[08:47:40] 10Lift-Wing, 06Machine-Learning-Team: [LLM] quantization: allow loading model weights as int8/int4 with HF - https://phabricator.wikimedia.org/T377848#10399827 (10achou) **GPTQ** I tried the kevinbazira/aya-expanse-8b-gptq-4bit and it performed fast. The full inference code and outputs are available in P71700... [08:54:22] (03CR) 10Nik Gkountas: [C:03+2] Extra logging in the cache_updater task [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102369 (https://phabricator.wikimedia.org/T381889) (owner: 10Sbisson) [08:56:14] (03Merged) 10jenkins-bot: Extra logging in the cache_updater task [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102369 (https://phabricator.wikimedia.org/T381889) (owner: 10Sbisson) [09:21:40] hello! [12:58:22] o/ [14:31:20] hello [14:45:05] hi Chris! [15:09:45] (03PS1) 10Nik Gkountas: remove unused "custom_generate_unique_id" method [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102873 [15:11:25] (03PS1) 10Nik Gkountas: store in diskcache the process id of the worker that updates the cache [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102875 [15:12:03] (03CR) 10CI reject: [V:04-1] store in diskcache the process id of the worker that updates the cache [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102875 (owner: 10Nik Gkountas) [15:14:31] Me and Stephane are deploying rec-api in staging.. [15:31:52] (03PS2) 10Nik Gkountas: store in diskcache the process id of the worker that updates the cache [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102875 [16:06:55] (03CR) 10Sbisson: [C:03+2] shuffle recommendations for search and popular cases [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102354 (owner: 10Nik Gkountas) [16:07:34] (03CR) 10CI reject: [V:04-1] shuffle recommendations for search and popular cases [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102354 (owner: 10Nik Gkountas) [16:10:41] (03CR) 10Sbisson: [C:03+2] remove unused "custom_generate_unique_id" method [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102873 (owner: 10Nik Gkountas) [16:11:20] (03Merged) 10jenkins-bot: remove unused "custom_generate_unique_id" method [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102873 (owner: 10Nik Gkountas) [17:13:56] (03PS2) 10Nik Gkountas: shuffle recommendations for search and popular cases [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102354 [17:14:14] (03PS3) 10Nik Gkountas: store in diskcache the process id of the worker that updates the cache [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102875 [18:07:06] (03PS4) 10Nik Gkountas: store in diskcache the process id of the worker that updates the cache [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102875 [18:07:53] (03PS5) 10Nik Gkountas: store in diskcache the process id of the worker that updates the cache [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102875 [18:08:40] (03CR) 10Nik Gkountas: "recheck" [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102354 (owner: 10Nik Gkountas) [18:10:10] (03CR) 10Nik Gkountas: [V:03+2 C:03+2] shuffle recommendations for search and popular cases [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102354 (owner: 10Nik Gkountas) [18:10:50] (03Merged) 10jenkins-bot: shuffle recommendations for search and popular cases [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102354 (owner: 10Nik Gkountas) [19:03:20] (03PS1) 10Sbisson: Run cache updater task in all workers [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102926 [19:04:00] (03CR) 10CI reject: [V:04-1] Run cache updater task in all workers [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102926 (owner: 10Sbisson) [19:18:15] (03PS2) 10Sbisson: Run cache updater task in all workers [research/recommendation-api] - 10https://gerrit.wikimedia.org/r/1102926