[V1] V1 Enablement Oracle (#13726)

Signed-off-by: rshaw@neuralmagic.com <rshaw@neuralmagic.com>
Co-authored-by: rshaw@neuralmagic.com <rshaw@neuralmagic.com>
Co-authored-by: Nicolò Lucchesi <nlucches@redhat.com>
Co-authored-by: Tyler Michael Smith <tyler@neuralmagic.com>
Co-authored-by: Michael Goin <michael@neuralmagic.com>
This commit is contained in:
Robert Shaw
2025-03-14 22:02:20 -07:00
committed by GitHub
co-authored by rshaw@neuralmagic.com <rshaw@neuralmagic.com> Nicolò Lucchesi Tyler Michael Smith Michael Goin
parent 8c0d15d5c5
commit d4d93db2c5
96 changed files with 1537 additions and 512 deletions
+4
View File
@@ -80,6 +80,8 @@ def v1(run_with_both_engines_lora):
pass
# V1 Test: Failing due to numerics on V1.
@pytest.mark.skip_v1
@fork_new_process_for_each_test
def test_llama_lora(sql_lora_files):
@@ -123,6 +125,8 @@ def test_llama_lora_warmup(sql_lora_files):
"less when using lora than when not using lora")
# V1 Test: Failing due to numerics on V1.
@pytest.mark.skip_v1
@multi_gpu_test(num_gpus=4)
@fork_new_process_for_each_test
def test_llama_lora_tp4(sql_lora_files):
+2 -2
View File
@@ -8,7 +8,7 @@ import os
import pytest
from vllm.engine.arg_utils import AsyncEngineArgs, EngineArgs
from vllm.entrypoints.llm import LLM
from vllm.engine.llm_engine import LLMEngine
from vllm.lora.request import LoRARequest
MODEL_PATH = "meta-llama/Llama-2-7b-hf"
@@ -43,7 +43,7 @@ def test_lora_functions_sync():
gpu_memory_utilization=0.8,
enforce_eager=True)
llm = LLM.get_engine_class().from_engine_args(engine_args)
llm = LLMEngine.from_engine_args(engine_args)
def run_check(fn, args, expected: list):
fn(args)
+3
View File
@@ -7,6 +7,7 @@ import torch
from safetensors.torch import load_file
from torch import nn
from vllm import envs
from vllm.config import LoRAConfig
from vllm.lora.layers import (ColumnParallelLinearWithLoRA,
MergedColumnParallelLinearWithLoRA,
@@ -410,6 +411,7 @@ def test_lru_lora_model_manager(dist_init, dummy_model, device):
assert manager.device == device
@pytest.mark.skipif(envs.VLLM_USE_V1, reason="Test leverages V0 internals.")
@pytest.mark.parametrize("device", DEVICES)
def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings,
sql_lora_files, device):
@@ -489,6 +491,7 @@ def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings,
device)
@pytest.mark.skipif(envs.VLLM_USE_V1, reason="Test leverages V0 internals.")
@pytest.mark.parametrize("device", DEVICES)
def test_worker_adapter_manager(llama_2_7b_model_extra_embeddings,
sql_lora_files, device):