sgl-project--sglang
94057c3d3e
PR Test (NPU) / check-changes (push) Has been cancelled
PR Test (NPU) / pr-gate (push) Has been cancelled
PR Test (NPU) / set-image-config (push) Has been cancelled
PR Test (NPU) / stage-b-test-1-npu-a2 (0) (push) Has been cancelled
PR Test (NPU) / stage-b-test-1-npu-a2 (1) (push) Has been cancelled
PR Test (NPU) / stage-b-test-2-npu-a2 (0) (push) Has been cancelled
PR Test (NPU) / stage-b-test-2-npu-a2 (1) (push) Has been cancelled
PR Test (NPU) / stage-b-test-4-npu-a3 (push) Has been cancelled
PR Test (NPU) / stage-b-test-16-npu-a3 (push) Has been cancelled
PR Test (NPU) / multimodal-gen-test-1-npu-a3 (push) Has been cancelled
PR Test (NPU) / multimodal-gen-test-2-npu-a3 (push) Has been cancelled
PR Test (Arm64) / pr-gate (push) Has been cancelled
PR Test (Arm64) / check-changes (push) Has been cancelled
PR Test (Arm64) / build-test (push) Has been cancelled
PR Test (sgl-router) / gate (push) Has been cancelled
PR Test (sgl-router) / tier-1 — lint (push) Has been cancelled
PR Test (sgl-router) / tier-2 — build + test (push) Has been cancelled
PR Test (sgl-router) / tier-3 — docker (placeholder) (push) Has been cancelled
PR Test (sgl-router) / tier-3 — k8s integration (push) Has been cancelled
PR Test (sgl-router) / tier-3 — e2e (push) Has been cancelled
PR Test (sgl-router) / finish (push) Has been cancelled
PR Test (NPU) / single-node-poc (map[name:qwen3_6_27b_w8a8_1p_in64k_out1k_50ms runner:linux-aarch64-a3-2 test_case:test/registered/ascend/performance/qwen3_6_27b/test_npu_qwen3_6_27b_w8a8_1p_in64k_out1k_50ms.py test_type:perf]) (push) Has been cancelled
PR Test (NPU) / pr-test-npu-finish (push) Has been cancelled
PR Test (Xeon) / pr-gate (push) Has been cancelled
PR Test (Xeon) / check-changes (push) Has been cancelled
PR Test (Xeon) / build-test (, xeon-gnr, base-b-test-cpu) (push) Has been cancelled
PR Test (XPU) / check-changes (push) Has been cancelled
PR Test (XPU) / pr-gate (push) Has been cancelled
PR Test (XPU) / stage-a-test-1-gpu-xpu (push) Has been cancelled
PR Test (XPU) / wait-for-stage-a (push) Has been cancelled
PR Test (XPU) / stage-b-test-1-gpu-xpu (push) Has been cancelled
PR Test (XPU) / finish (push) Has been cancelled
CI Model Inventory / build-inventory (push) Has been cancelled
Lint / lint (push) Has been cancelled
PR Benchmark (SMG Components) / Benchmark Compilation Check (push) Has been cancelled
PR Benchmark (SMG Components) / Benchmark - Manual Policy (push) Has been cancelled
PR Benchmark (SMG Components) / Benchmark - Request Processing (push) Has been cancelled
PR Benchmark (SMG Components) / Benchmark Summary (push) Has been cancelled
PR Test (SMG) / build-wheel (push) Has been cancelled
Release SGLang Model Gateway to PyPI / build on windows (x86_64 - auto) (push) Has been cancelled
Release SGLang Model Gateway to PyPI / build on macos (x86_64 - auto) (push) Has been cancelled
PR Test (SMG) / python-unit-tests (push) Has been cancelled
PR Test (SMG) / unit-tests (push) Has been cancelled
PR Test (SMG) / benchmarks (push) Has been cancelled
PR Test (SMG) / chat-completions (push) Has been cancelled
PR Test (SMG) / chat-completions-4gpu (push) Has been cancelled
PR Test (SMG) / e2e (push) Has been cancelled
PR Test (SMG) / docker-build-test (push) Has been cancelled
PR Test (SMG) / k8s-integration (push) Has been cancelled
PR Test (SMG) / finish (push) Has been cancelled
PR Test (SMG) / summarize-benchmarks (push) Has been cancelled
Release SGLang Model Gateway Docker Image / publish (push) Has been cancelled
Release SGLang Model Gateway to PyPI / build on macos (aarch64 - auto) (push) Has been cancelled
Release SGLang Model Gateway to PyPI / build on linux (aarch64 - auto) (push) Has been cancelled
Release SGLang Model Gateway to PyPI / build on linux (x86_64 - auto) (push) Has been cancelled
Release SGLang Model Gateway to PyPI / build on linux (aarch64 - musllinux_1_1) (push) Has been cancelled
Release SGLang Model Gateway to PyPI / build on linux (x86_64 - musllinux_1_1) (push) Has been cancelled
Release SGLang Model Gateway to PyPI / Build SDist (push) Has been cancelled
Release SGLang Model Gateway to PyPI / Upload to PyPI (push) Has been cancelled
Release SGLang Kernels / build-cu129-matrix (aarch64, 12.9, 3.10, arm-kernel-build-node) (push) Has been cancelled
Release SGLang Kernels / build-cu129-matrix (x86_64, 12.9, 3.10, x64-kernel-build-node) (push) Has been cancelled
Release SGLang Kernels / release-cu129 (push) Has been cancelled
Release SGLang Kernels / build-cu130-matrix (aarch64, 13.0, 3.10, arm-kernel-build-node) (push) Has been cancelled
Release SGLang Kernels / build-cu130-matrix (x86_64, 13.0, 3.10, x64-kernel-build-node) (push) Has been cancelled
Release SGLang Kernels / release-cu130 (push) Has been cancelled
Release SGLang Kernels / build-rocm-matrix (3.10, 700) (push) Has been cancelled
Release SGLang Kernels / build-rocm-matrix (3.10, 720) (push) Has been cancelled
Release SGLang Kernels / release-rocm700 (push) Has been cancelled
Release SGLang Kernels / release-rocm720 (push) Has been cancelled
Release SGLang Kernels / build-musa43 (43, 3.10) (push) Has been cancelled
Release SGLang Kernels / release-musa43 (push) Has been cancelled
83 行
3.3 KiB
Python
83 行
3.3 KiB
Python
"""Tests for caller-supplied mm_hashes plumbing.
|
|
|
|
Verifies the contract that:
|
|
1. GenerateReqInput.mm_hashes is an optional list of hex strings.
|
|
2. MultimodalDataItem.set_pad_value() honors a pre-set hash and does NOT
|
|
overwrite it via hash_feature().
|
|
3. The derived pad_value is deterministic across requests with identical
|
|
mm_hashes — the property external KV routers depend on.
|
|
|
|
The wiring step that copies GenerateReqInput.mm_hashes into per-item
|
|
MultimodalDataItem.hash lives in tokenizer_manager.py and is exercised by
|
|
the e2e serve tests; this file pins the unit-level invariants the wiring
|
|
relies on.
|
|
"""
|
|
|
|
import unittest
|
|
from unittest.mock import patch
|
|
|
|
from sglang.srt.managers.io_struct import GenerateReqInput
|
|
from sglang.srt.managers.schedule_batch import (
|
|
Modality,
|
|
MultimodalDataItem,
|
|
_compute_pad_value,
|
|
)
|
|
from sglang.test.ci.ci_register import register_amd_ci, register_cuda_ci
|
|
from sglang.test.test_utils import CustomTestCase
|
|
|
|
register_cuda_ci(est_time=2, stage="base-b", runner_config="1-gpu-small")
|
|
register_amd_ci(est_time=2, suite="stage-b-test-1-gpu-small-amd")
|
|
|
|
|
|
class TestMmHashesContract(CustomTestCase):
|
|
def test_generate_req_input_accepts_mm_hashes(self):
|
|
"""GenerateReqInput exposes mm_hashes as an optional field."""
|
|
req = GenerateReqInput(
|
|
text="hi",
|
|
image_data=["http://example.com/img.png"],
|
|
mm_hashes=["deadbeefcafe1234"],
|
|
)
|
|
self.assertEqual(req.mm_hashes, ["deadbeefcafe1234"])
|
|
|
|
def test_generate_req_input_defaults_mm_hashes_to_none(self):
|
|
"""Absent mm_hashes preserves existing (None) behavior."""
|
|
req = GenerateReqInput(text="hi")
|
|
self.assertIsNone(req.mm_hashes)
|
|
|
|
def test_set_pad_value_honors_preset_hash(self):
|
|
"""set_pad_value() must use a pre-set hash without recomputing."""
|
|
item = MultimodalDataItem(modality=Modality.IMAGE, hash=0xDEADBEEF)
|
|
# If hash_feature is invoked, the test fails — we patch it to
|
|
# raise so any accidental recompute is loud.
|
|
with patch(
|
|
"sglang.srt.managers.mm_utils.hash_feature",
|
|
side_effect=AssertionError(
|
|
"hash_feature must NOT be called when hash is preset"
|
|
),
|
|
):
|
|
item.set_pad_value()
|
|
self.assertEqual(item.hash, 0xDEADBEEF)
|
|
self.assertEqual(item.pad_value, _compute_pad_value(0xDEADBEEF))
|
|
|
|
def test_set_pad_value_is_deterministic_across_items(self):
|
|
"""Two items with the same preset hash must derive the same pad_value."""
|
|
a = MultimodalDataItem(modality=Modality.IMAGE, hash=0x123456789ABCDEF0)
|
|
b = MultimodalDataItem(modality=Modality.IMAGE, hash=0x123456789ABCDEF0)
|
|
# No feature payload — set_pad_value uses the preset hash.
|
|
a.set_pad_value()
|
|
b.set_pad_value()
|
|
self.assertEqual(a.pad_value, b.pad_value)
|
|
self.assertEqual(a.hash, b.hash)
|
|
|
|
def test_set_pad_value_distinguishes_different_preset_hashes(self):
|
|
"""Distinct preset hashes must produce distinct pad_values."""
|
|
a = MultimodalDataItem(modality=Modality.IMAGE, hash=0xAAAA)
|
|
b = MultimodalDataItem(modality=Modality.IMAGE, hash=0xBBBB)
|
|
a.set_pad_value()
|
|
b.set_pad_value()
|
|
self.assertNotEqual(a.pad_value, b.pad_value)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
unittest.main()
|