train_cb_1757340266

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: 1.9157
• Num Input Tokens Seen: 359824

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: 101112
• 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
0.5529	0.5088	29	0.2530	19872
0.425	1.0175	58	0.3134	36432
0.1547	1.5263	87	0.2663	53680
0.1254	2.0351	116	0.1659	72160
0.0558	2.5439	145	0.0878	91904
0.354	3.0526	174	0.0978	108856
0.3631	3.5614	203	0.0429	128056
0.0179	4.0702	232	0.0784	146952
0.097	4.5789	261	0.0163	165128
0.0996	5.0877	290	0.0240	183224
0.0087	5.5965	319	0.0313	202424
0.0045	6.1053	348	0.0421	220000
0.0108	6.6140	377	0.0190	238272
0.0252	7.1228	406	0.0611	255984
0.0008	7.6316	435	0.0516	275536
0.0008	8.1404	464	0.0267	293296
0.01	8.6491	493	0.0293	312304
0.0004	9.1579	522	0.0468	329216
0.0003	9.6667	551	0.0471	346944

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.8.0+cu128
• Datasets 3.6.0
• Tokenizers 0.21.1