MedLLaMA-3
This model is developed by Basel Anaya.
π» Usage
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "Reverb/MedLLaMA-3"
messages = [{"role": "user", "content": "What is a large language model?"}]
tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
"text-generation",
model=model,
torch_dtype=torch.float16,
device_map="auto",
)
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
π Evaluation
Tasks |
Version |
Filter |
n-shot |
Metric |
Value |
|
Stderr |
stem |
N/A |
none |
0 |
acc |
0.6466 |
Β± |
0.0056 |
|
|
none |
0 |
acc_norm |
0.6124 |
Β± |
0.0066 |
- medmcqa |
Yaml |
none |
0 |
acc |
0.6118 |
Β± |
0.0075 |
|
|
none |
0 |
acc_norm |
0.6118 |
Β± |
0.0075 |
- medqa_4options |
Yaml |
none |
0 |
acc |
0.6143 |
Β± |
0.0136 |
|
|
none |
0 |
acc_norm |
0.6143 |
Β± |
0.0136 |
- anatomy (mmlu) |
0 |
none |
0 |
acc |
0.7185 |
Β± |
0.0389 |
- clinical_knowledge (mmlu) |
0 |
none |
0 |
acc |
0.7811 |
Β± |
0.0254 |
- college_biology (mmlu) |
0 |
none |
0 |
acc |
0.8264 |
Β± |
0.0317 |
- college_medicine (mmlu) |
0 |
none |
0 |
acc |
0.7110 |
Β± |
0.0346 |
- medical_genetics (mmlu) |
0 |
none |
0 |
acc |
0.8300 |
Β± |
0.0378 |
- professional_medicine (mmlu) |
0 |
none |
0 |
acc |
0.7868 |
Β± |
0.0249 |
- pubmedqa |
1 |
none |
0 |
acc |
0.7420 |
Β± |
0.0196 |
Groups |
Version |
Filter |
n-shot |
Metric |
Value |
|
Stderr |
stem |
N/A |
none |
0 |
acc |
0.6466 |
Β± |
0.0056 |
|
|
none |
0 |
acc_norm |
0.6124 |
Β± |
0.0066 |