helldivers2-jarvis-asr

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: 59.4560
• Wer: 0.2597
• Cer: 0.8461

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: 1e-05
• train_batch_size: 12
• eval_batch_size: 12
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• num_epochs: 35

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
592.8701	1.0	33	309.4361	0.4890	0.8633
451.2683	2.0	66	203.9322	0.3950	0.8575
379.4145	3.0	99	167.4042	0.3591	0.8544
336.9402	4.0	132	150.8170	0.3398	0.8525
321.581	5.0	165	134.3422	0.3232	0.8515
286.1988	6.0	198	118.7198	0.3232	0.8504
275.0421	7.0	231	98.2645	0.3232	0.8503
271.4439	8.0	264	96.4090	0.3149	0.8501
257.0019	9.0	297	90.3913	0.3094	0.8490
237.0949	10.0	330	86.4160	0.2928	0.8484
245.8697	11.0	363	86.0826	0.2901	0.8485
237.8742	12.0	396	84.3617	0.2901	0.8478
234.2587	13.0	429	78.6872	0.2845	0.8481
209.8943	14.0	462	81.1480	0.2873	0.8478
226.1856	15.0	495	86.4198	0.2790	0.8481
200.5639	16.0	528	73.4982	0.2735	0.8473
206.5734	17.0	561	70.4848	0.2735	0.8473
190.7183	18.0	594	69.2320	0.2818	0.8474
224.0161	19.0	627	63.0107	0.2680	0.8467
202.4105	20.0	660	69.8898	0.2790	0.8467
189.4804	21.0	693	64.9169	0.2790	0.8469
194.0889	22.0	726	73.2138	0.2735	0.8464
177.1985	23.0	759	69.4830	0.2735	0.8464
179.883	24.0	792	67.9310	0.2624	0.8462
184.1469	25.0	825	65.2100	0.2735	0.8464
180.9639	26.0	858	65.6959	0.2707	0.8465
161.8609	27.0	891	64.2684	0.2597	0.8461
176.6417	28.0	924	64.3977	0.2652	0.8459
162.5974	29.0	957	61.5234	0.2597	0.8458
174.1075	30.0	990	64.5171	0.2597	0.8459
165.3279	31.0	1023	64.0766	0.2597	0.8458
174.7633	32.0	1056	70.9056	0.2680	0.8464
166.3098	33.0	1089	61.5759	0.2707	0.8460
176.9802	34.0	1122	59.8946	0.2652	0.8461
161.3542	35.0	1155	59.4560	0.2597	0.8461

Framework versions

• Transformers 4.51.3
• Pytorch 2.4.1+cu118
• Datasets 3.5.1
• Tokenizers 0.21.1