You can deploy this model and start earning money today!

nik1509/
telugu_wav2vec_epoch_ablation_ep30

Name: telugu_wav2vec_epoch_ablation_ep30
Availability: InStock

transformers

safetensors

wav2vec2

automatic-speech-recognition

generated_from_trainer

nik1509/telugu_wav2vecmedium_basemodel

finetune

apache-2.0

endpoints_compatible

Views

No views yet

telugu_wav2vec_epoch_ablation_ep30

This model is a fine-tuned version of nik1509/telugu_wav2vecmedium_basemodel on the None dataset. It achieves the following results on the evaluation set:

Loss: 769.0142
Wer: 0.3949
Cer: 0.1737

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: 5e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 16
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
lr_scheduler_warmup_ratio: 0.1
num_epochs: 30
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
312.4978	1.0	1152	458.7804	0.4080	0.1691
302.6044	2.0	2304	466.0941	0.4071	0.1694
290.0246	3.0	3456	480.7546	0.4008	0.1671
288.1721	4.0	4608	485.0590	0.4037	0.1678
276.2847	5.0	5760	493.4674	0.4004	0.1686
274.3864	6.0	6912	509.0528	0.3995	0.1679
271.1487	7.0	8064	519.3875	0.4001	0.1672
261.3909	8.0	9216	529.6630	0.4011	0.1686
254.278	9.0	10368	538.4158	0.3987	0.1682
256.6805	10.0	11520	539.9032	0.4013	0.1695
247.624	11.0	12672	552.4019	0.3977	0.1688
248.5292	12.0	13824	556.6013	0.4009	0.1695
234.2797	13.0	14976	575.8162	0.3977	0.1695
230.7405	14.0	16128	585.4182	0.3972	0.1694
229.7057	15.0	17280	594.6546	0.3956	0.1700
220.506	16.0	18432	613.9507	0.3977	0.1710
216.3434	17.0	19584	623.5539	0.3964	0.1710
208.7105	18.0	20736	623.8289	0.4000	0.1723
198.9742	19.0	21888	645.6901	0.3940	0.1716
202.0175	20.0	23040	657.6493	0.3983	0.1718
197.4541	21.0	24192	666.8010	0.3976	0.1722
192.2262	22.0	25344	670.3865	0.3994	0.1741
180.9265	23.0	26496	691.3597	0.3940	0.1722
178.3302	24.0	27648	704.4398	0.3949	0.1736
171.9031	25.0	28800	707.5822	0.3933	0.1730
170.2448	26.0	29952	721.4016	0.3935	0.1732
164.3709	27.0	31104	733.5203	0.3955	0.1738
162.439	28.0	32256	745.4579	0.3949	0.1732
156.3425	29.0	33408	760.7412	0.3933	0.1730
149.772	30.0	34560	769.0142	0.3949	0.1737

Framework versions

Transformers 4.53.0
Pytorch 2.8.0.dev20250609+cu118
Datasets 3.6.0
Tokenizers 0.21.2

nik1509/telugu_wav2vec_epoch_ablation_ep30

telugu_wav2vec_epoch_ablation_ep30

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

nik1509/
telugu_wav2vec_epoch_ablation_ep30