mistral_fine_tuned / README.md
jjayaraj's picture
sri-lasya/gst-taxing-llm
2806553 verified
metadata
base_model: mistralai/Mistral-7B-v0.3
library_name: peft
license: apache-2.0
tags:
  - generated_from_trainer
model-index:
  - name: mistral_fine_tuned
    results: []

Visualize in Weights & Biases

mistral_fine_tuned

This model is a fine-tuned version of mistralai/Mistral-7B-v0.3 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 1.7497

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
  • train_batch_size: 15
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: constant
  • lr_scheduler_warmup_steps: 0.03
  • training_steps: 100

Training results

Training Loss Epoch Step Validation Loss
3.3433 0.1 10 2.8619
2.3427 0.2 20 2.3762
1.9334 0.3 30 2.0743
1.796 0.4 40 1.9626
1.8265 0.5 50 1.9076
1.7717 0.6 60 1.9241
1.597 0.7 70 1.8398
1.5922 0.8 80 1.8226
1.5468 0.9 90 1.7813
1.695 1.0 100 1.7497

Framework versions

  • PEFT 0.11.1
  • Transformers 4.41.0
  • Pytorch 2.3.0+cu121
  • Datasets 2.19.0
  • Tokenizers 0.19.1