mirror of
https://github.com/BerriAI/litellm.git
synced 2026-09-08 22:21:35 +00:00
fix(ocr): send each provider a health-check document it accepts
Health checks probed every OCR deployment with a PDF, which Cohere Parse rejects, so /health, background health checks, and the UI Test Connection button marked Cohere Parse deployments unhealthy. BaseOCRConfig gains a get_health_check_document hook (PDF by default) that CohereParseConfig overrides with a 1x1 PNG data URI. cohere also gains ocr in the provider endpoint matrix
This commit is contained in:
parent
d3a179f988
commit
004a820116
8 changed files with 87 additions and 7 deletions
|
|
@ -6,14 +6,13 @@ import base64
|
|||
from collections.abc import Awaitable, Callable
|
||||
from typing import TYPE_CHECKING, Final, Literal
|
||||
|
||||
from litellm.types.utils import LIST_BATCHES_SUPPORTED_PROVIDERS
|
||||
from litellm.llms.base_llm.ocr.transformation import BaseOCRConfig, DocumentType
|
||||
from litellm.types.utils import LIST_BATCHES_SUPPORTED_PROVIDERS, LlmProviders
|
||||
|
||||
if TYPE_CHECKING:
|
||||
from litellm.litellm_core_utils.litellm_logging import Logging
|
||||
from litellm.types.utils import ImageResponse
|
||||
|
||||
# Minimal PDF for health checks - base64 encoded 1-page PDF with just "test"
|
||||
TEST_PDF_URL = "data:application/pdf;base64,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"
|
||||
|
||||
# Minimal image for health checks - base64 encoded 512x512 blue circle on a white background PNG
|
||||
TEST_IMAGE_BASE64 = "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"
|
||||
|
|
@ -29,6 +28,14 @@ def get_image_file_for_health_check() -> bytes:
|
|||
return base64.b64decode(TEST_IMAGE_BASE64)
|
||||
|
||||
|
||||
def _ocr_health_check_document(model: str, custom_llm_provider: str) -> DocumentType:
|
||||
from litellm.utils import ProviderConfigManager
|
||||
|
||||
provider: Final = next((known for known in LlmProviders if known.value == custom_llm_provider), None)
|
||||
config: Final = ProviderConfigManager.get_provider_ocr_config(model=model, provider=provider) if provider else None
|
||||
return (config or BaseOCRConfig()).get_health_check_document()
|
||||
|
||||
|
||||
class HealthCheckHelpers:
|
||||
@staticmethod
|
||||
async def ahealth_check_wildcard_models(
|
||||
|
|
@ -247,9 +254,6 @@ class HealthCheckHelpers:
|
|||
),
|
||||
"ocr": lambda: litellm.aocr(
|
||||
**_filter_model_params(model_params=model_params),
|
||||
document={
|
||||
"type": "document_url",
|
||||
"document_url": TEST_PDF_URL,
|
||||
},
|
||||
document=_ocr_health_check_document(model=model, custom_llm_provider=custom_llm_provider),
|
||||
),
|
||||
}
|
||||
|
|
|
|||
|
|
@ -33,6 +33,8 @@ OCR_REQUEST_FORMAT_HEADER: Final = "x-req-format"
|
|||
|
||||
PROVIDER_NATIVE_RESPONSE_KEY: Final = "provider_native_response"
|
||||
|
||||
HEALTH_CHECK_PDF_DATA_URI: Final = "data:application/pdf;base64,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"
|
||||
|
||||
|
||||
def parse_ocr_request_format(value: object) -> OCRRequestFormat:
|
||||
if value == "litellm":
|
||||
|
|
@ -146,6 +148,12 @@ class BaseOCRConfig:
|
|||
"""Whether the Rust OCR bridge may serve this config when it is enabled for the provider."""
|
||||
return True
|
||||
|
||||
def get_health_check_document(self) -> DocumentType:
|
||||
return { # mutable-ok: litellm.aocr rejects any document that is not a dict
|
||||
"type": "document_url",
|
||||
"document_url": HEALTH_CHECK_PDF_DATA_URI,
|
||||
}
|
||||
|
||||
def map_ocr_params(
|
||||
self,
|
||||
non_default_params: dict,
|
||||
|
|
|
|||
|
|
@ -34,6 +34,9 @@ COHERE_PARSE_OUTPUT_FORMAT_PARAM: Final = "output_format"
|
|||
COHERE_PARSE_OUTPUT_FORMATS: Final = ("markdown", "blocks")
|
||||
COHERE_PARSE_DEFAULT_OUTPUT_FORMAT: Final = "markdown"
|
||||
COHERE_PARSE_SUPPORTED_PARAMS: Final = (COHERE_PARSE_OUTPUT_FORMAT_PARAM, OCR_REQUEST_FORMAT_PARAM)
|
||||
COHERE_PARSE_HEALTH_CHECK_IMAGE_DATA_URI: Final = (
|
||||
"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAIAAACQd1PeAAAADElEQVR4nGP4//8/AAX+Av4N70a4AAAAAElFTkSuQmCC"
|
||||
)
|
||||
COHERE_PARSE_IMAGE_ONLY_MESSAGE: Final = (
|
||||
"Cohere Parse only accepts `image_url` documents (an image URL or a base64 image data URI); "
|
||||
"`document_url` and PDF inputs are not supported."
|
||||
|
|
@ -144,6 +147,12 @@ class CohereParseConfig(BaseOCRConfig):
|
|||
def supports_rust_bridge(self) -> bool:
|
||||
return False
|
||||
|
||||
def get_health_check_document(self) -> DocumentType:
|
||||
return { # mutable-ok: litellm.aocr rejects any document that is not a dict
|
||||
"type": "image_url",
|
||||
"image_url": COHERE_PARSE_HEALTH_CHECK_IMAGE_DATA_URI,
|
||||
}
|
||||
|
||||
def _llm_provider(self) -> str:
|
||||
return "cohere"
|
||||
|
||||
|
|
|
|||
|
|
@ -559,6 +559,7 @@
|
|||
"moderations": false,
|
||||
"batches": false,
|
||||
"rerank": true,
|
||||
"ocr": true,
|
||||
"a2a": true,
|
||||
"interactions": true
|
||||
}
|
||||
|
|
|
|||
|
|
@ -594,6 +594,7 @@
|
|||
"moderations": false,
|
||||
"batches": false,
|
||||
"rerank": true,
|
||||
"ocr": true,
|
||||
"a2a": true,
|
||||
"interactions": true
|
||||
}
|
||||
|
|
|
|||
|
|
@ -453,3 +453,32 @@ async def test_realtime_health_check_uses_model_level_vertex_params():
|
|||
"Authorization": "Bearer model-level-token",
|
||||
"x-goog-user-project": "model-level-project",
|
||||
}
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
@pytest.mark.parametrize(
|
||||
"model, custom_llm_provider, expected_document_type, expected_uri_prefix",
|
||||
[
|
||||
("mistral/mistral-ocr-latest", "mistral", "document_url", "data:application/pdf;base64,"),
|
||||
("azure_ai/mistral-document-ai-2512", "azure_ai", "document_url", "data:application/pdf;base64,"),
|
||||
("cohere/parse-v5.0", "cohere", "image_url", "data:image/png;base64,"),
|
||||
("azure_ai/Cohere-parse-v5", "azure_ai", "image_url", "data:image/png;base64,"),
|
||||
],
|
||||
)
|
||||
async def test_ocr_health_check_sends_the_document_kind_the_provider_config_accepts(
|
||||
model, custom_llm_provider, expected_document_type, expected_uri_prefix
|
||||
):
|
||||
handlers = HealthCheckHelpers.get_mode_handlers(
|
||||
model=model,
|
||||
custom_llm_provider=custom_llm_provider,
|
||||
model_params={"model": model, "api_key": "sk-test"},
|
||||
)
|
||||
|
||||
with patch( # test-quality-ok: the public health-check path has no dependency injection seam
|
||||
"litellm.aocr", new_callable=AsyncMock, return_value={}
|
||||
) as mock_aocr:
|
||||
await handlers["ocr"]()
|
||||
|
||||
document = mock_aocr.call_args.kwargs["document"]
|
||||
assert document["type"] == expected_document_type
|
||||
assert document[expected_document_type].startswith(expected_uri_prefix)
|
||||
|
|
|
|||
|
|
@ -166,3 +166,19 @@ async def test_aocr_rejects_pdf_before_calling_foundry(disable_aiohttp_transport
|
|||
|
||||
assert exc_info.value.llm_provider == "azure_ai"
|
||||
assert not route.called
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_ahealth_check_ocr_sends_an_image_to_the_foundry_cohere_parse_deployment(
|
||||
disable_aiohttp_transport, respx_mock
|
||||
):
|
||||
route = respx_mock.post(PARSE_URL).respond(json=_parse_response())
|
||||
|
||||
result = await litellm.ahealth_check(
|
||||
model_params={"model": MODEL, "api_base": API_BASE, "api_key": "test-key"}, mode="ocr"
|
||||
)
|
||||
|
||||
document = json.loads(route.calls.last.request.content)["document"]
|
||||
assert document["type"] == "image_url"
|
||||
assert document["image_url"].startswith("data:image/png;base64,")
|
||||
assert "error" not in result
|
||||
|
|
|
|||
|
|
@ -215,3 +215,15 @@ async def test_aocr_without_api_key_names_the_env_var(disable_aiohttp_transport,
|
|||
await litellm.aocr(model=MODEL, document=IMAGE_DOCUMENT)
|
||||
|
||||
assert not route.called
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_ahealth_check_ocr_sends_an_image_cohere_parse_accepts(disable_aiohttp_transport, respx_mock):
|
||||
route = respx_mock.post(PARSE_URL).respond(json=_markdown_response())
|
||||
|
||||
result = await litellm.ahealth_check(model_params={"model": MODEL, "api_key": "test-key"}, mode="ocr")
|
||||
|
||||
document = json.loads(route.calls.last.request.content)["document"]
|
||||
assert document["type"] == "image_url"
|
||||
assert document["image_url"].startswith("data:image/png;base64,")
|
||||
assert "error" not in result
|
||||
|
|
|
|||
Loading…
Add table
Reference in a new issue