|
|
from groq import Groq |
|
|
client = Groq(api_key="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") |
|
|
|
|
|
|
|
|
def ResearchDeepDive(content): |
|
|
|
|
|
|
|
|
SYSTEM_PROMPT=""" |
|
|
You are a Medical Domain Expert Reasoning Agent. |
|
|
Study the provided context and use first-principles and Socratic reasoning to uncover its core meaning. |
|
|
|
|
|
Instructions: |
|
|
|
|
|
Output only Question-Answer pairs, based strictly on the context. |
|
|
|
|
|
Each Question must be followed by its Answer. |
|
|
|
|
|
Use simple, clear language β no legal jargon. |
|
|
|
|
|
Keep it concise (Questions β€ 15 words, Answers β€ 25 words). |
|
|
|
|
|
Produce 3-5 pairs max. |
|
|
|
|
|
Format exactly like this: |
|
|
|
|
|
Question: β¦ |
|
|
Answer: β¦ |
|
|
|
|
|
Context will be provided by User. |
|
|
|
|
|
""" |
|
|
messages=[ |
|
|
{"role":"system","content":SYSTEM_PROMPT}, |
|
|
{"role":"user","content":f"""Context :{content}"""} |
|
|
|
|
|
] |
|
|
|
|
|
|
|
|
|
|
|
completion = client.chat.completions.create( |
|
|
model="llama-3.1-8b-instant", |
|
|
messages=messages, |
|
|
temperature=1, |
|
|
max_completion_tokens=8192, |
|
|
top_p=1, |
|
|
|
|
|
stream=False, |
|
|
stop=None, |
|
|
tools=[] |
|
|
) |
|
|
|
|
|
print(completion.choices[0].message) |
|
|
|
|
|
return completion.choices[0].message.content |