rlt_2409_1450

This model is a fine-tuned version of svalabs/gbert-large-zeroshot-nli on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0518
• F1: 0.9826

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: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	F1
No log	0.99	36	0.5165	0.8542
No log	1.99	72	0.1459	0.9599
No log	2.99	108	0.0733	0.9882
No log	3.99	144	0.1385	0.9502
No log	4.99	180	0.0948	0.9806
No log	5.99	216	0.0699	0.9822
No log	6.99	252	0.0582	0.9859
No log	7.99	288	0.0340	0.9933
No log	8.99	324	0.0475	0.9826
No log	9.99	360	0.0518	0.9826

Framework versions

• Transformers 4.22.1
• Pytorch 1.12.1+cu113
• Datasets 2.5.1
• Tokenizers 0.12.1