engineer-barc-llama3.2-3b-instruct-fft-transduction-lr1e-5_epoch3

This model is a fine-tuned version of meta-llama/Llama-3.2-3B-Instruct on the barc0/transduction_concept_library, the barc0/transduction_angmented_100k-gpt4-description-gpt4omini-code_generated_problems, the barc0/transduction_angmented_100k_gpt4o-mini_generated_problems and the barc0/transduction_rearc_dataset_400k datasets. It achieves the following results on the evaluation set:

• Loss: 0.0255

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 2
• total_train_batch_size: 128
• total_eval_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
0.0341	1.0	2586	0.0395
0.0199	2.0	5172	0.0275
0.0238	3.0	7758	0.0255

Framework versions

• Transformers 4.45.0.dev0
• Pytorch 2.4.0+cu121
• Datasets 3.0.1
• Tokenizers 0.19.1