Royal-lobster's picture
Update README.md
e25c56f
metadata
license: apache-2.0
base_model: teknium/OpenHermes-2.5-Mistral-7B
tags:
  - generated_from_trainer
model-index:
  - name: out
    results: []

Built with Axolotl

Enhanced Slither Auditor

This model is a fine-tuned version of teknium/OpenHermes-2.5-Mistral-7B on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.1923

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: 5e-06
  • train_batch_size: 2
  • eval_batch_size: 2
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 2
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 16
  • total_eval_batch_size: 4
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 10
  • num_epochs: 2

Training results

Training Loss Epoch Step Validation Loss
1.1498 0.0 1 1.1953
0.321 0.1 31 0.3176
0.2693 0.2 62 0.2712
0.2701 0.31 93 0.2523
0.27 0.41 124 0.2362
0.2244 0.51 155 0.2284
0.2227 0.61 186 0.2260
0.2167 0.71 217 0.2171
0.2098 0.81 248 0.2082
0.1842 0.92 279 0.2047
0.1917 1.02 310 0.2013
0.1639 1.12 341 0.1982
0.1835 1.22 372 0.1968
0.1666 1.32 403 0.1953
0.1694 1.43 434 0.1932
0.1461 1.53 465 0.1929
0.1535 1.63 496 0.1927
0.1419 1.73 527 0.1925
0.1612 1.83 558 0.1923
0.1857 1.93 589 0.1923

Framework versions

  • Transformers 4.34.1
  • Pytorch 2.0.1+cu118
  • Datasets 2.14.6
  • Tokenizers 0.14.1