Hugging Face's logo

Aharneish
/
mistral-test_1

PEFT
Safetensors
trl
sft
Generated from Trainer
Model card Files Community
mistral-test_1 / README.md
Aharneish's picture
Aharneish
End of training
7c148ef verified
preview code
|
raw
history blame contribute delete
2.07 kB
metadata
license: apache-2.0
library_name: peft
tags:
  - trl
  - sft
  - generated_from_trainer
base_model: mistralai/Mistral-7B-Instruct-v0.1
model-index:
  - name: mistral-test_1
    results: []

mistral-test_1

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: 2.2947

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: 1e-05
  • train_batch_size: 4
  • eval_batch_size: 4
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 5
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
1.8128 1.36 2500 2.1855
1.7961 1.63 3000 2.1808
1.7701 1.9 3500 2.2271
1.7186 2.17 4000 2.2265
1.6784 2.44 4500 2.2547
1.6692 2.71 5000 2.2547
1.6752 2.98 5500 2.2470
1.6206 3.26 6000 2.2842
1.599 3.53 6500 2.2663
1.6054 3.8 7000 2.2560
1.593 4.07 7500 2.3039
1.5771 4.34 8000 2.2797
1.5636 4.61 8500 2.2915
1.5551 4.88 9000 2.2947

Framework versions

  • PEFT 0.9.0
  • Transformers 4.38.1
  • Pytorch 2.1.2
  • Datasets 2.16.0
  • Tokenizers 0.15.2