CoolFace
Modelpublic

nrishabh/llama3-8b-instruct-qlora-mini

sourceHugging Facemitupdated 2y agoView on Hugging Face
0likes7downloads
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. -->

llama3-8b-instruct-qlora-mini

This model is a fine-tuned version of LoftQ/Meta-Llama-3-8B-Instruct-4bit-64rank on an unknown dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.8668

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: 2e-05
  • —trainbatchsize: 8
  • —evalbatchsize: 8
  • —seed: 42
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: cosine
  • —num_epochs: 30

Training results

Training LossEpochStepValidation Loss
2.39091.0531.5473
2.19372.01061.2690
2.09153.01591.0977
1.99274.02121.0320
1.90585.02651.0046
1.80326.03180.9885
1.66887.03710.9754
1.52158.04240.9745
1.36179.04770.9640
1.207410.05300.9579
1.042911.05830.9441
0.901312.06360.9355
0.796913.06890.9278
0.709214.07420.9171
0.627215.07950.9070
0.568816.08480.9052
0.512817.09010.8942
0.46918.09540.8894
0.429419.010070.8871
0.395320.010600.8807
0.37121.011130.8756
0.353322.011660.8750
0.333523.012190.8730
0.321224.012720.8699
0.310825.013250.8687
0.308926.013780.8676
0.303127.014310.8678
0.301428.014840.8675
0.301329.015370.8666
0.297830.015900.8668

Framework versions

  • —PEFT 0.10.0
  • —Transformers 4.40.0
  • —Pytorch 2.2.1+cu121
  • —Datasets 2.18.0
  • —Tokenizers 0.19.1