Helldivers2ASR_V2

This model is a fine-tuned version of facebook/wav2vec2-base-960h on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 403.7162
• Wer: 0.1140

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: 7e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1549.7986	1.0	239	1010.8287	0.4509
1073.833	2.0	478	847.1273	0.3799
863.4314	3.0	717	669.3406	0.3011
735.351	4.0	956	701.8651	0.3046
646.4708	5.0	1195	516.7015	0.2290
555.7994	6.0	1434	563.6837	0.2194
511.6518	7.0	1673	463.4145	0.1902
442.7085	8.0	1912	473.8917	0.1862
419.5254	9.0	2151	513.1998	0.1893
372.6168	10.0	2390	424.3353	0.1622
321.958	11.0	2629	430.3576	0.1550
321.6388	12.0	2868	422.5235	0.1624
293.641	13.0	3107	359.6511	0.1239
270.1068	14.0	3346	455.4551	0.1461
258.8057	15.0	3585	466.8189	0.1465
256.9584	16.0	3824	402.8066	0.1282
231.2268	17.0	4063	456.3300	0.1415
222.786	18.0	4302	329.8922	0.1127
213.7656	19.0	4541	463.8095	0.1426
188.9466	20.0	4780	337.8596	0.1132
199.4221	21.0	5019	390.5159	0.1212
189.1897	22.0	5258	384.9171	0.1190
182.3199	23.0	5497	375.2046	0.1143
180.674	24.0	5736	426.1663	0.1223
160.5728	25.0	5975	411.1659	0.1204
161.0371	26.0	6214	417.6654	0.1158
169.4026	27.0	6453	412.9657	0.1141
162.3533	28.0	6692	407.0595	0.1143
169.8538	29.0	6931	404.1884	0.1143
187.099	30.0	7170	403.7162	0.1140

Framework versions

• Transformers 4.44.0
• Pytorch 2.5.1+cu121
• Datasets 3.6.0
• Tokenizers 0.19.1