WizardLM-13B-V1.2 Open-Source Language Model - Free Deployment, Focused on Complex Instruction-Following Ability

Wizardlm 13B V1.2

Developed by WizardLMTeam

WizardLM-13B V1.2 is a large language model trained on Llama-2 13b, focusing on complex instruction-following capabilities.

Large Language Model

Transformers

#Complex instruction understanding #Enhanced mathematical reasoning #Optimized code generation

Downloads 989

Release Time : 7/25/2023

Model Overview

This model is part of the WizardLM series, designed to empower large pre-trained language models to better understand and execute complex instructions.

Model Features

Complex instruction understanding

Specially optimized for understanding and executing complex multi-step instructions

High performance

Outperforms similar open-source models in multiple benchmarks

Llama-2 foundation

Fine-tuned based on the powerful Llama-2 13B model

Model Capabilities

Text generation

Instruction understanding and execution

Code generation

Mathematical problem solving

Complex task decomposition

Use Cases

Programming assistance

Code generation

Generate functional code from natural language descriptions

Achieved 36.6 pass@1 on HumanEval benchmark

Mathematical problem solving

Mathematical reasoning

Solve complex mathematical problems

Excellent performance on GSM8K and MATH benchmarks

General AI assistant

Multi-turn dialogue

Engage in complex multi-turn conversations and task decomposition

Achieved 89.17% on AlpacaEval

🚀 WizardLM-13B V1.2 Full-Weight

This is the Full-Weight of the WizardLM-13B V1.2 model, which is trained from Llama-2 13b. It aims to empower large pre - trained language models to follow complex instructions, providing high - quality language interaction capabilities.

✨ Features

Model Release News

WizardCoder Release:
- [2023/08/26] We released WizardCoder - Python - 34B - V1.0, achieving 73.2 pass@1 and surpassing GPT4 (2023/03/15), ChatGPT - 3.5, and Claude2 on the [HumanEval Benchmarks](https://github.com/openai/human - eval). For more details, refer to WizardCoder.
- [2023/06/16] We released WizardCoder - 15B - V1.0, surpassing Claude - Plus (+6.8), Bard (+15.3) and InstructCodeT5+ (+22.3) on the [HumanEval Benchmarks](https://github.com/openai/human - eval). For more details, refer to WizardCoder.
WizardMath Release:
- [08/11/2023] We released WizardMath Models.
- Our WizardMath - 70B - V1.0 model slightly outperforms some closed - source LLMs on the GSM8K, including ChatGPT 3.5, Claude Instant 1 and PaLM 2 540B.
- It achieves 81.6 pass@1 on the [GSM8k Benchmarks](https://github.com/openai/grade - school - math), which is 24.8 points higher than the SOTA open - source LLM.
- It also achieves 22.7 pass@1 on the MATH Benchmarks, which is 9.2 points higher than the SOTA open - source LLM.
WizardLM Release:
- [7/25/2023] We released WizardLM V1.2 models. The WizardLM - 13B - V1.2 is available (Demo_13B - V1.2, Demo_13B - V1.2_bak - 1, [Full Model Weight](https://huggingface.co/WizardLM/WizardLM - 13B - V1.2)). Check out the paper.
- The WizardLM - 13B - V1.2 achieves 7.06 on MT - Bench Leaderboard, 89.17% on [AlpacaEval Leaderboard](https://tatsu - lab.github.io/alpaca_eval/), and 101.4% on WizardLM Eval. (Note: MT - Bench and AlpacaEval are all self - test, will push update and request review. All tests are completed under their official settings.)

Model Performance Tables

WizardCoder Models

Model	Checkpoint	Paper	HumanEval	MBPP	Demo	License
WizardCoder - Python - 34B - V1.0	🤗 HF Link	📃 [WizardCoder]	73.2	61.2	Demo	Llama2
WizardCoder - 15B - V1.0	🤗 HF Link	📃 [WizardCoder]	59.8	50.6	--	OpenRAIL - M
WizardCoder - Python - 13B - V1.0	🤗 HF Link	📃 [WizardCoder]	64.0	55.6	--	Llama2
WizardCoder - Python - 7B - V1.0	🤗 HF Link	📃 [WizardCoder]	55.5	51.6	Demo	Llama2
WizardCoder - 3B - V1.0	🤗 HF Link	📃 [WizardCoder]	34.8	37.4	--	OpenRAIL - M
WizardCoder - 1B - V1.0	🤗 HF Link	📃 [WizardCoder]	23.8	28.6	--	OpenRAIL - M

WizardMath Models

Model	Checkpoint	Paper	GSM8k	MATH	Online Demo	License
WizardMath - 70B - V1.0	🤗 HF Link	📃 [WizardMath]	81.6	22.7	Demo	Llama 2
WizardMath - 13B - V1.0	🤗 HF Link	📃 [WizardMath]	63.9	14.0	Demo	Llama 2
WizardMath - 7B - V1.0	🤗 HF Link	📃 [WizardMath]	54.9	10.7	Demo	Llama 2

WizardLM Models

Model	Checkpoint	Paper	MT - Bench	AlpacaEval	WizardEval	HumanEval	License
WizardLM - 13B - V1.2	🤗 HF Link		7.06	89.17%	101.4%	36.6 pass@1	Llama 2 License
WizardLM - 13B - V1.1	🤗 HF Link		6.76	86.32%	99.3%	25.0 pass@1	Non - commercial
WizardLM - 30B - V1.0	🤗 HF Link		7.01		97.8%	37.8 pass@1	Non - commercial
WizardLM - 13B - V1.0	🤗 HF Link		6.35	75.31%	89.1%	24.0 pass@1	Non - commercial
WizardLM - 7B - V1.0	🤗 HF Link	📃 [WizardLM]			78.0%	19.1 pass@1	Non - commercial

🚀 Quick Start

Model Prompt Format

WizardLM adopts the prompt format from Vicuna and supports multi - turn conversation. The prompt should be as follows:

A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. USER: Hi ASSISTANT: Hello.</s>USER: Who are you? ASSISTANT: I am WizardLM.</s>......

Inference Demo Script

We provide the inference WizardLM demo code here.

📚 Documentation

Citation

Please cite the paper if you use the data or code from WizardLM.

@article{xu2023wizardlm,
  title={Wizardlm: Empowering large language models to follow complex instructions},
  author={Xu, Can and Sun, Qingfeng and Zheng, Kai and Geng, Xiubo and Zhao, Pu and Feng, Jiazhan and Tao, Chongyang and Jiang, Daxin},
  journal={arXiv preprint arXiv:2304.12244},
  year={2023}
}

Dataset Concern

Recently, there have been clear changes in the open - source policy and regulations of our overall organization's code, data, and models. Despite this, we have still worked hard to obtain opening the weights of the model first, but the data involves stricter auditing and is in review with our legal team. Our researchers have no authority to publicly release them without authorization. Thank you for your understanding.

📄 License

The model uses the Llama2 license.