Gemini Try

app.py

@@ -1,45 +1,31 @@
 from fastapi import FastAPI
 from pydantic import BaseModel
-from transformers import AutoTokenizer, AutoModelForCausalLM
 from fastapi.middleware.cors import CORSMiddleware
-import torch
+import google.generativeai as genai
 import os
+from dotenv import load_dotenv
 
-# Ensure Hugging Face cache uses a writable path
-os.environ["TRANSFORMERS_CACHE"] = "/app/.cache"
-os.environ["HF_HOME"] = "/app/.cache"
+# Load variables from .env file
+load_dotenv()
+# ✅ Configure API Key (set GOOGLE_API_KEY in environment variables)
+genai.configure(api_key=os.environ["GOOGLE_API_KEY"])
 
 app = FastAPI()
 
 # ✅ Allow all origins
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=["*"],  # allow all origins
+    allow_origins=["*"],  
     allow_credentials=True,
-    allow_methods=["*"],  # allow all HTTP methods
-    allow_headers=["*"],  # allow all headers
+    allow_methods=["*"],
+    allow_headers=["*"],
 )
 
-
 class ChatRequest(BaseModel):
     message: str
 
-# Load DeepSeek model (small one for local use)
-model_name = "deepseek-ai/deepseek-coder-1.3b-base"
-
-# model_name = "deepseek-ai/deepseek-llm-7b-base"
-
-#model_name="Qwen/Qwen2.5-1.5B-Instruct"
-#model_name="TinyLlama/TinyLlama-1.1B-Chat-v1.0"
-
-print("Loading model... this may take a minute ⏳")
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForCausalLM.from_pretrained(
-    model_name,
-    torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32,
-    device_map="auto"
-)
-print("Model loaded ✅")
+# ✅ Load Gemini model (example: gemini-1.5-flash is lightweight & fast)
+model = genai.GenerativeModel("gemini-1.5-flash")
 
 @app.get("/")
 def root():
@@ -47,11 +33,6 @@ def root():
 
 @app.post("/chat")
 def chat(request: ChatRequest):
-    """Chat endpoint using DeepSeek model"""
-    inputs = tokenizer(request.message, return_tensors="pt").to(model.device)
-    outputs = model.generate(**inputs, max_new_tokens=200)
-   
-    reply = tokenizer.decode(outputs[0], skip_special_tokens=True)
-    
-   
-    return {"reply": reply}
+    """Chat endpoint using Gemini"""
+    response = model.generate_content(request.message)
+    return {"reply": response.text}

app2.py

@@ -0,0 +1,57 @@
+from fastapi import FastAPI
+from pydantic import BaseModel
+from transformers import AutoTokenizer, AutoModelForCausalLM
+from fastapi.middleware.cors import CORSMiddleware
+import torch
+import os
+
+# Ensure Hugging Face cache uses a writable path
+os.environ["TRANSFORMERS_CACHE"] = "/app/.cache"
+os.environ["HF_HOME"] = "/app/.cache"
+
+app = FastAPI()
+
+# ✅ Allow all origins
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],  # allow all origins
+    allow_credentials=True,
+    allow_methods=["*"],  # allow all HTTP methods
+    allow_headers=["*"],  # allow all headers
+)
+
+
+class ChatRequest(BaseModel):
+    message: str
+
+# Load DeepSeek model (small one for local use)
+model_name = "deepseek-ai/deepseek-coder-1.3b-base"
+
+# model_name = "deepseek-ai/deepseek-llm-7b-base"
+
+#model_name="Qwen/Qwen2.5-1.5B-Instruct"
+#model_name="TinyLlama/TinyLlama-1.1B-Chat-v1.0"
+
+print("Loading model... this may take a minute ⏳")
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForCausalLM.from_pretrained(
+    model_name,
+    torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32,
+    device_map="auto"
+)
+print("Model loaded ✅")
+
+@app.get("/")
+def root():
+    return {"status": "ok"}
+
+@app.post("/chat")
+def chat(request: ChatRequest):
+    """Chat endpoint using DeepSeek model"""
+    inputs = tokenizer(request.message, return_tensors="pt").to(model.device)
+    outputs = model.generate(**inputs, max_new_tokens=200)
+   
+    reply = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    
+   
+    return {"reply": reply}