Hugging Face's logo

SocialCompUW
/
CHAST

PEFT
Safetensors
lora
Model card Files Community
CHAST / README.md
SocialCompUW's picture
SocialCompUW
Update README.md
795663a verified
preview code
|
raw
history blame
917 Bytes
metadata
license: apache-2.0
library_name: peft
tags:
  - llama-factory
  - lora
base_model: lmsys/vicuna-13b-v1.5-16k
model-index:
  - name: CHAST
    results: []

CHAST (Finetuned Vicuna13b_16K_yaml_lora32)

This model is a fine-tuned version of lmsys/vicuna-13b-v1.5-16k.

Model description

Computes Covert Harms and Social Threats (CHAST) metrics for conversational data.

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 5e-05
  • train_batch_size: 8
  • eval_batch_size: 64
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 32
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • num_epochs: 2.0
  • mixed_precision_training: Native AMP

Framework versions

  • PEFT 0.9.0
  • Transformers 4.37.2
  • Pytorch 2.2.1+cu121
  • Datasets 2.18.0
  • Tokenizers 0.15.2