Use monarch's distributed setup utility and colocate vLLM with its workers#409

Merged

allenwang28 merged 22 commits intometa-pytorch:mainfrom

allenwang28:titan_setup

Oct 15, 2025

Commits on Oct 14, 2025

Commits on Oct 15, 2025

vllm colocation works
allenwang28
committed
Zero to Forge Tutorials (#300)

init27
and
allenwang28
committed
Flatten GRPO main: Group and Episode (#400)

Jack-Khuu
authored and
allenwang28
committed
Remove use_dcp option in qwen3_8b (#415)

casteryh
authored and
allenwang28
committed
rebase
allenwang28
committed
Auto-track and globally shut down all Forge actors and services (#357)

DNXie
authored and
allenwang28
committed
Add CI badges to README.md (#412)

JenniferWang
authored and
allenwang28
committed
[EZ] remove extraneous `@pytest.mark.asyncio` (#418)

JenniferWang
authored and
allenwang28
committed
Fix policy update test (#365)

JenniferWang
authored and
allenwang28
committed
mark vLLM policy integration tests as async (#416)

ebsmothers
authored and
allenwang28
committed
Metric Logging updates 4/N - better actor name (#351)

authored and
allenwang28
committed
CI badge should point to main branch (#422)

casteryh
authored and
allenwang28
committed
Policy cleaner launch / setup (#401)

joecummings
authored and
allenwang28
committed
Rename service / actor spawning stuff (#423)

joecummings
authored and
allenwang28
committed
Configurable ReplayBuffer (#410)

pbontrager
authored and
allenwang28
committed
merge
allenwang28
committed
dict
allenwang28
committed
Merge branch 'main' into titan_setup
allenwang28
committed
gate setup dist
allenwang28
committed