Merge pull request #30 from i4Ds/feature/additional-config

adds more configs, adds wandb utils

config/ce-vae/defaults.yaml

@@ -34,7 +34,7 @@ data:
     value: 8
     desc: "Number of samples loaded in advance by each worker. 2 means there will be a total of 2 * num_workers samples prefetched across all workers."
   pin_memory:
-    value: true
+    value: false
     desc: "If true, the data loader will copy Tensors into CUDA pinned memory before returning them"
   sdo_ml_v2:
     storage_driver:

config/ce-vae/run-fhnw-1.yaml

@@ -3,7 +3,7 @@ model:
     value: null
 data:
   data_dir:
-    value: /mnt/nas05/astrodata01/astroml_data/sdomlv2.zarr
+    value: /mnt/nas05/astrodata01/astroml_data/sdomlv2_full/sdomlv2.zarr
   dataset:
     value: SDOMLDatasetV2
   num_data_loader_workers:
@@ -39,3 +39,6 @@ predict:
     value: "./output/predictions"
 log_dir:
   value: "./output/train-sdo-ml"
+train:
+  early_stopping:
+    value: false

config/ce-vae/run-fhnw-2-use-geco.yaml

@@ -0,0 +1,46 @@
+model:
+  load_path:
+    value: null
+data:
+  data_dir:
+    value: /mnt/nas05/astrodata01/astroml_data/sdomlv2_full/sdomlv2.zarr
+  dataset:
+    value: SDOMLDatasetV2
+  num_data_loader_workers:
+    value: 16
+  prefetch_factor:
+    value: 8
+  batch_size:
+    value: 32
+  sdo_ml_v2:
+    train_year:
+      value:
+        - "2010"
+        - "2011"
+    train_start_date:
+      value: null
+    train_end_date:
+      value: null
+    test_year:
+      value:
+        - "2012"
+    test_start_date:
+      value: null
+    test_end_date:
+      value: null
+    train_val_split_ratio:
+      value: 0.8
+    train_val_split_temporal_chunk_size:
+      value: 3d
+predict:
+  mode:
+    value: sample
+  pred_dir:
+    value: "./output/predictions"
+log_dir:
+  value: "./output/train-sdo-ml"
+train:
+  early_stopping:
+    value: false
+  use_geco:
+    value: true

config/ce-vae/run-fhnw-full-1.yaml

@@ -0,0 +1,49 @@
+model:
+  load_path:
+    value: null
+data:
+  data_dir:
+    value: /mnt/nas05/astrodata01/astroml_data/sdomlv2_full/sdomlv2.zarr
+  dataset:
+    value: SDOMLDatasetV2
+  num_data_loader_workers:
+    value: 16
+  prefetch_factor:
+    value: 8
+  batch_size:
+    value: 32
+  sdo_ml_v2:
+    train_year:
+      value:
+        - "2010"
+        - "2012"
+        - "2013"
+        - "2015"
+        - "2017"
+        - "2018"
+        - "2020"
+    train_start_date:
+      value: null
+    train_end_date:
+      value: null
+    test_year:
+      value:
+        - "2011"
+        - "2014"
+        - "2016"
+        - "2019"
+    test_start_date:
+      value: null
+    test_end_date:
+      value: null
+    train_val_split_ratio:
+      value: 0.8
+    train_val_split_temporal_chunk_size:
+      value: 14d
+predict:
+  mode:
+    value: sample
+  pred_dir:
+    value: "./output/predictions"
+log_dir:
+  value: "./output/train-sdo-ml"

notebooks/wandb_utils.ipynb

@@ -0,0 +1,200 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "01e3c258",
+   "metadata": {},
+   "source": [
+    "## W&B Utils\n",
+    "\n",
+    "- https://docs.wandb.ai/guides/track/public-api-guide\n",
+    "- https://docs.wandb.ai/ref/python/public-api/run"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "68d45422",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import wandb\n",
+    "\n",
+    "api = wandb.Api()\n",
+    "runs = api.runs(\"mariusgiger/sdo-sood\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "c47e1ca5",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "run 2nuu15v4 with state finished\n",
+      "run 38xbbrf0 with state failed\n",
+      "run 3ctkpft0 with state failed\n",
+      "run l6lejey1 with state failed\n",
+      "run te606vkj with state finished\n",
+      "run 10rtm2tv with state failed\n",
+      "run 2qvg5dhc with state failed\n",
+      "run 11lp21yf with state crashed\n",
+      "run lv2aex4j with state crashed\n",
+      "run 2zqrqpp5 with state failed\n",
+      "run 1fzn9092 with state failed\n",
+      "run 71p99trq with state failed\n",
+      "run 19dqvasr with state failed\n",
+      "run 1j3cgzq3 with state failed\n",
+      "run l8kgeqbn with state finished\n",
+      "run 21m3ucxd with state crashed\n",
+      "run 1k4qhlud with state crashed\n",
+      "run 34eoz7ka with state crashed\n",
+      "run 106bp0qy with state crashed\n",
+      "run 1tgfv6rt with state failed\n",
+      "run 329oose2 with state crashed\n",
+      "run 1ws7f4rc with state finished\n",
+      "run 913zh6ar with state crashed\n",
+      "run 1ygauldv with state crashed\n",
+      "run 28n7ocav with state finished\n",
+      "run b6qqdqo5 with state finished\n",
+      "run 2n3ofmuw with state finished\n",
+      "run 1np3by0g with state finished\n",
+      "run 5zupatxn with state finished\n",
+      "run 3ehknwtk with state finished\n",
+      "run 3k0flaze with state finished\n",
+      "run 1lkm6ftd with state failed\n",
+      "run 36tfoo4q with state finished\n",
+      "run zh2feqlv with state finished\n",
+      "run 2pcd3hdj with state failed\n",
+      "run 3oyydq7x with state finished\n",
+      "run jzo18qkw with state failed\n",
+      "run 357id1hx with state failed\n",
+      "run 1xnegvzy with state failed\n",
+      "run 3m9oz6rl with state finished\n",
+      "run 39uy76fo with state failed\n",
+      "run 33xn5hcx with state failed\n",
+      "run gqr86jwj with state failed\n",
+      "run 3ce9byck with state failed\n",
+      "run 3ov4ixu2 with state failed\n",
+      "run 1uqu0ukw with state failed\n",
+      "run 1k4cpb9p with state failed\n",
+      "run 3ewoto0c with state finished\n",
+      "run xaaxibor with state finished\n",
+      "run 3l3x9gtu with state finished\n",
+      "run 2mh7qvup with state crashed\n",
+      "run 1pydotta with state crashed\n",
+      "run 3ljg7ntt with state crashed\n",
+      "run 2bgltqvu with state finished\n",
+      "run 2vy9wmmj with state crashed\n",
+      "run p2whcsa2 with state crashed\n",
+      "run 2yibaqs3 with state crashed\n",
+      "run 1lf0wfcr with state crashed\n",
+      "run g5q6wzg1 with state failed\n",
+      "run 3ochsc1e with state crashed\n",
+      "run 3lwpsznj with state crashed\n",
+      "run 1bxx8kc4 with state crashed\n",
+      "run 1vuwhzl7 with state crashed\n",
+      "run 3pe2sjgv with state crashed\n",
+      "run 10qwufqr with state crashed\n",
+      "run 1mq33277 with state failed\n",
+      "run 175qf053 with state crashed\n",
+      "run 2pkajuoi with state finished\n",
+      "run 1trtx7b9 with state finished\n",
+      "run 61fvu79a with state finished\n",
+      "run 2tf5fl2i with state finished\n",
+      "run 2nop4en9 with state crashed\n",
+      "run gnwy0bgq with state failed\n",
+      "run 3l6ulb81 with state failed\n",
+      "run 23l9wkl4 with state failed\n",
+      "run 6nhiyno7 with state failed\n",
+      "run blta96jh with state failed\n",
+      "run 3fvwxq7r with state failed\n",
+      "run gjut85w8 with state crashed\n",
+      "run 2y1uxe3g with state failed\n",
+      "run 1t453lx9 with state crashed\n",
+      "run 3pz9nrut with state failed\n",
+      "run 10rpx2e9 with state failed\n",
+      "run 8sfx2dk8 with state failed\n",
+      "run 2ig8kb8m with state failed\n",
+      "run 19v4g4la with state failed\n",
+      "run 20csxu2n with state failed\n",
+      "run 3hsnu271 with state failed\n",
+      "run 3hp3q8bv with state failed\n",
+      "run 174b65vr with state failed\n",
+      "run 2qrtndq3 with state failed\n",
+      "run srdsl9bo with state crashed\n",
+      "run 3idsuzra with state finished\n",
+      "run 3szlcuro with state crashed\n",
+      "run 3g8l62fm with state crashed\n",
+      "run 2gursnum with state crashed\n",
+      "run 22bdbjt7 with state crashed\n",
+      "run 389bo70b with state crashed\n",
+      "run tlh0o8w0 with state failed\n",
+      "run 3b3p9e71 with state failed\n",
+      "run 29cpbtho with state failed\n",
+      "run e7qqqj88 with state failed\n",
+      "run 2s643qei with state failed\n",
+      "run 1c64xxjb with state failed\n",
+      "run 1spxk4ls with state failed\n",
+      "run 19eosjzt with state failed\n",
+      "run ovcv7y75 with state failed\n",
+      "run 176gt7i6 with state failed\n",
+      "run u0yxay0f with state failed\n",
+      "run 2565yf2t with state crashed\n",
+      "run 8vhefnkf with state finished\n",
+      "run 2ahoewvq with state finished\n"
+     ]
+    }
+   ],
+   "source": [
+    "for run in runs:\n",
+    "    print(f\"run {run.id} with state {run.state}\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "9289ca86",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# delete invalid runs\n",
+    "for run in runs:\n",
+    "    if run.state == \"failed\" or run.state == \"crashed\":\n",
+    "        run.delete(delete_artifacts=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "658ee585",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "sdo-cli",
+   "language": "python",
+   "name": "sdo-cli"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.13"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}