From f3b198c1b7aaa890c6a17b1bb02e65750370231f Mon Sep 17 00:00:00 2001 From: mateo-berri <277851410+mateo-berri@users.noreply.github.com> Date: Fri, 18 Sep 2026 21:56:43 -0700 Subject: [PATCH] fix(mistral): accept user_data as the OCR file purpose and keep OCR cost warnings single-line --- litellm/cost_calculator.py | 12 ++++--- litellm/llms/custom_httpx/llm_http_handler.py | 5 ++- litellm/llms/mistral/files/transformation.py | 31 +++++++++++-------- .../test_mistral_files_transformation.py | 22 ++++++++++++- 4 files changed, 51 insertions(+), 19 deletions(-) diff --git a/litellm/cost_calculator.py b/litellm/cost_calculator.py index 352d95a4fdf..30f2cb8489c 100644 --- a/litellm/cost_calculator.py +++ b/litellm/cost_calculator.py @@ -2146,8 +2146,8 @@ def ocr_batch_cost( if resolved_info is None: verbose_logger.warning( "OCR batch cost: model=%s custom_llm_provider=%s has no pricing entry; returning 0.0 cost.", - model, - custom_llm_provider, + _single_log_line(model), + _single_log_line(custom_llm_provider), ) return 0.0, 0.0 @@ -2159,14 +2159,18 @@ def ocr_batch_cost( verbose_logger.warning( "OCR batch cost: model=%s custom_llm_provider=%s reported pages_processed=%s but no " "ocr_cost_per_page is configured; returning 0.0 cost for those pages.", - model, - custom_llm_provider, + _single_log_line(model), + _single_log_line(custom_llm_provider), pages_processed, ) effective_annotation_rate: Final = annotation_rate if annotation_rate is not None else page_rate return (page_rate or 0.0) * pages_processed + (effective_annotation_rate or 0.0) * annotation_pages, 0.0 +def _single_log_line(value: str | None) -> str: + return str(value).replace("\n", "").replace("\r", "") + + def _lookup_model_info_or_none(model: str, custom_llm_provider: str | None) -> ModelInfo | None: try: return litellm.get_model_info(model=model, custom_llm_provider=custom_llm_provider) diff --git a/litellm/llms/custom_httpx/llm_http_handler.py b/litellm/llms/custom_httpx/llm_http_handler.py index 196b437c66b..8e0cdf547a2 100644 --- a/litellm/llms/custom_httpx/llm_http_handler.py +++ b/litellm/llms/custom_httpx/llm_http_handler.py @@ -307,7 +307,10 @@ def _mask_presigned_request_headers(transformed_request: bytes | str | dict) -> _get_masked_values, # pyright: ignore[reportPrivateUsage] # the shared header-masking helper has no public name ) - return {**transformed_request, "headers": _get_masked_values(request_headers)} + return { # mutable-ok: logging's curl and raw-request builders take dict + **transformed_request, + "headers": _get_masked_values(request_headers), + } def _aws_signing_overrides(optional_params: Mapping[str, Any], litellm_params: Mapping[str, Any]) -> Mapping[str, Any]: diff --git a/litellm/llms/mistral/files/transformation.py b/litellm/llms/mistral/files/transformation.py index bf1ef7cb69f..6e64961485a 100644 --- a/litellm/llms/mistral/files/transformation.py +++ b/litellm/llms/mistral/files/transformation.py @@ -8,6 +8,7 @@ Mistral only accepts ``fine-tune``, ``batch`` and ``ocr`` as upload purposes. import time from collections.abc import Mapping, Sequence +from types import MappingProxyType from typing import Final, Literal, TypeAlias import httpx @@ -33,6 +34,14 @@ from ..common_utils import get_mistral_api_base, get_mistral_auth_headers, mistr MistralFilePurpose: TypeAlias = Literal["fine-tune", "batch", "ocr"] +_OPENAI_PURPOSE_BY_MISTRAL: Final[Mapping[MistralFilePurpose, OpenAIFilesPurpose]] = MappingProxyType( + {"fine-tune": "fine-tune", "batch": "batch", "ocr": "user_data"} +) +_MISTRAL_PURPOSE_BY_OPENAI: Final[Mapping[str, MistralFilePurpose]] = MappingProxyType( + {"fine-tune": "fine-tune", "batch": "batch", "ocr": "ocr", "user_data": "ocr"} +) +_SUPPORTED_PURPOSES: Final = ", ".join(_MISTRAL_PURPOSE_BY_OPENAI) + _NO_QUERY_PARAMS: Final[dict[str, str]] = {} # mutable-ok: BaseFilesConfig request transforms return tuple[str, dict] @@ -81,22 +90,18 @@ def _to_openai_file_object(file: MistralFile) -> OpenAIFileObject: def _to_openai_purpose(purpose: MistralFilePurpose) -> OpenAIFilesPurpose: - match purpose: - case "fine-tune" | "batch": - return purpose - case "ocr": - return "user_data" + return _OPENAI_PURPOSE_BY_MISTRAL[purpose] def _to_mistral_purpose(purpose: str) -> MistralFilePurpose: - """Only Mistral's own purposes pass through. Silently mapping anything else to ``batch`` - would let an upload skip the proxy's batch-file validation and guardrails, which only - run when the caller says ``purpose=batch``.""" - match purpose: - case "batch" | "fine-tune" | "ocr": - return purpose - case _: - raise ValueError(f"Mistral does not support purpose={purpose!r}. Use one of: batch, fine-tune, ocr") + """``user_data`` is what an OCR file reads back as, since OpenAI's purpose literal has no ``ocr``, + so it maps back onto ``ocr``. Every other purpose Mistral lacks is rejected: silently rewriting + it to ``batch`` would let an upload skip the proxy's batch-file validation and guardrails, which + only run when the caller says ``purpose=batch``.""" + mistral_purpose: Final = _MISTRAL_PURPOSE_BY_OPENAI.get(purpose) + if mistral_purpose is None: + raise ValueError(f"Mistral does not support purpose={purpose!r}. Use one of: {_SUPPORTED_PURPOSES}") + return mistral_purpose def _api_base_from(litellm_params: Mapping[str, object]) -> str: diff --git a/tests/test_litellm/llms/mistral/files/test_mistral_files_transformation.py b/tests/test_litellm/llms/mistral/files/test_mistral_files_transformation.py index b81740c0429..1dcd6d92a4b 100644 --- a/tests/test_litellm/llms/mistral/files/test_mistral_files_transformation.py +++ b/tests/test_litellm/llms/mistral/files/test_mistral_files_transformation.py @@ -102,7 +102,17 @@ def test_upload_request_passes_mistral_purposes_through(config, purpose): assert body["purpose"] == (None, purpose) -@pytest.mark.parametrize("purpose", ["assistants", "user_data", "vision", "evals"]) +def test_upload_request_maps_user_data_onto_ocr(config): + body = config.transform_create_file_request( + model="", + create_file_data=CreateFileRequest(file=("scan.pdf", b"%PDF"), purpose="user_data"), + optional_params={}, + litellm_params={}, + ) + assert body["purpose"] == (None, "ocr") + + +@pytest.mark.parametrize("purpose", ["assistants", "vision", "evals"]) def test_upload_request_rejects_purposes_mistral_lacks(config, purpose): """Regression: these used to be silently rewritten to ``batch``, so an upload that skipped the proxy's batch-only validation and guardrails still landed on Mistral as a batch input file.""" @@ -191,6 +201,16 @@ def test_list_request_filters_by_mapped_purpose(config): assert no_params == {} +def test_list_request_accepts_the_purpose_an_ocr_file_reads_back_as(config): + """Regression: an OCR file reads back as ``purpose=user_data``, and listing with that purpose + used to raise, so ``files.list(purpose=file.purpose)`` could never find OCR files.""" + ocr_file = config.transform_retrieve_file_response( + raw_response=_response(_file(purpose="ocr")), logging_obj=None, litellm_params={} + ) + _, params = config.transform_list_files_request(purpose=ocr_file.purpose, optional_params={}, litellm_params={}) + assert params == {"purpose": "ocr"} + + def test_list_request_rejects_purposes_mistral_lacks(config): with pytest.raises(ValueError, match="purpose='assistants'"): config.transform_list_files_request(purpose="assistants", optional_params={}, litellm_params={})