Update readme

README.md

@@ -7,13 +7,20 @@ Zachary Teed and Jia Deng<br/>
 
 <img src="RAFT.png">
 
+## Todos
+ - [ ] Train supervised to 200 (See if scratch is good enough)
+ - [ ] Train self-supervised to 200 (Using scratch / transfer)
+ - [ ] Run experiments of ctx size = { 96, 64 }
+ - [ ] Add self-collected dataset experiments
+ - [ ] Try-out context attention module ([GMFlowNet](https://github.com/xiaofeng94/GMFlowNet))
+
 ## Requirements
 The code has been tested with PyTorch 1.6 and Cuda 10.1.
 ```Shell
 conda create --name raft
 conda activate raft
-conda install pytorch=1.6.0 torchvision=0.7.0 cudatoolkit=10.1 matplotlib tensorboard scipy opencv -c pytorch
 ```
+Then follow the commands inside `setup-env.sh`
 
 ## Demos
 Pretrained models can be downloaded by running

core/raft.py

@@ -27,14 +27,14 @@ def __init__(self, args):
         self.args = args
 
         if args.small:
-            self.hidden_dim = hdim = 96
-            self.context_dim = cdim = 64
+            self.hidden_dim = hdim = (args.hidden // 4) * 3
+            self.context_dim = cdim = args.context // 2
             args.corr_levels = 4
             args.corr_radius = 3
 
         else:
-            self.hidden_dim = hdim = 128
-            self.context_dim = cdim = 128
+            self.hidden_dim = hdim = args.hidden
+            self.context_dim = cdim = args.context
             args.corr_levels = 4
             args.corr_radius = 4
 

setup-env.sh

@@ -1,8 +1,10 @@
+#!/usr/bin/env bash
+
+echo Please run this script manually
 exit 1
-conda create -n raft-dl2023
-conda activate raft-dl2023
+
 conda install -y python=3.8
 conda install -y cudatoolkit=11.1 -c conda-forge
 conda install -y pytorch==1.8.0 torchvision==0.9.0 -c pytorch
-conda install -y tensorboard=2.10.0 matplotlib scipy tqdm
-pip install opencv-python
+conda install -y matplotlib scipy tensorboard=2.10.0 tqdm
+pip install opencv-python

train-selfsupervised.py

@@ -6,8 +6,9 @@
 
 import argparse
 import os
+from collections import OrderedDict
 from pathlib import Path
-from typing import List
+from typing import Any, List
 
 import numpy as np
 import torch
@@ -101,8 +102,14 @@ def train(args):
     print("Parameter Count: %d" % count_parameters(model))
 
     if args.restore_ckpt is not None:
-        model.load_state_dict(torch.load(args.restore_ckpt),
-                              strict=(not args.allow_nonstrict))
+        checkpoint: OrderedDict[str, Any] = torch.load(args.restore_ckpt)
+        if args.reset_context:
+            weight = OrderedDict()
+            for key, val in checkpoint.items():
+                if '.cnet.' not in key:
+                    weight[key] = val
+            checkpoint = weight
+        model.load_state_dict(checkpoint, strict=(not args.allow_nonstrict))
 
     model.cuda()
     model.train()
@@ -206,7 +213,16 @@ def train(args):
     parser.add_argument('--gamma', type=float, default=0.8,
                         help='exponential weighting')
     parser.add_argument('--add_noise', action='store_true')
+
+    parser.add_argument('--hidden', type=int, default=128,
+                        help='The hidden size of the updater')
+    parser.add_argument('--context', type=int, default=128,
+                        help='The context size of the updater')
+    parser.add_argument('--reset_context', action='store_true')
+
     args = parser.parse_args()
+    if args.hidden != 128 or args.context != 128:
+        args.reset_context = True
 
     torch.manual_seed(1234)
     np.random.seed(1234)

train-supervised.py

@@ -6,7 +6,9 @@
 
 import argparse
 import os
+from collections import OrderedDict
 from pathlib import Path
+from typing import Any
 
 import numpy as np
 import torch
@@ -94,8 +96,14 @@ def train(args):
     print("Parameter Count: %d" % count_parameters(model))
 
     if args.restore_ckpt is not None:
-        model.load_state_dict(torch.load(args.restore_ckpt),
-                              strict=(not args.allow_nonstrict))
+        checkpoint: OrderedDict[str, Any] = torch.load(args.restore_ckpt)
+        if args.reset_context:
+            weight = OrderedDict()
+            for key, val in checkpoint.items():
+                if '.cnet.' not in key:
+                    weight[key] = val
+            checkpoint = weight
+        model.load_state_dict(checkpoint, strict=(not args.allow_nonstrict))
 
     model.cuda()
     model.train()
@@ -199,7 +207,16 @@ def train(args):
     parser.add_argument('--gamma', type=float, default=0.8,
                         help='exponential weighting')
     parser.add_argument('--add_noise', action='store_true')
+
+    parser.add_argument('--hidden', type=int, default=128,
+                        help='The hidden size of the updater')
+    parser.add_argument('--context', type=int, default=128,
+                        help='The context size of the updater')
+    parser.add_argument('--reset_context', action='store_true')
+
     args = parser.parse_args()
+    if args.hidden != 128 or args.context != 128:
+        args.reset_context = True
 
     torch.manual_seed(1234)
     np.random.seed(1234)