Create app.py

app.py

@@ -0,0 +1,76 @@
+from fastapi import FastAPI, Request
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+import os
+
+app = FastAPI()
+
+# --- MODEL SETUP ---
+MODEL_ID = "Piyush-boss/Nexari-Qwen-3B-Full"
+
+print("🔄 Loading Nexari Model... (This takes time on CPU)")
+tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
+model = AutoModelForCausalLM.from_pretrained(
+    MODEL_ID, 
+    torch_dtype=torch.float32, # CPU ke liye float32 safe hai
+    device_map="cpu",          # Force CPU
+    low_cpu_mem_usage=True
+)
+print("✅ Nexari Loaded Successfully!")
+
+@app.get("/")
+def home():
+    return {"status": "Nexari Server is Running!"}
+
+@app.post("/v1/chat/completions")
+async def chat_completions(request: Request):
+    data = await request.json()
+    messages = data.get("messages", [])
+    
+    # 1. Prompt Format
+    prompt = ""
+    for msg in messages:
+        role = msg["role"]
+        content = msg["content"]
+        if role == "system":
+            prompt += f"<|im_start|>system\n{content}<|im_end|>\n"
+        elif role == "user":
+            prompt += f"<|im_start|>user\n{content}<|im_end|>\n"
+        elif role == "assistant":
+            prompt += f"<|im_start|>assistant\n{content}<|im_end|>\n"
+            
+    prompt += "<|im_start|>assistant\n"
+    
+    # 2. Tokenize & Generate
+    inputs = tokenizer(prompt, return_tensors="pt")
+    
+    # CPU Generation (Thoda slow hoga, par chalega)
+    outputs = model.generate(
+        **inputs, 
+        max_new_tokens=512,
+        temperature=0.7,
+        do_sample=True
+    )
+    
+    # 3. Decode
+    generated_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    # Prompt hata kar sirf naya text nikalo
+    response_text = generated_text.replace(prompt, "").split("<|im_end|>")[0].strip()
+    # Fallback agar prompt replace theek se na ho
+    if "assistant" in response_text:
+        response_text = response_text.split("assistant")[-1].strip()
+
+    # 4. Return OpenAI JSON
+    return {
+        "id": "chatcmpl-nexari",
+        "object": "chat.completion",
+        "created": 1234567890,
+        "choices": [{
+            "index": 0,
+            "message": {
+                "role": "assistant",
+                "content": response_text
+            },
+            "finish_reason": "stop"
+        }]
+    }