feat: add initial AutoGPTQ backend implementation

This commit is contained in:
Ettore Di Giacinto 2023-08-07 22:39:10 +02:00
parent 91d49cfe9f
commit a843e64fc2
37 changed files with 660 additions and 148 deletions

View file

@ -18,7 +18,6 @@ class BackendServicer(backend_pb2_grpc.BackendServicer):
return backend_pb2.Reply(message=bytes("OK", 'utf-8'))
def LoadModel(self, request, context):
model_name = request.Model
model_name = os.path.basename(model_name)
try:
self.model = SentenceTransformer(model_name)
except Exception as err: