train_cb_789_1760637869

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: 0.0699
• Num Input Tokens Seen: 711112

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: 789
• 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
0.1004	1.0	57	0.1302	35448
0.2242	2.0	114	0.0921	70496
0.1286	3.0	171	0.0699	106416
0.1525	4.0	228	0.1624	142480
0.0366	5.0	285	0.3202	177224
0.0001	6.0	342	0.1976	212000
0.0	7.0	399	0.2687	248272
0.0	8.0	456	0.2064	284248
0.0001	9.0	513	0.2067	319488
0.0	10.0	570	0.2093	354472
0.0	11.0	627	0.2140	389408
0.0	12.0	684	0.2185	425328
0.0	13.0	741	0.2167	461216
0.0	14.0	798	0.2229	496704
0.0	15.0	855	0.2184	532504
0.0	16.0	912	0.2199	567952
0.0	17.0	969	0.2230	603760
0.0	18.0	1026	0.2177	639784
0.0	19.0	1083	0.2234	675800
0.0	20.0	1140	0.2218	711112

Framework versions

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