squall/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[core][distributed] fix ray worker rank assignment (#6235)

Browse Source
This commit is contained in:
youkaichao 2024-07-08 21:31:44 -07:00 committed by GitHub
parent a3c9435d93
commit 70c232f85a
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 27 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

32
vllm/executor/ray_gpu_executor.py
View File

@ -134,11 +134,32 @@ class RayGPUExecutor(DistributedGPUExecutor):
worker_node_and_gpu_ids = self._run_workers("get_node_and_gpu_ids", worker_node_and_gpu_ids = self._run_workers("get_node_and_gpu_ids",
use_dummy_driver=True) use_dummy_driver=True)
node_workers = defaultdict(list) # the order in `worker_node_and_gpu_ids` does not necessarily match
node_gpus = defaultdict(list) # the machine boundaries. We need to make sure that workers in the
# same node are assigned consecutive ranks.
# examples:
# [('852a09a13c7503ef126d7c828454c741494b1be33a8627a5206604d9', [0]), ('dfaad7adfdae57a694cc74490db45bd112c9f31243523e43ddc2e7f0', [0]), ('dfaad7adfdae57a694cc74490db45bd112c9f31243523e43ddc2e7f0', [1]), ('dfaad7adfdae57a694cc74490db45bd112c9f31243523e43ddc2e7f0', [2]), ('dfaad7adfdae57a694cc74490db45bd112c9f31243523e43ddc2e7f0', [3]), ('852a09a13c7503ef126d7c828454c741494b1be33a8627a5206604d9', [1]), ('852a09a13c7503ef126d7c828454c741494b1be33a8627a5206604d9', [2]), ('852a09a13c7503ef126d7c828454c741494b1be33a8627a5206604d9', [3])] # noqa
for i, (node_id, gpu_ids) in enumerate(worker_node_and_gpu_ids): # initialize worker ranks with -1 (unassigned)
node_workers[node_id].append(i) worker_ranks = [-1 for x in worker_node_and_gpu_ids]
current_rank = 0
while -1 in worker_ranks:
# whenever we find an unassigned worker, find the node
index = worker_ranks.index(-1)
current_node_id = worker_node_and_gpu_ids[index][0]
# assign ranks to all workers in the same node
for i, (node_id, _) in enumerate(worker_node_and_gpu_ids):
if node_id == current_node_id:
worker_ranks[i] = current_rank
current_rank += 1
# with the above example, worker_ranks will be [0, 4, 5, 6, 7, 1, 2, 3]
node_workers = defaultdict(list) # node id -> list of worker ranks
node_gpus = defaultdict(list) # node id -> list of gpu ids
for worker_rank, (node_id, gpu_ids) in zip(worker_ranks,
worker_node_and_gpu_ids):
node_workers[node_id].append(worker_rank)
# `gpu_ids` can be a list of strings or integers. # `gpu_ids` can be a list of strings or integers.
# convert them to integers for consistency. # convert them to integers for consistency.
# NOTE: gpu_ids can be larger than 9 (e.g. 16 GPUs), # NOTE: gpu_ids can be larger than 9 (e.g. 16 GPUs),
@ -184,7 +205,8 @@ class RayGPUExecutor(DistributedGPUExecutor):
local_rank=node_workers[node_id].index(rank), local_rank=node_workers[node_id].index(rank),
rank=rank, rank=rank,
distributed_init_method=distributed_init_method, distributed_init_method=distributed_init_method,
) for rank, (node_id, _) in enumerate(worker_node_and_gpu_ids) ) for rank, (node_id,
_) in zip(worker_ranks, worker_node_and_gpu_ids)
] ]
self._run_workers("init_worker", all_kwargs=init_worker_all_kwargs) self._run_workers("init_worker", all_kwargs=init_worker_all_kwargs)