pattonma's picture
ai-maker-space/llama381binstruct_summarize_short
21a2131 verified
metadata
base_model: NousResearch/Meta-Llama-3.1-8B-Instruct
datasets:
  - generator
library_name: peft
license: llama3.1
tags:
  - trl
  - sft
  - generated_from_trainer
model-index:
  - name: llama381binstruct_summarize_short
    results: []

llama381binstruct_summarize_short

This model is a fine-tuned version of NousResearch/Meta-Llama-3.1-8B-Instruct on the generator dataset. It achieves the following results on the evaluation set:

  • Loss: 2.0084

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: 1
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 30
  • training_steps: 500

Training results

Training Loss Epoch Step Validation Loss
1.7525 1.25 25 1.5271
0.7826 2.5 50 1.4348
0.395 3.75 75 1.3611
0.2184 5.0 100 1.4174
0.103 6.25 125 1.5257
0.0332 7.5 150 1.7178
0.0237 8.75 175 1.7389
0.026 10.0 200 1.6446
0.0165 11.25 225 1.8576
0.007 12.5 250 1.7998
0.0048 13.75 275 1.8755
0.0061 15.0 300 1.8949
0.003 16.25 325 1.9242
0.002 17.5 350 1.9594
0.0025 18.75 375 1.9789
0.0022 20.0 400 1.9908
0.002 21.25 425 1.9993
0.0017 22.5 450 2.0042
0.0021 23.75 475 2.0071
0.0024 25.0 500 2.0084

Framework versions

  • PEFT 0.12.0
  • Transformers 4.44.2
  • Pytorch 2.4.0+cu121
  • Datasets 3.0.0
  • Tokenizers 0.19.1