File size: 12,642 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 |
import json
import os
from typing import TYPE_CHECKING, Any, Dict, Optional, Tuple, Union
from litellm._logging import verbose_logger
from litellm.integrations.custom_batch_logger import CustomBatchLogger
from litellm.llms.custom_httpx.http_handler import (
get_async_httpx_client,
httpxSpecialProvider,
)
from litellm.types.integrations.gcs_bucket import *
from litellm.types.utils import StandardCallbackDynamicParams, StandardLoggingPayload
if TYPE_CHECKING:
from litellm.llms.vertex_ai.vertex_llm_base import VertexBase
else:
VertexBase = Any
IAM_AUTH_KEY = "IAM_AUTH"
class GCSBucketBase(CustomBatchLogger):
def __init__(self, bucket_name: Optional[str] = None, **kwargs) -> None:
self.async_httpx_client = get_async_httpx_client(
llm_provider=httpxSpecialProvider.LoggingCallback
)
_path_service_account = os.getenv("GCS_PATH_SERVICE_ACCOUNT")
_bucket_name = bucket_name or os.getenv("GCS_BUCKET_NAME")
self.path_service_account_json: Optional[str] = _path_service_account
self.BUCKET_NAME: Optional[str] = _bucket_name
self.vertex_instances: Dict[str, VertexBase] = {}
super().__init__(**kwargs)
async def construct_request_headers(
self,
service_account_json: Optional[str],
vertex_instance: Optional[VertexBase] = None,
) -> Dict[str, str]:
from litellm import vertex_chat_completion
if vertex_instance is None:
vertex_instance = vertex_chat_completion
_auth_header, vertex_project = await vertex_instance._ensure_access_token_async(
credentials=service_account_json,
project_id=None,
custom_llm_provider="vertex_ai",
)
auth_header, _ = vertex_instance._get_token_and_url(
model="gcs-bucket",
auth_header=_auth_header,
vertex_credentials=service_account_json,
vertex_project=vertex_project,
vertex_location=None,
gemini_api_key=None,
stream=None,
custom_llm_provider="vertex_ai",
api_base=None,
)
verbose_logger.debug("constructed auth_header %s", auth_header)
headers = {
"Authorization": f"Bearer {auth_header}", # auth_header
"Content-Type": "application/json",
}
return headers
def sync_construct_request_headers(self) -> Dict[str, str]:
from litellm import vertex_chat_completion
_auth_header, vertex_project = vertex_chat_completion._ensure_access_token(
credentials=self.path_service_account_json,
project_id=None,
custom_llm_provider="vertex_ai",
)
auth_header, _ = vertex_chat_completion._get_token_and_url(
model="gcs-bucket",
auth_header=_auth_header,
vertex_credentials=self.path_service_account_json,
vertex_project=vertex_project,
vertex_location=None,
gemini_api_key=None,
stream=None,
custom_llm_provider="vertex_ai",
api_base=None,
)
verbose_logger.debug("constructed auth_header %s", auth_header)
headers = {
"Authorization": f"Bearer {auth_header}", # auth_header
"Content-Type": "application/json",
}
return headers
def _handle_folders_in_bucket_name(
self,
bucket_name: str,
object_name: str,
) -> Tuple[str, str]:
"""
Handles when the user passes a bucket name with a folder postfix
Example:
- Bucket name: "my-bucket/my-folder/dev"
- Object name: "my-object"
- Returns: bucket_name="my-bucket", object_name="my-folder/dev/my-object"
"""
if "/" in bucket_name:
bucket_name, prefix = bucket_name.split("/", 1)
object_name = f"{prefix}/{object_name}"
return bucket_name, object_name
return bucket_name, object_name
async def get_gcs_logging_config(
self, kwargs: Optional[Dict[str, Any]] = {}
) -> GCSLoggingConfig:
"""
This function is used to get the GCS logging config for the GCS Bucket Logger.
It checks if the dynamic parameters are provided in the kwargs and uses them to get the GCS logging config.
If no dynamic parameters are provided, it uses the default values.
"""
if kwargs is None:
kwargs = {}
standard_callback_dynamic_params: Optional[StandardCallbackDynamicParams] = (
kwargs.get("standard_callback_dynamic_params", None)
)
bucket_name: str
path_service_account: Optional[str]
if standard_callback_dynamic_params is not None:
verbose_logger.debug("Using dynamic GCS logging")
verbose_logger.debug(
"standard_callback_dynamic_params: %s", standard_callback_dynamic_params
)
_bucket_name: Optional[str] = (
standard_callback_dynamic_params.get("gcs_bucket_name", None)
or self.BUCKET_NAME
)
_path_service_account: Optional[str] = (
standard_callback_dynamic_params.get("gcs_path_service_account", None)
or self.path_service_account_json
)
if _bucket_name is None:
raise ValueError(
"GCS_BUCKET_NAME is not set in the environment, but GCS Bucket is being used as a logging callback. Please set 'GCS_BUCKET_NAME' in the environment."
)
bucket_name = _bucket_name
path_service_account = _path_service_account
vertex_instance = await self.get_or_create_vertex_instance(
credentials=path_service_account
)
else:
# If no dynamic parameters, use the default instance
if self.BUCKET_NAME is None:
raise ValueError(
"GCS_BUCKET_NAME is not set in the environment, but GCS Bucket is being used as a logging callback. Please set 'GCS_BUCKET_NAME' in the environment."
)
bucket_name = self.BUCKET_NAME
path_service_account = self.path_service_account_json
vertex_instance = await self.get_or_create_vertex_instance(
credentials=path_service_account
)
return GCSLoggingConfig(
bucket_name=bucket_name,
vertex_instance=vertex_instance,
path_service_account=path_service_account,
)
async def get_or_create_vertex_instance(
self, credentials: Optional[str]
) -> VertexBase:
"""
This function is used to get the Vertex instance for the GCS Bucket Logger.
It checks if the Vertex instance is already created and cached, if not it creates a new instance and caches it.
"""
from litellm.llms.vertex_ai.vertex_llm_base import VertexBase
_in_memory_key = self._get_in_memory_key_for_vertex_instance(credentials)
if _in_memory_key not in self.vertex_instances:
vertex_instance = VertexBase()
await vertex_instance._ensure_access_token_async(
credentials=credentials,
project_id=None,
custom_llm_provider="vertex_ai",
)
self.vertex_instances[_in_memory_key] = vertex_instance
return self.vertex_instances[_in_memory_key]
def _get_in_memory_key_for_vertex_instance(self, credentials: Optional[str]) -> str:
"""
Returns key to use for caching the Vertex instance in-memory.
When using Vertex with Key based logging, we need to cache the Vertex instance in-memory.
- If a credentials string is provided, it is used as the key.
- If no credentials string is provided, "IAM_AUTH" is used as the key.
"""
return credentials or IAM_AUTH_KEY
async def download_gcs_object(self, object_name: str, **kwargs):
"""
Download an object from GCS.
https://cloud.google.com/storage/docs/downloading-objects#download-object-json
"""
try:
gcs_logging_config: GCSLoggingConfig = await self.get_gcs_logging_config(
kwargs=kwargs
)
headers = await self.construct_request_headers(
vertex_instance=gcs_logging_config["vertex_instance"],
service_account_json=gcs_logging_config["path_service_account"],
)
bucket_name = gcs_logging_config["bucket_name"]
bucket_name, object_name = self._handle_folders_in_bucket_name(
bucket_name=bucket_name,
object_name=object_name,
)
url = f"https://storage.googleapis.com/storage/v1/b/{bucket_name}/o/{object_name}?alt=media"
# Send the GET request to download the object
response = await self.async_httpx_client.get(url=url, headers=headers)
if response.status_code != 200:
verbose_logger.error(
"GCS object download error: %s", str(response.text)
)
return None
verbose_logger.debug(
"GCS object download response status code: %s", response.status_code
)
# Return the content of the downloaded object
return response.content
except Exception as e:
verbose_logger.error("GCS object download error: %s", str(e))
return None
async def delete_gcs_object(self, object_name: str, **kwargs):
"""
Delete an object from GCS.
"""
try:
gcs_logging_config: GCSLoggingConfig = await self.get_gcs_logging_config(
kwargs=kwargs
)
headers = await self.construct_request_headers(
vertex_instance=gcs_logging_config["vertex_instance"],
service_account_json=gcs_logging_config["path_service_account"],
)
bucket_name = gcs_logging_config["bucket_name"]
bucket_name, object_name = self._handle_folders_in_bucket_name(
bucket_name=bucket_name,
object_name=object_name,
)
url = f"https://storage.googleapis.com/storage/v1/b/{bucket_name}/o/{object_name}"
# Send the DELETE request to delete the object
response = await self.async_httpx_client.delete(url=url, headers=headers)
if (response.status_code != 200) or (response.status_code != 204):
verbose_logger.error(
"GCS object delete error: %s, status code: %s",
str(response.text),
response.status_code,
)
return None
verbose_logger.debug(
"GCS object delete response status code: %s, response: %s",
response.status_code,
response.text,
)
# Return the content of the downloaded object
return response.text
except Exception as e:
verbose_logger.error("GCS object download error: %s", str(e))
return None
async def _log_json_data_on_gcs(
self,
headers: Dict[str, str],
bucket_name: str,
object_name: str,
logging_payload: Union[StandardLoggingPayload, str],
):
"""
Helper function to make POST request to GCS Bucket in the specified bucket.
"""
if isinstance(logging_payload, str):
json_logged_payload = logging_payload
else:
json_logged_payload = json.dumps(logging_payload, default=str)
bucket_name, object_name = self._handle_folders_in_bucket_name(
bucket_name=bucket_name,
object_name=object_name,
)
response = await self.async_httpx_client.post(
headers=headers,
url=f"https://storage.googleapis.com/upload/storage/v1/b/{bucket_name}/o?uploadType=media&name={object_name}",
data=json_logged_payload,
)
if response.status_code != 200:
verbose_logger.error("GCS Bucket logging error: %s", str(response.text))
verbose_logger.debug("GCS Bucket response %s", response)
verbose_logger.debug("GCS Bucket status code %s", response.status_code)
verbose_logger.debug("GCS Bucket response.text %s", response.text)
return response.json()
|