Automated sync from GitHub Actions

.github/workflows/sync_to_hf_space.yml

@@ -0,0 +1,66 @@
+name: Sync to Hugging Face Space
+
+on:
+  push:
+    branches:
+      - main # Or whichever branch you want to trigger deployment from
+
+jobs:
+  sync-to-hub:
+    runs-on: ubuntu-latest
+    steps:
+      # Step 1: Check out the code from the GitHub repository
+      - name: Checkout repository
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 0  # Fetch all history for proper syncing
+
+      # Step 2: Setup Git LFS
+      - name: Setup Git LFS
+        run: |
+          git lfs install
+
+      # Step 3: Configure Git
+      - name: Configure Git
+        run: |
+          git config --global user.name "github-actions"
+          git config --global user.email "[email protected]"
+          
+      # Step 4: List files to confirm what's available
+      - name: List files
+        run: |
+          ls -la
+          
+      # Step 5: Create and push to Hugging Face Space
+      - name: Push to Hugging Face Space
+        env:
+          HF_TOKEN: ${{ secrets.HF_TOKEN_WRITE }}
+        run: |
+          # Clean any existing hf_repo directory
+          rm -rf hf_repo
+          
+          # Create a new directory for the HF repo
+          mkdir -p hf_repo
+          
+          # Copy specific files to the HF repo
+          cp -r README.md app.py requirements.txt .github hf_repo/
+          
+          # Initialize git in the new directory
+          cd hf_repo
+          git init
+          git branch -m main
+          
+          # Show what was copied
+          ls -la
+          
+          # Set up the remote
+          git remote add origin https://USER:[email protected]/spaces/sikeaditya/LLM-Cloud-Demo
+          
+          # Add all files
+          git add .
+          
+          # Commit files
+          git commit -m "Automated sync from GitHub Actions"
+          
+          # Force push to HF Space
+          git push -f origin main

README.md

@@ -0,0 +1,9 @@
+---
+title: LLM-Cloud-Demo
+sdk: gradio
+emoji: ⚡
+colorFrom: yellow
+colorTo: red
+---
+
+# LLM-Cloud-Demo

app.py

@@ -0,0 +1,111 @@
+import gradio as gr
+import requests
+import os
+import json # Import json for better error message handling
+
+# --- Configuration ---
+# Using the specific Nebius/HF router URL from your snippet
+API_URL = "https://router.huggingface.co/nebius/v1/chat/completions"
+# Using the model from your snippet
+MODEL_ID = "google/gemma-3-27b-it-fast"
+# Get Hugging Face token from environment variable/secrets
+HF_TOKEN = os.getenv("HF_TOKEN")
+
+if not HF_TOKEN:
+    raise ValueError("Hugging Face token not found. Please set the HF_TOKEN environment variable or secret.")
+
+HEADERS = {"Authorization": f"Bearer {HF_TOKEN}", "Content-Type": "application/json"}
+
+# --- No Local Model Loading Needed ---
+print(f"Application configured to use Hugging Face Inference API.")
+print(f"Target Model: AgriAssist_LLM")
+print(f"API Endpoint: {API_URL}")
+
+# --- Inference Function (Using Hugging Face API) ---
+def generate_response(prompt, max_new_tokens=512): # Using max_tokens from your snippet
+    print(f"Received prompt: {prompt}")
+    print("Preparing payload for API...")
+
+    # Construct the payload based on the API requirements
+    # NOTE: This version assumes text-only input matching the Gradio interface.
+    #       To handle image input like your snippet, the Gradio interface
+    #       and payload structure would need modification.
+    payload = {
+        "messages": [
+            {
+                "role": "user",
+                "content": prompt
+                # Example for multimodal if Gradio input changes:
+                # "content": [
+                #    {"type": "text", "text": prompt},
+                #    {"type": "image_url", "image_url": {"url": "some_image_url.jpg"}}
+                # ]
+            }
+        ],
+        "model": MODEL_ID,
+        "max_tokens": max_new_tokens,
+        # Optional parameters you might want to add:
+        # "temperature": 0.7,
+        # "top_p": 0.9,
+        # "stream": False # Set to True for streaming responses if API supports it
+    }
+
+    print(f"Sending request to API for model AgriAssist_LLM...")
+    try:
+        # Make the POST request
+        response = requests.post(API_URL, headers=HEADERS, json=payload)
+
+        # Raise an exception for bad status codes (like 4xx or 5xx)
+        response.raise_for_status()
+
+        # Parse the JSON response
+        result = response.json()
+        print("API Response Received Successfully.")
+
+        # Extract the generated text - Structure matches your snippet's expectation
+        if "choices" in result and len(result["choices"]) > 0 and "message" in result["choices"][0] and "content" in result["choices"][0]["message"]:
+            api_response_content = result["choices"][0]["message"]["content"]
+            print(f"API generated content: {api_response_content}")
+            return api_response_content
+        else:
+            # Handle unexpected response structure
+            print(f"Unexpected API response structure: {result}")
+            return f"Error: Unexpected API response structure. Full response: {json.dumps(result)}"
+
+    except requests.exceptions.RequestException as e:
+        # Handle network errors, timeout errors, invalid responses, etc.
+        error_message = f"Error calling Hugging Face API: {e}"
+        # Try to get more details from the response body if it exists
+        error_detail = ""
+        if e.response is not None:
+            try:
+                error_detail = e.response.json() # Try parsing JSON error
+            except json.JSONDecodeError:
+                error_detail = e.response.text # Fallback to raw text
+        print(f"{error_message}\nResponse details: {error_detail}")
+        return f"{error_message}\nDetails: {error_detail}"
+
+    except Exception as e:
+        # Handle other potential errors during processing
+        print(f"An unexpected error occurred: {e}")
+        return f"An unexpected error occurred: {e}"
+
+# --- Gradio Interface ---
+iface = gr.Interface(
+    fn=generate_response,
+    inputs=gr.Textbox(lines=5, label="Enter your prompt", placeholder="Type your question or instruction here..."),
+    outputs=gr.Textbox(lines=8, label=f"AgriAssist_LLM Says (via API):"), # Updated label
+    title=f"Chat with AgriAssist_LLM via Inference API", # Updated title
+    description=("This demo sends your text to a remote server for processing."), # Updated description
+    allow_flagging="never",
+    examples=[ # Examples should still be relevant
+        ["Explain the concept of cloud computing in simple terms."],
+        ["Write Python code to list files in a directory."],
+        ["What are the main benefits of using Generative AI?"],
+        ["Translate 'Cloud computing offers scalability' to German."],
+    ]
+)
+
+# --- Launch the App ---
+# You can add share=True if you want to create a temporary public link (use with caution)
+iface.launch(server_name="0.0.0.0", server_port=7860) # Makes it accessible in Codespaces/docker

requirements.txt

@@ -0,0 +1,7 @@
+transformers>=4.38.0 # Or a recent version supporting Gemma
+accelerate
+bitsandbytes # Often needed for efficient loading
+torch
+gradio
+# Add sentencepiece if not included by transformers
+sentencepiece