Dev (clean edition) #1095
Merged
Dev (clean edition) #1095
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
* fix: scikit-image version downgrade * show reqs * show reqs * upd reqs * fix: remove kornia in tests with minimal reqs * update imageio version * check hydra error * fix: update torch for hydra tests * Revert "update imageio version" This reverts commit d3320fb. * update imageio version
* SubRegistry PoC * fix defaultdict * upd registries * update __init__.py * fix __getitem__ * replace registries with registry * fix registry.get_from_params * fix registry.get_from_params * fix registry.get_from_params * replace REGISTRY with OPTIMIZERS * fix registry * fix * upd registry * upd registry * upd registry * upd registry * upd registry * upd registry * upd registry * upd registry * upd registry * use single registry * fix configs * fix kornia * fix kornia * fix kornia * upd kornia callback * ignore kornia segmentation test * fix wizard * fix: replace "name" keyword with "_target_", fix pytest tests * fix: typos in configs * mv Registry into tools * fix codestyle * update docs
* backward fix * codestyle
* backward fix * extra abs * codestyle
* vision * without pr * cuda available info * simple engine usage * codestyle * ignore flake in engine * enine changes * test prototype * dummy model * loss * loss as main metric * fixed loss * cuda devices tests * codestyle * init & destroy process group * reduce functions and tests * using engine * minization part * codestyle * config experiments + engine * ignore flake8 in engine tests * codestyle fix * engine: model initialization fix * devices * renamed function for understanding engine * dp * ddp updates * config + ddp prototype * naming and merging * kitty - init commit * kitty - cleanup * kitty - demo * kitty - demo 2 * metrics update * 99 line check * Metric update * Arch – IEngine (#1001) * fixed tests & ddp (without loader sampling) for python & config * renamed variable * docs & noqa * docs * removed unused import * sync metric fn * fixed imports & test for functional Co-authored-by: Dmytro Doroshenko <dimdoroshenko@gmail.com> * test work * device work * device work - 2 * dp2 * ddp trial * ddp trial todo fix * a few seed fixes * callbacks arch * scheduler callback * warn -> warning * less steps to check * import parts from device tests * extractid nn.Module from wrapper * skip config tests * run in ddp setup * world size checker * lr finer * merged * callbacks list * demo update * DDP works * callbacks cleanup * checkpointer works * checkpointer fix * tools * metircs and demo update * metircs - confusion matrix * tempdir instead of fixed name of logdir * metrics and confusion matrix * ddp fix * initial * engine extended with autocast context * accuracy metric update + hold for old callback-metrics * metrics arch * settings update * metrics test arch update * Python API - experiment * autocast fixes * supervised runner * Classification metrics (#1057) * Start classification metrics * add to __all__ * More output, more average * np * Move, rename * Precision/recall/f1/support v2 * After merge fixes * Kittylyst (#1056) * add DynamicBalanceClassSampler * add DynamicBalanceClassSampler: add usage example * add DynamicBalanceClassSampler: add tests * Update catalyst/data/tests/test_sampler.py * Update catalyst/data/tests/test_sampler.py * add DynamicBalanceClassSampler: debag tests * update sampler: add mode * add example notebook * sampler: fixes * samler: docs * DynamicBalanceClassSampler: fixes * change import order * change import order * iou metric * fix trevsky loss * add IOU, Dice, Trevsky metrics Co-authored-by: Sergey Kolesnikov <scitator@gmail.com> * classification metics merged * segmentation metics merged * update * update * wrapped `handle_batch` in autocast; renamed test files; apex engine * tests for model finetuning * removed some imports in tests * [WIP] Classification metrics (#1058) * Start classification metrics * add to __all__ * More output, more average * np * Move, rename * Precision/recall/f1/support v2 * After merge fixes * Classification metrics -> batch * Add docs to some tests * metric + callbacks; legacy removed * new examples * exp-runner-engine sync update * naming * comments * upd config experiment (#1055) * upd config experiment * upd callbacks imports * Classification metrics (#1063) * Start classification metrics * add to __all__ * More output, more average * np * Move, rename * Precision/recall/f1/support v2 * After merge fixes * Classification metrics -> batch * Add docs to some tests * Accuracy start * Add some docs * Some style fixes * Add tests for multiclass accuracy * Add warning to tests * Add std for accuracy@k to tests * exp-run merge (#1061) * exp-run merge * exp-run merge 2 * exp-run merge 3 - cmd * exp-run merge 3 - working example * exp-run merge 3 - working example 2 * exp-run merge 3 - working example 3 * exp-run merge 3 - working example 4 * exp-run merge 3 - working example 5 * exp-run merge 3 - working example 6 * [WIP] Add MRR metric (#1059) * Add MRR metric * Add MRR metric * add hitrate, ndcg * edit init * edit typo * add map * codestyle * Kittylyst (#1065) * add DynamicBalanceClassSampler * add DynamicBalanceClassSampler: add usage example * add DynamicBalanceClassSampler: add tests * Update catalyst/data/tests/test_sampler.py * Update catalyst/data/tests/test_sampler.py * add DynamicBalanceClassSampler: debag tests * update sampler: add mode * add example notebook * sampler: fixes * samler: docs * DynamicBalanceClassSampler: fixes * change import order * change import order * iou metric * fix trevsky loss * add IOU, Dice, Trevsky metrics * add segmentation losses update metrics * fix pep8 * fix pep8 * fix pep8 * add tests * Delete misc.py Co-authored-by: Sergey Kolesnikov <scitator@gmail.com> * merged & cleanup * RL: dqn and ddpg * RL: reinforce * simplified Co-authored-by: Dmytro Doroshenko <dimdoroshenko@gmail.com> Co-authored-by: Julia Shenshina <yuliya.shenshina@phystech.edu> Co-authored-by: Dokholyan <37542142+Dokholyan@users.noreply.github.com> Co-authored-by: Yauheni Kachan <19803638+bagxi@users.noreply.github.com> Co-authored-by: Dany Chepenko <raveforlive@gmail.com>
* add DynamicBalanceClassSampler * add DynamicBalanceClassSampler: add usage example * add DynamicBalanceClassSampler: add tests * Update catalyst/data/tests/test_sampler.py * Update catalyst/data/tests/test_sampler.py * add DynamicBalanceClassSampler: debag tests * update sampler: add mode * add example notebook * sampler: fixes * samler: docs * DynamicBalanceClassSampler: fixes * change import order * change import order * vision * without pr * cuda available info * simple engine usage * codestyle * ignore flake in engine * enine changes * test prototype * dummy model * loss * loss as main metric * fixed loss * cuda devices tests * codestyle * init & destroy process group * reduce functions and tests * using engine * minization part * codestyle * config experiments + engine * ignore flake8 in engine tests * codestyle fix * engine: model initialization fix * devices * renamed function for understanding engine * dp * ddp updates * config + ddp prototype * naming and merging * kitty - init commit * kitty - cleanup * kitty - demo * kitty - demo 2 * metrics update * 99 line check * Metric update * Arch – IEngine (#1001) * fixed tests & ddp (without loader sampling) for python & config * renamed variable * docs & noqa * docs * removed unused import * sync metric fn * fixed imports & test for functional Co-authored-by: Dmytro Doroshenko <dimdoroshenko@gmail.com> * test work * device work * device work - 2 * dp2 * ddp trial * ddp trial todo fix * a few seed fixes * callbacks arch * scheduler callback * warn -> warning * less steps to check * import parts from device tests * extractid nn.Module from wrapper * skip config tests * run in ddp setup * world size checker * lr finer * merged * callbacks list * demo update * DDP works * callbacks cleanup * checkpointer works * checkpointer fix * tools * metircs and demo update * metircs - confusion matrix * tempdir instead of fixed name of logdir * metrics and confusion matrix * ddp fix * iou metric * initial * engine extended with autocast context * accuracy metric update + hold for old callback-metrics * metrics arch * settings update * metrics test arch update * Python API - experiment * fix trevsky loss * add IOU, Dice, Trevsky metrics * autocast fixes * supervised runner * Classification metrics (#1057) * Start classification metrics * add to __all__ * More output, more average * np * Move, rename * Precision/recall/f1/support v2 * After merge fixes * Kittylyst (#1056) * add DynamicBalanceClassSampler * add DynamicBalanceClassSampler: add usage example * add DynamicBalanceClassSampler: add tests * Update catalyst/data/tests/test_sampler.py * Update catalyst/data/tests/test_sampler.py * add DynamicBalanceClassSampler: debag tests * update sampler: add mode * add example notebook * sampler: fixes * samler: docs * DynamicBalanceClassSampler: fixes * change import order * change import order * iou metric * fix trevsky loss * add IOU, Dice, Trevsky metrics Co-authored-by: Sergey Kolesnikov <scitator@gmail.com> * classification metics merged * segmentation metics merged * update * update * wrapped `handle_batch` in autocast; renamed test files; apex engine * tests for model finetuning * removed some imports in tests * [WIP] Classification metrics (#1058) * Start classification metrics * add to __all__ * More output, more average * np * Move, rename * Precision/recall/f1/support v2 * After merge fixes * Classification metrics -> batch * Add docs to some tests * metric + callbacks; legacy removed * new examples * exp-runner-engine sync update * naming * comments * upd config experiment (#1055) * upd config experiment * upd callbacks imports * Classification metrics (#1063) * Start classification metrics * add to __all__ * More output, more average * np * Move, rename * Precision/recall/f1/support v2 * After merge fixes * Classification metrics -> batch * Add docs to some tests * Accuracy start * Add some docs * Some style fixes * Add tests for multiclass accuracy * Add warning to tests * Add std for accuracy@k to tests * exp-run merge (#1061) * exp-run merge * exp-run merge 2 * exp-run merge 3 - cmd * exp-run merge 3 - working example * exp-run merge 3 - working example 2 * exp-run merge 3 - working example 3 * exp-run merge 3 - working example 4 * exp-run merge 3 - working example 5 * exp-run merge 3 - working example 6 * [WIP] Add MRR metric (#1059) * Add MRR metric * Add MRR metric * add hitrate, ndcg * edit init * edit typo * add map * codestyle * add segmentation losses update metrics * fix pep8 * fix pep8 * fix pep8 * add tests * Kittylyst (#1065) * add DynamicBalanceClassSampler * add DynamicBalanceClassSampler: add usage example * add DynamicBalanceClassSampler: add tests * Update catalyst/data/tests/test_sampler.py * Update catalyst/data/tests/test_sampler.py * add DynamicBalanceClassSampler: debag tests * update sampler: add mode * add example notebook * sampler: fixes * samler: docs * DynamicBalanceClassSampler: fixes * change import order * change import order * iou metric * fix trevsky loss * add IOU, Dice, Trevsky metrics * add segmentation losses update metrics * fix pep8 * fix pep8 * fix pep8 * add tests * Delete misc.py Co-authored-by: Sergey Kolesnikov <scitator@gmail.com> * merged & cleanup * RL: dqn and ddpg * RL: reinforce * simplified * rename tests * uodate tests * delete misc Co-authored-by: Sergey Kolesnikov <scitator@gmail.com> Co-authored-by: Dmytro Doroshenko <dimdoroshenko@gmail.com> Co-authored-by: Julia Shenshina <yuliya.shenshina@phystech.edu> Co-authored-by: Yauheni Kachan <19803638+bagxi@users.noreply.github.com> Co-authored-by: Dany Chepenko <raveforlive@gmail.com>
* tests 2.0 * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+
* tests 2.0 * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+
* tests 2.0 * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+
* ddp fix; apex ddp * renamed tests * blacked engines
* tests 2.0 * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+
* tests 2.0 * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+
* tests 2.0 * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+
* tests 2.0 * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * one line change * many line change * many line change * many line change
* vision * without pr * cuda available info * simple engine usage * codestyle * ignore flake in engine * enine changes * test prototype * dummy model * loss * loss as main metric * fixed loss * cuda devices tests * codestyle * init & destroy process group * reduce functions and tests * using engine * minization part * codestyle * config experiments + engine * ignore flake8 in engine tests * codestyle fix * engine: model initialization fix * devices * renamed function for understanding engine * dp * ddp updates * config + ddp prototype * naming and merging * kitty - init commit * kitty - cleanup * kitty - demo * kitty - demo 2 * metrics update * 99 line check * Metric update * Arch – IEngine (#1001) * fixed tests & ddp (without loader sampling) for python & config * renamed variable * docs & noqa * docs * removed unused import * sync metric fn * fixed imports & test for functional Co-authored-by: Dmytro Doroshenko <dimdoroshenko@gmail.com> * test work * device work * device work - 2 * dp2 * ddp trial * ddp trial todo fix * a few seed fixes * callbacks arch * scheduler callback * warn -> warning * less steps to check * import parts from device tests * extractid nn.Module from wrapper * skip config tests * run in ddp setup * world size checker * lr finer * merged * callbacks list * demo update * DDP works * callbacks cleanup * checkpointer works * checkpointer fix * tools * metircs and demo update * metircs - confusion matrix * tempdir instead of fixed name of logdir * metrics and confusion matrix * ddp fix * initial * engine extended with autocast context * accuracy metric update + hold for old callback-metrics * metrics arch * settings update * metrics test arch update * Python API - experiment * autocast fixes * supervised runner * Classification metrics (#1057) * Start classification metrics * add to __all__ * More output, more average * np * Move, rename * Precision/recall/f1/support v2 * After merge fixes * Kittylyst (#1056) * add DynamicBalanceClassSampler * add DynamicBalanceClassSampler: add usage example * add DynamicBalanceClassSampler: add tests * Update catalyst/data/tests/test_sampler.py * Update catalyst/data/tests/test_sampler.py * add DynamicBalanceClassSampler: debag tests * update sampler: add mode * add example notebook * sampler: fixes * samler: docs * DynamicBalanceClassSampler: fixes * change import order * change import order * iou metric * fix trevsky loss * add IOU, Dice, Trevsky metrics Co-authored-by: Sergey Kolesnikov <scitator@gmail.com> * classification metics merged * segmentation metics merged * update * update * bump quantization * add example * add onnx to settings * onnx import * remove arg from pruning * add model_loading.py * improve example * wrapped `handle_batch` in autocast; renamed test files; apex engine * tests for model finetuning * removed some imports in tests * [WIP] Classification metrics (#1058) * Start classification metrics * add to __all__ * More output, more average * np * Move, rename * Precision/recall/f1/support v2 * After merge fixes * Classification metrics -> batch * Add docs to some tests * metric + callbacks; legacy removed * new examples * improve example and fixes * Update catalyst/utils/onnx.py Co-authored-by: Sergey Kolesnikov <scitator@gmail.com> * fix path in load_model * exp-runner-engine sync update * fix imports * fix naming * naming * comments * tmp fix * upd config experiment (#1055) * upd config experiment * upd callbacks imports * Classification metrics (#1063) * Start classification metrics * add to __all__ * More output, more average * np * Move, rename * Precision/recall/f1/support v2 * After merge fixes * Classification metrics -> batch * Add docs to some tests * Accuracy start * Add some docs * Some style fixes * Add tests for multiclass accuracy * Add warning to tests * Add std for accuracy@k to tests * exp-run merge (#1061) * exp-run merge * exp-run merge 2 * exp-run merge 3 - cmd * exp-run merge 3 - working example * exp-run merge 3 - working example 2 * exp-run merge 3 - working example 3 * exp-run merge 3 - working example 4 * exp-run merge 3 - working example 5 * exp-run merge 3 - working example 6 * [WIP] Add MRR metric (#1059) * Add MRR metric * Add MRR metric * add hitrate, ndcg * edit init * edit typo * add map * codestyle * Kittylyst (#1065) * add DynamicBalanceClassSampler * add DynamicBalanceClassSampler: add usage example * add DynamicBalanceClassSampler: add tests * Update catalyst/data/tests/test_sampler.py * Update catalyst/data/tests/test_sampler.py * add DynamicBalanceClassSampler: debag tests * update sampler: add mode * add example notebook * sampler: fixes * samler: docs * DynamicBalanceClassSampler: fixes * change import order * change import order * iou metric * fix trevsky loss * add IOU, Dice, Trevsky metrics * add segmentation losses update metrics * fix pep8 * fix pep8 * fix pep8 * add tests * Delete misc.py Co-authored-by: Sergey Kolesnikov <scitator@gmail.com> * merged & cleanup * RL: dqn and ddpg * RL: reinforce * simplified * add method param to onnx * fix arg name * model loading * model loading refactoring * experiment fix * amp availability fix * add docstrings * fix merges * remove some functions from tracing * fix f-str * quantization fix * add test and fix * fix pruning * docstrings * fix quantization * format * fix quotes * codestyle * move forwarf override method to tools * codestyle Co-authored-by: Sergey Kolesnikov <scitator@gmail.com> Co-authored-by: Dmytro Doroshenko <dimdoroshenko@gmail.com> Co-authored-by: Julia Shenshina <yuliya.shenshina@phystech.edu> Co-authored-by: Dokholyan <37542142+Dokholyan@users.noreply.github.com> Co-authored-by: Yauheni Kachan <19803638+bagxi@users.noreply.github.com> Co-authored-by: Dany Chepenko <raveforlive@gmail.com>
* tests 2.0 * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * one line change * many line change * many line change * many line change * codestyle * many line change * docs
* tests 2.0 * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * one line change * many line change * many line change * many line change * codestyle * many line change * docs * docs
* tests 2.0 * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * one line change * many line change * many line change * many line change * codestyle * many line change * docs * docs * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle + * codestyle + * codestyle + * codestyle + * codestyle + * codestyle + * version update * version update * version update * version update
* tests 2.0 * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * one line change * many line change * many line change * many line change * codestyle * many line change * docs * docs * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle + * codestyle + * codestyle + * codestyle + * codestyle + * codestyle + * version update * version update * version update * version update * callback
* tests 2.0 * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 2.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 3.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * tests 4.0+ * one line change * many line change * many line change * many line change * codestyle * many line change * docs * docs * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle * codestyle + * codestyle + * codestyle + * codestyle + * codestyle + * codestyle + * version update * version update * version update * version update * callback * codestyle + ddp * settings * runner fix * test fix * settings update * settings update * settings update * settings update * settings update * settings update * settings update * settings update * settings update * settings update
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
* settings update * settings update * settings update * settings update
* add DynamicBalanceClassSampler * add DynamicBalanceClassSampler: add usage example * add DynamicBalanceClassSampler: add tests * Update catalyst/data/tests/test_sampler.py * Update catalyst/data/tests/test_sampler.py * add DynamicBalanceClassSampler: debag tests * update sampler: add mode * add example notebook * sampler: fixes * samler: docs * DynamicBalanceClassSampler: fixes * change import order * change import order * add custom mode for MetricAggregationCallback * use mode as custom_function * fix codestyle * add aggregation * add MetricAggregationCallback * add full example * codestyle * add aggregation tests * codestyle * Update catalyst/callbacks/aggregation.py * Update docs/api/callbacks.rst * codestyle * fix example * codestyle(I hate him) * docs fix * torch==1.3 support fixes Co-authored-by: Sergey Kolesnikov <scitator@gmail.com>
* tracing docs and refactoring * pruning docs * fix codestyle * add test for onnx * fix codestyle * fix * remove files * fix * fix * fix docs * fix docs * docs and onnx * fix * fix * fix docs * fix * fix init file * req fix
* requirements update * requirements update * requirements update * utils update * utils update
* config tests for device engine * tests for dp engine * new engine property - `is_master_process` * amp & apex from string; tests for configs * tests * fixed tests * reformatted * load engine from config & reformatted * codestyle * fixed usage of settings properties * using settings * fixed config tests; model packing & unpacking * codestyle * fixed apex * apex import fix * codestyle * codestyle fixes * return in engine docs * removed unused import * imports instead of using `dl` * again codestyle * DDP: delete loggers for non master processes * fixed tests, imports and removed redundant typings * apex: link to an issue with O1 * removed apex & ddp "replacements" * config runner: criterion from params * config runner: scheduler from params * codestyle * renamed class * removed process_engine tests * test metrics sync: initial * disabled inputs & targets sync * device: args and kwargs for sync_tensor * switch to sgd and cross entropy * extra checks * fix * fix * fix Co-authored-by: Dmytro Doroshenko <dimdoroshenko@gmail.com>
* docs for map * docs for map * docs for mrr * docs for ndcg * docs for hitrate * pep8 Co-authored-by: Даниил <zkid18@MacBook-Pro.Dlink>
* config tests for device engine * tests for dp engine * new engine property - `is_master_process` * amp & apex from string; tests for configs * tests * fixed tests * reformatted * load engine from config & reformatted * codestyle * fixed usage of settings properties * using settings * fixed config tests; model packing & unpacking * codestyle * fixed apex * apex import fix * codestyle * codestyle fixes * return in engine docs * removed unused import * imports instead of using `dl` * again codestyle * DDP: delete loggers for non master processes * fixed tests, imports and removed redundant typings * apex: link to an issue with O1 * removed apex & ddp "replacements" * config runner: criterion from params * config runner: scheduler from params * codestyle * renamed class * removed process_engine tests * test metrics sync: initial * disabled inputs & targets sync * device: args and kwargs for sync_tensor * switch to sgd and cross entropy * ddp * ddp * ddp * ddp * wrap model in data parallel * model type checks * ddp * ddp * ddp * ddp * ddp Co-authored-by: Dmytro Doroshenko <dimdoroshenko@gmail.com>
* cmc checked * segmentation checked * fix * codestyle * classification * recsys * classification ddp * ddp * codestyle * codestyle * codestyle
* add DynamicBalanceClassSampler * add DynamicBalanceClassSampler: add usage example * add DynamicBalanceClassSampler: add tests * Update catalyst/data/tests/test_sampler.py * Update catalyst/data/tests/test_sampler.py * add DynamicBalanceClassSampler: debag tests * update sampler: add mode * add example notebook * sampler: fixes * samler: docs * DynamicBalanceClassSampler: fixes * change import order * change import order * add custom mode for MetricAggregationCallback * use mode as custom_function * fix codestyle * add aggregation * add MetricAggregationCallback * add full example * codestyle * add aggregation tests * codestyle * Update catalyst/callbacks/aggregation.py * Update docs/api/callbacks.rst * codestyle * fix example * codestyle(I hate him) * docs fix * torch==1.3 support fixes * Add BatchFuntionalMetric for custom metrics * Add tests for functional_metric: mae, mse, dice * Fix BatchFuntionalMetric by using AdditiveValueMetric * fix codestyle * rename file _funtional_metric * Add CustomMetricCallback * fix name * Add functional_metric in Docs * remove aggregation callback * Update docs/api/metrics.rst * Update catalyst/callbacks/__init__.py * Update docs/api/metrics.rst * rename file Co-authored-by: Sergey Kolesnikov <scitator@gmail.com>
* AMP DP * APEX DP * renamed functions * codestyle & fixed blobs * fixed imports * renamed functions * removed redundant variable * codestyle * dp in `__all__` * fixed import in test
* add DynamicBalanceClassSampler * add DynamicBalanceClassSampler: add usage example * add DynamicBalanceClassSampler: add tests * Update catalyst/data/tests/test_sampler.py * Update catalyst/data/tests/test_sampler.py * add DynamicBalanceClassSampler: debag tests * update sampler: add mode * add example notebook * sampler: fixes * samler: docs * DynamicBalanceClassSampler: fixes * change import order * change import order * add custom mode for MetricAggregationCallback * use mode as custom_function * fix codestyle * add aggregation * add MetricAggregationCallback * add full example * codestyle * add aggregation tests * codestyle * Update catalyst/callbacks/aggregation.py * Update docs/api/callbacks.rst * codestyle * fix example * codestyle(I hate him) * docs fix * torch==1.3 support fixes * add micro and macro mode in update_key_value * fix segmentation metrics * Fix codestyle * Fix confising macro and micro * remove aggregation.py * prefix and suffixfixes * Fix tests * prefix and suffix fixes * Codestyle fixes * Fix metric name on the end of the loader * fix docstrings Co-authored-by: Sergey Kolesnikov <scitator@gmail.com>
* dpp fix thx ditwoo * dpp fix thx ditwoo * segmentaiton check
* utils update * utils update
* callbacks * dosc * dosc * dosc
* tests for loading state on stage start * raise exception on missed files * codestyle & naming * sorted imports * noqa * codestyle * codestyle & mkl option * check for first stage * raises section * codestyle
* tests for loading state on stage start * raise exception on missed files * codestyle & naming * sorted imports * noqa * codestyle * codestyle & mkl option * check for first stage * raises section * codestyle
* tests for loading state on stage start * raise exception on missed files * codestyle & naming * sorted imports * noqa * codestyle * codestyle & mkl option * check for first stage * raises section * codestyle
* runner args * small fix * codestyle * apex flag * sorted imports * codestyle * absolute engine imports * get_engine moved to engines * removed get_engine from all * test marks & codestyle * ddp: save checkpoint from master process * utils * fix * fix * fix Co-authored-by: Sergey Kolesnikov <scitator@gmail.com>
* Added MLflow logger. * codestyle + docstrings + removed unnecessary functions * removed unnecessary parts [2] * added mlflow to settings and setup * bugfixes and removed unnecessary lines * added requirements * removed unnecessary imports * fixed docstring * fixes for github ci * renamed requirements * added mlflow to settings, setup and some small fixes * bugfix * fixed scope and run name * codestyle * moved everything from functional to mlflow * added docs * fixed name + added link to docstring * added stage_key to log_hparams, so in case of nested runs in mlflow, the parent run will be named as experiment_key, and chil runs will be named as stage_key * added test * some fixes after testing * fixed image type and shape for logging * comments taken into account * codestyle * moved 'image_to_tensor' to loggers.functional
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.
Before submitting (checklist)
catalyst-make-codestyle && catalyst-check-codestyle(pip install -U catalyst-codestyle).make check-docs?pytest .?Description
Related Issue
Type of Change
PR review
Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in Github issues there's a high chance it will not be merged.
PS