430f5ee6f91d23aefefa1f084927fd78

This model is a fine-tuned version of meta-llama/Llama-3.2-3B on the nyu-mll/glue [sst2] dataset. It achieves the following results on the evaluation set:

• Loss: 2.1119
• Data Size: 1.0
• Epoch Runtime: 502.8370
• Accuracy: 0.8530
• F1 Macro: 0.8530
• Rouge1: 0.8519
• Rouge2: 0.0
• Rougel: 0.8530
• Rougelsum: 0.8530

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	Rouge1	Rouge2	Rougel	Rougelsum
No log	0	0	3.2234	0	2.9307	0.5451	0.4794	0.5451	0.0	0.5451	0.5451
No log	1	2104	4.1434	0.0078	7.3677	0.5093	0.3374	0.5093	0.0	0.5093	0.5081
No log	2	4208	1.4480	0.0156	13.6764	0.8461	0.8461	0.8449	0.0	0.8461	0.8461
0.0693	3	6312	1.6838	0.0312	26.1015	0.8160	0.8151	0.8148	0.0	0.8160	0.8160
1.4141	4	8416	1.3928	0.0625	44.1734	0.8657	0.8657	0.8657	0.0	0.8669	0.8657
1.2379	5	10520	1.4230	0.125	74.9340	0.8507	0.8503	0.8507	0.0	0.8507	0.8507
0.9471	6	12624	1.3091	0.25	135.5735	0.8576	0.8575	0.8565	0.0	0.8576	0.8565
0.9611	7	14728	1.4165	0.5	255.2976	0.8669	0.8666	0.8669	0.0	0.8669	0.8669
0.7278	8.0	16832	1.3745	1.0	506.8193	0.8519	0.8515	0.8519	0.0	0.8507	0.8519
0.5976	9.0	18936	1.7965	1.0	501.1181	0.8414	0.8414	0.8414	0.0	0.8414	0.8414
0.4746	10.0	21040	2.1119	1.0	502.8370	0.8530	0.8530	0.8519	0.0	0.8530	0.8530

Framework versions

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