Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/NVIDIA/FasterTransformer
/ data
Method
data
examples/cpp/xlnet/cnpy.h:36–39 ·
view source on GitHub ↗
Source
from the content-addressed store, hash-verified
34
35
template<typename T>
36
T* data()
37
{
38
return reinterpret_cast<T*>(&(*data_holder)[0]);
39
}
40
41
template<typename T>
42
const T* data() const
Callers
15
DqMmaPipelined
Class · 0.80
DqMmaMultistage
Class · 0.80
operand_A_ref
Method · 0.80
operand_B_ref
Method · 0.80
run_kernel
Method · 0.80
Params
Method · 0.80
MmaTensorOpDequantizer
Method · 0.80
MmaTensorOpDequantizer
Method · 0.80
MmaTensorOpDequantizer
Method · 0.80
MmaTensorOpDequantizer
Method · 0.80
SharedLoadIteratorMixed
Method · 0.80
malloc
Method · 0.80
Calls
no outgoing calls
Tested by
4
test_fp8_masked_softmax
Function · 0.64
main
Function · 0.64
reshape_key_cache
Function · 0.64
test_masked_multihead_attention
Function · 0.64