chore(main): release 0.0.9 (#359)

🤖 I have created a release *beep* *boop*
---


##
[0.0.9](v0.0.8...v0.0.9)
(2024-07-12)

### Bugfix

* fix the decode kernel segfault in cudagraph mode
([#368](https://github.com/flashinfer-ai/flashinfer/pull/368))([c69cfa](https://github.com/flashinfer-ai/flashinfer/commit/c69cfabc540e4a7edd991713df10d575ff3b0c21))
- fix decode kernels output for empty kv cache
([#363](https://github.com/flashinfer-ai/flashinfer/pull/363))([ac72b1](https://github.com/flashinfer-ai/flashinfer/commit/ac72b1cc14a6474d601f371c8d69e2600ac28d2f))
- check gpu id in PyTorch APIs and use input tensor's gpu default stream
([#361](https://github.com/flashinfer-ai/flashinfer/pull/361))([1b84fa](https://github.com/flashinfer-ai/flashinfer/commit/1b84fab3e4f53fb4fa26952fdb46fa8018634057))

### Performance Improvements

* accelerate alibi
([#365](#365))
([4f0a9f9](4f0a9f9))
* accelerate gqa performance
([#356](#356))
([e56ddad](e56ddad))
* Optimize tensor conversions in C++ code to avoid unnecessary copies
([#366](#366))
([1116237](1116237))

### Acknowledgement

We thank [@Yard1](https://github.com/Yard1),
[@Ying1123](https://github.com/Ying1123) and
[@zhyncs](https://github.com/zhyncs) for their contributions.

---
This PR was generated with [Release
Please](https://github.com/googleapis/release-please). See
[documentation](https://github.com/googleapis/release-please#release-please).

---------

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: Zihao Ye <expye@outlook.com>

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "0.0.8"
+  ".": "0.0.9"
 }

CHANGELOG.md

@@ -1,5 +1,23 @@
 # Changelog
 
+## [0.0.9](https://github.com/flashinfer-ai/flashinfer/compare/v0.0.8...v0.0.9) (2024-07-12)
+
+### Bugfix
+
+* fix the decode kernel segfault in cudagraph mode ([#368](https://github.com/flashinfer-ai/flashinfer/pull/368))([c69cfa](https://github.com/flashinfer-ai/flashinfer/commit/c69cfabc540e4a7edd991713df10d575ff3b0c21))
+- fix decode kernels output for empty kv cache ([#363](https://github.com/flashinfer-ai/flashinfer/pull/363))([ac72b1](https://github.com/flashinfer-ai/flashinfer/commit/ac72b1cc14a6474d601f371c8d69e2600ac28d2f))
+- check gpu id in PyTorch APIs and use input tensor's gpu default stream ([#361](https://github.com/flashinfer-ai/flashinfer/pull/361))([1b84fa](https://github.com/flashinfer-ai/flashinfer/commit/1b84fab3e4f53fb4fa26952fdb46fa8018634057))
+
+### Performance Improvements
+
+* accelerate alibi ([#365](https://github.com/flashinfer-ai/flashinfer/issues/365)) ([4f0a9f9](https://github.com/flashinfer-ai/flashinfer/commit/4f0a9f987ad2036f3c466257459de823be85fcc6))
+* accelerate gqa performance ([#356](https://github.com/flashinfer-ai/flashinfer/issues/356)) ([e56ddad](https://github.com/flashinfer-ai/flashinfer/commit/e56ddadf4bdbb164c3f1a03f9f69cb8a25621ef5))
+* Optimize tensor conversions in C++ code to avoid unnecessary copies ([#366](https://github.com/flashinfer-ai/flashinfer/issues/366)) ([1116237](https://github.com/flashinfer-ai/flashinfer/commit/1116237ac1e5690cf404841327b58b1d268d9951))
+
+### Acknowledgement
+
+We thank [@Yard1](https://github.com/Yard1), [@Ying1123](https://github.com/Ying1123) and [@zhyncs](https://github.com/zhyncs) for their contributions.
+
 ## [0.0.8](https://github.com/flashinfer-ai/flashinfer/compare/v0.0.7...v0.0.8) (2024-07-03)
 
 ### Bugfix

docs/conf.py

@@ -18,8 +18,8 @@
 author = "FlashInfer Contributors"
 copyright = "2023-2024, {}".format(author)
 
-version = "0.0.8"
-release = "0.0.8"
+version = "0.0.9"
+release = "0.0.9"
 
 # -- General configuration ---------------------------------------------------
 # https://www.sphinx-doc.org/en/master/usage/configuration.html#general-configuration

version.txt

@@ -1 +1 @@
-0.0.8
+0.0.9