train_hellaswag_123_1760637740

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

• Loss: 0.4621
• Num Input Tokens Seen: 218506144

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: 123
• 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
0.4688	1.0	8979	0.4636	10932896
0.4573	2.0	17958	0.4627	21856400
0.46	3.0	26937	0.4627	32797696
0.4644	4.0	35916	0.4622	43715520
0.4609	5.0	44895	0.4626	54639040
0.4619	6.0	53874	0.4626	65562352
0.4554	7.0	62853	0.4621	76495264
0.4586	8.0	71832	0.4625	87424000
0.4637	9.0	80811	0.4625	98355744
0.4647	10.0	89790	0.4625	109279616
0.4614	11.0	98769	0.4625	120190896
0.4631	12.0	107748	0.4623	131118336
0.4628	13.0	116727	0.4624	142033584
0.4598	14.0	125706	0.4621	152960704
0.4625	15.0	134685	0.4621	163884192
0.461	16.0	143664	0.4624	174816592
0.4611	17.0	152643	0.4621	185740864
0.4607	18.0	161622	0.4621	196657440
0.4621	19.0	170601	0.4621	207581424
0.4597	20.0	179580	0.4621	218506144

Framework versions

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