replica-full-heavy-barc-ttft

This model is a fine-tuned version of barc0/engineer1-heavy-barc-llama3.1-8b-ins-fft-transduction_lr1e-5_epoch3 on the barc0/transduction_formatted_test_time_finetune_for_evaluation, the barc0/transduction_formatted_rearc_dataset_100k and the barc0/transduction_heavy_100k_jsonl datasets. It achieves the following results on the evaluation set:

• Loss: 0.0342

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: 0.0002
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 2
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• total_eval_batch_size: 4
• 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
0.0307	1.0	2668	0.0473
0.0584	2.0	5336	0.0337
0.0015	3.0	8004	0.0342

Framework versions

• PEFT 0.13.2
• Transformers 4.45.0.dev0
• Pytorch 2.4.0+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1