Views
No views yet
qtype)qtype: Type of medical question (e.g., Treatment, Symptoms).Question: Patient's medical question.Answer: Expert response (from doctors, nurses, and pharmacists).qtype for domain-specific fine-tuning.turquise/MedQA_q8. Below are several ways to use the model:1from llama_cpp import Llama
2
3# Load the model
4llm = Llama.from_pretrained(
5 repo_id="turquise/MedQA_q8",
6 filename="medQA.Q8_0.gguf",
7)
8
9# Query the model
10output = llm(
11 "What is Medullary Sponge Kidney?",
12 max_tokens=512,
13 echo=True
14)
15print(output)1brew install llama.cpp
2
3llama-cli \
4 --hf-repo "turquise/MedQA_q8" \
5 --hf-file medQA.Q8_0.gguf \
6 -p "What is Medullary Sponge Kidney?"1# Download pre-built binary from:
2# https://github.com/ggerganov/llama.cpp/releases
3
4./llama-cli \
5 --hf-repo "turquise/MedQA_q8" \
6 --hf-file medQA.Q8_0.gguf \
7 -p "What is Medullary Sponge Kidney?"1git clone https://github.com/ggerganov/llama.cpp.git
2cd llama.cpp
3cmake -B build -DLLAMA_CURL=ON
4cmake --build build -j --target llama-cli
5
6./build/bin/llama-cli \
7 --hf-repo "turquise/MedQA_q8" \
8 --hf-file medQA.Q8_0.gguf \
9 -p "What is Medullary Sponge Kidney?"1question = "What are the symptoms of diabetes?"
2output = llm(question, max_tokens=512)
3print(output)