🚀 feat(inspect): Add progressbar #3045

ashwinvaidya17 · 2025-10-23T06:07:37Z

📝 Description

Provide a clear summary of the changes and the issue that has been addressed.
🛠️ Fixes 📋 [TASK] Add training progress #3073

✨ Changes

Select what type of change your PR is:

✅ Checklist

Before you submit your pull request, please make sure you have completed the following steps:

📚 I have made the necessary updates to the documentation (if applicable).
🧪 I have written tests that support my changes and prove that my fix is effective or my feature works (if applicable).
🏷️ My PR title follows conventional commit format.

For more information about code review checklists, see the Code Review Checklist.

Signed-off-by: Ashwin Vaidya <[email protected]>

application/backend/src/db/schema.py

application/backend/src/services/job_service.py

application/backend/src/services/training_service.py

maxxgx

Thanks for the PR @ashwinvaidya17. Looks good overall, left some comments

application/backend/src/pydantic_models/job.py

application/backend/src/api/endpoints/job_endpoints.py

application/backend/src/pydantic_models/job.py

application/backend/src/services/job_service.py

application/backend/src/services/training_service.py

application/backend/src/utils/callbacks.py

application/backend/src/services/training_service.py

Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot

Pull Request Overview

This PR adds a progress bar feature to the inspect interface that displays real-time training progress with the ability to cancel jobs. The implementation includes both frontend UI components and backend progress tracking infrastructure.

Key changes:

Adds a new status bar component with integrated progress display for training jobs
Implements real-time progress tracking using Server-Sent Events (SSE) and Lightning callbacks
Introduces job cancellation functionality with corresponding API endpoints

Reviewed Changes

Copilot reviewed 9 out of 10 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
application/ui/src/routes/inspect/inspect.tsx	Adds StatusBar component to the inspect layout
application/ui/src/features/inspect/statusbar/statusbar.component.tsx	Creates status bar container component
application/ui/src/features/inspect/statusbar/items/progressbar.component.tsx	Implements progress bar with job polling and cancellation
application/backend/src/utils/callbacks.py	Adds Lightning callback for progress synchronization
application/backend/src/services/training_service.py	Integrates progress tracking into training workflow
application/backend/src/services/job_service.py	Adds progress streaming and job cancellation methods
application/backend/src/pydantic_models/job.py	Extends job model with stage field and cancellation response
application/backend/src/db/schema.py	Adds stage column to job database schema
application/backend/src/api/endpoints/job_endpoints.py	Adds progress streaming and cancellation endpoints

Files not reviewed (1)

application/ui/package-lock.json: Language not supported

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2025-10-31T08:45:44Z

application/backend/src/utils/callbacks.py

+    def setup(self, trainer: Trainer, pl_module: LightningModule, stage: str) -> None:
+        self._send_progress(0, stage)


The stage parameter should be of type RunningStage to match the _send_progress method signature, but it's declared as str. This will cause a type error.

Copilot · 2025-10-31T08:45:45Z

application/backend/src/utils/callbacks.py

+    def teardown(self, trainer: Trainer, pl_module: LightningModule, stage: RunningStage) -> None:
+        self._send_progress(1.0, stage)


The teardown method signature uses RunningStage type but the setup method uses str type. These should be consistent, and both should use RunningStage type.

application/backend/src/services/training_service.py

Copilot · 2025-10-31T08:45:46Z

application/backend/src/services/job_service.py

+                    cached_still_running = await cls.is_job_still_running(job_id=job_id)
+                    last_status_check = now
+                still_running = cached_still_running
+                yield json.dumps({"progress": job.progress, "stage": job.stage})


The job object is fetched once outside the loop but never refreshed. Progress and stage values will remain static throughout the streaming. The job should be refetched from the database in each iteration.

application/backend/src/api/endpoints/job_endpoints.py

Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot

Pull Request Overview

Copilot reviewed 9 out of 10 changed files in this pull request and generated 5 comments.

Files not reviewed (1)

application/ui/package-lock.json: Language not supported

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

application/backend/src/services/training_service.py

application/ui/src/features/inspect/statusbar/items/progressbar.component.tsx

application/backend/src/services/job_service.py

application/backend/src/services/training_service.py

Co-authored-by: Copilot <[email protected]> Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot

Pull Request Overview

Copilot reviewed 9 out of 10 changed files in this pull request and generated 3 comments.

Files not reviewed (1)

application/ui/package-lock.json: Language not supported

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

application/backend/src/services/training_service.py

application/backend/src/services/job_service.py

maxxgx · 2025-10-31T16:06:08Z

application/backend/src/db/schema.py

    id: Mapped[str] = mapped_column(primary_key=True, default=lambda: str(uuid4()))
    project_id: Mapped[str] = mapped_column(ForeignKey("projects.id"))
    type: Mapped[str] = mapped_column(String(64), nullable=False)
+    stage: Mapped[JobStage] = mapped_column(Enum(JobStage), nullable=False)


What's the advantage of declaring an Enum column in the DB schema?

What is the values that is actually stored in DB, e.g. int or str?

I wonder what happens if you add/remove job stages, will queries still work?

Hmm not sure how it will respond when we modify it. I was following Mark's suggestion #3045 (comment). Maybe we can achieve this in a different manner?

I've run some testing with Enum columns. It seems to be safe, at least with StrEnum, it saves the enum as a string. Adding/removing values in the code does not seem to require the DB to be updated.

maxxgx · 2025-10-31T16:12:16Z

application/backend/src/pydantic_models/job.py

+class JobStage(StrEnum):
+    """Job stages follow PyTorch Lightning stages with the addition of idle stage.
+
+    See ``lightning.pytorch.trainer.states.RunningStage`` for more details.
+    """
+
+    IDLE = "idle"
+    TRAINING = "train"
+    SANITY_CHECKING = "sanity_check"
+    VALIDATING = "validate"
+    TESTING = "test"
+    PREDICTING = "predict"


Not sure if we are going to have other types of jobs in the future, but these are only related to training and there is overlap with job.status. Perhaps, it's more appropriate to store the training stage as part of job.message string

We will have to extract it from the string if we want to show it in the progress bar.

The job message can be reset based on the current stage. So you could just use the whole message.

ATM, it seems you're not setting job.message at all.

In Geti classic, job.step.message was displayed:

Looks like JobStage corresponds to job.step in the image above. However, our job progress is global, so we don't have this mechanism of tracking individual stage progress. Not sure how well this works in terms of UX: users would see progress resetting from 100 to 0 (e.g. from stage "training" to "validate"), but they don't know the total number of stages elapsed or remaining.

application/backend/src/pydantic_models/job.py

maxxgx · 2025-10-31T16:17:19Z

application/backend/src/services/training_service.py

+            logger.debug("Syncing progress with db stopped")
+            synchronization_task.cancel()


You could put this 2 lines in a finally block

maxxgx · 2025-10-31T16:20:56Z

application/backend/src/utils/callbacks.py

+    # Test callbacks
+    def on_test_start(self, trainer: Trainer, pl_module: LightningModule) -> None:
+        """Called when testing starts."""
+        self._send_progress(0, JobStage.TESTING)


How come the progress is 0 here?

The idea is to treat each stage as having different progress instead of clubbing train->test as a single progress.
Currently the progressbar will first show the training progress, and then the testing progress. This just makes the progress a bit verbose.

Added a comment above regarding this "staged" progress tracking approach.

Co-authored-by: Max Xiang <[email protected]> Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot

Pull Request Overview

Copilot reviewed 9 out of 10 changed files in this pull request and generated 1 comment.

Files not reviewed (1)

application/ui/package-lock.json: Language not supported

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2025-11-03T10:15:24Z

application/backend/src/services/training_service.py

        # Capture pytorch stdout logs into logger
        with redirect_stdout(LoggerStdoutWriter()):  # type: ignore[type-var]
-            engine.train(model=anomalib_model, datamodule=datamodule)
+            engine.fit(model=anomalib_model, datamodule=datamodule)


Changed from engine.train() to engine.fit() - this appears to be an API update but should be verified that this method exists and provides the same functionality.

Signed-off-by: Ashwin Vaidya <[email protected]>

MarkRedeman · 2025-11-03T13:14:03Z

application/ui/src/routes/inspect/inspect.tsx

-                </InferenceProvider>
-            </SelectedMediaItemProvider>
-        </Grid>
+        <div style={{ display: 'flex', flexDirection: 'column', height: '100%' }}>


Let's revert this change, the extra div is not needed.

MarkRedeman · 2025-11-03T13:15:39Z

application/ui/src/routes/inspect/inspect.tsx

-        </Grid>
+        <div style={{ display: 'flex', flexDirection: 'column', height: '100%' }}>
+            <Grid
+                areas={['toolbar sidebar', 'canvas sidebar']}


Suggested change

areas={['toolbar sidebar', 'canvas sidebar']}

areas={['toolbar sidebar', 'canvas sidebar', 'footer sidebar']}

The Grid component is essentially a wrapper around css's grid properties we can add a new footer grid cell that is used for displaying the sidebar.

MarkRedeman · 2025-11-03T13:16:25Z

application/ui/src/routes/inspect/inspect.tsx

+        <div style={{ display: 'flex', flexDirection: 'column', height: '100%' }}>
+            <Grid
+                areas={['toolbar sidebar', 'canvas sidebar']}
+                rows={['size-800', 'minmax(0, 1fr)']}


With the above change this should make it so that the footer takes up a minimal height, (do check this though, we might need to tweak this)

Suggested change

rows={['size-800', 'minmax(0, 1fr)']}

rows={['size-800', 'minmax(0, 1fr)', 'auto']}

MarkRedeman · 2025-11-03T13:17:38Z

application/ui/src/routes/inspect/inspect.tsx

+                        <Sidebar />
+                    </InferenceProvider>
+                </SelectedMediaItemProvider>
+            </Grid>
+            <StatusBar />


Suggested change

<Sidebar />

</InferenceProvider>

</SelectedMediaItemProvider>

</Grid>

<StatusBar />

<Sidebar />

<Footer />

</InferenceProvider>

</SelectedMediaItemProvider>

</Grid>

Let's move the statusbar into a Footer component and put it inside of the grid.

MarkRedeman · 2025-11-03T13:20:44Z

application/ui/src/features/inspect/statusbar/statusbar.component.tsx

+export const StatusBar = () => {
+    return (
+        <View gridArea={'statusbar'} backgroundColor={'gray-100'} width={'100%'} height={'30px'} overflow={'hidden'}>


Suggested change

export const StatusBar = () => {

return (

<View gridArea={'statusbar'} backgroundColor={'gray-100'} width={'100%'} height={'30px'} overflow={'hidden'}>

export const Footer = () => {

return (

<View gridArea={'footer'} backgroundColor={'gray-100'} width={'100%'} height={'size-400'} overflow={'hidden'}>

For spacing we try to adhere to spectrum's styling guidelines. This makes it so that the applications' UI is more consistent.

MarkRedeman · 2025-11-03T13:22:12Z

application/ui/src/features/inspect/statusbar/items/progressbar.component.tsx

Currently this component feels very non-idomatic react code. I will send you a component that should follow more of our guidelines.

Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot

Pull Request Overview

Copilot reviewed 10 out of 11 changed files in this pull request and generated 5 comments.

Files not reviewed (1)

application/ui/package-lock.json: Language not supported

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2025-11-04T10:11:52Z

application/ui/src/features/inspect/jobs/show-job-logs.component.tsx

    return (
        <Flex direction='column' gap='size-25'>
-            {query.data?.map((line, idx) => <Text key={idx}> {line}</Text>)}
+            {query.data?.map((line, idx) => <Text key={idx}> {line.text}</Text>)}


The code attempts to access line.text but the fetchSSE function now yields the entire parsed JSON object. This will cause a runtime error if the object doesn't have a text property.

Suggested change

{query.data?.map((line, idx) => <Text key={idx}> {line.text}</Text>)}

{query.data?.map((line, idx) => (

<Text key={idx}>

{typeof line === 'object' && line !== null && 'text' in line

? line.text

: JSON.stringify(line)}

</Text>

))}

Copilot · 2025-11-04T10:11:53Z

application/backend/src/services/training_service.py

        engine = Engine(
            default_root_dir=model.export_path,
            logger=[trackio, tensorboard],
+            devices=[0],  # Only single GPU training is supported for now


Hardcoded device index [0] should be derived from the device parameter or made configurable. This creates inconsistency with the device parameter handling.

Copilot · 2025-11-04T10:11:53Z

application/backend/src/services/training_service.py

+        finally:
+            logger.debug("Syncing progress with db stopped")
+            synchronization_task.cancel()


The synchronization_task variable is referenced in the finally block but may not be defined if an exception occurs before line 79-83. This will cause an UnboundLocalError.

Copilot · 2025-11-04T10:11:53Z

application/backend/src/utils/callbacks.py

+
+    def on_test_epoch_end(self, trainer: Trainer, pl_module: LightningModule) -> None:
+        """Called when a test epoch ends."""
+        progress = (trainer.current_epoch + 1) / trainer.max_epochs if trainer.max_epochs else 0.5


Magic number 0.5 (50%) is used as fallback progress when max_epochs is not available. This should be documented or made configurable to clarify the intended behavior.

Copilot · 2025-11-04T10:11:54Z

application/backend/src/utils/callbacks.py

+
+    def on_predict_epoch_end(self, trainer: Trainer, pl_module: LightningModule) -> None:
+        """Called when a prediction epoch ends."""
+        progress = (trainer.current_epoch + 1) / trainer.max_epochs if trainer.max_epochs else 0.5


Duplicated progress calculation logic with magic number 0.5. This should be extracted to a helper method to avoid code duplication and ensure consistency.

Copilot

Pull Request Overview

Copilot reviewed 9 out of 10 changed files in this pull request and generated 3 comments.

Files not reviewed (1)

application/ui/package-lock.json: Language not supported

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

application/backend/src/pydantic_models/job.py

application/backend/src/services/training_service.py

Copilot · 2025-11-05T09:40:30Z

application/backend/src/services/training_service.py

+        finally:
+            logger.debug("Syncing progress with db stopped")


The raise e statement at line 106 inside the except block causes the finally block to execute, which attempts to cancel synchronization_task. However, if the exception occurred before synchronization_task was created (line 79-83), this will raise an UnboundLocalError. Move the synchronization_task initialization outside the try block or add a check for its existence in the finally block.

Co-authored-by: Copilot <[email protected]> Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot

Pull Request Overview

Copilot reviewed 9 out of 10 changed files in this pull request and generated 3 comments.

Files not reviewed (1)

application/ui/package-lock.json: Language not supported

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2025-11-05T09:45:11Z

application/backend/src/services/training_service.py

+            raise e
+        finally:
+            logger.debug("Syncing progress with db stopped")
+            synchronization_task.cancel()


The synchronization_task.cancel() call is duplicated in both the except and finally blocks. This could cause an error if the task was already cancelled in the except block. Remove the cancellation from the except block and only keep it in the finally block to ensure it's always executed once.

application/backend/src/services/training_service.py

Copilot · 2025-11-05T09:45:11Z

application/ui/src/features/inspect/footer/footer.component.tsx

+                backgroundColor,
+            }}
+        >
+            <Flex direction='row' alignItems='center' width='100px' justifyContent='space-between'>


The hardcoded width of '100px' is repeated in multiple places (lines 110, 138). Extract this to a constant to maintain consistency and make it easier to adjust the width across all components.

Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot

Pull Request Overview

Copilot reviewed 9 out of 10 changed files in this pull request and generated 3 comments.

Files not reviewed (1)

application/ui/package-lock.json: Language not supported

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2025-11-05T15:15:13Z

application/backend/src/services/training_service.py

        engine = Engine(
            default_root_dir=model.export_path,
            logger=[trackio, tensorboard],
+            devices=[0],  # Only single GPU training is supported for now


Hardcoding devices=[0] overrides the device parameter logic that was previously used. This comment indicates a limitation but the implementation ignores the device parameter that gets passed to this function, which could confuse users who expect that parameter to work.

Suggested change

engine = Engine(

default_root_dir=model.export_path,

logger=[trackio, tensorboard],

devices=[0], # Only single GPU training is supported for now

# Use training_device to set devices dynamically

# If training_device is 'cpu', set devices=['cpu'], else set devices=[0] or as appropriate

devices = ['cpu'] if training_device == 'cpu' else [0]

engine = Engine(

default_root_dir=model.export_path,

logger=[trackio, tensorboard],

devices=devices, # Devices now set based on training_device

Copilot · 2025-11-05T15:15:14Z

application/ui/src/features/inspect/footer/footer.component.tsx

+    const handleCancel = async () => {
+        try {
+            if (trainingJob.id === undefined) {
+                throw Error('TODO: jobs should always have an ID');


The error message contains 'TODO' which suggests incomplete implementation. This should either be a proper error message or the condition should be handled differently if jobs can legitimately be without IDs at this point.

Suggested change

throw Error('TODO: jobs should always have an ID');

throw Error(`Cannot cancel job: missing job ID for job ${JSON.stringify(trainingJob)}`);

Copilot · 2025-11-05T15:15:14Z

application/backend/src/services/job_service.py

                # Only check job status every status_check_interval seconds
                if now - last_status_check > status_check_interval:
-                    cached_still_running = await is_job_still_running()
+                    cached_still_running = await cls.is_job_still_running(job_id=job_id)


The is_job_still_running method now raises ResourceNotFoundException if the job is not found, but this is called within a streaming context where the job might be deleted. Consider handling this exception or documenting the expected behavior when a job is deleted mid-stream.

Suggested change

cached_still_running = await cls.is_job_still_running(job_id=job_id)

try:

cached_still_running = await cls.is_job_still_running(job_id=job_id)

except ResourceNotFoundException:

# Job was deleted mid-stream; terminate gracefully

break

Signed-off-by: Ashwin Vaidya <[email protected]>

maxxgx · 2025-11-05T15:23:30Z

application/backend/src/services/training_service.py

+            logger.debug("Syncing progress with db stopped")
+            synchronization_task.cancel()


I think the log would show twice because finally block is run in any case. You can move synchronization_task.cancel() under finally and remove the conditional task cancellation there.

Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot

Pull Request Overview

Copilot reviewed 9 out of 10 changed files in this pull request and generated 3 comments.

Files not reviewed (1)

application/ui/package-lock.json: Language not supported

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2025-11-05T15:36:55Z

application/backend/src/services/training_service.py

        # Capture pytorch stdout logs into logger
        with redirect_stdout(LoggerStdoutWriter()):  # type: ignore[type-var]
-            engine.train(model=anomalib_model, datamodule=datamodule)
+            engine.fit(model=anomalib_model, datamodule=datamodule)


Changed from engine.train() to engine.fit(). Verify that this API change is intentional and that both methods exist in the anomalib library, as this could be a breaking change if train() was the correct method.

Copilot · 2025-11-05T15:36:55Z

application/ui/src/features/inspect/footer/footer.component.tsx

+                backgroundColor,
+            }}
+        >
+            <Flex direction='row' alignItems='center' width='100px' justifyContent='space-between'>


Hard-coded width of '100px' may cause layout issues with longer job messages. Consider using a more flexible width or making this responsive to content.

Copilot · 2025-11-05T15:36:55Z

application/ui/src/features/inspect/footer/footer.component.tsx

+                    {message}
+                </Text>
+            </Flex>
+            <ProgressBar value={progress} aria-label={message} width='100px' showValueLabel={false} />


Hard-coded width of '100px' should match or be derived from the parent Flex container width (also 100px on line 110) to avoid inconsistencies. Consider using a constant or CSS variable.

Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot

Pull Request Overview

Copilot reviewed 11 out of 12 changed files in this pull request and generated 1 comment.

Files not reviewed (1)

application/ui/package-lock.json: Language not supported

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2025-11-06T08:52:59Z

application/backend/src/services/job_service.py


    @staticmethod
-    async def get_job_by_id(job_id: UUID) -> Job | None:
+    async def get_job_by_id(job_id: UUID | str) -> Job | None:


The function signature changed to accept UUID | str but there's no validation or conversion of string inputs to UUID. This could lead to runtime errors if a string is passed that isn't a valid UUID.

Suggested change

async def get_job_by_id(job_id: UUID | str) -> Job | None:

async def get_job_by_id(job_id: UUID | str) -> Job | None:

if isinstance(job_id, str):

try:

job_id = UUID(job_id)

except (ValueError, AttributeError, TypeError):

raise ResourceNotFoundException(resource_id=job_id, resource_name="job")

Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot

Pull Request Overview

Copilot reviewed 10 out of 10 changed files in this pull request and generated 3 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2025-11-06T10:41:29Z

application/ui/src/api/fetch-sse.ts

+                    }
+
+                    try {
+                        yield JSON.parse(message);


The code yields the entire parsed JSON object, but in show-job-logs.component.tsx line 35, it accesses line.text property. The parsed object should have a text property for consistency, or the access pattern needs adjustment.

Suggested change

yield JSON.parse(message);

const parsed = JSON.parse(message);

if (typeof parsed === 'string') {

yield { text: parsed };

} else if (parsed && typeof parsed === 'object' && 'text' in parsed) {

yield parsed;

} else {

yield { text: JSON.stringify(parsed) };

}

Copilot · 2025-11-06T10:41:29Z

application/backend/src/services/job_service.py


    @staticmethod
-    async def get_job_by_id(job_id: UUID) -> Job | None:
+    async def get_job_by_id(job_id: UUID | str) -> Job | None:


Changing the parameter type from UUID to UUID | str is an API change. Ensure this is intentional and that all callers properly handle string UUIDs, as this could lead to validation issues if invalid strings are passed.

Copilot · 2025-11-06T10:41:30Z

application/backend/src/services/training_service.py

        export_format = ExportType.OPENVINO

        # Capture pytorch stdout logs into logger
        with redirect_stdout(LoggerStdoutWriter()):  # type: ignore[type-var]


The change from engine.train() to engine.fit() suggests an API update in the anomalib library. Consider adding a comment explaining this API change or updating related documentation to help future maintainers understand the transition.

Suggested change

with redirect_stdout(LoggerStdoutWriter()): # type: ignore[type-var]

with redirect_stdout(LoggerStdoutWriter()): # type: ignore[type-var]

# Note: anomalib API changed from `engine.train()` to `engine.fit()`.

# See anomalib release notes for details. This ensures compatibility with the latest version.

* Add progress Signed-off-by: Ashwin Vaidya <[email protected]> * Merge fixes Signed-off-by: Ashwin Vaidya <[email protected]> * Fix progress bar Signed-off-by: Ashwin Vaidya <[email protected]> * Update application/backend/src/services/job_service.py Co-authored-by: Copilot <[email protected]> Signed-off-by: Ashwin Vaidya <[email protected]> * Update application/backend/src/pydantic_models/job.py Co-authored-by: Max Xiang <[email protected]> Signed-off-by: Ashwin Vaidya <[email protected]> * Use finally block Signed-off-by: Ashwin Vaidya <[email protected]> * Add Mark's changes Signed-off-by: Ashwin Vaidya <[email protected]> * Use job.message for informing training stage Signed-off-by: Ashwin Vaidya <[email protected]> * Restore callback Signed-off-by: Ashwin Vaidya <[email protected]> * Update application/backend/src/pydantic_models/job.py Co-authored-by: Copilot <[email protected]> Signed-off-by: Ashwin Vaidya <[email protected]> * Add check for synchronization task Signed-off-by: Ashwin Vaidya <[email protected]> * Fix conflicts Signed-off-by: Ashwin Vaidya <[email protected]> * Fix prek issues Signed-off-by: Ashwin Vaidya <[email protected]> * Fix mypy issues Signed-off-by: Ashwin Vaidya <[email protected]> * Fix tests Signed-off-by: Ashwin Vaidya <[email protected]> * Cosmetics Signed-off-by: Ashwin Vaidya <[email protected]> * Update playwright in action Signed-off-by: Ashwin Vaidya <[email protected]> * Revert playwright changes Signed-off-by: Ashwin Vaidya <[email protected]> * Target package-lock to feature branch Signed-off-by: Ashwin Vaidya <[email protected]> --------- Signed-off-by: Ashwin Vaidya <[email protected]> Signed-off-by: Ashwin Vaidya <[email protected]> Co-authored-by: Copilot <[email protected]> Co-authored-by: Max Xiang <[email protected]>

Add progress

4ff613c

Signed-off-by: Ashwin Vaidya <[email protected]>

ashwinvaidya17 requested review from MarkRedeman and maxxgx October 23, 2025 06:07

ashwinvaidya17 requested a review from samet-akcay as a code owner October 23, 2025 06:07

Merge branch 'feature/geti-inspect' into ashwin/feat/progress_bar_sse

12e5c52

MarkRedeman reviewed Oct 23, 2025

View reviewed changes

application/backend/src/db/schema.py Outdated Show resolved Hide resolved

application/backend/src/services/job_service.py Outdated Show resolved Hide resolved

application/backend/src/services/training_service.py Outdated Show resolved Hide resolved

maxxgx reviewed Oct 23, 2025

View reviewed changes

MarkRedeman reviewed Oct 23, 2025

View reviewed changes

application/backend/src/services/training_service.py Show resolved Hide resolved

ashwinvaidya17 added 2 commits October 31, 2025 09:22

Merge branch 'feature/geti-inspect' into ashwin/feat/progress_bar_sse

66b2f46

Merge fixes

f3855a2

Signed-off-by: Ashwin Vaidya <[email protected]>

ashwinvaidya17 marked this pull request as draft October 31, 2025 08:44

Copilot AI review requested due to automatic review settings October 31, 2025 08:44

Copilot AI reviewed Oct 31, 2025

View reviewed changes

Fix progress bar

12130d2

Signed-off-by: Ashwin Vaidya <[email protected]>

ashwinvaidya17 marked this pull request as ready for review October 31, 2025 13:01

Copilot AI review requested due to automatic review settings October 31, 2025 13:01

Copilot AI reviewed Oct 31, 2025

View reviewed changes

Update application/backend/src/services/job_service.py

4e05737

Co-authored-by: Copilot <[email protected]> Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot AI review requested due to automatic review settings October 31, 2025 13:13

Copilot AI reviewed Oct 31, 2025

View reviewed changes

application/backend/src/services/training_service.py Show resolved Hide resolved

application/backend/src/services/training_service.py Show resolved Hide resolved

application/backend/src/services/job_service.py Show resolved Hide resolved

maxxgx reviewed Oct 31, 2025

View reviewed changes

Update application/backend/src/pydantic_models/job.py

004e880

Co-authored-by: Max Xiang <[email protected]> Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot AI review requested due to automatic review settings November 3, 2025 10:14

Copilot AI reviewed Nov 3, 2025

View reviewed changes

Use finally block

9776eb4

Signed-off-by: Ashwin Vaidya <[email protected]>

maxxgx linked an issue Nov 3, 2025 that may be closed by this pull request

📋 [TASK] Add training progress #3073

Open

MarkRedeman reviewed Nov 3, 2025

View reviewed changes

Add Mark's changes

e29cb09

Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot AI review requested due to automatic review settings November 4, 2025 10:10

Copilot AI reviewed Nov 4, 2025

View reviewed changes

Copilot AI reviewed Nov 5, 2025

View reviewed changes

Update application/backend/src/pydantic_models/job.py

5e25b9a

Co-authored-by: Copilot <[email protected]> Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot AI review requested due to automatic review settings November 5, 2025 09:43

Copilot AI reviewed Nov 5, 2025

View reviewed changes

ashwinvaidya17 added 3 commits November 5, 2025 10:50

Add check for synchronization task

873e5de

Signed-off-by: Ashwin Vaidya <[email protected]>

Merge branch 'feature/geti-inspect' into ashwin/feat/progress_bar_sse

00a9b71

Fix conflicts

6791180

Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot AI review requested due to automatic review settings November 5, 2025 15:14

Copilot AI reviewed Nov 5, 2025

View reviewed changes

Fix prek issues

eb057e2

Signed-off-by: Ashwin Vaidya <[email protected]>

maxxgx reviewed Nov 5, 2025

View reviewed changes

maxxgx approved these changes Nov 5, 2025

View reviewed changes

Fix mypy issues

b5a87d6

Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot AI review requested due to automatic review settings November 5, 2025 15:35

Copilot AI reviewed Nov 5, 2025

View reviewed changes

ashwinvaidya17 added 3 commits November 6, 2025 09:24

Fix tests

48d762a

Signed-off-by: Ashwin Vaidya <[email protected]>

Cosmetics

ddc1152

Signed-off-by: Ashwin Vaidya <[email protected]>

Update playwright in action

422a38c

Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot AI review requested due to automatic review settings November 6, 2025 08:52

ashwinvaidya17 requested review from AlexanderBarabanov and mramotowski as code owners November 6, 2025 08:52

Copilot AI reviewed Nov 6, 2025

View reviewed changes

AlexanderBarabanov approved these changes Nov 6, 2025

View reviewed changes

ashwinvaidya17 added 2 commits November 6, 2025 11:22

Revert playwright changes

b19bfb0

Signed-off-by: Ashwin Vaidya <[email protected]>

Target package-lock to feature branch

db05d27

Signed-off-by: Ashwin Vaidya <[email protected]>

Copilot AI review requested due to automatic review settings November 6, 2025 10:40

Copilot AI reviewed Nov 6, 2025

View reviewed changes

MarkRedeman approved these changes Nov 6, 2025

View reviewed changes

ashwinvaidya17 merged commit 32f527a into open-edge-platform:feature/geti-inspect Nov 6, 2025
17 of 18 checks passed

		def setup(self, trainer: Trainer, pl_module: LightningModule, stage: str) -> None:
		self._send_progress(0, stage)

		def teardown(self, trainer: Trainer, pl_module: LightningModule, stage: RunningStage) -> None:
		self._send_progress(1.0, stage)

		logger.debug("Syncing progress with db stopped")
		synchronization_task.cancel()

	areas={['toolbar sidebar', 'canvas sidebar']}
	areas={['toolbar sidebar', 'canvas sidebar', 'footer sidebar']}

	rows={['size-800', 'minmax(0, 1fr)']}
	rows={['size-800', 'minmax(0, 1fr)', 'auto']}

-            {query.data?.map((line, idx) => <Text key={idx}> {line.text}</Text>)}
+            {query.data?.map((line, idx) => (
+                <Text key={idx}>
+                    {typeof line === 'object' && line !== null && 'text' in line
+                        ? line.text
+                        : JSON.stringify(line)}
+                </Text>
+            ))}

	throw Error('TODO: jobs should always have an ID');
	throw Error(`Cannot cancel job: missing job ID for job ${JSON.stringify(trainingJob)}`);

-    async def get_job_by_id(job_id: UUID | str) -> Job | None:
+    async def get_job_by_id(job_id: UUID | str) -> Job | None:
+        if isinstance(job_id, str):
+            try:
+                job_id = UUID(job_id)
+            except (ValueError, AttributeError, TypeError):
+                raise ResourceNotFoundException(resource_id=job_id, resource_name="job")

-                        yield JSON.parse(message);
+                        const parsed = JSON.parse(message);
+                        if (typeof parsed === 'string') {
+                            yield { text: parsed };
+                        } else if (parsed && typeof parsed === 'object' && 'text' in parsed) {
+                            yield parsed;
+                        } else {
+                            yield { text: JSON.stringify(parsed) };
+                        }

🚀 feat(inspect): Add progressbar #3045

🚀 feat(inspect): Add progressbar #3045

Uh oh!

Conversation

ashwinvaidya17 commented Oct 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

📝 Description

✨ Changes

✅ Checklist

Uh oh!

Uh oh!

Uh oh!

Uh oh!

maxxgx left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Reviewed Changes

Uh oh!

Copilot AI Oct 31, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Oct 31, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Copilot AI Oct 31, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

maxxgx Oct 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ashwinvaidya17 commented Oct 23, 2025 •

edited

Loading

maxxgx Oct 31, 2025 •

edited

Loading