Upload 3 files

Dockerfile

@@ -0,0 +1,6 @@
+FROM python:3.10
+WORKDIR /app
+COPY . /app
+RUN pip install --no-cache-dir -r requirements.txt
+EXPOSE 7860
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

app.py

@@ -0,0 +1,74 @@
+import os
+import torch
+from fastapi import FastAPI
+from pydantic import BaseModel
+from transformers import AutoTokenizer, AutoModelForCausalLM
+from peft import PeftModel
+from fastapi.middleware.cors import CORSMiddleware
+
+app = FastAPI()
+
+# Security Pass for your HTML file
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+model_id = "Qwen/Qwen2.5-1.5B-Instruct"
+
+print("Nexus is initializing reasoning cores...")
+
+# 1. Load Tokenizer
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+
+# 2. Load Base Brain (CPU Optimized)
+base_model = AutoModelForCausalLM.from_pretrained(
+    model_id,
+    torch_dtype=torch.float32,
+    device_map="cpu",
+    low_cpu_mem_usage=True
+)
+
+# 3. Attach YOUR Nexus Training (Xirrik's Work)
+# We use "." because the files were uploaded to the root of the Space
+try:
+    model = PeftModel.from_pretrained(base_model, ".")
+    print("Nexus Training Attached Successfully!")
+except Exception as e:
+    print(f"Loading files from root failed: {e}. Trying folder...")
+    model = PeftModel.from_pretrained(base_model, "./zenvy-nexus-model")
+
+class ChatRequest(BaseModel):
+    message: str
+
+@app.get("/")
+def home():
+    return {"status": "Zenvy Nexus is ONLINE", "model": "Qwen-2.5-1.5B-Custom"}
+
+@app.post("/chat")
+async def chat(request: ChatRequest):
+    try:
+        # Nexus Logic Prompt
+        prompt = f"<|im_start|>user\n{request.message}<|im_end|>\n<|im_start|>assistant\n"
+        
+        inputs = tokenizer(prompt, return_tensors="pt").to("cpu")
+        
+        with torch.no_grad():
+            outputs = model.generate(
+                **inputs,
+                max_new_tokens=200,
+                temperature=0.7,
+                do_sample=True,
+                repetition_penalty=1.1
+            )
+        
+        full_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
+        reply = full_text.split("assistant")[-1].strip()
+        
+        return {"reply": reply}
+
+    except Exception as e:
+        return {"reply": f"Nexus Core Error: {str(e)}"}

requirements.txt

@@ -0,0 +1,7 @@
+fastapi
+uvicorn
+transformers
+torch
+peft
+accelerate
+pydantic