modelling_ms_convstar_dlbs.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DLBS - MSConvstar Experiments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from src.dlbs import train_dlbs\n",
    "from src import test\n",
    "import numpy as np\n",
    "import torch\n",
    "import wandb"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Overfit Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output = train_dlbs.main(\n",
    "    datadir = \"../scratch/AgroLuege/ZueriCrop/ZueriCrop.hdf5\",\n",
    "    workers= 0,\n",
    "    batchsize=4,\n",
    "    epochs=1000,\n",
    "    lr=1e-4,\n",
    "    snapshot=None,\n",
    "    checkpoint_dir=\"./checkpoint/\",\n",
    "    layer=3,\n",
    "    lambda_1=0.0,\n",
    "    lambda_2=0.0,\n",
    "    weight_decay=0.0,\n",
    "    hidden=64,\n",
    "    lrS=2,\n",
    "    name=\"msConvSTAR\",\n",
    "    dropout=0.0,\n",
    "    stage=3,\n",
    "    clip=5,\n",
    "    seed=0,\n",
    "    gt_path=\"./raw_data/ZueriCrop/labels.csv\",\n",
    "    cell=\"star\",\n",
    "    input_dim=4,\n",
    "    fold_num='all',\n",
    "    project= \"DLBS-MS-Convstar\",\n",
    "    run_group=\"Overfit-Test\",\n",
    "    one_batch_training = True,\n",
    "    augment_rate=0.,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## First Run"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output = train_dlbs.main(\n",
    "    datadir = \"../scratch/AgroLuege/ZueriCrop/ZueriCrop.hdf5\",\n",
    "    workers= 0,\n",
    "    batchsize=4,\n",
    "    epochs=30,\n",
    "    lr=1e-4,\n",
    "    snapshot=None,\n",
    "    checkpoint_dir=\"./checkpoint/\",\n",
    "    layer=6,\n",
    "    lambda_1=0.1,\n",
    "    lambda_2=0.5,\n",
    "    weight_decay=0,\n",
    "    hidden=64,\n",
    "    lrS=2,\n",
    "    name=\"msConvSTAR\",\n",
    "    dropout=0,\n",
    "    stage=3,\n",
    "    clip=5,\n",
    "    seed=0,\n",
    "    gt_path=\"./raw_data/ZueriCrop/labels.csv\",\n",
    "    cell=\"star\",\n",
    "    input_dim=4,\n",
    "    fold_num='all',\n",
    "    project= \"DLBS-MS-Convstar\",\n",
    "    run_group=\"Run-Unregularized\",\n",
    "    augment_rate=0.,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Regularisations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Batchsize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output = train_dlbs.main(\n",
    "    datadir = \"../scratch/AgroLuege/ZueriCrop/ZueriCrop.hdf5\",\n",
    "    workers= 0,\n",
    "    batchsize=16,\n",
    "    epochs=30,\n",
    "    lr=1e-4,\n",
    "    snapshot=None,\n",
    "    checkpoint_dir=\"./checkpoint/\",\n",
    "    layer=6,\n",
    "    lambda_1=0.1,\n",
    "    lambda_2=0.5,\n",
    "    weight_decay=0,\n",
    "    hidden=64,\n",
    "    lrS=2,\n",
    "    name=\"msConvSTAR\",\n",
    "    dropout=0,\n",
    "    stage=3,\n",
    "    clip=5,\n",
    "    seed=0,\n",
    "    gt_path=\"./raw_data/ZueriCrop/labels.csv\",\n",
    "    cell=\"star\",\n",
    "    input_dim=4,\n",
    "    fold_num='all',\n",
    "    project= \"DLBS-MS-Convstar\",\n",
    "    run_group=\"Run-Batchsizes-16\",\n",
    "    augment_rate=0.,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output = train_dlbs.main(\n",
    "    datadir = \"../scratch/AgroLuege/ZueriCrop/ZueriCrop.hdf5\",\n",
    "    workers= 0,\n",
    "    batchsize=4,\n",
    "    epochs=30,\n",
    "    lr=1e-4,\n",
    "    snapshot=None,\n",
    "    checkpoint_dir=\"./checkpoint/\",\n",
    "    layer=6,\n",
    "    lambda_1=0.1,\n",
    "    lambda_2=0.5,\n",
    "    weight_decay=0,\n",
    "    hidden=64,\n",
    "    lrS=2,\n",
    "    name=\"msConvSTAR\",\n",
    "    dropout=0.3,\n",
    "    stage=3,\n",
    "    clip=5,\n",
    "    seed=0,\n",
    "    gt_path=\"./raw_data/ZueriCrop/labels.csv\",\n",
    "    cell=\"star\",\n",
    "    input_dim=4,\n",
    "    fold_num='all',\n",
    "    project= \"DLBS-MS-Convstar\",\n",
    "    run_group=\"Run-Dropout-0.3\",\n",
    "    augment_rate=0.,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output = train_dlbs.main(\n",
    "    datadir = \"../scratch/AgroLuege/ZueriCrop/ZueriCrop.hdf5\",\n",
    "    workers= 0,\n",
    "        #TODO: change according to first run\n",
    "    batchsize=4,\n",
    "    epochs=30,\n",
    "    lr=1e-4,\n",
    "    snapshot=None,\n",
    "    checkpoint_dir=\"./checkpoint/\",\n",
    "    layer=6,\n",
    "    lambda_1=0.1,\n",
    "    lambda_2=0.5,\n",
    "    weight_decay=0,\n",
    "    hidden=64,\n",
    "    lrS=2,\n",
    "    name=\"msConvSTAR\",\n",
    "    dropout=0.5,\n",
    "    stage=3,\n",
    "    clip=5,\n",
    "    seed=0,\n",
    "    gt_path=\"./raw_data/ZueriCrop/labels.csv\",\n",
    "    cell=\"star\",\n",
    "    input_dim=4,\n",
    "    fold_num='all',\n",
    "    project= \"DLBS-MS-Convstar\",\n",
    "    run_group=\"Run-Dropout-0.5\",\n",
    "    augment_rate=0.,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output = train_dlbs.main(\n",
    "    datadir = \"../scratch/AgroLuege/ZueriCrop/ZueriCrop.hdf5\",\n",
    "    workers= 0,\n",
    "        #TODO: change according to first run\n",
    "    batchsize=4,\n",
    "    epochs=30,\n",
    "    lr=1e-4,\n",
    "    snapshot=None,\n",
    "    checkpoint_dir=\"./checkpoint/\",\n",
    "    layer=6,\n",
    "    lambda_1=0.1,\n",
    "    lambda_2=0.5,\n",
    "    weight_decay=0,\n",
    "    hidden=64,\n",
    "    lrS=2,\n",
    "    name=\"msConvSTAR\",\n",
    "    dropout=0.7,\n",
    "    stage=3,\n",
    "    clip=5,\n",
    "    seed=0,\n",
    "    gt_path=\"./raw_data/ZueriCrop/labels.csv\",\n",
    "    cell=\"star\",\n",
    "    input_dim=4,\n",
    "    fold_num='all',\n",
    "    project= \"DLBS-MS-Convstar\",\n",
    "    run_group=\"Run-Dropout-0.7\",\n",
    "    augment_rate=0.,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Weight Decay"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output = train_dlbs.main(\n",
    "    datadir = \"../scratch/AgroLuege/ZueriCrop/ZueriCrop.hdf5\",\n",
    "    workers= 0,\n",
    "        #TODO: change according to first run\n",
    "    batchsize=4,\n",
    "    epochs=30,\n",
    "    lr=1e-4,\n",
    "    snapshot=None,\n",
    "    checkpoint_dir=\"./checkpoint/\",\n",
    "    layer=6,\n",
    "    lambda_1=0.1,\n",
    "    lambda_2=0.5,\n",
    "    weight_decay=0.001,\n",
    "    hidden=64,\n",
    "    lrS=2,\n",
    "    name=\"msConvSTAR\",\n",
    "    dropout=0.,\n",
    "    stage=3,\n",
    "    clip=5,\n",
    "    seed=0,\n",
    "    gt_path=\"./raw_data/ZueriCrop/labels.csv\",\n",
    "    cell=\"star\",\n",
    "    input_dim=4,\n",
    "    fold_num='all',\n",
    "    project= \"DLBS-MS-Convstar\",\n",
    "    run_group=\"Run-Weigh-Decay-0.001\",\n",
    "    augment_rate=0.,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output = train_dlbs.main(\n",
    "    datadir = \"../scratch/AgroLuege/ZueriCrop/ZueriCrop.hdf5\",\n",
    "    workers= 0,\n",
    "        #TODO: change according to first run\n",
    "    batchsize=4,\n",
    "    epochs=30,\n",
    "    lr=1e-4,\n",
    "    snapshot=None,\n",
    "    checkpoint_dir=\"./checkpoint/\",\n",
    "    layer=6,\n",
    "    lambda_1=0.1,\n",
    "    lambda_2=0.5,\n",
    "    weight_decay=0.0001,\n",
    "    hidden=64,\n",
    "    lrS=2,\n",
    "    name=\"msConvSTAR\",\n",
    "    dropout=0.,\n",
    "    stage=3,\n",
    "    clip=5,\n",
    "    seed=0,\n",
    "    gt_path=\"./raw_data/ZueriCrop/labels.csv\",\n",
    "    cell=\"star\",\n",
    "    input_dim=4,\n",
    "    fold_num='all',\n",
    "    project= \"DLBS-MS-Convstar\",\n",
    "    run_group=\"Run-Weigh-Decay-0.0001\",\n",
    "    augment_rate=0.,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Augmentation Rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output = train_dlbs.main(\n",
    "    datadir = \"../scratch/AgroLuege/ZueriCrop/ZueriCrop.hdf5\",\n",
    "    workers= 0,\n",
    "        #TODO: change according to first run\n",
    "    batchsize=4,\n",
    "    epochs=30,\n",
    "    lr=1e-4,\n",
    "    snapshot=None,\n",
    "    checkpoint_dir=\"./checkpoint/\",\n",
    "    layer=6,\n",
    "    lambda_1=0.1,\n",
    "    lambda_2=0.5,\n",
    "    weight_decay=0.,\n",
    "    hidden=64,\n",
    "    lrS=2,\n",
    "    name=\"msConvSTAR\",\n",
    "    dropout=0.,\n",
    "    stage=3,\n",
    "    clip=5,\n",
    "    seed=0,\n",
    "    gt_path=\"./raw_data/ZueriCrop/labels.csv\",\n",
    "    cell=\"star\",\n",
    "    input_dim=4,\n",
    "    fold_num='all',\n",
    "    project= \"DLBS-MS-Convstar\",\n",
    "    run_group=\"Run-augment_rate-0.33\",\n",
    "    augment_rate=0.33,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output = train_dlbs.main(\n",
    "    datadir = \"../scratch/AgroLuege/ZueriCrop/ZueriCrop.hdf5\",\n",
    "    workers= 0,\n",
    "        #TODO: change according to first run\n",
    "    batchsize=4,\n",
    "    epochs=30,\n",
    "    lr=1e-4,\n",
    "    snapshot=None,\n",
    "    checkpoint_dir=\"./checkpoint/\",\n",
    "    layer=6,\n",
    "    lambda_1=0.1,\n",
    "    lambda_2=0.5,\n",
    "    weight_decay=0.,\n",
    "    hidden=64,\n",
    "    lrS=2,\n",
    "    name=\"msConvSTAR\",\n",
    "    dropout=0.,\n",
    "    stage=3,\n",
    "    clip=5,\n",
    "    seed=0,\n",
    "    gt_path=\"./raw_data/ZueriCrop/labels.csv\",\n",
    "    cell=\"star\",\n",
    "    input_dim=4,\n",
    "    fold_num='all',\n",
    "    project= \"DLBS-MS-Convstar\",\n",
    "    run_group=\"Run-augment_rate-0.66\",\n",
    "    augment_rate=0.66,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tune"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "lrs = [3e-4, 1e-5,1e-6]\n",
    "print(lrs)\n",
    "for lr in lrs:\n",
    "    output = train_dlbs.main(\n",
    "        datadir = \"../scratch/AgroLuege/ZueriCrop/ZueriCrop.hdf5\",\n",
    "        workers= 0,\n",
    "            #TODO: change according to first run\n",
    "        batchsize=4,\n",
    "        epochs=30,\n",
    "        lr=lr,\n",
    "        snapshot=None,\n",
    "        checkpoint_dir=\"./checkpoint/\",\n",
    "        layer=6,\n",
    "        lambda_1=0.1,\n",
    "        lambda_2=0.5,\n",
    "        weight_decay=0.,\n",
    "        hidden=64,\n",
    "        lrS=2,\n",
    "        name=\"msConvSTAR\",\n",
    "        dropout=0.,\n",
    "        stage=3,\n",
    "        clip=5,\n",
    "        seed=0,\n",
    "        gt_path=\"./raw_data/ZueriCrop/labels.csv\",\n",
    "        cell=\"star\",\n",
    "        input_dim=4,\n",
    "        fold_num='all',\n",
    "        project= \"DLBS-MS-Convstar\",\n",
    "        run_group=\"Run-augment_rate-0.66\",\n",
    "        augment_rate=0.66,\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}