Spaces:
Paused
Paused
File size: 1,326 Bytes
cbbe164 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 |
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
from rag.search import search_context
import os
BASE_DIR = os.path.dirname(os.path.dirname(__file__))
MODEL_PATH = os.path.join(BASE_DIR, "model", "final")
tokenizer = AutoTokenizer.from_pretrained(
MODEL_PATH,
local_files_only=True,
trust_remote_code=True,
fix_mistral_regex=True
)
model = AutoModelForCausalLM.from_pretrained(
MODEL_PATH,
local_files_only=True,
trust_remote_code=True,
device_map="auto",
dtype=torch.float16
)
def analyze(user_input: str):
context = search_context(user_input)
prompt = f"""
You are a cybersecurity malware analysis assistant.
Respond ONLY in valid JSON.
Use these fields exactly once:
- reasoning (array of strings)
- indicators (array)
- confidence (float 0-1)
- recommendation (string)
- mitre_attack (array)
Context:
{context}
Input:
{user_input}
Response:
"""
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
with torch.no_grad():
output = model.generate(
**inputs,
max_new_tokens=256,
do_sample=True,
temperature=0.2,
top_p=0.9
)
return tokenizer.decode(output[0], skip_special_tokens=True)
|