sizhkhy/Klystroglobal

llm3br256-v1.5

This model is a fine-tuned version of meta-llama/Llama-3.2-3B-Instruct on the Klystroglobal dataset. It achieves the following results on the evaluation set:

• Loss: 0.0154

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: 0.0001
• train_batch_size: 48
• eval_batch_size: 48
• seed: 42
• 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: 35

Training results

Training Loss	Epoch	Step	Validation Loss
0.0459	2.5	25	0.0362
0.0083	5.0	50	0.0184
0.0058	7.5	75	0.0171
0.0024	10.0	100	0.0154
0.0015	12.5	125	0.0164
0.0038	15.0	150	0.0210
0.0013	17.5	175	0.0282
0.0012	20.0	200	0.0279
0.0013	22.5	225	0.0306

Framework versions

• PEFT 0.12.0
• Transformers 4.46.1
• Pytorch 2.4.0+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3