Comparing dd9db6b4d2..efee86e49c - xc_validation_strategy - Gitea: Git with a cup of tea

zhoushasha/xc_validation_strategy

Author	SHA1	Message	Date
zhousha	efee86e49c	submit 24 models on zhoushasha: ppu_zw_810e(18)/hygon_k100-ai(4)/MetaX_c-500(1)/Biren_166m(1) Candidate pool is the 24711 already-downloaded models, which satisfies mechanism A's precondition that the model be present in platform storage. Strict and relaxed condition-2 give identical results this round: of the 24007 pooled models with verify records, 24006 already have 已验证 on some card, so relaxing to include 验证中 adds just one. The binding constraint is now how few models lack a record on each card - Kunlunxin_p-800, Cambricon_mlu-370-x8 and Iluvatar_bi-150 are all at zero and are left out of GPU_JOBS. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-09-20 18:27:33 +08:00
zhousha	793c49aea3	submit round-24 filter results on zhoushasha: MetaX_c-500(63)/Kunlunxin_p-800(1)/Cambricon_mlu-370-x8(1)/Biren_166m(95) + ppu_zw_810e(57), 217 models total Also add a hygon_k100-ai config branch and HYGON_MODELS list (kept available but excluded from GPU_JOBS per request). Refresh the expired AUTH_TOKEN. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-09-15 17:25:35 +08:00
zhousha	d5f576fac7	submit MetaX_c-500(7) + Biren_166m(95) non-quantized models; refresh AUTH_TOKEN	2026-08-31 15:18:39 +08:00
zhousha	965783957e	add self-looping quota retry: retry quota-blocked models every 30min in-process, no redeploy needed	2026-08-19 13:59:54 +08:00
zhousha	cd5c7a2435	submit entire remaining ppu_zw_810e candidate pool (5424 models); run to natural quota exhaustion	2026-08-19 11:19:18 +08:00
zhousha	cdbacf5a46	submit next batch of ppu_zw_810e (200 models)	2026-08-19 01:15:33 +08:00
zhousha	5bfc0fc53e	submit next batch of ppu_zw_810e (200 models)	2026-08-19 01:07:58 +08:00
zhousha	606655b876	submit entire remaining ppu_zw_810e candidate pool (6220 models); run to natural quota exhaustion	2026-08-18 15:55:51 +08:00
zhousha	eee9e5813c	submit next batch of ppu_zw_810e (250 models, matches remaining ~2000-slot quota headroom)	2026-08-18 15:43:55 +08:00
zhousha	32122cd866	submit next batch of ppu_zw_810e (360 models, supersedes stuck v1.0.28 build)	2026-08-18 15:34:19 +08:00
zhousha	ebc9f400e0	submit next batch of ppu_zw_810e (250 models); refresh AUTH_TOKEN	2026-08-18 11:13:33 +08:00
zhousha	b546ad980f	submit next batch of ppu_zw_810e (300 models)	2026-08-13 16:56:17 +08:00
zhousha	54adf4f956	submit next batch of ppu_zw_810e (200 models)	2026-08-11 14:30:02 +08:00
zhousha	66f378bdc0	submit next batch of ppu_zw_810e (200 models); refresh AUTH_TOKEN	2026-08-10 17:07:48 +08:00
zhousha	d49bf186ae	submit next batch of ppu_zw_810e (200 models)	2026-08-07 11:23:46 +08:00
zhousha	1b95e92f72	add next batch of ppu_zw_810e submission (500 models, lines 601-1100 of source list)	2026-08-06 13:06:23 +08:00

1 changed files with 152 additions and 1221 deletions

1373

main.py

View File

File diff suppressed because it is too large Load Diff