{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":730371795,"defaultBranch":"main","name":"QuantFour_AdamW_Cuda","ownerLogin":"lessw2020","currentUserCanPush":false,"isFork":false,"isEmpty":false,"createdAt":"2023-12-11T19:26:47.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/46302957?v=4","public":true,"private":false,"isOrgOwned":false},"refInfo":{"name":"","listCacheKey":"v0:1704844482.0","currentOid":""},"activityList":{"items":[{"before":"1af2b2d7578dde8cda1cb973f02647359bf1eed3","after":"6825de1007ce7776c504ef58f53c59435e146419","ref":"refs/heads/main","pushedAt":"2024-02-20T03:25:26.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"improve gmem access (coalesce, ldg intrinsics)","shortMessageHtmlLink":"improve gmem access (coalesce, ldg intrinsics)"}},{"before":"988e3274e2740a815d14f824f79f7689a084271a","after":"1af2b2d7578dde8cda1cb973f02647359bf1eed3","ref":"refs/heads/main","pushedAt":"2024-02-19T20:41:55.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"Update fused_quantfour.py","shortMessageHtmlLink":"Update fused_quantfour.py"}},{"before":"ae7b86c0734b9f125255582ad55cb800dcad7e85","after":"988e3274e2740a815d14f824f79f7689a084271a","ref":"refs/heads/main","pushedAt":"2024-02-19T20:39:25.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"Create perf_profiling.py","shortMessageHtmlLink":"Create perf_profiling.py"}},{"before":"4a08e30f1a2ec681bf8ec38206de1ee829b00007","after":"ae7b86c0734b9f125255582ad55cb800dcad7e85","ref":"refs/heads/main","pushedAt":"2024-02-08T17:39:27.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"update comments - double warp shuffle down synch","shortMessageHtmlLink":"update comments - double warp shuffle down synch"}},{"before":"c26ba52bf2a65e6a3019f2a4e0af2c3f4d8a97e3","after":"4a08e30f1a2ec681bf8ec38206de1ee829b00007","ref":"refs/heads/main","pushedAt":"2024-02-08T17:33:41.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"double warp shuffle down synch","shortMessageHtmlLink":"double warp shuffle down synch"}},{"before":"bcfe8f9e50e32401d177c7c1c3b062e99d63a859","after":"c26ba52bf2a65e6a3019f2a4e0af2c3f4d8a97e3","ref":"refs/heads/main","pushedAt":"2024-02-08T04:48:20.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"use syncwarp for second synch","shortMessageHtmlLink":"use syncwarp for second synch"}},{"before":"f3126f89d859b0ef11c9210cfc6479b632fa1aa0","after":"bcfe8f9e50e32401d177c7c1c3b062e99d63a859","ref":"refs/heads/main","pushedAt":"2024-02-08T04:41:26.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"use shuffle down synch warp primitive for seq_parallel_reduction","shortMessageHtmlLink":"use shuffle down synch warp primitive for seq_parallel_reduction"}},{"before":"1c7c6024d5073dc227ad746e5e303475c3fbf67a","after":"f3126f89d859b0ef11c9210cfc6479b632fa1aa0","ref":"refs/heads/main","pushedAt":"2024-02-08T01:44:59.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"cleanup - remove more atomicMax related code","shortMessageHtmlLink":"cleanup - remove more atomicMax related code"}},{"before":"896a1e2c224ddcb920518c44b8a8e89dbcafe50c","after":"1c7c6024d5073dc227ad746e5e303475c3fbf67a","ref":"refs/heads/main","pushedAt":"2024-02-07T23:47:38.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"cleanup - remove old atomicMax related code","shortMessageHtmlLink":"cleanup - remove old atomicMax related code"}},{"before":"771fa22dd2fba9ecd9924ea1db791fb5f7ffb37b","after":"896a1e2c224ddcb920518c44b8a8e89dbcafe50c","ref":"refs/heads/main","pushedAt":"2024-02-07T23:41:52.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"functionalize seq_threads_max_reducer","shortMessageHtmlLink":"functionalize seq_threads_max_reducer"}},{"before":"e53633435adaa32fd09e5e62fc45601377fe8db4","after":"771fa22dd2fba9ecd9924ea1db791fb5f7ffb37b","ref":"refs/heads/main","pushedAt":"2024-02-07T22:50:17.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"sequential threads parallel reduction implemented and working!","shortMessageHtmlLink":"sequential threads parallel reduction implemented and working!"}},{"before":"7a3345699dc3904c37335f8cc6f963d50c8263b1","after":"e53633435adaa32fd09e5e62fc45601377fe8db4","ref":"refs/heads/main","pushedAt":"2024-02-07T04:38:47.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"absmax can assume pos only","shortMessageHtmlLink":"absmax can assume pos only"}},{"before":"fde24c87eb48aae472bc120cef3af4df9d54089d","after":"7a3345699dc3904c37335f8cc6f963d50c8263b1","ref":"refs/heads/main","pushedAt":"2024-02-06T22:48:30.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"expand unit tests - passing with 10 iters, quantized is within ~2.5% vs fp32 AdamW","shortMessageHtmlLink":"expand unit tests - passing with 10 iters, quantized is within ~2.5% …"}},{"before":"f88e293f78ede99fd2c351190f5465ac8084b7fa","after":"fde24c87eb48aae472bc120cef3af4df9d54089d","ref":"refs/heads/main","pushedAt":"2024-02-05T18:13:41.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"start repo cleanup","shortMessageHtmlLink":"start repo cleanup"}},{"before":"4ddaa19b382af79f6a0b80f47fc2f995cb87bf08","after":"f88e293f78ede99fd2c351190f5465ac8084b7fa","ref":"refs/heads/main","pushedAt":"2024-02-05T04:01:44.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"add resid_beta1/2 for reduced computations","shortMessageHtmlLink":"add resid_beta1/2 for reduced computations"}},{"before":"ba6b86866a89938ec3d5b1c63b757a5bd2a6144b","after":"4ddaa19b382af79f6a0b80f47fc2f995cb87bf08","ref":"refs/heads/main","pushedAt":"2024-02-05T03:46:34.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"add const for bitmask, volatile for atomicPosMax","shortMessageHtmlLink":"add const for bitmask, volatile for atomicPosMax"}},{"before":"b1d1b9d027a357e710a71f0f19636a1ebfd0ade5","after":"ba6b86866a89938ec3d5b1c63b757a5bd2a6144b","ref":"refs/heads/main","pushedAt":"2024-02-04T23:39:53.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"add pragma unroll","shortMessageHtmlLink":"add pragma unroll"}},{"before":"be616a50d36fa4e289cd840bd9d83be8b2d205ad","after":"b1d1b9d027a357e710a71f0f19636a1ebfd0ade5","ref":"refs/heads/main","pushedAt":"2024-02-04T23:34:30.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"header and credits","shortMessageHtmlLink":"header and credits"}},{"before":"618d1681ba03da26940584911e835bd9a3fa6d05","after":"be616a50d36fa4e289cd840bd9d83be8b2d205ad","ref":"refs/heads/main","pushedAt":"2024-02-04T23:29:46.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"create right pack bitmask in const memory","shortMessageHtmlLink":"create right pack bitmask in const memory"}},{"before":"839686a1281fadabbd408fe77b11b62cd852e516","after":"618d1681ba03da26940584911e835bd9a3fa6d05","ref":"refs/heads/main","pushedAt":"2024-02-04T22:54:34.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"move bitmask to const memory","shortMessageHtmlLink":"move bitmask to const memory"}},{"before":"6552478f8bbd8780aeffec66ee7f21f045e4b708","after":"839686a1281fadabbd408fe77b11b62cd852e516","ref":"refs/heads/main","pushedAt":"2024-02-04T22:46:37.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"single compute for decoupled weight decay","shortMessageHtmlLink":"single compute for decoupled weight decay"}},{"before":"b2291aba5b27b7e0011e1b4224cc733f98194b3b","after":"6552478f8bbd8780aeffec66ee7f21f045e4b708","ref":"refs/heads/main","pushedAt":"2024-02-04T22:38:15.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"fail fast on thread index check, remove extra syncthreads","shortMessageHtmlLink":"fail fast on thread index check, remove extra syncthreads"}},{"before":"a273819c34216674d694f6efb842c798cd4b1090","after":"b2291aba5b27b7e0011e1b4224cc733f98194b3b","ref":"refs/heads/main","pushedAt":"2024-02-04T21:57:25.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"update corrected atomic float max","shortMessageHtmlLink":"update corrected atomic float max"}},{"before":"25f7e056e36b461927943cf1a40f61773f3d24e4","after":"a273819c34216674d694f6efb842c798cd4b1090","ref":"refs/heads/main","pushedAt":"2024-02-04T21:45:41.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"add corrected atomic float max","shortMessageHtmlLink":"add corrected atomic float max"}},{"before":"ab1c66fcf4dcb2a5c33f973a584433d3a3be7f0b","after":"25f7e056e36b461927943cf1a40f61773f3d24e4","ref":"refs/heads/main","pushedAt":"2024-02-04T21:38:31.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"improve formatting, add few comments","shortMessageHtmlLink":"improve formatting, add few comments"}},{"before":"7ea9cc0bdc5c5a29aa48c2c3aad9a068157bae44","after":"ab1c66fcf4dcb2a5c33f973a584433d3a3be7f0b","ref":"refs/heads/main","pushedAt":"2024-02-04T21:34:12.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"directly inline absmax comparisons","shortMessageHtmlLink":"directly inline absmax comparisons"}},{"before":"742f2866a5c73b4faf7ea08050b96d076b4377dc","after":"7ea9cc0bdc5c5a29aa48c2c3aad9a068157bae44","ref":"refs/heads/main","pushedAt":"2024-02-04T21:12:21.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"Update README.md","shortMessageHtmlLink":"Update README.md"}},{"before":"812961da666287477cf55b9ac24f7808884dc536","after":"742f2866a5c73b4faf7ea08050b96d076b4377dc","ref":"refs/heads/main","pushedAt":"2024-02-04T21:11:37.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"Update README.md","shortMessageHtmlLink":"Update README.md"}},{"before":"08d284a28034cfacd3bc28e138294166bb73b671","after":"812961da666287477cf55b9ac24f7808884dc536","ref":"refs/heads/main","pushedAt":"2024-02-04T21:10:53.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"Update README.md","shortMessageHtmlLink":"Update README.md"}},{"before":"865e3e66068951b1dc3fe934115a802b3e76fe79","after":"08d284a28034cfacd3bc28e138294166bb73b671","ref":"refs/heads/main","pushedAt":"2024-02-04T21:05:32.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"lessw2020","name":"Less Wright","path":"/lessw2020","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/46302957?s=80&v=4"},"commit":{"message":"Update README.md","shortMessageHtmlLink":"Update README.md"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAD_z3cBAA","startCursor":null,"endCursor":null}},"title":"Activity · lessw2020/QuantFour_AdamW_Cuda"}