train_cb_789_1760637866

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: 0.2356
• Num Input Tokens Seen: 622560

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: 1e-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
0.4162	2.0	100	0.2347	62752
0.2069	4.0	200	0.3462	126016
0.1345	6.0	300	0.2122	188192
0.31	8.0	400	0.2302	249984
0.0348	10.0	500	0.2131	311904
0.0007	12.0	600	0.2538	374240
0.0002	14.0	700	0.2395	436160
0.0002	16.0	800	0.2310	498688
0.0002	18.0	900	0.2360	560544
0.0002	20.0	1000	0.2356	622560

Framework versions

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