Spaces:
Paused
Paused
File size: 1,581 Bytes
334fdf5 e0d816d 334fdf5 63b9234 c72165a ebf8f51 c72165a 07099ff c72165a 334fdf5 c72165a 334fdf5 c72165a 334fdf5 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 |
from http.server import HTTPServer, BaseHTTPRequestHandler
from urllib.parse import urlparse
import os
import json
#from huggingface_hub.file_download import http_get
from llama_cpp import Llama
directory = "/home/oluser/olapp/"
model_url = "https://huggingface.co/IlyaGusev/saiga_mistral_7b_gguf/resolve/main/model-q8_0.gguf"
model_name = "model-q8_0.gguf"
final_model_path = os.path.join(directory, model_name)
print("Downloading all files...")
#rm_files = [os.path.join(directory, f) for f in os.listdir(directory)]
#for f in rm_files:
# if os.path.isfile(f):
# os.remove(f)
# else:
# shutil.rmtree(f)
if not os.path.exists(final_model_path):
with open(final_model_path, "wb") as f:
http_get(model_url, f)
os.chmod(final_model_path, 0o777)
print("Files downloaded!")
model = Llama(
model_path=final_model_path,
n_ctx=4096,
n_parts=1,
)
print("Model loaded!")
class OlHandler(BaseHTTPRequestHandler):
def do_GET(self):
query = urlparse(self.path).query
query_components = dict(qc.split("=") for qc in query.split("&"))
q = query_components["q"]
# message = '-=# ' + q + ' #=-'
output = llm(
q,
max_tokens=32, # Generate up to 32 tokens
echo=False
)
self.send_response(200)
self.end_headers()
self.wfile.write(output.encode('utf-8'))
return
if __name__ == '__main__':
olserver = HTTPServer(('0.0.0.0', 7860), OlHandler)
print('Starting server at http://0.0.0.0:7860')
olserver.serve_forever()
|