Tttalalma

Runtime error

App Files Files Community

dexsecon commited on Jul 12

Commit

d02f5a2

verified ·

1 Parent(s): d86d3ea

Update app.py

Browse files

Files changed (1) hide show

app.py +11 -62

app.py CHANGED Viewed

@@ -1,4 +1,4 @@
-from flask import Flask, request, jsonify, render_template_string
 from llama_cpp import Llama
 from huggingface_hub import hf_hub_download
 import threading, time
@@ -6,85 +6,34 @@ import threading, time
 app = Flask(__name__)
 start_time = time.time()
-# 🔧 CONFIG
-REPO = "sens2010/law_llama3_8B_8bit_GUFF"
-FILE = "unsloth.Q8_0.gguf"  # ~9 GB quant
-print("🔽 Downloading model…")
 MODEL_PATH = hf_hub_download(REPO, FILE, local_dir=".", local_dir_use_symlinks=False)
-print("🔄 Loading model… (this may take a while)")
 llm = Llama(model_path=MODEL_PATH, n_ctx=2048, n_threads=8)
-# 🏠 Serve HTML directly from code
 @app.route("/", methods=["GET"])
-def homepage():
-    return render_template_string("""
-<!DOCTYPE html>
-<html lang="en">
-<head>
-  <meta charset="UTF-8" />
-  <title>🤖 Chat with LawLLaMA</title>
-  <meta name="viewport" content="width=device-width, initial-scale=1.0" />
-  <style>
-    body { font-family: sans-serif; background: #f4f4f4; padding: 20px; margin: 0; }
-    h2 { text-align: center; color: #333; }
-    textarea { width: 100%; font-size: 16px; padding: 10px; border: 1px solid #ccc; border-radius: 6px; margin-top: 10px; }
-    button { width: 100%; padding: 12px; background: #4CAF50; color: white; border: none; border-radius: 6px; font-size: 16px; margin-top: 10px; cursor: pointer; }
-    button:hover { background: #45a049; }
-    #response { margin-top: 20px; background: #fff; padding: 15px; border-radius: 6px; border-left: 5px solid #4CAF50; white-space: pre-wrap; color: #333; }
-  </style>
-</head>
-<body>
-  <h2>💬 LawLLaMA 8B Chat</h2>
-  <textarea id="msg" rows="4" placeholder="Type your legal or coding question..."></textarea>
-  <button onclick="send()">Send</button>
-  <div id="response">Reply will appear here...</div>
-  <script>
-    async function send() {
-      const msg = document.getElementById("msg").value.trim();
-      const resBox = document.getElementById("response");
-      if (!msg) {
-        alert("Please enter a message.");
-        return;
-      }
-      resBox.innerText = "⏳ Thinking...";
-      try {
-        const res = await fetch("/chat", {
-          method: "POST",
-          headers: { "Content-Type": "application/json" },
-          body: JSON.stringify({ message: msg })
-        });
-        const data = await res.json();
-        resBox.innerText = data.reply ? "🧠 " + data.reply : "⚠️ " + data.error;
-      } catch (err) {
-        resBox.innerText = "❌ Error: " + err.message;
-      }
-    }
-  </script>
-</body>
-</html>
-    """)
 @app.route("/chat", methods=["POST"])
 def chat():
     msg = request.json.get("message", "").strip()
     if not msg:
         return jsonify({"error": "Empty message"}), 400
-    prompt = f"# User:\n{msg}\n# Assistant:\n"
-    out = llm(prompt, max_tokens=256, temperature=0.2, stop=["# User:", "# Assistant:"])
     return jsonify({"reply": out["choices"][0]["text"].strip()})
 @app.route("/status")
 def status():
     return jsonify({
-        "uptime_s": int(time.time() - start_time),
         "model": FILE
     })
-def run_app():
-    app.run(host="0.0.0.0", port=7860)
 if __name__ == "__main__":
-    threading.Thread(target=run_app).start()

+from flask import Flask, request, jsonify, send_file
 from llama_cpp import Llama
 from huggingface_hub import hf_hub_download
 import threading, time
 app = Flask(__name__)
 start_time = time.time()
+REPO = "TheBloke/Qwen2.5-1.8B-Chat-GGUF"
+FILE = "qwen2_5-1.8b-chat.Q4_K_M.gguf"
+print("🔽 Downloading model...")
 MODEL_PATH = hf_hub_download(REPO, FILE, local_dir=".", local_dir_use_symlinks=False)
+print("🔄 Loading model...")
 llm = Llama(model_path=MODEL_PATH, n_ctx=2048, n_threads=8)
 @app.route("/", methods=["GET"])
+def root():
+    return send_file("index.html")
 @app.route("/chat", methods=["POST"])
 def chat():
     msg = request.json.get("message", "").strip()
     if not msg:
         return jsonify({"error": "Empty message"}), 400
+    prompt = f"<|user|>\n{msg}\n<|assistant|>"
+    out = llm(prompt, max_tokens=300, temperature=0.7, stop=["<|user|>", "<|assistant|>"])
     return jsonify({"reply": out["choices"][0]["text"].strip()})
 @app.route("/status")
 def status():
     return jsonify({
+        "uptime": round(time.time() - start_time),
         "model": FILE
     })
 if __name__ == "__main__":
+    app.run(host="0.0.0.0", port=7860)