{ "cells": [ { "cell_type": "markdown", "id": "be7a6c7e", "metadata": {}, "source": [ "# Explore a completed training run\n", "\n", "Use this notebook to sanity-check the artifacts written by `airfrans-frontier train`: final metrics, training curves, split manifest, normalization stats, checkpoint contents, and reloaded-model predictions. It reads existing artifacts only.\n" ] }, { "cell_type": "markdown", "id": "23b531cc", "metadata": {}, "source": [ "## Setup\n", "\n", "Run from the repository root or from `notebooks/`. Select the repository `.venv` kernel. If imports fail, run `uv sync --dev` from the repository root and restart the kernel.\n" ] }, { "cell_type": "code", "execution_count": 1, "id": "18ae45d0", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "python: /home/aaron/data/airfrans/.venv/bin/python\n", "repo: /home/aaron/data/airfrans\n" ] } ], "source": [ "from __future__ import annotations\n", "\n", "import json\n", "import os\n", "import sys\n", "from pathlib import Path\n", "\n", "for path in list(sys.path):\n", " if \"swactor-mvp/pydeps\" in path:\n", " sys.path.remove(path)\n", "if \"PYTHONPATH\" in os.environ:\n", " os.environ[\"PYTHONPATH\"] = os.pathsep.join(\n", " entry for entry in os.environ[\"PYTHONPATH\"].split(os.pathsep) if \"swactor-mvp/pydeps\" not in entry\n", " )\n", "\n", "print(f\"python: {sys.executable}\")\n", "\n", "try:\n", " import matplotlib.pyplot as plt\n", " import numpy as np\n", " import torch\n", "except (ImportError, ModuleNotFoundError) as exc:\n", " raise ImportError(\n", " f\"Required notebook dependencies are not importable in this kernel: {sys.executable}. \"\n", " \"From the repository root, run `uv sync --dev`, then restart the notebook kernel.\"\n", " ) from exc\n", "\n", "\n", "def find_repo_root(start: Path) -> Path:\n", " for candidate in (start, *start.parents):\n", " if (candidate / \"pyproject.toml\").exists() and (candidate / \"src\" / \"airfrans_frontier\").exists():\n", " return candidate\n", " raise RuntimeError(f\"Could not find repository root from {start}\")\n", "\n", "\n", "REPO_ROOT = find_repo_root(Path.cwd())\n", "SRC_DIR = REPO_ROOT / \"src\"\n", "if str(SRC_DIR) not in sys.path:\n", " sys.path.insert(0, str(SRC_DIR))\n", "\n", "from airfrans_frontier.models.mlp import PointwiseMLP\n", "\n", "print(f\"repo: {REPO_ROOT}\")\n" ] }, { "cell_type": "markdown", "id": "1388d69e", "metadata": {}, "source": [ "## Select a run\n", "\n", "By default this picks the newest directory under `artifacts/runs`. Override `RUN_DIR` manually if you want an older run.\n" ] }, { "cell_type": "code", "execution_count": 2, "id": "fd3c05a7", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "selected run: artifacts/runs/20260721T082353Z_mlp_tiny\n", "available runs:\n", " artifacts/runs/20260721T082151Z_mlp_tiny\n", "* artifacts/runs/20260721T082353Z_mlp_tiny\n" ] } ], "source": [ "RUNS_DIR = REPO_ROOT / \"artifacts\" / \"runs\"\n", "run_dirs = sorted([path for path in RUNS_DIR.glob(\"*\") if path.is_dir()], key=lambda path: path.stat().st_mtime)\n", "if not run_dirs:\n", " raise FileNotFoundError(f\"No training runs found under {RUNS_DIR}\")\n", "\n", "RUN_DIR = run_dirs[-1]\n", "print(f\"selected run: {RUN_DIR.relative_to(REPO_ROOT)}\")\n", "print(\"available runs:\")\n", "for path in run_dirs:\n", " marker = \"*\" if path == RUN_DIR else \" \"\n", " print(f\"{marker} {path.relative_to(REPO_ROOT)}\")\n" ] }, { "cell_type": "code", "execution_count": 3, "id": "e7e15a27", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "config.toml 438 bytes\n", "split_manifest.json 153 bytes\n", "normalization.json 733 bytes\n", "metrics.jsonl 725 bytes\n", "final_metrics.json 1,074 bytes\n", "checkpoint.pt 623,285 bytes\n" ] } ], "source": [ "expected_files = [\n", " \"config.toml\",\n", " \"split_manifest.json\",\n", " \"normalization.json\",\n", " \"metrics.jsonl\",\n", " \"final_metrics.json\",\n", " \"checkpoint.pt\",\n", "]\n", "missing = [name for name in expected_files if not (RUN_DIR / name).exists()]\n", "if missing:\n", " raise FileNotFoundError(f\"Run is missing expected artifacts: {missing}\")\n", "\n", "for name in expected_files:\n", " path = RUN_DIR / name\n", " print(f\"{name:20} {path.stat().st_size:>10,} bytes\")\n" ] }, { "cell_type": "markdown", "id": "da64dacb", "metadata": {}, "source": [ "## Final metrics\n", "\n", "This is the end-of-run summary written by the training loop. The key checks are: train loss dropped from the initial loss, validation/test losses are finite, and device/GPU fields match the intended hardware.\n" ] }, { "cell_type": "code", "execution_count": 4, "id": "cb352094", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "{'device': 'cuda:0',\n", " 'elapsed_seconds': 1.7905638560005173,\n", " 'gpu_memory_peak_allocated_mb': 17,\n", " 'gpu_memory_total_mb': 3717,\n", " 'gpu_name': 'NVIDIA T550 Laptop GPU',\n", " 'initial_train_loss': 1.0026862990200476,\n", " 'parameter_count': 50820,\n", " 'points_per_case': 128,\n", " 'steps': 500,\n", " 'test_cases': 1,\n", " 'test_loss': 0.00022942002189996227,\n", " 'test_mse_per_channel': {'pressure': 0.000408856492614153,\n", " 'turbulent_viscosity': 0.00035512470154015644,\n", " 'velocity_x': 5.778009158681787e-05,\n", " 'velocity_y': 9.591880185872174e-05},\n", " 'train_cases': 4,\n", " 'train_loss': 0.0001373240501380092,\n", " 'train_mse_per_channel': {'pressure': 0.00023612300953799542,\n", " 'turbulent_viscosity': 0.00018737132594797248,\n", " 'velocity_x': 5.6441453127828415e-05,\n", " 'velocity_y': 6.936041193824046e-05},\n", " 'val_cases': 1,\n", " 'val_loss': 0.0007143189445776568,\n", " 'val_mse_per_channel': {'pressure': 0.0012492708046041544,\n", " 'turbulent_viscosity': 0.0010587173853462636,\n", " 'velocity_x': 0.00028163288487499037,\n", " 'velocity_y': 0.0002676547034852188}}" ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "final_metrics = json.loads((RUN_DIR / \"final_metrics.json\").read_text())\n", "final_metrics\n" ] }, { "cell_type": "code", "execution_count": 5, "id": "42010827", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "initial_train_loss: 1.00269\n", "train_loss: 0.000137324\n", "val_loss: 0.000714319\n", "test_loss: 0.00022942\n", "loss improvement: 7301.6x\n", "device: cuda:0\n", "gpu: NVIDIA T550 Laptop GPU\n", "parameters: 50,820\n", "basic metric checks: ok\n" ] } ], "source": [ "initial = final_metrics[\"initial_train_loss\"]\n", "train = final_metrics[\"train_loss\"]\n", "val = final_metrics.get(\"val_loss\")\n", "test = final_metrics.get(\"test_loss\")\n", "\n", "print(f\"initial_train_loss: {initial:.6g}\")\n", "print(f\"train_loss: {train:.6g}\")\n", "print(f\"val_loss: {val:.6g}\" if val is not None else \"val_loss: None\")\n", "print(f\"test_loss: {test:.6g}\" if test is not None else \"test_loss: None\")\n", "print(f\"loss improvement: {initial / train:.1f}x\")\n", "print(f\"device: {final_metrics.get('device')}\")\n", "print(f\"gpu: {final_metrics.get('gpu_name')}\")\n", "print(f\"parameters: {final_metrics.get('parameter_count'):,}\")\n", "\n", "assert np.isfinite(train), \"train loss is not finite\"\n", "assert train < initial, \"train loss did not improve\"\n", "if val is not None:\n", " assert np.isfinite(val), \"val loss is not finite\"\n", "if test is not None:\n", " assert np.isfinite(test), \"test loss is not finite\"\n", "print(\"basic metric checks: ok\")\n" ] }, { "cell_type": "markdown", "id": "e4cf0126", "metadata": {}, "source": [ "## Training curves\n", "\n", "`metrics.jsonl` has one JSON row per log interval. A sanity run should show train loss falling; validation loss should be finite and usually trend down on this tiny baseline.\n" ] }, { "cell_type": "code", "execution_count": 6, "id": "b4d36a9d", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "{'elapsed_seconds': 0.0, 'step': 0, 'train_loss': 1.0026862990200476, 'val_loss': 0.8999216645838499}\n", "{'elapsed_seconds': 0.6329330740009027, 'step': 100, 'train_loss': 0.005394324883408844, 'val_loss': 0.008841235591800927}\n", "{'elapsed_seconds': 0.9306512370003475, 'step': 200, 'train_loss': 0.0008050906900035648, 'val_loss': 0.002610322449175867}\n", "{'elapsed_seconds': 1.130966939001155, 'step': 300, 'train_loss': 0.00029600711922912065, 'val_loss': 0.0015054571447265593}\n", "{'elapsed_seconds': 1.4691060960012692, 'step': 400, 'train_loss': 0.00019286412112246845, 'val_loss': 0.0012647405838277085}\n", "{'elapsed_seconds': 1.7875807079999504, 'step': 500, 'train_loss': 0.0001373240501380092, 'val_loss': 0.0007143189445776568}\n" ] } ], "source": [ "metric_rows = [json.loads(line) for line in (RUN_DIR / \"metrics.jsonl\").read_text().splitlines() if line.strip()]\n", "steps = np.array([row[\"step\"] for row in metric_rows])\n", "train_loss = np.array([row[\"train_loss\"] for row in metric_rows], dtype=float)\n", "val_loss = np.array([row[\"val_loss\"] if row[\"val_loss\"] is not None else np.nan for row in metric_rows], dtype=float)\n", "\n", "for row in metric_rows:\n", " print(row)\n" ] }, { "cell_type": "code", "execution_count": 7, "id": "c678379b", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "fig, ax = plt.subplots(figsize=(8, 4.5))\n", "ax.plot(steps, train_loss, marker=\"o\", label=\"train\")\n", "if np.isfinite(val_loss).any():\n", " ax.plot(steps, val_loss, marker=\"o\", label=\"val\")\n", "ax.set_yscale(\"log\")\n", "ax.set_xlabel(\"optimization step\")\n", "ax.set_ylabel(\"normalized MSE\")\n", "ax.set_title(RUN_DIR.name)\n", "ax.grid(True, which=\"both\", alpha=0.25)\n", "ax.legend()\n", "fig.tight_layout()\n" ] }, { "cell_type": "markdown", "id": "804c8844", "metadata": {}, "source": [ "## Split and normalization\n", "\n", "The split manifest tells you exactly which `.npz` cases produced train/validation/test metrics. Normalization is computed from the train split and stored so the checkpoint can be used later.\n" ] }, { "cell_type": "code", "execution_count": 8, "id": "d993f11f", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[run]\n", "name = \"mlp_tiny\"\n", "seed = 0\n", "artifact_dir = \"artifacts/runs\"\n", "\n", "[data]\n", "root = \"data/processed/minimal\"\n", "train_cases = 4\n", "val_cases = 1\n", "test_cases = 1\n", "points_per_case = 128\n", "batch_size = 128\n", "\n", "[model]\n", "type = \"mlp\"\n", "hidden_width = 128\n", "depth = 4\n", "activation = \"gelu\"\n", "\n", "[optim]\n", "lr = 0.001\n", "weight_decay = 0.0\n", "steps = 500\n", "log_interval = 100\n", "\n", "[device]\n", "type = \"cuda\"\n", "allow_cpu_fallback = false\n", "benchmark_kernels = true\n", "\n", "[loss]\n", "type = \"normalized_mse\"\n", "\n", "split manifest:\n", "{\n", " \"test_ids\": [\n", " \"case_03\"\n", " ],\n", " \"train_ids\": [\n", " \"case_04\",\n", " \"case_02\",\n", " \"case_01\",\n", " \"case_00\"\n", " ],\n", " \"val_ids\": [\n", " \"case_05\"\n", " ]\n", "}\n", "normalization target names: ['velocity_x', 'velocity_y', 'pressure', 'turbulent_viscosity']\n" ] } ], "source": [ "split_manifest = json.loads((RUN_DIR / \"split_manifest.json\").read_text())\n", "normalization = json.loads((RUN_DIR / \"normalization.json\").read_text())\n", "config_text = (RUN_DIR / \"config.toml\").read_text()\n", "\n", "print(config_text)\n", "print(\"split manifest:\")\n", "print(json.dumps(split_manifest, indent=2))\n", "print(\"normalization target names:\", normalization[\"target_names\"])\n" ] }, { "cell_type": "code", "execution_count": 9, "id": "b16e05bc", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "feature_mean = np.array(normalization[\"feature_mean\"], dtype=np.float32)\n", "feature_std = np.array(normalization[\"feature_std\"], dtype=np.float32)\n", "target_mean = np.array(normalization[\"target_mean\"], dtype=np.float32)\n", "target_std = np.array(normalization[\"target_std\"], dtype=np.float32)\n", "\n", "fig, axes = plt.subplots(1, 2, figsize=(12, 4))\n", "axes[0].bar(normalization[\"feature_names\"], feature_mean)\n", "axes[0].set_title(\"feature means\")\n", "axes[0].tick_params(axis=\"x\", rotation=45)\n", "axes[0].grid(axis=\"y\", alpha=0.25)\n", "\n", "axes[1].bar(normalization[\"target_names\"], target_std)\n", "axes[1].set_title(\"target stds\")\n", "axes[1].tick_params(axis=\"x\", rotation=45)\n", "axes[1].grid(axis=\"y\", alpha=0.25)\n", "fig.tight_layout()\n" ] }, { "cell_type": "markdown", "id": "32f03cf3", "metadata": {}, "source": [ "## What the input features mean\n", "\n", "Each row is one sampled point from one case. The MLP sees five scalar inputs for that point:\n", "\n", "| feature | meaning | varies per |\n", "|---|---|---|\n", "| `re_norm` | normalized Reynolds-number-like case condition; constant across all points in one case | case |\n", "| `aoa_norm` | normalized angle-of-attack-like case condition; constant across all points in one case | case |\n", "| `x` | point x-coordinate in the local 2D domain | point |\n", "| `y` | point y-coordinate in the local 2D domain | point |\n", "| `sdf` | signed-distance-like geometry feature; here computed as radial distance from a radius-0.5 body proxy, so negative means inside/near the proxy and positive means farther away | point |\n", "\n", "For the current `data/processed/minimal` sanity dataset, these are deliberately compact synthetic features from the training-framework spec, not the full AirfRANS graph feature set. The target formulas used by the synthetic sanity data are simple functions of these columns, which is why the tiny MLP should learn them quickly." ] }, { "cell_type": "code", "execution_count": 10, "id": "c9b3b2fe", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "re_norm - case-level normalized Reynolds-number-like condition\n", "aoa_norm - case-level normalized angle-of-attack-like condition\n", "x - sampled point x-coordinate\n", "y - sampled point y-coordinate\n", "sdf - signed-distance-like geometry signal; negative near/inside radius-0.5 proxy\n" ] } ], "source": [ "feature_descriptions = {\n", " \"re_norm\": \"case-level normalized Reynolds-number-like condition\",\n", " \"aoa_norm\": \"case-level normalized angle-of-attack-like condition\",\n", " \"x\": \"sampled point x-coordinate\",\n", " \"y\": \"sampled point y-coordinate\",\n", " \"sdf\": \"signed-distance-like geometry signal; negative near/inside radius-0.5 proxy\",\n", "}\n", "\n", "for name in normalization[\"feature_names\"]:\n", " print(f\"{name:8} - {feature_descriptions.get(name, 'no description recorded')}\")\n" ] }, { "cell_type": "code", "execution_count": 11, "id": "2a31f98e", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[{'split': 'test',\n", " 'case_id': 'case_03',\n", " 'points': 256,\n", " 're_norm_min': 0.10000000149011612,\n", " 're_norm_mean': 0.10000001639127731,\n", " 're_norm_max': 0.10000000149011612,\n", " 'aoa_norm_min': 0.10000000149011612,\n", " 'aoa_norm_mean': 0.10000001639127731,\n", " 'aoa_norm_max': 0.10000000149011612,\n", " 'x_min': -0.992418110370636,\n", " 'x_mean': -0.028246253728866577,\n", " 'x_max': 0.9915499091148376,\n", " 'y_min': -0.999594509601593,\n", " 'y_mean': 0.00812380388379097,\n", " 'y_max': 0.9907281994819641,\n", " 'sdf_min': -0.483142614364624,\n", " 'sdf_mean': 0.24713799357414246,\n", " 'sdf_max': 0.855535626411438},\n", " {'split': 'train',\n", " 'case_id': 'case_04',\n", " 'points': 256,\n", " 're_norm_min': 0.30000001192092896,\n", " 're_norm_mean': 0.30000001192092896,\n", " 're_norm_max': 0.30000001192092896,\n", " 'aoa_norm_min': 0.20000000298023224,\n", " 'aoa_norm_mean': 0.20000003278255463,\n", " 'aoa_norm_max': 0.20000000298023224,\n", " 'x_min': -0.9973810315132141,\n", " 'x_mean': 0.016297580674290657,\n", " 'x_max': 0.9969028830528259,\n", " 'y_min': -0.9971365332603455,\n", " 'y_mean': -0.03903472423553467,\n", " 'y_max': 0.988767683506012,\n", " 'sdf_min': -0.4517866373062134,\n", " 'sdf_mean': 0.29825884103775024,\n", " 'sdf_max': 0.8830513954162598},\n", " {'split': 'train',\n", " 'case_id': 'case_02',\n", " 'points': 256,\n", " 're_norm_min': -0.10000000149011612,\n", " 're_norm_mean': -0.10000001639127731,\n", " 're_norm_max': -0.10000000149011612,\n", " 'aoa_norm_min': 0.0,\n", " 'aoa_norm_mean': 0.0,\n", " 'aoa_norm_max': 0.0,\n", " 'x_min': -0.9979289770126343,\n", " 'x_mean': 0.03891447186470032,\n", " 'x_max': 0.9994633793830872,\n", " 'y_min': -0.9955872893333435,\n", " 'y_mean': -0.018903596326708794,\n", " 'y_max': 0.9769426584243774,\n", " 'sdf_min': -0.4469470679759979,\n", " 'sdf_mean': 0.2548476755619049,\n", " 'sdf_max': 0.8520128726959229},\n", " {'split': 'train',\n", " 'case_id': 'case_01',\n", " 'points': 256,\n", " 're_norm_min': -0.30000001192092896,\n", " 're_norm_mean': -0.30000001192092896,\n", " 're_norm_max': -0.30000001192092896,\n", " 'aoa_norm_min': -0.10000000149011612,\n", " 'aoa_norm_mean': -0.10000001639127731,\n", " 'aoa_norm_max': -0.10000000149011612,\n", " 'x_min': -0.9986386895179749,\n", " 'x_mean': -0.039472322911024094,\n", " 'x_max': 0.991798460483551,\n", " 'y_min': -0.9973899126052856,\n", " 'y_mean': 0.029526807367801666,\n", " 'y_max': 0.998464822769165,\n", " 'sdf_min': -0.4342630207538605,\n", " 'sdf_mean': 0.2738206684589386,\n", " 'sdf_max': 0.9004930257797241},\n", " {'split': 'train',\n", " 'case_id': 'case_00',\n", " 'points': 256,\n", " 're_norm_min': -0.5,\n", " 're_norm_mean': -0.5,\n", " 're_norm_max': -0.5,\n", " 'aoa_norm_min': -0.20000000298023224,\n", " 'aoa_norm_mean': -0.20000003278255463,\n", " 'aoa_norm_max': -0.20000000298023224,\n", " 'x_min': -0.9898697733879089,\n", " 'x_mean': -0.004540023393929005,\n", " 'x_max': 0.9964226484298706,\n", " 'y_min': -0.9919538497924805,\n", " 'y_mean': -0.0564127042889595,\n", " 'y_max': 0.9963796734809875,\n", " 'sdf_min': -0.40864098072052,\n", " 'sdf_mean': 0.25872567296028137,\n", " 'sdf_max': 0.8634771108627319},\n", " {'split': 'val',\n", " 'case_id': 'case_05',\n", " 'points': 256,\n", " 're_norm_min': 0.5,\n", " 're_norm_mean': 0.5,\n", " 're_norm_max': 0.5,\n", " 'aoa_norm_min': 0.30000001192092896,\n", " 'aoa_norm_mean': 0.30000001192092896,\n", " 'aoa_norm_max': 0.30000001192092896,\n", " 'x_min': -0.986579179763794,\n", " 'x_mean': -0.012095458805561066,\n", " 'x_max': 0.9900404810905457,\n", " 'y_min': -0.9959282875061035,\n", " 'y_mean': 0.022574514150619507,\n", " 'y_max': 0.9964964985847473,\n", " 'sdf_min': -0.4551549255847931,\n", " 'sdf_mean': 0.25610190629959106,\n", " 'sdf_max': 0.8619978427886963}]" ] }, "execution_count": 11, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import tomllib\n", "\n", "config = tomllib.loads(config_text)\n", "DATA_ROOT = REPO_ROOT / config[\"data\"][\"root\"]\n", "\n", "\n", "def load_case(case_id: str) -> tuple[np.ndarray, np.ndarray]:\n", " data = np.load(DATA_ROOT / f\"{case_id}.npz\")\n", " return data[\"features\"].astype(np.float32), data[\"targets\"].astype(np.float32)\n", "\n", "\n", "feature_rows = []\n", "for split_name, ids in split_manifest.items():\n", " for case_id in ids:\n", " features, _ = load_case(case_id)\n", " row = {\"split\": split_name.replace(\"_ids\", \"\"), \"case_id\": case_id, \"points\": len(features)}\n", " for index, name in enumerate(normalization[\"feature_names\"]):\n", " values = features[:, index]\n", " row[f\"{name}_min\"] = float(values.min())\n", " row[f\"{name}_mean\"] = float(values.mean())\n", " row[f\"{name}_max\"] = float(values.max())\n", " feature_rows.append(row)\n", "\n", "feature_rows\n" ] }, { "cell_type": "code", "execution_count": 12, "id": "ccab3feb", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "case-level conditions for case_05:\n", "re_norm = 0.500\n", "aoa_norm = 0.300\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Visualize one case's point-level features. re_norm and aoa_norm are constant for the case;\n", "# x, y, and sdf vary over sampled points.\n", "case_id = (split_manifest.get(\"val_ids\") or split_manifest.get(\"test_ids\") or split_manifest.get(\"train_ids\"))[0]\n", "features, _ = load_case(case_id)\n", "feature_index = {name: idx for idx, name in enumerate(normalization[\"feature_names\"])}\n", "\n", "x_values = features[:, feature_index[\"x\"]]\n", "y_values = features[:, feature_index[\"y\"]]\n", "sdf_values = features[:, feature_index[\"sdf\"]]\n", "\n", "fig, axes = plt.subplots(1, 2, figsize=(11.5, 4.3))\n", "scatter = axes[0].scatter(x_values, y_values, c=sdf_values, s=18, cmap=\"coolwarm\")\n", "axes[0].set_title(f\"{case_id}: sampled points colored by sdf\")\n", "axes[0].set_xlabel(\"x\")\n", "axes[0].set_ylabel(\"y\")\n", "axes[0].set_aspect(\"equal\", adjustable=\"box\")\n", "axes[0].grid(alpha=0.25)\n", "fig.colorbar(scatter, ax=axes[0], label=\"sdf\")\n", "\n", "axes[1].hist(sdf_values, bins=30, edgecolor=\"white\")\n", "axes[1].set_title(\"sdf distribution for selected case\")\n", "axes[1].set_xlabel(\"sdf\")\n", "axes[1].set_ylabel(\"sampled points\")\n", "axes[1].grid(alpha=0.25)\n", "fig.tight_layout()\n", "\n", "print(f\"case-level conditions for {case_id}:\")\n", "print(f\"re_norm = {features[0, feature_index['re_norm']]:.3f}\")\n", "print(f\"aoa_norm = {features[0, feature_index['aoa_norm']]:.3f}\")\n" ] }, { "cell_type": "markdown", "id": "3c22f84b", "metadata": {}, "source": [ "## Checkpoint contents\n", "\n", "The checkpoint is saved with CPU-compatible tensors. This section verifies it can be loaded, shows metadata, and reconstructs the model.\n" ] }, { "cell_type": "code", "execution_count": 13, "id": "ed0720ec", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "checkpoint keys:\n", " step: int\n", " model_type: str\n", " input_dim: int\n", " output_dim: int\n", " model_config: dict\n", " model_state_dict: 10 tensors\n", " optimizer_state_dict: 2 tensors\n", " normalization: dict\n", " target_names: tuple\n", " feature_names: tuple\n", " final_metrics: dict\n", "\n", "step: 500\n", "model_type: mlp\n", "input_dim: 5\n", "output_dim: 4\n", "feature_names: ('re_norm', 'aoa_norm', 'x', 'y', 'sdf')\n", "target_names: ('velocity_x', 'velocity_y', 'pressure', 'turbulent_viscosity')\n", "model_config: {'type': 'mlp', 'hidden_width': 128, 'depth': 4, 'activation': 'gelu'}\n", "\n", "first model tensors:\n", "network.0.weight (128, 5) torch.float32\n", "network.0.bias (128,) torch.float32\n", "network.2.weight (128, 128) torch.float32\n", "network.2.bias (128,) torch.float32\n", "network.4.weight (128, 128) torch.float32\n", "network.4.bias (128,) torch.float32\n", "network.6.weight (128, 128) torch.float32\n", "network.6.bias (128,) torch.float32\n" ] } ], "source": [ "checkpoint = torch.load(RUN_DIR / \"checkpoint.pt\", map_location=\"cpu\")\n", "\n", "print(\"checkpoint keys:\")\n", "for key, value in checkpoint.items():\n", " if key.endswith(\"state_dict\"):\n", " print(f\" {key}: {len(value)} tensors\")\n", " else:\n", " print(f\" {key}: {type(value).__name__}\")\n", "\n", "print()\n", "print(\"step:\", checkpoint[\"step\"])\n", "print(\"model_type:\", checkpoint[\"model_type\"])\n", "print(\"input_dim:\", checkpoint[\"input_dim\"])\n", "print(\"output_dim:\", checkpoint[\"output_dim\"])\n", "print(\"feature_names:\", checkpoint[\"feature_names\"])\n", "print(\"target_names:\", checkpoint[\"target_names\"])\n", "print(\"model_config:\", checkpoint[\"model_config\"])\n", "\n", "print()\n", "print(\"first model tensors:\")\n", "for name, tensor in list(checkpoint[\"model_state_dict\"].items())[:8]:\n", " print(name, tuple(tensor.shape), tensor.dtype)\n" ] }, { "cell_type": "markdown", "id": "94b4e5d8", "metadata": {}, "source": [ "## MLP architecture\n", "\n", "The training loop derives `input_dim` from normalized feature columns and `output_dim` from normalized target columns. This cell expands that into the concrete network shape, layer table, and parameter count so the abstraction is visible at a glance." ] }, { "cell_type": "code", "execution_count": 14, "id": "e48801e5", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "MLP shape at a glance\n", "----------------------\n", "inputs (5): ['re_norm', 'aoa_norm', 'x', 'y', 'sdf']\n", "outputs (4): ['velocity_x', 'velocity_y', 'pressure', 'turbulent_viscosity']\n", "architecture: 5 -> 128 -> 128 -> 128 -> 128 -> 4\n", "activation after hidden layers: gelu\n", "hidden layers: 4\n", "hidden width: 128\n", "\n", "layer table:\n", " 0: Linear 5 -> 128 params=768\n", " 1: GELU -> params=0\n", " 2: Linear 128 -> 128 params=16512\n", " 3: GELU -> params=0\n", " 4: Linear 128 -> 128 params=16512\n", " 5: GELU -> params=0\n", " 6: Linear 128 -> 128 params=16512\n", " 7: GELU -> params=0\n", " 8: Linear 128 -> 4 params=516\n", "\n", "total parameters: 50,820\n" ] } ], "source": [ "input_names = list(checkpoint[\"feature_names\"])\n", "target_names = list(checkpoint[\"target_names\"])\n", "model_config = checkpoint[\"model_config\"]\n", "\n", "print(\"MLP shape at a glance\")\n", "print(\"----------------------\")\n", "print(f\"inputs ({len(input_names)}): {input_names}\")\n", "print(f\"outputs ({len(target_names)}): {target_names}\")\n", "print(\n", " \"architecture: \"\n", " f\"{checkpoint['input_dim']} -> \"\n", " + \" -> \".join([str(model_config[\"hidden_width\"])] * model_config[\"depth\"])\n", " + f\" -> {checkpoint['output_dim']}\"\n", ")\n", "print(f\"activation after hidden layers: {model_config['activation']}\")\n", "print(f\"hidden layers: {model_config['depth']}\")\n", "print(f\"hidden width: {model_config['hidden_width']}\")\n", "print()\n", "\n", "model = PointwiseMLP(\n", " input_dim=checkpoint[\"input_dim\"],\n", " output_dim=checkpoint[\"output_dim\"],\n", " hidden_width=model_config[\"hidden_width\"],\n", " depth=model_config[\"depth\"],\n", " activation=model_config[\"activation\"],\n", ")\n", "model.load_state_dict(checkpoint[\"model_state_dict\"])\n", "model.eval()\n", "\n", "layer_rows = []\n", "for index, layer in enumerate(model.network):\n", " row = {\"index\": index, \"type\": layer.__class__.__name__}\n", " if isinstance(layer, torch.nn.Linear):\n", " row[\"in_features\"] = layer.in_features\n", " row[\"out_features\"] = layer.out_features\n", " row[\"parameters\"] = layer.weight.numel() + layer.bias.numel()\n", " else:\n", " row[\"in_features\"] = \"\"\n", " row[\"out_features\"] = \"\"\n", " row[\"parameters\"] = 0\n", " layer_rows.append(row)\n", "\n", "print(\"layer table:\")\n", "for row in layer_rows:\n", " print(\n", " f\"{row['index']:>2}: {row['type']:<8} \"\n", " f\"{str(row['in_features']):>4} -> {str(row['out_features']):<4} \"\n", " f\"params={row['parameters']}\"\n", " )\n", "\n", "parameter_count = sum(parameters.numel() for parameters in model.parameters())\n", "print()\n", "print(f\"total parameters: {parameter_count:,}\")\n", "assert parameter_count == final_metrics[\"parameter_count\"]\n" ] }, { "cell_type": "markdown", "id": "6f638448", "metadata": {}, "source": [ "## Reloaded-model predictions\n", "\n", "This section runs the checkpoint on real `.npz` cases from the split. Predictions are de-normalized back to the target scale before comparison.\n" ] }, { "cell_type": "code", "execution_count": 15, "id": "09e67ecb", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "/home/aaron/data/airfrans/data/processed/minimal\n", "['case_00.npz', 'case_01.npz', 'case_02.npz', 'case_03.npz', 'case_04.npz', 'case_05.npz']\n" ] } ], "source": [ "def config_value(config: str, section: str, key: str) -> str:\n", " current_section = None\n", " for raw_line in config.splitlines():\n", " line = raw_line.strip()\n", " if not line or line.startswith(\"#\"):\n", " continue\n", " if line.startswith(\"[\") and line.endswith(\"]\"):\n", " current_section = line[1:-1]\n", " continue\n", " if current_section == section and line.startswith(f\"{key} =\"):\n", " return line.split(\"=\", 1)[1].strip().strip('\\\"')\n", " raise KeyError(f\"{section}.{key}\")\n", "\n", "\n", "DATA_ROOT = REPO_ROOT / config_value(config_text, \"data\", \"root\")\n", "print(DATA_ROOT)\n", "print(sorted(path.name for path in DATA_ROOT.glob(\"*.npz\")))\n" ] }, { "cell_type": "code", "execution_count": 16, "id": "924e2cef", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[{'split': 'test',\n", " 'case_id': 'case_03',\n", " 'rows': 256,\n", " 'normalized_mse': 0.00025704235304147005,\n", " 'mse_velocity_x': 5.847604370501358e-06,\n", " 'mse_velocity_y': 6.842037691967562e-06,\n", " 'mse_pressure': 4.433627691469155e-05,\n", " 'mse_turbulent_viscosity': 1.3116708032612223e-05},\n", " {'split': 'train',\n", " 'case_id': 'case_04',\n", " 'rows': 256,\n", " 'normalized_mse': 0.00018982301116921008,\n", " 'mse_velocity_x': 6.853222657809965e-06,\n", " 'mse_velocity_y': 6.716706138831796e-06,\n", " 'mse_pressure': 2.8528093025670387e-05,\n", " 'mse_turbulent_viscosity': 9.420269634574652e-06},\n", " {'split': 'train',\n", " 'case_id': 'case_02',\n", " 'rows': 256,\n", " 'normalized_mse': 0.0001731793163344264,\n", " 'mse_velocity_x': 5.711051926482469e-06,\n", " 'mse_velocity_y': 4.464081939659081e-06,\n", " 'mse_pressure': 2.741067874012515e-05,\n", " 'mse_turbulent_viscosity': 9.072812645172235e-06},\n", " {'split': 'train',\n", " 'case_id': 'case_01',\n", " 'rows': 256,\n", " 'normalized_mse': 0.00011508660099934787,\n", " 'mse_velocity_x': 3.957362423534505e-06,\n", " 'mse_velocity_y': 2.683263573999284e-06,\n", " 'mse_pressure': 2.2211312170838937e-05,\n", " 'mse_turbulent_viscosity': 5.00482428833493e-06},\n", " {'split': 'train',\n", " 'case_id': 'case_00',\n", " 'rows': 256,\n", " 'normalized_mse': 0.00016892868734430522,\n", " 'mse_velocity_x': 8.186969353118911e-06,\n", " 'mse_velocity_y': 7.417508186335908e-06,\n", " 'mse_pressure': 3.0236758902901784e-05,\n", " 'mse_turbulent_viscosity': 5.846786280017113e-06},\n", " {'split': 'val',\n", " 'case_id': 'case_05',\n", " 'rows': 256,\n", " 'normalized_mse': 0.0007373533444479108,\n", " 'mse_velocity_x': 2.7578798835747875e-05,\n", " 'mse_velocity_y': 1.9889514078386128e-05,\n", " 'mse_pressure': 0.00013202131958678365,\n", " 'mse_turbulent_viscosity': 3.3118620194727555e-05}]" ] }, "execution_count": 16, "metadata": {}, "output_type": "execute_result" } ], "source": [ "def load_case(case_id: str) -> tuple[np.ndarray, np.ndarray]:\n", " data = np.load(DATA_ROOT / f\"{case_id}.npz\")\n", " return data[\"features\"].astype(np.float32), data[\"targets\"].astype(np.float32)\n", "\n", "\n", "def predict_targets(features: np.ndarray, batch_size: int = 4096) -> np.ndarray:\n", " features_norm = (features - feature_mean) / feature_std\n", " outputs = []\n", " with torch.no_grad():\n", " for start in range(0, len(features_norm), batch_size):\n", " batch = torch.from_numpy(features_norm[start : start + batch_size])\n", " pred_norm = model(batch).numpy()\n", " outputs.append(pred_norm)\n", " pred_norm_all = np.concatenate(outputs, axis=0)\n", " return pred_norm_all * target_std + target_mean\n", "\n", "\n", "def normalized_mse(predictions: np.ndarray, targets: np.ndarray) -> float:\n", " pred_norm = (predictions - target_mean) / target_std\n", " target_norm = (targets - target_mean) / target_std\n", " return float(np.mean((pred_norm - target_norm) ** 2))\n", "\n", "\n", "case_rows = []\n", "for split_name, ids in split_manifest.items():\n", " for case_id in ids:\n", " features, targets = load_case(case_id)\n", " predictions = predict_targets(features)\n", " per_channel_mse = np.mean((predictions - targets) ** 2, axis=0)\n", " case_rows.append(\n", " {\n", " \"split\": split_name.replace(\"_ids\", \"\"),\n", " \"case_id\": case_id,\n", " \"rows\": len(features),\n", " \"normalized_mse\": normalized_mse(predictions, targets),\n", " **{f\"mse_{name}\": float(value) for name, value in zip(checkpoint[\"target_names\"], per_channel_mse)},\n", " }\n", " )\n", "\n", "case_rows\n" ] }, { "cell_type": "code", "execution_count": 17, "id": "4fe67c98", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "case: case_05\n", "target names: ('velocity_x', 'velocity_y', 'pressure', 'turbulent_viscosity')\n", "first 5 predictions:\n", "[[ 0.44094524 -0.02950683 0.19028291 0.0982319 ]\n", " [ 0.02009592 0.14751981 -0.21558036 0.03777217]\n", " [ 0.45918038 -0.00707006 0.2556273 0.12591958]\n", " [ 0.31692797 0.19948506 -0.60716456 0.3552236 ]\n", " [ 0.23262833 0.05736347 -0.33526948 0.07037625]]\n", "first 5 actual targets:\n", "[[ 0.44482228 -0.03109705 0.18664566 0.09674599]\n", " [ 0.01824543 0.1526474 -0.20674932 0.03895786]\n", " [ 0.4640298 -0.00911091 0.25503194 0.12455965]\n", " [ 0.32461593 0.1904991 -0.6266293 0.35012734]\n", " [ 0.23149356 0.05860274 -0.33879325 0.06851949]]\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Pick a validation case if present, otherwise test, otherwise train.\n", "case_id = (split_manifest.get(\"val_ids\") or split_manifest.get(\"test_ids\") or split_manifest.get(\"train_ids\"))[0]\n", "features, targets = load_case(case_id)\n", "predictions = predict_targets(features)\n", "\n", "print(f\"case: {case_id}\")\n", "print(\"target names:\", checkpoint[\"target_names\"])\n", "print(\"first 5 predictions:\")\n", "print(predictions[:5])\n", "print(\"first 5 actual targets:\")\n", "print(targets[:5])\n", "\n", "fig, axes = plt.subplots(1, len(checkpoint[\"target_names\"]), figsize=(4 * len(checkpoint[\"target_names\"]), 3.8))\n", "if len(checkpoint[\"target_names\"]) == 1:\n", " axes = [axes]\n", "for index, (ax, name) in enumerate(zip(axes, checkpoint[\"target_names\"])):\n", " ax.scatter(targets[:, index], predictions[:, index], s=12, alpha=0.7)\n", " low = float(min(targets[:, index].min(), predictions[:, index].min()))\n", " high = float(max(targets[:, index].max(), predictions[:, index].max()))\n", " ax.plot([low, high], [low, high], color=\"black\", linewidth=1)\n", " ax.set_title(name)\n", " ax.set_xlabel(\"actual\")\n", " ax.set_ylabel(\"predicted\")\n", " ax.grid(alpha=0.25)\n", "fig.suptitle(f\"Reloaded checkpoint predictions for {case_id}\", y=1.03)\n", "fig.tight_layout()\n" ] }, { "cell_type": "markdown", "id": "5901d5d7", "metadata": {}, "source": [ "## What good looks like\n", "\n", "For this tiny baseline sanity run:\n", "\n", "- `train_loss` is finite and much lower than `initial_train_loss`.\n", "- `val_loss` and `test_loss` are finite.\n", "- The loss curve falls on a log-scale plot rather than staying flat or exploding.\n", "- `checkpoint.pt` loads on CPU and reconstructs `PointwiseMLP` without missing/unexpected keys.\n", "- Reloaded predictions have the expected target channels and roughly follow the identity line in predicted-vs-actual plots.\n", "\n", "This does not prove the model is scientifically useful. It proves the training framework can run end-to-end, write readable artifacts, reload the model, and produce numerically sensible outputs on the processed mini dataset.\n" ] } ], "metadata": { "kernelspec": { "display_name": ".venv", "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": "3.12.12" } }, "nbformat": 4, "nbformat_minor": 5 }