llama-3.1-8b-cola-lora

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

• Loss: 1.3813
• Accuracy: 0.8791
• Precision: 0.9213
• Recall: 0.9036
• F1: 0.9124

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: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 2
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 10
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.3774	0.9999	4275	0.4300	0.8733	0.8877	0.9366	0.9115
0.3087	2.0	8551	0.6652	0.8752	0.9005	0.9229	0.9116
0.2726	2.9999	12826	0.6527	0.8829	0.9441	0.8843	0.9132
0.1244	4.0	17102	0.8652	0.8676	0.9106	0.8981	0.9043
0.0698	4.9999	21377	1.0741	0.8810	0.9263	0.9008	0.9134
0.1921	6.0	25653	0.8503	0.8772	0.9074	0.9174	0.9123
0.0451	6.9999	29928	1.0125	0.8810	0.9288	0.8981	0.9132
0.0001	8.0	34204	0.9989	0.8752	0.9093	0.9118	0.9106
0.0188	8.9999	38479	1.1504	0.8810	0.9192	0.9091	0.9141
0.0	9.9988	42750	1.3813	0.8791	0.9213	0.9036	0.9124

Framework versions

• PEFT 0.15.0
• Transformers 4.44.2
• Pytorch 2.3.1+cu121
• Datasets 3.0.1
• Tokenizers 0.19.1