llama-7b-sst-2

This model is a fine-tuned version of meta-llama/Llama-3.1-8B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2342
• Accuracy: 0.9117
• F1: 0.9126

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: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.1900	100	0.4794	0.7810	0.7975
No log	0.3800	200	0.3449	0.8555	0.8591
No log	0.5701	300	0.3052	0.8796	0.8832
No log	0.7601	400	0.2865	0.8819	0.8847
1.638	0.9501	500	0.2738	0.8922	0.8932
1.638	1.1387	600	0.2604	0.9014	0.9025
1.638	1.3287	700	0.2683	0.9060	0.9040
1.638	1.5188	800	0.2525	0.9106	0.9099
1.638	1.7088	900	0.2596	0.9083	0.9119
0.9792	1.8988	1000	0.2422	0.9128	0.9126
0.9792	2.0874	1100	0.2426	0.9106	0.9101
0.9792	2.2774	1200	0.2465	0.9151	0.9176
0.9792	2.4675	1300	0.2411	0.9117	0.9118
0.9792	2.6575	1400	0.2356	0.9106	0.9114
0.8907	2.8475	1500	0.2342	0.9117	0.9126

Framework versions

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