metadata
tags:
- llama-3-8b
- sft
- medical
base_model:
- meta-llama/Meta-Llama-3-8B
license: cc-by-nc-nd-4.0
JSL-MedLlama-3-8B-v1.0
This model is developed by John Snow Labs.
This model is available under a CC-BY-NC-ND license and must also conform to this Acceptable Use Policy. If you need to license this model for commercial use, please contact us at [email protected].
💻 Usage
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "johnsnowlabs/JSL-MedLlama-3-8B-v1.0"
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"])
🏆 Evaluation
Tasks | Version | Filter | n-shot | Metric | Value | Stderr | |
---|---|---|---|---|---|---|---|
stem | N/A | none | 0 | acc | 0.6217 | ± | 0.0057 |
none | 0 | acc_norm | 0.5847 | ± | 0.0066 | ||
- medmcqa | Yaml | none | 0 | acc | 0.5563 | ± | 0.0077 |
none | 0 | acc_norm | 0.5563 | ± | 0.0077 | ||
- medqa_4options | Yaml | none | 0 | acc | 0.6779 | ± | 0.0131 |
none | 0 | acc_norm | 0.6779 | ± | 0.0131 | ||
- anatomy (mmlu) | 0 | none | 0 | acc | 0.6963 | ± | 0.0397 |
- clinical_knowledge (mmlu) | 0 | none | 0 | acc | 0.7509 | ± | 0.0266 |
- college_biology (mmlu) | 0 | none | 0 | acc | 0.7986 | ± | 0.0335 |
- college_medicine (mmlu) | 0 | none | 0 | acc | 0.6590 | ± | 0.0361 |
- medical_genetics (mmlu) | 0 | none | 0 | acc | 0.8500 | ± | 0.0359 |
- professional_medicine (mmlu) | 0 | none | 0 | acc | 0.7868 | ± | 0.0249 |
- pubmedqa | 1 | none | 0 | acc | 0.7380 | ± | 0.0197 |
Groups | Version | Filter | n-shot | Metric | Value | Stderr | |
---|---|---|---|---|---|---|---|
stem | N/A | none | 0 | acc | 0.6217 | ± | 0.0057 |
none | 0 | acc_norm | 0.5847 | ± | 0.0066 |