train_boolq_123_1762598659

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

• Loss: 0.4221
• Num Input Tokens Seen: 42678144

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: 0.001
• 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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Input Tokens Seen
0.3124	1.0	2121	0.3276	2131904
0.308	2.0	4242	0.3278	4264768
0.3357	3.0	6363	0.3351	6404896
0.3423	4.0	8484	0.3320	8537088
0.3169	5.0	10605	0.3238	10677088
0.2902	6.0	12726	0.3203	12814080
0.2245	7.0	14847	0.3191	14950432
0.3277	8.0	16968	0.3129	17082336
0.2917	9.0	19089	0.3303	19211360
0.2396	10.0	21210	0.3094	21342336
0.3376	11.0	23331	0.3155	23472352
0.2367	12.0	25452	0.3056	25602144
0.2787	13.0	27573	0.3165	27739072
0.2327	14.0	29694	0.3215	29880544
0.188	15.0	31815	0.3245	32013760
0.2431	16.0	33936	0.3384	34138272
0.2415	17.0	36057	0.3464	36269152
0.1752	18.0	38178	0.3552	38408800
0.2795	19.0	40299	0.3594	40541312
0.1345	20.0	42420	0.3606	42678144

Framework versions

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