Adds FFCV support to trainer + restores old def file

Singularity.v0.1.def

@@ -0,0 +1,33 @@
+Bootstrap: oras
+From: ghcr.io/sinzlab/pytorch-singularity:v3.9-torch1.10.2-dj0.12.7.def
+
+%labels
+MAINTAINER Arne Nix <[email protected]>
+
+%post
+    # install third-party libraries
+    # needed for vim extension in jupyter and tex export in matplotlib:
+    apt update && apt install -y \
+                          texlive-latex-extra \
+                          texlive-fonts-recommended \
+                          texlive-base \
+                          dvipng \
+                          zsh \
+                          python3-venv
+
+    python3.9 -m pip --no-cache-dir install \
+                               checkout_code \
+                               requests \
+                               imageio \
+                               scikit-image \
+                               einops \
+                               vit-pytorch
+
+%environment
+    export SHELL=/usr/bin/zsh
+
+%startscript
+    exec "$@"
+
+%runscript
+    exec "$@"

Singularity.v0.2.def

@@ -1,80 +1,31 @@
-Bootstrap: docker
-
-From: continuumio/miniconda3
-
-%files
-    environment.yml
+Bootstrap: oras
+From: ghcr.io/sinzlab/pytorch-singularity:v3.9-torch1.10.2-dj0.12.7.def
 
+%labels
+    MAINTAINER Arne Nix <[email protected]>
 %post
-    apt update && apt install -y \
-        build-essential \
-        git \
-        wget \
-        vim \
-        curl \
-        zip \
-        zlib1g-dev \
-        unzip \
-        pkg-config \
-        libblas-dev \
-        liblapack-dev \
-        python3-tk \
-        python3-wheel \
-        graphviz \
-        libhdf5-dev \
-        python3.9 \
-        python3.9-dev \
-        python3.9-distutils \
-        python3-testresources \
-        software-properties-common \
-        swig \
-        ffmpeg \
-        texlive-latex-extra \
-        texlive-fonts-recommended \
-        texlive-base \
-        dvipng
-         #zsh
-    apt-get clean
+    # install third-party libraries
+    # needed for vim extension in jupyter and tex export in matplotlib:
+    apt update && apt install -y libturbojpeg \
+    libturbojpeg-dev \
+        libopencv-dev \
+                          python3-venv \
+                          zsh \
+         python3-opencv
+                          # texlive-latex-extra \
+                          # texlive-fonts-recommended \
+                          # texlive-base \
+                          # dvipng \
 
-    ln -s /usr/bin/python3.9 /usr/local/bin/python
-    ln -s /usr/bin/python3.9 /usr/local/bin/python3
-
-    curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py
-    python3.9 get-pip.py
-    rm get-pip.py
-    python3.9 -m pip --no-cache-dir install --upgrade pip
 
     python3.9 -m pip --no-cache-dir install \
-        blackcellmagic\
-        pytest \
-        pytest-cov \
-        numpy \
-        matplotlib \
-        scipy \
-        pandas \
-        jupyter \
-        scikit-learn \
-        scikit-image \
-        seaborn \
-        graphviz \
-        gpustat \
-        h5py \
-        gitpython \
-        Pillow==8.0.1 \
-        jupyterlab \
-        datajoint==0.12.7\
         ipykernel \
         requests \
         imageio \
         scikit-image \
         einops \
         vit-pytorch
 
-    python3.9 -m pip --no-cache-dir  install \
-        torch==1.10.2+cu113 \
-        torchvision==0.11.3+cu113 \
-        torchaudio==0.10.2+cu113 \
-        -f https://download.pytorch.org/whl/cu113/torch_stable.html
 
 %runscript
     exec "$@"

nntransfer/trainer/trainer.py

@@ -1,4 +1,5 @@
 import sys
+from contextlib import nullcontext
 from functools import partial
 
 from tqdm import tqdm
@@ -59,7 +60,7 @@ def __init__(self, dataloaders, model, seed, uid, cb, **kwargs):
         self.task_keys = dataloaders["train"].keys()
         self.optimizer, self.stop_closure, self.criterion = self.get_training_controls()
         self.lr_scheduler = self.prepare_lr_schedule()
-        if self.use_ffcv:
+        if self.config.use_ffcv:
             self.scaler = GradScaler()
 
         # Potentially reset parts of the model (after loading pretrained parameters)
@@ -215,7 +216,7 @@ def main_loop(
                 shared_memory = {}  # e.g. to remember where which noise was applied
                 model_ = self.model
 
-                forward_context = autocast() if self.config.use_ffcv else nullcontext
+                forward_context = autocast() if self.config.use_ffcv else nullcontext()
                 with forward_context:
                     for module in self.main_loop_modules:
                         model_, inputs = module.pre_forward(