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.

Steve Luo a8fd3b0342 [Inference/Kernel] Optimize paged attention: Refactor key cache layout (#5643) ... * optimize flashdecodingattention: refactor code with different key cache layout(from [num_blocks, num_kv_heads, block_size, head_size] to [num_blocks, num_kv_heads, head_size/x, block_size, x]) * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>		7 months ago
..
_C	…
_analyzer	[hotfix] quick fixes to make legacy tutorials runnable (#5559)	8 months ago
accelerator	[hotfix] fix typo change MoECheckpintIO to MoECheckpointIO (#5335)	9 months ago
amp	[npu] change device to accelerator api (#5239)	11 months ago
auto_parallel	[hotfix] Fix wrong import in meta_registry (#5392)	9 months ago
autochunk	[misc] update pre-commit and run all files (#4752)	1 year ago
booster	[devops] remove post commit ci (#5566)	8 months ago
checkpoint_io	[devops] remove post commit ci (#5566)	8 months ago
cli	[devops] fix extention building (#5427)	9 months ago
cluster	[devops] remove post commit ci (#5566)	8 months ago
context	[moe] merge moe into main (#4978)	1 year ago
device	[npu] add npu support for hybrid plugin and llama (#5090)	1 year ago
fx	[misc] update pre-commit and run all files (#4752)	1 year ago
inference	[Inference/Kernel] Optimize paged attention: Refactor key cache layout (#5643)	7 months ago
interface	[lazy] support from_pretrained (#4801)	1 year ago
kernel	[Fix/Inference] Fix GQA Triton and Support Llama3 (#5624)	7 months ago
lazy	[doc] add lazy init docs (#4808)	1 year ago
legacy	[Fix] resolve conflicts of merging main	8 months ago
logging	[misc] update pre-commit and run all files (#4752)	1 year ago
moe	[hotfix] fix typo change MoECheckpintIO to MoECheckpointIO (#5335)	9 months ago
nn	[hotfix] quick fixes to make legacy tutorials runnable (#5559)	8 months ago
pipeline	[shardformer, pipeline] add `gradient_checkpointing_ratio` and heterogenous shard policy for llama (#5508)	8 months ago
shardformer	[Fix] resolve conflicts of merging main	8 months ago
tensor	[devops] remove post commit ci (#5566)	8 months ago
testing	[shardformer] update colo attention to support custom mask (#5510)	8 months ago
utils	Merge pull request #5310 from hpcaitech/feature/npu	10 months ago
zero	[devops] remove post commit ci (#5566)	8 months ago
__init__.py	[devops] remove post commit ci (#5566)	8 months ago
initialize.py	[npu] change device to accelerator api (#5239)	11 months ago