Llama-3.3-70B-Instruct-3d-500K-50K-0.1-reverse-padzero-plus-mul-sub-99-512D-1L-8H-2048I

This model is a fine-tuned version of meta-llama/Llama-3.3-70B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: nan

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: 128
• eval_batch_size: 128
• seed: 42
• 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: 0.05
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0	0	nan
0.0	0.1280	500	nan
0.0	0.2560	1000	nan
0.0	0.3839	1500	nan
0.0	0.5119	2000	nan
0.0	0.6399	2500	nan
0.0	0.7679	3000	nan
0.0	0.8958	3500	nan
0.0	1.0238	4000	nan
0.0	1.1518	4500	nan
0.0	1.2798	5000	nan
0.0	1.4077	5500	nan
0.0	1.5357	6000	nan
0.0	1.6637	6500	nan
0.0	1.7917	7000	nan
0.0	1.9196	7500	nan
0.0	2.0476	8000	nan
0.0	2.1756	8500	nan
0.0	2.3036	9000	nan
0.0	2.4315	9500	nan
0.0	2.5595	10000	nan
0.0	2.6875	10500	nan
0.0	2.8155	11000	nan
0.0	2.9434	11500	nan
0.0	3.0714	12000	nan
0.0	3.1994	12500	nan
0.0	3.3274	13000	nan
0.0	3.4553	13500	nan
0.0	3.5833	14000	nan
0.0	3.7113	14500	nan
0.0	3.8393	15000	nan
0.0	3.9672	15500	nan
0.0	4.0952	16000	nan
0.0	4.2232	16500	nan
0.0	4.3512	17000	nan
0.0	4.4791	17500	nan
0.0	4.6071	18000	nan
0.0	4.7351	18500	nan
0.0	4.8631	19000	nan
0.0	4.9910	19500	nan

Framework versions

• Transformers 4.57.1
• Pytorch 2.9.0+cu128
• Datasets 4.5.0
• Tokenizers 0.22.1