lg-digits-classification

This model is a fine-tuned version of facebook/wav2vec2-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.5405
• Accuracy: 0.6087

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: 4
• eval_batch_size: 1
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.9211	1.0	47	2.0718	0.2174
1.9034	2.0	94	2.0263	0.2174
1.8375	3.0	141	2.1374	0.2609
1.7839	4.0	188	2.0091	0.3043
1.6964	5.0	235	2.0135	0.2609
1.6485	6.0	282	1.9707	0.4348
1.5452	7.0	329	1.8629	0.3043
1.4662	8.0	376	1.8037	0.4348
1.3512	9.0	423	1.6884	0.5217
1.2567	10.0	470	1.7180	0.5652
1.1825	11.0	517	1.5855	0.5652
1.0535	12.0	564	1.6781	0.4783
0.9519	13.0	611	1.5681	0.5652
0.9136	14.0	658	1.4914	0.6087
0.7924	15.0	705	1.7370	0.5652
0.6775	16.0	752	1.6007	0.5652
0.7597	17.0	799	1.5242	0.6087
0.5731	18.0	846	1.7712	0.6087
0.5235	19.0	893	1.5915	0.5652
0.5463	20.0	940	1.7188	0.5652
0.4434	21.0	987	1.8435	0.5652
0.3921	22.0	1034	1.8180	0.6522
0.4408	23.0	1081	1.9594	0.5652
0.3323	24.0	1128	1.9446	0.5652
0.2806	25.0	1175	1.9466	0.5652
0.2228	26.0	1222	2.1430	0.5652
0.3144	27.0	1269	2.2630	0.5652
0.2841	28.0	1316	2.1039	0.5652
0.2392	29.0	1363	2.1887	0.6087
0.1964	30.0	1410	2.1597	0.6522
0.2014	31.0	1457	2.2470	0.6087
0.194	32.0	1504	2.3058	0.5652
0.202	33.0	1551	2.4327	0.5652
0.1968	34.0	1598	2.4571	0.5652
0.1675	35.0	1645	2.4031	0.6087
0.2261	36.0	1692	2.4036	0.6087
0.279	37.0	1739	2.4073	0.6087
0.2378	38.0	1786	2.4481	0.6087
0.2414	39.0	1833	2.4839	0.6087
0.189	40.0	1880	2.5730	0.5652
0.2018	41.0	1927	2.6159	0.5652
0.2418	42.0	1974	2.4725	0.5652
0.1489	43.0	2021	2.4596	0.5652
0.2277	44.0	2068	2.4913	0.6087
0.2313	45.0	2115	2.4922	0.6087
0.2333	46.0	2162	2.4951	0.6087
0.2201	47.0	2209	2.5213	0.6087
0.2116	48.0	2256	2.5164	0.6087
0.2122	49.0	2303	2.5424	0.6087
0.1965	50.0	2350	2.5405	0.6087

Framework versions

• Transformers 4.56.2
• Pytorch 2.8.0+cu128
• Datasets 3.6.0
• Tokenizers 0.22.1