train_math_qa_42_1760637605

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

• Loss: 0.8015
• Num Input Tokens Seen: 77902976

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: 0.03
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Use 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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Input Tokens Seen
0.7875	1.0	6714	0.8115	3894552
0.8276	2.0	13428	0.8128	7790784
0.8419	3.0	20142	0.8133	11684296
0.822	4.0	26856	0.8200	15578848
0.859	5.0	33570	0.8048	19476576
0.7725	6.0	40284	0.8028	23368392
0.8174	7.0	46998	0.8046	27263880
0.8153	8.0	53712	0.8094	31154960
0.8042	9.0	60426	0.8066	35053760
0.8402	10.0	67140	0.8033	38947216
0.8065	11.0	73854	0.8024	42844400
0.7968	12.0	80568	0.8049	46741816
0.7543	13.0	87282	0.8061	50638456
0.7896	14.0	93996	0.8030	54533112
0.7867	15.0	100710	0.8024	58429624
0.7691	16.0	107424	0.8025	62323952
0.8042	17.0	114138	0.8015	66220752
0.7946	18.0	120852	0.8019	70113640
0.8079	19.0	127566	0.8023	74009160
0.7752	20.0	134280	0.8016	77902976

Framework versions

• PEFT 0.17.1
• Transformers 4.51.3
• Pytorch 2.9.0+cu128
• Datasets 4.0.0
• Tokenizers 0.21.4