Columbia-NLP_-_gemma-2b-zephyr-sft-4bits

Quantization made by Richard Erkhov.

Github

Discord

Request more models

gemma-2b-zephyr-sft - bnb 4bits

• Model creator: https://huggingface.co/Columbia-NLP/
• Original model: https://huggingface.co/Columbia-NLP/gemma-2b-zephyr-sft/

Original model description:

license: other license_name: gemma-terms-of-use license_link: https://ai.google.dev/gemma/terms base_model: google/gemma-2b tags:

• alignment-handbook
• trl
• sft
• generated_from_trainer datasets:
• HuggingFaceH4/deita-10k-v0-sft model-index:
• name: gemma-2b-zephyr-sft results:
  • task: type: text-generation name: Text Generation dataset: name: AI2 Reasoning Challenge (25-Shot) type: ai2_arc config: ARC-Challenge split: test args: num_few_shot: 25 metrics:
    • type: acc_norm value: 51.88 name: normalized accuracy
  • task: type: text-generation name: Text Generation dataset: name: HellaSwag (10-Shot) type: hellaswag split: validation args: num_few_shot: 10 metrics:
    • type: acc_norm value: 72.63 name: normalized accuracy
  • task: type: text-generation name: Text Generation dataset: name: MMLU (5-Shot) type: cais/mmlu config: all split: test args: num_few_shot: 5 metrics:
    • type: acc value: 42.20 name: accuracy
  • task: type: text-generation name: Text Generation dataset: name: TruthfulQA (0-shot) type: truthful_qa config: multiple_choice split: validation args: num_few_shot: 0 metrics:
    • type: mc2 value: 41.96
  • task: type: text-generation name: Text Generation dataset: name: Winogrande (5-shot) type: winogrande config: winogrande_xl split: validation args: num_few_shot: 5 metrics:
    • type: acc value: 63.85 name: accuracy
  • task: type: text-generation name: Text Generation dataset: name: GSM8k (5-shot) type: gsm8k config: main split: test args: num_few_shot: 5 metrics:
    • type: acc value: 20.09 name: accuracy

---

Model Card for Gemma 2B Zephyr SFT

We trained the google/gemma-2b with deita-10k-v0-sft. We carefully selected the hyper-parameters and masked the user tokens during training to achieve the best supervised fine-tuning performance.

Model description

• Model type: A 2.5B parameter GPT-like model fine-tuned on a mix of publicly available, synthetic datasets.
• Language(s) (NLP): Primarily English
• License: Gemma Terms of Use
• Finetuned from model: google/gemma-2b

License

This model has the same license as the original Gemma model collection

OpenLLM Leaderboard Performance

MT-Bench

GPT-4-0125-preview as Judge