metadata

license: apache-2.0
base_model: facebook/hubert-base-ls960
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: hubert-classifier
    results: []

hubert-classifier

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: 2.6644
Accuracy: 0.4576
Precision: 0.4384
Recall: 0.4576
F1: 0.3988
Binary: 0.6165

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: 5e-05
train_batch_size: 32
eval_batch_size: 32
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 128
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Binary
No log	0.86	50	4.1527	0.0484	0.0321	0.0484	0.0218	0.3266
No log	1.72	100	3.7623	0.0605	0.0356	0.0605	0.0253	0.3361
No log	2.59	150	3.5557	0.1211	0.1031	0.1211	0.0914	0.3731
No log	3.45	200	3.3630	0.2131	0.1912	0.2131	0.1632	0.4375
No log	4.31	250	3.1590	0.2809	0.2107	0.2809	0.2096	0.4915
No log	5.17	300	3.0569	0.3414	0.3012	0.3414	0.2825	0.5312
No log	6.03	350	2.8798	0.3995	0.3279	0.3995	0.3298	0.5751
No log	6.9	400	2.7761	0.4189	0.3461	0.4189	0.3522	0.5918
No log	7.76	450	2.7086	0.4383	0.3693	0.4383	0.3689	0.6036
3.4503	8.62	500	2.6644	0.4576	0.4384	0.4576	0.3988	0.6165
3.4503	9.48	550	2.6426	0.4649	0.4400	0.4649	0.4021	0.6211

Framework versions

Transformers 4.38.2
Pytorch 2.3.0
Datasets 2.19.1
Tokenizers 0.15.1