We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent c9b0789 commit b43edc8Copy full SHA for b43edc8
CHANGELOG.md
@@ -1,4 +1,14 @@
1
# CHANGELOG
2
+# [Version v1.8.2](https://github.com/intel/xFasterTransformer/releases/tag/v1.8.2)
3
+v1.8.2
4
+
5
+## Performance
6
+- Enable flash attention by default for `W8A8` dtype to accelerate the performance of the 1st token.
7
8
+## Benchmark
9
+- When the number of ranks is 1, run in single mode to avoid the dependency on `mpirun`.
10
+- Support `SNC-3` platform.
11
12
# [Version v1.8.1](https://github.com/intel/xFasterTransformer/releases/tag/v1.8.1)
13
v1.8.1
14
VERSION
@@ -1 +1 @@
-1.8.1
+1.8.2
0 commit comments