xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-06-08 02:09:08 +08:00
Code Issues Packages Projects Releases Wiki Activity

[CI/Build] Cleanup LoRA test (#26752)

Browse Source 
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
This commit is contained in:
Jee Jee Li 2025-10-14 20:06:35 +08:00 committed by GitHub
parent c715ba3735
commit fdd32750f0
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 3 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
tests/lora/test_chatglm3_tp.py
View File

@ -58,7 +58,6 @@ def test_chatglm3_lora(chatglm3_lora_files):
max_loras=4, max_loras=4,
max_lora_rank=64, max_lora_rank=64,
trust_remote_code=True, trust_remote_code=True,
enable_chunked_prefill=True,
) )
output1 = do_sample(llm, chatglm3_lora_files, lora_id=1) output1 = do_sample(llm, chatglm3_lora_files, lora_id=1)
@ -70,7 +69,6 @@ def test_chatglm3_lora(chatglm3_lora_files):
@multi_gpu_test(num_gpus=4) @multi_gpu_test(num_gpus=4)
@create_new_process_for_each_test()
def test_chatglm3_lora_tp4(chatglm3_lora_files): def test_chatglm3_lora_tp4(chatglm3_lora_files):
llm = vllm.LLM( llm = vllm.LLM(
MODEL_PATH, MODEL_PATH,
@ -81,7 +79,6 @@ def test_chatglm3_lora_tp4(chatglm3_lora_files):
tensor_parallel_size=4, tensor_parallel_size=4,
trust_remote_code=True, trust_remote_code=True,
fully_sharded_loras=False, fully_sharded_loras=False,
enable_chunked_prefill=True,
) )
output1 = do_sample(llm, chatglm3_lora_files, lora_id=1) output1 = do_sample(llm, chatglm3_lora_files, lora_id=1)
@ -93,7 +90,6 @@ def test_chatglm3_lora_tp4(chatglm3_lora_files):
@multi_gpu_test(num_gpus=4) @multi_gpu_test(num_gpus=4)
@create_new_process_for_each_test()
def test_chatglm3_lora_tp4_fully_sharded_loras(chatglm3_lora_files): def test_chatglm3_lora_tp4_fully_sharded_loras(chatglm3_lora_files):
# https://github.com/NVIDIA/nccl/issues/1790, set a lower value for # https://github.com/NVIDIA/nccl/issues/1790, set a lower value for
# gpu_memory_utilization here because NCCL >= 2.26.3 seems to use # gpu_memory_utilization here because NCCL >= 2.26.3 seems to use
@ -107,7 +103,6 @@ def test_chatglm3_lora_tp4_fully_sharded_loras(chatglm3_lora_files):
tensor_parallel_size=4, tensor_parallel_size=4,
trust_remote_code=True, trust_remote_code=True,
fully_sharded_loras=True, fully_sharded_loras=True,
enable_chunked_prefill=True,
gpu_memory_utilization=0.85, gpu_memory_utilization=0.85,
) )
output1 = do_sample(llm, chatglm3_lora_files, lora_id=1) output1 = do_sample(llm, chatglm3_lora_files, lora_id=1)

3
tests/lora/test_llama_tp.py
View File

@ -113,7 +113,6 @@ def test_llama_lora(sql_lora_files):
@multi_gpu_test(num_gpus=4) @multi_gpu_test(num_gpus=4)
@create_new_process_for_each_test()
def test_llama_lora_tp4(sql_lora_files): def test_llama_lora_tp4(sql_lora_files):
llm = vllm.LLM( llm = vllm.LLM(
MODEL_PATH, MODEL_PATH,
@ -127,7 +126,6 @@ def test_llama_lora_tp4(sql_lora_files):
@multi_gpu_test(num_gpus=4) @multi_gpu_test(num_gpus=4)
@create_new_process_for_each_test()
def test_llama_lora_tp4_fully_sharded_loras(sql_lora_files): def test_llama_lora_tp4_fully_sharded_loras(sql_lora_files):
llm = vllm.LLM( llm = vllm.LLM(
MODEL_PATH, MODEL_PATH,
@ -142,7 +140,6 @@ def test_llama_lora_tp4_fully_sharded_loras(sql_lora_files):
@multi_gpu_test(num_gpus=2) @multi_gpu_test(num_gpus=2)
@create_new_process_for_each_test()
def test_tp2_serialize_and_deserialize_lora( def test_tp2_serialize_and_deserialize_lora(
tmp_path, sql_lora_files, sql_lora_huggingface_id tmp_path, sql_lora_files, sql_lora_huggingface_id
): ):

6
tests/lora/test_minicpmv_tp.py
View File

@ -8,7 +8,7 @@ from vllm.assets.image import ImageAsset
from vllm.lora.request import LoRARequest from vllm.lora.request import LoRARequest
from vllm.platforms import current_platform from vllm.platforms import current_platform
from ..utils import create_new_process_for_each_test from ..utils import multi_gpu_test
MODEL_PATH = "openbmb/MiniCPM-Llama3-V-2_5" MODEL_PATH = "openbmb/MiniCPM-Llama3-V-2_5"
@ -88,7 +88,7 @@ def test_minicpmv_lora(minicpmv_lora_files):
current_platform.is_rocm(), current_platform.is_rocm(),
reason="MiniCPM-V dependency xformers incompatible with ROCm", reason="MiniCPM-V dependency xformers incompatible with ROCm",
) )
@create_new_process_for_each_test() @multi_gpu_test(num_gpus=4)
def test_minicpmv_tp4_wo_fully_sharded_loras(minicpmv_lora_files): def test_minicpmv_tp4_wo_fully_sharded_loras(minicpmv_lora_files):
llm = vllm.LLM( llm = vllm.LLM(
MODEL_PATH, MODEL_PATH,
@ -112,7 +112,7 @@ def test_minicpmv_tp4_wo_fully_sharded_loras(minicpmv_lora_files):
current_platform.is_rocm(), current_platform.is_rocm(),
reason="MiniCPM-V dependency xformers incompatible with ROCm", reason="MiniCPM-V dependency xformers incompatible with ROCm",
) )
@create_new_process_for_each_test() @multi_gpu_test(num_gpus=4)
def test_minicpmv_tp4_fully_sharded_loras(minicpmv_lora_files): def test_minicpmv_tp4_fully_sharded_loras(minicpmv_lora_files):
llm = vllm.LLM( llm = vllm.LLM(
MODEL_PATH, MODEL_PATH,