{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":620583116,"defaultBranch":"master","name":"ray","ownerLogin":"ollie-iterators","currentUserCanPush":false,"isFork":true,"isEmpty":false,"createdAt":"2023-03-29T01:07:28.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/46904826?v=4","public":true,"private":false,"isOrgOwned":false},"refInfo":{"name":"","listCacheKey":"v0:1690238619.0","currentOid":""},"activityList":{"items":[{"before":"d34e81885fdd10d04027fb1c4d5c98c8285d0a3b","after":"05232b531de447e8d596c6489309cb5a4251a28e","ref":"refs/heads/master","pushedAt":"2024-04-18T21:06:05.000Z","pushType":"push","commitsCount":377,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[Tune] Remove spammy log for \"new output engine\" (#44824)\n\nRemove spammy log output about the `RAY_AIR_NEW_OUTPUT` feature flag environment variable.\r\n\r\nSigned-off-by: Hongpeng Guo <hpguo@anyscale.com>","shortMessageHtmlLink":"[Tune] Remove spammy log for \"new output engine\" (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2249506716\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/44824\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/44824/hovercard\" href=\"https://github.com/ray-project/ray/pull/44824\">ray-project#44824</a>)"}},{"before":"671cbabc553c4b4d547d4777014756331d9835ef","after":"d34e81885fdd10d04027fb1c4d5c98c8285d0a3b","ref":"refs/heads/master","pushedAt":"2024-03-11T20:39:22.000Z","pushType":"push","commitsCount":361,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[ci] upgrade dataplane image (#43877)\n\nUpgrade dataplane image\r\n\r\nSigned-off-by: can <can@anyscale.com>","shortMessageHtmlLink":"[ci] upgrade dataplane image (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2179863420\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/43877\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/43877/hovercard\" href=\"https://github.com/ray-project/ray/pull/43877\">ray-project#43877</a>)"}},{"before":"fae8d2ff814377eb027d63d73a23d5c5bf3b02bd","after":"671cbabc553c4b4d547d4777014756331d9835ef","ref":"refs/heads/master","pushedAt":"2024-02-08T17:54:01.000Z","pushType":"push","commitsCount":139,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"update graph serve dashboard error graph to sum by error code (#42883)\n\n---------\r\n\r\nSigned-off-by: Alan Guo <aguo@anyscale.com>","shortMessageHtmlLink":"update graph serve dashboard error graph to sum by error code (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2110868293\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/42883\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/42883/hovercard\" href=\"https://github.com/ray-project/ray/pull/42883\">ray-pr…</a>"}},{"before":"8bc7660373d99843c85a998e4c1f7b9f42464e61","after":"fae8d2ff814377eb027d63d73a23d5c5bf3b02bd","ref":"refs/heads/master","pushedAt":"2024-01-25T19:56:18.000Z","pushType":"push","commitsCount":601,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"Revert \"[ci] temp skip linkcheck on premerge, which is failing\" (#42709)\n\nReverts #42707\r\n\r\nissue already fixed on master.","shortMessageHtmlLink":"Revert \"[ci] temp skip linkcheck on premerge, which is failing\" (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2101007564\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/42709\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/42709/hovercard\" href=\"https://github.com/ray-project/ray/pull/42709\">ray-…</a>"}},{"before":"c1282229473a8ab75a6f80d7e363e477c0290704","after":"8bc7660373d99843c85a998e4c1f7b9f42464e61","ref":"refs/heads/master","pushedAt":"2023-11-20T18:28:10.000Z","pushType":"push","commitsCount":678,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[ci] skip flaky test on premerge (#41274)\n\nSkip flaky tests on premerge, as signed off in https://docs.google.com/document/d/1IVgI7LNo21bPJO8kdYTwdWwEdZr5D9VtNrabUeX8gCc/edit#heading=h.isxl96agy1it. This will help save good amount of cost on PR, especially due to those gpu test runs.\r\n\r\nSigned-off-by: can <can@anyscale.com>","shortMessageHtmlLink":"[ci] skip flaky test on premerge (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2002438067\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/41274\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/41274/hovercard\" href=\"https://github.com/ray-project/ray/pull/41274\">ray-project#41274</a>)"}},{"before":"a7cccff522b108307c174dbad064092e6dfdf188","after":"c1282229473a8ab75a6f80d7e363e477c0290704","ref":"refs/heads/master","pushedAt":"2023-09-22T17:39:32.000Z","pushType":"push","commitsCount":17,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[Serve] Enable isort for ray serve lib (#39604)\n\n- Enable isort for future prs.\r\n- isort all serve files (one time)","shortMessageHtmlLink":"[Serve] Enable isort for ray serve lib (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1893312801\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/39604\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/39604/hovercard\" href=\"https://github.com/ray-project/ray/pull/39604\">ray-project#39604</a>)"}},{"before":"32c73a319d61864513b385a9d4d2e37bc7716ae2","after":"a7cccff522b108307c174dbad064092e6dfdf188","ref":"refs/heads/master","pushedAt":"2023-09-21T13:18:12.000Z","pushType":"push","commitsCount":49,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"Fix data docker file (#39780)\n\nARROW_VERSION is currently declared before the FROM keyword so it is not read in properly into the docker image. As a result, all data tests are currently running pyarrow=12 (which is part of requirement files).\r\n\r\nFix by moving the arg to below the FROM keyword\r\n\r\nSigned-off-by: can <can@anyscale.com>","shortMessageHtmlLink":"Fix data docker file (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1905851463\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/39780\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/39780/hovercard\" href=\"https://github.com/ray-project/ray/pull/39780\">ray-project#39780</a>)"}},{"before":"99bf18975d1e0cf21c908078e5b36926d959f6fd","after":"32c73a319d61864513b385a9d4d2e37bc7716ae2","ref":"refs/heads/master","pushedAt":"2023-09-12T13:08:35.000Z","pushType":"push","commitsCount":269,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[RLlib] Issue 39031: SlateQ example script bug. (#39550)","shortMessageHtmlLink":"[RLlib] Issue 39031: SlateQ example script bug. (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1890960362\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/39550\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/39550/hovercard\" href=\"https://github.com/ray-project/ray/pull/39550\">ray-project#39550</a>)"}},{"before":"921ec470177c73c0e605c1041ecc5e6780974fa2","after":"99bf18975d1e0cf21c908078e5b36926d959f6fd","ref":"refs/heads/master","pushedAt":"2023-08-25T22:08:20.000Z","pushType":"push","commitsCount":244,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[Docs][Serve] Add model serve using AWS NeuronCore (#38811)\n\n* [Docs] Add model serve using AWS NeuronCore\r\n\r\nSigned-off-by: maheedhar reddy chappidi <chappidm@amazon.com>","shortMessageHtmlLink":"[Docs][Serve] Add model serve using AWS NeuronCore (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1864156474\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/38811\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/38811/hovercard\" href=\"https://github.com/ray-project/ray/pull/38811\">ray-project#38811</a>)"}},{"before":"e39b18cec7215816960c76cf3115ca4e30d90b72","after":"921ec470177c73c0e605c1041ecc5e6780974fa2","ref":"refs/heads/master","pushedAt":"2023-08-15T19:08:23.000Z","pushType":"push","commitsCount":38,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[air] `pyarrow.fs` persistence (11/n): Support pausing trials (and certain schedulers) (#38355)\n\nThis PR gets pausing trials to work without in-memory checkpoints. This is needed for pausing scheduler support (PBT, BOHB).\r\n\r\nInstead of placing a `save_to_object` future in the trial data and immediately stopping the trial, this PR now waits to handle the `save` future before stopping the trial. Otherwise, the save future gets immediately cleared when `schedule_trial_stop` is called, causing no checkpoint to be saved --> possibly starting training from scratch.\r\n\r\nSigned-off-by: Justin Yu <justinvyu@anyscale.com>","shortMessageHtmlLink":"[air] <code>pyarrow.fs</code> persistence (11/n): Support pausing trials (and ce…"}},{"before":"3fcdc9284ed09edae2334398c46b2adcc33bb26e","after":"e39b18cec7215816960c76cf3115ca4e30d90b72","ref":"refs/heads/master","pushedAt":"2023-08-12T18:30:54.000Z","pushType":"push","commitsCount":167,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[air] `pyarrow.fs` persistence (10/n): Unify Tune and Train sessions to support new persistence path in `FunctionTrainable` (#38284)","shortMessageHtmlLink":"[air] <code>pyarrow.fs</code> persistence (10/n): Unify Tune and Train sessions …"}},{"before":"dcf499f768ae04d17b4d5e5afa616d644b446ad2","after":"3fcdc9284ed09edae2334398c46b2adcc33bb26e","ref":"refs/heads/master","pushedAt":"2023-08-02T22:53:58.511Z","pushType":"push","commitsCount":127,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"Revert \"[RLlib] DreamerV3: Minor fixes and enhancements. (#37977)\" (#38035)\n\nThis reverts commit 48488f50de39a9d7ef51094b5c5eefabee3ba08c.","shortMessageHtmlLink":"Revert \"[RLlib] DreamerV3: Minor fixes and enhancements. (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1831554197\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/37977\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/37977/hovercard\" href=\"https://github.com/ray-project/ray/pull/37977\">ray-project…</a>"}},{"before":"dcf499f768ae04d17b4d5e5afa616d644b446ad2","after":"3fcdc9284ed09edae2334398c46b2adcc33bb26e","ref":"refs/heads/master","pushedAt":"2023-08-02T22:53:58.000Z","pushType":"push","commitsCount":127,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"Revert \"[RLlib] DreamerV3: Minor fixes and enhancements. (#37977)\" (#38035)\n\nThis reverts commit 48488f50de39a9d7ef51094b5c5eefabee3ba08c.","shortMessageHtmlLink":"Revert \"[RLlib] DreamerV3: Minor fixes and enhancements. (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1831554197\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/37977\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/37977/hovercard\" href=\"https://github.com/ray-project/ray/pull/37977\">ray-project…</a>"}},{"before":null,"after":"a67b5110b0b4acd935463ed4f9adf8d720ae736c","ref":"refs/heads/revert-37504-air/examples/dreambooth_batch_fix","pushedAt":"2023-07-24T22:43:39.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"Revert \"[air][example] Dreambooth fine-tuning example fixes (#37504)\"\n\nThis reverts commit 6ac4521f2a666b1dff8912f455ac928cbe91c5f4.","shortMessageHtmlLink":"Revert \"[air][example] Dreambooth fine-tuning example fixes (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1809349082\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/37504\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/37504/hovercard\" href=\"https://github.com/ray-project/ray/pull/37504\">ray-proj…</a>"}},{"before":"a6f13e35b86b68084dd88ee5cd8c34ef0d7440b2","after":"dcf499f768ae04d17b4d5e5afa616d644b446ad2","ref":"refs/heads/master","pushedAt":"2023-07-24T13:46:25.940Z","pushType":"push","commitsCount":187,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[RLlib][doc-test] Improved doc examples inside core folder (#37558)\n\nSigned-off-by: Artur Niederfahrenhorst <attaismyname@googlemail.com>","shortMessageHtmlLink":"[RLlib][doc-test] Improved doc examples inside core folder (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1811022499\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/37558\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/37558/hovercard\" href=\"https://github.com/ray-project/ray/pull/37558\">ray-proje…</a>"}},{"before":"a6f13e35b86b68084dd88ee5cd8c34ef0d7440b2","after":"dcf499f768ae04d17b4d5e5afa616d644b446ad2","ref":"refs/heads/master","pushedAt":"2023-07-24T13:46:25.000Z","pushType":"push","commitsCount":187,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[RLlib][doc-test] Improved doc examples inside core folder (#37558)\n\nSigned-off-by: Artur Niederfahrenhorst <attaismyname@googlemail.com>","shortMessageHtmlLink":"[RLlib][doc-test] Improved doc examples inside core folder (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1811022499\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/37558\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/37558/hovercard\" href=\"https://github.com/ray-project/ray/pull/37558\">ray-proje…</a>"}},{"before":"c07ed7ecfd3345c0cfcd0c5ba6819f3e0c325fe4","after":"a6f13e35b86b68084dd88ee5cd8c34ef0d7440b2","ref":"refs/heads/master","pushedAt":"2023-07-07T01:31:54.518Z","pushType":"push","commitsCount":46,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[Docs] De-flake doctests (#37162)\n\nSigned-off-by: Balaji Veeramani <balaji@anyscale.com>\r\n\r\n:book: Doctest (CPU) fails 25% of runs due to a few flaky tests. This PR deflakes those tests.","shortMessageHtmlLink":"[Docs] De-flake doctests (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1792194365\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/37162\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/37162/hovercard\" href=\"https://github.com/ray-project/ray/pull/37162\">ray-project#37162</a>)"}},{"before":"c07ed7ecfd3345c0cfcd0c5ba6819f3e0c325fe4","after":"a6f13e35b86b68084dd88ee5cd8c34ef0d7440b2","ref":"refs/heads/master","pushedAt":"2023-07-07T01:31:54.000Z","pushType":"push","commitsCount":46,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[Docs] De-flake doctests (#37162)\n\nSigned-off-by: Balaji Veeramani <balaji@anyscale.com>\r\n\r\n:book: Doctest (CPU) fails 25% of runs due to a few flaky tests. This PR deflakes those tests.","shortMessageHtmlLink":"[Docs] De-flake doctests (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1792194365\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/37162\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/37162/hovercard\" href=\"https://github.com/ray-project/ray/pull/37162\">ray-project#37162</a>)"}},{"before":"0c13e5466e45b2ef7e278f53aba3ffa6771116bc","after":"c07ed7ecfd3345c0cfcd0c5ba6819f3e0c325fe4","ref":"refs/heads/master","pushedAt":"2023-07-01T13:21:24.316Z","pushType":"push","commitsCount":92,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"Pin pydantic to < 2.0.0 (#37000)\n\n2.0.0 brings breaking changes\r\n\r\nHere's a failed CI run: https://buildkite.com/ray-project/oss-ci-build-pr/builds/27338#01890d26-94ff-42f2-80be-c2ac0a86e8d3/447-762","shortMessageHtmlLink":"Pin pydantic to &lt; 2.0.0 (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1782994464\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/37000\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/37000/hovercard\" href=\"https://github.com/ray-project/ray/pull/37000\">ray-project#37000</a>)"}},{"before":"0c13e5466e45b2ef7e278f53aba3ffa6771116bc","after":"c07ed7ecfd3345c0cfcd0c5ba6819f3e0c325fe4","ref":"refs/heads/master","pushedAt":"2023-07-01T13:21:24.000Z","pushType":"push","commitsCount":92,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"Pin pydantic to < 2.0.0 (#37000)\n\n2.0.0 brings breaking changes\r\n\r\nHere's a failed CI run: https://buildkite.com/ray-project/oss-ci-build-pr/builds/27338#01890d26-94ff-42f2-80be-c2ac0a86e8d3/447-762","shortMessageHtmlLink":"Pin pydantic to &lt; 2.0.0 (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1782994464\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/37000\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/37000/hovercard\" href=\"https://github.com/ray-project/ray/pull/37000\">ray-project#37000</a>)"}},{"before":"2cf264c7b482c305fd36375f86745a8cb805ee80","after":"0c13e5466e45b2ef7e278f53aba3ffa6771116bc","ref":"refs/heads/master","pushedAt":"2023-06-28T00:27:06.219Z","pushType":"push","commitsCount":179,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[serve] Power of two random choices routing (#36501)\n\nAdds a feature flag for \"power of two random choices\" routing that will replace our existing naiive round-robin approach.\r\n\r\nThe main motivations for this are to:\r\n\r\n1. properly enforce `max_concurrent_queries` independent of the number of callers. Note that we can still exceed , but should only be by a constant factor rather than `O(number of callers)`.\r\n2. improve our routing to lower tail latency (particularly when there are variable-length processing times).\r\n3. enforce `max_concurrent_queries` w/ streaming responses (not possible w/ our current implementation).\r\n\r\nBenchmark results are [here](https://docs.google.com/spreadsheets/d/1Fha6fDAyB2yFxH7BSGaFvlTzwniJClH9L4APy6nt-6o/edit?usp=sharing), but TL;DR:\r\n\r\n- Average & tail latencies decreased and throughput increased over a number of conditions, increasing with variable processing times.\r\n- No-op latencies increased by ~0.5ms (small regression).\r\n- Throughput decreased for condition with a single caller and no-op requests (stressful microbenchmark).\r\n\r\nOverall we expect this solution to perform better in real-world scenarios.\r\n\r\nCopying from the docstring:\r\n\r\n>     \"\"\"Chooses a replica for each request using the \"power of two choices\" procedure.\r\n> \r\n>     Requests are scheduled in FIFO order.\r\n> \r\n>     When a request comes in, two replicas are chosen randomly as candidates. The queue\r\n>     length of each replicas is requested from it via a control message.\r\n> \r\n>     The replica responds with two items: (queue_length, accepted). Only replicas that\r\n>     accept the request are considered; between those, the one with the lower queue\r\n>     length is chosen.\r\n> \r\n>     In the case when neither replica accepts the request (e.g., their queues are full),\r\n>     the procedure is repeated with backoff. This backoff repeats indefinitely until a\r\n>     replica is chosen, so the caller should use timeouts and cancellation to avoid\r\n>     hangs.\r\n> \r\n>     Each request being scheduled may spawn an independent task that runs the scheduling\r\n>     procedure concurrently. This task will not necessarily satisfy the request that\r\n>     started it (in order to maintain the FIFO order). The total number of tasks is\r\n>     capped at (2 * num_replicas).\r\n>     \"\"\"\r\n\r\nJava and multiplexing support are not included in this PR and will be added in follow-ups.","shortMessageHtmlLink":"[serve] Power of two random choices routing (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1760902677\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/36501\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/36501/hovercard\" href=\"https://github.com/ray-project/ray/pull/36501\">ray-project#36501</a>)"}},{"before":"2cf264c7b482c305fd36375f86745a8cb805ee80","after":"0c13e5466e45b2ef7e278f53aba3ffa6771116bc","ref":"refs/heads/master","pushedAt":"2023-06-28T00:27:06.000Z","pushType":"push","commitsCount":179,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[serve] Power of two random choices routing (#36501)\n\nAdds a feature flag for \"power of two random choices\" routing that will replace our existing naiive round-robin approach.\r\n\r\nThe main motivations for this are to:\r\n\r\n1. properly enforce `max_concurrent_queries` independent of the number of callers. Note that we can still exceed , but should only be by a constant factor rather than `O(number of callers)`.\r\n2. improve our routing to lower tail latency (particularly when there are variable-length processing times).\r\n3. enforce `max_concurrent_queries` w/ streaming responses (not possible w/ our current implementation).\r\n\r\nBenchmark results are [here](https://docs.google.com/spreadsheets/d/1Fha6fDAyB2yFxH7BSGaFvlTzwniJClH9L4APy6nt-6o/edit?usp=sharing), but TL;DR:\r\n\r\n- Average & tail latencies decreased and throughput increased over a number of conditions, increasing with variable processing times.\r\n- No-op latencies increased by ~0.5ms (small regression).\r\n- Throughput decreased for condition with a single caller and no-op requests (stressful microbenchmark).\r\n\r\nOverall we expect this solution to perform better in real-world scenarios.\r\n\r\nCopying from the docstring:\r\n\r\n>     \"\"\"Chooses a replica for each request using the \"power of two choices\" procedure.\r\n> \r\n>     Requests are scheduled in FIFO order.\r\n> \r\n>     When a request comes in, two replicas are chosen randomly as candidates. The queue\r\n>     length of each replicas is requested from it via a control message.\r\n> \r\n>     The replica responds with two items: (queue_length, accepted). Only replicas that\r\n>     accept the request are considered; between those, the one with the lower queue\r\n>     length is chosen.\r\n> \r\n>     In the case when neither replica accepts the request (e.g., their queues are full),\r\n>     the procedure is repeated with backoff. This backoff repeats indefinitely until a\r\n>     replica is chosen, so the caller should use timeouts and cancellation to avoid\r\n>     hangs.\r\n> \r\n>     Each request being scheduled may spawn an independent task that runs the scheduling\r\n>     procedure concurrently. This task will not necessarily satisfy the request that\r\n>     started it (in order to maintain the FIFO order). The total number of tasks is\r\n>     capped at (2 * num_replicas).\r\n>     \"\"\"\r\n\r\nJava and multiplexing support are not included in this PR and will be added in follow-ups.","shortMessageHtmlLink":"[serve] Power of two random choices routing (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1760902677\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/36501\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/36501/hovercard\" href=\"https://github.com/ray-project/ray/pull/36501\">ray-project#36501</a>)"}},{"before":"5fa92ccaeae23d47022baf3181a1ddf96f54ee52","after":"2cf264c7b482c305fd36375f86745a8cb805ee80","ref":"refs/heads/master","pushedAt":"2023-06-15T17:46:34.190Z","pushType":"push","commitsCount":20,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[train] Re-upgrade mlflow to 2.x for python 3.8+ (#36446)\n\nMlflow was downgraded for python 3.8 in our recent migration. This PR updates mlflow to the latest release again for Python 3.8+ images.\r\n\r\nSigned-off-by: Kai Fricke <kai@anyscale.com>","shortMessageHtmlLink":"[train] Re-upgrade mlflow to 2.x for python 3.8+ (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1758232119\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/36446\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/36446/hovercard\" href=\"https://github.com/ray-project/ray/pull/36446\">ray-project#36446</a>)"}},{"before":"5fa92ccaeae23d47022baf3181a1ddf96f54ee52","after":"2cf264c7b482c305fd36375f86745a8cb805ee80","ref":"refs/heads/master","pushedAt":"2023-06-15T17:46:34.109Z","pushType":"push","commitsCount":20,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[train] Re-upgrade mlflow to 2.x for python 3.8+ (#36446)\n\nMlflow was downgraded for python 3.8 in our recent migration. This PR updates mlflow to the latest release again for Python 3.8+ images.\r\n\r\nSigned-off-by: Kai Fricke <kai@anyscale.com>","shortMessageHtmlLink":"[train] Re-upgrade mlflow to 2.x for python 3.8+ (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1758232119\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/36446\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/36446/hovercard\" href=\"https://github.com/ray-project/ray/pull/36446\">ray-project#36446</a>)"}},{"before":"c1c466a8642aba0df4ebc87ecf2ace5dde532fca","after":"5fa92ccaeae23d47022baf3181a1ddf96f54ee52","ref":"refs/heads/master","pushedAt":"2023-06-14T15:20:52.134Z","pushType":"push","commitsCount":48,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[docs] create advanced guides directory; rename files 1/N (#36297)\n\nImplement new directory structure per plan. This is one of many to come. I wanted to start with just the directory structure changes and get that merged asap to head off merge conflicts.\r\n\r\n@edoakes, @akshay-anyscale and I discussed removing the User Guides directory. With the restructuring, the number of non-Advanced guides was only 5-6 guides and we thought they would be more discoverable at a higher level. Take a look and let us know what you think, please.","shortMessageHtmlLink":"[docs] create advanced guides directory; rename files 1/N (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1750725024\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/36297\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/36297/hovercard\" href=\"https://github.com/ray-project/ray/pull/36297\">ray-projec…</a>"}},{"before":"c1c466a8642aba0df4ebc87ecf2ace5dde532fca","after":"5fa92ccaeae23d47022baf3181a1ddf96f54ee52","ref":"refs/heads/master","pushedAt":"2023-06-14T15:20:52.072Z","pushType":"push","commitsCount":48,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[docs] create advanced guides directory; rename files 1/N (#36297)\n\nImplement new directory structure per plan. This is one of many to come. I wanted to start with just the directory structure changes and get that merged asap to head off merge conflicts.\r\n\r\n@edoakes, @akshay-anyscale and I discussed removing the User Guides directory. With the restructuring, the number of non-Advanced guides was only 5-6 guides and we thought they would be more discoverable at a higher level. Take a look and let us know what you think, please.","shortMessageHtmlLink":"[docs] create advanced guides directory; rename files 1/N (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1750725024\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/36297\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/36297/hovercard\" href=\"https://github.com/ray-project/ray/pull/36297\">ray-projec…</a>"}},{"before":"e9a1c6d814fb1a81033809f56695030d651388f5","after":"c1c466a8642aba0df4ebc87ecf2ace5dde532fca","ref":"refs/heads/master","pushedAt":"2023-06-09T20:11:12.092Z","pushType":"push","commitsCount":58,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[Data] Skip writing a file for an empty block in `write_datasource` (#36134)\n\n## Why are these changes needed?\r\n\r\n<!-- Please give a short summary of the change and the problem this solves. -->\r\nWhen writing a Ray Dataset which contains blocks with no rows (e.g. after filtering) with `write_` methods, Ray currently creates an empty file for each block, which can lead to a growing collection of empty files. This PR skips this write for empty blocks.\r\n\r\nResult of running reproducible in the original issue (now does not write any empty files):\r\n```\r\nimport ray\r\n\r\nds = ray.data.range(10)\r\nhas_all = ds.filter(lambda x: x[\"id\"] < 1000).materialize()\r\nhas_none = ds.filter(lambda x: x['id'] > 1000).materialize()\r\n\r\npath = \"output_data/\"\r\n\r\nhas_all.write_json(f\"{path}/has_all/\")\r\nhas_none.write_json(f\"{path}/has_none/\")\r\n```\r\n![Screenshot at Jun 06 15-56-33](https://github.com/ray-project/ray/assets/5122851/1d8de498-a835-401f-81d5-f5e4e7b54d82)\r\n\r\n## Related issue number\r\n\r\n<!-- For example: \"Closes #1234\" -->\r\nCloses #36123\r\n\r\nSigned-off-by: Scott Lee <sjl@anyscale.com>","shortMessageHtmlLink":"[Data] Skip writing a file for an empty block in <code>write_datasource</code> (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1744746670\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/36134\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/36134/hovercard\" href=\"https://github.com/ray-project/ray/pull/36134\">r…</a>"}},{"before":"e9a1c6d814fb1a81033809f56695030d651388f5","after":"c1c466a8642aba0df4ebc87ecf2ace5dde532fca","ref":"refs/heads/master","pushedAt":"2023-06-09T20:11:11.986Z","pushType":"push","commitsCount":58,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[Data] Skip writing a file for an empty block in `write_datasource` (#36134)\n\n## Why are these changes needed?\r\n\r\n<!-- Please give a short summary of the change and the problem this solves. -->\r\nWhen writing a Ray Dataset which contains blocks with no rows (e.g. after filtering) with `write_` methods, Ray currently creates an empty file for each block, which can lead to a growing collection of empty files. This PR skips this write for empty blocks.\r\n\r\nResult of running reproducible in the original issue (now does not write any empty files):\r\n```\r\nimport ray\r\n\r\nds = ray.data.range(10)\r\nhas_all = ds.filter(lambda x: x[\"id\"] < 1000).materialize()\r\nhas_none = ds.filter(lambda x: x['id'] > 1000).materialize()\r\n\r\npath = \"output_data/\"\r\n\r\nhas_all.write_json(f\"{path}/has_all/\")\r\nhas_none.write_json(f\"{path}/has_none/\")\r\n```\r\n![Screenshot at Jun 06 15-56-33](https://github.com/ray-project/ray/assets/5122851/1d8de498-a835-401f-81d5-f5e4e7b54d82)\r\n\r\n## Related issue number\r\n\r\n<!-- For example: \"Closes #1234\" -->\r\nCloses #36123\r\n\r\nSigned-off-by: Scott Lee <sjl@anyscale.com>","shortMessageHtmlLink":"[Data] Skip writing a file for an empty block in <code>write_datasource</code> (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1744746670\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/36134\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/36134/hovercard\" href=\"https://github.com/ray-project/ray/pull/36134\">r…</a>"}},{"before":"9e3fc55917f1237f7d518302f1802a207972d424","after":"e9a1c6d814fb1a81033809f56695030d651388f5","ref":"refs/heads/master","pushedAt":"2023-06-07T21:34:49.983Z","pushType":"push","commitsCount":47,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[docs] fix last warning in the test (#36169)\n\nSigned-off-by: Huaiwei Sun <scottsun94@gmail.com>","shortMessageHtmlLink":"[docs] fix last warning in the test (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1746495834\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/36169\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/36169/hovercard\" href=\"https://github.com/ray-project/ray/pull/36169\">ray-project#36169</a>)"}},{"before":"9e3fc55917f1237f7d518302f1802a207972d424","after":"e9a1c6d814fb1a81033809f56695030d651388f5","ref":"refs/heads/master","pushedAt":"2023-06-07T21:34:49.926Z","pushType":"push","commitsCount":47,"pusher":{"login":"ollie-iterators","name":"Ollie","path":"/ollie-iterators","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46904826?s=80&v=4"},"commit":{"message":"[docs] fix last warning in the test (#36169)\n\nSigned-off-by: Huaiwei Sun <scottsun94@gmail.com>","shortMessageHtmlLink":"[docs] fix last warning in the test (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"1746495834\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ray-project/ray/issues/36169\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ray-project/ray/pull/36169/hovercard\" href=\"https://github.com/ray-project/ray/pull/36169\">ray-project#36169</a>)"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAENJGs_AA","startCursor":null,"endCursor":null}},"title":"Activity · ollie-iterators/ray"}