train_cb_1753094175

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

• Loss: 0.1446
• Num Input Tokens Seen: 367864

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: 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: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss	Input Tokens Seen
0.3206	0.5088	29	0.2182	20064
0.1923	1.0175	58	0.2405	37832
0.0899	1.5263	87	0.2054	57288
0.0185	2.0351	116	0.1621	74520
0.2439	2.5439	145	0.2253	93080
0.0828	3.0526	174	0.1446	111928
0.0129	3.5614	203	0.1694	131160
0.0193	4.0702	232	0.1753	150056
0.0002	4.5789	261	0.1988	167208
0.001	5.0877	290	0.2456	186160
0.0001	5.5965	319	0.2628	206000
0.0001	6.1053	348	0.2836	224064
0.0001	6.6140	377	0.2813	243840
0.0	7.1228	406	0.2790	261504
0.0001	7.6316	435	0.2830	280352
0.0001	8.1404	464	0.2781	299344
0.0001	8.6491	493	0.2798	318672
0.0	9.1579	522	0.2795	337480
0.0001	9.6667	551	0.2841	356456

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.7.1+cu126
• Datasets 3.6.0
• Tokenizers 0.21.1