biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

Convert formatting to use ruff instead of yapf + isort (#26247)

Browse Source 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-10-05 15:06:22 +01:00
committed by GitHub
parent 17edd8a807
commit d6953beb91
1508 changed files with 115244 additions and 94146 deletions
Download Patch File Download Diff File Expand all files Collapse all files

119
tests/lora/conftest.py
View File

@@ -10,12 +10,16 @@ import torch
import torch.nn as nn
from huggingface_hub import snapshot_download
from vllm.distributed import (cleanup_dist_env_and_memory,
init_distributed_environment,
initialize_model_parallel)
from vllm.model_executor.layers.linear import (ColumnParallelLinear,
MergedColumnParallelLinear,
RowParallelLinear)
from vllm.distributed import (
cleanup_dist_env_and_memory,
init_distributed_environment,
initialize_model_parallel,
)
from vllm.model_executor.layers.linear import (
ColumnParallelLinear,
MergedColumnParallelLinear,
RowParallelLinear,
)
from vllm.model_executor.layers.logits_processor import LogitsProcessor
from vllm.model_executor.layers.vocab_parallel_embedding import ParallelLMHead
from vllm.model_executor.models.interfaces import SupportsLoRA
@@ -47,11 +51,13 @@ def dist_init():
if current_platform.is_cpu() or current_platform.is_tpu():
backend = "gloo"
init_distributed_environment(world_size=1,
rank=0,
distributed_init_method=f"file://{temp_file}",
local_rank=0,
backend=backend)
init_distributed_environment(
world_size=1,
rank=0,
distributed_init_method=f"file://{temp_file}",
local_rank=0,
backend=backend,
)
initialize_model_parallel(1, 1)
yield
cleanup_dist_env_and_memory(shutdown_ray=True)
@@ -66,10 +72,9 @@ def dist_init_torch_only():
backend = "gloo"
temp_file = tempfile.mkstemp()[1]
torch.distributed.init_process_group(world_size=1,
rank=0,
init_method=f"file://{temp_file}",
backend=backend)
torch.distributed.init_process_group(
world_size=1, rank=0, init_method=f"file://{temp_file}", backend=backend
)
class DummyLoRAModel(nn.Sequential, SupportsLoRA):
@@ -79,24 +84,30 @@ class DummyLoRAModel(nn.Sequential, SupportsLoRA):
@pytest.fixture
def dummy_model() -> nn.Module:
model = DummyLoRAModel(
OrderedDict([
("dense1", ColumnParallelLinear(764, 100)),
("dense2", RowParallelLinear(100, 50)),
(
"layer1",
nn.Sequential(
OrderedDict([
("dense1", ColumnParallelLinear(100, 10)),
("dense2", RowParallelLinear(10, 50)),
])),
),
("act2", nn.ReLU()),
("output", ColumnParallelLinear(50, 10)),
("outact", nn.Sigmoid()),
# Special handling for lm_head & sampler
("lm_head", ParallelLMHead(512, 10)),
("logits_processor", LogitsProcessor(512)),
]))
OrderedDict(
[
("dense1", ColumnParallelLinear(764, 100)),
("dense2", RowParallelLinear(100, 50)),
(
"layer1",
nn.Sequential(
OrderedDict(
[
("dense1", ColumnParallelLinear(100, 10)),
("dense2", RowParallelLinear(10, 50)),
]
)
),
),
("act2", nn.ReLU()),
("output", ColumnParallelLinear(50, 10)),
("outact", nn.Sigmoid()),
# Special handling for lm_head & sampler
("lm_head", ParallelLMHead(512, 10)),
("logits_processor", LogitsProcessor(512)),
]
)
)
model.config = MagicMock()
model.embedding_modules = {"lm_head": "lm_head"}
model.unpadded_vocab_size = 32000
@@ -106,24 +117,30 @@ def dummy_model() -> nn.Module:
@pytest.fixture
def dummy_model_gate_up() -> nn.Module:
model = DummyLoRAModel(
OrderedDict([
("dense1", ColumnParallelLinear(764, 100)),
("dense2", RowParallelLinear(100, 50)),
(
"layer1",
nn.Sequential(
OrderedDict([
("dense1", ColumnParallelLinear(100, 10)),
("dense2", RowParallelLinear(10, 50)),
])),
),
("act2", nn.ReLU()),
("gate_up_proj", MergedColumnParallelLinear(50, [5, 5])),
("outact", nn.Sigmoid()),
# Special handling for lm_head & sampler
("lm_head", ParallelLMHead(512, 10)),
("logits_processor", LogitsProcessor(512)),
]))
OrderedDict(
[
("dense1", ColumnParallelLinear(764, 100)),
("dense2", RowParallelLinear(100, 50)),
(
"layer1",
nn.Sequential(
OrderedDict(
[
("dense1", ColumnParallelLinear(100, 10)),
("dense2", RowParallelLinear(10, 50)),
]
)
),
),
("act2", nn.ReLU()),
("gate_up_proj", MergedColumnParallelLinear(50, [5, 5])),
("outact", nn.Sigmoid()),
# Special handling for lm_head & sampler
("lm_head", ParallelLMHead(512, 10)),
("logits_processor", LogitsProcessor(512)),
]
)
)
model.config = MagicMock()
model.packed_modules_mapping = {
"gate_up_proj": [