modified batch_size and update_freq & add more tipc_test configs

ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_S.yaml

@@ -55,7 +55,6 @@ Optimizer:
     eta_min: 0.0002
     warmup_epoch: 1  # 3000 iterations
     warmup_start_lr: 0.0002
-    # by_epoch: True
 
 # data loader for train and eval
 DataLoader:

ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_S_L2.yaml

@@ -14,6 +14,7 @@ Global:
   image_shape: [3, 256, 256]
   save_inference_dir: ./inference
   use_dali: False
+  update_freq: 3  # for 4 gpus
 
 # mixed precision training
 AMP:
@@ -49,13 +50,11 @@ Optimizer:
   epsilon: 1e-8
   weight_decay: 0.01
   lr:
-    # for 8 cards
     name: Cosine
-    learning_rate: 0.002
+    learning_rate: 0.002  # for total batch size 384
     eta_min: 0.0002
     warmup_epoch: 1  # 3000 iterations
     warmup_start_lr: 0.0002
-    # by_epoch: True
 
 # data loader for train and eval
 DataLoader:
@@ -86,7 +85,7 @@ DataLoader:
       scales: [256, 160, 192, 224, 288, 320]
       # first_bs: batch size for the first image resolution in the scales list
       # divide_factor: to ensure the width and height dimensions can be devided by downsampling multiple
-      first_bs: 48
+      first_bs: 32
       divided_factor: 32
       is_training: True
     loader:

ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_XS.yaml

@@ -14,6 +14,7 @@ Global:
   image_shape: [3, 256, 256]
   save_inference_dir: ./inference
   use_dali: False
+  update_freq: 3  # for 4 gpus
 
 # mixed precision training
 AMP:
@@ -49,13 +50,11 @@ Optimizer:
   epsilon: 1e-8
   weight_decay: 0.01
   lr:
-    # for 8 cards
     name: Cosine
-    learning_rate: 0.002
+    learning_rate: 0.002  # for total batch size 384
     eta_min: 0.0002
     warmup_epoch: 1  # 3000 iterations
     warmup_start_lr: 0.0002
-    # by_epoch: True
 
 # data loader for train and eval
 DataLoader:
@@ -86,7 +85,7 @@ DataLoader:
       scales: [256, 160, 192, 224, 288, 320]
       # first_bs: batch size for the first image resolution in the scales list
       # divide_factor: to ensure the width and height dimensions can be devided by downsampling multiple
-      first_bs: 48
+      first_bs: 32
       divided_factor: 32
       is_training: True
     loader:

ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_XS_L2.yaml

@@ -14,6 +14,7 @@ Global:
   image_shape: [3, 256, 256]
   save_inference_dir: ./inference
   use_dali: False
+  update_freq: 3  # for 4 gpus
 
 # mixed precision training
 AMP:
@@ -49,13 +50,11 @@ Optimizer:
   epsilon: 1e-8
   weight_decay: 0.01
   lr:
-    # for 8 cards
     name: Cosine
-    learning_rate: 0.002
+    learning_rate: 0.002  # for total batch size 384
     eta_min: 0.0002
     warmup_epoch: 1  # 3000 iterations
     warmup_start_lr: 0.0002
-    # by_epoch: True
 
 # data loader for train and eval
 DataLoader:
@@ -86,7 +85,7 @@ DataLoader:
       scales: [256, 160, 192, 224, 288, 320]
       # first_bs: batch size for the first image resolution in the scales list
       # divide_factor: to ensure the width and height dimensions can be devided by downsampling multiple
-      first_bs: 48
+      first_bs: 32
       divided_factor: 32
       is_training: True
     loader:

ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_XXS.yaml

@@ -14,6 +14,7 @@ Global:
   image_shape: [3, 256, 256]
   save_inference_dir: ./inference
   use_dali: False
+  update_freq: 3  # for 4 gpus
 
 # mixed precision training
 AMP:
@@ -49,13 +50,11 @@ Optimizer:
   epsilon: 1e-8
   weight_decay: 0.01
   lr:
-    # for 8 cards
     name: Cosine
-    learning_rate: 0.002
+    learning_rate: 0.002  # for total batch size 384
     eta_min: 0.0002
     warmup_epoch: 1  # 3000 iterations
     warmup_start_lr: 0.0002
-    # by_epoch: True
 
 # data loader for train and eval
 DataLoader:
@@ -86,7 +85,7 @@ DataLoader:
       scales: [256, 160, 192, 224, 288, 320]
       # first_bs: batch size for the first image resolution in the scales list
       # divide_factor: to ensure the width and height dimensions can be devided by downsampling multiple
-      first_bs: 48
+      first_bs: 32
       divided_factor: 32
       is_training: True
     loader:

ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_XXS_L2.yaml

@@ -14,6 +14,7 @@ Global:
   image_shape: [3, 256, 256]
   save_inference_dir: ./inference
   use_dali: False
+  update_freq: 3  # for 4 gpus
 
 # mixed precision training
 AMP:
@@ -49,13 +50,11 @@ Optimizer:
   epsilon: 1e-8
   weight_decay: 0.01
   lr:
-    # for 8 cards
     name: Cosine
-    learning_rate: 0.002
+    learning_rate: 0.002  # for total batch size 384
     eta_min: 0.0002
     warmup_epoch: 1  # 3000 iterations
     warmup_start_lr: 0.0002
-    # by_epoch: True
 
 # data loader for train and eval
 DataLoader:
@@ -86,7 +85,7 @@ DataLoader:
       scales: [256, 160, 192, 224, 288, 320]
       # first_bs: batch size for the first image resolution in the scales list
       # divide_factor: to ensure the width and height dimensions can be devided by downsampling multiple
-      first_bs: 48
+      first_bs: 32
       divided_factor: 32
       is_training: True
     loader:

ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_x0_5.yaml

@@ -14,6 +14,7 @@ Global:
   image_shape: [3, 256, 256]
   save_inference_dir: ./inference
   use_dali: False
+  update_freq: 3  # for 4 gpus
 
 # mixed precision training
 AMP:
@@ -50,13 +51,11 @@ Optimizer:
   weight_decay: 0.05
   one_dim_param_no_weight_decay: True
   lr:
-    # for 8 cards
     name: Cosine
-    learning_rate: 0.002
+    learning_rate: 0.002  # for total batch size 1020
     eta_min: 0.0002
     warmup_epoch: 16  # 20000 iterations
     warmup_start_lr: 1e-6
-    # by_epoch: True
   clip_norm: 10
 
 # data loader for train and eval
@@ -104,7 +103,7 @@ DataLoader:
               prob: 0.25
     sampler:
       name: DistributedBatchSampler
-      batch_size: 128
+      batch_size: 85
       drop_last: False
       shuffle: True
     loader:

ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_x0_75.yaml

@@ -14,6 +14,7 @@ Global:
   image_shape: [3, 256, 256]
   save_inference_dir: ./inference
   use_dali: False
+  update_freq: 3  # for 4 gpus
 
 # mixed precision training
 AMP:
@@ -50,13 +51,11 @@ Optimizer:
   weight_decay: 0.05
   one_dim_param_no_weight_decay: True
   lr:
-    # for 8 cards
     name: Cosine
-    learning_rate: 0.002
+    learning_rate: 0.002  # for total batch size 1020
     eta_min: 0.0002
     warmup_epoch: 16  # 20000 iterations
     warmup_start_lr: 1e-6
-    # by_epoch: True
   clip_norm: 10
 
 # data loader for train and eval
@@ -104,7 +103,7 @@ DataLoader:
               prob: 0.25
     sampler:
       name: DistributedBatchSampler
-      batch_size: 128
+      batch_size: 85
       drop_last: False
       shuffle: True
     loader:

ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_x1_0.yaml

@@ -14,6 +14,7 @@ Global:
   image_shape: [3, 256, 256]
   save_inference_dir: ./inference
   use_dali: False
+  update_freq: 3  # for 4 gpus
 
 # mixed precision training
 AMP:
@@ -50,13 +51,11 @@ Optimizer:
   weight_decay: 0.05
   one_dim_param_no_weight_decay: True
   lr:
-    # for 8 cards
     name: Cosine
-    learning_rate: 0.002
+    learning_rate: 0.002  # for total batch size 1020
     eta_min: 0.0002
     warmup_epoch: 16  # 20000 iterations
     warmup_start_lr: 1e-6
-    # by_epoch: True
   clip_norm: 10
 
 # data loader for train and eval
@@ -104,7 +103,7 @@ DataLoader:
               prob: 0.25
     sampler:
       name: DistributedBatchSampler
-      batch_size: 128
+      batch_size: 85
       drop_last: False
       shuffle: True
     loader:

test_tipc/configs/MobileViTv3/MobileViTv3_S_L2_train_infer_python.txt

+===========================train_params===========================
+model_name:MobileViTv3_S_L2
+python:python3.7
+gpu_list:0|0,1
+-o Global.device:gpu
+-o Global.auto_cast:null
+-o Global.epochs:lite_train_lite_infer=2|whole_train_whole_infer=120
+-o Global.output_dir:./output/
+-o DataLoader.Train.sampler.first_bs:8
+-o Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./dataset/ILSVRC2012/val
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_S_L2.yaml -o Global.seed=1234 -o DataLoader.Train.sampler.shuffle=False -o DataLoader.Train.loader.num_workers=0 -o DataLoader.Train.loader.use_shared_memory=False -o Global.print_batch_step=1 -o Global.eval_during_train=False -o Global.save_interval=2
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:tools/eval.py -c ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_S_L2.yaml
+null:null
+##
+===========================infer_params==========================
+-o Global.save_inference_dir:./inference
+-o Global.pretrained_model:
+norm_export:tools/export_model.py -c ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_S_L2.yaml
+quant_export:null
+fpgm_export:null
+distill_export:null
+kl_quant:null
+export2:null
+inference_dir:null
+infer_model:../inference/
+infer_export:True
+infer_quant:Fasle
+inference:python/predict_cls.py -c configs/inference_cls.yaml -o PreProcess.transform_ops.0.ResizeImage.resize_short=288 -o PreProcess.transform_ops.1.CropImage.size=256 -o PreProcess.transform_ops.2.NormalizeImage.mean=[0.,0.,0.] -o PreProcess.transform_ops.2.NormalizeImage.std=[1.,1.,1.]
+-o Global.use_gpu:True|False
+-o Global.enable_mkldnn:False
+-o Global.cpu_num_threads:1
+-o Global.batch_size:1
+-o Global.use_tensorrt:False
+-o Global.use_fp16:False
+-o Global.inference_model_dir:../inference
+-o Global.infer_imgs:../dataset/ILSVRC2012/val/ILSVRC2012_val_00000001.JPEG
+-o Global.save_log_path:null
+-o Global.benchmark:False
+null:null
+null:null
+===========================train_benchmark_params==========================
+batch_size:128
+fp_items:fp32
+epoch:1
+model_type:norm_train
+--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
+flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,256,256]}]

test_tipc/configs/MobileViTv3/MobileViTv3_XS_L2_train_infer_python.txt

+===========================train_params===========================
+model_name:MobileViTv3_XS_L2
+python:python3.7
+gpu_list:0|0,1
+-o Global.device:gpu
+-o Global.auto_cast:null
+-o Global.epochs:lite_train_lite_infer=2|whole_train_whole_infer=120
+-o Global.output_dir:./output/
+-o DataLoader.Train.sampler.first_bs:8
+-o Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./dataset/ILSVRC2012/val
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_XS_L2.yaml -o Global.seed=1234 -o DataLoader.Train.sampler.shuffle=False -o DataLoader.Train.loader.num_workers=0 -o DataLoader.Train.loader.use_shared_memory=False -o Global.print_batch_step=1 -o Global.eval_during_train=False -o Global.save_interval=2
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:tools/eval.py -c ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_XS_L2.yaml
+null:null
+##
+===========================infer_params==========================
+-o Global.save_inference_dir:./inference
+-o Global.pretrained_model:
+norm_export:tools/export_model.py -c ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_XS_L2.yaml
+quant_export:null
+fpgm_export:null
+distill_export:null
+kl_quant:null
+export2:null
+inference_dir:null
+infer_model:../inference/
+infer_export:True
+infer_quant:Fasle
+inference:python/predict_cls.py -c configs/inference_cls.yaml -o PreProcess.transform_ops.0.ResizeImage.resize_short=288 -o PreProcess.transform_ops.1.CropImage.size=256 -o PreProcess.transform_ops.2.NormalizeImage.mean=[0.,0.,0.] -o PreProcess.transform_ops.2.NormalizeImage.std=[1.,1.,1.]
+-o Global.use_gpu:True|False
+-o Global.enable_mkldnn:False
+-o Global.cpu_num_threads:1
+-o Global.batch_size:1
+-o Global.use_tensorrt:False
+-o Global.use_fp16:False
+-o Global.inference_model_dir:../inference
+-o Global.infer_imgs:../dataset/ILSVRC2012/val/ILSVRC2012_val_00000001.JPEG
+-o Global.save_log_path:null
+-o Global.benchmark:False
+null:null
+null:null
+===========================train_benchmark_params==========================
+batch_size:128
+fp_items:fp32
+epoch:1
+model_type:norm_train
+--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
+flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,256,256]}]

test_tipc/configs/MobileViTv3/MobileViTv3_XXS_L2_train_infer_python.txt

+===========================train_params===========================
+model_name:MobileViTv3_XXS_L2
+python:python3.7
+gpu_list:0|0,1
+-o Global.device:gpu
+-o Global.auto_cast:null
+-o Global.epochs:lite_train_lite_infer=2|whole_train_whole_infer=120
+-o Global.output_dir:./output/
+-o DataLoader.Train.sampler.first_bs:8
+-o Global.pretrained_model:null
+train_model_name:latest
+train_infer_img_dir:./dataset/ILSVRC2012/val
+null:null
+##
+trainer:norm_train
+norm_train:tools/train.py -c ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_XXS_L2.yaml -o Global.seed=1234 -o DataLoader.Train.sampler.shuffle=False -o DataLoader.Train.loader.num_workers=0 -o DataLoader.Train.loader.use_shared_memory=False -o Global.print_batch_step=1 -o Global.eval_during_train=False -o Global.save_interval=2
+pact_train:null
+fpgm_train:null
+distill_train:null
+null:null
+null:null
+##
+===========================eval_params===========================
+eval:tools/eval.py -c ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_XXS_L2.yaml
+null:null
+##
+===========================infer_params==========================
+-o Global.save_inference_dir:./inference
+-o Global.pretrained_model:
+norm_export:tools/export_model.py -c ppcls/configs/ImageNet/MobileViTv3/MobileViTv3_XXS_L2.yaml
+quant_export:null
+fpgm_export:null
+distill_export:null
+kl_quant:null
+export2:null
+inference_dir:null
+infer_model:../inference/
+infer_export:True
+infer_quant:Fasle
+inference:python/predict_cls.py -c configs/inference_cls.yaml -o PreProcess.transform_ops.0.ResizeImage.resize_short=288 -o PreProcess.transform_ops.1.CropImage.size=256 -o PreProcess.transform_ops.2.NormalizeImage.mean=[0.,0.,0.] -o PreProcess.transform_ops.2.NormalizeImage.std=[1.,1.,1.]
+-o Global.use_gpu:True|False
+-o Global.enable_mkldnn:False
+-o Global.cpu_num_threads:1
+-o Global.batch_size:1
+-o Global.use_tensorrt:False
+-o Global.use_fp16:False
+-o Global.inference_model_dir:../inference
+-o Global.infer_imgs:../dataset/ILSVRC2012/val/ILSVRC2012_val_00000001.JPEG
+-o Global.save_log_path:null
+-o Global.benchmark:False
+null:null
+null:null
+===========================train_benchmark_params==========================
+batch_size:128
+fp_items:fp32
+epoch:1
+model_type:norm_train
+--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
+flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
+===========================infer_benchmark_params==========================
+random_infer_input:[{float32,[3,256,256]}]