You can deploy this model and start earning money today!

nik1509/
xlsr300m_hiddendropout_ablation_hd_02

Name: xlsr300m_hiddendropout_ablation_hd_02
Availability: InStock

Views

No views yet

xlsr300m_hiddendropout_ablation_hd_02

This model is a fine-tuned version of nik1509/xlsr300m_finetuned_telugu_basemodel on the None dataset. It achieves the following results on the evaluation set:

More information needed

More information needed

More information needed

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: 15
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
336.711	0.6076	700	451.9146	0.4021	0.1670
284.9517	1.2153	1400	456.9569	0.4011	0.1668
299.4901	1.8229	2100	451.4269	0.4066	0.1678
297.3845	2.4306	2800	457.5531	0.4023	0.1677
306.1488	3.0382	3500	461.1308	0.4046	0.1685
241.4591	3.6458	4200	469.7468	0.4003	0.1677
267.8068	4.2535	4900	474.7147	0.4035	0.1682
304.2234	4.8611	5600	469.3616	0.4047	0.1679
269.7648	5.4688	6300	481.7546	0.4031	0.1680
280.8081	6.0764	7000	492.9857	0.3991	0.1670
288.6179	6.6840	7700	491.6023	0.3975	0.1665
295.0424	7.2917	8400	497.6441	0.3999	0.1670
285.9945	7.8993	9100	495.3789	0.3978	0.1672
238.1933	8.5069	9800	507.9428	0.4003	0.1672
255.7765	9.1146	10500	512.9609	0.3994	0.1676
303.4657	9.7222	11200	514.5809	0.3967	0.1672
267.0756	10.3299	11900	522.4242	0.3957	0.1669
249.5104	10.9375	12600	514.8857	0.3945	0.1666
282.4019	11.5451	13300	527.7849	0.3954	0.1666
235.1887	12.1528	14000	527.1183	0.3967	0.1668
287.2251	12.7604	14700	530.1781	0.3983	0.1666
251.9085	13.3681	15400	535.0036	0.3946	0.1661
280.9337	13.9757	16100	530.5440	0.3972	0.1666
229.736	14.5833	16800	532.7958	0.3972	0.1669