neel-nanonets/goavanto_1

llm3br256

This model is a fine-tuned version of meta-llama/Llama-3.2-3B-Instruct on the goavanto-oneshot-train dataset. It achieves the following results on the evaluation set:

• Loss: 0.0060

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: 0.0001
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• 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: 5.0

Training results

Training Loss	Epoch	Step	Validation Loss
0.0165	0.2548	50	0.0163
0.0137	0.5096	100	0.0120
0.01	0.7643	150	0.0105
0.0094	1.0191	200	0.0094
0.0086	1.2739	250	0.0088
0.0082	1.5287	300	0.0081
0.0072	1.7834	350	0.0076
0.0075	2.0382	400	0.0075
0.0056	2.2930	450	0.0071
0.0054	2.5478	500	0.0069
0.0048	2.8025	550	0.0067
0.0037	3.0573	600	0.0066
0.0029	3.3121	650	0.0062
0.0031	3.5669	700	0.0064
0.0029	3.8217	750	0.0060
0.0025	4.0764	800	0.0062
0.0021	4.3312	850	0.0063
0.0024	4.5860	900	0.0061
0.0025	4.8408	950	0.0061

Framework versions

• PEFT 0.12.0
• Transformers 4.46.1
• Pytorch 2.4.0+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3