Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / ludwig-ai/ludwig issues and pull requests

#2896 - Added support for using embedding features in GBMs (one-hot, BERT, etc.)

Pull Request - State: closed - Opened by tgaddair about 2 years ago

#2895 - Confused defined train_step in Trainer and evaluation_step/predict_step in Model?

Issue - State: closed - Opened by ericxsun about 2 years ago - 3 comments

#2893 - WIP: Pytorch 2.0 compiled model

Pull Request - State: closed - Opened by tgaddair about 2 years ago - 1 comment

#2889 - Transient error: Metadata mismatch found in `from_delayed`

Issue - State: closed - Opened by tgaddair about 2 years ago - 2 comments
Labels: bug

#2871 - Add `preprocessing.most_common_percentile`

Pull Request - State: closed - Opened by abidwael about 2 years ago - 2 comments

#2859 - LudwigModel `create_model` static method requires a `config_obj`, but it should take a dict or config obj

Issue - State: closed - Opened by arnavgarg1 about 2 years ago - 3 comments
Labels: bug, easy

#2841 - Allow early stop based on walltime since improvement

Pull Request - State: closed - Opened by jppgks about 2 years ago - 1 comment

#2840 - Bump py39 test to Ray 2.2

Pull Request - State: closed - Opened by geoffreyangus about 2 years ago - 2 comments

#2837 - Support separate `model_resume_path` and `output_directory`

Issue - State: closed - Opened by tgaddair about 2 years ago
Labels: feature

#2822 - NLU error

Issue - State: closed - Opened by martinb-ai about 2 years ago - 9 comments

#2807 - WIP AVMnist dataset

Pull Request - State: closed - Opened by skanjila about 2 years ago - 1 comment

#2745 - Remove extra warnings when images cannot be read.

Pull Request - State: closed - Opened by justinxzhao about 2 years ago - 1 comment

#2727 - Serve example doesn't work (empty feature name?)

Issue - State: closed - Opened by Jeffwan over 2 years ago - 5 comments

#2686 - [test] Ray determinism tests are flaky

Issue - State: closed - Opened by tgaddair over 2 years ago
Labels: tests

#2683 - [fix] Only drop rows where Nan is not in target column

Pull Request - State: closed - Opened by ShreyaR over 2 years ago - 1 comment

#2657 - Add Kaggle tabular dataset and multiscale benchmarksfor multimodal learning

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

#2656 - Expose hyperopt trial output directory

Issue - State: closed - Opened by jeffkinnison over 2 years ago
Labels: feature

#2643 - Modin ray tests failing

Issue - State: closed - Opened by geoffreyangus over 2 years ago
Labels: bug

#2638 - Add new datasets to Ludwig Dataset Zoo

Issue - State: closed - Opened by dantreiman over 2 years ago - 1 comment
Labels: help wanted

#2633 - Presence of hyperopt in config causes issue with model.train

Issue - State: closed - Opened by arnavgarg1 over 2 years ago - 1 comment
Labels: bug

#2588 - `H3` input features fail to save in Dask/Ray Backend

Issue - State: closed - Opened by geoffreyangus over 2 years ago
Labels: bug

#2587 - `Set` output features fail to save in Dask/Ray Backend

Issue - State: closed - Opened by geoffreyangus over 2 years ago
Labels: bug

#2586 - `Text`/`Sequence` output features fail to save in Dask/Ray Backend

Issue - State: closed - Opened by geoffreyangus over 2 years ago
Labels: bug

#2498 - best practice to tune configs to avoid OOM issue

Issue - State: closed - Opened by Jeffwan over 2 years ago - 3 comments

#2412 - additional benchmark configs

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

#2407 - Adds a DotProductCombiner

Pull Request - State: closed - Opened by w4nderlust over 2 years ago - 5 comments

#2402 - Training percentage is not changed

Issue - State: closed - Opened by Jeffwan over 2 years ago - 12 comments

#2375 - Write out data split information as a separate file, i.e. splits.csv, separate from preprocessed data.

Issue - State: closed - Opened by justinxzhao over 2 years ago - 2 comments
Labels: feature, help wanted

#2344 - Return trial parameters from hyperopt before training starts

Issue - State: closed - Opened by arnavgarg1 over 2 years ago - 1 comment
Labels: feature

#2338 - Expose method to pick metrics for output features

Issue - State: closed - Opened by arnavgarg1 over 2 years ago
Labels: feature

#2336 - Categorical output feature doesn't support numerical values

Issue - State: closed - Opened by arnavgarg1 over 2 years ago
Labels: bug

#2330 - Unpinning Dask causes issue when saving image features to parquet

Issue - State: closed - Opened by geoffreyangus over 2 years ago
Labels: bug

#2324 - Ludwig does not gracefully handle empty partitions during saving

Issue - State: closed - Opened by geoffreyangus over 2 years ago - 1 comment
Labels: bug

#2297 - Make `ParameterMetadata` JSON-serializable

Issue - State: closed - Opened by ksbrar over 2 years ago
Labels: bug

#2280 - Save and load pre-trained encoders (transfer learning support)

Pull Request - State: closed - Opened by dantreiman over 2 years ago - 1 comment

#2277 - Use fixture for ray clusters in tests

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

#2221 - KeyError: 'class_weights'

Issue - State: closed - Opened by jiangweiatgithub over 2 years ago - 6 comments

#2216 - [IterML] Add parameter metadata for Feature parameters.

Issue - State: closed - Opened by justinxzhao over 2 years ago

#2215 - [IterML] Add parameter metadata for Combiner parameters.

Issue - State: closed - Opened by justinxzhao over 2 years ago

#2207 - Make preprocessing for predictions backend-aware

Pull Request - State: closed - Opened by hungcs over 2 years ago - 1 comment

#2204 - Add captum dependency within an extra explain package

Pull Request - State: closed - Opened by jeffreyftang over 2 years ago - 1 comment

#2176 - Remove backwards compatibility hacks for older versions of Ray.

Issue - State: open - Opened by justinxzhao over 2 years ago
Labels: productivity and code quality, ray, dependency

#2174 - Informative error message when split col is missing train data

Issue - State: closed - Opened by jppgks over 2 years ago

#2168 - [Hyperopt] Config is merged with defaults twice unnecessarily

Issue - State: closed - Opened by arnavgarg1 over 2 years ago
Labels: productivity and code quality

#2158 - Add the ability to visualize and select a threshold for binary postprocessing

Issue - State: closed - Opened by geoffreyangus over 2 years ago - 1 comment
Labels: feature

#2155 - [Hyperopt] Experiment Management Integrations

Issue - State: closed - Opened by ShreyaR over 2 years ago

#2138 - Category Support for Random Oversampling/Undersampling

Issue - State: closed - Opened by connor-mccorm over 2 years ago
Labels: feature

#2127 - Add validation for mutually exclusive sets of fields in trainer

Issue - State: closed - Opened by ksbrar over 2 years ago - 1 comment
Labels: looking into it

#2099 - Standardize on a preprocessing caching format

Issue - State: closed - Opened by hungcs over 2 years ago

#2092 - Enables logging Torchscript models with MLFlow callback

Pull Request - State: closed - Opened by geoffreyangus over 2 years ago - 1 comment

#2091 - Updates the export_torchscript CLI to match the API

Pull Request - State: closed - Opened by geoffreyangus over 2 years ago - 1 comment

#2090 - Enables saving Torchscript modules during training

Pull Request - State: closed - Opened by geoffreyangus over 2 years ago - 1 comment

#2087 - Add unit tests for `ludwig/utils/audio_utils.py`

Issue - State: closed - Opened by geoffreyangus over 2 years ago
Labels: productivity and code quality

#2085 - Speeds up Torchscript SequenceFeature with parallelism

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

#2011 - refactor/feat: Convert pieces of Ludwig schema to marshmallow-style schemas

Issue - State: closed - Opened by ksbrar over 2 years ago - 1 comment
Labels: feature, productivity and code quality

#1955 - Adds ability to log TorchScript models to MLFlow

Pull Request - State: closed - Opened by geoffreyangus almost 3 years ago - 1 comment

#1794 - Add s4 module

Pull Request - State: closed - Opened by geoffreyangus almost 3 years ago

#1753 - Debugging no-GPU issue

Pull Request - State: closed - Opened by ShreyaR almost 3 years ago - 2 comments

#1750 - Changes for Criteo Benchmarks

Pull Request - State: closed - Opened by ShreyaR almost 3 years ago - 2 comments

#1736 - Add json schema logic for validating 'conditional' fields in combiners

Issue - State: closed - Opened by ksbrar about 3 years ago - 1 comment
Labels: feature

#1695 - Added ATIS to Ludwig datasets

Pull Request - State: closed - Opened by ShreyaR about 3 years ago - 2 comments

#1384 - Added prototype of end to end model training test

Pull Request - State: closed - Opened by ShreyaR over 3 years ago - 1 comment

#1343 - Add Ludwig schema definition to JSON Schema Store

Issue - State: open - Opened by tgaddair over 3 years ago - 2 comments
Labels: feature, help wanted, nice to have

#1275 - Sampled softmax

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

#1131 - Initial time series implementation

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

#1084 - Feature Proposal: Allow numpy array to be accepted as vector feature

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

#1048 - Transfer weights from a pretrained model when the a new feature is added

Pull Request - State: closed - Opened by jenishah about 4 years ago - 1 comment

#1035 - Image Captioning Example

Pull Request - State: closed - Opened by michaelbzhu about 4 years ago - 2 comments

#1025 - Beginning italian readme translation

Pull Request - State: closed - Opened by w4nderlust about 4 years ago

#1024 - feat(progress_tracker): save current learning rate

Pull Request - State: closed - Opened by ifokeev about 4 years ago - 2 comments

#522 - GAN and AutoEncoder

Issue - State: open - Opened by Kantshun over 5 years ago - 5 comments
Labels: feature

#85 - KeyError: 'text'

Issue - State: closed - Opened by loretoparisi almost 6 years ago - 10 comments
Labels: waiting for answer