hubert-base-ls960-fsc-h

gokuls/hubert-base-ls960-fsc-h

hubert-base-ls960-fsc-h is a machine learning model from gokuls. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.

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

Downloads · 30 days

36% of all-time downloads

All-time downloads

Public

Parameters

91.1M

4.4 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors365 MB · 100%

At a glance

License: apache-2.0
Model type: hubert
Access: Public
Created: Oct 6, 2024
Updated: Oct 6, 2024
SHA: 889f27df

Base models

facebook/hubert-base-ls960

Model card

From the Hugging Face model README

hubert-base-ls960-fsc-h

This model is a fine-tuned version of facebook/hubert-base-ls960 on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 1.7232
Accuracy: 0.2557

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: 0.0005
train_batch_size: 48
eval_batch_size: 48
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 192
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.9959	120	1.7232	0.2557
No log	2.0	241	1.7236	0.2557
No log	2.9959	361	1.7201	0.2383
No log	4.0	482	1.7198	0.2557
No log	4.9959	602	1.7183	0.2557
No log	6.0	723	1.7203	0.2557

Framework versions

Transformers 4.43.3
Pytorch 2.2.2+cu121
Datasets 2.18.0
Tokenizers 0.19.1

Type: hubert
License: apache-2.0
Created: Oct 6, 2024
Updated: Oct 6, 2024