FT-S2

JuanPajero/FT-S2

FT-S2 is a machine learning model from JuanPajero. Use it for the machine learning task on the model card, and read the license before you ship it in a product.

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

Downloads · 30 days

2% of all-time downloads

All-time downloads

180

Public

Repo size

11 GB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.pt1.4 GB · 55%

At a glance

Model type: wav2vec2
Access: Public
Created: Mar 10, 2026
Updated: Mar 14, 2026
SHA: df39d6e4

Model card

From the Hugging Face model README

FT-S2

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.6682
Wer: 0.2795

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-05
train_batch_size: 16
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 3
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
2.2181	0.3	4300	0.8826	0.4381
0.5238	0.6	8600	0.7311	0.3478
0.4351	0.9	12900	0.7427	0.3232
0.3871	1.2	17200	0.6747	0.3098
0.3601	1.5	21500	0.6746	0.2992
0.3471	1.79	25800	0.6689	0.2918
0.3299	2.09	30100	0.6708	0.2862
0.3111	2.39	34400	0.6614	0.2852
0.3032	2.69	38700	0.6848	0.2815
0.2973	2.99	43000	0.6682	0.2795

Framework versions

Transformers 4.17.0
Pytorch 2.10.0+cu128
Datasets 3.6.0
Tokenizers 0.22.2

Type: wav2vec2
Created: Mar 10, 2026
Updated: Mar 14, 2026