CoolFace
Modelpublic

Galatea007/llama381binstruct_summarize_short

sourceHugging Facellama3.1updated 2y agoView on Hugging Face
0likes4downloads
Model Card

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

llama381binstructsummarizeshort

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: 2.7961

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
  • —trainbatchsize: 1
  • —evalbatchsize: 8
  • —seed: 42
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 30
  • —training_steps: 500

Training results

Training LossEpochStepValidation Loss
1.58311.25251.6479
0.80862.5501.6711
0.33333.75751.8202
0.18345.01001.9703
0.0836.251252.2035
0.03817.51502.3948
0.02128.751752.4440
0.019710.02002.4377
0.018811.252252.3988
0.009412.52502.5389
0.006213.752752.5661
0.004815.03002.6391
0.003616.253252.6828
0.002117.53502.7307
0.001718.753752.7560
0.002320.04002.7722
0.002421.254252.7839
0.00222.54502.7908
0.002123.754752.7946
0.001825.05002.7961

Framework versions

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