File size: 7,104 Bytes
e3278e4 |
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 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 |
import json
from typing import Any, Coroutine, Dict, Optional, Union
import httpx
import litellm
from litellm.llms.custom_httpx.http_handler import (
_get_httpx_client,
get_async_httpx_client,
)
from litellm.llms.vertex_ai.gemini.vertex_and_google_ai_studio_gemini import VertexLLM
from litellm.types.llms.openai import Batch, CreateBatchRequest
from litellm.types.llms.vertex_ai import VertexAIBatchPredictionJob
from .transformation import VertexAIBatchTransformation
class VertexAIBatchPrediction(VertexLLM):
def __init__(self, gcs_bucket_name: str, *args, **kwargs):
super().__init__(*args, **kwargs)
self.gcs_bucket_name = gcs_bucket_name
def create_batch(
self,
_is_async: bool,
create_batch_data: CreateBatchRequest,
api_base: Optional[str],
vertex_credentials: Optional[str],
vertex_project: Optional[str],
vertex_location: Optional[str],
timeout: Union[float, httpx.Timeout],
max_retries: Optional[int],
) -> Union[Batch, Coroutine[Any, Any, Batch]]:
sync_handler = _get_httpx_client()
access_token, project_id = self._ensure_access_token(
credentials=vertex_credentials,
project_id=vertex_project,
custom_llm_provider="vertex_ai",
)
default_api_base = self.create_vertex_url(
vertex_location=vertex_location or "us-central1",
vertex_project=vertex_project or project_id,
)
if len(default_api_base.split(":")) > 1:
endpoint = default_api_base.split(":")[-1]
else:
endpoint = ""
_, api_base = self._check_custom_proxy(
api_base=api_base,
custom_llm_provider="vertex_ai",
gemini_api_key=None,
endpoint=endpoint,
stream=None,
auth_header=None,
url=default_api_base,
)
headers = {
"Content-Type": "application/json; charset=utf-8",
"Authorization": f"Bearer {access_token}",
}
vertex_batch_request: VertexAIBatchPredictionJob = (
VertexAIBatchTransformation.transform_openai_batch_request_to_vertex_ai_batch_request(
request=create_batch_data
)
)
if _is_async is True:
return self._async_create_batch(
vertex_batch_request=vertex_batch_request,
api_base=api_base,
headers=headers,
)
response = sync_handler.post(
url=api_base,
headers=headers,
data=json.dumps(vertex_batch_request),
)
if response.status_code != 200:
raise Exception(f"Error: {response.status_code} {response.text}")
_json_response = response.json()
vertex_batch_response = VertexAIBatchTransformation.transform_vertex_ai_batch_response_to_openai_batch_response(
response=_json_response
)
return vertex_batch_response
async def _async_create_batch(
self,
vertex_batch_request: VertexAIBatchPredictionJob,
api_base: str,
headers: Dict[str, str],
) -> Batch:
client = get_async_httpx_client(
llm_provider=litellm.LlmProviders.VERTEX_AI,
)
response = await client.post(
url=api_base,
headers=headers,
data=json.dumps(vertex_batch_request),
)
if response.status_code != 200:
raise Exception(f"Error: {response.status_code} {response.text}")
_json_response = response.json()
vertex_batch_response = VertexAIBatchTransformation.transform_vertex_ai_batch_response_to_openai_batch_response(
response=_json_response
)
return vertex_batch_response
def create_vertex_url(
self,
vertex_location: str,
vertex_project: str,
) -> str:
"""Return the base url for the vertex garden models"""
# POST https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT_ID/locations/LOCATION/batchPredictionJobs
return f"https://{vertex_location}-aiplatform.googleapis.com/v1/projects/{vertex_project}/locations/{vertex_location}/batchPredictionJobs"
def retrieve_batch(
self,
_is_async: bool,
batch_id: str,
api_base: Optional[str],
vertex_credentials: Optional[str],
vertex_project: Optional[str],
vertex_location: Optional[str],
timeout: Union[float, httpx.Timeout],
max_retries: Optional[int],
) -> Union[Batch, Coroutine[Any, Any, Batch]]:
sync_handler = _get_httpx_client()
access_token, project_id = self._ensure_access_token(
credentials=vertex_credentials,
project_id=vertex_project,
custom_llm_provider="vertex_ai",
)
default_api_base = self.create_vertex_url(
vertex_location=vertex_location or "us-central1",
vertex_project=vertex_project or project_id,
)
# Append batch_id to the URL
default_api_base = f"{default_api_base}/{batch_id}"
if len(default_api_base.split(":")) > 1:
endpoint = default_api_base.split(":")[-1]
else:
endpoint = ""
_, api_base = self._check_custom_proxy(
api_base=api_base,
custom_llm_provider="vertex_ai",
gemini_api_key=None,
endpoint=endpoint,
stream=None,
auth_header=None,
url=default_api_base,
)
headers = {
"Content-Type": "application/json; charset=utf-8",
"Authorization": f"Bearer {access_token}",
}
if _is_async is True:
return self._async_retrieve_batch(
api_base=api_base,
headers=headers,
)
response = sync_handler.get(
url=api_base,
headers=headers,
)
if response.status_code != 200:
raise Exception(f"Error: {response.status_code} {response.text}")
_json_response = response.json()
vertex_batch_response = VertexAIBatchTransformation.transform_vertex_ai_batch_response_to_openai_batch_response(
response=_json_response
)
return vertex_batch_response
async def _async_retrieve_batch(
self,
api_base: str,
headers: Dict[str, str],
) -> Batch:
client = get_async_httpx_client(
llm_provider=litellm.LlmProviders.VERTEX_AI,
)
response = await client.get(
url=api_base,
headers=headers,
)
if response.status_code != 200:
raise Exception(f"Error: {response.status_code} {response.text}")
_json_response = response.json()
vertex_batch_response = VertexAIBatchTransformation.transform_vertex_ai_batch_response_to_openai_batch_response(
response=_json_response
)
return vertex_batch_response
|