Update data/chat.py
Browse files- data/chat.py +17 -22
data/chat.py
CHANGED
|
@@ -1,17 +1,26 @@
|
|
|
|
|
| 1 |
import spaces
|
| 2 |
import os
|
| 3 |
import json
|
| 4 |
-
import argparse
|
| 5 |
from huggingface_hub import InferenceClient, login
|
| 6 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
api_key = os.getenv("UCODE_SECRET")
|
| 8 |
login(api_key)
|
| 9 |
-
|
| 10 |
-
# Initialize the InferenceClient with the specified model
|
| 11 |
client = InferenceClient("meta-llama/Meta-Llama-3-70B-Instruct")
|
| 12 |
|
|
|
|
| 13 |
@spaces.GPU()
|
| 14 |
-
def chat_completion(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
try:
|
| 16 |
response = ""
|
| 17 |
for message in client.chat_completion(
|
|
@@ -24,26 +33,12 @@ def chat_completion(user_input, max_tokens, temperature, top_p):
|
|
| 24 |
token = message.choices[0].delta.get("content", "")
|
| 25 |
response += token
|
| 26 |
|
| 27 |
-
return
|
| 28 |
except Exception as e:
|
| 29 |
-
return
|
| 30 |
|
| 31 |
def main():
|
| 32 |
-
|
| 33 |
-
parser.add_argument("user_input", type=str, help="The input text for the chat model")
|
| 34 |
-
parser.add_argument("--max_tokens", type=int, default=50, help="Maximum number of tokens in the response")
|
| 35 |
-
parser.add_argument("--temperature", type=float, default=0.7, help="Sampling temperature")
|
| 36 |
-
parser.add_argument("--top_p", type=float, default=0.9, help="Top-p sampling value")
|
| 37 |
-
|
| 38 |
-
args = parser.parse_args()
|
| 39 |
-
|
| 40 |
-
result = chat_completion(
|
| 41 |
-
user_input=args.user_input,
|
| 42 |
-
max_tokens=args.max_tokens,
|
| 43 |
-
temperature=args.temperature,
|
| 44 |
-
top_p=args.top_p
|
| 45 |
-
)
|
| 46 |
-
print(result)
|
| 47 |
|
| 48 |
if __name__ == "__main__":
|
| 49 |
-
main()
|
|
|
|
| 1 |
+
from flask import Flask, request, jsonify
|
| 2 |
import spaces
|
| 3 |
import os
|
| 4 |
import json
|
|
|
|
| 5 |
from huggingface_hub import InferenceClient, login
|
| 6 |
|
| 7 |
+
# Initialize Flask app
|
| 8 |
+
app = Flask(__name__)
|
| 9 |
+
|
| 10 |
+
# Load the API key and initialize the InferenceClient
|
| 11 |
api_key = os.getenv("UCODE_SECRET")
|
| 12 |
login(api_key)
|
|
|
|
|
|
|
| 13 |
client = InferenceClient("meta-llama/Meta-Llama-3-70B-Instruct")
|
| 14 |
|
| 15 |
+
@app.route('/chat', methods=['POST'])
|
| 16 |
@spaces.GPU()
|
| 17 |
+
def chat_completion():
|
| 18 |
+
data = request.json
|
| 19 |
+
user_input = data.get('user_input', '')
|
| 20 |
+
max_tokens = data.get('max_tokens', 512)
|
| 21 |
+
temperature = data.get('temperature', 0.7)
|
| 22 |
+
top_p = data.get('top_p', 0.95)
|
| 23 |
+
|
| 24 |
try:
|
| 25 |
response = ""
|
| 26 |
for message in client.chat_completion(
|
|
|
|
| 33 |
token = message.choices[0].delta.get("content", "")
|
| 34 |
response += token
|
| 35 |
|
| 36 |
+
return jsonify({"status": "success", "output": response})
|
| 37 |
except Exception as e:
|
| 38 |
+
return jsonify({"status": "error", "message": str(e)})
|
| 39 |
|
| 40 |
def main():
|
| 41 |
+
app.run(host='localhost', port=3000)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 42 |
|
| 43 |
if __name__ == "__main__":
|
| 44 |
+
main()
|