fix some bug

recognition/arcface_paddle/README.md

@@ -67,19 +67,15 @@ If you want to use customed dataset, you can arrange your data according to the
 python -m paddle.distributed.launch --gpus=0 tools/train.py \
     --config_file configs/ms1mv2_mobileface.py \
     --is_static False \
-    --backbone MobileFaceNet_128 \
-    --classifier LargeScaleClassifier \
     --embedding_size 128 \
     --sample_ratio 1.0 \
     --loss ArcFace \
-    --batch_size 1024 \
+    --batch_size 512 \
     --dataset MS1M_v2 \
     --num_classes 85742 \
     --data_dir MS1M_v2/ \
     --label_file MS1M_v2/label.txt \
-    --fp16 False \
-    --train_unit 'epoch' \
-    --output ./MS1M_v2_arcface_MobileFaceNet_128_1.0_fp32
+    --fp16 False
 ```
 
 ### 4.2 Single node, 8 GPUs:
@@ -156,7 +152,7 @@ sh scripts/inference.sh
 
 | Model structure           | lfw    | cfp_fp  | agedb30 | CPU time cost | GPU time cost | Inference model |
 | ------------------------- | ------ | ------- | ------- | -------| -------- |---- |
-| MobileFace-Paddle      | 0.9945 | 0.9343  | 0.9613  | 4.3ms  | 2.3ms    | [download link](https://paddle-model-ecology.bj.bcebos.com/model/insight-face/mobileface_v1.0_infer.tar)  |
+| MobileFace-Paddle      | 0.9952 | 0.9280  | 0.9612  | 4.3ms  | 2.3ms    | [download link](https://paddle-model-ecology.bj.bcebos.com/model/insight-face/mobileface_v1.0_infer.tar)  |
 | MobileFace-mxnet       | 0.9950 | 0.8894  | 0.9591  | 7.3ms  | 4.7ms    | -   |
 
 * Note: MobileFaceNet-Paddle training using MobileFaceNet_128

recognition/arcface_paddle/configs/ms1mv2_mobileface.py

@@ -0,0 +1,54 @@
+# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from easydict import EasyDict as edict
+
+config = edict()
+config.is_static = False
+config.backbone = 'MobileFaceNet_128'
+config.classifier = 'LargeScaleClassifier'
+config.embedding_size = 128
+config.model_parallel = True
+config.sample_ratio = 1.0
+config.loss = 'ArcFace'
+config.dropout = 0.0
+
+config.lr = 0.1  # for global batch size = 512
+config.lr_decay = 0.1
+config.weight_decay = 5e-4
+config.momentum = 0.9
+config.train_unit = 'epoch'  # 'step' or 'epoch'
+config.warmup_num = 0
+config.train_num = 25
+config.decay_boundaries = [10, 16, 22]
+
+config.use_synthetic_dataset = False
+config.dataset = "MS1M_v2"
+config.data_dir = "./MS1M_v2"
+config.label_file = "./MS1M_v2/label.txt"
+config.is_bin = False
+config.num_classes = 85742  # 85742 for MS1M_v2, 93431 for MS1M_v3
+config.batch_size = 128  # global batch size 1024 of 8 GPU
+config.num_workers = 8
+
+config.do_validation_while_train = True
+config.validation_interval_step = 2000
+config.val_targets = ["lfw", "cfp_fp", "agedb_30"]
+
+config.logdir = './log'
+config.log_interval_step = 100
+config.output = './MS1M_v2_arcface_MobileFaceNet_128_0.1'
+config.resume = False
+config.checkpoint_dir = None
+config.max_num_last_checkpoint = 1

recognition/arcface_paddle/dynamic/backbones/mobilefacenet.py

@@ -110,18 +110,18 @@ class MobileFaceNet(nn.Layer):
                 n = m.weight.shape[1] * m.weight.shape[2] * m.weight.shape[3]
                 m.weight = paddle.create_parameter(
                     shape=m.weight.shape,
-                    dtype='float32',
+                    dtype=m.weight.dtype,
                     default_initializer=nn.initializer.Normal(
                         mean=0.0, std=math.sqrt(2.0 / n)))
-                # nn.init.normal_(m.weight, 0, 0.1)
+                
             elif isinstance(m, (nn.BatchNorm, nn.BatchNorm2D, nn.GroupNorm)):
                 m.weight = paddle.create_parameter(
                     shape=m.weight.shape,
-                    dtype='float32',
+                    dtype=m.weight.dtype,
                     default_initializer=nn.initializer.Constant(value=1.0))
                 m.bias = paddle.create_parameter(
                     shape=m.bias.shape,
-                    dtype='float32',
+                    dtype=m.bias.dtype,
                     default_initializer=nn.initializer.Constant(value=0.0))
 
     def _make_layer(self, block, setting):

recognition/arcface_paddle/dynamic/classifiers/lsc.py

@@ -18,7 +18,6 @@ import os
 import paddle
 import paddle.nn as nn
 
-
 class LargeScaleClassifier(nn.Layer):
     """
     Author: {Xiang An, Yang Xiao, XuHan Zhu} in DeepGlint,
@@ -77,10 +76,11 @@ class LargeScaleClassifier(nn.Layer):
             self.weight.stop_gradient = True
 
     def step(self, optimizer):
-        warnings.warn(
-            "Explicitly call the function paddle._C_ops.sparse_momentum is a temporary manner. "
-            "We will merge it to optimizer in the future, please don't follow.")
         if int(self.sample_ratio) < 1:
+            warnings.warn(
+                "Explicitly call the function paddle._C_ops.sparse_momentum is a temporary manner. "
+                "We will merge it to optimizer in the future, please don't follow.")
+            
             found_inf = paddle.logical_not(
                 paddle.all(paddle.isfinite(self._parameter_list[0].grad)))
             if found_inf:

recognition/arcface_paddle/dynamic/utils/amp.py

@@ -43,7 +43,7 @@ class LSCGradScaler(GradScaler):
 #         if self._scale >= self.max_loss_scaling:
 #             self._scale = paddle.to_tensor([self.max_loss_scaling], dtype='float32')
 
-#  unscale the grad
+        # unscale the grad
         self._unscale(optimizer)
 
         if self._found_inf:
@@ -92,6 +92,7 @@ class LSCGradScaler(GradScaler):
                 self._found_inf = paddle.logical_not(
                     paddle.all(paddle.isfinite(grad)))
                 if self._found_inf:
+                    print('Found inf or nan in classifier, dtype is', dtype)
                     return
 
         for dtype in param_grads_dict:

recognition/arcface_paddle/dynamic/utils/io.py

@@ -143,7 +143,7 @@ class Checkpoint(object):
             tensor = paddle.load(path, return_numpy=True)
             if dtype:
                 assert dtype in ['float32', 'float16']
-                tensor = tensor.astype('float32')
+                tensor = tensor.astype(dtype)
 
             if 'dist@' in name and '@rank@' in name:
                 if '.w' in name and 'velocity' not in name:

recognition/arcface_paddle/scripts/inference.sh

@@ -16,9 +16,9 @@ python tools/inference.py \
     --export_type paddle \
     --model_file MS1M_v3_arcface_static_128_fp16_0.1/FresResNet50/exported_model/FresResNet50.pdmodel \
     --params_file MS1M_v3_arcface_static_128_fp16_0.1/FresResNet50/exported_model/FresResNet50.pdiparams \
-    --image_path /wangguoxia/plsc/MS1M_v3/images/00000001.jpg
+    --image_path MS1M_v3/images/00000001.jpg
 
 python tools/inference.py \
     --export_type onnx \
     --onnx_file MS1M_v3_arcface_static_128_fp16_0.1/FresResNet50/exported_model/FresResNet50.onnx \
-    --image_path /wangguoxia/plsc/MS1M_v3/images/00000001.jpg
+    --image_path MS1M_v3/images/00000001.jpg

recognition/arcface_paddle/scripts/train_dynamic.sh

@@ -25,8 +25,8 @@ python -m paddle.distributed.launch --gpus=0,1,2,3,4,5,6,7 tools/train.py \
     --batch_size 128 \
     --dataset MS1M_v3 \
     --num_classes 93431 \
-    --data_dir /wangguoxia/plsc/MS1M_v3/ \
-    --label_file /wangguoxia/plsc/MS1M_v3/label.txt \
+    --data_dir MS1M_v3/ \
+    --label_file MS1M_v3/label.txt \
     --is_bin False \
     --log_interval_step 100 \
     --validation_interval_step 2000 \

recognition/arcface_paddle/scripts/train_static.sh

@@ -25,8 +25,8 @@ python -m paddle.distributed.launch --gpus=0,1,2,3,4,5,6,7 tools/train.py \
     --batch_size 128 \
     --dataset MS1M_v3 \
     --num_classes 93431 \
-    --data_dir /wangguoxia/plsc/MS1M_v3/ \
-    --label_file /wangguoxia/plsc/MS1M_v3/label.txt \
+    --data_dir MS1M_v3/ \
+    --label_file MS1M_v3/label.txt \
     --is_bin False \
     --log_interval_step 100 \
     --validation_interval_step 2000 \

recognition/arcface_paddle/scripts/validation_dynamic.sh

@@ -17,6 +17,6 @@ python tools/validation.py \
     --backbone FresResNet50 \
     --embedding_size 512 \
     --checkpoint_dir MS1M_v3_arcface_dynamic_128_fp16_0.1/FresResNet50/24 \
-    --data_dir /wangguoxia/plsc/MS1M_v3/ \
+    --data_dir MS1M_v3/ \
     --val_targets lfw,cfp_fp,agedb_30 \
     --batch_size 128

recognition/arcface_paddle/scripts/validation_static.sh

@@ -17,6 +17,6 @@ python tools/validation.py \
     --backbone FresResNet50 \
     --embedding_size 512 \
     --checkpoint_dir MS1M_v3_arcface_static_128_fp16_0.1/FresResNet50/24 \
-    --data_dir /wangguoxia/plsc/MS1M_v3/ \
+    --data_dir MS1M_v3/ \
     --val_targets lfw,cfp_fp,agedb_30 \
     --batch_size 128

recognition/arcface_paddle/static/utils/io.py

@@ -120,7 +120,7 @@ class Checkpoint(object):
             tensor = paddle.load(path, return_numpy=True)
             if dtype:
                 assert dtype in ['float32', 'float16']
-                tensor = tensor.astype('float32')
+                tensor = tensor.astype(dtype)
 
             if 'dist@' in name and '@rank@' in name:
                 if '.w' in name and 'velocity' not in name: