Upload folder using huggingface_hub

README.md

@@ -1,3 +1,14 @@
+---
+title: amkyaw-coder
+emoji: 🤖
+colorFrom: blue
+colorTo: purple
+sdk: gradio
+sdk_version: "4.0"
+app_file: app.py
+pinned: false
+---
+
 # amkyawdev/amkyaw-dev-v1
 
 ## Model Overview
@@ -12,9 +23,6 @@
 ```bash
 # Run the model
 ollama run amkyawdev/amkyaw-dev-v1
-
-# Or run with specific tag
-ollama run amkyawdev/amkyaw-dev-v1:latest
 ```
 
 ## Features

app.py

@@ -1,30 +1,25 @@
 import gradio as gr
-import ollama
+from huggingface_hub import InferenceClient
 import os
 
+# Use environment variable for token (set as secret in Space)
+token = os.environ.get("HF_TOKEN", "")
+client = InferenceClient("amkyawdev/amkyaw-dev-v1", token=token)
+
 def generate(prompt, temperature=0.8, max_tokens=512):
     try:
-        response = ollama.generate(
-            model='amkyawdev/amkyaw-dev-v1',
+        response = client.text_generation(
             prompt=prompt,
-            options={
-                'temperature': temperature,
-                'num_predict': max_tokens
-            }
+            temperature=temperature,
+            max_new_tokens=max_tokens,
+            do_sample=True if temperature > 0 else False
         )
-        return response['response']
+        return response
     except Exception as e:
         return f"Error: {str(e)}"
 
-# Check if Ollama is accessible
-try:
-    ollama.list()
-    status = "✅ Ollama is running"
-except:
-    status = "❌ Ollama is not running"
-
 with gr.Blocks(title="amkyaw-coder") as demo:
-    gr.Markdown(f"# amkyaw-coder-1.5b-instruct\n{status}")
+    gr.Markdown("# amkyaw-coder-1.5b-instruct\n🤖 Code Generation Model")
     
     with gr.Row():
         with gr.Column():

requirements.txt

@@ -1,2 +1,2 @@
 gradio>=4.0.0
-ollama>=0.1.0
+huggingface_hub>=0.20.0