ReWiz-7B/README.md

language, license, tags, base_model, datasets, model-index

language	license	tags	base_model	datasets	model-index
en	apache-2.0	text-generation-inference transformers unsloth mistral trl sft theprint	unsloth/mistral-7b-instruct-v0.3-bnb-4bit	KingNish/reasoning-base-20k arcee-ai/EvolKit-20k cognitivecomputations/WizardLM_alpaca_evol_instruct_70k_unfiltered	name results ReWiz-7B task dataset metrics source type name text-generation Text Generation name type args IFEval (0-Shot) HuggingFaceH4/ifeval num_few_shot 0 type value name inst_level_strict_acc and prompt_level_strict_acc 40.48 strict accuracy url name https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=theprint/ReWiz-7B Open LLM Leaderboard task dataset metrics source type name text-generation Text Generation name type args BBH (3-Shot) BBH num_few_shot 3 type value name acc_norm 23.5 normalized accuracy url name https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=theprint/ReWiz-7B Open LLM Leaderboard task dataset metrics source type name text-generation Text Generation name type args MATH Lvl 5 (4-Shot) hendrycks/competition_math num_few_shot 4 type value name exact_match 2.57 exact match url name https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=theprint/ReWiz-7B Open LLM Leaderboard task dataset metrics source type name text-generation Text Generation name type args GPQA (0-shot) Idavidrein/gpqa num_few_shot 0 type value name acc_norm 3.36 acc_norm url name https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=theprint/ReWiz-7B Open LLM Leaderboard task dataset metrics source type name text-generation Text Generation name type args MuSR (0-shot) TAUR-Lab/MuSR num_few_shot 0 type value name acc_norm 16.74 acc_norm url name https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=theprint/ReWiz-7B Open LLM Leaderboard task dataset metrics source type name text-generation Text Generation name type config split args MMLU-PRO (5-shot) TIGER-Lab/MMLU-Pro main test num_few_shot 5 type value name acc 18.56 accuracy url name https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=theprint/ReWiz-7B Open LLM Leaderboard

ReWiz-7B

This is a fine tune of Mistral 7B Instruct (0.3). Half the data was geared towards better reasoning (EvolKit-20k and reasoning-base-20k), the other half will help to de-censor the model (WizardLM data set).

Uploaded model

• Developed by: theprint
• License: apache-2.0
• Finetuned from model : unsloth/mistral-7b-instruct-v0.3-bnb-4bit

This mistral model was trained 2x faster with Unsloth and Huggingface's TRL library.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	17.54
IFEval (0-Shot)	40.48
BBH (3-Shot)	23.50
MATH Lvl 5 (4-Shot)	2.57
GPQA (0-shot)	3.36
MuSR (0-shot)	16.74
MMLU-PRO (5-shot)	18.56