CoolFace
Modelpublic

wintercoming6/mistral-journal-finetune-aipi

sourceHugging Faceapache-2.0updated 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. -->

mistral-journal-finetune-aipi

This model is a fine-tuned version of mistralai/Mistral-7B-v0.1 on an unknown dataset. It achieves the following results on the evaluation set:

  • —Loss: 2.7849

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: 2.5e-05
  • —trainbatchsize: 2
  • —evalbatchsize: 8
  • —seed: 42
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 1
  • —training_steps: 500
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation Loss
1.35410.6757251.4800
0.86441.3514501.4804
0.84492.0270751.5426
0.55012.70271001.6733
0.26913.37841251.8077
0.21874.05411501.9603
0.14084.72971752.0132
0.08925.40542002.2048
0.08916.08112252.2487
0.06336.75682502.2303
0.05167.43242752.2637
0.04798.10813002.2366
0.04868.78383252.3266
0.04359.45953502.5567
0.038910.13513752.6020
0.0410.81084002.7252
0.034111.48654252.6801
0.035712.16224502.7342
0.036312.83784752.7706
0.034213.51355002.7849

Framework versions

  • —PEFT 0.10.1.dev0
  • —Transformers 4.40.0.dev0
  • —Pytorch 2.2.1+cu121
  • —Datasets 2.18.0
  • —Tokenizers 0.19.0