AllenInstitute/croissant issues and pull requests

#93 - Bump mlflow from 1.10.0 to 2.5.0

Pull Request - State: open - Opened by dependabot[bot] over 1 year ago
Labels: dependencies

#92 - Bump mlflow from 1.10.0 to 2.3.1

Pull Request - State: closed - Opened by dependabot[bot] over 1 year ago - 1 comment
Labels: dependencies

#91 - Bump mlflow from 1.10.0 to 2.2.1

Pull Request - State: closed - Opened by dependabot[bot] almost 2 years ago - 1 comment
Labels: dependencies

#90 - Bump mlflow from 1.10.0 to 1.23.1

Pull Request - State: closed - Opened by dependabot[bot] almost 3 years ago - 1 comment
Labels: dependencies

#89 - separates out dependencies in extras

Pull Request - State: closed - Opened by djkapner over 3 years ago

#88 - bumps argschema version

Pull Request - State: closed - Opened by djkapner almost 4 years ago

#87 - just-in-time labeling job(s) and visualizations

Issue - State: closed - Opened by wbwakeman about 4 years ago - 15 comments

#86 - Improve and present performance on CNN using transfer learning

Issue - State: closed - Opened by wbwakeman about 4 years ago

#85 - Evaluate and prepare additional features for classifier

Issue - State: closed - Opened by wbwakeman about 4 years ago

#84 - Make curve of model performance versus amount of data using existing labeled data

Issue - State: closed - Opened by wbwakeman about 4 years ago - 1 comment

#83 - Make curve of model performance versus amount of data using production ROI set

Issue - State: open - Opened by wbwakeman about 4 years ago

#82 - Summarize disagreement in previous labeling efforts

Issue - State: closed - Opened by wbwakeman about 4 years ago

#81 - Investigate transfer learning for CNN classifier

Issue - State: closed - Opened by wbwakeman about 4 years ago

#80 - Investigate CNN deep learning solution for segmentation classifier

Issue - State: closed - Opened by wbwakeman about 4 years ago - 1 comment

#79 - Report classifier progress to stakeholders

Issue - State: closed - Opened by wbwakeman about 4 years ago - 1 comment

#78 - Targeted training and feature engineering for Slc set

Issue - State: closed - Opened by wbwakeman about 4 years ago - 2 comments

#77 - Add some more helpful info in readme regarding training.

Pull Request - State: closed - Opened by kschelonka about 4 years ago

#76 - Update train methods

Pull Request - State: closed - Opened by kschelonka about 4 years ago

#75 - Incorporate classifier into production pipeline

Issue - State: open - Opened by wbwakeman about 4 years ago

#74 - adds tag

Pull Request - State: closed - Opened by djkapner over 4 years ago

#73 - Train model based on 500 ROI Slc labeling job

Issue - State: closed - Opened by wbwakeman over 4 years ago - 2 comments

#72 - Additional features from old segmentation + new

Pull Request - State: closed - Opened by kschelonka over 4 years ago - 1 comment

#71 - Investigate how models are mis-classifying ROIs in dense experiments

Issue - State: closed - Opened by wbwakeman over 4 years ago - 2 comments

#70 - Investigate cell matching and train classifier using current production data

Issue - State: closed - Opened by wbwakeman over 4 years ago - 1 comment
Labels: wontfix

#69 - Run models with new labeling data and report performance

Issue - State: closed - Opened by wbwakeman over 4 years ago

#68 - Ingest new labeling outputs

Issue - State: closed - Opened by wbwakeman over 4 years ago

#67 - Plan additional features based on older segmentation and implemented selection

Issue - State: closed - Opened by wbwakeman over 4 years ago

#66 - Engage with SMEs for feature engineering effort

Issue - State: closed - Opened by kschelonka over 4 years ago - 1 comment

#65 - Rerun models without trace-derived features

Issue - State: closed - Opened by djkapner over 4 years ago

#64 - auto-deploy to pypi

Issue - State: open - Opened by djkapner over 4 years ago - 1 comment

#63 - Update init.py

Pull Request - State: closed - Opened by djkapner over 4 years ago

#62 - updates doctring for training

Pull Request - State: closed - Opened by djkapner over 4 years ago - 1 comment

#61 - Optionally remove features from classifier training/prediction set

Pull Request - State: closed - Opened by kschelonka over 4 years ago - 4 comments

#60 - adds ability to specify image from CLI

Pull Request - State: closed - Opened by djkapner over 4 years ago - 1 comment

#59 - GH 41 & 53: Additional features from feature exploration, with columns in deterministic order

Pull Request - State: closed - Opened by kschelonka over 4 years ago - 2 comments

#58 - Low priority: Increase online_training_helper timeout to 40 seconds or add retry

Issue - State: open - Opened by kschelonka over 4 years ago

#57 - Summarize model generalization performance

Issue - State: closed - Opened by kschelonka over 4 years ago - 2 comments

#56 - Update fargate to take docker image as a parameter to enable easier prototyping

Issue - State: closed - Opened by kschelonka over 4 years ago

#55 - Feature extractor should probably check for (and deal with) NaN values when computing features

Issue - State: closed - Opened by njmei over 4 years ago - 1 comment
Labels: bug

#54 - Consider removing `full_genotype` and `targeted_structure` fields from consideration of classifier

Issue - State: closed - Opened by njmei over 4 years ago - 1 comment

#53 - Ordering of metadata columns is not deterministic in FeatureExtractor

Issue - State: closed - Opened by kschelonka over 4 years ago
Labels: bug

#52 - Add methods to select hyperparameters for LogisticRegression and RandomForestClassifier models

Pull Request - State: closed - Opened by kschelonka over 4 years ago - 4 comments

#51 - makes sure stack container task can R/W its own bucket

Pull Request - State: closed - Opened by djkapner over 4 years ago - 1 comment

#50 - Update init.py

Pull Request - State: closed - Opened by djkapner over 4 years ago

#49 - Update setup.py

Pull Request - State: closed - Opened by djkapner over 4 years ago

#48 - missing comma

Pull Request - State: closed - Opened by djkapner over 4 years ago - 1 comment

#47 - Feature/support py377

Pull Request - State: closed - Opened by djkapner over 4 years ago

#46 - adds codecov config file to stop flagging codecov issues with red x

Pull Request - State: closed - Opened by djkapner over 4 years ago

#45 - Classifier feature engineering incorporating neuropil traces

Issue - State: closed - Opened by wbwakeman over 4 years ago - 4 comments

#44 - pins mlflow

Pull Request - State: closed - Opened by djkapner over 4 years ago

#43 - Containerizes and sets up fargate

Pull Request - State: closed - Opened by djkapner over 4 years ago - 5 comments

#42 - Classifier training includes hyperparameter tuning and model selection.

Issue - State: closed - Opened by kschelonka over 4 years ago

#41 - Features from data exploration and prototype are implemented in production code

Issue - State: closed - Opened by kschelonka over 4 years ago

#40 - Classifier feature engineering/evaluation

Issue - State: closed - Opened by wbwakeman over 4 years ago - 2 comments

#39 - Create Feature Extraction and Inference Module

Issue - State: closed - Opened by isaak-willett over 4 years ago - 1 comment
Labels: wontfix

#38 - ~Containerize Croissant Croissant and MlFlow~

Issue - State: closed - Opened by isaak-willett over 4 years ago - 1 comment
Labels: wontfix

#37 - LIMS work for new Feature Extraction and Inference workflow step

Issue - State: closed - Opened by isaak-willett over 4 years ago - 1 comment

#36 - adds training instructions to README

Pull Request - State: closed - Opened by djkapner over 4 years ago - 1 comment

#35 - adds ability to load test/train data from s3

Pull Request - State: closed - Opened by djkapner over 4 years ago - 1 comment

#34 - Tweaks/aws postgres mlflow tracking backend

Pull Request - State: closed - Opened by djkapner over 4 years ago - 2 comments

#33 - makes read_jsonlines a generator and moves to utils

Pull Request - State: closed - Opened by djkapner over 4 years ago - 1 comment

#32 - Feature/merge sgt outputs

Pull Request - State: closed - Opened by djkapner over 4 years ago - 1 comment

#31 - Access rules for security group and roles are secure and only allow desired traffic.

Issue - State: closed - Opened by kschelonka over 4 years ago - 1 comment

#30 - MLFlow Experiment Cloud Requirements

Pull Request - State: closed - Opened by kschelonka over 4 years ago - 1 comment

#29 - Developers can launch model training on AWS

Issue - State: closed - Opened by kschelonka over 4 years ago

#28 - Build infrastructure for cross-validating model algorithms

Issue - State: closed - Opened by wbwakeman over 4 years ago - 1 comment

#27 - Classifier feature evaluation/engineering

Issue - State: closed - Opened by wbwakeman over 4 years ago - 1 comment

#26 - Exclude punctuation in mlflow db secret

Pull Request - State: closed - Opened by kschelonka over 4 years ago - 1 comment

#25 - The final labeled dataset from SageMaker jobs is available to use for training.

Issue - State: closed - Opened by kschelonka over 4 years ago - 1 comment

#24 - Add cloudformation stacks for deploying resources for MLFlow

Pull Request - State: closed - Opened by kschelonka over 4 years ago - 3 comments

#23 - Feature/ingest s3 labels

Pull Request - State: closed - Opened by kschelonka over 4 years ago - 2 comments

#22 - Adds MVP Training Module and Input Schema

Pull Request - State: closed - Opened by isaak-willett over 4 years ago - 2 comments

#21 - Add mlflow docker file and README instructions.

Pull Request - State: closed - Opened by kschelonka over 4 years ago - 1 comment

#20 - Feature: MVP Feature generation

Pull Request - State: closed - Opened by kschelonka over 4 years ago - 2 comments

#19 - Added dev data for testing and validating functions

Pull Request - State: closed - Opened by isaak-willett over 4 years ago - 15 comments

#18 - Productionize MLFlow Cloud deployment

Issue - State: closed - Opened by kschelonka over 4 years ago - 1 comment

#17 - AWS Training jobs can connect to cloud resources for MLFlow

Issue - State: closed - Opened by kschelonka over 4 years ago - 1 comment

#16 - Prototype MLFlow cloud deployment

Issue - State: closed - Opened by kschelonka over 4 years ago

#15 - Setting up CI and MLFlow Project Structure

Pull Request - State: closed - Opened by isaak-willett over 4 years ago

#14 - Update README.md

Pull Request - State: closed - Opened by isaak-willett over 4 years ago

#13 - Adding CI and MLProject files

Pull Request - State: closed - Opened by isaak-willett over 4 years ago

#12 - MVP Train/Tune

Issue - State: closed - Opened by kschelonka over 4 years ago - 1 comment

#11 - Full CI/CD setup

Issue - State: open - Opened by kschelonka over 4 years ago

#10 - Classifier full development

Issue - State: open - Opened by kschelonka over 4 years ago - 1 comment
Labels: Epic

#9 - MVP Feature Generation

Issue - State: closed - Opened by kschelonka over 4 years ago

#8 - Classifier Full Development

Issue - State: closed - Opened by kschelonka over 4 years ago

#7 - Generate/pull data for development and testing

Issue - State: closed - Opened by kschelonka over 4 years ago - 3 comments

#6 - Finalize design for MLFlow Cloud Deployment

Issue - State: closed - Opened by kschelonka over 4 years ago - 1 comment

#5 - Ingest human-labeled data from s3

Issue - State: closed - Opened by kschelonka over 4 years ago

#4 - Define functional MVP requirements

Issue - State: closed - Opened by kschelonka over 4 years ago

#3 - MLFlow Cloud Deployment

Issue - State: closed - Opened by kschelonka over 4 years ago
Labels: Epic

#2 - Classifier MVP

Issue - State: closed - Opened by kschelonka over 4 years ago - 1 comment
Labels: Epic

#1 - Set up github repository, CI/CD

Issue - State: closed - Opened by kschelonka over 4 years ago

GitHub / AllenInstitute/croissant issues and pull requests