Skip to content

Build AI methods with your coding agent, then run them anywhere: as an MCP for chatbots, as a webapp for people, or via API for your software.

Check extract model capability captions and skip tests if not (#553)

Louis Choquel923076fdcd80
diff --git a/.pipelex/inference/backends/mistral.toml b/.pipelex/inference/backends/mistral.toml
index c8c4b05d9..b695ac70f 100644
--- a/.pipelex/inference/backends/mistral.toml
+++ b/.pipelex/inference/backends/mistral.toml
@@ -135,7 +135,6 @@ costs = { input = 0.4, output = 2.0 }
# TODO: add support to pricing per page
-# --- OCR Models ---------------------------------------------------------------
[mistral-ocr-2503]
model_type = "text_extractor"
model_id = "mistral-ocr-2503"
diff --git a/pipelex/cogt/extract/extract_worker_abstract.py b/pipelex/cogt/extract/extract_worker_abstract.py
index c14af0c9a..0f76db261 100644
--- a/pipelex/cogt/extract/extract_worker_abstract.py
+++ b/pipelex/cogt/extract/extract_worker_abstract.py
@@ -41,6 +41,10 @@ class ExtractWorkerAbstract(InferenceWorkerAbstract):
def is_image_supported(self) -> bool:
return self.inference_model.is_image_supported_for_extract
+ @property
+ def is_caption_supported(self) -> bool:
+ return self.inference_model.is_caption_supported_for_extract
+
def _check_can_perform_job(self, extract_job: ExtractJob):
# This can be overridden by subclasses for specific checks
extract_input = extract_job.extract_input
@@ -52,6 +56,10 @@ class ExtractWorkerAbstract(InferenceWorkerAbstract):
if not self.inference_model.is_pdf_supported_for_extract:
msg = f"Extract engine '{self.inference_model.tag}' does not support PDF extraction."
raise ExtractCapabilityError(msg)
+ if extract_job.job_params.should_caption_images:
+ if not self.inference_model.is_caption_supported_for_extract:
+ msg = f"Extract engine '{self.inference_model.tag}' does not support image captioning."
+ raise ExtractCapabilityError(msg)
async def extract_pages(
self,
diff --git a/pipelex/cogt/model_backends/model_spec.py b/pipelex/cogt/model_backends/model_spec.py
index c40878bd0..d4ba2b641 100644
--- a/pipelex/cogt/model_backends/model_spec.py
+++ b/pipelex/cogt/model_backends/model_spec.py
@@ -56,6 +56,10 @@ class InferenceModelSpec(ConfigModel):
def is_image_supported_for_extract(self) -> bool:
return "image" in self.inputs
+ @property
+ def is_caption_supported_for_extract(self) -> bool:
+ return "captions" in self.outputs
+
def get_instructor_mode(self) -> InstructorMode | None:
if self.structure_method:
return self.structure_method.as_instructor_mode()
diff --git a/pipelex/kit/configs/inference/backends/mistral.toml b/pipelex/kit/configs/inference/backends/mistral.toml
index c8c4b05d9..b695ac70f 100644
--- a/pipelex/kit/configs/inference/backends/mistral.toml
+++ b/pipelex/kit/configs/inference/backends/mistral.toml
@@ -135,7 +135,6 @@ costs = { input = 0.4, output = 2.0 }
# TODO: add support to pricing per page
-# --- OCR Models ---------------------------------------------------------------
[mistral-ocr-2503]
model_type = "text_extractor"
model_id = "mistral-ocr-2503"
diff --git a/tests/integration/pipelex/cogt/test_extract.py b/tests/integration/pipelex/cogt/test_extract.py
index 2ec259503..a26d21a52 100644
--- a/tests/integration/pipelex/cogt/test_extract.py
+++ b/tests/integration/pipelex/cogt/test_extract.py
@@ -30,6 +30,9 @@ class TestExtract:
if not extract_worker.is_pdf_supported:
msg = f"PDF extraction is not supported for this extract worker: '{extract_worker.desc}'"
pytest.skip(msg)
+ if extract_job_params.should_caption_images and not extract_worker.is_caption_supported:
+ msg = f"Image captioning is not supported for this extract worker: '{extract_worker.desc}'"
+ pytest.skip(msg)
extract_job = ExtractJobFactory.make_extract_job(
extract_input=ExtractInput(pdf_uri=file_path),
extract_job_params=extract_job_params,
@@ -58,6 +61,9 @@ class TestExtract:
if not extract_worker.is_pdf_supported:
msg = f"PDF extraction is not supported for this extract worker: '{extract_worker.desc}'"
pytest.skip(msg)
+ if extract_job_params.should_caption_images and not extract_worker.is_caption_supported:
+ msg = f"Image captioning is not supported for this extract worker: '{extract_worker.desc}'"
+ pytest.skip(msg)
extract_job = ExtractJobFactory.make_extract_job(
extract_input=ExtractInput(pdf_uri=url),
extract_job_params=extract_job_params,
@@ -80,6 +86,9 @@ class TestExtract:
if not extract_worker.is_image_supported:
msg = f"Image extraction is not supported for this extract worker: '{extract_worker.desc}'"
pytest.skip(msg)
+ if extract_job_params.should_caption_images and not extract_worker.is_caption_supported:
+ msg = f"Image captioning is not supported for this extract worker: '{extract_worker.desc}'"
+ pytest.skip(msg)
extract_job = ExtractJobFactory.make_extract_job(
extract_input=ExtractInput(image_uri=file_path),
extract_job_params=extract_job_params,
@@ -101,6 +110,9 @@ class TestExtract:
if not extract_worker.is_image_supported:
msg = f"Image extraction is not supported for this extract worker: '{extract_worker.desc}'"
pytest.skip(msg)
+ if extract_job_params.should_caption_images and not extract_worker.is_caption_supported:
+ msg = f"Image captioning is not supported for this extract worker: '{extract_worker.desc}'"
+ pytest.skip(msg)
extract_job = ExtractJobFactory.make_extract_job(
extract_input=ExtractInput(image_uri=url),
extract_job_params=extract_job_params,