fix k700 (#1986)

Author: hukkai

configs/recognition/slowonly/README.md

@@ -34,8 +34,8 @@ We present SlowFast networks for video recognition. Our model involves (i) a Slo
 
 | frame sampling strategy |    scheduler     |   resolution   | gpus | backbone | pretrain | top1 acc | top5 acc | testing protocol  | inference time(video/s) | gpu_mem(M) |         config         |         ckpt         |         log         |
 | :---------------------: | :--------------: | :------------: | :--: | :------: | :------: | :------: | :------: | :---------------: | :---------------------: | :--------: | :--------------------: | :------------------: | :-----------------: |
-|         4x16x1          | Linear+MultiStep | short-side 320 | 8x2  | ResNet50 | ImageNet |  65.18   |  86.05   | 10 clips x 3 crop |            x            |    5797    | [config](/configs/recognition/slowonly/slowonly_imagenet-pretrained-r50_16xb16-4x16x1-steplr-150e_kinetics700-rgb.py) | [ckpt](https://download.openmmlab.com/mmaction/v1.0/recognition/slowonly/slowonly_imagenet-pretrained-r50_8xb16-4x16x1-steplr-150e_kinetics700-rgb/slowonly_imagenet-pretrained-r50_16xb16-4x16x1-steplr-150e_kinetics700-rgb_20220901-f73b3e89.pth) | [log](https://download.openmmlab.com/mmaction/v1.0/recognition/slowonly/slowonly_imagenet-pretrained-r50_16xb16-4x16x1-steplr-150e_kinetics700-rgb/slowonly_imagenet-pretrained-r50_16xb16-4x16x1-steplr-150e_kinetics700-rgb.log) |
-|          8x8x1          | Linear+MultiStep | short-side 320 | 8x2  | ResNet50 | ImageNet |  66.93   |  87.47   | 10 clips x 3 crop |            x            |   11089    | [config](/configs/recognition/slowonly/slowonly_imagenet-pretrained-r50_16xb16-8x8x1-steplr-150e_kinetics700-rgb.py) | [ckpt](https://download.openmmlab.com/mmaction/v1.0/recognition/slowonly/slowonly_imagenet-pretrained-r50_8xb16-8x8x1-steplr-150e_kinetics700-rgb/slowonly_imagenet-pretrained-r50_16xb16-8x8x1-steplr-150e_kinetics700-rgb_20220901-4098e1eb.pth) | [log](https://download.openmmlab.com/mmaction/v1.0/recognition/slowonly/slowonly_imagenet-pretrained-r50_16xb16-8x8x1-steplr-150e_kinetics700-rgb/slowonly_imagenet-pretrained-r50_16xb16-8x8x1-steplr-150e_kinetics700-rgb.log) |
+|         4x16x1          | Linear+MultiStep | short-side 320 | 8x2  | ResNet50 | ImageNet |  65.23   |  86.14   | 10 clips x 3 crop |            x            |    5797    | [config](/configs/recognition/slowonly/slowonly_imagenet-pretrained-r50_16xb16-4x16x1-steplr-150e_kinetics700-rgb.py) | [ckpt](https://download.openmmlab.com/mmaction/v1.0/recognition/slowonly/slowonly_imagenet-pretrained-r50_16xb16-4x16x1-steplr-150e_kinetics700-rgb/slowonly_imagenet-pretrained-r50_16xb16-4x16x1-steplr-150e_kinetics700-rgb_20221013-98b1b0a7.pth) | [log](https://download.openmmlab.com/mmaction/v1.0/recognition/slowonly/slowonly_imagenet-pretrained-r50_16xb16-4x16x1-steplr-150e_kinetics700-rgb/slowonly_imagenet-pretrained-r50_16xb16-4x16x1-steplr-150e_kinetics700-rgb.log) |
+|          8x8x1          | Linear+MultiStep | short-side 320 | 8x2  | ResNet50 | ImageNet |  67.36   |  87.64   | 10 clips x 3 crop |            x            |   11089    | [config](/configs/recognition/slowonly/slowonly_imagenet-pretrained-r50_16xb16-8x8x1-steplr-150e_kinetics700-rgb.py) | [ckpt](https://download.openmmlab.com/mmaction/v1.0/recognition/slowonly/slowonly_imagenet-pretrained-r50_16xb16-8x8x1-steplr-150e_kinetics700-rgb/slowonly_imagenet-pretrained-r50_16xb16-8x8x1-steplr-150e_kinetics700-rgb_20221013-15b93b10.pth) | [log](https://download.openmmlab.com/mmaction/v1.0/recognition/slowonly/slowonly_imagenet-pretrained-r50_16xb16-8x8x1-steplr-150e_kinetics700-rgb/slowonly_imagenet-pretrained-r50_16xb16-8x8x1-steplr-150e_kinetics700-rgb.log) |
 
 Note:
 

configs/recognition/slowonly/slowonly_imagenet-pretrained-r50_16xb16-4x16x1-steplr-150e_kinetics700-rgb.py

@@ -1,4 +1,4 @@
-_base_ = ['slowonly_r50_8xb16-4x16x1-256e_kinetics700-rgb.py']
+_base_ = ['slowonly_r50_8xb16-4x16x1-256e_kinetics400-rgb.py']
 
 model = dict(
     backbone=dict(