 zhang-prog
|
5b71ecdafb
opt processors (#3714)
|
hai 7 meses |
 Lin Manhui
|
bc54f1bb90
[Infra] Prune and sort imports (#3694)
|
hai 7 meses |
|
Lin Manhui
|
a65d269b21
[Feat] Support multi-backend inference (#2860)
|
hai 7 meses |
|
zhang-prog
|
eb3e60685b
add cache func (#3509)
|
hai 8 meses |
|
zhang-prog
|
21519ccc29
add operation's location information in benchmark (#3487)
|
hai 8 meses |
|
zhang-prog
|
3c9400e6dd
Enhance benchmark (#3475)
|
hai 8 meses |
|
zhang-prog
|
5158085f23
[Feat] upgrade benchmark (#3416)
|
hai 8 meses |
 Tingquan Gao
|
5fb83f71f3
deprecate NEW_PREDICTOR & USE_NEW_INFERENCE flags (#2932)
|
hai 9 meses |
|
Tingquan Gao
|
e98fde5b8c
Refactor v2 (#2585)
|
hai 11 meses |
|
gaotingquan
|
7e680c7b4f
fix unit of e2e and warmup in benchmark
|
hai 1 ano |
|
gaotingquan
|
990eb14903
fix unit to ms in benchmark csv file
|
hai 1 ano |
|
gaotingquan
|
660a3ca14c
update benchmark info
|
hai 1 ano |
|
gaotingquan
|
b73c5036e3
separate Paddle Predictor to copy2gpu, infer, copy2cpu
|
hai 1 ano |
|
gaotingquan
|
a12f265f8f
update doc
|
hai 1 ano |
|
gaotingquan
|
18ba0d79f4
support warmup elapse
|
hai 1 ano |
|
gaotingquan
|
27225ed673
benchmark support end to end elapse
|
hai 1 ano |
|
gaotingquan
|
f20662bc7e
support benchmark
|
hai 1 ano |
