Llama-3.3-70B-Instruct-3d-1M-100K-0.1-reverse-padzero-plus-mul-sub-99-512D-3L-4H-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	0.0640	500	nan
0.0	0.1280	1000	nan
0.0	0.1920	1500	nan
0.0	0.2560	2000	nan
0.0	0.3200	2500	nan
0.0	0.3840	3000	nan
0.0	0.4480	3500	nan
0.0	0.5120	4000	nan
0.0	0.5760	4500	nan
0.0	0.6400	5000	nan
0.0	0.7040	5500	nan
0.0	0.7680	6000	nan
0.0	0.8319	6500	nan
0.0	0.8959	7000	nan
0.0	0.9599	7500	nan
0.0	1.0239	8000	nan
0.0	1.0879	8500	nan
0.0	1.1519	9000	nan
0.0	1.2159	9500	nan
0.0	1.2799	10000	nan
0.0	1.3439	10500	nan
0.0	1.4079	11000	nan
0.0	1.4719	11500	nan
0.0	1.5359	12000	nan
0.0	1.5999	12500	nan
0.0	1.6639	13000	nan
0.0	1.7279	13500	nan
0.0	1.7919	14000	nan
0.0	1.8559	14500	nan
0.0	1.9199	15000	nan
0.0	1.9839	15500	nan
0.0	2.0479	16000	nan
0.0	2.1119	16500	nan
0.0	2.1759	17000	nan
0.0	2.2399	17500	nan
0.0	2.3039	18000	nan
0.0	2.3678	18500	nan
0.0	2.4318	19000	nan
0.0	2.4958	19500	nan
0.0	2.5598	20000	nan
0.0	2.6238	20500	nan
0.0	2.6878	21000	nan
0.0	2.7518	21500	nan
0.0	2.8158	22000	nan
0.0	2.8798	22500	nan
0.0	2.9438	23000	nan
0.0	3.0078	23500	nan
0.0	3.0718	24000	nan
0.0	3.1358	24500	nan
0.0	3.1998	25000	nan
0.0	3.2638	25500	nan
0.0	3.3278	26000	nan
0.0	3.3918	26500	nan
0.0	3.4558	27000	nan
0.0	3.5198	27500	nan
0.0	3.5838	28000	nan
0.0	3.6478	28500	nan
0.0	3.7118	29000	nan
0.0	3.7758	29500	nan
0.0	3.8398	30000	nan
0.0	3.9038	30500	nan
0.0	3.9677	31000	nan
0.0	4.0317	31500	nan
0.0	4.0957	32000	nan
0.0	4.1597	32500	nan
0.0	4.2237	33000	nan
0.0	4.2877	33500	nan
0.0	4.3517	34000	nan
0.0	4.4157	34500	nan
0.0	4.4797	35000	nan
0.0	4.5437	35500	nan
0.0	4.6077	36000	nan
0.0	4.6717	36500	nan
0.0	4.7357	37000	nan
0.0	4.7997	37500	nan
0.0	4.8637	38000	nan
0.0	4.9277	38500	nan
0.0	4.9917	39000	nan

Framework versions

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