jialicheng/superb-si_hubert-large

superb_si_42

This model is a fine-tuned version of facebook/hubert-large-ll60k on the superb dataset. It achieves the following results on the evaluation set:

• Loss: 3.5572
• Accuracy: 0.2026

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: 4
• 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
6.9439	1.0	4324	7.0073	0.0019
5.7711	2.0	8648	5.9516	0.0101
5.3911	3.0	12972	5.6571	0.0167
4.9798	4.0	17296	5.1595	0.0365
4.577	5.0	21620	4.6393	0.0671
4.23	6.0	25944	4.2066	0.1057
3.8768	7.0	30268	3.8820	0.1564
3.6853	8.0	34592	3.6688	0.1811
3.5543	9.0	38916	3.5760	0.1973
3.4964	10.0	43240	3.5572	0.2026

Framework versions

• Transformers 4.40.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1