hlillemark/llama3_8b_sft_mc

sft_mc

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on the identity and the data_mc datasets. It achieves the following results on the evaluation set:

• Loss: 2.3011

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: 1e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• total_eval_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss
1.0534	0.7463	50	1.2635
0.8118	1.4925	100	1.3805
0.3889	2.2388	150	1.6007
0.4361	2.9851	200	1.5327
0.265	3.7313	250	1.6067
0.1347	4.4776	300	1.8177
0.0857	5.2239	350	1.9771
0.0709	5.9701	400	1.9008
0.0474	6.7164	450	2.1317
0.0286	7.4627	500	2.2199
0.0091	8.2090	550	2.2086
0.0054	8.9552	600	2.2865
0.0038	9.7015	650	2.3016

Framework versions

• Transformers 4.49.0
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0