README.md

metadata

license: mit
library_name: peft
tags:
  - alignment-handbook
  - trl
  - sft
  - generated_from_trainer
  - trl
  - sft
  - generated_from_trainer
datasets:
  - ChenWu98/skills_red_herring_chat
base_model: HuggingFaceH4/zephyr-7b-beta
model-index:
  - name: skills_red_herring_chat-lora
    results: []

skills_red_herring_chat-lora

This model is a fine-tuned version of HuggingFaceH4/zephyr-7b-beta on the ChenWu98/skills_red_herring_chat dataset. It achieves the following results on the evaluation set:

• Loss: 0.2756

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
• distributed_type: multi-GPU
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 2.0

Training results

Training Loss	Epoch	Step	Validation Loss
1.8942	0.96	9	0.3455
0.2839	1.92	18	0.2756

Framework versions

• PEFT 0.7.1
• Transformers 4.37.2
• Pytorch 2.1.2+cu121
• Datasets 2.14.6
• Tokenizers 0.15.1