Merge remote-tracking branch 'origin/dev-1.x' into 1.x

.circleci/test.yml

@@ -31,7 +31,7 @@ jobs:
           name: Check docstring coverage
           command: |
             pip install interrogate
-            interrogate -v --ignore-init-method --ignore-module --ignore-nested-functions --ignore-regex "__repr__" --fail-under 60 mmcls
+            interrogate -v --ignore-init-method --ignore-module --ignore-nested-functions --ignore-magic --ignore-regex "__repr__" --fail-under 60 mmcls
   build_cpu:
     parameters:
       # The python version must match available image tags in
@@ -42,8 +42,6 @@ jobs:
         type: string
       torchvision:
         type: string
-      mmcv:
-        type: string
     docker:
       - image: cimg/python:<< parameters.python >>
     resource_class: large
@@ -57,31 +55,32 @@ jobs:
       - run:
           name: Configure Python & pip
           command: |
-            python -m pip install --upgrade pip
-            python -m pip install wheel
+            pip install --upgrade pip
+            pip install wheel
       - run:
           name: Install PyTorch
           command: |
             python -V
-            python -m pip install torch==<< parameters.torch >>+cpu torchvision==<< parameters.torchvision >>+cpu -f https://download.pytorch.org/whl/torch_stable.html
+            pip install torch==<< parameters.torch >>+cpu torchvision==<< parameters.torchvision >>+cpu -f https://download.pytorch.org/whl/torch_stable.html
       - run:
           name: Install mmcls dependencies
           command: |
-            python -m pip install git+ssh://git@github.com/open-mmlab/mmengine.git@main
-            python -m pip install << parameters.mmcv >>
-            python -m pip install timm
-            python -m pip install -r requirements.txt
+            pip install git+https://github.com/open-mmlab/mmengine.git@main
+            pip install -U openmim
+            mim install 'mmcv >= 2.0.0rc1'
+            pip install timm
+            pip install -r requirements.txt
             python -c 'import mmcv; print(mmcv.__version__)'
       - run:
           name: Build and install
           command: |
-            python -m pip install -e .
+            pip install -e .
       - run:
           name: Run unittests
           command: |
-            python -m coverage run --branch --source mmcls -m pytest tests/
-            python -m coverage xml
-            python -m coverage report -m
+            coverage run --branch --source mmcls -m pytest tests/
+            coverage xml
+            coverage report -m
 
   build_cuda:
     machine:
@@ -96,15 +95,13 @@ jobs:
       cudnn:
         type: integer
         default: 7
-      mmcv:
-        type: string
     steps:
       - checkout
       - run:
           # Cloning repos in VM since Docker doesn't have access to the private key
           name: Clone Repos
           command: |
-            git clone -b main --depth 1 ssh://git@github.com/open-mmlab/mmengine.git /home/circleci/mmengine
+            git clone -b main --depth 1 https://github.com/open-mmlab/mmengine.git /home/circleci/mmengine
       - run:
           name: Build Docker image
           command: |
@@ -114,7 +111,8 @@ jobs:
           name: Install mmcls dependencies
           command: |
             docker exec mmcls pip install -e /mmengine
-            docker exec mmcls pip install << parameters.mmcv >>
+            docker exec mmcls pip install -U openmim
+            docker exec mmcls mim install 'mmcv >= 2.0.0rc1'
             docker exec mmcls pip install -r requirements.txt
             docker exec mmcls python -c 'import mmcv; print(mmcv.__version__)'
       - run:
@@ -124,7 +122,7 @@ jobs:
       - run:
           name: Run unittests
           command: |
-            docker exec mmcls python -m pytest tests/ --ignore tests/test_models/test_backbones/test_timm_backbone.py
+            docker exec mmcls python -m pytest tests/ -k 'not timm'
 
 # Invoke jobs via workflows
 # See: https://circleci.com/docs/2.0/configuration-reference/#workflows
@@ -138,6 +136,7 @@ workflows:
             branches:
               ignore:
                 - dev-1.x
+                - 1.x
   pr_stage_test:
     when:
       not:
@@ -154,15 +153,13 @@ workflows:
           torch: 1.6.0
           torchvision: 0.7.0
           python: 3.6.9  # The lowest python 3.6.x version available on CircleCI images
-          mmcv: https://download.openmmlab.com/mmcv/dev-2.x/cpu/torch1.6.0/mmcv_full-2.0.0rc0-cp36-cp36m-manylinux1_x86_64.whl
           requires:
             - lint
       - build_cpu:
           name: maximum_version_cpu
-          torch: 1.9.0  # TODO: Update the version after mmcv provides more pre-compiled packages.
-          torchvision: 0.10.0
+          torch: 1.12.1
+          torchvision: 0.13.1
           python: 3.9.0
-          mmcv: https://download.openmmlab.com/mmcv/dev-2.x/cpu/torch1.9.0/mmcv_full-2.0.0rc0-cp39-cp39-manylinux1_x86_64.whl
           requires:
             - minimum_version_cpu
       - hold:
@@ -175,7 +172,6 @@ workflows:
           # Use double quotation mark to explicitly specify its type
           # as string instead of number
           cuda: "10.2"
-          mmcv: https://download.openmmlab.com/mmcv/dev-2.x/cu102/torch1.8.0/mmcv_full-2.0.0rc0-cp37-cp37m-manylinux1_x86_64.whl
           requires:
             - hold
   merge_stage_test:
@@ -188,7 +184,6 @@ workflows:
           torch: 1.6.0
           # Use double quotation mark to explicitly specify its type
           # as string instead of number
-          mmcv: https://download.openmmlab.com/mmcv/dev-2.x/cu101/torch1.6.0/mmcv_full-2.0.0rc0-cp37-cp37m-manylinux1_x86_64.whl
           cuda: "10.1"
           filters:
             branches:

.dev_scripts/benchmark_regression/1-benchmark_valid.py

@@ -2,15 +2,17 @@ import logging
 import re
 import tempfile
 from argparse import ArgumentParser
+from collections import OrderedDict
 from pathlib import Path
 from time import time
-from typing import OrderedDict
 
 import mmcv
 import numpy as np
 import torch
-from mmengine import Config, MMLogger, Runner
-from mmengine.dataset import Compose
+from mmengine import Config, DictAction, MMLogger
+from mmengine.dataset import Compose, default_collate
+from mmengine.fileio import FileClient
+from mmengine.runner import Runner
 from modelindex.load_model_index import load
 from rich.console import Console
 from rich.table import Table
@@ -52,6 +54,16 @@ def parse_args():
         '--flops-str',
         action='store_true',
         help='Output FLOPs and params counts in a string form.')
+    parser.add_argument(
+        '--cfg-options',
+        nargs='+',
+        action=DictAction,
+        help='override some settings in the used config, the key-value pair '
+        'in xxx=yyy format will be merged into config file. If the value to '
+        'be overwritten is a list, it should be like key="[a,b]" or key=a,b '
+        'It also allows nested list/tuple values, e.g. key="[(a,b),(c,d)]" '
+        'Note that the quotation marks are necessary and that no white space '
+        'is allowed.')
     args = parser.parse_args()
     return args
 
@@ -62,6 +74,8 @@ def inference(config_file, checkpoint, work_dir, args, exp_name):
     cfg.load_from = checkpoint
     cfg.log_level = 'WARN'
     cfg.experiment_name = exp_name
+    if args.cfg_options is not None:
+        cfg.merge_from_dict(args.cfg_options)
 
     # build the data pipeline
     test_dataset = cfg.test_dataloader.dataset
@@ -72,7 +86,8 @@ def inference(config_file, checkpoint, work_dir, args, exp_name):
         test_dataset.pipeline.insert(1, dict(type='Resize', scale=32))
 
     data = Compose(test_dataset.pipeline)({'img_path': args.img})
-    resolution = tuple(data['inputs'].shape[1:])
+    data = default_collate([data])
+    resolution = tuple(data['inputs'].shape[-2:])
 
     runner: Runner = Runner.from_cfg(cfg)
     model = runner.model
@@ -83,26 +98,30 @@ def inference(config_file, checkpoint, work_dir, args, exp_name):
         if args.inference_time:
             time_record = []
             for _ in range(10):
+                model.val_step(data)  # warmup before profiling
+                torch.cuda.synchronize()
                 start = time()
-                model.val_step([data])
+                model.val_step(data)
+                torch.cuda.synchronize()
                 time_record.append((time() - start) * 1000)
             result['time_mean'] = np.mean(time_record[1:-1])
             result['time_std'] = np.std(time_record[1:-1])
         else:
-            model.val_step([data])
+            model.val_step(data)
 
     result['model'] = config_file.stem
 
     if args.flops:
-        from mmcv.cnn.utils import get_model_complexity_info
+        from fvcore.nn import FlopCountAnalysis, parameter_count
+        from fvcore.nn.print_model_statistics import _format_size
+        _format_size = _format_size if args.flops_str else lambda x: x
         with torch.no_grad():
             if hasattr(model, 'extract_feat'):
                 model.forward = model.extract_feat
-                flops, params = get_model_complexity_info(
-                    model,
-                    input_shape=(3, ) + resolution,
-                    print_per_layer_stat=False,
-                    as_strings=args.flops_str)
+                model.to('cpu')
+                inputs = (torch.randn((1, 3, *resolution)), )
+                flops = _format_size(FlopCountAnalysis(model, inputs).total())
+                params = _format_size(parameter_count(model)[''])
                 result['flops'] = flops if args.flops_str else int(flops)
                 result['params'] = params if args.flops_str else int(params)
             else:
@@ -184,7 +203,6 @@ def main(args):
         if args.checkpoint_root is not None:
             root = args.checkpoint_root
             if 's3://' in args.checkpoint_root:
-                from mmcv.fileio import FileClient
                 from petrel_client.common.exception import AccessDeniedError
                 file_client = FileClient.infer_client(uri=root)
                 checkpoint = file_client.join_path(

.dev_scripts/benchmark_regression/2-benchmark_test.py

@@ -62,6 +62,12 @@ def parse_args():
         action='store_true',
         help='Summarize benchmark test results.')
     parser.add_argument('--save', action='store_true', help='Save the summary')
+    parser.add_argument(
+        '--cfg-options',
+        nargs='+',
+        type=str,
+        default=[],
+        help='Config options for all config files.')
 
     args = parser.parse_args()
     return args
@@ -76,7 +82,7 @@ def create_test_job_batch(commands, model_info, args, port, script_name):
 
     http_prefix = 'https://download.openmmlab.com/mmclassification/'
     if 's3://' in args.checkpoint_root:
-        from mmcv.fileio import FileClient
+        from mmengine.fileio import FileClient
         from petrel_client.common.exception import AccessDeniedError
         file_client = FileClient.infer_client(uri=args.checkpoint_root)
         checkpoint = file_client.join_path(
@@ -125,6 +131,7 @@ def create_test_job_batch(commands, model_info, args, port, script_name):
                   f'--work-dir={work_dir} '
                   f'--out={result_file} '
                   f'--cfg-option dist_params.port={port} '
+                  f'{" ".join(args.cfg_options)} '
                   f'--launcher={launcher}\n')
 
     with open(work_dir / 'job.sh', 'w') as f:

.dev_scripts/benchmark_regression/3-benchmark_train.py

@@ -3,22 +3,48 @@ import json
 import os
 import os.path as osp
 import re
+from collections import OrderedDict
 from datetime import datetime
 from pathlib import Path
 from zipfile import ZipFile
 
+import yaml
 from modelindex.load_model_index import load
 from rich.console import Console
 from rich.syntax import Syntax
 from rich.table import Table
 
 console = Console()
+MMCLS_ROOT = Path(__file__).absolute().parents[2]
+CYCLE_LEVELS = ['month', 'quarter', 'half-year', 'no-training']
 METRICS_MAP = {
     'Top 1 Accuracy': 'accuracy/top1',
     'Top 5 Accuracy': 'accuracy/top5'
 }
 
 
+class RangeAction(argparse.Action):
+
+    def __call__(self, parser, namespace, values: str, option_string):
+        matches = re.match(r'([><=]*)([-\w]+)', values)
+        if matches is None:
+            raise ValueError(f'Unavailable range option {values}')
+        symbol, range_str = matches.groups()
+        assert range_str in CYCLE_LEVELS, \
+            f'{range_str} are not in {CYCLE_LEVELS}.'
+        level = CYCLE_LEVELS.index(range_str)
+        symbol = symbol or '<='
+        ranges = set()
+        if '=' in symbol:
+            ranges.add(level)
+        if '>' in symbol:
+            ranges.update(range(level + 1, len(CYCLE_LEVELS)))
+        if '<' in symbol:
+            ranges.update(range(level))
+        assert len(ranges) > 0, 'No range are selected.'
+        setattr(namespace, self.dest, ranges)
+
+
 def parse_args():
     parser = argparse.ArgumentParser(
         description='Train models (in bench_train.yml) and compare accuracy.')
@@ -32,6 +58,14 @@ def parse_args():
     parser.add_argument('--port', type=int, default=29666, help='dist port')
     parser.add_argument(
         '--models', nargs='+', type=str, help='Specify model names to run.')
+    parser.add_argument(
+        '--range',
+        type=str,
+        default={0},
+        action=RangeAction,
+        metavar='{month,quarter,half-year,no-training}',
+        help='The training benchmark range, "no-training" means all models '
+        "including those we haven't trained.")
     parser.add_argument(
         '--work-dir',
         default='work_dirs/benchmark_train',
@@ -63,18 +97,33 @@ def parse_args():
         '--save',
         action='store_true',
         help='Save the summary and archive log files.')
+    parser.add_argument(
+        '--cfg-options',
+        nargs='+',
+        type=str,
+        default=[],
+        help='Config options for all config files.')
 
     args = parser.parse_args()
     return args
 
 
+def get_gpu_number(model_info):
+    config = osp.basename(model_info.config)
+    matches = re.match(r'.*[-_](\d+)xb(\d+).*', config)
+    if matches is None:
+        raise ValueError(
+            'Cannot get gpu numbers from the config name {config}')
+    gpus = int(matches.groups()[0])
+    return gpus
+
+
 def create_train_job_batch(commands, model_info, args, port, script_name):
 
     fname = model_info.name
 
-    assert 'Gpus' in model_info.data, \
-        f"Haven't specify gpu numbers for {fname}"
-    gpus = model_info.data['Gpus']
+    gpus = get_gpu_number(model_info)
+    gpus_per_node = min(gpus, 8)
 
     config = Path(model_info.config)
     assert config.exists(), f'"{fname}": {config} not found.'
@@ -101,15 +150,17 @@ def create_train_job_batch(commands, model_info, args, port, script_name):
                   f'#SBATCH --output {work_dir}/job.%j.out\n'
                   f'#SBATCH --partition={args.partition}\n'
                   f'#SBATCH --job-name {job_name}\n'
-                  f'#SBATCH --gres=gpu:8\n'
+                  f'#SBATCH --gres=gpu:{gpus_per_node}\n'
                   f'{mail_cfg}{quota_cfg}'
-                  f'#SBATCH --ntasks-per-node=8\n'
+                  f'#SBATCH --ntasks-per-node={gpus_per_node}\n'
                   f'#SBATCH --ntasks={gpus}\n'
                   f'#SBATCH --cpus-per-task=5\n\n'
                   f'{runner} -u {script_name} {config} '
                   f'--work-dir={work_dir} --cfg-option '
-                  f'dist_params.port={port} '
-                  f'checkpoint_config.max_keep_ckpts=10 '
+                  f'env_cfg.dist_cfg.port={port} '
+                  f'{" ".join(args.cfg_options)} '
+                  f'default_hooks.checkpoint.max_keep_ckpts=2 '
+                  f'default_hooks.checkpoint.save_best="auto" '
                   f'--launcher={launcher}\n')
 
     with open(work_dir / 'job.sh', 'w') as f:
@@ -124,33 +175,16 @@ def create_train_job_batch(commands, model_info, args, port, script_name):
     return work_dir / 'job.sh'
 
 
-def train(args):
-    models_cfg = load(str(Path(__file__).parent / 'bench_train.yml'))
-    models_cfg.build_models_with_collections()
-    models = {model.name: model for model in models_cfg.models}
-
+def train(models, args):
     script_name = osp.join('tools', 'train.py')
     port = args.port
 
     commands = []
-    if args.models:
-        patterns = [re.compile(pattern) for pattern in args.models]
-        filter_models = {}
-        for k, v in models.items():
-            if any([re.match(pattern, k) for pattern in patterns]):
-                filter_models[k] = v
-        if len(filter_models) == 0:
-            print('No model found, please specify models in:')
-            print('\n'.join(models.keys()))
-            return
-        models = filter_models
 
     for model_info in models.values():
-        months = model_info.data.get('Months', range(1, 13))
-        if datetime.now().month in months:
-            script_path = create_train_job_batch(commands, model_info, args,
-                                                 port, script_name)
-            port += 1
+        script_path = create_train_job_batch(commands, model_info, args, port,
+                                             script_name)
+        port += 1
 
     command_str = '\n'.join(commands)
 
@@ -245,12 +279,14 @@ def show_summary(summary_data):
                 metric = summary[metric_key]
                 expect = metric['expect']
                 last = metric['last']
+                last_epoch = metric['last_epoch']
                 last_color = set_color(last, expect)
                 best = metric['best']
                 best_color = set_color(best, expect)
                 best_epoch = metric['best_epoch']
                 row.append(f'{expect:.2f}')
-                row.append(f'[{last_color}]{last:.2f}[/{last_color}]')
+                row.append(
+                    f'[{last_color}]{last:.2f}[/{last_color}] ({last_epoch})')
                 row.append(
                     f'[{best_color}]{best:.2f}[/{best_color}] ({best_epoch})')
         table.add_row(*row)
@@ -258,25 +294,11 @@ def show_summary(summary_data):
     console.print(table)
 
 
-def summary(args):
-    models_cfg = load(str(Path(__file__).parent / 'bench_train.yml'))
-    models = {model.name: model for model in models_cfg.models}
+def summary(models, args):
 
     work_dir = Path(args.work_dir)
     dir_map = {p.name: p for p in work_dir.iterdir() if p.is_dir()}
 
-    if args.models:
-        patterns = [re.compile(pattern) for pattern in args.models]
-        filter_models = {}
-        for k, v in models.items():
-            if any([re.match(pattern, k) for pattern in patterns]):
-                filter_models[k] = v
-        if len(filter_models) == 0:
-            print('No model found, please specify models in:')
-            print('\n'.join(models.keys()))
-            return
-        models = filter_models
-
     summary_data = {}
     for model_name, model_info in models.items():
 
@@ -287,17 +309,19 @@ def summary(args):
 
         # Skip if not found any vis_data folder.
         sub_dir = dir_map[model_name]
-        vis_folders = [d for d in sub_dir.iterdir() if d.is_dir()]
-        if len(vis_folders) == 0:
-            continue
-        log_file = sorted(vis_folders)[-1] / 'vis_data' / 'scalars.json'
-        if not log_file.exists():
+        log_files = [f for f in sub_dir.glob('*/vis_data/scalars.json')]
+        if len(log_files) == 0:
             continue
+        log_file = sorted(log_files)[-1]
 
         # parse train log
         with open(log_file) as f:
             json_logs = [json.loads(s) for s in f.readlines()]
-            val_logs = [log for log in json_logs if 'loss' not in log]
+            val_logs = [
+                log for log in json_logs
+                # TODO: need a better method to extract validate log
+                if 'loss' not in log and 'accuracy/top1' in log
+            ]
 
         if len(val_logs) == 0:
             continue
@@ -320,9 +344,10 @@ def summary(args):
                 summary[key_yml] = dict(
                     expect=expect_result,
                     last=last,
+                    last_epoch=len(val_logs),
                     best=best,
-                    best_epoch=best_epoch)
-        summary_data[model_name] = summary
+                    best_epoch=best_epoch + 1)
+        summary_data[model_name].update(summary)
 
     show_summary(summary_data)
     if args.save:
@@ -332,10 +357,39 @@ def summary(args):
 def main():
     args = parse_args()
 
+    model_index_file = MMCLS_ROOT / 'model-index.yml'
+    model_index = load(str(model_index_file))
+    model_index.build_models_with_collections()
+    all_models = {model.name: model for model in model_index.models}
+
+    with open(Path(__file__).parent / 'bench_train.yml', 'r') as f:
+        train_items = yaml.safe_load(f)
+    models = OrderedDict()
+    for item in train_items:
+        name = item['Name']
+        model_info = all_models[name]
+        model_info.cycle = item.get('Cycle', None)
+        cycle = getattr(model_info, 'cycle', 'month')
+        cycle_level = CYCLE_LEVELS.index(cycle)
+        if cycle_level in args.range:
+            models[name] = model_info
+
+    if args.models:
+        patterns = [re.compile(pattern) for pattern in args.models]
+        filter_models = {}
+        for k, v in models.items():
+            if any([re.match(pattern, k) for pattern in patterns]):
+                filter_models[k] = v
+        if len(filter_models) == 0:
+            print('No model found, please specify models in:')
+            print('\n'.join(models.keys()))
+            return
+        models = filter_models
+
     if args.summary:
-        summary(args)
+        summary(models, args)
     else:
-        train(args)
+        train(models, args)
 
 
 if __name__ == '__main__':

.dev_scripts/benchmark_regression/bench_train.yml

@@ -1,88 +1,86 @@
-Models:
-  - Name: resnet50
-    Results:
-      - Dataset: ImageNet-1k
-        Metrics:
-          Top 1 Accuracy: 76.55
-          Top 5 Accuracy: 93.06
-    Config: configs/resnet/resnet50_8xb32_in1k.py
-    Gpus: 8
+- Name: mobilenet-v2_8xb32_in1k
+  Cycle: month
 
-  - Name: seresnet50
-    Results:
-      - Dataset: ImageNet-1k
-        Metrics:
-          Top 1 Accuracy: 77.74
-          Top 5 Accuracy: 93.84
-    Config: configs/seresnet/seresnet50_8xb32_in1k.py
-    Gpus: 8
+- Name: resnet50_8xb32_in1k
+  Cycle: month
 
-  - Name: vit-base
-    Results:
-      - Dataset: ImageNet-1k
-        Metrics:
-          Top 1 Accuracy: 82.37
-          Top 5 Accuracy: 96.15
-    Config: configs/vision_transformer/vit-base-p16_pt-32xb128-mae_in1k-224.py
-    Gpus: 32
+- Name: seresnet50_8xb32_in1k
+  Cycle: month
 
-  - Name: mobilenetv2
-    Results:
-      - Dataset: ImageNet-1k
-        Metrics:
-          Top 1 Accuracy: 71.86
-          Top 5 Accuracy: 90.42
-    Config: configs/mobilenet_v2/mobilenet-v2_8xb32_in1k.py
-    Gpus: 8
+- Name: swin-small_16xb64_in1k
+  Cycle: month
 
-  - Name: swin_tiny
-    Results:
-    - Dataset: ImageNet
-      Metrics:
-        Top 1 Accuracy: 81.18
-        Top 5 Accuracy: 95.61
-    Weights: https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_tiny_224_b16x64_300e_imagenet_20210616_090925-66df6be6.pth
-    Config: configs/swin_transformer/swin-tiny_16xb64_in1k.py
-    Gpus: 16
+- Name: vit-base-p16_pt-32xb128-mae_in1k
+  Cycle: month
 
-  - Name: vgg16
-    Results:
-      - Dataset: ImageNet-1k
-        Metrics:
-          Top 1 Accuracy: 71.62
-          Top 5 Accuracy: 90.49
-    Config: configs/vgg/vgg16_8xb32_in1k.py
-    Gpus: 8
-    Months:
-      - 1
-      - 4
-      - 7
-      - 10
+- Name: resnet50_8xb256-rsb-a1-600e_in1k
+  Cycle: quarter
 
-  - Name: shufflenet_v2
-    Results:
-      - Dataset: ImageNet-1k
-        Metrics:
-          Top 1 Accuracy: 69.55
-          Top 5 Accuracy: 88.92
-    Config: configs/shufflenet_v2/shufflenet-v2-1x_16xb64_in1k.py
-    Gpus: 16
-    Months:
-      - 2
-      - 5
-      - 8
-      - 11
+- Name: resnext50-32x4d_8xb32_in1k
+  Cycle: quarter
 
-  - Name: resnet-rsb
-    Results:
-      - Dataset: ImageNet-1k
-        Metrics:
-          Top 1 Accuracy: 80.12
-          Top 5 Accuracy: 94.78
-    Config: configs/resnet/resnet50_8xb256-rsb-a1-600e_in1k.py
-    Gpus: 8
-    Months:
-      - 3
-      - 6
-      - 9
-      - 12
+- Name: shufflenet-v2-1x_16xb64_in1k
+  Cycle: quarter
+
+- Name: vgg16_8xb32_in1k
+  Cycle: quarter
+
+- Name: shufflenet-v1-1x_16xb64_in1k
+  Cycle: half-year
+
+- Name: t2t-vit-t-14_8xb64_in1k
+  Cycle: half-year
+
+- Name: regnetx-1.6gf_8xb128_in1k
+  Cycle: half-year
+
+- Name: van-small_8xb128_in1k
+  Cycle: no-training
+
+- Name: res2net50-w14-s8_3rdparty_8xb32_in1k
+  Cycle: no-training
+
+- Name: repvgg-A2_3rdparty_4xb64-coslr-120e_in1k
+  Cycle: no-training
+
+- Name: tnt-small-p16_3rdparty_in1k
+  Cycle: no-training
+
+- Name: mlp-mixer-base-p16_3rdparty_64xb64_in1k
+  Cycle: no-training
+
+- Name: conformer-small-p16_3rdparty_8xb128_in1k
+  Cycle: no-training
+
+- Name: twins-pcpvt-base_3rdparty_8xb128_in1k
+  Cycle: no-training
+
+- Name: efficientnet-b0_3rdparty_8xb32_in1k
+  Cycle: no-training
+
+- Name: convnext-small_3rdparty_32xb128_in1k
+  Cycle: no-training
+
+- Name: hrnet-w18_3rdparty_8xb32_in1k
+  Cycle: no-training
+
+- Name: repmlp-base_3rdparty_8xb64_in1k
+  Cycle: no-training
+
+- Name: wide-resnet50_3rdparty_8xb32_in1k
+  Cycle: no-training
+
+- Name: cspresnet50_3rdparty_8xb32_in1k
+  Cycle: no-training
+
+- Name: convmixer-768-32_10xb64_in1k
+  Cycle: no-training
+
+- Name: densenet169_4xb256_in1k
+  Cycle: no-training
+
+- Name: poolformer-s24_3rdparty_32xb128_in1k
+  Cycle: no-training
+
+- Name: inception-v3_3rdparty_8xb32_in1k
+  Cycle: no-training

.github/workflows/lint.yml

@@ -0,0 +1,27 @@
+name: lint
+
+on: [push, pull_request]
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  lint:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v2
+      - name: Set up Python 3.7
+        uses: actions/setup-python@v2
+        with:
+          python-version: 3.7
+      - name: Install pre-commit hook
+        run: |
+          pip install pre-commit
+          pre-commit install
+      - name: Linting
+        run: pre-commit run --all-files
+      - name: Check docstring coverage
+        run: |
+          pip install interrogate
+          interrogate -v --ignore-init-method --ignore-module --ignore-nested-functions --ignore-magic --ignore-regex "__repr__" --fail-under 60 mmcls

.github/workflows/pr_stage_test.yml

@@ -0,0 +1,87 @@
+name: pr_stage_test
+
+on:
+  pull_request:
+    paths-ignore:
+      - 'README.md'
+      - 'README_zh-CN.md'
+      - 'docs/**'
+      - 'demo/**'
+      - 'tools/**'
+      - 'configs/**'
+      - '.dev_scripts/**'
+      - '.circleci/**'
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  build:
+    runs-on: ubuntu-18.04
+    strategy:
+      matrix:
+        python-version: [3.7]
+        include:
+          - torch: 1.8.1
+            torchvision: 0.9.1
+    steps:
+      - uses: actions/checkout@v2
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v2
+        with:
+          python-version: ${{ matrix.python-version }}
+      - name: Upgrade pip
+        run: pip install pip --upgrade
+      - name: Install PyTorch
+        run: pip install torch==${{matrix.torch}}+cpu torchvision==${{matrix.torchvision}}+cpu -f https://download.pytorch.org/whl/torch_stable.html
+      - name: Install mmcls dependencies
+        run: |
+          pip install git+https://github.com/open-mmlab/mmengine.git@main
+          pip install -U openmim
+          mim install 'mmcv >= 2.0.0rc1'
+          pip install -r requirements.txt
+      - name: Build and install
+        run: pip install -e .
+      - name: Run unittests and generate coverage report
+        run: |
+          coverage run --branch --source mmcls -m pytest tests/ -k 'not timm'
+          coverage xml
+          coverage report -m
+      # Upload coverage report for python3.7 && pytorch1.8.1 cpu
+      - name: Upload coverage to Codecov
+        uses: codecov/codecov-action@v1.0.14
+        with:
+          file: ./coverage.xml
+          flags: unittests
+          env_vars: OS,PYTHON
+          name: codecov-umbrella
+          fail_ci_if_error: false
+
+  build_windows:
+    runs-on: windows-2022
+    strategy:
+      matrix:
+        python: [3.7]
+        platform: [cu111]
+    steps:
+      - uses: actions/checkout@v2
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v2
+        with:
+          python-version: ${{ matrix.python-version }}
+      - name: Upgrade pip
+        run: pip install pip --upgrade
+      - name: Install PyTorch
+        run: pip install torch==1.8.2+${{matrix.platform}} torchvision==0.9.2+${{matrix.platform}} -f https://download.pytorch.org/whl/lts/1.8/torch_lts.html
+      - name: Install mmcls dependencies
+        run: |
+          pip install git+https://github.com/open-mmlab/mmengine.git@main
+          pip install -U openmim
+          mim install 'mmcv >= 2.0.0rc1'
+          pip install -r requirements.txt
+      - name: Build and install
+        run: pip install -e .
+      - name: Run unittests
+        run: |
+          pytest tests/ -k 'not timm' --ignore tests/test_models/test_backbones

.github/workflows/publish-to-pypi.yml

@@ -0,0 +1,22 @@
+name: deploy
+
+on: push
+
+jobs:
+  build-n-publish:
+    runs-on: ubuntu-latest
+    if: startsWith(github.event.ref, 'refs/tags')
+    steps:
+      - uses: actions/checkout@v2
+      - name: Set up Python 3.7
+        uses: actions/setup-python@v2
+        with:
+          python-version: 3.7
+      - name: Build MMClassification
+        run: |
+          pip install wheel
+          python setup.py sdist bdist_wheel
+      - name: Publish distribution to PyPI
+        run: |
+          pip install twine
+          twine upload dist/* -u __token__ -p ${{ secrets.pypi_password }}

.github/workflows/test_mim.yml

@@ -0,0 +1,44 @@
+name: test-mim
+
+on:
+  push:
+    paths:
+      - 'model-index.yml'
+      - 'configs/**'
+
+  pull_request:
+    paths:
+      - 'model-index.yml'
+      - 'configs/**'
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  build_cpu:
+    runs-on: ubuntu-18.04
+    strategy:
+      matrix:
+        python-version: [3.7]
+        torch: [1.8.0]
+        include:
+          - torch: 1.8.0
+            torch_version: torch1.8
+            torchvision: 0.9.0
+    steps:
+      - uses: actions/checkout@v2
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v2
+        with:
+          python-version: ${{ matrix.python-version }}
+      - name: Upgrade pip
+        run: pip install pip --upgrade
+      - name: Install PyTorch
+        run: pip install torch==${{matrix.torch}}+cpu torchvision==${{matrix.torchvision}}+cpu -f https://download.pytorch.org/whl/torch_stable.html
+      - name: Install openmim
+        run: pip install openmim
+      - name: Build and install
+        run: mim install -e .
+      - name: test commands of mim
+        run: mim search mmcls

.pre-commit-config.yaml

@@ -44,10 +44,10 @@ repos:
       - id: docformatter
         args: ["--in-place", "--wrap-descriptions", "79"]
   - repo: https://github.com/open-mmlab/pre-commit-hooks
-    rev: v0.2.0
+    rev: v0.4.0
     hooks:
     -   id: check-copyright
-        args: ["mmcls", "tests", "demo", "tools"]
+        args: ["mmcls", "tests", "demo", "tools", "--excludes", "mmcls/.mim/", "--ignore-file-not-found-error"]
   # - repo: local
   #   hooks:
   #     - id: clang-format

README.md

@@ -58,6 +58,11 @@ The `1.x` branch works with **PyTorch 1.6+**.
 
 ## What's new
 
+v1.0.0rc1 was released in 30/9/2022.
+
+- Support MViT, EdgeNeXt, Swin-Transformer V2, EfficientFormer and MobileOne.
+- Support BEiT type transformer layer.
+
 v1.0.0rc0 was released in 31/8/2022.
 
 This release introduced a brand new and flexible training & test engine, but it's still in progress. Welcome
@@ -115,6 +120,7 @@ Results and models are available in the [model zoo](https://mmclassification.rea
 - [x] [MobileNetV2](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/mobilenet_v2)
 - [x] [MobileNetV3](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/mobilenet_v3)
 - [x] [Swin-Transformer](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/swin_transformer)
+- [x] [Swin-Transformer V2](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/swin_transformer_v2)
 - [x] [RepVGG](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/repvgg)
 - [x] [Vision-Transformer](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/vision_transformer)
 - [x] [Transformer-in-Transformer](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/tnt)
@@ -125,6 +131,7 @@ Results and models are available in the [model zoo](https://mmclassification.rea
 - [x] [T2T-ViT](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/t2t_vit)
 - [x] [Twins](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/twins)
 - [x] [EfficientNet](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/efficientnet)
+- [x] [EdgeNeXt](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/edgenext)
 - [x] [ConvNeXt](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/convnext)
 - [x] [HRNet](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/hrnet)
 - [x] [VAN](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/van)
@@ -132,6 +139,9 @@ Results and models are available in the [model zoo](https://mmclassification.rea
 - [x] [CSPNet](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/cspnet)
 - [x] [PoolFormer](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/poolformer)
 - [x] [Inception V3](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/inception_v3)
+- [x] [MobileOne](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/mobileone)
+- [x] [EfficientFormer](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/efficientformer)
+- [x] [MViT](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/mvit)
 
 </details>
 

README_zh-CN.md

@@ -57,6 +57,11 @@ MMClassification 是一款基于 PyTorch 的开源图像分类工具箱，是 [O
 
 ## 更新日志
 
+2022/9/30 发布了 v1.0.0rc1 版本
+
+- 支持了 MViT，EdgeNeXt，Swin-Transformer V2，EfficientFormer，MobileOne 等主干网络。
+- 支持了 BEiT 风格的 transformer 层。
+
 2022/8/31 发布了 v1.0.0rc0 版本
 
 这个版本引入一个全新的，可扩展性强的训练和测试引擎，但目前仍在开发中。欢迎根据[文档](https://mmclassification.readthedocs.io/zh_CN/1.x/)进行试用。
@@ -114,6 +119,7 @@ mim install -e .
 - [x] [MobileNetV2](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/mobilenet_v2)
 - [x] [MobileNetV3](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/mobilenet_v3)
 - [x] [Swin-Transformer](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/swin_transformer)
+- [x] [Swin-Transformer V2](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/swin_transformer_v2)
 - [x] [RepVGG](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/repvgg)
 - [x] [Vision-Transformer](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/vision_transformer)
 - [x] [Transformer-in-Transformer](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/tnt)
@@ -124,6 +130,7 @@ mim install -e .
 - [x] [T2T-ViT](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/t2t_vit)
 - [x] [Twins](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/twins)
 - [x] [EfficientNet](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/efficientnet)
+- [x] [EdgeNeXt](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/edgenext)
 - [x] [ConvNeXt](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/convnext)
 - [x] [HRNet](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/hrnet)
 - [x] [VAN](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/van)
@@ -131,6 +138,9 @@ mim install -e .
 - [x] [CSPNet](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/cspnet)
 - [x] [PoolFormer](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/poolformer)
 - [x] [Inception V3](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/inception_v3)
+- [x] [MobileOne](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/mobileone)
+- [x] [EfficientFormer](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/efficientformer)
+- [x] [MViT](https://github.com/open-mmlab/mmclassification/tree/1.x/configs/mvit)
 
 </details>
 

configs/_base_/datasets/imagenet_bs64_edgenext_256.py

@@ -0,0 +1,83 @@
+# dataset settings
+dataset_type = 'ImageNet'
+data_preprocessor = dict(
+    # RGB format normalization parameters
+    mean=[123.675, 116.28, 103.53],
+    std=[58.395, 57.12, 57.375],
+    # convert image from BGR to RGB
+    to_rgb=True,
+)
+
+bgr_mean = data_preprocessor['mean'][::-1]
+bgr_std = data_preprocessor['std'][::-1]
+
+train_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(
+        type='RandomResizedCrop',
+        scale=256,
+        backend='pillow',
+        interpolation='bicubic'),
+    dict(type='RandomFlip', prob=0.5, direction='horizontal'),
+    dict(
+        type='RandAugment',
+        policies='timm_increasing',
+        num_policies=2,
+        total_level=10,
+        magnitude_level=9,
+        magnitude_std=0.5,
+        hparams=dict(
+            pad_val=[round(x) for x in bgr_mean], interpolation='bicubic')),
+    dict(
+        type='RandomErasing',
+        erase_prob=0.25,
+        mode='rand',
+        min_area_ratio=0.02,
+        max_area_ratio=1 / 3,
+        fill_color=bgr_mean,
+        fill_std=bgr_std),
+    dict(type='PackClsInputs'),
+]
+
+test_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(
+        type='ResizeEdge',
+        scale=292,
+        edge='short',
+        backend='pillow',
+        interpolation='bicubic'),
+    dict(type='CenterCrop', crop_size=256),
+    dict(type='PackClsInputs')
+]
+
+train_dataloader = dict(
+    batch_size=64,
+    num_workers=5,
+    dataset=dict(
+        type=dataset_type,
+        data_root='data/imagenet',
+        ann_file='meta/train.txt',
+        data_prefix='train',
+        pipeline=train_pipeline),
+    sampler=dict(type='DefaultSampler', shuffle=True),
+    persistent_workers=True,
+)
+
+val_dataloader = dict(
+    batch_size=64,
+    num_workers=5,
+    dataset=dict(
+        type=dataset_type,
+        data_root='data/imagenet',
+        ann_file='meta/val.txt',
+        data_prefix='val',
+        pipeline=test_pipeline),
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    persistent_workers=True,
+)
+val_evaluator = dict(type='Accuracy', topk=(1, 5))
+
+# If you want standard test, please manually configure the test dataset
+test_dataloader = val_dataloader
+test_evaluator = val_evaluator

configs/_base_/datasets/imagenet_bs64_swin_256.py

@@ -0,0 +1,83 @@
+# dataset settings
+dataset_type = 'ImageNet'
+data_preprocessor = dict(
+    # RGB format normalization parameters
+    mean=[123.675, 116.28, 103.53],
+    std=[58.395, 57.12, 57.375],
+    # convert image from BGR to RGB
+    to_rgb=True,
+)
+
+bgr_mean = data_preprocessor['mean'][::-1]
+bgr_std = data_preprocessor['std'][::-1]
+
+train_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(
+        type='RandomResizedCrop',
+        scale=256,
+        backend='pillow',
+        interpolation='bicubic'),
+    dict(type='RandomFlip', prob=0.5, direction='horizontal'),
+    dict(
+        type='RandAugment',
+        policies='timm_increasing',
+        num_policies=2,
+        total_level=10,
+        magnitude_level=9,
+        magnitude_std=0.5,
+        hparams=dict(
+            pad_val=[round(x) for x in bgr_mean], interpolation='bicubic')),
+    dict(
+        type='RandomErasing',
+        erase_prob=0.25,
+        mode='rand',
+        min_area_ratio=0.02,
+        max_area_ratio=1 / 3,
+        fill_color=bgr_mean,
+        fill_std=bgr_std),
+    dict(type='PackClsInputs'),
+]
+
+test_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(
+        type='ResizeEdge',
+        scale=292,  # ( 256 / 224 * 256 )
+        edge='short',
+        backend='pillow',
+        interpolation='bicubic'),
+    dict(type='CenterCrop', crop_size=256),
+    dict(type='PackClsInputs'),
+]
+
+train_dataloader = dict(
+    batch_size=64,
+    num_workers=5,
+    dataset=dict(
+        type=dataset_type,
+        data_root='data/imagenet',
+        ann_file='meta/train.txt',
+        data_prefix='train',
+        pipeline=train_pipeline),
+    sampler=dict(type='DefaultSampler', shuffle=True),
+    persistent_workers=True,
+)
+
+val_dataloader = dict(
+    batch_size=64,
+    num_workers=5,
+    dataset=dict(
+        type=dataset_type,
+        data_root='data/imagenet',
+        ann_file='meta/val.txt',
+        data_prefix='val',
+        pipeline=test_pipeline),
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    persistent_workers=True,
+)
+val_evaluator = dict(type='Accuracy', topk=(1, 5))
+
+# If you want standard test, please manually configure the test dataset
+test_dataloader = val_dataloader
+test_evaluator = val_evaluator

configs/_base_/models/edgenext/edgenext-base.py

@@ -0,0 +1,23 @@
+# Model settings
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='EdgeNeXt',
+        arch='base',
+        out_indices=(3, ),
+        drop_path_rate=0.1,
+        gap_before_final_norm=True,
+        init_cfg=[
+            dict(
+                type='TruncNormal',
+                layer=['Conv2d', 'Linear'],
+                std=.02,
+                bias=0.),
+            dict(type='Constant', layer=['LayerNorm'], val=1., bias=0.),
+        ]),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=584,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+    ))

configs/_base_/models/edgenext/edgenext-small.py

@@ -0,0 +1,23 @@
+# Model settings
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='EdgeNeXt',
+        arch='small',
+        out_indices=(3, ),
+        drop_path_rate=0.1,
+        gap_before_final_norm=True,
+        init_cfg=[
+            dict(
+                type='TruncNormal',
+                layer=['Conv2d', 'Linear'],
+                std=.02,
+                bias=0.),
+            dict(type='Constant', layer=['LayerNorm'], val=1., bias=0.),
+        ]),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=304,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+    ))

configs/_base_/models/edgenext/edgenext-xsmall.py

@@ -0,0 +1,23 @@
+# Model settings
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='EdgeNeXt',
+        arch='xsmall',
+        out_indices=(3, ),
+        drop_path_rate=0.1,
+        gap_before_final_norm=True,
+        init_cfg=[
+            dict(
+                type='TruncNormal',
+                layer=['Conv2d', 'Linear'],
+                std=.02,
+                bias=0.),
+            dict(type='Constant', layer=['LayerNorm'], val=1., bias=0.),
+        ]),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=192,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+    ))

configs/_base_/models/edgenext/edgenext-xxsmall.py

@@ -0,0 +1,23 @@
+# Model settings
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='EdgeNeXt',
+        arch='xxsmall',
+        out_indices=(3, ),
+        drop_path_rate=0.1,
+        gap_before_final_norm=True,
+        init_cfg=[
+            dict(
+                type='TruncNormal',
+                layer=['Conv2d', 'Linear'],
+                std=.02,
+                bias=0.),
+            dict(type='Constant', layer=['LayerNorm'], val=1., bias=0.),
+        ]),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=168,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+    ))

configs/_base_/models/efficientformer-l1.py

@@ -0,0 +1,18 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='EfficientFormer',
+        arch='l1',
+        drop_path_rate=0,
+        init_cfg=[
+            dict(
+                type='TruncNormal',
+                layer=['Conv2d', 'Linear'],
+                std=.02,
+                bias=0.),
+            dict(type='Constant', layer=['GroupNorm'], val=1., bias=0.),
+            dict(type='Constant', layer=['LayerScale'], val=1e-5)
+        ]),
+    neck=dict(type='GlobalAveragePooling', dim=1),
+    head=dict(
+        type='EfficientFormerClsHead', in_channels=448, num_classes=1000))

configs/_base_/models/mobileone/mobileone_s0.py

@@ -0,0 +1,19 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='MobileOne',
+        arch='s0',
+        out_indices=(3, ),
+    ),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=1024,
+        loss=dict(
+            type='LabelSmoothLoss',
+            label_smooth_val=0.1,
+            mode='original',
+        ),
+        topk=(1, 5),
+    ))

configs/_base_/models/mobileone/mobileone_s1.py

@@ -0,0 +1,19 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='MobileOne',
+        arch='s1',
+        out_indices=(3, ),
+    ),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=1280,
+        loss=dict(
+            type='LabelSmoothLoss',
+            label_smooth_val=0.1,
+            mode='original',
+        ),
+        topk=(1, 5),
+    ))

configs/_base_/models/mobileone/mobileone_s2.py

@@ -0,0 +1,19 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='MobileOne',
+        arch='s2',
+        out_indices=(3, ),
+    ),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=2048,
+        loss=dict(
+            type='LabelSmoothLoss',
+            label_smooth_val=0.1,
+            mode='original',
+        ),
+        topk=(1, 5),
+    ))

configs/_base_/models/mobileone/mobileone_s3.py

@@ -0,0 +1,19 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='MobileOne',
+        arch='s3',
+        out_indices=(3, ),
+    ),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=2048,
+        loss=dict(
+            type='LabelSmoothLoss',
+            label_smooth_val=0.1,
+            mode='original',
+        ),
+        topk=(1, 5),
+    ))

configs/_base_/models/mobileone/mobileone_s4.py

@@ -0,0 +1,19 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='MobileOne',
+        arch='s4',
+        out_indices=(3, ),
+    ),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=2048,
+        loss=dict(
+            type='LabelSmoothLoss',
+            label_smooth_val=0.1,
+            mode='original',
+        ),
+        topk=(1, 5),
+    ))

configs/_base_/models/mvit/mvitv2-base.py

@@ -0,0 +1,19 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(type='MViT', arch='base', drop_path_rate=0.3),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        in_channels=768,
+        num_classes=1000,
+        loss=dict(
+            type='LabelSmoothLoss', label_smooth_val=0.1, mode='original'),
+    ),
+    init_cfg=[
+        dict(type='TruncNormal', layer='Linear', std=0.02, bias=0.),
+        dict(type='Constant', layer='LayerNorm', val=1., bias=0.)
+    ],
+    train_cfg=dict(augments=[
+        dict(type='Mixup', alpha=0.8, num_classes=1000),
+        dict(type='CutMix', alpha=1.0, num_classes=1000)
+    ]))

configs/_base_/models/mvit/mvitv2-large.py

@@ -0,0 +1,23 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='MViT',
+        arch='large',
+        drop_path_rate=0.5,
+        dim_mul_in_attention=False),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        in_channels=1152,
+        num_classes=1000,
+        loss=dict(
+            type='LabelSmoothLoss', label_smooth_val=0.1, mode='original'),
+    ),
+    init_cfg=[
+        dict(type='TruncNormal', layer='Linear', std=0.02, bias=0.),
+        dict(type='Constant', layer='LayerNorm', val=1., bias=0.)
+    ],
+    train_cfg=dict(augments=[
+        dict(type='Mixup', alpha=0.8, num_classes=1000),
+        dict(type='CutMix', alpha=1.0, num_classes=1000)
+    ]))

configs/_base_/models/mvit/mvitv2-small.py

@@ -0,0 +1,19 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(type='MViT', arch='small', drop_path_rate=0.1),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        in_channels=768,
+        num_classes=1000,
+        loss=dict(
+            type='LabelSmoothLoss', label_smooth_val=0.1, mode='original'),
+    ),
+    init_cfg=[
+        dict(type='TruncNormal', layer='Linear', std=0.02, bias=0.),
+        dict(type='Constant', layer='LayerNorm', val=1., bias=0.)
+    ],
+    train_cfg=dict(augments=[
+        dict(type='Mixup', alpha=0.8, num_classes=1000),
+        dict(type='CutMix', alpha=1.0, num_classes=1000)
+    ]))

configs/_base_/models/mvit/mvitv2-tiny.py

@@ -0,0 +1,19 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(type='MViT', arch='tiny', drop_path_rate=0.1),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        in_channels=768,
+        num_classes=1000,
+        loss=dict(
+            type='LabelSmoothLoss', label_smooth_val=0.1, mode='original'),
+    ),
+    init_cfg=[
+        dict(type='TruncNormal', layer='Linear', std=0.02, bias=0.),
+        dict(type='Constant', layer='LayerNorm', val=1., bias=0.)
+    ],
+    train_cfg=dict(augments=[
+        dict(type='Mixup', alpha=0.8, num_classes=1000),
+        dict(type='CutMix', alpha=1.0, num_classes=1000)
+    ]))

configs/_base_/models/swin_transformer_v2/base_256.py

@@ -0,0 +1,26 @@
+# model settings
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='SwinTransformerV2',
+        arch='base',
+        img_size=256,
+        drop_path_rate=0.5),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=1024,
+        init_cfg=None,  # suppress the default init_cfg of LinearClsHead.
+        loss=dict(
+            type='LabelSmoothLoss', label_smooth_val=0.1, mode='original'),
+        cal_acc=False),
+    init_cfg=[
+        dict(type='TruncNormal', layer='Linear', std=0.02, bias=0.),
+        dict(type='Constant', layer='LayerNorm', val=1., bias=0.)
+    ],
+    train_cfg=dict(augments=[
+        dict(type='Mixup', alpha=0.8, num_classes=1000),
+        dict(type='CutMix', alpha=1.0, num_classes=1000)
+    ]),
+)

configs/_base_/models/swin_transformer_v2/base_384.py

@@ -0,0 +1,17 @@
+# model settings
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='SwinTransformerV2',
+        arch='base',
+        img_size=384,
+        drop_path_rate=0.2),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=1024,
+        init_cfg=None,  # suppress the default init_cfg of LinearClsHead.
+        loss=dict(
+            type='LabelSmoothLoss', label_smooth_val=0.1, mode='original'),
+        cal_acc=False))

configs/_base_/models/swin_transformer_v2/large_256.py

@@ -0,0 +1,16 @@
+# model settings
+# Only for evaluation
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='SwinTransformerV2',
+        arch='large',
+        img_size=256,
+        drop_path_rate=0.2),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=1536,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+        topk=(1, 5)))

configs/_base_/models/swin_transformer_v2/large_384.py

@@ -0,0 +1,16 @@
+# model settings
+# Only for evaluation
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='SwinTransformerV2',
+        arch='large',
+        img_size=384,
+        drop_path_rate=0.2),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=1536,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+        topk=(1, 5)))

configs/_base_/models/swin_transformer_v2/small_256.py

@@ -0,0 +1,26 @@
+# model settings
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='SwinTransformerV2',
+        arch='small',
+        img_size=256,
+        drop_path_rate=0.3),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=768,
+        init_cfg=None,  # suppress the default init_cfg of LinearClsHead.
+        loss=dict(
+            type='LabelSmoothLoss', label_smooth_val=0.1, mode='original'),
+        cal_acc=False),
+    init_cfg=[
+        dict(type='TruncNormal', layer='Linear', std=0.02, bias=0.),
+        dict(type='Constant', layer='LayerNorm', val=1., bias=0.)
+    ],
+    train_cfg=dict(augments=[
+        dict(type='Mixup', alpha=0.8, num_classes=1000),
+        dict(type='CutMix', alpha=1.0, num_classes=1000)
+    ]),
+)

configs/_base_/models/swin_transformer_v2/tiny_256.py

@@ -0,0 +1,26 @@
+# model settings
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='SwinTransformerV2',
+        arch='tiny',
+        img_size=256,
+        drop_path_rate=0.2),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=768,
+        init_cfg=None,  # suppress the default init_cfg of LinearClsHead.
+        loss=dict(
+            type='LabelSmoothLoss', label_smooth_val=0.1, mode='original'),
+        cal_acc=False),
+    init_cfg=[
+        dict(type='TruncNormal', layer='Linear', std=0.02, bias=0.),
+        dict(type='Constant', layer='LayerNorm', val=1., bias=0.)
+    ],
+    train_cfg=dict(augments=[
+        dict(type='Mixup', alpha=0.8, num_classes=1000),
+        dict(type='CutMix', alpha=1.0, num_classes=1000)
+    ]),
+)

configs/_base_/schedules/imagenet_bs1024_adamw_swin.py

@@ -23,18 +23,11 @@ param_scheduler = [
         type='LinearLR',
         start_factor=1e-3,
         by_epoch=True,
-        begin=0,
         end=20,
         # update by iter
         convert_to_iter_based=True),
     # main learning rate scheduler
-    dict(
-        type='CosineAnnealingLR',
-        T_max=280,
-        eta_min=1e-5,
-        by_epoch=True,
-        begin=20,
-        end=300)
+    dict(type='CosineAnnealingLR', eta_min=1e-5, by_epoch=True, begin=20)
 ]
 
 # train, val, test setting

configs/_base_/schedules/imagenet_bs1024_linearlr_bn_nowd.py

@@ -6,14 +6,8 @@ optim_wrapper = dict(
 
 # learning policy
 param_scheduler = [
-    dict(
-        type='ConstantLR',
-        factor=0.1,
-        by_epoch=True,
-        begin=0,
-        end=5,
-        convert_to_iter_based=True),
-    dict(type='PolyLR', eta_min=0, by_epoch=True, begin=5, end=300)
+    dict(type='ConstantLR', factor=0.1, by_epoch=False, begin=0, end=5000),
+    dict(type='PolyLR', eta_min=0, by_epoch=False, begin=5000)
 ]
 
 # train, val, test setting

configs/_base_/schedules/imagenet_bs2048_AdamW.py

@@ -11,25 +11,22 @@ optim_wrapper = dict(
 )
 
 # learning policy
+warmup_epochs = 15  # about 10000 iterations for ImageNet-1k
 param_scheduler = [
     # warm up learning rate scheduler
     dict(
         type='LinearLR',
         start_factor=1e-3,
         by_epoch=True,
-        begin=0,
-        # about 10000 iterations for ImageNet-1k
-        end=15,
+        end=warmup_epochs,
         # update by iter
         convert_to_iter_based=True),
     # main learning rate scheduler
     dict(
         type='CosineAnnealingLR',
-        T_max=285,
         eta_min=1e-5,
         by_epoch=True,
-        begin=15,
-        end=300)
+        begin=warmup_epochs)
 ]
 
 # train, val, test setting

configs/conformer/README.md

@@ -16,12 +16,12 @@ Within Convolutional Neural Network (CNN), the convolution operations are good a
 
 ### ImageNet-1k
 
-|         Model         | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                 Config                                  |                                 Download                                  |
-| :-------------------: | :-------: | :------: | :-------: | :-------: | :---------------------------------------------------------------------: | :-----------------------------------------------------------------------: |
-| Conformer-tiny-p16\*  |   23.52   |   4.90   |   81.31   |   95.60   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/conformer/conformer-tiny-p16_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/conformer/conformer-tiny-p16_3rdparty_8xb128_in1k_20211206-f6860372.pth) |
-| Conformer-small-p32\* |   38.85   |   7.09   |   81.96   |   96.02   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/conformer/conformer-small-p32_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/conformer/conformer-small-p32_8xb128_in1k_20211206-947a0816.pth) |
-| Conformer-small-p16\* |   37.67   |  10.31   |   83.32   |   96.46   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/conformer/conformer-small-p16_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/conformer/conformer-small-p16_3rdparty_8xb128_in1k_20211206-3065dcf5.pth) |
-| Conformer-base-p16\*  |   83.29   |  22.89   |   83.82   |   96.59   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/conformer/conformer-base-p16_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/conformer/conformer-base-p16_3rdparty_8xb128_in1k_20211206-bfdf8637.pth) |
+|         Model         | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                     Config                     |                                              Download                                              |
+| :-------------------: | :-------: | :------: | :-------: | :-------: | :--------------------------------------------: | :------------------------------------------------------------------------------------------------: |
+| Conformer-tiny-p16\*  |   23.52   |   4.90   |   81.31   |   95.60   | [config](./conformer-tiny-p16_8xb128_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/conformer/conformer-tiny-p16_3rdparty_8xb128_in1k_20211206-f6860372.pth) |
+| Conformer-small-p32\* |   38.85   |   7.09   |   81.96   |   96.02   | [config](./conformer-small-p32_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/conformer/conformer-small-p32_8xb128_in1k_20211206-947a0816.pth) |
+| Conformer-small-p16\* |   37.67   |  10.31   |   83.32   |   96.46   | [config](./conformer-small-p16_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/conformer/conformer-small-p16_3rdparty_8xb128_in1k_20211206-3065dcf5.pth) |
+| Conformer-base-p16\*  |   83.29   |  22.89   |   83.82   |   96.59   | [config](./conformer-base-p16_8xb128_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/conformer/conformer-base-p16_3rdparty_8xb128_in1k_20211206-bfdf8637.pth) |
 
 *Models with * are converted from the [official repo](https://github.com/pengzhiliang/Conformer). The config files of these models are only for validation. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/convmixer/README.md

@@ -20,11 +20,11 @@ Although convolutional networks have been the dominant architecture for vision t
 
 ### ImageNet-1k
 
-|        Model        | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                  Config                                  |                                  Download                                  |
-| :-----------------: | :-------: | :------: | :-------: | :-------: | :----------------------------------------------------------------------: | :------------------------------------------------------------------------: |
-| ConvMixer-768/32\*  |   21.11   |  19.62   |   80.16   |   95.08   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/convmixer/convmixer-768-32_10xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convmixer/convmixer-768-32_3rdparty_10xb64_in1k_20220323-bca1f7b8.pth) |
-| ConvMixer-1024/20\* |   24.38   |   5.55   |   76.94   |   93.36   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/convmixer/convmixer-1024-20_10xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convmixer/convmixer-1024-20_3rdparty_10xb64_in1k_20220323-48f8aeba.pth) |
-| ConvMixer-1536/20\* |   51.63   |  48.71   |   81.37   |   95.61   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/convmixer/convmixer-1536-20_10xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convmixer/convmixer-1536_20_3rdparty_10xb64_in1k_20220323-ea5786f3.pth) |
+|        Model        | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                    Config                    |                                                Download                                                |
+| :-----------------: | :-------: | :------: | :-------: | :-------: | :------------------------------------------: | :----------------------------------------------------------------------------------------------------: |
+| ConvMixer-768/32\*  |   21.11   |  19.62   |   80.16   |   95.08   | [config](./convmixer-768-32_10xb64_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/convmixer/convmixer-768-32_3rdparty_10xb64_in1k_20220323-bca1f7b8.pth) |
+| ConvMixer-1024/20\* |   24.38   |   5.55   |   76.94   |   93.36   | [config](./convmixer-1024-20_10xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convmixer/convmixer-1024-20_3rdparty_10xb64_in1k_20220323-48f8aeba.pth) |
+| ConvMixer-1536/20\* |   51.63   |  48.71   |   81.37   |   95.61   | [config](./convmixer-1536-20_10xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convmixer/convmixer-1536_20_3rdparty_10xb64_in1k_20220323-ea5786f3.pth) |
 
 *Models with * are converted from the [official repo](https://github.com/locuslab/convmixer). The config files of these models are only for inference. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/convmixer/metafile.yml

@@ -15,7 +15,7 @@ Models:
     Metadata:
       FLOPs: 19623051264
       Parameters: 21110248
-    In Collections: ConvMixer
+    In Collection: ConvMixer
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -31,7 +31,7 @@ Models:
     Metadata:
       FLOPs: 5550112768
       Parameters: 24383464
-    In Collections: ConvMixer
+    In Collection: ConvMixer
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -47,7 +47,7 @@ Models:
     Metadata:
       FLOPs: 48713170944
       Parameters: 51625960
-    In Collections: ConvMixer
+    In Collection: ConvMixer
     Results:
       - Dataset: ImageNet-1k
         Metrics:

configs/convnext/README.md

@@ -20,15 +20,15 @@ The "Roaring 20s" of visual recognition began with the introduction of Vision Tr
 
 ### ImageNet-1k
 
-|     Model     |   Pretrain   | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                Config                                 |                                Download                                 |
-| :-----------: | :----------: | :-------: | :------: | :-------: | :-------: | :-------------------------------------------------------------------: | :---------------------------------------------------------------------: |
-| ConvNeXt-T\*  | From scratch |   28.59   |   4.46   |   82.05   |   95.86   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/convnext/convnext-tiny_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-tiny_3rdparty_32xb128_in1k_20220124-18abde00.pth) |
-| ConvNeXt-S\*  | From scratch |   50.22   |   8.69   |   83.13   |   96.44   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/convnext/convnext-small_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-small_3rdparty_32xb128_in1k_20220124-d39b5192.pth) |
-| ConvNeXt-B\*  | From scratch |   88.59   |  15.36   |   83.85   |   96.74   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/convnext/convnext-base_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-base_3rdparty_32xb128_in1k_20220124-d0915162.pth) |
-| ConvNeXt-B\*  | ImageNet-21k |   88.59   |  15.36   |   85.81   |   97.86   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/convnext/convnext-base_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-base_in21k-pre-3rdparty_32xb128_in1k_20220124-eb2d6ada.pth) |
-| ConvNeXt-L\*  | From scratch |  197.77   |  34.37   |   84.30   |   96.89   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/convnext/convnext-large_64xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-large_3rdparty_64xb64_in1k_20220124-f8a0ded0.pth) |
-| ConvNeXt-L\*  | ImageNet-21k |  197.77   |  34.37   |   86.61   |   98.04   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/convnext/convnext-large_64xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-large_in21k-pre-3rdparty_64xb64_in1k_20220124-2412403d.pth) |
-| ConvNeXt-XL\* | ImageNet-21k |  350.20   |  60.93   |   86.97   |   98.20   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/convnext/convnext-xlarge_64xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-xlarge_in21k-pre-3rdparty_64xb64_in1k_20220124-76b6863d.pth) |
+|     Model     |   Pretrain   | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                   Config                   |                                              Download                                              |
+| :-----------: | :----------: | :-------: | :------: | :-------: | :-------: | :----------------------------------------: | :------------------------------------------------------------------------------------------------: |
+| ConvNeXt-T\*  | From scratch |   28.59   |   4.46   |   82.05   |   95.86   | [config](./convnext-tiny_32xb128_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-tiny_3rdparty_32xb128_in1k_20220124-18abde00.pth) |
+| ConvNeXt-S\*  | From scratch |   50.22   |   8.69   |   83.13   |   96.44   | [config](./convnext-small_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-small_3rdparty_32xb128_in1k_20220124-d39b5192.pth) |
+| ConvNeXt-B\*  | From scratch |   88.59   |  15.36   |   83.85   |   96.74   | [config](./convnext-base_32xb128_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-base_3rdparty_32xb128_in1k_20220124-d0915162.pth) |
+| ConvNeXt-B\*  | ImageNet-21k |   88.59   |  15.36   |   85.81   |   97.86   | [config](./convnext-base_32xb128_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-base_in21k-pre-3rdparty_32xb128_in1k_20220124-eb2d6ada.pth) |
+| ConvNeXt-L\*  | From scratch |  197.77   |  34.37   |   84.30   |   96.89   | [config](./convnext-large_64xb64_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-large_3rdparty_64xb64_in1k_20220124-f8a0ded0.pth) |
+| ConvNeXt-L\*  | ImageNet-21k |  197.77   |  34.37   |   86.61   |   98.04   | [config](./convnext-large_64xb64_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-large_in21k-pre-3rdparty_64xb64_in1k_20220124-2412403d.pth) |
+| ConvNeXt-XL\* | ImageNet-21k |  350.20   |  60.93   |   86.97   |   98.20   | [config](./convnext-xlarge_64xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/convnext/convnext-xlarge_in21k-pre-3rdparty_64xb64_in1k_20220124-76b6863d.pth) |
 
 *Models with * are converted from the [official repo](https://github.com/facebookresearch/ConvNeXt). The config files of these models are only for inference. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/convnext/metafile.yml

@@ -18,7 +18,7 @@ Models:
     Metadata:
       FLOPs: 4457472768
       Parameters: 28589128
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -35,7 +35,7 @@ Models:
       Training Data: ImageNet-1k
       FLOPs: 4457472768
       Parameters: 28589128
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -49,9 +49,10 @@ Models:
       Code: https://github.com/facebookresearch/ConvNeXt
   - Name: convnext-small_3rdparty_32xb128_in1k
     Metadata:
+      Training Data: ImageNet-1k
       FLOPs: 8687008512
       Parameters: 50223688
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -68,7 +69,7 @@ Models:
       Training Data: ImageNet-1k
       FLOPs: 8687008512
       Parameters: 50223688
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -82,9 +83,10 @@ Models:
       Code: https://github.com/facebookresearch/ConvNeXt
   - Name: convnext-base_3rdparty_32xb128_in1k
     Metadata:
+      Training Data: ImageNet-1k
       FLOPs: 15359124480
       Parameters: 88591464
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -101,7 +103,7 @@ Models:
       Training Data: ImageNet-1k
       FLOPs: 15359124480
       Parameters: 88591464
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -118,7 +120,7 @@ Models:
       Training Data: ImageNet-21k
       FLOPs: 15359124480
       Parameters: 88591464
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results: null
     Weights: https://download.openmmlab.com/mmclassification/v0/convnext/convnext-base_3rdparty_in21k_20220124-13b83eec.pth
     Converted From:
@@ -131,7 +133,7 @@ Models:
         - ImageNet-1k
       FLOPs: 15359124480
       Parameters: 88591464
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -145,9 +147,10 @@ Models:
       Code: https://github.com/facebookresearch/ConvNeXt
   - Name: convnext-large_3rdparty_64xb64_in1k
     Metadata:
+      Training Data: ImageNet-1k
       FLOPs: 34368026112
       Parameters: 197767336
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -164,7 +167,7 @@ Models:
       Training Data: ImageNet-21k
       FLOPs: 34368026112
       Parameters: 197767336
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results: null
     Weights: https://download.openmmlab.com/mmclassification/v0/convnext/convnext-large_3rdparty_in21k_20220124-41b5a79f.pth
     Converted From:
@@ -177,7 +180,7 @@ Models:
         - ImageNet-1k
       FLOPs: 34368026112
       Parameters: 197767336
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -194,7 +197,7 @@ Models:
       Training Data: ImageNet-21k
       FLOPs: 60929820672
       Parameters: 350196968
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results: null
     Weights: https://download.openmmlab.com/mmclassification/v0/convnext/convnext-xlarge_3rdparty_in21k_20220124-f909bad7.pth
     Converted From:
@@ -207,7 +210,7 @@ Models:
         - ImageNet-1k
       FLOPs: 60929820672
       Parameters: 350196968
-    In Collections: ConvNeXt
+    In Collection: ConvNeXt
     Results:
       - Dataset: ImageNet-1k
         Metrics:

configs/cspnet/README.md

@@ -20,11 +20,11 @@ Neural networks have enabled state-of-the-art approaches to achieve incredible r
 
 ### ImageNet-1k
 
-|     Model      |   Pretrain   | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                Config                                |                                Download                                 |
-| :------------: | :----------: | :-------: | :------: | :-------: | :-------: | :------------------------------------------------------------------: | :---------------------------------------------------------------------: |
-| CSPDarkNet50\* | From scratch |   27.64   |   5.04   |   80.05   |   95.07   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/cspnet/cspdarknet50_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/cspnet/cspdarknet50_3rdparty_8xb32_in1k_20220329-bd275287.pth) |
-| CSPResNet50\*  | From scratch |   21.62   |   3.48   |   79.55   |   94.68   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/cspnet/cspresnet50_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/cspnet/cspresnet50_3rdparty_8xb32_in1k_20220329-dd6dddfb.pth) |
-| CSPResNeXt50\* | From scratch |   20.57   |   3.11   |   79.96   |   94.96   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/cspnet/cspresnext50_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/cspnet/cspresnext50_3rdparty_8xb32_in1k_20220329-2cc84d21.pth) |
+|     Model      |   Pretrain   | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                 Config                 |                                               Download                                                |
+| :------------: | :----------: | :-------: | :------: | :-------: | :-------: | :------------------------------------: | :---------------------------------------------------------------------------------------------------: |
+| CSPDarkNet50\* | From scratch |   27.64   |   5.04   |   80.05   |   95.07   | [config](./cspdarknet50_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/cspnet/cspdarknet50_3rdparty_8xb32_in1k_20220329-bd275287.pth) |
+| CSPResNet50\*  | From scratch |   21.62   |   3.48   |   79.55   |   94.68   | [config](./cspresnet50_8xb32_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/cspnet/cspresnet50_3rdparty_8xb32_in1k_20220329-dd6dddfb.pth) |
+| CSPResNeXt50\* | From scratch |   20.57   |   3.11   |   79.96   |   94.96   | [config](./cspresnext50_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/cspnet/cspresnext50_3rdparty_8xb32_in1k_20220329-2cc84d21.pth) |
 
 *Models with * are converted from the [timm repo](https://github.com/rwightman/pytorch-image-models). The config files of these models are only for inference. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/cspnet/metafile.yml

@@ -17,7 +17,7 @@ Models:
     Metadata:
       FLOPs: 5040000000
       Parameters: 27640000
-    In Collections: CSPNet
+    In Collection: CSPNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -34,7 +34,7 @@ Models:
       Training Data: ImageNet-1k
       FLOPs: 3480000000
       Parameters: 21620000
-    In Collections: CSPNet
+    In Collection: CSPNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -50,7 +50,7 @@ Models:
     Metadata:
       FLOPs: 3110000000
       Parameters: 20570000
-    In Collections: CSPNet
+    In Collection: CSPNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:

configs/deit/README.md

@@ -18,17 +18,17 @@ Recently, neural networks purely based on attention were shown to address image
 
 The teacher of the distilled version DeiT is RegNetY-16GF.
 
-|            Model            |   Pretrain   | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                             Config                             |                             Download                             |
-| :-------------------------: | :----------: | :-------: | :------: | :-------: | :-------: | :------------------------------------------------------------: | :--------------------------------------------------------------: |
-|          DeiT-tiny          | From scratch |   5.72    |   1.08   |   74.50   |   92.24   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/deit/deit-tiny_pt-4xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-tiny_pt-4xb256_in1k_20220218-13b382a0.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/deit/deit-tiny_pt-4xb256_in1k_20220218-13b382a0.log.json) |
-|    DeiT-tiny distilled\*    | From scratch |   5.72    |   1.08   |   74.51   |   91.90   | [config](https://github.com/open-mmlab/mmclassification/tree/master/configs/deit/deit-tiny-distilled_pt-4xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-tiny-distilled_3rdparty_pt-4xb256_in1k_20211216-c429839a.pth) |
-|         DeiT-small          | From scratch |   22.05   |   4.24   |   80.69   |   95.06   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/deit/deit-small_pt-4xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-small_pt-4xb256_in1k_20220218-9425b9bb.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/deit/deit-small_pt-4xb256_in1k_20220218-9425b9bb.log.json) |
-|   DeiT-small distilled\*    | From scratch |   22.05   |   4.24   |   81.17   |   95.40   | [config](https://github.com/open-mmlab/mmclassification/tree/master/configs/deit/deit-small-distilled_pt-4xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-small-distilled_3rdparty_pt-4xb256_in1k_20211216-4de1d725.pth) |
-|          DeiT-base          | From scratch |   86.57   |  16.86   |   81.76   |   95.81   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/deit/deit-base_pt-16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-base_pt-16xb64_in1k_20220216-db63c16c.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/deit/deit-base_pt-16xb64_in1k_20220216-db63c16c.log.json) |
-|         DeiT-base\*         | From scratch |   86.57   |  16.86   |   81.79   |   95.59   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/deit/deit-base_pt-16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-base_3rdparty_pt-16xb64_in1k_20211124-6f40c188.pth) |
-|    DeiT-base distilled\*    | From scratch |   86.57   |  16.86   |   83.33   |   96.49   | [config](https://github.com/open-mmlab/mmclassification/tree/master/configs/deit/deit-base-distilled_pt-16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-base-distilled_3rdparty_pt-16xb64_in1k_20211216-42891296.pth) |
-|      DeiT-base 384px\*      | ImageNet-1k  |   86.86   |  49.37   |   83.04   |   96.31   | [config](https://github.com/open-mmlab/mmclassification/tree/master/configs/deit/deit-base_ft-16xb32_in1k-384px.py) | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-base_3rdparty_ft-16xb32_in1k-384px_20211124-822d02f2.pth) |
-| DeiT-base distilled 384px\* | ImageNet-1k  |   86.86   |  49.37   |   85.55   |   97.35   | [config](https://github.com/open-mmlab/mmclassification/tree/master/configs/deit/deit-base-distilled_ft-16xb32_in1k-384px.py) | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-base-distilled_3rdparty_ft-16xb32_in1k-384px_20211216-e48d6000.pth) |
+|            Model            |   Pretrain   | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                         Config                          |                                Download                                 |
+| :-------------------------: | :----------: | :-------: | :------: | :-------: | :-------: | :-----------------------------------------------------: | :---------------------------------------------------------------------: |
+|          DeiT-tiny          | From scratch |   5.72    |   1.08   |   74.50   |   92.24   |         [config](./deit-tiny_pt-4xb256_in1k.py)         | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-tiny_pt-4xb256_in1k_20220218-13b382a0.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/deit/deit-tiny_pt-4xb256_in1k_20220218-13b382a0.log.json) |
+|    DeiT-tiny distilled\*    | From scratch |   5.72    |   1.08   |   74.51   |   91.90   |    [config](./deit-tiny-distilled_pt-4xb256_in1k.py)    | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-tiny-distilled_3rdparty_pt-4xb256_in1k_20211216-c429839a.pth) |
+|         DeiT-small          | From scratch |   22.05   |   4.24   |   80.69   |   95.06   |        [config](./deit-small_pt-4xb256_in1k.py)         | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-small_pt-4xb256_in1k_20220218-9425b9bb.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/deit/deit-small_pt-4xb256_in1k_20220218-9425b9bb.log.json) |
+|   DeiT-small distilled\*    | From scratch |   22.05   |   4.24   |   81.17   |   95.40   |   [config](./deit-small-distilled_pt-4xb256_in1k.py)    | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-small-distilled_3rdparty_pt-4xb256_in1k_20211216-4de1d725.pth) |
+|          DeiT-base          | From scratch |   86.57   |  16.86   |   81.76   |   95.81   |         [config](./deit-base_pt-16xb64_in1k.py)         | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-base_pt-16xb64_in1k_20220216-db63c16c.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/deit/deit-base_pt-16xb64_in1k_20220216-db63c16c.log.json) |
+|         DeiT-base\*         | From scratch |   86.57   |  16.86   |   81.79   |   95.59   |         [config](./deit-base_pt-16xb64_in1k.py)         | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-base_3rdparty_pt-16xb64_in1k_20211124-6f40c188.pth) |
+|    DeiT-base distilled\*    | From scratch |   86.57   |  16.86   |   83.33   |   96.49   |    [config](./deit-base-distilled_pt-16xb64_in1k.py)    | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-base-distilled_3rdparty_pt-16xb64_in1k_20211216-42891296.pth) |
+|      DeiT-base 384px\*      | ImageNet-1k  |   86.86   |  49.37   |   83.04   |   96.31   |      [config](./deit-base_ft-16xb32_in1k-384px.py)      | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-base_3rdparty_ft-16xb32_in1k-384px_20211124-822d02f2.pth) |
+| DeiT-base distilled 384px\* | ImageNet-1k  |   86.86   |  49.37   |   85.55   |   97.35   | [config](./deit-base-distilled_ft-16xb32_in1k-384px.py) | [model](https://download.openmmlab.com/mmclassification/v0/deit/deit-base-distilled_3rdparty_ft-16xb32_in1k-384px_20211216-e48d6000.pth) |
 
 *Models with * are converted from the [official repo](https://github.com/facebookresearch/deit). The config files of these models are only for validation. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/densenet/README.md

@@ -16,12 +16,12 @@ Recent work has shown that convolutional networks can be substantially deeper, m
 
 ### ImageNet-1k
 
-|     Model     | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                   Config                                    |                                   Download                                    |
-| :-----------: | :-------: | :------: | :-------: | :-------: | :-------------------------------------------------------------------------: | :---------------------------------------------------------------------------: |
-| DenseNet121\* |   7.98    |   2.88   |   74.96   |   92.21   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/densenet/densenet121_4xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/densenet/densenet121_4xb256_in1k_20220426-07450f99.pth) |
-| DenseNet169\* |   14.15   |   3.42   |   76.08   |   93.11   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/densenet/densenet169_4xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/densenet/densenet169_4xb256_in1k_20220426-a2889902.pth) |
-| DenseNet201\* |   20.01   |   4.37   |   77.32   |   93.64   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/densenet/densenet201_4xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/densenet/densenet201_4xb256_in1k_20220426-05cae4ef.pth) |
-| DenseNet161\* |   28.68   |   7.82   |   77.61   |   93.83   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/densenet/densenet161_4xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/densenet/densenet161_4xb256_in1k_20220426-ee6a80a9.pth) |
+|     Model     | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                 Config                 |                                                      Download                                                      |
+| :-----------: | :-------: | :------: | :-------: | :-------: | :------------------------------------: | :----------------------------------------------------------------------------------------------------------------: |
+| DenseNet121\* |   7.98    |   2.88   |   74.96   |   92.21   | [config](./densenet121_4xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/densenet/densenet121_4xb256_in1k_20220426-07450f99.pth) |
+| DenseNet169\* |   14.15   |   3.42   |   76.08   |   93.11   | [config](./densenet169_4xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/densenet/densenet169_4xb256_in1k_20220426-a2889902.pth) |
+| DenseNet201\* |   20.01   |   4.37   |   77.32   |   93.64   | [config](./densenet201_4xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/densenet/densenet201_4xb256_in1k_20220426-05cae4ef.pth) |
+| DenseNet161\* |   28.68   |   7.82   |   77.61   |   93.83   | [config](./densenet161_4xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/densenet/densenet161_4xb256_in1k_20220426-ee6a80a9.pth) |
 
 *Models with * are converted from [pytorch](https://pytorch.org/vision/stable/models.html), guided by [original repo](https://github.com/liuzhuang13/DenseNet). The config files of these models are only for inference. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/densenet/metafile.yml

@@ -14,7 +14,7 @@ Models:
     Metadata:
       FLOPs: 2881695488
       Parameters: 7978856
-    In Collections: DenseNet
+    In Collection: DenseNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -30,7 +30,7 @@ Models:
     Metadata:
       FLOPs: 3416860160
       Parameters: 14149480
-    In Collections: DenseNet
+    In Collection: DenseNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -46,7 +46,7 @@ Models:
     Metadata:
       FLOPs: 4365236736
       Parameters: 20013928
-    In Collections: DenseNet
+    In Collection: DenseNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -62,7 +62,7 @@ Models:
     Metadata:
       FLOPs: 7816363968
       Parameters: 28681000
-    In Collections: DenseNet
+    In Collection: DenseNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:

configs/edgenext/README.md

@@ -0,0 +1,43 @@
+# EdgeNeXt
+
+> [EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications](https://arxiv.org/abs/2206.10589)
+
+<!-- [ALGORITHM] -->
+
+## Abstract
+
+<!-- [ABSTRACT] -->
+
+In the pursuit of achieving ever-increasing accuracy, large and complex neural networks are usually developed. Such models demand high computational resources and therefore cannot be deployed on edge devices. It is of great interest to build resource-efficient general purpose networks due to their usefulness in several application areas. In this work, we strive to effectively combine the strengths of both CNN and Transformer models and propose a new efficient hybrid architecture EdgeNeXt. Specifically in EdgeNeXt, we introduce split depth-wise transpose attention (SDTA) encoder that splits input tensors into multiple channel groups and utilizes depth-wise convolution along with self-attention across channel dimensions to implicitly increase the receptive field and encode multi-scale features. Our extensive experiments on classification, detection and segmentation tasks, reveal the merits of the proposed approach, outperforming state-of-the-art methods with comparatively lower compute requirements. Our EdgeNeXt model with 1.3M parameters achieves 71.2% top-1 accuracy on ImageNet-1K, outperforming MobileViT with an absolute gain of 2.2% with 28% reduction in FLOPs. Further, our EdgeNeXt model with 5.6M parameters achieves 79.4% top-1 accuracy on ImageNet-1K.
+
+<!-- [IMAGE] -->
+
+<div align=center>
+<img src="https://github.com/mmaaz60/EdgeNeXt/raw/main/images/EdgeNext.png" width="100%"/>
+</div>
+
+## Results and models
+
+### ImageNet-1k
+
+|        Model         |   Pretrain   | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                    Config                     |                                         Download                                         |
+| :------------------: | :----------: | :-------: | :------: | :-------: | :-------: | :-------------------------------------------: | :--------------------------------------------------------------------------------------: |
+| EdgeNeXt-Base-usi\*  | From scratch |   18.51   |   3.84   |   83.67   |   96.7    | [config](./edgenext-base_8xb256-usi_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/edgenext/edgenext-base_3rdparty-usi_in1k_20220801-909e8939.pth) |
+|   EdgeNeXt-Base\*    | From scratch |   18.51   |   3.84   |   82.48   |   96.2    |   [config](./edgenext-base_8xb256_in1k.py)    | [model](https://download.openmmlab.com/mmclassification/v0/edgenext/edgenext-base_3rdparty_in1k_20220801-9ade408b.pth) |
+| EdgeNeXt-Small-usi\* | From scratch |   5.59    |   1.26   |   81.06   |   95.34   | [config](./edgenext-small_8xb256-usi_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/edgenext/edgenext-small_3rdparty-usi_in1k_20220801-ae6d8dd3.pth) |
+|   EdgeNeXt-Small\*   | From scratch |   5.59    |   1.26   |   79.41   |   94.53   |   [config](./edgenext-small_8xb256_in1k.py)   | [model](https://download.openmmlab.com/mmclassification/v0/edgenext/edgenext-small_3rdparty_in1k_20220801-d00db5f8.pth) |
+|  EdgeNeXt-X-Small\*  | From scratch |   2.34    |  0.538   |   74.86   |   92.31   |  [config](./edgenext-xsmall_8xb256_in1k.py)   | [model](https://download.openmmlab.com/mmclassification/v0/edgenext/edgenext-xsmall_3rdparty_in1k_20220801-974f9fe7.pth) |
+| EdgeNeXt-XX-Small\*  | From scratch |   1.33    |  0.261   |   71.2    |   89.91   |  [config](./edgenext-xxsmall_8xb256_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/edgenext/edgenext-xxsmall_3rdparty_in1k_20220801-7ca8a81d.pth) |
+
+*Models with * are converted from the [official repo](https://github.com/mmaaz60/EdgeNeXt). The config files of these models are only for inference. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
+
+## Citation
+
+```bibtex
+@article{Maaz2022EdgeNeXt,
+    title={EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications},
+    author={Muhammad Maaz and Abdelrahman Shaker and Hisham Cholakkal and Salman Khan and Syed Waqas Zamir and Rao Muhammad Anwer and Fahad Shahbaz Khan},
+    journal={2206.10589},
+    year={2022}
+}
+```

configs/edgenext/edgenext-base_8xb256-usi_in1k.py

@@ -0,0 +1,19 @@
+_base_ = ['./edgenext-base_8xb256_in1k.py']
+
+# dataset setting
+
+test_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(
+        type='ResizeEdge',
+        scale=269,
+        edge='short',
+        backend='pillow',
+        interpolation='bicubic'),
+    dict(type='CenterCrop', crop_size=256),
+    dict(type='PackClsInputs')
+]
+
+val_dataloader = dict(dataset=dict(pipeline=test_pipeline))
+
+test_dataloader = val_dataloader

configs/edgenext/edgenext-base_8xb256_in1k.py

@@ -0,0 +1,20 @@
+_base_ = [
+    '../_base_/models/edgenext/edgenext-base.py',
+    '../_base_/datasets/imagenet_bs64_edgenext_256.py',
+    '../_base_/schedules/imagenet_bs1024_adamw_swin.py',
+    '../_base_/default_runtime.py',
+]
+
+# schedule setting
+optim_wrapper = dict(
+    optimizer=dict(lr=6e-3),
+    clip_grad=dict(max_norm=5.0),
+)
+
+# runtime setting
+custom_hooks = [dict(type='EMAHook', momentum=4e-5, priority='ABOVE_NORMAL')]
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR
+# based on the actual training batch size.
+# base_batch_size = (32 GPUs) x (128 samples per GPU)
+auto_scale_lr = dict(base_batch_size=4096)

configs/edgenext/edgenext-small_8xb256-usi_in1k.py

@@ -0,0 +1,19 @@
+_base_ = ['./edgenext-small_8xb256_in1k.py']
+
+# dataset setting
+
+test_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(
+        type='ResizeEdge',
+        scale=269,
+        edge='short',
+        backend='pillow',
+        interpolation='bicubic'),
+    dict(type='CenterCrop', crop_size=256),
+    dict(type='PackClsInputs')
+]
+
+val_dataloader = dict(dataset=dict(pipeline=test_pipeline))
+
+test_dataloader = val_dataloader

configs/edgenext/edgenext-small_8xb256_in1k.py

@@ -0,0 +1,20 @@
+_base_ = [
+    '../_base_/models/edgenext/edgenext-small.py',
+    '../_base_/datasets/imagenet_bs64_edgenext_256.py',
+    '../_base_/schedules/imagenet_bs1024_adamw_swin.py',
+    '../_base_/default_runtime.py',
+]
+
+# schedule setting
+optim_wrapper = dict(
+    optimizer=dict(lr=6e-3),
+    clip_grad=dict(max_norm=5.0),
+)
+
+# runtime setting
+custom_hooks = [dict(type='EMAHook', momentum=4e-5, priority='ABOVE_NORMAL')]
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR
+# based on the actual training batch size.
+# base_batch_size = (32 GPUs) x (128 samples per GPU)
+auto_scale_lr = dict(base_batch_size=4096)

configs/edgenext/edgenext-xsmall_8xb256_in1k.py

@@ -0,0 +1,20 @@
+_base_ = [
+    '../_base_/models/edgenext/edgenext-xsmall.py',
+    '../_base_/datasets/imagenet_bs64_edgenext_256.py',
+    '../_base_/schedules/imagenet_bs1024_adamw_swin.py',
+    '../_base_/default_runtime.py',
+]
+
+# schedule setting
+optim_wrapper = dict(
+    optimizer=dict(lr=6e-3),
+    clip_grad=dict(max_norm=5.0),
+)
+
+# runtime setting
+custom_hooks = [dict(type='EMAHook', momentum=4e-5, priority='ABOVE_NORMAL')]
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR
+# based on the actual training batch size.
+# base_batch_size = (32 GPUs) x (128 samples per GPU)
+auto_scale_lr = dict(base_batch_size=4096)

configs/edgenext/edgenext-xxsmall_8xb256_in1k.py

@@ -0,0 +1,20 @@
+_base_ = [
+    '../_base_/models/edgenext/edgenext-xxsmall.py',
+    '../_base_/datasets/imagenet_bs64_edgenext_256.py',
+    '../_base_/schedules/imagenet_bs1024_adamw_swin.py',
+    '../_base_/default_runtime.py',
+]
+
+# schedule setting
+optim_wrapper = dict(
+    optimizer=dict(lr=6e-3),
+    clip_grad=dict(max_norm=5.0),
+)
+
+# runtime setting
+custom_hooks = [dict(type='EMAHook', momentum=4e-5, priority='ABOVE_NORMAL')]
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR
+# based on the actual training batch size.
+# base_batch_size = (32 GPUs) x (128 samples per GPU)
+auto_scale_lr = dict(base_batch_size=4096)

configs/edgenext/metafile.yml

@@ -0,0 +1,118 @@
+Collections:
+  - Name: EdgeNeXt
+    Metadata:
+      Training Data: ImageNet-1k
+      Architecture:
+        - SDTA
+        - 1x1 Convolution
+        - Channel Self-attention
+    Paper:
+      URL: https://arxiv.org/abs/2206.10589
+      Title: 'EdgeNeXt: Efficiently Amalgamated CNN-Transformer Architecture for Mobile Vision Applications'
+    README: configs/edgenext/README.md
+    Code:
+      Version: v1.0.0rc1
+      URL: https://github.com/open-mmlab/mmclassification/blob/v0.23.2/mmcls/models/backbones/edgenext.py
+
+Models:
+  - Name: edgenext-xxsmall_3rdparty_in1k
+    Metadata:
+      FLOPs: 255640144
+      Parameters: 1327216
+    In Collection: EdgeNeXt
+    Results:
+      - Dataset: ImageNet-1k
+        Metrics:
+          Top 1 Accuracy: 71.20
+          Top 5 Accuracy: 89.91
+        Task: Image Classification
+    Weights: https://download.openmmlab.com/mmclassification/v0/edgenext/edgenext-xxsmall_3rdparty_in1k_20220801-7ca8a81d.pth
+    Config: configs/edgenext/edgenext-xxsmall_8xb256_in1k.py
+    Converted From:
+      Weights: https://github.com/mmaaz60/EdgeNeXt/releases/download/v1.0/edgenext_xxsmall.pth
+      Code: https://github.com/mmaaz60/EdgeNeXt
+  - Name: edgenext-xsmall_3rdparty_in1k
+    Metadata:
+      Training Data: ImageNet-1k
+      FLOPs: 529970560
+      Parameters: 2336804
+    In Collection: EdgeNeXt
+    Results:
+      - Dataset: ImageNet-1k
+        Metrics:
+          Top 1 Accuracy: 74.86
+          Top 5 Accuracy: 92.31
+        Task: Image Classification
+    Weights: https://download.openmmlab.com/mmclassification/v0/edgenext/edgenext-xsmall_3rdparty_in1k_20220801-974f9fe7.pth
+    Config: configs/edgenext/edgenext-xsmall_8xb256_in1k.py
+    Converted From:
+      Weights: https://github.com/mmaaz60/EdgeNeXt/releases/download/v1.0/edgenext_xsmall.pth
+      Code: https://github.com/mmaaz60/EdgeNeXt
+  - Name: edgenext-small_3rdparty_in1k
+    Metadata:
+      Training Data: ImageNet-1k
+      FLOPs: 1249788000
+      Parameters: 5586832
+    In Collection: EdgeNeXt
+    Results:
+      - Dataset: ImageNet-1k
+        Metrics:
+          Top 1 Accuracy: 79.41
+          Top 5 Accuracy: 94.53
+        Task: Image Classification
+    Weights: https://download.openmmlab.com/mmclassification/v0/edgenext/edgenext-small_3rdparty_in1k_20220801-d00db5f8.pth
+    Config: configs/edgenext/edgenext-small_8xb256_in1k.py
+    Converted From:
+      Weights: https://github.com/mmaaz60/EdgeNeXt/releases/download/v1.0/edgenext_small.pth
+      Code: https://github.com/mmaaz60/EdgeNeXt
+  - Name: edgenext-small-usi_3rdparty_in1k
+    Metadata:
+      Training Data: ImageNet-1k
+      FLOPs: 1249788000
+      Parameters: 5586832
+    In Collection: EdgeNeXt
+    Results:
+      - Dataset: ImageNet-1k
+        Metrics:
+          Top 1 Accuracy: 81.06
+          Top 5 Accuracy: 95.34
+        Task: Image Classification
+    Weights: https://download.openmmlab.com/mmclassification/v0/edgenext/edgenext-small_3rdparty-usi_in1k_20220801-ae6d8dd3.pth
+    Config: configs/edgenext/edgenext-small_8xb256-usi_in1k.py
+    Converted From:
+      Weights: https://github.com/mmaaz60/EdgeNeXt/releases/download/v1.1/edgenext_small_usi.pth
+      Code: https://github.com/mmaaz60/EdgeNeXt
+  - Name: edgenext-base_3rdparty_in1k
+    Metadata:
+      Training Data: ImageNet-1k
+      FLOPs: 3814395280
+      Parameters: 18511292
+    In Collection: EdgeNeXt
+    Results:
+      - Dataset: ImageNet-1k
+        Metrics:
+          Top 1 Accuracy: 82.48
+          Top 5 Accuracy: 96.2
+        Task: Image Classification
+    Weights: https://download.openmmlab.com/mmclassification/v0/edgenext/edgenext-base_3rdparty_in1k_20220801-9ade408b.pth
+    Config: configs/edgenext/edgenext-base_8xb256_in1k.py
+    Converted From:
+      Weights: https://github.com/mmaaz60/EdgeNeXt/releases/download/v1.2/edgenext_base.pth
+      Code: https://github.com/mmaaz60/EdgeNeXt
+  - Name: edgenext-base_3rdparty-usi_in1k
+    Metadata:
+      Training Data: ImageNet-1k
+      FLOPs: 3814395280
+      Parameters: 18511292
+    In Collection: EdgeNeXt
+    Results:
+      - Dataset: ImageNet-1k
+        Metrics:
+          Top 1 Accuracy: 83.67
+          Top 5 Accuracy: 96.7
+        Task: Image Classification
+    Weights: https://download.openmmlab.com/mmclassification/v0/edgenext/edgenext-base_3rdparty-usi_in1k_20220801-909e8939.pth
+    Config: configs/edgenext/edgenext-base_8xb256-usi_in1k.py
+    Converted From:
+      Weights: https://github.com/mmaaz60/EdgeNeXt/releases/download/v1.2/edgenext_base_usi.pth
+      Code: https://github.com/mmaaz60/EdgeNeXt

configs/efficientformer/README.md

@@ -0,0 +1,47 @@
+# EfficientFormer
+
+> [EfficientFormer: Vision Transformers at MobileNet Speed](https://arxiv.org/abs/2206.01191)
+
+<!-- [ALGORITHM] -->
+
+## Abstract
+
+Vision Transformers (ViT) have shown rapid progress in computer vision tasks, achieving promising results on various benchmarks. However, due to the massive number of parameters and model design, e.g., attention mechanism, ViT-based models are generally times slower than lightweight convolutional networks. Therefore, the deployment of ViT for real-time applications is particularly challenging, especially on resource-constrained hardware such as mobile devices. Recent efforts try to reduce the computation complexity of ViT through network architecture search or hybrid design with MobileNet block, yet the inference speed is still unsatisfactory. This leads to an important question: can transformers run as fast as MobileNet while obtaining high performance? To answer this, we first revisit the network architecture and operators used in ViT-based models and identify inefficient designs. Then we introduce a dimension-consistent pure transformer (without MobileNet blocks) as a design paradigm.  Finally, we perform latency-driven slimming to get a series of final models dubbed EfficientFormer. Extensive experiments show the superiority of EfficientFormer in performance and speed on mobile devices. Our fastest model, EfficientFormer-L1, achieves 79.2% top-1 accuracy on ImageNet-1K with only 1.6 ms inference latency on iPhone 12 (compiled with CoreML), which runs as fast as MobileNetV2×1.4 (1.6 ms, 74.7% top-1), and our largest model, EfficientFormer-L7, obtains 83.3% accuracy with only 7.0 ms latency. Our work proves that properly designed transformers can reach extremely low latency on mobile devices while maintaining high performance.
+
+<div align=center>
+<img src="https://user-images.githubusercontent.com/18586273/180713426-9d3d77e3-3584-42d8-9098-625b4170d796.png" width="100%"/>
+</div>
+
+## Results and models
+
+### ImageNet-1k
+
+|        Model         | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                    Config                     |                                               Download                                               |
+| :------------------: | :-------: | :------: | :-------: | :-------: | :-------------------------------------------: | :--------------------------------------------------------------------------------------------------: |
+| EfficientFormer-l1\* |   12.19   |   1.30   |   80.46   |   94.99   | [config](./efficientformer-l1_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientformer/efficientformer-l1_3rdparty_in1k_20220915-cc3e1ac6.pth) |
+| EfficientFormer-l3\* |   31.41   |   3.93   |   82.45   |   96.18   | [config](./efficientformer-l3_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientformer/efficientformer-l3_3rdparty_in1k_20220915-466793d6.pth) |
+| EfficientFormer-l7\* |   82.23   |  10.16   |   83.40   |   96.60   | [config](./efficientformer-l7_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientformer/efficientformer-l7_3rdparty_in1k_20220915-185e30af.pth) |
+
+*Models with * are converted from the [official repo](https://github.com/snap-research/EfficientFormer). The config files of these models are only for inference. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
+
+## Citation
+
+```bibtex
+@misc{https://doi.org/10.48550/arxiv.2206.01191,
+  doi = {10.48550/ARXIV.2206.01191},
+
+  url = {https://arxiv.org/abs/2206.01191},
+
+  author = {Li, Yanyu and Yuan, Geng and Wen, Yang and Hu, Eric and Evangelidis, Georgios and Tulyakov, Sergey and Wang, Yanzhi and Ren, Jian},
+
+  keywords = {Computer Vision and Pattern Recognition (cs.CV), FOS: Computer and information sciences, FOS: Computer and information sciences},
+
+  title = {EfficientFormer: Vision Transformers at MobileNet Speed},
+
+  publisher = {arXiv},
+
+  year = {2022},
+
+  copyright = {Creative Commons Attribution 4.0 International}
+}
+```

configs/efficientformer/efficientformer-l1_8xb128_in1k.py

@@ -0,0 +1,6 @@
+_base_ = [
+    '../_base_/models/efficientformer-l1.py',
+    '../_base_/datasets/imagenet_bs128_poolformer_small_224.py',
+    '../_base_/schedules/imagenet_bs1024_adamw_swin.py',
+    '../_base_/default_runtime.py',
+]

configs/efficientformer/efficientformer-l3_8xb128_in1k.py

@@ -0,0 +1,3 @@
+_base_ = './efficientformer-l1_8xb128_in1k.py'
+
+model = dict(backbone=dict(arch='l3'), head=dict(in_channels=512))

configs/efficientformer/efficientformer-l7_8xb128_in1k.py

@@ -0,0 +1,3 @@
+_base_ = './efficientformer-l1_8xb128_in1k.py'
+
+model = dict(backbone=dict(arch='l7'), head=dict(in_channels=768))

configs/efficientformer/metafile.yml

@@ -0,0 +1,67 @@
+Collections:
+  - Name: EfficientFormer
+    Metadata:
+      Training Data: ImageNet-1k
+      Architecture:
+        - Pooling
+        - 1x1 Convolution
+        - LayerScale
+        - MetaFormer
+    Paper:
+      URL: https://arxiv.org/pdf/2206.01191.pdf
+      Title: "EfficientFormer: Vision Transformers at MobileNet Speed"
+    README: configs/efficientformer/README.md
+    Code:
+      Version: v1.0.0rc1
+      URL: https://github.com/open-mmlab/mmclassification/blob/v1.0.0rc1/configs/efficientformer/metafile.yml
+
+Models:
+  - Name: efficientformer-l1_3rdparty_8xb128_in1k
+    Metadata:
+      FLOPs: 1304601088     # 1.3G
+      Parameters: 12278696   # 12M
+    In Collection: EfficientFormer
+    Results:
+      - Dataset: ImageNet-1k
+        Metrics:
+          Top 1 Accuracy: 80.46
+          Top 5 Accuracy: 94.99
+        Task: Image Classification
+    Weights: https://download.openmmlab.com/mmclassification/v0/efficientformer/efficientformer-l1_3rdparty_in1k_20220915-cc3e1ac6.pth
+    Config: configs/efficientformer/efficientformer-l1_8xb128_in1k.py
+    Converted From:
+      Weights: https://drive.google.com/file/d/11SbX-3cfqTOc247xKYubrAjBiUmr818y/view?usp=sharing
+      Code: https://github.com/snap-research/EfficientFormer
+  - Name: efficientformer-l3_3rdparty_8xb128_in1k
+    Metadata:
+      Training Data: ImageNet-1k
+      FLOPs: 3737045760        # 3.7G
+      Parameters: 31406000     # 31M
+    In Collection: EfficientFormer
+    Results:
+      - Dataset: ImageNet-1k
+        Metrics:
+          Top 1 Accuracy: 82.45
+          Top 5 Accuracy: 96.18
+        Task: Image Classification
+    Weights: https://download.openmmlab.com/mmclassification/v0/efficientformer/efficientformer-l3_3rdparty_in1k_20220915-466793d6.pth
+    Config: configs/efficientformer/efficientformer-l3_8xb128_in1k.py
+    Converted From:
+      Weights: https://drive.google.com/file/d/1OyyjKKxDyMj-BcfInp4GlDdwLu3hc30m/view?usp=sharing
+      Code: https://github.com/snap-research/EfficientFormer
+  - Name: efficientformer-l7_3rdparty_8xb128_in1k
+    Metadata:
+      FLOPs: 10163951616      # 10.2G
+      Parameters: 82229328    # 82M
+    In Collection: EfficientFormer
+    Results:
+      - Dataset: ImageNet-1k
+        Metrics:
+          Top 1 Accuracy: 83.40
+          Top 5 Accuracy: 96.60
+        Task: Image Classification
+    Weights: https://download.openmmlab.com/mmclassification/v0/efficientformer/efficientformer-l7_3rdparty_in1k_20220915-185e30af.pth
+    Config: configs/efficientformer/efficientformer-l7_8xb128_in1k.py
+    Converted From:
+      Weights: https://drive.google.com/file/d/1cVw-pctJwgvGafeouynqWWCwgkcoFMM5/view?usp=sharing
+      Code: https://github.com/snap-research/EfficientFormer

configs/efficientnet/README.md

@@ -20,31 +20,31 @@ In the result table, AA means trained with AutoAugment pre-processing, more deta
 
 Note: In MMClassification, we support training with AutoAugment, don't support AdvProp by now.
 
-|              Model               | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                              Config                               |                               Download                               |
-| :------------------------------: | :-------: | :------: | :-------: | :-------: | :---------------------------------------------------------------: | :------------------------------------------------------------------: |
-|        EfficientNet-B0\*         |   5.29    |   0.02   |   76.74   |   93.17   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b0_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b0_3rdparty_8xb32_in1k_20220119-a7e2a0b1.pth) |
-|      EfficientNet-B0 (AA)\*      |   5.29    |   0.02   |   77.26   |   93.41   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b0_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b0_3rdparty_8xb32-aa_in1k_20220119-8d939117.pth) |
-| EfficientNet-B0 (AA + AdvProp)\* |   5.29    |   0.02   |   77.53   |   93.61   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b0_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b0_3rdparty_8xb32-aa-advprop_in1k_20220119-26434485.pth) |
-|        EfficientNet-B1\*         |   7.79    |   0.03   |   78.68   |   94.28   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b1_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b1_3rdparty_8xb32_in1k_20220119-002556d9.pth) |
-|      EfficientNet-B1 (AA)\*      |   7.79    |   0.03   |   79.20   |   94.42   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b1_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b1_3rdparty_8xb32-aa_in1k_20220119-619d8ae3.pth) |
-| EfficientNet-B1 (AA + AdvProp)\* |   7.79    |   0.03   |   79.52   |   94.43   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b1_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b1_3rdparty_8xb32-aa-advprop_in1k_20220119-5715267d.pth) |
-|        EfficientNet-B2\*         |   9.11    |   0.03   |   79.64   |   94.80   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b2_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b2_3rdparty_8xb32_in1k_20220119-ea374a30.pth) |
-|      EfficientNet-B2 (AA)\*      |   9.11    |   0.03   |   80.21   |   94.96   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b2_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b2_3rdparty_8xb32-aa_in1k_20220119-dd61e80b.pth) |
-| EfficientNet-B2 (AA + AdvProp)\* |   9.11    |   0.03   |   80.45   |   95.07   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b2_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b2_3rdparty_8xb32-aa-advprop_in1k_20220119-1655338a.pth) |
-|        EfficientNet-B3\*         |   12.23   |   0.06   |   81.01   |   95.34   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b3_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b3_3rdparty_8xb32_in1k_20220119-4b4d7487.pth) |
-|      EfficientNet-B3 (AA)\*      |   12.23   |   0.06   |   81.58   |   95.67   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b3_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b3_3rdparty_8xb32-aa_in1k_20220119-5b4887a0.pth) |
-| EfficientNet-B3 (AA + AdvProp)\* |   12.23   |   0.06   |   81.81   |   95.69   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b3_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b3_3rdparty_8xb32-aa-advprop_in1k_20220119-53b41118.pth) |
-|        EfficientNet-B4\*         |   19.34   |   0.12   |   82.57   |   96.09   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b4_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b4_3rdparty_8xb32_in1k_20220119-81fd4077.pth) |
-|      EfficientNet-B4 (AA)\*      |   19.34   |   0.12   |   82.95   |   96.26   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b4_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b4_3rdparty_8xb32-aa_in1k_20220119-45b8bd2b.pth) |
-| EfficientNet-B4 (AA + AdvProp)\* |   19.34   |   0.12   |   83.25   |   96.44   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b4_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b4_3rdparty_8xb32-aa-advprop_in1k_20220119-38c2238c.pth) |
-|        EfficientNet-B5\*         |   30.39   |   0.24   |   83.18   |   96.47   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b5_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b5_3rdparty_8xb32_in1k_20220119-e9814430.pth) |
-|      EfficientNet-B5 (AA)\*      |   30.39   |   0.24   |   83.82   |   96.76   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b5_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b5_3rdparty_8xb32-aa_in1k_20220119-2cab8b78.pth) |
-| EfficientNet-B5 (AA + AdvProp)\* |   30.39   |   0.24   |   84.21   |   96.98   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b5_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b5_3rdparty_8xb32-aa-advprop_in1k_20220119-f57a895a.pth) |
-|      EfficientNet-B6 (AA)\*      |   43.04   |   0.41   |   84.05   |   96.82   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b6_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b6_3rdparty_8xb32-aa_in1k_20220119-45b03310.pth) |
-| EfficientNet-B6 (AA + AdvProp)\* |   43.04   |   0.41   |   84.74   |   97.14   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b6_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b6_3rdparty_8xb32-aa-advprop_in1k_20220119-bfe3485e.pth) |
-|      EfficientNet-B7 (AA)\*      |   66.35   |   0.72   |   84.38   |   96.88   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b7_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b7_3rdparty_8xb32-aa_in1k_20220119-bf03951c.pth) |
-| EfficientNet-B7 (AA + AdvProp)\* |   66.35   |   0.72   |   85.14   |   97.23   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b7_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b7_3rdparty_8xb32-aa-advprop_in1k_20220119-c6dbff10.pth) |
-| EfficientNet-B8 (AA + AdvProp)\* |   87.41   |   1.09   |   85.38   |   97.28   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/efficientnet/efficientnet-b8_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b8_3rdparty_8xb32-aa-advprop_in1k_20220119-297ce1b7.pth) |
+|              Model               | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                      Config                      |                                       Download                                        |
+| :------------------------------: | :-------: | :------: | :-------: | :-------: | :----------------------------------------------: | :-----------------------------------------------------------------------------------: |
+|        EfficientNet-B0\*         |   5.29    |   0.02   |   76.74   |   93.17   |    [config](./efficientnet-b0_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b0_3rdparty_8xb32_in1k_20220119-a7e2a0b1.pth) |
+|      EfficientNet-B0 (AA)\*      |   5.29    |   0.02   |   77.26   |   93.41   |    [config](./efficientnet-b0_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b0_3rdparty_8xb32-aa_in1k_20220119-8d939117.pth) |
+| EfficientNet-B0 (AA + AdvProp)\* |   5.29    |   0.02   |   77.53   |   93.61   | [config](./efficientnet-b0_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b0_3rdparty_8xb32-aa-advprop_in1k_20220119-26434485.pth) |
+|        EfficientNet-B1\*         |   7.79    |   0.03   |   78.68   |   94.28   |    [config](./efficientnet-b1_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b1_3rdparty_8xb32_in1k_20220119-002556d9.pth) |
+|      EfficientNet-B1 (AA)\*      |   7.79    |   0.03   |   79.20   |   94.42   |    [config](./efficientnet-b1_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b1_3rdparty_8xb32-aa_in1k_20220119-619d8ae3.pth) |
+| EfficientNet-B1 (AA + AdvProp)\* |   7.79    |   0.03   |   79.52   |   94.43   | [config](./efficientnet-b1_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b1_3rdparty_8xb32-aa-advprop_in1k_20220119-5715267d.pth) |
+|        EfficientNet-B2\*         |   9.11    |   0.03   |   79.64   |   94.80   |    [config](./efficientnet-b2_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b2_3rdparty_8xb32_in1k_20220119-ea374a30.pth) |
+|      EfficientNet-B2 (AA)\*      |   9.11    |   0.03   |   80.21   |   94.96   |    [config](./efficientnet-b2_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b2_3rdparty_8xb32-aa_in1k_20220119-dd61e80b.pth) |
+| EfficientNet-B2 (AA + AdvProp)\* |   9.11    |   0.03   |   80.45   |   95.07   | [config](./efficientnet-b2_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b2_3rdparty_8xb32-aa-advprop_in1k_20220119-1655338a.pth) |
+|        EfficientNet-B3\*         |   12.23   |   0.06   |   81.01   |   95.34   |    [config](./efficientnet-b3_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b3_3rdparty_8xb32_in1k_20220119-4b4d7487.pth) |
+|      EfficientNet-B3 (AA)\*      |   12.23   |   0.06   |   81.58   |   95.67   |    [config](./efficientnet-b3_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b3_3rdparty_8xb32-aa_in1k_20220119-5b4887a0.pth) |
+| EfficientNet-B3 (AA + AdvProp)\* |   12.23   |   0.06   |   81.81   |   95.69   | [config](./efficientnet-b3_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b3_3rdparty_8xb32-aa-advprop_in1k_20220119-53b41118.pth) |
+|        EfficientNet-B4\*         |   19.34   |   0.12   |   82.57   |   96.09   |    [config](./efficientnet-b4_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b4_3rdparty_8xb32_in1k_20220119-81fd4077.pth) |
+|      EfficientNet-B4 (AA)\*      |   19.34   |   0.12   |   82.95   |   96.26   |    [config](./efficientnet-b4_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b4_3rdparty_8xb32-aa_in1k_20220119-45b8bd2b.pth) |
+| EfficientNet-B4 (AA + AdvProp)\* |   19.34   |   0.12   |   83.25   |   96.44   | [config](./efficientnet-b4_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b4_3rdparty_8xb32-aa-advprop_in1k_20220119-38c2238c.pth) |
+|        EfficientNet-B5\*         |   30.39   |   0.24   |   83.18   |   96.47   |    [config](./efficientnet-b5_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b5_3rdparty_8xb32_in1k_20220119-e9814430.pth) |
+|      EfficientNet-B5 (AA)\*      |   30.39   |   0.24   |   83.82   |   96.76   |    [config](./efficientnet-b5_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b5_3rdparty_8xb32-aa_in1k_20220119-2cab8b78.pth) |
+| EfficientNet-B5 (AA + AdvProp)\* |   30.39   |   0.24   |   84.21   |   96.98   | [config](./efficientnet-b5_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b5_3rdparty_8xb32-aa-advprop_in1k_20220119-f57a895a.pth) |
+|      EfficientNet-B6 (AA)\*      |   43.04   |   0.41   |   84.05   |   96.82   |    [config](./efficientnet-b6_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b6_3rdparty_8xb32-aa_in1k_20220119-45b03310.pth) |
+| EfficientNet-B6 (AA + AdvProp)\* |   43.04   |   0.41   |   84.74   |   97.14   | [config](./efficientnet-b6_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b6_3rdparty_8xb32-aa-advprop_in1k_20220119-bfe3485e.pth) |
+|      EfficientNet-B7 (AA)\*      |   66.35   |   0.72   |   84.38   |   96.88   |    [config](./efficientnet-b7_8xb32_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b7_3rdparty_8xb32-aa_in1k_20220119-bf03951c.pth) |
+| EfficientNet-B7 (AA + AdvProp)\* |   66.35   |   0.72   |   85.14   |   97.23   | [config](./efficientnet-b7_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b7_3rdparty_8xb32-aa-advprop_in1k_20220119-c6dbff10.pth) |
+| EfficientNet-B8 (AA + AdvProp)\* |   87.41   |   1.09   |   85.38   |   97.28   | [config](./efficientnet-b8_8xb32-01norm_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/efficientnet/efficientnet-b8_3rdparty_8xb32-aa-advprop_in1k_20220119-297ce1b7.pth) |
 
 *Models with * are converted from the [official repo](https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet). The config files of these models are only for inference. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/efficientnet/metafile.yml

@@ -25,7 +25,7 @@ Models:
     Metadata:
       FLOPs: 16481180
       Parameters: 5288548
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -41,7 +41,7 @@ Models:
     Metadata:
       FLOPs: 16481180
       Parameters: 5288548
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -57,7 +57,7 @@ Models:
     Metadata:
       FLOPs: 16481180
       Parameters: 5288548
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -73,7 +73,7 @@ Models:
     Metadata:
       FLOPs: 27052224
       Parameters: 7794184
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -89,7 +89,7 @@ Models:
     Metadata:
       FLOPs: 27052224
       Parameters: 7794184
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -105,7 +105,7 @@ Models:
     Metadata:
       FLOPs: 27052224
       Parameters: 7794184
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -121,7 +121,7 @@ Models:
     Metadata:
       FLOPs: 34346386
       Parameters: 9109994
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -137,7 +137,7 @@ Models:
     Metadata:
       FLOPs: 34346386
       Parameters: 9109994
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -153,7 +153,7 @@ Models:
     Metadata:
       FLOPs: 34346386
       Parameters: 9109994
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -169,7 +169,7 @@ Models:
     Metadata:
       FLOPs: 58641904
       Parameters: 12233232
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -185,7 +185,7 @@ Models:
     Metadata:
       FLOPs: 58641904
       Parameters: 12233232
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -201,7 +201,7 @@ Models:
     Metadata:
       FLOPs: 58641904
       Parameters: 12233232
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -217,7 +217,7 @@ Models:
     Metadata:
       FLOPs: 121870624
       Parameters: 19341616
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -233,7 +233,7 @@ Models:
     Metadata:
       FLOPs: 121870624
       Parameters: 19341616
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -249,7 +249,7 @@ Models:
     Metadata:
       FLOPs: 121870624
       Parameters: 19341616
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -265,7 +265,7 @@ Models:
     Metadata:
       FLOPs: 243879440
       Parameters: 30389784
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -281,7 +281,7 @@ Models:
     Metadata:
       FLOPs: 243879440
       Parameters: 30389784
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -297,7 +297,7 @@ Models:
     Metadata:
       FLOPs: 243879440
       Parameters: 30389784
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -313,7 +313,7 @@ Models:
     Metadata:
       FLOPs: 412002408
       Parameters: 43040704
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -329,7 +329,7 @@ Models:
     Metadata:
       FLOPs: 412002408
       Parameters: 43040704
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -345,7 +345,7 @@ Models:
     Metadata:
       FLOPs: 715526512
       Parameters: 66347960
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -361,7 +361,7 @@ Models:
     Metadata:
       FLOPs: 715526512
       Parameters: 66347960
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -377,7 +377,7 @@ Models:
     Metadata:
       FLOPs: 1092755326
       Parameters: 87413142
-    In Collections: EfficientNet
+    In Collection: EfficientNet
     Results:
       - Dataset: ImageNet-1k
         Metrics:

configs/hrnet/README.md

@@ -16,17 +16,17 @@ High-resolution representations are essential for position-sensitive vision prob
 
 ## ImageNet-1k
 
-|       Model        | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                  Config                                  |                                  Download                                   |
-| :----------------: | :-------: | :------: | :-------: | :-------: | :----------------------------------------------------------------------: | :-------------------------------------------------------------------------: |
-|    HRNet-W18\*     |   21.30   |   4.33   |   76.75   |   93.44   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/hrnet/hrnet-w18_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w18_3rdparty_8xb32_in1k_20220120-0c10b180.pth) |
-|    HRNet-W30\*     |   37.71   |   8.17   |   78.19   |   94.22   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/hrnet/hrnet-w30_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w30_3rdparty_8xb32_in1k_20220120-8aa3832f.pth) |
-|    HRNet-W32\*     |   41.23   |   8.99   |   78.44   |   94.19   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/hrnet/hrnet-w32_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w32_3rdparty_8xb32_in1k_20220120-c394f1ab.pth) |
-|    HRNet-W40\*     |   57.55   |  12.77   |   78.94   |   94.47   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/hrnet/hrnet-w40_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w40_3rdparty_8xb32_in1k_20220120-9a2dbfc5.pth) |
-|    HRNet-W44\*     |   67.06   |  14.96   |   78.88   |   94.37   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/hrnet/hrnet-w44_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w44_3rdparty_8xb32_in1k_20220120-35d07f73.pth) |
-|    HRNet-W48\*     |   77.47   |  17.36   |   79.32   |   94.52   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/hrnet/hrnet-w48_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w48_3rdparty_8xb32_in1k_20220120-e555ef50.pth) |
-|    HRNet-W64\*     |  128.06   |  29.00   |   79.46   |   94.65   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/hrnet/hrnet-w64_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w64_3rdparty_8xb32_in1k_20220120-19126642.pth) |
-| HRNet-W18 (ssld)\* |   21.30   |   4.33   |   81.06   |   95.70   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/hrnet/hrnet-w18_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w18_3rdparty_8xb32-ssld_in1k_20220120-455f69ea.pth) |
-| HRNet-W48 (ssld)\* |   77.47   |  17.36   |   83.63   |   96.79   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/hrnet/hrnet-w48_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w48_3rdparty_8xb32-ssld_in1k_20220120-d0459c38.pth) |
+|       Model        | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |               Config                |                                                     Download                                                     |
+| :----------------: | :-------: | :------: | :-------: | :-------: | :---------------------------------: | :--------------------------------------------------------------------------------------------------------------: |
+|    HRNet-W18\*     |   21.30   |   4.33   |   76.75   |   93.44   | [config](./hrnet-w18_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w18_3rdparty_8xb32_in1k_20220120-0c10b180.pth) |
+|    HRNet-W30\*     |   37.71   |   8.17   |   78.19   |   94.22   | [config](./hrnet-w30_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w30_3rdparty_8xb32_in1k_20220120-8aa3832f.pth) |
+|    HRNet-W32\*     |   41.23   |   8.99   |   78.44   |   94.19   | [config](./hrnet-w32_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w32_3rdparty_8xb32_in1k_20220120-c394f1ab.pth) |
+|    HRNet-W40\*     |   57.55   |  12.77   |   78.94   |   94.47   | [config](./hrnet-w40_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w40_3rdparty_8xb32_in1k_20220120-9a2dbfc5.pth) |
+|    HRNet-W44\*     |   67.06   |  14.96   |   78.88   |   94.37   | [config](./hrnet-w44_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w44_3rdparty_8xb32_in1k_20220120-35d07f73.pth) |
+|    HRNet-W48\*     |   77.47   |  17.36   |   79.32   |   94.52   | [config](./hrnet-w48_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w48_3rdparty_8xb32_in1k_20220120-e555ef50.pth) |
+|    HRNet-W64\*     |  128.06   |  29.00   |   79.46   |   94.65   | [config](./hrnet-w64_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w64_3rdparty_8xb32_in1k_20220120-19126642.pth) |
+| HRNet-W18 (ssld)\* |   21.30   |   4.33   |   81.06   |   95.70   | [config](./hrnet-w18_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w18_3rdparty_8xb32-ssld_in1k_20220120-455f69ea.pth) |
+| HRNet-W48 (ssld)\* |   77.47   |  17.36   |   83.63   |   96.79   | [config](./hrnet-w48_4xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/hrnet/hrnet-w48_3rdparty_8xb32-ssld_in1k_20220120-d0459c38.pth) |
 
 *Models with * are converted from the [official repo](https://github.com/HRNet/HRNet-Image-Classification). The config files of these models are only for inference. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/inception_v3/README.md

@@ -16,9 +16,9 @@ Convolutional networks are at the core of most state-of-the-art computer vision
 
 ### ImageNet-1k
 
-|     Model      | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                   Config                                   |                                   Download                                    |
-| :------------: | :-------: | :------: | :-------: | :-------: | :------------------------------------------------------------------------: | :---------------------------------------------------------------------------: |
-| Inception V3\* |   23.83   |   5.75   |   77.57   |   93.58   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/inception_v3/inception-v3_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/inception-v3/inception-v3_3rdparty_8xb32_in1k_20220615-dcd4d910.pth) |
+|     Model      | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                 Config                 |                                                     Download                                                      |
+| :------------: | :-------: | :------: | :-------: | :-------: | :------------------------------------: | :---------------------------------------------------------------------------------------------------------------: |
+| Inception V3\* |   23.83   |   5.75   |   77.57   |   93.58   | [config](./inception-v3_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/inception-v3/inception-v3_3rdparty_8xb32_in1k_20220615-dcd4d910.pth) |
 
 *Models with * are converted from the [official repo](https://github.com/pytorch/vision/blob/main/torchvision/models/inception.py#L28). The config files of these models are only for inference. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/inception_v3/metafile.yml

@@ -15,8 +15,8 @@ Collections:
       Title: "Rethinking the Inception Architecture for Computer Vision"
     README: configs/inception_v3/README.md
     Code:
-      URL: TODO
-      Version: TODO
+      URL: https://github.com/open-mmlab/mmclassification/blob/v1.0.0rc1/configs/inception_v3/metafile.yml
+      Version: v1.0.0rc1
 
 Models:
   - Name: inception-v3_3rdparty_8xb32_in1k

configs/mlp_mixer/README.md

@@ -16,10 +16,10 @@ Convolutional Neural Networks (CNNs) are the go-to model for computer vision. Re
 
 ### ImageNet-1k
 
-|    Model     | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                   Config                                    |                                    Download                                    |
-| :----------: | :-------: | :------: | :-------: | :-------: | :-------------------------------------------------------------------------: | :----------------------------------------------------------------------------: |
-| Mixer-B/16\* |   59.88   |  12.61   |   76.68   |   92.25   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/mlp_mixer/mlp-mixer-base-p16_64xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mlp-mixer/mixer-base-p16_3rdparty_64xb64_in1k_20211124-1377e3e0.pth) |
-| Mixer-L/16\* |   208.2   |  44.57   |   72.34   |   88.02   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/mlp_mixer/mlp-mixer-large-p16_64xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mlp-mixer/mixer-large-p16_3rdparty_64xb64_in1k_20211124-5a2519d2.pth) |
+|    Model     | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                     Config                     |                                                  Download                                                   |
+| :----------: | :-------: | :------: | :-------: | :-------: | :--------------------------------------------: | :---------------------------------------------------------------------------------------------------------: |
+| Mixer-B/16\* |   59.88   |  12.61   |   76.68   |   92.25   | [config](./mlp-mixer-base-p16_64xb64_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/mlp-mixer/mixer-base-p16_3rdparty_64xb64_in1k_20211124-1377e3e0.pth) |
+| Mixer-L/16\* |   208.2   |  44.57   |   72.34   |   88.02   | [config](./mlp-mixer-large-p16_64xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mlp-mixer/mixer-large-p16_3rdparty_64xb64_in1k_20211124-5a2519d2.pth) |
 
 *Models with * are converted from [timm](https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/mlp_mixer.py). The config files of these models are only for validation. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/mobilenet_v2/README.md

@@ -18,9 +18,9 @@ The MobileNetV2 architecture is based on an inverted residual structure where th
 
 ### ImageNet-1k
 
-|    Model     | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                   Config                                    |                                    Download                                    |
-| :----------: | :-------: | :------: | :-------: | :-------: | :-------------------------------------------------------------------------: | :----------------------------------------------------------------------------: |
-| MobileNet V2 |    3.5    |  0.319   |   71.86   |   90.42   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/mobilenet_v2/mobilenet-v2_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mobilenet_v2/mobilenet_v2_batch256_imagenet_20200708-3b2dc3af.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/mobilenet_v2/mobilenet_v2_batch256_imagenet_20200708-3b2dc3af.log.json) |
+|    Model     | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                 Config                 |                                                      Download                                                       |
+| :----------: | :-------: | :------: | :-------: | :-------: | :------------------------------------: | :-----------------------------------------------------------------------------------------------------------------: |
+| MobileNet V2 |    3.5    |  0.319   |   71.86   |   90.42   | [config](./mobilenet-v2_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mobilenet_v2/mobilenet_v2_batch256_imagenet_20200708-3b2dc3af.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/mobilenet_v2/mobilenet_v2_batch256_imagenet_20200708-3b2dc3af.log.json) |
 
 ## Citation
 

configs/mobilenet_v3/README.md

@@ -16,10 +16,10 @@ We present the next generation of MobileNets based on a combination of complemen
 
 ### ImageNet-1k
 
-|        Model        | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                  Config                                  |                                  Download                                  |
-| :-----------------: | :-------: | :------: | :-------: | :-------: | :----------------------------------------------------------------------: | :------------------------------------------------------------------------: |
-| MobileNetV3-Small\* |   2.54    |   0.06   |   67.66   |   87.41   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/mobilenet_v3/mobilenet-v3-small_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mobilenet_v3/convert/mobilenet_v3_small-8427ecf0.pth) |
-| MobileNetV3-Large\* |   5.48    |   0.23   |   74.04   |   91.34   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/mobilenet_v3/mobilenet-v3-large_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mobilenet_v3/convert/mobilenet_v3_large-3ea3c186.pth) |
+|        Model        | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                    Config                     |                                               Download                                                |
+| :-----------------: | :-------: | :------: | :-------: | :-------: | :-------------------------------------------: | :---------------------------------------------------------------------------------------------------: |
+| MobileNetV3-Small\* |   2.54    |   0.06   |   67.66   |   87.41   | [config](./mobilenet-v3-small_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mobilenet_v3/convert/mobilenet_v3_small-8427ecf0.pth) |
+| MobileNetV3-Large\* |   5.48    |   0.23   |   74.04   |   91.34   | [config](./mobilenet-v3-large_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mobilenet_v3/convert/mobilenet_v3_large-3ea3c186.pth) |
 
 *Models with * are converted from [torchvision](https://pytorch.org/vision/stable/_modules/torchvision/models/mobilenetv3.html). The config files of these models are only for validation. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/mobileone/README.md

@@ -0,0 +1,132 @@
+# MobileOne
+
+> [An Improved One millisecond Mobile Backbone](https://arxiv.org/abs/2206.04040)
+
+<!-- [ALGORITHM] -->
+
+## Abstract
+
+Efficient neural network backbones for mobile devices are often optimized for metrics such as FLOPs or parameter count. However, these metrics may not correlate well with latency of the network when deployed on a mobile device. Therefore, we perform extensive analysis of different metrics by deploying several mobile-friendly networks on a mobile device. We identify and analyze architectural and optimization bottlenecks in recent efficient neural networks and provide ways to mitigate these bottlenecks. To this end, we design an efficient backbone MobileOne, with variants achieving an inference time under 1 ms on an iPhone12 with 75.9% top-1 accuracy on ImageNet. We show that MobileOne achieves state-of-the-art performance within the efficient architectures while being many times faster on mobile. Our best model obtains similar performance on ImageNet as MobileFormer while being 38x faster. Our model obtains 2.3% better top-1 accuracy on ImageNet than EfficientNet at similar latency. Furthermore, we show that our model generalizes to multiple tasks - image classification, object detection, and semantic segmentation with significant improvements in latency and accuracy as compared to existing efficient architectures when deployed on a mobile device.
+
+<div align=center>
+<img src="https://user-images.githubusercontent.com/18586273/183552452-74657532-f461-48f7-9aa7-c23f006cdb07.png" width="40%"/>
+</div>
+
+## Results and models
+
+### ImageNet-1k
+
+|     Model      |            Params(M)            |            Flops(G)            | Top-1 (%) | Top-5 (%) |                        Config                        |                        Download                         |
+| :------------: | :-----------------------------: | :----------------------------: | :-------: | :-------: | :--------------------------------------------------: | :-----------------------------------------------------: |
+| MobileOne-s0\* |  5.29（train) \| 2.08 (deploy)  | 1.09 (train) \| 0.28 (deploy)  |   71.36   |   89.87   | [config (train)](./mobileone-s0_8xb128_in1k.py) \| [config (deploy)](./deploy/mobileone-s0_deploy_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mobileone/mobileone-s0_3rdparty_in1k_20220915-007ae971.pth) |
+| MobileOne-s1\* |  4.83 (train) \| 4.76 (deploy)  | 0.86 (train) \| 0.84 (deploy)  |   75.76   |   92.77   | [config (train)](./mobileone-s1_8xb128_in1k.py) \| [config (deploy)](./deploy/mobileone-s1_deploy_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mobileone/mobileone-s1_3rdparty_in1k_20220915-473c8469.pth) |
+| MobileOne-s2\* |  7.88 (train) \| 7.88 (deploy)  | 1.34 (train)  \| 1.31 (deploy) |   77.39   |   93.63   | [config (train)](./mobileone-s2_8xb128_in1k.py) \|[config (deploy)](./deploy/mobileone-s2_deploy_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mobileone/mobileone-s2_3rdparty_in1k_20220915-ed2e4c30.pth) |
+| MobileOne-s3\* | 10.17 (train) \| 10.08 (deploy) | 1.95 (train)  \| 1.91 (deploy) |   77.93   |   93.89   | [config (train)](./mobileone-s3_8xb128_in1k.py) \|[config (deploy)](./deploy/mobileone-s3_deploy_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mobileone/mobileone-s3_3rdparty_in1k_20220915-84d6a02c.pth) |
+| MobileOne-s4\* | 14.95 (train) \| 14.84 (deploy) | 3.05 (train) \| 3.00 (deploy)  |   79.30   |   94.37   | [config (train)](./mobileone-s4_8xb128_in1k.py) \|[config (deploy)](./deploy/mobileone-s4_deploy_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mobileone/mobileone-s4_3rdparty_in1k_20220915-ce9509ee.pth) |
+
+*Models with * are converted from the [official repo](https://github.com/apple/ml-mobileone). The config files of these models are only for validation. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
+
+*Because the [official repo.](https://github.com/apple/ml-mobileone) does not give a strategy for training and testing, the test data pipline of [RepVGG](https://github.com/open-mmlab/mmclassification/tree/master/configs/repvgg) is used here, and the result is about 0.1 lower than that in the paper. Refer to [this issue](https://github.com/apple/ml-mobileone/issues/2).*
+
+## How to use
+
+The checkpoints provided are all `training-time` models. Use the reparameterize tool to switch them to more efficient `inference-time` architecture, which not only has fewer parameters but also less calculations.
+
+### Use tool
+
+Use provided tool to reparameterize the given model and save the checkpoint:
+
+```bash
+python tools/convert_models/reparameterize_model.py ${CFG_PATH} ${SRC_CKPT_PATH} ${TARGET_CKPT_PATH}
+```
+
+`${CFG_PATH}` is the config file path, `${SRC_CKPT_PATH}` is the source chenpoint file path, `${TARGET_CKPT_PATH}` is the target deploy weight file path.
+
+For example:
+
+```shell
+python ./tools/convert_models/reparameterize_model.py ./configs/mobileone/mobileone-s0_8xb128_in1k.py https://download.openmmlab.com/mmclassification/v0/mobileone/mobileone-s0_3rdparty_in1k_20220811-db5ce29b.pth ./mobileone_s0_deploy.pth
+```
+
+To use reparameterized weights, the config file must switch to **the deploy config files**.
+
+```bash
+python tools/test.py ${Deploy_CFG} ${Deploy_Checkpoint} --metrics accuracy
+```
+
+For example of using the reparameterized weights above:
+
+```shell
+python ./tools/test.py ./configs/mobileone/deploy/mobileone-s0_deploy_8xb128_in1k.py  mobileone_s0_deploy.pth --metrics accuracy
+```
+
+### In the code
+
+Use the API `switch_to_deploy` of `MobileOne` backbone to to switch to the deploy mode. Usually called like `backbone.switch_to_deploy()` or `classificer.backbone.switch_to_deploy()`.
+
+For Backbones:
+
+```python
+from mmcls.models import build_backbone
+import torch
+
+x = torch.randn( (1, 3, 224, 224) )
+backbone_cfg=dict(type='MobileOne', arch='s0')
+backbone = build_backbone(backbone_cfg)
+backbone.init_weights()
+backbone.eval()
+outs_ori = backbone(x)
+
+backbone.switch_to_deploy()
+outs_dep = backbone(x)
+
+for out1, out2 in zip(outs_ori, outs_dep):
+    assert torch.allclose(out1, out2)
+```
+
+For ImageClassifiers:
+
+```python
+from mmcls.models import build_classifier
+import torch
+import numpy as np
+
+cfg = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='MobileOne',
+        arch='s0',
+        out_indices=(3, ),
+    ),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=1024,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+        topk=(1, 5),
+    ))
+
+x = torch.randn( (1, 3, 224, 224) )
+classifier = build_classifier(cfg)
+classifier.init_weights()
+classifier.eval()
+y_ori = classifier(x, return_loss=False)
+
+classifier.backbone.switch_to_deploy()
+y_dep = classifier(x, return_loss=False)
+
+for y1, y2 in zip(y_ori, y_dep):
+    assert np.allclose(y1, y2)
+```
+
+## Citation
+
+```bibtex
+@article{mobileone2022,
+  title={An Improved One millisecond Mobile Backbone},
+  author={Vasu, Pavan Kumar Anasosalu and Gabriel, James and Zhu, Jeff and Tuzel, Oncel and Ranjan, Anurag},
+  journal={arXiv preprint arXiv:2206.04040},
+  year={2022}
+}
+```

configs/mobileone/deploy/mobileone-s0_deploy_8xb128_in1k.py

@@ -0,0 +1,3 @@
+_base_ = ['../mobileone-s0_8xb128_in1k.py']
+
+model = dict(backbone=dict(deploy=True))

configs/mobileone/deploy/mobileone-s1_deploy_8xb128_in1k.py

@@ -0,0 +1,3 @@
+_base_ = ['../mobileone-s1_8xb128_in1k.py']
+
+model = dict(backbone=dict(deploy=True))

configs/mobileone/deploy/mobileone-s2_deploy_8xb128_in1k.py

@@ -0,0 +1,3 @@
+_base_ = ['../mobileone-s2_8xb128_in1k.py']
+
+model = dict(backbone=dict(deploy=True))

configs/mobileone/deploy/mobileone-s3_deploy_8xb128_in1k.py

@@ -0,0 +1,3 @@
+_base_ = ['../mobileone-s3_8xb128_in1k.py']
+
+model = dict(backbone=dict(deploy=True))

configs/mobileone/deploy/mobileone-s4_deploy_8xb128_in1k.py

@@ -0,0 +1,3 @@
+_base_ = ['../mobileone-s4_8xb128_in1k.py']
+
+model = dict(backbone=dict(deploy=True))

configs/mobileone/metafile.yml

@@ -0,0 +1,98 @@
+Collections:
+  - Name: MobileOne
+    Metadata:
+      Training Data: ImageNet-1k
+      Architecture:
+        - re-parameterization Convolution
+        - VGG-style Neural Network
+        - Depthwise Convolution
+        - Pointwise Convolution
+    Paper:
+      URL: https://arxiv.org/abs/2206.04040
+      Title: 'An Improved One millisecond Mobile Backbone'
+    README: configs/mobileone/README.md
+    Code:
+      URL: https://github.com/open-mmlab/mmclassification/blob/v1.0.0rc1/configs/mobileone/metafile.yml
+      Version: v1.0.0rc1
+
+Models:
+  - Name: mobileone-s0_3rdparty_8xb128_in1k
+    In Collection: MobileOne
+    Config: configs/mobileone/mobileone-s0_8xb128_in1k.py
+    Metadata:
+      FLOPs: 1091227648  # 1.09G
+      Parameters: 5293272  # 5.29M
+    Results:
+    - Dataset: ImageNet-1k
+      Task: Image Classification
+      Metrics:
+        Top 1 Accuracy: 71.36
+        Top 5 Accuracy: 89.87
+    Weights: https://download.openmmlab.com/mmclassification/v0/mobileone/mobileone-s0_3rdparty_in1k_20220915-007ae971.pth
+    Converted From:
+      Weights: https://docs-assets.developer.apple.com/ml-research/datasets/mobileone/mobileone_s0_unfused.pth.tar
+      Code: https://github.com/apple/ml-mobileone
+  - Name: mobileone-s1_3rdparty_8xb128_in1k
+    In Collection: MobileOne
+    Config: configs/mobileone/mobileone-s1_8xb128_in1k.py
+    Metadata:
+      FLOPs: 863491328    # 8.6G
+      Parameters: 4825192  # 4.82M
+    Results:
+    - Dataset: ImageNet-1k
+      Task: Image Classification
+      Metrics:
+        Top 1 Accuracy: 75.76
+        Top 5 Accuracy: 92.77
+    Weights: https://download.openmmlab.com/mmclassification/v0/mobileone/mobileone-s1_3rdparty_in1k_20220915-473c8469.pth
+    Converted From:
+      Weights: https://docs-assets.developer.apple.com/ml-research/datasets/mobileone/mobileone_s1_unfused.pth.tar
+      Code: https://github.com/apple/ml-mobileone
+  - Name: mobileone-s2_3rdparty_8xb128_in1k
+    In Collection: MobileOne
+    Config: configs/mobileone/mobileone-s2_8xb128_in1k.py
+    Metadata:
+      FLOPs: 1344083328
+      Parameters: 7884648
+    Results:
+    - Dataset: ImageNet-1k
+      Task: Image Classification
+      Metrics:
+        Top 1 Accuracy: 77.39
+        Top 5 Accuracy: 93.63
+    Weights: https://download.openmmlab.com/mmclassification/v0/mobileone/mobileone-s2_3rdparty_in1k_20220915-ed2e4c30.pth
+    Converted From:
+      Weights: https://docs-assets.developer.apple.com/ml-research/datasets/mobileone/mobileone_s2_unfused.pth.tar
+      Code: https://github.com/apple/ml-mobileone
+  - Name: mobileone-s3_3rdparty_8xb128_in1k
+    In Collection: MobileOne
+    Config: configs/mobileone/mobileone-s3_8xb128_in1k.py
+    Metadata:
+      FLOPs: 1951043584
+      Parameters: 10170600
+    Results:
+    - Dataset: ImageNet-1k
+      Task: Image Classification
+      Metrics:
+        Top 1 Accuracy: 77.93
+        Top 5 Accuracy: 93.89
+    Weights: https://download.openmmlab.com/mmclassification/v0/mobileone/mobileone-s3_3rdparty_in1k_20220915-84d6a02c.pth
+    Converted From:
+      Weights: https://docs-assets.developer.apple.com/ml-research/datasets/mobileone/mobileone_s3_unfused.pth.tar
+      Code: https://github.com/apple/ml-mobileone
+  - Name: mobileone-s4_3rdparty_8xb128_in1k
+    In Collection: MobileOne
+    Config: configs/mobileone/mobileone-s4_8xb128_in1k.py
+    Metadata:
+      FLOPs: 3052580688
+      Parameters: 14951248
+    Results:
+    - Dataset: ImageNet-1k
+      Task: Image Classification
+      Metrics:
+        Top 1 Accuracy: 79.30
+        Top 5 Accuracy: 94.37
+    Weights: https://download.openmmlab.com/mmclassification/v0/mobileone/mobileone-s4_3rdparty_in1k_20220915-ce9509ee.pth
+    Converted From:
+      Weights: https://docs-assets.developer.apple.com/ml-research/datasets/mobileone/mobileone_s4_unfused.pth.tar
+      Code: https://github.com/apple/ml-mobileone

configs/mobileone/mobileone-s0_8xb128_in1k.py

@@ -0,0 +1,56 @@
+_base_ = [
+    '../_base_/models/mobileone/mobileone_s0.py',
+    '../_base_/datasets/imagenet_bs32_pil_resize.py',
+    '../_base_/default_runtime.py'
+]
+
+# dataset settings
+train_dataloader = dict(batch_size=128)
+val_dataloader = dict(batch_size=128)
+test_dataloader = dict(batch_size=128)
+
+# schedule settings
+optim_wrapper = dict(
+    optimizer=dict(type='SGD', lr=0.1, momentum=0.9, weight_decay=0.0001),
+    paramwise_cfg=dict(bias_decay_mult=0., norm_decay_mult=0.),
+)
+
+# learning policy
+param_scheduler = [
+    # warm up learning rate scheduler
+    dict(
+        type='LinearLR',
+        start_factor=0.001,
+        by_epoch=True,
+        begin=0,
+        end=5,
+        # update by iter
+        convert_to_iter_based=True),
+    # main learning rate scheduler
+    dict(
+        type='CosineAnnealingLR',
+        T_max=295,
+        eta_min=1.0e-6,
+        by_epoch=True,
+        begin=5,
+        end=300),
+    dict(
+        type='CosineAnnealingParamScheduler',
+        param_name='weight_decay',
+        eta_min=0.00001,
+        by_epoch=True,
+        begin=0,
+        end=300)
+]
+
+# train, val, test setting
+train_cfg = dict(by_epoch=True, max_epochs=300, val_interval=1)
+val_cfg = dict()
+test_cfg = dict()
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR,
+# based on the actual training batch size.
+auto_scale_lr = dict(base_batch_size=1024)
+
+# runtime setting
+custom_hooks = [dict(type='EMAHook', momentum=5e-4, priority='ABOVE_NORMAL')]

configs/mobileone/mobileone-s1_8xb128_in1k.py

@@ -0,0 +1,15 @@
+_base_ = [
+    '../_base_/models/mobileone/mobileone_s1.py',
+    '../_base_/datasets/imagenet_bs32_pil_resize.py',
+    '../_base_/schedules/imagenet_bs256_coslr.py',
+    '../_base_/default_runtime.py'
+]
+
+# dataset settings
+train_dataloader = dict(batch_size=128)
+val_dataloader = dict(batch_size=128)
+test_dataloader = dict(batch_size=128)
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR,
+# based on the actual training batch size.
+auto_scale_lr = dict(base_batch_size=1024)

configs/mobileone/mobileone-s2_8xb128_in1k.py

@@ -0,0 +1,15 @@
+_base_ = [
+    '../_base_/models/mobileone/mobileone_s2.py',
+    '../_base_/datasets/imagenet_bs32_pil_resize.py',
+    '../_base_/schedules/imagenet_bs256_coslr.py',
+    '../_base_/default_runtime.py'
+]
+
+# dataset settings
+train_dataloader = dict(batch_size=128)
+val_dataloader = dict(batch_size=128)
+test_dataloader = dict(batch_size=128)
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR,
+# based on the actual training batch size.
+auto_scale_lr = dict(base_batch_size=1024)

configs/mobileone/mobileone-s3_8xb128_in1k.py

@@ -0,0 +1,15 @@
+_base_ = [
+    '../_base_/models/mobileone/mobileone_s3.py',
+    '../_base_/datasets/imagenet_bs64_pil_resize.py',
+    '../_base_/schedules/imagenet_bs256_coslr.py',
+    '../_base_/default_runtime.py'
+]
+
+# dataset settings
+train_dataloader = dict(batch_size=128)
+val_dataloader = dict(batch_size=128)
+test_dataloader = dict(batch_size=128)
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR,
+# based on the actual training batch size.
+auto_scale_lr = dict(base_batch_size=1024)

configs/mobileone/mobileone-s4_8xb128_in1k.py

@@ -0,0 +1,15 @@
+_base_ = [
+    '../_base_/models/mobileone/mobileone_s4.py',
+    '../_base_/datasets/imagenet_bs64_pil_resize.py',
+    '../_base_/schedules/imagenet_bs256_coslr.py',
+    '../_base_/default_runtime.py'
+]
+
+# dataset settings
+train_dataloader = dict(batch_size=128)
+val_dataloader = dict(batch_size=128)
+test_dataloader = dict(batch_size=128)
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR,
+# based on the actual training batch size.
+auto_scale_lr = dict(base_batch_size=1024)

configs/mvit/README.md

@@ -0,0 +1,44 @@
+# MViT V2
+
+> [MViTv2: Improved Multiscale Vision Transformers for Classification and Detection](http://openaccess.thecvf.com//content/CVPR2022/papers/Li_MViTv2_Improved_Multiscale_Vision_Transformers_for_Classification_and_Detection_CVPR_2022_paper.pdf)
+
+<!-- [ALGORITHM] -->
+
+## Abstract
+
+In this paper, we study Multiscale Vision Transformers (MViTv2) as a unified architecture for image and video
+classification, as well as object detection. We present an improved version of MViT that incorporates
+decomposed relative positional embeddings and residual pooling connections. We instantiate this architecture
+in five sizes and evaluate it for ImageNet classification, COCO detection and Kinetics video recognition where
+it outperforms prior work. We further compare MViTv2s' pooling attention to window attention mechanisms where
+it outperforms the latter in accuracy/compute. Without bells-and-whistles, MViTv2 has state-of-the-art
+performance in 3 domains: 88.8% accuracy on ImageNet classification, 58.7 boxAP on COCO object detection as
+well as 86.1% on Kinetics-400 video classification.
+
+<div align=center>
+<img src="https://user-images.githubusercontent.com/26739999/180376227-755243fa-158e-4068-940a-416036519665.png" width="50%"/>
+</div>
+
+## Results and models
+
+### ImageNet-1k
+
+|     Model      |   Pretrain   | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                 Config                  |                                               Download                                               |
+| :------------: | :----------: | :-------: | :------: | :-------: | :-------: | :-------------------------------------: | :--------------------------------------------------------------------------------------------------: |
+| MViTv2-tiny\*  | From scratch |   24.17   |   4.70   |   82.33   |   96.15   | [config](./mvitv2-tiny_8xb256_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/mvit/mvitv2-tiny_3rdparty_in1k_20220722-db7beeef.pth) |
+| MViTv2-small\* | From scratch |   34.87   |   7.00   |   83.63   |   96.51   | [config](./mvitv2-small_8xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mvit/mvitv2-small_3rdparty_in1k_20220722-986bd741.pth) |
+| MViTv2-base\*  | From scratch |   51.47   |  10.20   |   84.34   |   96.86   | [config](./mvitv2-base_8xb256_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/mvit/mvitv2-base_3rdparty_in1k_20220722-9c4f0a17.pth) |
+| MViTv2-large\* | From scratch |  217.99   |  42.10   |   85.25   |   97.14   | [config](./mvitv2-large_8xb256_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/mvit/mvitv2-large_3rdparty_in1k_20220722-2b57b983.pth) |
+
+*Models with * are converted from the [official repo](https://github.com/facebookresearch/mvit). The config files of these models are only for inference. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
+
+## Citation
+
+```bibtex
+@inproceedings{li2021improved,
+  title={MViTv2: Improved multiscale vision transformers for classification and detection},
+  author={Li, Yanghao and Wu, Chao-Yuan and Fan, Haoqi and Mangalam, Karttikeya and Xiong, Bo and Malik, Jitendra and Feichtenhofer, Christoph},
+  booktitle={CVPR},
+  year={2022}
+}
+```

configs/mvit/metafile.yml

@@ -0,0 +1,95 @@
+Collections:
+  - Name: MViT V2
+    Metadata:
+      Architecture:
+        - Attention Dropout
+        - Convolution
+        - Dense Connections
+        - GELU
+        - Layer Normalization
+        - Scaled Dot-Product Attention
+        - Attention Pooling
+    Paper:
+      URL: http://openaccess.thecvf.com//content/CVPR2022/papers/Li_MViTv2_Improved_Multiscale_Vision_Transformers_for_Classification_and_Detection_CVPR_2022_paper.pdf
+      Title: 'MViTv2: Improved Multiscale Vision Transformers for Classification and Detection'
+    README: configs/mvit/README.md
+    Code:
+      URL: https://github.com/open-mmlab/mmclassification/blob/v0.24.0/mmcls/models/backbones/mvit.py
+      Version: v0.24.0
+
+Models:
+  - Name: mvitv2-tiny_3rdparty_in1k
+    In Collection: MViT V2
+    Metadata:
+      FLOPs: 4703510768
+      Parameters: 24173320
+      Training Data:
+        - ImageNet-1k
+    Results:
+    - Dataset: ImageNet-1k
+      Task: Image Classification
+      Metrics:
+        Top 1 Accuracy: 82.33
+        Top 5 Accuracy: 96.15
+    Weights: https://download.openmmlab.com/mmclassification/v0/mvit/mvitv2-tiny_3rdparty_in1k_20220722-db7beeef.pth
+    Converted From:
+      Weights: https://dl.fbaipublicfiles.com/mvit/mvitv2_models/MViTv2_T_in1k.pyth
+      Code: https://github.com/facebookresearch/mvit
+    Config: configs/mvit/mvitv2-tiny_8xb256_in1k.py
+
+  - Name: mvitv2-small_3rdparty_in1k
+    In Collection: MViT V2
+    Metadata:
+      FLOPs: 6997555136
+      Parameters: 34870216
+      Training Data:
+        - ImageNet-1k
+    Results:
+    - Dataset: ImageNet-1k
+      Task: Image Classification
+      Metrics:
+        Top 1 Accuracy: 83.63
+        Top 5 Accuracy: 96.51
+    Weights: https://download.openmmlab.com/mmclassification/v0/mvit/mvitv2-small_3rdparty_in1k_20220722-986bd741.pth
+    Converted From:
+      Weights: https://dl.fbaipublicfiles.com/mvit/mvitv2_models/MViTv2_S_in1k.pyth
+      Code: https://github.com/facebookresearch/mvit
+    Config: configs/mvit/mvitv2-small_8xb256_in1k.py
+
+  - Name: mvitv2-base_3rdparty_in1k
+    In Collection: MViT V2
+    Metadata:
+      FLOPs: 10157964400
+      Parameters: 51472744
+      Training Data:
+        - ImageNet-1k
+    Results:
+    - Dataset: ImageNet-1k
+      Task: Image Classification
+      Metrics:
+        Top 1 Accuracy: 84.34
+        Top 5 Accuracy: 96.86
+    Weights: https://download.openmmlab.com/mmclassification/v0/mvit/mvitv2-base_3rdparty_in1k_20220722-9c4f0a17.pth
+    Converted From:
+      Weights: https://dl.fbaipublicfiles.com/mvit/mvitv2_models/MViTv2_B_in1k.pyth
+      Code: https://github.com/facebookresearch/mvit
+    Config: configs/mvit/mvitv2-base_8xb256_in1k.py
+
+  - Name: mvitv2-large_3rdparty_in1k
+    In Collection: MViT V2
+    Metadata:
+      FLOPs: 43868151412
+      Parameters: 217992952
+      Training Data:
+        - ImageNet-1k
+    Results:
+    - Dataset: ImageNet-1k
+      Task: Image Classification
+      Metrics:
+        Top 1 Accuracy: 85.25
+        Top 5 Accuracy: 97.14
+    Weights: https://download.openmmlab.com/mmclassification/v0/mvit/mvitv2-large_3rdparty_in1k_20220722-2b57b983.pth
+    Converted From:
+      Weights: https://dl.fbaipublicfiles.com/mvit/mvitv2_models/MViTv2_L_in1k.pyth
+      Code: https://github.com/facebookresearch/mvit
+    Config: configs/mvit/mvitv2-large_8xb256_in1k.py

configs/mvit/mvitv2-base_8xb256_in1k.py

@@ -0,0 +1,43 @@
+_base_ = [
+    '../_base_/models/mvit/mvitv2-base.py',
+    '../_base_/datasets/imagenet_bs64_swin_224.py',
+    '../_base_/schedules/imagenet_bs1024_adamw_swin.py',
+    '../_base_/default_runtime.py'
+]
+
+# dataset settings
+train_dataloader = dict(batch_size=256)
+val_dataloader = dict(batch_size=256)
+test_dataloader = dict(batch_size=256)
+
+# schedule settings
+optim_wrapper = dict(
+    optimizer=dict(lr=2.5e-4),
+    paramwise_cfg=dict(
+        norm_decay_mult=0.0,
+        bias_decay_mult=0.0,
+        custom_keys={
+            '.pos_embed': dict(decay_mult=0.0),
+            '.rel_pos_h': dict(decay_mult=0.0),
+            '.rel_pos_w': dict(decay_mult=0.0)
+        }),
+    clip_grad=dict(max_norm=1.0),
+)
+
+# learning policy
+param_scheduler = [
+    # warm up learning rate scheduler
+    dict(
+        type='LinearLR',
+        start_factor=1e-3,
+        by_epoch=True,
+        end=70,
+        # update by iter
+        convert_to_iter_based=True),
+    # main learning rate scheduler
+    dict(type='CosineAnnealingLR', eta_min=1e-5, by_epoch=True, begin=70)
+]
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR,
+# based on the actual training batch size.
+auto_scale_lr = dict(base_batch_size=2048)

configs/mvit/mvitv2-large_8xb256_in1k.py

@@ -0,0 +1,43 @@
+_base_ = [
+    '../_base_/models/mvit/mvitv2-large.py',
+    '../_base_/datasets/imagenet_bs64_swin_224.py',
+    '../_base_/schedules/imagenet_bs2048_AdamW.py',
+    '../_base_/default_runtime.py'
+]
+
+# dataset settings
+train_dataloader = dict(batch_size=256)
+val_dataloader = dict(batch_size=256)
+test_dataloader = dict(batch_size=256)
+
+# schedule settings
+optim_wrapper = dict(
+    optimizer=dict(lr=2.5e-4),
+    paramwise_cfg=dict(
+        norm_decay_mult=0.0,
+        bias_decay_mult=0.0,
+        custom_keys={
+            '.pos_embed': dict(decay_mult=0.0),
+            '.rel_pos_h': dict(decay_mult=0.0),
+            '.rel_pos_w': dict(decay_mult=0.0)
+        }),
+    clip_grad=dict(max_norm=1.0),
+)
+
+# learning policy
+param_scheduler = [
+    # warm up learning rate scheduler
+    dict(
+        type='LinearLR',
+        start_factor=1e-3,
+        by_epoch=True,
+        end=70,
+        # update by iter
+        convert_to_iter_based=True),
+    # main learning rate scheduler
+    dict(type='CosineAnnealingLR', eta_min=1e-5, by_epoch=True, begin=70)
+]
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR,
+# based on the actual training batch size.
+auto_scale_lr = dict(base_batch_size=2048)

configs/mvit/mvitv2-small_8xb256_in1k.py

@@ -0,0 +1,43 @@
+_base_ = [
+    '../_base_/models/mvit/mvitv2-small.py',
+    '../_base_/datasets/imagenet_bs64_swin_224.py',
+    '../_base_/schedules/imagenet_bs2048_AdamW.py',
+    '../_base_/default_runtime.py'
+]
+
+# dataset settings
+train_dataloader = dict(batch_size=256)
+val_dataloader = dict(batch_size=256)
+test_dataloader = dict(batch_size=256)
+
+# schedule settings
+optim_wrapper = dict(
+    optimizer=dict(lr=2.5e-4),
+    paramwise_cfg=dict(
+        norm_decay_mult=0.0,
+        bias_decay_mult=0.0,
+        custom_keys={
+            '.pos_embed': dict(decay_mult=0.0),
+            '.rel_pos_h': dict(decay_mult=0.0),
+            '.rel_pos_w': dict(decay_mult=0.0)
+        }),
+    clip_grad=dict(max_norm=1.0),
+)
+
+# learning policy
+param_scheduler = [
+    # warm up learning rate scheduler
+    dict(
+        type='LinearLR',
+        start_factor=1e-3,
+        by_epoch=True,
+        end=70,
+        # update by iter
+        convert_to_iter_based=True),
+    # main learning rate scheduler
+    dict(type='CosineAnnealingLR', eta_min=1e-5, by_epoch=True, begin=70)
+]
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR,
+# based on the actual training batch size.
+auto_scale_lr = dict(base_batch_size=2048)

configs/mvit/mvitv2-tiny_8xb256_in1k.py

@@ -0,0 +1,43 @@
+_base_ = [
+    '../_base_/models/mvit/mvitv2-tiny.py',
+    '../_base_/datasets/imagenet_bs64_swin_224.py',
+    '../_base_/schedules/imagenet_bs2048_AdamW.py',
+    '../_base_/default_runtime.py'
+]
+
+# dataset settings
+train_dataloader = dict(batch_size=256)
+val_dataloader = dict(batch_size=256)
+test_dataloader = dict(batch_size=256)
+
+# schedule settings
+optim_wrapper = dict(
+    optimizer=dict(lr=2.5e-4),
+    paramwise_cfg=dict(
+        norm_decay_mult=0.0,
+        bias_decay_mult=0.0,
+        custom_keys={
+            '.pos_embed': dict(decay_mult=0.0),
+            '.rel_pos_h': dict(decay_mult=0.0),
+            '.rel_pos_w': dict(decay_mult=0.0)
+        }),
+    clip_grad=dict(max_norm=1.0),
+)
+
+# learning policy
+param_scheduler = [
+    # warm up learning rate scheduler
+    dict(
+        type='LinearLR',
+        start_factor=1e-3,
+        by_epoch=True,
+        end=70,
+        # update by iter
+        convert_to_iter_based=True),
+    # main learning rate scheduler
+    dict(type='CosineAnnealingLR', eta_min=1e-5, by_epoch=True, begin=70)
+]
+
+# NOTE: `auto_scale_lr` is for automatically scaling LR,
+# based on the actual training batch size.
+auto_scale_lr = dict(base_batch_size=2048)

configs/poolformer/README.md

@@ -16,13 +16,13 @@ Transformers have shown great potential in computer vision tasks. A common belie
 
 ### ImageNet-1k
 
-|      Model       | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                  Config                                   |                                   Download                                   |
-| :--------------: | :-------: | :------: | :-------: | :-------: | :-----------------------------------------------------------------------: | :--------------------------------------------------------------------------: |
-| PoolFormer-S12\* |   11.92   |   1.87   |   77.24   |   93.51   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/poolformer/poolformer-s12_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/poolformer/poolformer-s12_3rdparty_32xb128_in1k_20220414-f8d83051.pth) |
-| PoolFormer-S24\* |   21.39   |   3.51   |   80.33   |   95.05   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/poolformer/poolformer-s24_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/poolformer/poolformer-s24_3rdparty_32xb128_in1k_20220414-d7055904.pth) |
-| PoolFormer-S36\* |   30.86   |   5.15   |   81.43   |   95.45   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/poolformer/poolformer-s36_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/poolformer/poolformer-s36_3rdparty_32xb128_in1k_20220414-d78ff3e8.pth) |
-| PoolFormer-M36\* |   56.17   |   8.96   |   82.14   |   95.71   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/poolformer/poolformer-m36_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/poolformer/poolformer-m36_3rdparty_32xb128_in1k_20220414-c55e0949.pth) |
-| PoolFormer-M48\* |   73.47   |  11.80   |   82.51   |   95.95   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/poolformer/poolformer-m48_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/poolformer/poolformer-m48_3rdparty_32xb128_in1k_20220414-9378f3eb.pth) |
+|      Model       | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                   Config                   |                                                  Download                                                   |
+| :--------------: | :-------: | :------: | :-------: | :-------: | :----------------------------------------: | :---------------------------------------------------------------------------------------------------------: |
+| PoolFormer-S12\* |   11.92   |   1.87   |   77.24   |   93.51   | [config](./poolformer-s12_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/poolformer/poolformer-s12_3rdparty_32xb128_in1k_20220414-f8d83051.pth) |
+| PoolFormer-S24\* |   21.39   |   3.51   |   80.33   |   95.05   | [config](./poolformer-s24_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/poolformer/poolformer-s24_3rdparty_32xb128_in1k_20220414-d7055904.pth) |
+| PoolFormer-S36\* |   30.86   |   5.15   |   81.43   |   95.45   | [config](./poolformer-s36_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/poolformer/poolformer-s36_3rdparty_32xb128_in1k_20220414-d78ff3e8.pth) |
+| PoolFormer-M36\* |   56.17   |   8.96   |   82.14   |   95.71   | [config](./poolformer-m36_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/poolformer/poolformer-m36_3rdparty_32xb128_in1k_20220414-c55e0949.pth) |
+| PoolFormer-M48\* |   73.47   |  11.80   |   82.51   |   95.95   | [config](./poolformer-m48_32xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/poolformer/poolformer-m48_3rdparty_32xb128_in1k_20220414-9378f3eb.pth) |
 
 *Models with * are converted from the [official repo](https://github.com/sail-sg/poolformer). The config files of these models are only for inference. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/poolformer/metafile.yml

@@ -19,7 +19,7 @@ Models:
     Metadata:
       FLOPs: 1871399424
       Parameters: 11915176
-    In Collections: PoolFormer
+    In Collection: PoolFormer
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -36,7 +36,7 @@ Models:
       Training Data: ImageNet-1k
       FLOPs: 3510411008
       Parameters: 21388968
-    In Collections: PoolFormer
+    In Collection: PoolFormer
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -52,7 +52,7 @@ Models:
     Metadata:
       FLOPs: 5149422592
       Parameters: 30862760
-    In Collections: PoolFormer
+    In Collection: PoolFormer
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -69,7 +69,7 @@ Models:
       Training Data: ImageNet-1k
       FLOPs: 8960175744
       Parameters: 56172520
-    In Collections: PoolFormer
+    In Collection: PoolFormer
     Results:
       - Dataset: ImageNet-1k
         Metrics:
@@ -85,7 +85,7 @@ Models:
     Metadata:
       FLOPs: 11801805696
       Parameters: 73473448
-    In Collections: PoolFormer
+    In Collection: PoolFormer
     Results:
       - Dataset: ImageNet-1k
         Metrics:

configs/regnet/README.md

@@ -16,24 +16,24 @@ In this work, we present a new network design paradigm. Our goal is to help adva
 
 ### ImageNet-1k
 
-|      Model      | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                   Config                                   |                                   Download                                   |
-| :-------------: | :-------: | :------: | :-------: | :-------: | :------------------------------------------------------------------------: | :--------------------------------------------------------------------------: |
-|  RegNetX-400MF  |   5.16    |   0.41   |   72.56   |   90.78   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-400mf_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-400mf_8xb128_in1k_20211213-89bfc226.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-400mf_8xb128_in1k_20211208_143316.log.json) |
-|  RegNetX-800MF  |   7.26    |   0.81   |   74.76   |   92.32   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-800mf_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-800mf_8xb128_in1k_20211213-222b0f11.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-800mf_8xb128_in1k_20211207_143037.log.json) |
-|  RegNetX-1.6GF  |   9.19    |   1.63   |   76.84   |   93.31   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-1.6gf_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-1.6gf_8xb128_in1k_20211213-d1b89758.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-1.6gf_8xb128_in1k_20211208_143018.log.json) |
-|  RegNetX-3.2GF  |   15.3    |   3.21   |   78.09   |   94.08   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-3.2gf_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-3.2gf_8xb64_in1k_20211213-1fdd82ae.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-3.2gf_8xb64_in1k_20211208_142720.log.json) |
-|  RegNetX-4.0GF  |   22.12   |   4.0    |   78.60   |   94.17   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-4.0gf_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-4.0gf_8xb64_in1k_20211213-efed675c.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-4.0gf_8xb64_in1k_20211207_150431.log.json) |
-|  RegNetX-6.4GF  |   26.21   |   6.51   |   79.38   |   94.65   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-6.4gf_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-6.4gf_8xb64_in1k_20211215-5c6089da.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-6.4gf_8xb64_in1k_20211213_172748.log.json) |
-|  RegNetX-8.0GF  |   39.57   |   8.03   |   79.12   |   94.51   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-8.0gf_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-8.0gf_8xb64_in1k_20211213-9a9fcc76.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-8.0gf_8xb64_in1k_20211208_103250.log.json) |
-|  RegNetX-12GF   |   46.11   |  12.15   |   79.67   |   95.03   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-12gf_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-12gf_8xb64_in1k_20211213-5df8c2f8.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-12gf_8xb64_in1k_20211208_143713.log.json) |
-| RegNetX-400MF\* |   5.16    |   0.41   |   72.55   |   90.91   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-400mf_8xb128_in1k) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-400MF-0db9f35c.pth) |
-| RegNetX-800MF\* |   7.26    |   0.81   |   75.21   |   92.37   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-800mf_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-800MF-4f9d1e8a.pth) |
-| RegNetX-1.6GF\* |   9.19    |   1.63   |   77.04   |   93.51   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-1.6gf_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-1.6GF-cfb32375.pth) |
-| RegNetX-3.2GF\* |   15.3    |   3.21   |   78.26   |   94.20   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-3.2gf_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-3.2GF-82c43fd5.pth) |
-| RegNetX-4.0GF\* |   22.12   |   4.0    |   78.72   |   94.22   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-4.0gf_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-4.0GF-ef8bb32c.pth) |
-| RegNetX-6.4GF\* |   26.21   |   6.51   |   79.22   |   94.61   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-6.4gf_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-6.4GF-6888c0ea.pth) |
-| RegNetX-8.0GF\* |   39.57   |   8.03   |   79.31   |   94.57   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-8.0gf_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-8.0GF-cb4c77ec.pth) |
-| RegNetX-12GF\*  |   46.11   |  12.15   |   79.91   |   94.78   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/regnet/regnetx-12gf_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-12GF-0574538f.pth) |
+|      Model      | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                  Config                  |                                                    Download                                                    |
+| :-------------: | :-------: | :------: | :-------: | :-------: | :--------------------------------------: | :------------------------------------------------------------------------------------------------------------: |
+|  RegNetX-400MF  |   5.16    |   0.41   |   72.56   |   90.78   | [config](./regnetx-400mf_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-400mf_8xb128_in1k_20211213-89bfc226.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-400mf_8xb128_in1k_20211208_143316.log.json) |
+|  RegNetX-800MF  |   7.26    |   0.81   |   74.76   |   92.32   | [config](./regnetx-800mf_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-800mf_8xb128_in1k_20211213-222b0f11.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-800mf_8xb128_in1k_20211207_143037.log.json) |
+|  RegNetX-1.6GF  |   9.19    |   1.63   |   76.84   |   93.31   | [config](./regnetx-1.6gf_8xb128_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-1.6gf_8xb128_in1k_20211213-d1b89758.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-1.6gf_8xb128_in1k_20211208_143018.log.json) |
+|  RegNetX-3.2GF  |   15.3    |   3.21   |   78.09   |   94.08   | [config](./regnetx-3.2gf_8xb64_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-3.2gf_8xb64_in1k_20211213-1fdd82ae.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-3.2gf_8xb64_in1k_20211208_142720.log.json) |
+|  RegNetX-4.0GF  |   22.12   |   4.0    |   78.60   |   94.17   | [config](./regnetx-4.0gf_8xb64_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-4.0gf_8xb64_in1k_20211213-efed675c.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-4.0gf_8xb64_in1k_20211207_150431.log.json) |
+|  RegNetX-6.4GF  |   26.21   |   6.51   |   79.38   |   94.65   | [config](./regnetx-6.4gf_8xb64_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-6.4gf_8xb64_in1k_20211215-5c6089da.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-6.4gf_8xb64_in1k_20211213_172748.log.json) |
+|  RegNetX-8.0GF  |   39.57   |   8.03   |   79.12   |   94.51   | [config](./regnetx-8.0gf_8xb64_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-8.0gf_8xb64_in1k_20211213-9a9fcc76.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-8.0gf_8xb64_in1k_20211208_103250.log.json) |
+|  RegNetX-12GF   |   46.11   |  12.15   |   79.67   |   95.03   |  [config](./regnetx-12gf_8xb64_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-12gf_8xb64_in1k_20211213-5df8c2f8.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/regnet/regnetx-12gf_8xb64_in1k_20211208_143713.log.json) |
+| RegNetX-400MF\* |   5.16    |   0.41   |   72.55   |   90.91   | [config](./regnetx-400mf_8xb128_in1k.py) |     [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-400MF-0db9f35c.pth)      |
+| RegNetX-800MF\* |   7.26    |   0.81   |   75.21   |   92.37   | [config](./regnetx-800mf_8xb128_in1k.py) |     [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-800MF-4f9d1e8a.pth)      |
+| RegNetX-1.6GF\* |   9.19    |   1.63   |   77.04   |   93.51   | [config](./regnetx-1.6gf_8xb128_in1k.py) |     [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-1.6GF-cfb32375.pth)      |
+| RegNetX-3.2GF\* |   15.3    |   3.21   |   78.26   |   94.20   | [config](./regnetx-3.2gf_8xb64_in1k.py)  |     [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-3.2GF-82c43fd5.pth)      |
+| RegNetX-4.0GF\* |   22.12   |   4.0    |   78.72   |   94.22   | [config](./regnetx-4.0gf_8xb64_in1k.py)  |     [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-4.0GF-ef8bb32c.pth)      |
+| RegNetX-6.4GF\* |   26.21   |   6.51   |   79.22   |   94.61   | [config](./regnetx-6.4gf_8xb64_in1k.py)  |     [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-6.4GF-6888c0ea.pth)      |
+| RegNetX-8.0GF\* |   39.57   |   8.03   |   79.31   |   94.57   | [config](./regnetx-8.0gf_8xb64_in1k.py)  |     [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-8.0GF-cb4c77ec.pth)      |
+| RegNetX-12GF\*  |   46.11   |  12.15   |   79.91   |   94.78   |  [config](./regnetx-12gf_8xb64_in1k.py)  |      [model](https://download.openmmlab.com/mmclassification/v0/regnet/convert/RegNetX-12GF-0574538f.pth)      |
 
 *Models with * are converted from [pycls](https://github.com/facebookresearch/pycls/blob/master/MODEL_ZOO.md). The config files of these models are only for validation.*
 

configs/repmlp/README.md

@@ -18,8 +18,8 @@ We propose RepMLP, a multi-layer-perceptron-style neural network building block
 
 |     Model     | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                   Config                                    |                                   Download                                    |
 | :-----------: | :-------: | :------: | :-------: | :-------: | :-------------------------------------------------------------------------: | :---------------------------------------------------------------------------: |
-| RepMLP-B224\* |   68.24   |   6.71   |   80.41   |   95.12   | [train_cfg](https://github.com/open-mmlab/mmclassification/blob/master/configs/repmlp/repmlp-base_8xb64_in1k.py) \| [deploy_cfg](https://github.com/open-mmlab/mmclassification/blob/master/configs/repmlp/repmlp-base_delopy_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repmlp/repmlp-base_3rdparty_8xb64_in1k_20220330-1cb1f11b.pth) |
-| RepMLP-B256\* |   96.45   |   9.69   |   81.11   |   95.5    | [train_cfg](https://github.com/open-mmlab/mmclassification/blob/master/configs/repmlp/repmlp-base_8xb64_in1k-256px.py) \| [deploy_cfg](https://github.com/open-mmlab/mmclassification/blob/master/configs/repmlp/repmlp-b256_deploy_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repmlp/repmlp-base_3rdparty_8xb64_in1k-256px_20220330-7c5a91ce.pth) |
+| RepMLP-B224\* |   68.24   |   6.71   |   80.41   |   95.12   | [train_cfg](./repmlp-base_8xb64_in1k.py) \| [deploy_cfg](./repmlp-base_delopy_8xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repmlp/repmlp-base_3rdparty_8xb64_in1k_20220330-1cb1f11b.pth) |
+| RepMLP-B256\* |   96.45   |   9.69   |   81.11   |   95.5    | [train_cfg](./repmlp-base_8xb64_in1k-256px.py) \| [deploy_cfg](./repmlp-base_deploy_8xb64_in1k-256px.py) | [model](https://download.openmmlab.com/mmclassification/v0/repmlp/repmlp-base_3rdparty_8xb64_in1k-256px_20220330-7c5a91ce.pth) |
 
 *Models with * are converted from [the official repo.](https://github.com/DingXiaoH/RepMLP). The config files of these models are only for validation. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/repvgg/README.md

@@ -18,18 +18,18 @@ We present a simple but powerful architecture of convolutional neural network, w
 
 |     Model     | Epochs |             Params(M)             |            Flops(G)             | Top-1 (%) | Top-5 (%) |                      Config                      |                      Download                       |
 | :-----------: | :----: | :-------------------------------: | :-----------------------------: | :-------: | :-------: | :----------------------------------------------: | :-------------------------------------------------: |
-|  RepVGG-A0\*  |  120   |   9.11（train) \| 8.31 (deploy)   |  1.52 (train) \| 1.36 (deploy)  |   72.41   |   90.50   | [config (train)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/repvgg-A0_4xb64-coslr-120e_in1k.py) \| [config (deploy)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/deploy/repvgg-A0_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-A0_3rdparty_4xb64-coslr-120e_in1k_20210909-883ab98c.pth) |
-|  RepVGG-A1\*  |  120   |  14.09 (train) \| 12.79 (deploy)  |  2.64 (train) \| 2.37 (deploy)  |   74.47   |   91.85   | [config (train)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/repvgg-A1_4xb64-coslr-120e_in1k.py) \| [config (deploy)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/deploy/repvgg-A1_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-A1_3rdparty_4xb64-coslr-120e_in1k_20210909-24003a24.pth) |
-|  RepVGG-A2\*  |  120   |  28.21 (train) \| 25.5 (deploy)   |  5.7 (train)  \| 5.12 (deploy)  |   76.48   |   93.01   | [config (train)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/repvgg-A2_4xb64-coslr-120e_in1k.py) \|[config (deploy)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/deploy/repvgg-A2_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-A2_3rdparty_4xb64-coslr-120e_in1k_20210909-97d7695a.pth) |
-|  RepVGG-B0\*  |  120   |  15.82 (train) \| 14.34 (deploy)  |  3.42 (train) \| 3.06 (deploy)  |   75.14   |   92.42   | [config (train)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/repvgg-B0_4xb64-coslr-120e_in1k.py) \|[config (deploy)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/deploy/repvgg-B0_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B0_3rdparty_4xb64-coslr-120e_in1k_20210909-446375f4.pth) |
-|  RepVGG-B1\*  |  120   |  57.42 (train) \| 51.83 (deploy)  | 13.16 (train) \| 11.82 (deploy) |   78.37   |   94.11   | [config (train)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/repvgg-B1_4xb64-coslr-120e_in1k.py) \|[config (deploy)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/deploy/repvgg-B1_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B1_3rdparty_4xb64-coslr-120e_in1k_20210909-750cdf67.pth) |
-| RepVGG-B1g2\* |  120   |  45.78 (train) \| 41.36 (deploy)  |  9.82 (train) \| 8.82 (deploy)  |   77.79   |   93.88   | [config (train)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/repvgg-B1g2_4xb64-coslr-120e_in1k.py) \|[config (deploy)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/deploy/repvgg-B1g2_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B1g2_3rdparty_4xb64-coslr-120e_in1k_20210909-344f6422.pth) |
-| RepVGG-B1g4\* |  120   |  39.97 (train) \| 36.13 (deploy)  |  8.15 (train) \| 7.32 (deploy)  |   77.58   |   93.84   | [config (train)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/repvgg-B1g4_4xb64-coslr-120e_in1k.py) \|[config (deploy)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/deploy/repvgg-B1g4_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B1g4_3rdparty_4xb64-coslr-120e_in1k_20210909-d4c1a642.pth) |
-|  RepVGG-B2\*  |  120   |  89.02 (train) \| 80.32 (deploy)  | 20.46 (train) \| 18.39 (deploy) |   78.78   |   94.42   | [config (train)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/repvgg-B2_4xb64-coslr-120e_in1k.py) \|[config (deploy)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/deploy/repvgg-B2_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B2_3rdparty_4xb64-coslr-120e_in1k_20210909-bd6b937c.pth) |
-| RepVGG-B2g4\* |  200   |  61.76 (train) \| 55.78 (deploy)  | 12.63 (train) \| 11.34 (deploy) |   79.38   |   94.68   | [config (train)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/repvgg-B2g4_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) \|[config (deploy)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/deploy/repvgg-B2g4_deploy_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B2g4_3rdparty_4xb64-autoaug-lbs-mixup-coslr-200e_in1k_20210909-7b7955f0.pth) |
-|  RepVGG-B3\*  |  200   | 123.09 (train) \| 110.96 (deploy) | 29.17 (train) \| 26.22 (deploy) |   80.52   |   95.26   | [config (train)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/repvgg-B3_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) \|[config (deploy)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/deploy/repvgg-B3_deploy_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B3_3rdparty_4xb64-autoaug-lbs-mixup-coslr-200e_in1k_20210909-dda968bf.pth) |
-| RepVGG-B3g4\* |  200   |  83.83 (train) \| 75.63 (deploy)  | 17.9 (train) \| 16.08 (deploy)  |   80.22   |   95.10   | [config (train)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/repvgg-B3g4_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) \|[config (deploy)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/deploy/repvgg-B3g4_deploy_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B3g4_3rdparty_4xb64-autoaug-lbs-mixup-coslr-200e_in1k_20210909-4e54846a.pth) |
-| RepVGG-D2se\* |  200   | 133.33 (train) \| 120.39 (deploy) | 36.56 (train) \| 32.85 (deploy) |   81.81   |   95.94   | [config (train)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/repvgg-D2se_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) \|[config (deploy)](https://github.com/open-mmlab/mmclassification/blob/master/configs/repvgg/deploy/repvgg-D2se_deploy_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-D2se_3rdparty_4xb64-autoaug-lbs-mixup-coslr-200e_in1k_20210909-cf3139b7.pth) |
+|  RepVGG-A0\*  |  120   |   9.11（train) \| 8.31 (deploy)   |  1.52 (train) \| 1.36 (deploy)  |   72.41   |   90.50   | [config (train)](./repvgg-A0_4xb64-coslr-120e_in1k.py) \| [config (deploy)](./deploy/repvgg-A0_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-A0_3rdparty_4xb64-coslr-120e_in1k_20210909-883ab98c.pth) |
+|  RepVGG-A1\*  |  120   |  14.09 (train) \| 12.79 (deploy)  |  2.64 (train) \| 2.37 (deploy)  |   74.47   |   91.85   | [config (train)](./repvgg-A1_4xb64-coslr-120e_in1k.py) \| [config (deploy)](./deploy/repvgg-A1_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-A1_3rdparty_4xb64-coslr-120e_in1k_20210909-24003a24.pth) |
+|  RepVGG-A2\*  |  120   |  28.21 (train) \| 25.5 (deploy)   |  5.7 (train)  \| 5.12 (deploy)  |   76.48   |   93.01   | [config (train)](./repvgg-A2_4xb64-coslr-120e_in1k.py) \|[config (deploy)](./deploy/repvgg-A2_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-A2_3rdparty_4xb64-coslr-120e_in1k_20210909-97d7695a.pth) |
+|  RepVGG-B0\*  |  120   |  15.82 (train) \| 14.34 (deploy)  |  3.42 (train) \| 3.06 (deploy)  |   75.14   |   92.42   | [config (train)](./repvgg-B0_4xb64-coslr-120e_in1k.py) \|[config (deploy)](./deploy/repvgg-B0_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B0_3rdparty_4xb64-coslr-120e_in1k_20210909-446375f4.pth) |
+|  RepVGG-B1\*  |  120   |  57.42 (train) \| 51.83 (deploy)  | 13.16 (train) \| 11.82 (deploy) |   78.37   |   94.11   | [config (train)](./repvgg-B1_4xb64-coslr-120e_in1k.py) \|[config (deploy)](./deploy/repvgg-B1_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B1_3rdparty_4xb64-coslr-120e_in1k_20210909-750cdf67.pth) |
+| RepVGG-B1g2\* |  120   |  45.78 (train) \| 41.36 (deploy)  |  9.82 (train) \| 8.82 (deploy)  |   77.79   |   93.88   | [config (train)](./repvgg-B1g2_4xb64-coslr-120e_in1k.py) \|[config (deploy)](./deploy/repvgg-B1g2_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B1g2_3rdparty_4xb64-coslr-120e_in1k_20210909-344f6422.pth) |
+| RepVGG-B1g4\* |  120   |  39.97 (train) \| 36.13 (deploy)  |  8.15 (train) \| 7.32 (deploy)  |   77.58   |   93.84   | [config (train)](./repvgg-B1g4_4xb64-coslr-120e_in1k.py) \|[config (deploy)](./deploy/repvgg-B1g4_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B1g4_3rdparty_4xb64-coslr-120e_in1k_20210909-d4c1a642.pth) |
+|  RepVGG-B2\*  |  120   |  89.02 (train) \| 80.32 (deploy)  | 20.46 (train) \| 18.39 (deploy) |   78.78   |   94.42   | [config (train)](./repvgg-B2_4xb64-coslr-120e_in1k.py) \|[config (deploy)](./deploy/repvgg-B2_deploy_4xb64-coslr-120e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B2_3rdparty_4xb64-coslr-120e_in1k_20210909-bd6b937c.pth) |
+| RepVGG-B2g4\* |  200   |  61.76 (train) \| 55.78 (deploy)  | 12.63 (train) \| 11.34 (deploy) |   79.38   |   94.68   | [config (train)](./repvgg-B2g4_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) \|[config (deploy)](./deploy/repvgg-B2g4_deploy_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B2g4_3rdparty_4xb64-autoaug-lbs-mixup-coslr-200e_in1k_20210909-7b7955f0.pth) |
+|  RepVGG-B3\*  |  200   | 123.09 (train) \| 110.96 (deploy) | 29.17 (train) \| 26.22 (deploy) |   80.52   |   95.26   | [config (train)](./repvgg-B3_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) \|[config (deploy)](./deploy/repvgg-B3_deploy_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B3_3rdparty_4xb64-autoaug-lbs-mixup-coslr-200e_in1k_20210909-dda968bf.pth) |
+| RepVGG-B3g4\* |  200   |  83.83 (train) \| 75.63 (deploy)  | 17.9 (train) \| 16.08 (deploy)  |   80.22   |   95.10   | [config (train)](./repvgg-B3g4_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) \|[config (deploy)](./deploy/repvgg-B3g4_deploy_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B3g4_3rdparty_4xb64-autoaug-lbs-mixup-coslr-200e_in1k_20210909-4e54846a.pth) |
+| RepVGG-D2se\* |  200   | 133.33 (train) \| 120.39 (deploy) | 36.56 (train) \| 32.85 (deploy) |   81.81   |   95.94   | [config (train)](./repvgg-D2se_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) \|[config (deploy)](./deploy/repvgg-D2se_deploy_4xb64-autoaug-lbs-mixup-coslr-200e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-D2se_3rdparty_4xb64-autoaug-lbs-mixup-coslr-200e_in1k_20210909-cf3139b7.pth) |
 
 *Models with * are converted from the [official repo](https://github.com/DingXiaoH/RepVGG). The config files of these models are only for validation. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/res2net/README.md

@@ -16,11 +16,11 @@ Representing features at multiple scales is of great importance for numerous vis
 
 ### ImageNet-1k
 
-|        Model         | resolution | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                               Config                               |                               Download                                |
-| :------------------: | :--------: | :-------: | :------: | :-------: | :-------: | :----------------------------------------------------------------: | :-------------------------------------------------------------------: |
-| Res2Net-50-14w-8s\*  |  224x224   |   25.06   |   4.22   |   78.14   |   93.85   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/res2net/res2net50-w14-s8_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/res2net/res2net50-w14-s8_3rdparty_8xb32_in1k_20210927-bc967bf1.pth) |
-| Res2Net-50-26w-8s\*  |  224x224   |   48.40   |   8.39   |   79.20   |   94.36   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/res2net/res2net50-w26-s8_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/res2net/res2net50-w26-s8_3rdparty_8xb32_in1k_20210927-f547a94b.pth) |
-| Res2Net-101-26w-4s\* |  224x224   |   45.21   |   8.12   |   79.19   |   94.44   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/res2net/res2net101-w26-s4_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/res2net/res2net101-w26-s4_3rdparty_8xb32_in1k_20210927-870b6c36.pth) |
+|        Model         | resolution | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                   Config                    |                                           Download                                           |
+| :------------------: | :--------: | :-------: | :------: | :-------: | :-------: | :-----------------------------------------: | :------------------------------------------------------------------------------------------: |
+| Res2Net-50-14w-8s\*  |  224x224   |   25.06   |   4.22   |   78.14   |   93.85   | [config](./res2net50-w14-s8_8xb32_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/res2net/res2net50-w14-s8_3rdparty_8xb32_in1k_20210927-bc967bf1.pth) |
+| Res2Net-50-26w-8s\*  |  224x224   |   48.40   |   8.39   |   79.20   |   94.36   | [config](./res2net50-w26-s8_8xb32_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/res2net/res2net50-w26-s8_3rdparty_8xb32_in1k_20210927-f547a94b.pth) |
+| Res2Net-101-26w-4s\* |  224x224   |   45.21   |   8.12   |   79.19   |   94.44   | [config](./res2net101-w26-s4_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/res2net/res2net101-w26-s4_3rdparty_8xb32_in1k_20210927-870b6c36.pth) |
 
 *Models with * are converted from the [official repo](https://github.com/Res2Net/Res2Net-PretrainedModels). The config files of these models are only for validation. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 

configs/resnet/README.md

@@ -26,41 +26,41 @@ The pre-trained models on ImageNet-21k are used to fine-tune, and therefore don'
 
 ### Cifar10
 
-|   Model    | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                    Config                                    |                                    Download                                     |
-| :--------: | :-------: | :------: | :-------: | :-------: | :--------------------------------------------------------------------------: | :-----------------------------------------------------------------------------: |
-| ResNet-18  |   11.17   |   0.56   |   94.82   |   99.87   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet18_8xb16_cifar10.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet18_b16x8_cifar10_20210528-bd6371c8.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet18_b16x8_cifar10_20210528-bd6371c8.log.json) |
-| ResNet-34  |   21.28   |   1.16   |   95.34   |   99.87   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet34_8xb16_cifar10.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet34_b16x8_cifar10_20210528-a8aa36a6.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet34_b16x8_cifar10_20210528-a8aa36a6.log.json) |
-| ResNet-50  |   23.52   |   1.31   |   95.55   |   99.91   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet50_8xb16_cifar10.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_b16x8_cifar10_20210528-f54bfad9.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_b16x8_cifar10_20210528-f54bfad9.log.json) |
-| ResNet-101 |   42.51   |   2.52   |   95.58   |   99.87   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet101_8xb16_cifar10.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet101_b16x8_cifar10_20210528-2d29e936.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet101_b16x8_cifar10_20210528-2d29e936.log.json) |
-| ResNet-152 |   58.16   |   3.74   |   95.76   |   99.89   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet152_8xb16_cifar10.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet152_b16x8_cifar10_20210528-3e8e9178.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet152_b16x8_cifar10_20210528-3e8e9178.log.json) |
+|   Model    | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                 Config                 |                                                       Download                                                        |
+| :--------: | :-------: | :------: | :-------: | :-------: | :------------------------------------: | :-------------------------------------------------------------------------------------------------------------------: |
+| ResNet-18  |   11.17   |   0.56   |   94.82   |   99.87   | [config](./resnet18_8xb16_cifar10.py)  | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet18_b16x8_cifar10_20210528-bd6371c8.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet18_b16x8_cifar10_20210528-bd6371c8.log.json) |
+| ResNet-34  |   21.28   |   1.16   |   95.34   |   99.87   | [config](./resnet34_8xb16_cifar10.py)  | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet34_b16x8_cifar10_20210528-a8aa36a6.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet34_b16x8_cifar10_20210528-a8aa36a6.log.json) |
+| ResNet-50  |   23.52   |   1.31   |   95.55   |   99.91   | [config](./resnet50_8xb16_cifar10.py)  | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_b16x8_cifar10_20210528-f54bfad9.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_b16x8_cifar10_20210528-f54bfad9.log.json) |
+| ResNet-101 |   42.51   |   2.52   |   95.58   |   99.87   | [config](./resnet101_8xb16_cifar10.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet101_b16x8_cifar10_20210528-2d29e936.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet101_b16x8_cifar10_20210528-2d29e936.log.json) |
+| ResNet-152 |   58.16   |   3.74   |   95.76   |   99.89   | [config](./resnet152_8xb16_cifar10.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet152_b16x8_cifar10_20210528-3e8e9178.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet152_b16x8_cifar10_20210528-3e8e9178.log.json) |
 
 ### Cifar100
 
-|   Model   | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                    Config                                     |                                    Download                                     |
-| :-------: | :-------: | :------: | :-------: | :-------: | :---------------------------------------------------------------------------: | :-----------------------------------------------------------------------------: |
-| ResNet-50 |   23.71   |   1.31   |   79.90   |   95.19   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet50_8xb16_cifar100.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_b16x8_cifar100_20210528-67b58a1b.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_b16x8_cifar100_20210528-67b58a1b.log.json) |
+|   Model   | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                 Config                 |                                                        Download                                                        |
+| :-------: | :-------: | :------: | :-------: | :-------: | :------------------------------------: | :--------------------------------------------------------------------------------------------------------------------: |
+| ResNet-50 |   23.71   |   1.31   |   79.90   |   95.19   | [config](./resnet50_8xb16_cifar100.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_b16x8_cifar100_20210528-67b58a1b.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_b16x8_cifar100_20210528-67b58a1b.log.json) |
 
 ### ImageNet-1k
 
-|       Model        | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                  Config                                  |                                  Download                                   |
-| :----------------: | :-------: | :------: | :-------: | :-------: | :----------------------------------------------------------------------: | :-------------------------------------------------------------------------: |
-|     ResNet-18      |   11.69   |   1.82   |   69.90   |   89.43   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet18_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet18_8xb32_in1k_20210831-fbbb1da6.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet18_8xb32_in1k_20210831-fbbb1da6.log.json) |
-|     ResNet-34      |   21.8    |   3.68   |   73.62   |   91.59   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet34_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet34_8xb32_in1k_20210831-f257d4e6.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet34_8xb32_in1k_20210831-f257d4e6.log.json) |
-|     ResNet-50      |   25.56   |   4.12   |   76.55   |   93.06   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet50_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb32_in1k_20210831-ea4938fc.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb32_in1k_20210831-ea4938fc.log.json) |
-|     ResNet-101     |   44.55   |   7.85   |   77.97   |   94.06   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet101_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet101_8xb32_in1k_20210831-539c63f8.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet101_8xb32_in1k_20210831-539c63f8.log.json) |
-|     ResNet-152     |   60.19   |  11.58   |   78.48   |   94.13   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet152_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet152_8xb32_in1k_20210901-4d7582fa.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet152_8xb32_in1k_20210901-4d7582fa.log.json) |
-|    ResNetV1C-50    |   25.58   |   4.36   |   77.01   |   93.58   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnetv1c50_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1c50_8xb32_in1k_20220214-3343eccd.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1c50_8xb32_in1k_20220214-3343eccd.log.json) |
-|   ResNetV1C-101    |   44.57   |   8.09   |   78.30   |   94.27   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnetv1c101_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1c101_8xb32_in1k_20220214-434fe45f.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1c101_8xb32_in1k_20220214-434fe45f.log.json) |
-|   ResNetV1C-152    |   60.21   |  11.82   |   78.76   |   94.41   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnetv1c152_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1c152_8xb32_in1k_20220214-c013291f.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1c152_8xb32_in1k_20220214-c013291f.log.json) |
-|    ResNetV1D-50    |   25.58   |   4.36   |   77.54   |   93.57   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnetv1d50_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d50_b32x8_imagenet_20210531-db14775a.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d50_b32x8_imagenet_20210531-db14775a.log.json) |
-|   ResNetV1D-101    |   44.57   |   8.09   |   78.93   |   94.48   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnetv1d101_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d101_b32x8_imagenet_20210531-6e13bcd3.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d101_b32x8_imagenet_20210531-6e13bcd3.log.json) |
-|   ResNetV1D-152    |   60.21   |  11.82   |   79.41   |   94.70   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnetv1d152_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d152_b32x8_imagenet_20210531-278cf22a.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d152_b32x8_imagenet_20210531-278cf22a.log.json) |
-|  ResNet-50 (fp16)  |   25.56   |   4.12   |   76.30   |   93.07   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet50_8xb32-fp16_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/fp16/resnet50_batch256_fp16_imagenet_20210320-b3964210.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/fp16/resnet50_batch256_fp16_imagenet_20210320-b3964210.log.json) |
-|  Wide-ResNet-50\*  |   68.88   |  11.44   |   78.48   |   94.08   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/wide-resnet50_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/wide-resnet50_3rdparty_8xb32_in1k_20220304-66678344.pth) |
-| Wide-ResNet-101\*  |  126.89   |  22.81   |   78.84   |   94.28   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet101_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/wide-resnet101_3rdparty_8xb32_in1k_20220304-8d5f9d61.pth) |
-| ResNet-50 (rsb-a1) |   25.56   |   4.12   |   80.12   |   94.78   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet50_8xb256-rsb-a1-600e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb256-rsb-a1-600e_in1k_20211228-20e21305.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb256-rsb-a1-600e_in1k_20211228-20e21305.log.json) |
-| ResNet-50 (rsb-a2) |   25.56   |   4.12   |   79.55   |   94.37   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet50_8xb256-rsb-a2-300e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb256-rsb-a2-300e_in1k_20211228-0fd8be6e.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb256-rsb-a2-300e_in1k_20211228-0fd8be6e.log.json) |
-| ResNet-50 (rsb-a3) |   25.56   |   4.12   |   78.30   |   93.80   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet50_8xb256-rsb-a3-100e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb256-rsb-a3-100e_in1k_20211228-3493673c.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb256-rsb-a3-100e_in1k_20211228-3493673c.log.json) |
+|       Model        | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                     Config                      |                                               Download                                               |
+| :----------------: | :-------: | :------: | :-------: | :-------: | :---------------------------------------------: | :--------------------------------------------------------------------------------------------------: |
+|     ResNet-18      |   11.69   |   1.82   |   69.90   |   89.43   |       [config](./resnet18_8xb32_in1k.py)        | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet18_8xb32_in1k_20210831-fbbb1da6.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet18_8xb32_in1k_20210831-fbbb1da6.log.json) |
+|     ResNet-34      |   21.8    |   3.68   |   73.62   |   91.59   |       [config](./resnet34_8xb32_in1k.py)        | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet34_8xb32_in1k_20210831-f257d4e6.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet34_8xb32_in1k_20210831-f257d4e6.log.json) |
+|     ResNet-50      |   25.56   |   4.12   |   76.55   |   93.06   |       [config](./resnet50_8xb32_in1k.py)        | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb32_in1k_20210831-ea4938fc.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb32_in1k_20210831-ea4938fc.log.json) |
+|     ResNet-101     |   44.55   |   7.85   |   77.97   |   94.06   |       [config](./resnet101_8xb32_in1k.py)       | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet101_8xb32_in1k_20210831-539c63f8.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet101_8xb32_in1k_20210831-539c63f8.log.json) |
+|     ResNet-152     |   60.19   |  11.58   |   78.48   |   94.13   |       [config](./resnet152_8xb32_in1k.py)       | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet152_8xb32_in1k_20210901-4d7582fa.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet152_8xb32_in1k_20210901-4d7582fa.log.json) |
+|    ResNetV1C-50    |   25.58   |   4.36   |   77.01   |   93.58   |      [config](./resnetv1c50_8xb32_in1k.py)      | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1c50_8xb32_in1k_20220214-3343eccd.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1c50_8xb32_in1k_20220214-3343eccd.log.json) |
+|   ResNetV1C-101    |   44.57   |   8.09   |   78.30   |   94.27   |     [config](./resnetv1c101_8xb32_in1k.py)      | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1c101_8xb32_in1k_20220214-434fe45f.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1c101_8xb32_in1k_20220214-434fe45f.log.json) |
+|   ResNetV1C-152    |   60.21   |  11.82   |   78.76   |   94.41   |     [config](./resnetv1c152_8xb32_in1k.py)      | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1c152_8xb32_in1k_20220214-c013291f.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1c152_8xb32_in1k_20220214-c013291f.log.json) |
+|    ResNetV1D-50    |   25.58   |   4.36   |   77.54   |   93.57   |      [config](./resnetv1d50_8xb32_in1k.py)      | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d50_b32x8_imagenet_20210531-db14775a.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d50_b32x8_imagenet_20210531-db14775a.log.json) |
+|   ResNetV1D-101    |   44.57   |   8.09   |   78.93   |   94.48   |     [config](./resnetv1d101_8xb32_in1k.py)      | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d101_b32x8_imagenet_20210531-6e13bcd3.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d101_b32x8_imagenet_20210531-6e13bcd3.log.json) |
+|   ResNetV1D-152    |   60.21   |  11.82   |   79.41   |   94.70   |     [config](./resnetv1d152_8xb32_in1k.py)      | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d152_b32x8_imagenet_20210531-278cf22a.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d152_b32x8_imagenet_20210531-278cf22a.log.json) |
+|  ResNet-50 (fp16)  |   25.56   |   4.12   |   76.30   |   93.07   |     [config](./resnet50_8xb32-fp16_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/fp16/resnet50_batch256_fp16_imagenet_20210320-b3964210.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/fp16/resnet50_batch256_fp16_imagenet_20210320-b3964210.log.json) |
+|  Wide-ResNet-50\*  |   68.88   |  11.44   |   78.48   |   94.08   |  [config](../wrn/wide-resnet50_8xb32_in1k.py)   | [model](https://download.openmmlab.com/mmclassification/v0/resnet/wide-resnet50_3rdparty_8xb32_in1k_20220304-66678344.pth) |
+| Wide-ResNet-101\*  |  126.89   |  22.81   |   78.84   |   94.28   |  [config](../wrn/wide-resnet101_8xb32_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/resnet/wide-resnet101_3rdparty_8xb32_in1k_20220304-8d5f9d61.pth) |
+| ResNet-50 (rsb-a1) |   25.56   |   4.12   |   80.12   |   94.78   | [config](./resnet50_8xb256-rsb-a1-600e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb256-rsb-a1-600e_in1k_20211228-20e21305.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb256-rsb-a1-600e_in1k_20211228-20e21305.log.json) |
+| ResNet-50 (rsb-a2) |   25.56   |   4.12   |   79.55   |   94.37   | [config](./resnet50_8xb256-rsb-a2-300e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb256-rsb-a2-300e_in1k_20211228-0fd8be6e.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb256-rsb-a2-300e_in1k_20211228-0fd8be6e.log.json) |
+| ResNet-50 (rsb-a3) |   25.56   |   4.12   |   78.30   |   93.80   | [config](./resnet50_8xb256-rsb-a3-100e_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb256-rsb-a3-100e_in1k_20211228-3493673c.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb256-rsb-a3-100e_in1k_20211228-3493673c.log.json) |
 
 *The "rsb" means using the training settings from [ResNet strikes back: An improved training procedure in timm](https://arxiv.org/abs/2110.00476).*
 
@@ -68,9 +68,9 @@ The pre-trained models on ImageNet-21k are used to fine-tune, and therefore don'
 
 ### CUB-200-2011
 
-|   Model   |                       Pretrain                       | resolution | Params(M) | Flops(G) | Top-1 (%) |                       Config                       |                       Download                        |
-| :-------: | :--------------------------------------------------: | :--------: | :-------: | :------: | :-------: | :------------------------------------------------: | :---------------------------------------------------: |
-| ResNet-50 | [ImageNet-21k-mill](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_3rdparty-mill_in21k_20220331-faac000b.pth) |  448x448   |   23.92   |  16.48   |   88.45   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet50_8xb8_cub.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb8_cub_20220307-57840e60.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb8_cub_20220307-57840e60.log.json) |
+|   Model   |                           Pretrain                            | resolution | Params(M) | Flops(G) | Top-1 (%) |              Config              |                            Download                            |
+| :-------: | :-----------------------------------------------------------: | :--------: | :-------: | :------: | :-------: | :------------------------------: | :------------------------------------------------------------: |
+| ResNet-50 | [ImageNet-21k-mill](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_3rdparty-mill_in21k_20220331-faac000b.pth) |  448x448   |   23.92   |  16.48   |   88.45   | [config](./resnet50_8xb8_cub.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb8_cub_20220307-57840e60.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb8_cub_20220307-57840e60.log.json) |
 
 ## Citation
 

configs/resnet/resnet50_8xb256-rsb-a1-600e_in1k.py

@@ -16,6 +16,7 @@ model = dict(
             type='LabelSmoothLoss',
             label_smooth_val=0.1,
             mode='original',
+            use_sigmoid=True,
         )),
     train_cfg=dict(augments=[
         dict(type='Mixup', alpha=0.2, num_classes=1000),

configs/resnext/README.md

@@ -16,12 +16,12 @@ We present a simple, highly modularized network architecture for image classific
 
 ### ImageNet-1k
 
-|       Model       | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                  Config                                   |                                  Download                                   |
-| :---------------: | :-------: | :------: | :-------: | :-------: | :-----------------------------------------------------------------------: | :-------------------------------------------------------------------------: |
-| ResNeXt-32x4d-50  |   25.03   |   4.27   |   77.90   |   93.66   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnext/resnext50-32x4d_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnext/resnext50_32x4d_b32x8_imagenet_20210429-56066e27.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnext/resnext50_32x4d_b32x8_imagenet_20210429-56066e27.log.json) |
-| ResNeXt-32x4d-101 |   44.18   |   8.03   |   78.61   |   94.17   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnext/resnext101-32x4d_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnext/resnext101_32x4d_b32x8_imagenet_20210506-e0fa3dd5.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnext/resnext101_32x4d_b32x8_imagenet_20210506-e0fa3dd5.log.json) |
-| ResNeXt-32x8d-101 |   88.79   |   16.5   |   79.27   |   94.58   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnext/resnext101-32x8d_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnext/resnext101_32x8d_b32x8_imagenet_20210506-23a247d5.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnext/resnext101_32x8d_b32x8_imagenet_20210506-23a247d5.log.json) |
-| ResNeXt-32x4d-152 |   59.95   |   11.8   |   78.88   |   94.33   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnext/resnext152-32x4d_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnext/resnext152_32x4d_b32x8_imagenet_20210524-927787be.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnext/resnext152_32x4d_b32x8_imagenet_20210524-927787be.log.json) |
+|       Model       | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                   Config                   |                                                  Download                                                  |
+| :---------------: | :-------: | :------: | :-------: | :-------: | :----------------------------------------: | :--------------------------------------------------------------------------------------------------------: |
+| ResNeXt-32x4d-50  |   25.03   |   4.27   |   77.90   |   93.66   | [config](./resnext50-32x4d_8xb32_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/resnext/resnext50_32x4d_b32x8_imagenet_20210429-56066e27.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnext/resnext50_32x4d_b32x8_imagenet_20210429-56066e27.log.json) |
+| ResNeXt-32x4d-101 |   44.18   |   8.03   |   78.61   |   94.17   | [config](./resnext101-32x4d_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnext/resnext101_32x4d_b32x8_imagenet_20210506-e0fa3dd5.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnext/resnext101_32x4d_b32x8_imagenet_20210506-e0fa3dd5.log.json) |
+| ResNeXt-32x8d-101 |   88.79   |   16.5   |   79.27   |   94.58   | [config](./resnext101-32x8d_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnext/resnext101_32x8d_b32x8_imagenet_20210506-23a247d5.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnext/resnext101_32x8d_b32x8_imagenet_20210506-23a247d5.log.json) |
+| ResNeXt-32x4d-152 |   59.95   |   11.8   |   78.88   |   94.33   | [config](./resnext152-32x4d_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnext/resnext152_32x4d_b32x8_imagenet_20210524-927787be.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/resnext/resnext152_32x4d_b32x8_imagenet_20210524-927787be.log.json) |
 
 ## Citation
 

configs/seresnet/README.md

@@ -16,10 +16,10 @@ The central building block of convolutional neural networks (CNNs) is the convol
 
 ### ImageNet-1k
 
-|     Model     | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                   Config                                    |                                   Download                                    |
-| :-----------: | :-------: | :------: | :-------: | :-------: | :-------------------------------------------------------------------------: | :---------------------------------------------------------------------------: |
-| SE-ResNet-50  |   28.09   |   4.13   |   77.74   |   93.84   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/seresnet/seresnet50_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/se-resnet/se-resnet50_batch256_imagenet_20200804-ae206104.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/se-resnet/se-resnet50_batch256_imagenet_20200708-657b3c36.log.json) |
-| SE-ResNet-101 |   49.33   |   7.86   |   78.26   |   94.07   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/seresnet/seresnet101_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/se-resnet/se-resnet101_batch256_imagenet_20200804-ba5b51d4.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/se-resnet/se-resnet101_batch256_imagenet_20200708-038a4d04.log.json) |
+|     Model     | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                Config                 |                                                      Download                                                       |
+| :-----------: | :-------: | :------: | :-------: | :-------: | :-----------------------------------: | :-----------------------------------------------------------------------------------------------------------------: |
+| SE-ResNet-50  |   28.09   |   4.13   |   77.74   |   93.84   | [config](./seresnet50_8xb32_in1k.py)  | [model](https://download.openmmlab.com/mmclassification/v0/se-resnet/se-resnet50_batch256_imagenet_20200804-ae206104.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/se-resnet/se-resnet50_batch256_imagenet_20200708-657b3c36.log.json) |
+| SE-ResNet-101 |   49.33   |   7.86   |   78.26   |   94.07   | [config](./seresnet101_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/se-resnet/se-resnet101_batch256_imagenet_20200804-ba5b51d4.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/se-resnet/se-resnet101_batch256_imagenet_20200708-038a4d04.log.json) |
 
 ## Citation
 

configs/shufflenet_v1/README.md

@@ -16,9 +16,9 @@ We introduce an extremely computation-efficient CNN architecture named ShuffleNe
 
 ### ImageNet-1k
 
-|            Model            | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                Config                                |                                Download                                |
-| :-------------------------: | :-------: | :------: | :-------: | :-------: | :------------------------------------------------------------------: | :--------------------------------------------------------------------: |
-| ShuffleNetV1 1.0x (group=3) |   1.87    |  0.146   |   68.13   |   87.81   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/shufflenet_v1/shufflenet-v1-1x_16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/shufflenet_v1/shufflenet_v1_batch1024_imagenet_20200804-5d6cec73.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/shufflenet_v1/shufflenet_v1_batch1024_imagenet_20200804-5d6cec73.log.json) |
+|            Model            | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                   Config                    |                                            Download                                             |
+| :-------------------------: | :-------: | :------: | :-------: | :-------: | :-----------------------------------------: | :---------------------------------------------------------------------------------------------: |
+| ShuffleNetV1 1.0x (group=3) |   1.87    |  0.146   |   68.13   |   87.81   | [config](./shufflenet-v1-1x_16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/shufflenet_v1/shufflenet_v1_batch1024_imagenet_20200804-5d6cec73.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/shufflenet_v1/shufflenet_v1_batch1024_imagenet_20200804-5d6cec73.log.json) |
 
 ## Citation
 

configs/shufflenet_v2/README.md

@@ -16,9 +16,9 @@ Currently, the neural network architecture design is mostly guided by the *indir
 
 ### ImageNet-1k
 
-|       Model       | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                                  Config                                   |                                  Download                                   |
-| :---------------: | :-------: | :------: | :-------: | :-------: | :-----------------------------------------------------------------------: | :-------------------------------------------------------------------------: |
-| ShuffleNetV2 1.0x |   2.28    |  0.149   |   69.55   |   88.92   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/shufflenet_v2/shufflenet-v2-1x_16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/shufflenet_v2/shufflenet_v2_batch1024_imagenet_20200812-5bf4721e.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/shufflenet_v2/shufflenet_v2_batch1024_imagenet_20200804-8860eec9.log.json) |
+|       Model       | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                   Config                    |                                                 Download                                                  |
+| :---------------: | :-------: | :------: | :-------: | :-------: | :-----------------------------------------: | :-------------------------------------------------------------------------------------------------------: |
+| ShuffleNetV2 1.0x |   2.28    |  0.149   |   69.55   |   88.92   | [config](./shufflenet-v2-1x_16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/shufflenet_v2/shufflenet_v2_batch1024_imagenet_20200812-5bf4721e.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/shufflenet_v2/shufflenet_v2_batch1024_imagenet_20200804-8860eec9.log.json) |
 
 ## Citation
 

configs/swin_transformer/README.md

@@ -27,26 +27,26 @@ The pre-trained models on ImageNet-21k are used to fine-tune, and therefore don'
 
 ### ImageNet-1k
 
-|  Model   |   Pretrain   | resolution | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                               Config                               |                               Download                                |
-| :------: | :----------: | :--------: | :-------: | :------: | :-------: | :-------: | :----------------------------------------------------------------: | :-------------------------------------------------------------------: |
-|  Swin-T  | From scratch |  224x224   |   28.29   |   4.36   |   81.18   |   95.61   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/swin_transformer/swin-tiny_16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_tiny_224_b16x64_300e_imagenet_20210616_090925-66df6be6.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_tiny_224_b16x64_300e_imagenet_20210616_090925.log.json) |
-|  Swin-S  | From scratch |  224x224   |   49.61   |   8.52   |   83.02   |   96.29   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/swin_transformer/swin-small_16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_small_224_b16x64_300e_imagenet_20210615_110219-7f9d988b.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_small_224_b16x64_300e_imagenet_20210615_110219.log.json) |
-|  Swin-B  | From scratch |  224x224   |   87.77   |  15.14   |   83.36   |   96.44   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/swin_transformer/swin_base_224_b16x64_300e_imagenet.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_base_224_b16x64_300e_imagenet_20210616_190742-93230b0d.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_base_224_b16x64_300e_imagenet_20210616_190742.log.json) |
-| Swin-S\* | From scratch |  224x224   |   49.61   |   8.52   |   83.21   |   96.25   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/swin_transformer/swin-small_16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_small_patch4_window7_224-cc7a01c9.pth) |
-| Swin-B\* | From scratch |  224x224   |   87.77   |  15.14   |   83.42   |   96.44   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/swin_transformer/swin-base_16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_base_patch4_window7_224-4670dd19.pth) |
-| Swin-B\* | From scratch |  384x384   |   87.90   |  44.49   |   84.49   |   96.95   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/swin_transformer/swin-base_16xb64_in1k-384px.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_base_patch4_window12_384-02c598a4.pth) |
-| Swin-B\* | ImageNet-21k |  224x224   |   87.77   |  15.14   |   85.16   |   97.50   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/swin_transformer/swin-base_16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_base_patch4_window7_224_22kto1k-f967f799.pth) |
-| Swin-B\* | ImageNet-21k |  384x384   |   87.90   |  44.49   |   86.44   |   98.05   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/swin_transformer/swin-base_16xb64_in1k-384px.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_base_patch4_window12_384_22kto1k-d59b0d1d.pth) |
-| Swin-L\* | ImageNet-21k |  224x224   |  196.53   |  34.04   |   86.24   |   97.88   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/swin_transformer/swin-large_16xb64_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_large_patch4_window7_224_22kto1k-5f0996db.pth) |
-| Swin-L\* | ImageNet-21k |  384x384   |  196.74   |  100.04  |   87.25   |   98.25   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/swin_transformer/swin-large_16xb64_in1k-384px.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_large_patch4_window12_384_22kto1k-0a40944b.pth) |
+|  Model   |   Pretrain   | resolution | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                   Config                    |                                           Download                                           |
+| :------: | :----------: | :--------: | :-------: | :------: | :-------: | :-------: | :-----------------------------------------: | :------------------------------------------------------------------------------------------: |
+|  Swin-T  | From scratch |  224x224   |   28.29   |   4.36   |   81.18   |   95.61   |    [config](./swin-tiny_16xb64_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_tiny_224_b16x64_300e_imagenet_20210616_090925-66df6be6.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_tiny_224_b16x64_300e_imagenet_20210616_090925.log.json) |
+|  Swin-S  | From scratch |  224x224   |   49.61   |   8.52   |   83.02   |   96.29   |    [config](./swin-small_16xb64_in1k.py)    | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_small_224_b16x64_300e_imagenet_20210615_110219-7f9d988b.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_small_224_b16x64_300e_imagenet_20210615_110219.log.json) |
+|  Swin-B  | From scratch |  224x224   |   87.77   |  15.14   |   83.36   |   96.44   |    [config](./swin-base_16xb64_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_base_224_b16x64_300e_imagenet_20210616_190742-93230b0d.pth)  \| [log](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_base_224_b16x64_300e_imagenet_20210616_190742.log.json) |
+| Swin-S\* | From scratch |  224x224   |   49.61   |   8.52   |   83.21   |   96.25   |    [config](./swin-small_16xb64_in1k.py)    | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_small_patch4_window7_224-cc7a01c9.pth) |
+| Swin-B\* | From scratch |  224x224   |   87.77   |  15.14   |   83.42   |   96.44   |    [config](./swin-base_16xb64_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_base_patch4_window7_224-4670dd19.pth) |
+| Swin-B\* | From scratch |  384x384   |   87.90   |  44.49   |   84.49   |   96.95   | [config](./swin-base_16xb64_in1k-384px.py)  | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_base_patch4_window12_384-02c598a4.pth) |
+| Swin-B\* | ImageNet-21k |  224x224   |   87.77   |  15.14   |   85.16   |   97.50   |    [config](./swin-base_16xb64_in1k.py)     | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_base_patch4_window7_224_22kto1k-f967f799.pth) |
+| Swin-B\* | ImageNet-21k |  384x384   |   87.90   |  44.49   |   86.44   |   98.05   | [config](./swin-base_16xb64_in1k-384px.py)  | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_base_patch4_window12_384_22kto1k-d59b0d1d.pth) |
+| Swin-L\* | ImageNet-21k |  224x224   |  196.53   |  34.04   |   86.24   |   97.88   |    [config](./swin-large_16xb64_in1k.py)    | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_large_patch4_window7_224_22kto1k-5f0996db.pth) |
+| Swin-L\* | ImageNet-21k |  384x384   |  196.74   |  100.04  |   87.25   |   98.25   | [config](./swin-large_16xb64_in1k-384px.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_large_patch4_window12_384_22kto1k-0a40944b.pth) |
 
 *Models with * are converted from the [official repo](https://github.com/microsoft/Swin-Transformer#main-results-on-imagenet-with-pretrained-models). The config files of these models are only for validation. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
 
 ### CUB-200-2011
 
-| Model  |                       Pretrain                        | resolution | Params(M) | Flops(G) | Top-1 (%) |                       Config                        |                        Download                        |
-| :----: | :---------------------------------------------------: | :--------: | :-------: | :------: | :-------: | :-------------------------------------------------: | :----------------------------------------------------: |
-| Swin-L | [ImageNet-21k](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin-base_3rdparty_in21k-384px.pth) |  384x384   |  195.51   |  100.04  |   91.87   | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/swin_transformer/swin-large_8xb8_cub_384px.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin-large_8xb8_cub_384px_20220307-1bbaee6a.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin-large_8xb8_cub_384px_20220307-1bbaee6a.log.json) |
+| Model  |                          Pretrain                           | resolution | Params(M) | Flops(G) | Top-1 (%) |                  Config                  |                          Download                           |
+| :----: | :---------------------------------------------------------: | :--------: | :-------: | :------: | :-------: | :--------------------------------------: | :---------------------------------------------------------: |
+| Swin-L | [ImageNet-21k](https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin-base_3rdparty_in21k-384px.pth) |  384x384   |  195.51   |  100.04  |   91.87   | [config](./swin-large_8xb8_cub_384px.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin-large_8xb8_cub_384px_20220307-1bbaee6a.pth) \| [log](https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin-large_8xb8_cub_384px_20220307-1bbaee6a.log.json) |
 
 ## Citation
 

configs/swin_transformer_v2/README.md

@@ -0,0 +1,58 @@
+# Swin Transformer V2
+
+> [Swin Transformer V2: Scaling Up Capacity and Resolution](https://arxiv.org/abs/2111.09883.pdf)
+
+<!-- [ALGORITHM] -->
+
+## Abstract
+
+Large-scale NLP models have been shown to significantly improve the performance on language tasks with no signs of saturation. They also demonstrate amazing few-shot capabilities like that of human beings. This paper aims to explore large-scale models in computer vision. We tackle three major issues in training and application of large vision models, including training instability, resolution gaps between pre-training and fine-tuning, and hunger on labelled data. Three main techniques are proposed: 1) a residual-post-norm method combined with cosine attention to improve training stability; 2) A log-spaced continuous position bias method to effectively transfer models pre-trained using low-resolution images to downstream tasks with high-resolution inputs; 3) A self-supervised pre-training method, SimMIM, to reduce the needs of vast labeled images. Through these techniques, this paper successfully trained a 3 billion-parameter Swin Transformer V2 model, which is the largest dense vision model to date, and makes it capable of training with images of up to 1,536×1,536 resolution. It set new performance records on 4 representative vision tasks, including ImageNet-V2 image classification, COCO object detection, ADE20K semantic segmentation, and Kinetics-400 video action classification. Also note our training is much more efficient than that in Google's billion-level visual models, which consumes 40 times less labelled data and 40 times less training time.
+
+<div align=center>
+<img src="https://user-images.githubusercontent.com/42952108/180748696-ee7ed23d-7fee-4ccf-9eb5-f117db228a42.png" width="100%"/>
+</div>
+
+## Results and models
+
+### ImageNet-21k
+
+The pre-trained models on ImageNet-21k are used to fine-tune, and therefore don't have evaluation results.
+
+|  Model   | resolution | Params(M) | Flops(G) |                                                                 Download                                                                 |
+| :------: | :--------: | :-------: | :------: | :--------------------------------------------------------------------------------------------------------------------------------------: |
+| Swin-B\* |  192x192   |   87.92   |   8.51   | [model](https://download.openmmlab.com/mmclassification/v0/swin-v2/pretrain/swinv2-base-w12_3rdparty_in21k-192px_20220803-f7dc9763.pth)  |
+| Swin-L\* |  192x192   |  196.74   |  19.04   | [model](https://download.openmmlab.com/mmclassification/v0/swin-v2/pretrain/swinv2-large-w12_3rdparty_in21k-192px_20220803-d9073fee.pth) |
+
+### ImageNet-1k
+
+|  Model   |   Pretrain   | resolution | window | Params(M) | Flops(G) | Top-1 (%) | Top-5 (%) |                           Config                            |                                Download                                |
+| :------: | :----------: | :--------: | :----: | :-------: | :------: | :-------: | :-------: | :---------------------------------------------------------: | :--------------------------------------------------------------------: |
+| Swin-T\* | From scratch |  256x256   |  8x8   |   28.35   |   4.35   |   81.76   |   95.87   |       [config](./swinv2-tiny-w8_16xb64_in1k-256px.py)       | [model](https://download.openmmlab.com/mmclassification/v0/swin-v2/swinv2-tiny-w8_3rdparty_in1k-256px_20220803-e318968f.pth) |
+| Swin-T\* | From scratch |  256x256   | 16x16  |   28.35   |   4.4    |   82.81   |   96.23   |      [config](./swinv2-tiny-w16_16xb64_in1k-256px.py)       | [model](https://download.openmmlab.com/mmclassification/v0/swin-v2/swinv2-tiny-w16_3rdparty_in1k-256px_20220803-9651cdd7.pth) |
+| Swin-S\* | From scratch |  256x256   |  8x8   |   49.73   |   8.45   |   83.74   |   96.6    |      [config](./swinv2-small-w8_16xb64_in1k-256px.py)       | [model](https://download.openmmlab.com/mmclassification/v0/swin-v2/swinv2-small-w8_3rdparty_in1k-256px_20220803-b01a4332.pth) |
+| Swin-S\* | From scratch |  256x256   | 16x16  |   49.73   |   8.57   |   84.13   |   96.83   |      [config](./swinv2-small-w16_16xb64_in1k-256px.py)      | [model](https://download.openmmlab.com/mmclassification/v0/swin-v2/swinv2-small-w16_3rdparty_in1k-256px_20220803-b707d206.pth) |
+| Swin-B\* | From scratch |  256x256   |  8x8   |   87.92   |  14.99   |   84.2    |   96.86   |       [config](./swinv2-base-w8_16xb64_in1k-256px.py)       | [model](https://download.openmmlab.com/mmclassification/v0/swin-v2/swinv2-base-w8_3rdparty_in1k-256px_20220803-8ff28f2b.pth) |
+| Swin-B\* | From scratch |  256x256   | 16x16  |   87.92   |  15.14   |   84.6    |   97.05   |      [config](./swinv2-base-w16_16xb64_in1k-256px.py)       | [model](https://download.openmmlab.com/mmclassification/v0/swin-v2/swinv2-base-w16_3rdparty_in1k-256px_20220803-5a1886b7.pth) |
+| Swin-B\* | ImageNet-21k |  256x256   | 16x16  |   87.92   |  15.14   |   86.17   |   97.88   | [config](./swinv2-base-w16_in21k-pre_16xb64_in1k-256px.py)  | [model](https://download.openmmlab.com/mmclassification/v0/swin-v2/swinv2-base-w16_in21k-pre_3rdparty_in1k-256px_20220803-8d7aa8ad.pth) |
+| Swin-B\* | ImageNet-21k |  384x384   | 24x24  |   87.92   |  34.07   |   87.14   |   98.23   | [config](./swinv2-base-w24_in21k-pre_16xb64_in1k-384px.py)  | [model](https://download.openmmlab.com/mmclassification/v0/swin-v2/swinv2-base-w24_in21k-pre_3rdparty_in1k-384px_20220803-44eb70f8.pth) |
+| Swin-L\* | ImageNet-21k |  256X256   | 16x16  |  196.75   |  33.86   |   86.93   |   98.06   | [config](./swinv2-large-w16_in21k-pre_16xb64_in1k-256px.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-v2/swinv2-large-w16_in21k-pre_3rdparty_in1k-256px_20220803-c40cbed7.pth) |
+| Swin-L\* | ImageNet-21k |  384x384   | 24x24  |  196.75   |   76.2   |   87.59   |   98.27   | [config](./swinv2-large-w24_in21k-pre_16xb64_in1k-384px.py) | [model](https://download.openmmlab.com/mmclassification/v0/swin-v2/swinv2-large-w24_in21k-pre_3rdparty_in1k-384px_20220803-3b36c165.pth) |
+
+*Models with * are converted from the [official repo](https://github.com/microsoft/Swin-Transformer#main-results-on-imagenet-with-pretrained-models). The config files of these models are only for validation. We don't ensure these config files' training accuracy and welcome you to contribute your reproduction results.*
+
+*ImageNet-21k pretrained models with input resolution of 256x256 and 384x384 both fine-tuned from the same pre-training model using a smaller input resolution of 192x192.*
+
+## Citation
+
+```
+@article{https://doi.org/10.48550/arxiv.2111.09883,
+  doi = {10.48550/ARXIV.2111.09883},
+  url = {https://arxiv.org/abs/2111.09883},
+  author = {Liu, Ze and Hu, Han and Lin, Yutong and Yao, Zhuliang and Xie, Zhenda and Wei, Yixuan and Ning, Jia and Cao, Yue and Zhang, Zheng and Dong, Li and Wei, Furu and Guo, Baining},
+  keywords = {Computer Vision and Pattern Recognition (cs.CV), FOS: Computer and information sciences, FOS: Computer and information sciences},
+  title = {Swin Transformer V2: Scaling Up Capacity and Resolution},
+  publisher = {arXiv},
+  year = {2021},
+  copyright = {Creative Commons Attribution 4.0 International}
+}
+```