You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

Zhongkai Zhao 75af66cd81 [Hotfix] Fix model policy matching strategy in ShardFormer (#5064) ... * hotfix/Fix get model policy strategy in ShardFormer * fix bug in auto policy		1 year ago
..
_C	…
_analyzer	…
amp	[npu] add npu support for gemini and zero (#5067)	1 year ago
auto_parallel	[npu] add npu support for gemini and zero (#5067)	1 year ago
autochunk	…
booster	[format] applied code formatting on changed files in pull request 5067 (#5072)	1 year ago
checkpoint_io	[pipeline,shardformer] Fix p2p efficiency in pipeline, allow skipping loading weight not in weight_map when `strict=False`, fix llama flash attention forward, add flop estimation by megatron in llama benchmark (#5017)	1 year ago
cli	…
cluster	[gemini] gemini support tensor parallelism. (#4942)	1 year ago
context	…
device	…
fx	…
inference	[Hotfix] Fix model policy matching strategy in ShardFormer (#5064)	1 year ago
interface	…
kernel	fix thrust-transform-reduce error (#5078)	1 year ago
lazy	…
legacy	[npu] add npu support for gemini and zero (#5067)	1 year ago
logging	…
moe	[hotfix]: modify create_ep_hierarchical_group and add test (#5032)	1 year ago
nn	[npu] add npu support for gemini and zero (#5067)	1 year ago
pipeline	[inference] refactor examples and fix schedule (#5077)	1 year ago
shardformer	[Hotfix] Fix model policy matching strategy in ShardFormer (#5064)	1 year ago
tensor	[hotfix]: modify create_ep_hierarchical_group and add test (#5032)	1 year ago
testing	…
utils	[npu] add npu support for gemini and zero (#5067)	1 year ago
zero	[gemini]fix gemini optimzer, saving Shardformer in Gemini got list assignment index out of range (#5085)	1 year ago
__init__.py	…
initialize.py	[npu] add npu support for gemini and zero (#5067)	1 year ago