flagos-ai/FlagScale

Python 483 stars

FlagScale is a large model toolkit based on open-sourced projects.

✓ Synced 3h ago

708 Merged PRs

2 days Avg Merge Time

0m Fastest PR

2 months Slowest PR

      #115
      Global Speed Rank
    

Top Reviewers

#	Title	Author	Time	Reviews	Blocks
#1116	[serve] Fix deploy check	@cyber-pioneer	1.5h	2	✓
#1113	Set logger stacklevel	@legitnull	2.6h	2	✓
#1109	Refactor pip install	@aoyulong	4 days	8	✓
#1110	[CICD] Update aquila tp2-pp2 ci for evaluation	@lxd-cumt	2 days	4	✓
#1108	[Tools]Update iluvatar patch history	@yanqingliu35-stack	2m	0	✓
#1104	[iluvatar]Upgrade Megatron to v0.15.0rc and integrate TE-FL	@DannyP0	2 days	1	✓
#1100	Refactor install scripts and add dockerfiles	@aoyulong	3 days	13	✓
#1105	[Kunlunxin] Support Qwen3-0.6B training	@Sanbao1052	58m	0	✓
#1101	Revert "[Kunlunxin] Support Qwen3-0.6B training"	@tengqm	1m	0	✓
#1085	[WIP] Refactor the dockerfile and install sytem	@aoyulong	4 days	0	✓
#1099	[Tools]Update patch history	@yanqingliu35-stack	1m	0	✓
#1080	[Kunlunxin] Support Qwen3-0.6B training	@ssuurrffaaccee	8 days	0	✓
#1087	[MLU] Support qwen3-0.6b training	@cifar10	3 days	0	✓
#1096	Update project banner	@tengqm	4.1h	0	✓
#1054	[BugFix] Resume partial tests for inference and serving	@zihugithub	20 days	30	✓
#1089	update runtime_env in rl configs	@physics31415926	1 day	5	✓
#1093	[Tools]Update Huawei patch history	@yanqingliu35-stack	1.1h	0	✓
#1092	[Huawei] Support X0 with lerobot dataset on Huawei Ascend 910C	@Hchnr	5m	0	✓
#1091	[Tools] Update hygon patch history	@yanqingliu35-stack	9m	1	✓
#1088	[Hygon] Support qwen3-0.6B training	@Leewzh	1 day	0	✓