feat(http): move from openai only to http frontend package

handler.py

@@ -1,17 +1,17 @@
 import platform
-from functools import reduce
 from operator import itemgetter
 from typing import Generator, Tuple
 
 import torch
-from hfendpoints.openai import Context, run
-from hfendpoints.openai.embeddings import Embedding, EmbeddingEndpoint, EmbeddingRequest, EmbeddingResponse, Usage
+from hfendpoints.http import Context, run
+from hfendpoints.http.embeddings import Embedding, EmbeddingEndpoint, EmbeddingRequest, EmbeddingResponse, Usage
 from intel_extension_for_pytorch.cpu.runtime import pin
 from loguru import logger
-from hfendpoints import EndpointConfig, Handler, __version__
 from sentence_transformers import SentenceTransformer
-from torch.nn import Module
 from torch.backends.mkldnn import VERBOSE_ON_CREATION, VERBOSE_OFF
+from torch.nn import Module
+
+from hfendpoints import EndpointConfig, Handler, __version__
 
 # Not used for now
 SUPPORTED_AMP_DTYPES = {torch.float32, torch.bfloat16}
@@ -59,7 +59,7 @@ def get_usage(mask: torch.IntTensor) -> Usage:
 
 
 class SentenceTransformerWithUsage(Module):
-    __slots__ = ("_model", )
+    __slots__ = ("_model",)
 
     def __init__(self, model: SentenceTransformer):
         super().__init__()
@@ -88,7 +88,6 @@ class SentenceTransformerHandler(Handler):
         dtype = torch.bfloat16 if has_bf16_support() else torch.float32
         model = SentenceTransformer(self._config.model_id, device="cpu", model_kwargs={"torch_dtype": dtype})
 
-
         if platform.machine() == "x86_64":
             import intel_extension_for_pytorch as ipex
             logger.info(f"x64 platform detected: {platform.processor()}")