Merge branch 'dev-1.x-zwf-rtm-speed' of github.com:zwfcrazy/mmpose in…

…to dev-1.x-zwf-rtm-speed
open-mmlab · Mar 10, 2023 · f7286fd · f7286fd
2 parents ab774ae + 1bba605
commit f7286fd
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 3 deletions.
diff --git a/projects/rtmpose/benchmark/README.md b/projects/rtmpose/benchmark/README.md
@@ -32,7 +32,7 @@ Currently tested:
 
 ### Speed Benchmark
 
-Numbers displayed in the table are inference latencies in millisecond.
+- Numbers displayed in the table are inference latencies in millisecond(ms).
 
 |   Config    | Input Size | ORT<sup><br>(i7-11700) | TRT-FP16<sup><br>(GTX 1660Ti) | TRT-FP16<sup><br>(RTX 3090) | ncnn-FP16<sup><br>(Snapdragon 865) | TRT-FP16<sup><br>(Jetson AGX Orin) | TRT-FP16<sup><br>(Jetson Orin NX) |
 | :---------: | :--------: | :--------------------: | :---------------------------: | :-------------------------: | :--------------------------------: | :--------------------------------: | :-------------------------------: |
@@ -55,7 +55,8 @@ Numbers displayed in the table are inference latencies in millisecond.
 
 ### Speed Benchmark
 
-Numbers displayed in the table are inference latencies in millisecond. Results contributed by different community users are separated by |.
+- Numbers displayed in the table are inference latencies in millisecond(ms).
+- Data from different community users are separated by `|`.
 
 | Config                                        | Input Size | ORT<sup><br>(i7-11700) | TRT-FP16<sup><br>(GTX 1660Ti) | TRT-FP16<sup><br>(RTX 3090) | TRT-FP16<sup><br>(Jetson AGX Orin) | TRT-FP16<sup><br>(Jetson Orin NX) |
 | :-------------------------------------------- | :--------: | :--------------------: | :---------------------------: | :-------------------------: | :--------------------------------: | :-------------------------------: |

diff --git a/projects/rtmpose/benchmark/README_CN.md b/projects/rtmpose/benchmark/README_CN.md
@@ -55,7 +55,8 @@
 
 ### Speed Benchmark
 
-图中所示为模型推理时间，单位毫秒。不同社区用户贡献的结果由|分割。
+- 图中所示为模型推理时间，单位毫秒。
+- 来自不同社区用户的测试数据用 `|` 分隔开。
 
 | Config                                        | Input Size | ORT<sup><br>(i7-11700) | TRT-FP16<sup><br>(GTX 1660Ti) | TRT-FP16<sup><br>(RTX 3090) | TRT-FP16<sup><br>(Jetson AGX Orin) | TRT-FP16<sup><br>(Jetson Orin NX) |
 | :-------------------------------------------- | :--------: | :--------------------: | :---------------------------: | :-------------------------: | :--------------------------------: | :-------------------------------: |