OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Huatuo Llama Med Chinese

Huatuo Llama Med Chinese is an open-source research repository containing large language models instruction-tuned with Chinese medical knowledge, along with model weights and scripts for inference and fine-tuning. Its documented model families include LLaMA, Alpaca-Chinese, Bloom and Huoz(i). The project describes building instruction data from medical knowledge graphs and literature with help from the ChatGPT API.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Huatuo Llama Med Chinese — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you need a research codebase and model artifacts for Chinese medical question-answering experiments, rather than a hosted consumer medical service.

Who it is for

  • Researchers prototyping Chinese-language medical question-answering systems.
  • Teams comparing medical responses across language-model families.
  • Developers studying knowledge-grounded generation or adapting models with their own formatted data.

Documented uses

  • Run inference using a selected base model, LoRA weights, test data and prompt template.
  • Compare knowledge-base and medical-literature inference workflows, including single- and multi-turn literature scripts.
  • Prepare a custom dataset in the referenced format and invoke the fine-tuning script.
  • Study the repository’s three-stage knowledge-tuning approach: identify retrieval parameters, retrieve knowledge, and generate an answer using it.
Before you choose

Requirements & limits

The public repository identifies its code license as Apache-2.0. Its README recommends Python 3.9 or later.

  • Inference requires compatible base-model and LoRA-weight paths, data formatted for the scripts, and a prompt template suited to the model.
  • The README provides distinct scripts for medical-knowledge inference and single- or multi-turn literature inference.
  • The project warns that its training examples and datasets have quality limitations and may contain errors or gaps.
  • The repository describes half-precision LoRA fine-tuning as its approach to balancing computing resources and model performance.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Run inference with a repository model

Prepare the documented inputs and run an inference script for Chinese medical question-answering research.

Before you start

  • Python environment; the repository recommends Python 3.9 or later.
  • Install dependencies with pip install -r requirements.txt.
  • A compatible base-model path, downloaded LoRA weights, inference-data path and model-appropriate prompt template.

Steps

  1. Install the repository dependencies in the Python environment.

  2. Download and extract the LoRA weights, then identify the base model and inference data paths.

  3. Choose the prompt template matching the model family and task: the README lists Bloom deployment template, medical-knowledge template, and literature template options.

  4. Replace the paths and template in the inference script configuration, then run the relevant supplied script: the command below, the command below, or the command below.

    Documented command
    bash ./scripts/infer.sh
    Documented command
    bash ./scripts/infer-literature-single.sh
    Documented command
    bash ./scripts/infer-literature-multi.sh

Expected outcome

The supplied inference workflow processes the selected test data using the chosen base model, LoRA weights and prompt template.

Guide limitations

  • The repository notes that prompt-template choice depends on the model.
  • The included training and evaluation data may contain errors or quality limitations; outputs are research material, not clinical guidance.
Documentation 1 ↗
Source-based practical guide

Prepare a custom fine-tuning dataset

Format a dataset for the repository’s documented fine-tuning script.

Before you start

  • A custom dataset formatted according to ./data/llama_data.json.

Steps

  1. Use the supplied LLaMA data file as the format reference when constructing the dataset.

  2. Run the command below as documented in the repository.

    Documented command
    bash ./scripts/finetune.sh

Expected outcome

The repository’s fine-tuning script is invoked with a dataset prepared in the referenced format.

Guide limitations

  • The README does not provide further custom-dataset execution details in the cited instructions.
  • The repository notes limitations in the quality of its training data.
Documentation 1 ↗
COMPARE OPTIONS

Compare options for Model inference and serving

These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.

ProjectDocumented purposePlatformAccess and requirements
Huatuo Llama Med ChineseRun inference using a selected base model, LoRA weights, test data and prompt template.The public repository identifies its code license as Apache-2.0. Its README recommends Python 3.9 or later.
Pulse Of MotionEstimate a video's physical frame rate from its visual dynamics.LinuxThe repository is public. Its README describes a released checkpoint that downloads from Hugging Face on first inference, with a manual download option. The strongest internal model is not fully released; the README directs interested users to contact the project about access. No license information is provided in the repository evidence.
Neural Image AssessmentTrain the model using AVA images and rating annotations.Cloud, Linux, macOS, Windows
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
7 October 2026
Original website title
SCIR-HI/Huatuo-Llama-Med-Chinese · Official homepage title · retained in its source language
How we review sourcesSuggest a correction

Related options

ChatGLM-6B-Med — A related medical ChatGLM model repository that the project identifies as trained using the same data.