en-tel-v2

This model is a fine-tuned version of unsloth/llama-3.2-3b-instruct-unsloth-bnb-4bit on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4413

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: 0.0003
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• 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: linear
• lr_scheduler_warmup_steps: 100
• num_epochs: 7
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.1951	8	1.6602
1.6739	0.3902	16	1.5405
1.545	0.5854	24	1.4408
1.4322	0.7805	32	1.2243
1.1047	0.9756	40	0.9248
1.1047	1.1707	48	0.6856
0.7786	1.3659	56	0.6219
0.6206	1.5610	64	0.6033
0.5775	1.7561	72	0.5886
0.5845	1.9512	80	0.5775
0.5845	2.1463	88	0.5674
0.5647	2.3415	96	0.5558
0.5466	2.5366	104	0.5470
0.5313	2.7317	112	0.5376
0.5236	2.9268	120	0.5310
0.5236	3.1220	128	0.5237
0.5262	3.3171	136	0.5184
0.4852	3.5122	144	0.5122
0.4957	3.7073	152	0.5062
0.4897	3.9024	160	0.5013
0.4897	4.0976	168	0.4956
0.5019	4.2927	176	0.4904
0.4622	4.4878	184	0.4851
0.4706	4.6829	192	0.4811
0.4403	4.8780	200	0.4763
0.4403	5.0732	208	0.4718
0.427	5.2683	216	0.4678
0.4407	5.4634	224	0.4627
0.4361	5.6585	232	0.4586
0.4234	5.8537	240	0.4538
0.4234	6.0488	248	0.4500
0.4319	6.2439	256	0.4475
0.4102	6.4390	264	0.4448
0.4048	6.6341	272	0.4425
0.3893	6.8293	280	0.4413

Framework versions

• PEFT 0.14.0
• Transformers 4.48.3
• Pytorch 2.6.0+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0