train_qqp_1752826675

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

• Loss: 0.1218
• Num Input Tokens Seen: 250787112

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: 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: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss	Input Tokens Seen
0.1748	0.5	40933	0.2304	12552544
0.2322	1.0	81866	0.1896	25087944
0.2315	1.5	122799	0.1723	37621672
0.1378	2.0	163732	0.1531	50164864
0.0891	2.5	204665	0.1446	62700096
0.1223	3.0	245598	0.1387	75242048
0.1895	3.5	286531	0.1417	87769248
0.2392	4.0	327464	0.1310	100320328
0.1416	4.5	368397	0.1299	112855464
0.2114	5.0	409330	0.1270	125387608
0.1165	5.5	450263	0.1258	137931704
0.1377	6.0	491196	0.1263	150463800
0.1099	6.5	532129	0.1239	163003576
0.1588	7.0	573062	0.1233	175543400
0.0516	7.5	613995	0.1224	188096200
0.0705	8.0	654928	0.1223	200622552
0.0748	8.5	695861	0.1220	213151000
0.2471	9.0	736794	0.1221	225701792
0.0588	9.5	777727	0.1220	238243968
0.0732	10.0	818660	0.1218	250787112

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.7.1+cu126
• Datasets 3.6.0
• Tokenizers 0.21.1