llama-3.1-correctness-reg-adapter

This model is a fine-tuned version of meta-llama/Llama-3.1-8B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.1128
• Mse: 1.1128
• Rmse: 1.0549
• Mae: 0.8341
• R2: 0.3010
• Rounded Accuracy: 0.3656
• Mae Class 0: 1.7920
• Mse Class 0: 4.0470
• Mae Class 1: 1.3730
• Mse Class 1: 2.4817
• Mae Class 2: 0.9256
• Mse Class 2: 1.1026
• Mae Class 3: 0.4746
• Mse Class 3: 0.3342
• Mae Class 4: 0.7528
• Mse Class 4: 0.8202
• Pred Count 0: 42
• Pred Percent 0: 1.0332
• Pred Count 1: 208
• Pred Percent 1: 5.1169
• Pred Count 2: 613
• Pred Percent 2: 15.0800
• Pred Count 3: 2137
• Pred Percent 3: 52.5707
• Pred Count 4: 1065
• Pred Percent 4: 26.1993

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: 2e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.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: 1
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Mse	Rmse	Mae	R2	Rounded Accuracy	Mae Class 0	Mse Class 0	Mae Class 1	Mse Class 1	Mae Class 2	Mse Class 2	Mae Class 3	Mse Class 3	Mae Class 4	Mse Class 4	Pred Count 0	Pred Percent 0	Pred Count 1	Pred Percent 1	Pred Count 2	Pred Percent 2	Pred Count 3	Pred Percent 3	Pred Count 4	Pred Percent 4
1.3585	0.2460	500	1.4487	1.4487	1.2036	0.9755	0.0901	0.2615	2.1072	4.9013	1.4185	2.4644	0.7422	0.8469	0.3903	0.3569	1.0780	1.4468	44	1.0824	142	3.4932	901	22.1648	2811	69.1513	167	4.1082
1.4516	0.4920	1000	1.1929	1.1929	1.0922	0.8503	0.2508	0.3811	1.7873	3.9885	1.3354	2.4576	0.9463	1.1696	0.5475	0.4553	0.7526	0.9228	44	1.0824	288	7.0849	621	15.2768	1838	45.2153	1274	31.3407
1.2905	0.7381	1500	1.1512	1.1512	1.0729	0.8684	0.2769	0.2940	1.6919	3.6392	1.2749	2.1948	0.8302	0.8937	0.3947	0.2967	0.9209	1.0826	43	1.0578	256	6.2977	659	16.2116	2810	69.1267	297	7.3063
1.2888	0.9841	2000	1.1128	1.1128	1.0549	0.8341	0.3010	0.3656	1.7920	4.0470	1.3730	2.4817	0.9256	1.1026	0.4746	0.3342	0.7528	0.8202	42	1.0332	208	5.1169	613	15.0800	2137	52.5707	1065	26.1993

Framework versions

• PEFT 0.13.2
• Transformers 4.49.0
• Pytorch 2.5.1+cu124
• Datasets 3.3.2
• Tokenizers 0.21.1