llama-1b-irony

This model is a fine-tuned version of meta-llama/Llama-3.2-1B-Instruct on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6571
• Accuracy: 0.6658
• F1: 0.5112

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: 0.0002
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	1.0	30	0.8135	0.6441	0.5114
No log	2.0	60	0.7566	0.6122	0.5529
No log	3.0	90	0.7246	0.6301	0.5338
No log	4.0	120	0.6850	0.6467	0.5249
No log	5.0	150	0.6739	0.6722	0.5353
No log	6.0	180	0.6852	0.6556	0.4706
No log	7.0	210	0.6446	0.6862	0.5669
No log	8.0	240	0.6489	0.6747	0.5487
No log	9.0	270	0.6449	0.6747	0.5422
No log	10.0	300	0.6571	0.6658	0.5112

Framework versions

• PEFT 0.14.0
• Transformers 4.47.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0