release: unilab 1.5.8

docs/sphinx/source/changelog.md

@@ -13,6 +13,22 @@ UniLab 遵循[语义化版本](https://semver.org/)。本共享页面以中英
 PyPI 版本变更与未发布变更；发布日期采用 PyPI 上传日期。完整提交历史请参阅
 [UniLab 仓库](https://github.com/unilabsim/UniLab)。
 
+## 1.5.8 (2026-11-05)
+
+### Changed / 变更
+
+- The default and ROCm profiles now pin `unisim-core==2.6.1`; this covers the
+  Motrix extra in both profiles and the SuperDex extra in the default profile.
+  The UniSim patch release speeds up MJWarp ordered body-pair contact
+  publication for sensor sets with at least eight views by aggregating unique
+  body pairs on device and reusing cached stable Torch references
+  ([UniSim #522](https://github.com/unilabsim/unisim/pull/522)).
+  默认与 ROCm profile 现在均钉定 `unisim-core==2.6.1`；两个 profile 均覆盖
+  Motrix extra，默认 profile 另覆盖 SuperDex extra。UniSim patch 版本通过
+  在 device 上聚合唯一 body pair 并复用缓存的稳定 Torch reference，提升了
+  至少八个 view 的 sensor 集合下 MJWarp 有序 body-pair contact 的发布性能
+  （[UniSim #522](https://github.com/unilabsim/unisim/pull/522)）。
+
 ## 1.5.7 (2026-11-04)
 
 ### Changed / 变更

pyproject.rocm.toml

@@ -4,7 +4,7 @@ build-backend = "uv_build"
 
 [project]
 name = "unilab"
-version = "1.5.7"
+version = "1.5.8"
 description = "Universal Lab for Robot Learning"
 readme = "README.md"
 license = "Apache-2.0"
@@ -23,7 +23,7 @@ classifiers = [
 requires-python = ">=3.10,<3.14"
 dependencies = [
     "numpy",
-    "unisim-core==2.6.0",
+    "unisim-core==2.6.1",
     "prettytable>=3.10",
     "torch==2.14.0",
     "triton-rocm==3.8.0 ; sys_platform == 'linux' and platform_machine == 'x86_64'",
@@ -68,7 +68,7 @@ mujoco = [
     "mujoco~=3.11.0",
     "mjbatch-uni~=0.2.5",
 ]
-motrix = ["unisim-core[motrix]==2.6.0"]
+motrix = ["unisim-core[motrix]==2.6.1"]
 genesis = ["genesis-world==1.3.3"]
 # Keep the training-stack extra aligned with pyproject.toml.
 unirl = [

pyproject.toml

@@ -14,7 +14,7 @@ source-exclude = [
 
 [project]
 name = "unilab"
-version = "1.5.7"
+version = "1.5.8"
 description = "Configurable, contract-driven robot learning across physics backends"
 readme = "README.md"
 license = "Apache-2.0"
@@ -33,7 +33,7 @@ classifiers = [
 requires-python = ">=3.10,<3.14"
 dependencies = [
     "numpy",
-    "unisim-core==2.6.0",
+    "unisim-core==2.6.1",
     "prettytable>=3.10",
     # A range (not an exact pin) lets ROCm users substitute a ROCm torch
     # build; uv.lock pins the tested CUDA builds via tool.uv.sources.
@@ -102,12 +102,12 @@ newton = [
     "pyglet>=2.1.6,<3",
     "imgui-bundle>=1.92.0",
 ]
-motrix = ["unisim-core[motrix]==2.6.0"]
+motrix = ["unisim-core[motrix]==2.6.1"]
 genesis = ["genesis-world==1.3.3"]
 # Wheels are CPython 3.12/3.13 Linux x86_64 only; the marker keeps uv lock
 # resolvable for the other required-environments.
 superdex = [
-    "unisim-core[superdex]==2.6.0 ; python_version >= '3.12' and sys_platform == 'linux' and platform_machine == 'x86_64'",
+    "unisim-core[superdex]==2.6.1 ; python_version >= '3.12' and sys_platform == 'linux' and platform_machine == 'x86_64'",
 ]
 # Tensor ray perception is optional and is discovered through UniSim's public
 # caster factory. The PyPI distribution is `uniray`; the import namespace is