train_stsb_789_1760637926

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

• Loss: 0.4683
• Num Input Tokens Seen: 8752512

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
0.4133	1.0	1294	0.5303	437440
0.553	2.0	2588	0.5284	875776
0.5114	3.0	3882	0.4840	1313392
0.3267	4.0	5176	0.4683	1753440
0.3727	5.0	6470	0.4904	2191488
0.3267	6.0	7764	0.5586	2629728
0.1728	7.0	9058	0.6214	3066368
0.3543	8.0	10352	0.6405	3502736
0.2155	9.0	11646	0.7220	3940080
0.2394	10.0	12940	0.8526	4376768
0.1277	11.0	14234	0.9907	4813216
0.1973	12.0	15528	1.2096	5251344
0.1222	13.0	16822	1.3117	5689568
0.0101	14.0	18116	1.6009	6127200
0.0016	15.0	19410	1.7984	6564352
0.0096	16.0	20704	1.9734	7002400
0.0376	17.0	21998	2.0502	7438560
0.0002	18.0	23292	2.1225	7876400
0.0001	19.0	24586	2.1361	8314032
0.0004	20.0	25880	2.1389	8752512

Framework versions

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