emptyngton's picture
emptyngton/mistral-instruct-generation
ed3fafd
metadata
license: apache-2.0
base_model: mistralai/Mistral-7B-Instruct-v0.1
tags:
  - generated_from_trainer
model-index:
  - name: mistral_instruct_generation
    results: []

mistral_instruct_generation

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

  • Loss: 1.3094

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: 4
  • 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
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
1.5255 0.02 20 1.3722
1.4519 0.03 40 1.3348
1.4296 0.05 60 1.3226
1.4289 0.06 80 1.3142
1.4278 0.08 100 1.3094

Framework versions

  • Transformers 4.36.0.dev0
  • Pytorch 2.1.0+cu118
  • Datasets 2.15.0
  • Tokenizers 0.15.0