokenizers 0.13.2