[Misc] Fix Current vLLM config is not set. warnings, assert to avoid issues in the future (#31747)

Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
This commit is contained in:
Lucas Wilkinson
2026-01-08 15:20:49 -08:00
committed by GitHub
co-authored by Luka Govedič
parent 5d3b6097ad
commit 6cdf015c3c
48 changed files with 380 additions and 240 deletions
+8 -8
View File
@@ -6,14 +6,14 @@ import torch
from vllm.v1.worker.utils import bind_kv_cache
def test_bind_kv_cache():
def test_bind_kv_cache(default_vllm_config):
from vllm.attention.layer import Attention
ctx = {
"layers.0.self_attn": Attention(32, 128, 0.1),
"layers.1.self_attn": Attention(32, 128, 0.1),
"layers.2.self_attn": Attention(32, 128, 0.1),
"layers.3.self_attn": Attention(32, 128, 0.1),
"layers.0.self_attn": Attention(32, 128, 0.1, prefix="layers.0.self_attn"),
"layers.1.self_attn": Attention(32, 128, 0.1, prefix="layers.1.self_attn"),
"layers.2.self_attn": Attention(32, 128, 0.1, prefix="layers.2.self_attn"),
"layers.3.self_attn": Attention(32, 128, 0.1, prefix="layers.3.self_attn"),
}
kv_cache = {
"layers.0.self_attn": torch.zeros((1,)),
@@ -34,13 +34,13 @@ def test_bind_kv_cache():
assert runner_kv_caches[3] is kv_cache["layers.3.self_attn"]
def test_bind_kv_cache_non_attention():
def test_bind_kv_cache_non_attention(default_vllm_config):
from vllm.attention.layer import Attention
# example from Jamba PP=2
ctx = {
"model.layers.20.attn": Attention(32, 128, 0.1),
"model.layers.28.attn": Attention(32, 128, 0.1),
"model.layers.20.attn": Attention(32, 128, 0.1, prefix="model.layers.20.attn"),
"model.layers.28.attn": Attention(32, 128, 0.1, prefix="model.layers.28.attn"),
}
kv_cache = {
"model.layers.20.attn": torch.zeros((1,)),