Nobunaga #782
Merged
Nobunaga #782
Conversation
This file contains 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
* constrain on the dataset to pull * added config file for test * constant.py approach instead * import error fix (#767) Co-authored-by: Eugene-hu <85906264+Eugene-hu@users.noreply.github.com>
* Add .circleci/config.yml * Add .circleci/config.yml * Updated config.yml * Updated config.yml * Updated config.yml * Updated config.yml * override default port testing server compression * 4-way parallelism * try smaller docker image * fix config * use small docker container * remove reference to undefined variable * add more parallelism * import error fix (#767) * remove local parallelism * make test_forward flaky * fix typo in keyword * Updated config.yml * Updated config.yml * Updated config.yml * Add .circleci/config.yml * Updated config.yml * better logs * better logs * might as well try sleeping * add sleep; see if flaky test is due to process still running * store durations * split timing by classname * use filename for timing * try two cores * use pytest-split to manage jobs across runners * add pytest-split dependency * fix syntax * try junit v1 with classname timings * try xunit2 with classnames split * try fixing relative test path names * work from parent dir * timing by filename * copy pytest.ini to root directory when testing * make timing type explicit * remove circleci docs from config file * try 4x parallelism * Updated config.yml * Updated config.yml * Updated config.yml * Updated config.yml * separate unit tests and integration tests * use pytest-split * Updated config.yml * use config file in its proper place * disable failing test that actually tests nothing * save durations to folder * don't overwrite test results * make test-timings if it doesnt exist * add version for test-timings * add mkdir -p * use xdist * add second job in workflow to collate results * add second job in workflow to collate results * add second job in workflow to collate results * use static durations file in VCS for now * fix typo * clean up for PR * don't base mock subtensor port off of pid * use relative import * distribute by module in local procs * use loadfile scope for unit tests * fix typo * parallelize across containers * why do the unit tests fail at night. are they scared of the dark? * disable buffering pytest * try to figure out where mock wallet test hangs in circleci * use random split rather than test durations * try larger docker container * try large docker container * try more procs on smaller executor * try using circleCI splitting * fix regex for circleci * remove pytest-split * don't run unit tests in parallel * revert to working config * dont use folder to store durations yet * don't use global subtensor in unit test * see if splitting by duration still causes suffering * use .test_durations * fix line break * 4 procs for integration tests * try different splitting algorithm * see if we can get away with small executor * tweak parallelism * tweak parallelism * remove ci build for github actions Co-authored-by: Joey Legere <joey@opentensor.ai> Co-authored-by: Eugene-hu <85906264+Eugene-hu@users.noreply.github.com>
* chain weight fix * uids must match weights Co-authored-by: Ala Shaabana <shaabana@gmail.com> Co-authored-by: joeylegere <joeylegere@gmail.com>
* check for replaced hotkey and reset the moving average score * Update __init__.py remove print statement Co-authored-by: Eugene-hu <85906264+Eugene-hu@users.noreply.github.com> Co-authored-by: Ala Shaabana <shaabana@gmail.com>
Co-authored-by: Joey Legere <joey@opentensor.ai>
* catch error when there are no wallets found display nicely * add "default" default name to regen_coldkey * added tests * added test for overview catch * added catch for overview * added mocking for wallet IO. TODO list with wallet * IO is mocked * changed wallet path to be unique to test Co-authored-by: joeylegere <joeylegere@gmail.com>
* added functions to validate ss58 and ed25519 str * test subtensor transfer invalid dest * added check for invalid dest on transfer * extract method to utils * change cli to use new validation method * added test for passing as bytes * can pass bytes to transfer for pubkey * extracted ss58_format to spec file Co-authored-by: joeylegere <joeylegere@gmail.com>
* Fix for dockerfile Dockerfile does not install python package (bitcli does not work), this should fix it. * Modify pip install Remove the -e flag, since its redundant. * import error fix (#767) Co-authored-by: Ala Shaabana <shaabana@gmail.com> Co-authored-by: Eugene-hu <85906264+Eugene-hu@users.noreply.github.com> Co-authored-by: joeylegere <joeylegere@gmail.com>
* merge master * Update __init__.py * shift solution bits to keep nonce range * remove unused import Co-authored-by: Thebes <jake@bittensor.com> Co-authored-by: Ala Shaabana <shaabana@gmail.com> Co-authored-by: camfairchild <cameron@fairchild.dev> Co-authored-by: joeylegere <joeylegere@gmail.com>
* standardize Python versions and fix CUDA repos * add PYTHONPATH to Bittensor library * change network to nobunaga Co-authored-by: joeylegere <joeylegere@gmail.com> Co-authored-by: Ala Shaabana <shaabana@gmail.com>
* fixed mp speed up * fix test * test fix * test fix * num_processes fix * undo use only half of the processes * print time * version checking * update * update * fixes Co-authored-by: isabella <isabella618033@gmail.com> Co-authored-by: isabella618033 <49876827+isabella618033@users.noreply.github.com> Co-authored-by: Unconst <32490803+unconst@users.noreply.github.com> Co-authored-by: Ala Shaabana <shaabana@gmail.com> Co-authored-by: joeylegere <joeylegere@gmail.com>
* added tests for overview using cache * added config argument for not using cache * added cached overview by default * add --wallet.hotkeys [] arg * add test to catch no flag specified * add sort_by arg for overview * add check_config for typing of new args * make default empty str * add wallet.sort_order arg * Only catch around retrieve_cached_neurons * remove fuzzy match sort_order * add better help message * make check more readable * add optional width arg for overview * add tests for width config option
* import error fix (#767) * test finetuning on last layer * added function to find the last layer * fix * log * fix * fix * fix * revert children * refine user config * Update run.py * change base * Revert "change base" This reverts commit 7156751. * change base * change base * change base * added test * added type hint * clean up * torch.nn.Module Co-authored-by: Eugene-hu <85906264+Eugene-hu@users.noreply.github.com> Co-authored-by: root <isabella618033> Co-authored-by: joeylegere <joeylegere@gmail.com>
* added tests for overview using cache * added config argument for not using cache * added cached overview by default * add --wallet.hotkeys [] arg * add test to catch no flag specified * add sort_by arg for overview * add check_config for typing of new args * make default empty str * add wallet.sort_order arg * Only catch around retrieve_cached_neurons * remove fuzzy match sort_order * add better help message * make check more readable * add optional width arg for overview * add tests for width config option * hotfix of numeric sorting * change fuzzy match to get max matching column
* bump pytorch version * bump msgpack version * run tests across multiple versions of python * setup.py reflects supported python versions * treat all tests as flaky * remove comment * try just making integration tests flaky * try different library to rerun tests * does where i put the arg matter??? Co-authored-by: Joey Legere <joey@opentensor.ai>
* google api reqs incompatible with our protobuf req * fix requirements * bump protobuf >=3.15.0 Co-authored-by: joeylegere <joeylegere@gmail.com>
* add coverage * change coverage output path * use node for coveralls image * try different python library for coveralls * fix syntax * fix typo * more reruns = fewer goats * ignore ./env * run against multiple python versions * coveralls uses .coveragerc * fix typo Co-authored-by: Joey Legere <joey@opentensor.ai>
Eugene-hu
approved these changes
Jun 13, 2022
shibshib
approved these changes
Jun 13, 2022
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
No description provided.