xlsr-a-nomimo

susmitabhatt/xlsr-a-nomimo

xlsr-a-nomimo is a automatic speech recognition model from susmitabhatt. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

19% of all-time downloads

All-time downloads

Public

Parameters

315M

65.6 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors1.3 GB · 100%

At a glance

Task: Automatic Speech Recognition
Library: transformers
License: apache-2.0
Model type: wav2vec2
Access: Public
Created: Nov 16, 2024
Updated: Nov 16, 2024
SHA: cf95e281

Base models

facebook/wav2vec2-large-xlsr-53

Model card

From the Hugging Face model README

xlsr-a-nomimo

This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.4323
Wer: 0.3434

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: 0.0004
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_steps: 132
num_epochs: 100
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
4.7158	1.9275	200	2.5326	1.0
1.6646	3.8502	400	0.5913	0.6836
0.4371	5.7729	600	0.2586	0.4437
0.2367	7.6957	800	0.3040	0.3974
0.1839	9.6184	1000	0.3419	0.3781
0.1751	11.5411	1200	0.3340	0.3804
0.1252	13.4638	1400	0.3298	0.3719
0.1281	15.3865	1600	0.3086	0.3519
0.1063	17.3092	1800	0.2961	0.3619
0.1074	19.2319	2000	0.3216	0.3673
0.0842	21.1546	2200	0.3414	0.3696
0.0876	23.0773	2400	0.3611	0.3549
0.075	25.0	2600	0.3633	0.3611
0.0714	26.9275	2800	0.3337	0.3565
0.072	28.8502	3000	0.3967	0.3696
0.0596	30.7729	3200	0.4276	0.3557
0.0547	32.6957	3400	0.3842	0.3611
0.0589	34.6184	3600	0.4189	0.3542
0.052	36.5411	3800	0.3832	0.3542
0.0445	38.4638	4000	0.3674	0.3627
0.0435	40.3865	4200	0.4747	0.3557
0.0492	42.3092	4400	0.4231	0.3526
0.0388	44.2319	4600	0.3696	0.3681
0.0485	46.1546	4800	0.4486	0.3526
0.0384	48.0773	5000	0.4647	0.3549
0.0355	50.0	5200	0.3833	0.3611
0.0365	51.9275	5400	0.4611	0.3549
0.0337	53.8502	5600	0.4655	0.3534
0.0378	55.7729	5800	0.4439	0.3534
0.0285	57.6957	6000	0.4738	0.3534
0.0312	59.6184	6200	0.4310	0.3619
0.0329	61.5411	6400	0.4504	0.3549
0.0253	63.4638	6600	0.4600	0.3542
0.0291	65.3865	6800	0.4838	0.3519
0.0272	67.3092	7000	0.4505	0.3495
0.0266	69.2319	7200	0.4152	0.3426
0.024	71.1546	7400	0.3947	0.3472
0.022	73.0773	7600	0.3965	0.3495
0.0215	75.0	7800	0.4233	0.3426
0.0225	76.9275	8000	0.3956	0.3449
0.0224	78.8502	8200	0.4179	0.3511
0.0207	80.7729	8400	0.4081	0.3434
0.0168	82.6957	8600	0.4523	0.3449
0.0194	84.6184	8800	0.3983	0.3457
0.0169	86.5411	9000	0.4381	0.3441
0.0173	88.4638	9200	0.4524	0.3465
0.0145	90.3865	9400	0.4308	0.3449
0.016	92.3092	9600	0.4272	0.3441
0.014	94.2319	9800	0.4203	0.3426
0.0152	96.1546	10000	0.4314	0.3457
0.0147	98.0773	10200	0.4323	0.3434

Framework versions

Transformers 4.47.0.dev0
Pytorch 2.4.0
Datasets 3.0.1
Tokenizers 0.20.0

Task: Automatic Speech Recognition
Library: transformers
Type: wav2vec2
License: apache-2.0
Created: Nov 16, 2024
Updated: Nov 16, 2024