feat(inference): make pipeline analysis-aware with dynamic model loading#8
Merged
Merged
Conversation
- _load_model() now accepts analysis_type and reads in_channels/num_classes from config.yaml - Per-analysis-type model cache prevents cross-contamination between deforestation/ice/flood models - _find_best_checkpoint() prefers config.yaml weight path per analysis type - run_inference() accepts analysis_type, pads/crops to correct n_channels, and returns dynamic class counts - run_inference_from_file() and run_inference_from_gee() propagate analysis_type parameter
- Pass body.analysis_type to run_inference_from_gee() in /api/predict - Pass analysis_type to run_inference_from_file() in /api/predict/upload - Enables the API to load the correct model and return correct class counts per analysis type
mvanhorn
pushed a commit
to mvanhorn/ClimateVision
that referenced
this pull request
May 17, 2026
…ing (Climate-Vision#8) * feat(inference): make pipeline analysis-aware with dynamic model loading - _load_model() now accepts analysis_type and reads in_channels/num_classes from config.yaml - Per-analysis-type model cache prevents cross-contamination between deforestation/ice/flood models - _find_best_checkpoint() prefers config.yaml weight path per analysis type - run_inference() accepts analysis_type, pads/crops to correct n_channels, and returns dynamic class counts - run_inference_from_file() and run_inference_from_gee() propagate analysis_type parameter * feat(api): wire analysis_type into prediction endpoints - Pass body.analysis_type to run_inference_from_gee() in /api/predict - Pass analysis_type to run_inference_from_file() in /api/predict/upload - Enables the API to load the correct model and return correct class counts per analysis type --------- Co-authored-by: Olufemi Taiwo <Olufemitaiwo23@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Makes the inference pipeline and API endpoints respect the
analysis_typeparameter, loading the correct model configuration per analysis type.Changes
pipeline.py:_load_model()now readsin_channelsandnum_classesfrom config.yaml per analysis typerun_inference()returns dynamic class counts instead of hardcoded forest metricsrun_inference_from_gee()now downloads real tiles via the GEE downloader and applies SCL cloud maskingapi/main.py: prediction endpoints passanalysis_typethrough to the inference pipelineImpact
analysis_typewas accepted by the API but ignored during inference