5bb528c701240fc5fa01d47fe240fd1c

This model is a fine-tuned version of meta-llama/Llama-3.2-1B on the dair-ai/emotion [split] dataset. It achieves the following results on the evaluation set:

• Loss: 1.2485
• Data Size: 1.0
• Epoch Runtime: 71.6250
• Accuracy: 0.9279
• F1 Macro: 0.8778

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: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro
No log	0	0	13.6535	0	3.5510	0.1431	0.0910
No log	1	500	11.9150	0.0078	3.9430	0.3483	0.0861
No log	2	1000	6.7104	0.0156	5.0038	0.3488	0.0863
No log	3	1500	6.6751	0.0312	8.1858	0.3488	0.0862
No log	4	2000	6.5261	0.0625	10.4282	0.3488	0.0862
0.3991	5	2500	3.3204	0.125	15.6649	0.7263	0.5431
1.7842	6	3000	1.4342	0.25	23.7644	0.9027	0.8434
0.1467	7	3500	2.0934	0.5	40.5385	0.9057	0.8148
0.7009	8.0	4000	0.7656	1.0	74.3755	0.9168	0.8604
0.6109	9.0	4500	0.8347	1.0	75.5958	0.9249	0.8791
0.5781	10.0	5000	0.8394	1.0	69.5502	0.9219	0.8862
0.5151	11.0	5500	1.0978	1.0	70.7940	0.9219	0.8823
0.4165	12.0	6000	1.2485	1.0	71.6250	0.9279	0.8778

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1