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

[v1] Add fp32 support to v1 engine through flex attn (#19319)

Browse Source 
Signed-off-by: Isotr0py <2037008807@qq.com>
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
This commit is contained in:
Isotr0py
2025-06-09 22:10:44 +08:00
committed by GitHub
parent 770e5dcdb8
commit b8089195b4
4 changed files with 38 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
tests/v1/worker/test_gpu_model_runner.py
View File

@@ -4,6 +4,7 @@
import random
import pytest
import torch
from vllm.attention import Attention
from vllm.config import (CacheConfig, ModelConfig, ParallelConfig,
@@ -399,6 +400,7 @@ def test_load_model_weights_inplace(dist_init, model_runner, model_runner_2):
def test_init_kv_cache_with_kv_sharing_invalid_target_layer_order():
torch.set_default_dtype(torch.float16)
layer_0 = "model.layers.0.self_attn.attn"
layer_1 = "model.layers.1.self_attn.attn"
error_msg = f"{layer_1} must come before the current layer"
@@ -427,6 +429,7 @@ def test_init_kv_cache_with_kv_sharing_invalid_target_layer_order():
def test_init_kv_cache_with_kv_sharing_target_layer_not_exist():
torch.set_default_dtype(torch.float16)
layer_0 = "model.layers.0.self_attn.attn"
layer_1 = "model.layers.1.self_attn.attn"
invalid_layer = "model.layers.0.cross_attn.attn"
@@ -455,6 +458,7 @@ def test_init_kv_cache_with_kv_sharing_target_layer_not_exist():
def test_init_kv_cache_with_kv_sharing_target_same_as_current():
torch.set_default_dtype(torch.float16)
layer_0 = "model.layers.0.self_attn.attn"
layer_1 = "model.layers.1.self_attn.attn"
error_msg = f"{layer_1} cannot be the same as the current layer"
@@ -483,6 +487,7 @@ def test_init_kv_cache_with_kv_sharing_target_same_as_current():
def test_init_kv_cache_without_kv_sharing():
torch.set_default_dtype(torch.float16)
layer_0 = "model.layers.0.self_attn.attn"
layer_1 = "model.layers.1.self_attn.attn"
vllm_config = get_vllm_config()
@@ -550,6 +555,7 @@ def test_init_kv_cache_without_kv_sharing():
def test_init_kv_cache_with_kv_sharing_valid():
torch.set_default_dtype(torch.float16)
layer_0 = "model.layers.0.self_attn.attn"
layer_1 = "model.layers.1.self_attn.attn"
vllm_config = get_vllm_config()