common_voice_clone_continued

This model is a fine-tuned version of Mehrdad-S/common_voice_clone on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4442

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-06
• train_batch_size: 8
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.4885	0.4002	100	0.4510
0.486	0.8004	200	0.4516
0.4836	1.2041	300	0.4500
0.4803	1.6043	400	0.4493
0.4873	2.0080	500	0.4510
0.4833	2.4082	600	0.4495
0.4869	2.8084	700	0.4486
0.4802	3.2121	800	0.4488
0.4758	3.6123	900	0.4470
0.4879	4.0160	1000	0.4472
0.4825	4.4162	1100	0.4480
0.4727	4.8164	1200	0.4457
0.4777	5.2201	1300	0.4485
0.4854	5.6203	1400	0.4488
0.4881	6.0240	1500	0.4472
0.481	6.4242	1600	0.4472
0.474	6.8244	1700	0.4471
0.4836	7.2281	1800	0.4468
0.4852	7.6283	1900	0.4480
0.479	8.0320	2000	0.4449
0.4805	8.4322	2100	0.4463
0.4743	8.8324	2200	0.4477
0.4792	9.2361	2300	0.4473
0.475	9.6363	2400	0.4451
0.4878	10.0400	2500	0.4456
0.478	10.4402	2600	0.4461
0.4805	10.8404	2700	0.4453
0.4773	11.2441	2800	0.4459
0.48	11.6443	2900	0.4453
0.479	12.0480	3000	0.4448
0.475	12.4482	3100	0.4437
0.4752	12.8484	3200	0.4461
0.4767	13.2521	3300	0.4434
0.4739	13.6523	3400	0.4458
0.4762	14.0560	3500	0.4431
0.4722	14.4562	3600	0.4450
0.4742	14.8564	3700	0.4442
0.4809	15.2601	3800	0.4448
0.475	15.6603	3900	0.4457
0.4789	16.0640	4000	0.4454
0.4709	16.4642	4100	0.4450
0.4826	16.8644	4200	0.4454
0.4735	17.2681	4300	0.4446
0.4727	17.6683	4400	0.4433
0.4867	18.0720	4500	0.4450
0.4804	18.4722	4600	0.4427
0.4802	18.8724	4700	0.4448
0.4798	19.2761	4800	0.4459
0.4788	19.6763	4900	0.4438
0.4772	20.0800	5000	0.4442

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.3.1
• Tokenizers 0.21.0