github
/
ColossalAI
mirror of https://github.com/hpcaitech/ColossalAI
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

[Bug FIx] import llama context ops fix (#4524) 

* added _vllm_rms_norm

* change place

* added tests

* added tests

* modify

* adding kernels

* added tests:

* adding kernels

* modify

* added

* updating kernels

* adding tests

* added tests

* kernel change

* submit

* modify

* added

* edit comments

* change name

* change commnets and fix import

* add

* added

* fix

* add ops into init.py

* add
pull/4532/head
Cuiqing Li 2023-08-28 13:41:57 +08:00 committed by GitHub
parent 2226c6836c
commit e937461312
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 14 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
colossalai/kernel/__init__.py
View File

@ -1,7 +1,14 @@
from .cuda_native import FusedScaleMaskSoftmax, LayerNorm, MultiHeadAttention
from .triton import llama_context_attn_fwd, bloom_context_attn_fwd
from .triton import softmax
from .triton import copy_kv_cache_to_dest
__all__ = [
"LayerNorm",
"FusedScaleMaskSoftmax",
"MultiHeadAttention",
"llama_context_attn_fwd",
"bloom_context_attn_fwd",
"softmax",
"copy_kv_cache_to_dest",
]

3
colossalai/kernel/triton/__init__.py Normal file
View File

@ -0,0 +1,3 @@
from .context_attention import llama_context_attn_fwd, bloom_context_attn_fwd
from .softmax import softmax
from .copy_kv_cache_dest import copy_kv_cache_to_dest

4
tests/test_infer_ops/triton/test_bloom_context_attention.py
View File

@ -9,8 +9,8 @@ from torch.nn import functional as F
try:
import triton
import triton.language as tl
from tests.test_kernels.triton.utils import benchmark, torch_context_attention
from colossalai.kernel.triton.context_attention import bloom_context_attn_fwd
from tests.test_infer_ops.triton.utils import benchmark, torch_context_attention
from colossalai.kernel.triton import bloom_context_attn_fwd
HAS_TRITON = True
except ImportError:
HAS_TRITON = False

4
tests/test_infer_ops/triton/test_llama_context_attention.py
View File

@ -9,8 +9,8 @@ from torch.nn import functional as F
try:
import triton
import triton.language as tl
from tests.test_kernels.triton.utils import benchmark, torch_context_attention
from colossalai.kernel.triton.context_attention import llama_context_attn_fwd
from tests.test_infer_ops.triton.utils import benchmark, torch_context_attention
from colossalai.kernel.triton import llama_context_attn_fwd
HAS_TRITON = True
except ImportError:
HAS_TRITON = False