wav2vec2-large-mms-1b-aft-cgg

robertp408/wav2vec2-large-mms-1b-aft-cgg

wav2vec2-large-mms-1b-aft-cgg is a automatic speech recognition model from robertp408. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as cc-by-nc-4.0.

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

Downloads · 30 days

21% of all-time downloads

All-time downloads

Public

Parameters

965M

178 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors3.9 GB · 100%

At a glance

Task: Automatic Speech Recognition
Library: transformers
License: cc-by-nc-4.0
Model type: wav2vec2
Access: Public
Created: Sep 28, 2025
Updated: Oct 7, 2025
SHA: bbc693c6

Base models

facebook/mms-1b-all

Datasets

audiofolder

Model card

From the Hugging Face model README

wav2vec2-large-mms-1b-aft-cgg

This model is a fine-tuned version of facebook/mms-1b-all on the audiofolder dataset. It achieves the following results on the evaluation set:

Loss: 0.3863
Wer: 0.4405

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: 0.001
train_batch_size: 2
eval_batch_size: 8
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 100
num_epochs: 5
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
8.3707	0.1130	100	4.4216	0.9999
1.2337	0.2260	200	0.6319	0.6490
0.5775	0.3390	300	0.5327	0.6029
0.6104	0.4520	400	0.5300	0.5887
0.5848	0.5650	500	0.5158	0.5771
0.5588	0.6780	600	0.4905	0.5513
0.54	0.7910	700	0.4668	0.5257
0.4982	0.9040	800	0.4728	0.5085
0.5231	1.0169	900	0.4614	0.5169
0.512	1.1299	1000	0.4479	0.5147
0.5573	1.2429	1100	0.4439	0.4923
0.4875	1.3559	1200	0.4475	0.4995
0.4518	1.4689	1300	0.4406	0.4950
0.5092	1.5819	1400	0.4272	0.4858
0.4673	1.6949	1500	0.4424	0.4881
0.4224	1.8079	1600	0.4270	0.4809
0.4596	1.9209	1700	0.4287	0.4812
0.4621	2.0339	1800	0.4185	0.4694
0.4703	2.1469	1900	0.4177	0.4731
0.4189	2.2599	2000	0.4091	0.4690
0.4126	2.3729	2100	0.4194	0.4782
0.4236	2.4859	2200	0.4183	0.4719
0.4504	2.5989	2300	0.4069	0.4690
0.4546	2.7119	2400	0.4041	0.4818
0.4969	2.8249	2500	0.4029	0.4674
0.4112	2.9379	2600	0.4043	0.4619
0.4431	3.0508	2700	0.4145	0.4637
0.4145	3.1638	2800	0.4048	0.4640
0.3467	3.2768	2900	0.4103	0.4600
0.4648	3.3898	3000	0.3990	0.4534
0.3663	3.5028	3100	0.3999	0.4619
0.4295	3.6158	3200	0.3953	0.4558
0.4842	3.7288	3300	0.3928	0.4588
0.341	3.8418	3400	0.4005	0.4525
0.4149	3.9548	3500	0.3926	0.4477
0.4282	4.0678	3600	0.3984	0.4460
0.4846	4.1808	3700	0.3904	0.4557
0.4152	4.2938	3800	0.3873	0.4533
0.4098	4.4068	3900	0.3895	0.4510
0.4065	4.5198	4000	0.3867	0.4466
0.3712	4.6328	4100	0.3886	0.4482
0.3766	4.7458	4200	0.3858	0.4440
0.3423	4.8588	4300	0.3863	0.4407
0.3358	4.9718	4400	0.3863	0.4405

Framework versions

Transformers 4.57.0
Pytorch 2.8.0+cu128
Datasets 4.1.1
Tokenizers 0.22.1

Task: Automatic Speech Recognition
Library: transformers
Type: wav2vec2
License: cc-by-nc-4.0
Created: Sep 28, 2025
Updated: Oct 7, 2025