Ecosyste.ms: Issues

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

GitHub / rstudio/cloudml issues and pull requests

#100 - Improve formatting for prediction response

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#99 - Set default region to us-central1 to fix #95

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 1 comment

#98 - Allow account changes in `cloudml_predict()` and `cloudml_train()`

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 4 comments

#97 - Use MNIST and keras in deployment vignette

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 1 comment

#96 - Support storing deployed models in storage bucket to fix #91

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#95 - consider `europe-west1` or `us-central1` as default regions

Issue - State: closed - Opened by javierluraschi almost 7 years ago - 5 comments

#94 - remove gcloud config options from R functions

Issue - State: closed - Opened by jjallaire almost 7 years ago - 4 comments

#93 - storage functions don't take gcloud argument

Issue - State: closed - Opened by jjallaire almost 7 years ago

#92 - more realistic hello world example

Issue - State: closed - Opened by jjallaire almost 7 years ago

#91 - storage location for model

Issue - State: closed - Opened by jjallaire almost 7 years ago

#90 - error during cloudml_predict

Issue - State: closed - Opened by jjallaire almost 7 years ago - 3 comments

#89 - Adds basic deployment use to deployment vignette

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 1 comment

#88 - Add support in cloudml_predict() for multiple prediction instances

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#87 - A few config changes over the machine types branch

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 1 comment

#86 - Add support for cloudml_deploy()

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 6 comments

#85 - Fix view_run() regression from trials change

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#84 - job data is only partially written in tfruns

Issue - State: closed - Opened by jjallaire almost 7 years ago - 4 comments

#83 - Trial collection to use `-###` folder names instead of subfolders

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 1 comment

#82 - Fix warning with missing keras folder to fix #77

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#81 - Add gcloud parameter on all gcloud related functions

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#80 - Fix keras cache creation and restore

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#79 - Fix keras storage caching and restore

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#78 - verify that keras cache is working correctly

Issue - State: closed - Opened by jjallaire almost 7 years ago - 1 comment

#77 - keras package warning in logs

Issue - State: closed - Opened by jjallaire almost 7 years ago

#76 - Fix training with alternate cloudml file

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#75 - does job_download handle multiple trials gracefully?

Issue - State: closed - Opened by jjallaire almost 7 years ago - 3 comments

#74 - ensure complete capture of hyperparameter trial info

Issue - State: closed - Opened by jjallaire almost 7 years ago - 5 comments

#73 - Fix warnings in windows to fix #55

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#72 - Read runtime version from google cloudml file

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#71 - Add default storage folder to bucket when storage not specified

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#70 - Add support to download multiple trials to fix #67

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#69 - Fix processx crash under windows

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 1 comment

#68 - Revamped specification of configuration info

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#67 - ability to download multiple / all trials

Issue - State: closed - Opened by jjallaire almost 7 years ago

#66 - error: cloudml package not available

Issue - State: closed - Opened by jjallaire almost 7 years ago - 4 comments

#65 - R version bucket name

Issue - State: closed - Opened by jjallaire almost 7 years ago

#64 - use subdirectory for default bucket

Issue - State: closed - Opened by jjallaire almost 7 years ago

#63 - clean out staging bucket?

Issue - State: closed - Opened by jjallaire almost 7 years ago - 2 comments

#62 - Refine hyperparameter tuning

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#61 - Add default region when not present

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#60 - tfruns files pane should show folder rather all files in folder

Issue - State: closed - Opened by jjallaire almost 7 years ago - 1 comment

#59 - tfruns should show cloudml info at the top

Issue - State: closed - Opened by jjallaire almost 7 years ago - 1 comment

#58 - gcloud_file function

Issue - State: closed - Opened by jjallaire almost 7 years ago - 1 comment

#57 - refine hyperparameter tuning

Issue - State: closed - Opened by jjallaire almost 7 years ago

#56 - revamped specification of configuration info

Issue - State: closed - Opened by jjallaire almost 7 years ago - 3 comments

#55 - comprehensive review/test on windows

Issue - State: closed - Opened by jjallaire almost 7 years ago - 1 comment

#54 - Make prints python3 compatible

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 1 comment

#53 - Fix model dir for hypertuning jobs

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#52 - support for training flags

Pull Request - State: closed - Opened by jjallaire almost 7 years ago - 3 comments

#50 - Set NOT_CRAN in travis

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 1 comment

#49 - Only auto-install devtools from CRAN

Pull Request - State: closed - Opened by jjallaire almost 7 years ago - 2 comments

#48 - revised version of gcloud-install

Pull Request - State: closed - Opened by jjallaire almost 7 years ago - 1 comment

#47 - zero-config for cloudml_train

Pull Request - State: closed - Opened by jjallaire almost 7 years ago

#46 - Caching, packrat and travis improvements

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 6 comments

#45 - Fix reference in application validation

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#44 - revised getting started vignette

Pull Request - State: closed - Opened by jjallaire almost 7 years ago

#43 - Address code TODO comments

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#42 - job fails to submit when containing directory includes a dash

Issue - State: closed - Opened by jjallaire almost 7 years ago - 3 comments

#41 - Consider defaults for `account` and `project`

Issue - State: closed - Opened by javierluraschi almost 7 years ago

#40 - Install and validate recent version of R is used in cloudml

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#39 - Move runtime-version field into cloudml section

Pull Request - State: closed - Opened by jjallaire almost 7 years ago - 2 comments

#38 - Example, test and fix to use keras

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 2 comments

#37 - Add summary_writer example to MNIST hyperparam trainer

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#36 - Adds training and initial tuning doc

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#35 - Old TODOs and enhancements

Issue - State: closed - Opened by javierluraschi almost 7 years ago

#34 - Support hyper parameter tuning

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#33 - Improve training time performance

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago - 1 comment

#32 - Improve README and pckdown site

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#31 - Add cloudml tests and travis integration

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#30 - Add packages cache and other train improvements

Pull Request - State: closed - Opened by javierluraschi almost 7 years ago

#29 - WIP: Implement a variety of Google SDK methods

Pull Request - State: closed - Opened by javierluraschi about 7 years ago - 5 comments

#28 - Warning: TensorFlow library wasn't compiled to use SSE4.1

Issue - State: closed - Opened by page1 about 7 years ago - 4 comments

#27 - Setting scale tier

Issue - State: closed - Opened by page1 about 7 years ago - 1 comment

#26 - Model Checkpoints - Missing h5py

Issue - State: closed - Opened by page1 about 7 years ago - 3 comments

#25 - How to install additional packages

Issue - State: closed - Opened by page1 about 7 years ago - 2 comments

#24 - gs_copy not working on windows

Issue - State: closed - Opened by page1 about 7 years ago - 2 comments

#23 - Is this package still NSFW?

Issue - State: closed - Opened by page1 about 7 years ago - 4 comments

#22 - after launching a training job, stream logs to (e.g.) a terminal tab

Issue - State: closed - Opened by kevinushey about 7 years ago - 1 comment

#21 - don't hardcode author (etc) fields in setup.py

Issue - State: closed - Opened by kevinushey about 7 years ago - 1 comment

#20 - avoid duplication of data retrieved from gs:// URLs

Issue - State: closed - Opened by kevinushey about 7 years ago - 6 comments

#19 - ensure that `training_run()` outputs get copied to bucket

Issue - State: closed - Opened by kevinushey about 7 years ago - 3 comments

#18 - mechanism for caching packages?

Issue - State: closed - Opened by kevinushey about 7 years ago - 2 comments

#17 - where should auxiliary configuration be specified?

Issue - State: closed - Opened by kevinushey about 7 years ago - 8 comments

#15 - verify credentials before a training run

Issue - State: closed - Opened by kevinushey about 7 years ago - 1 comment

#14 - provide install_gcloud_sdk() helper

Issue - State: closed - Opened by kevinushey about 7 years ago - 2 comments

#13 - integration with tfruns

Issue - State: closed - Opened by kevinushey about 7 years ago

#12 - consider using RStudio terminals for some commands

Issue - State: closed - Opened by kevinushey about 7 years ago - 1 comment

#11 - separate 'cloudml.yml', 'flags.yml'

Issue - State: closed - Opened by kevinushey about 7 years ago - 4 comments

#10 - cross platform implementation of copy_directory

Pull Request - State: closed - Opened by jjallaire over 7 years ago - 4 comments

#9 - refactoring around overlay construction

Pull Request - State: closed - Opened by kevinushey over 7 years ago - 4 comments

#8 - job_list function returning an error

Issue - State: closed - Opened by jjallaire over 7 years ago - 1 comment

#7 - consider new scheme for job directories

Issue - State: closed - Opened by jjallaire over 7 years ago - 9 comments

#6 - Any plans for JSON API implementation?

Issue - State: closed - Opened by MarkEdmondson1234 over 7 years ago - 7 comments

#5 - improve handling + construction of shell arguments

Pull Request - State: closed - Opened by kevinushey over 7 years ago

#4 - use 'cloudml' prefix for command line arguments we own

Pull Request - State: closed - Opened by kevinushey over 7 years ago - 4 comments

#3 - Introduce config function

Pull Request - State: closed - Opened by jjallaire over 7 years ago

#2 - Filter for custom config values

Pull Request - State: closed - Opened by jjallaire over 7 years ago - 1 comment

#1 - use a config-driven approach to training

Pull Request - State: closed - Opened by kevinushey over 7 years ago - 1 comment