{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"nbsphinx": "hidden"
},
"source": [
"It looks like you might be running this notebook in Colab! If you want to enable GPU acceleration, ensure you select a GPU runtime in the top-right dropdown menu 🔥"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Train a model\n",
"\n",
"> **FYI**, you can open this documentation as a [Google Colab notebook](https://colab.research.google.com/github/jla-gardner/graph-pes/blob/main/docs/source/quickstart/quickstart.ipynb) to follow along interactively\n",
"\n",
"[graph-pes-train](https://jla-gardner.github.io/graph-pes/cli/graph-pes-train/root.html) provides a unified interface to train any [GraphPESModel](https://jla-gardner.github.io/graph-pes/models/root.html#graph_pes.GraphPESModel), including those packaged within [graph_pes.models](https://jla-gardner.github.io/graph-pes/models/root.html), custom ones defined by you, and any of the wrapper interfaces that ``graph-pes`` provides to other machine learning frameworks.\n",
"\n",
"For more information on the ``graph-pes-train`` command, and the plethora of options available for specification in your ``config.yaml`` see the [CLI reference](https://jla-gardner.github.io/graph-pes/cli/graph-pes-train/root.html).\n",
"\n",
"Below, we train a lightweight [NequIP](https://jla-gardner.github.io/graph-pes/models/many-body/nequip.html) model on the [C-GAP-17](https://jla-gardner.github.io/load-atoms/datasets/C-GAP-17.html) dataset.\n",
"\n",
"## Installation\n"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Successfully installed graph-pes-0.0.34\n"
]
}
],
"source": [
"!pip install graph-pes"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We should now have access to the ``graph-pes-train`` command. We can check this by running:"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"usage: graph-pes-train [-h] [args ...]\n",
"\n",
"Train a GraphPES model using PyTorch Lightning.\n",
"\n",
"positional arguments:\n",
" args Config files and command line specifications. Config files\n",
" should be YAML (.yaml/.yml) files. Command line specifications\n",
" should be in the form my/nested/key=value. Final config is built\n",
" up from these items in a left to right manner, with later items\n",
" taking precedence over earlier ones in the case of conflicts.\n",
"\n",
"optional arguments:\n",
" -h, --help show this help message and exit\n",
"\n",
"Copyright 2023-25, John Gardner\n"
]
}
],
"source": [
"!graph-pes-train -h"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Data definition\n",
"\n",
"When training a model, we typically want 3 sets of data (i.e. labelled atomic structures): a training set, a validation set, and a test set.\n",
"\n",
"Below, we use [load-atoms](https://jla-gardner.github.io/load-atoms/) to download and split the C-GAP-17 dataset into training, validation and test datasets, and write these to `xyz` files. (``graph-pes`` supports other dataset formats too, including ``ase sqlite`` databases -- see [here](https://jla-gardner.github.io/graph-pes/data/datasets.html#useful-datasets) for more details)"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "4c9284a1589c41feb63442c9f5e05a53",
"version_major": 2,
"version_minor": 0
},
"text/plain": [
"Output()"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"\n",
"
\n",
"\n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" \n",
"\n",
"\n",
" \n",
"\n",
"
"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"from load_atoms import view\n",
"\n",
"view(train[0], show_bonds=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"As you can see, each structure has an ``energy`` label:"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"-5643.968171"
]
},
"execution_count": 17,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"train[0].info[\"energy\"]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"... as well as a ``forces`` label (one for each atom in the structure):"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(36, 3)"
]
},
"execution_count": 19,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"train[0].arrays[\"forces\"].shape"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"These properties are stored in the files we have just created:"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"36\n",
"Lattice=\"6.439806 0.0 0.0 0.0 6.439806 0.0 0.0 0.0 8.586408\" Properties=species:S:1:pos:R:3:forces:R:3 config_type=bulk_amo detailed_ct=iter4_2 split=train energy=-5643.968171 pbc=\"T T T\"\n",
"C -10.19681458 4.52108512 2.58260263 1.92054269 0.70905554 3.23398419\n",
"C 8.88245018 10.54923296 9.85602863 -8.61008207 4.76824471 10.54597273\n",
"C -12.37947091 3.12898582 0.00437048 0.40437923 -0.84438408 0.64039651\n",
"C -15.36751513 3.67112089 7.46005158 -2.19558355 -5.72081017 -10.70417213\n",
"C 1.48348659 8.44603096 -5.40849254 2.13894508 -3.77202448 2.30942937\n",
"C 6.68203286 2.50162636 -11.97770429 0.95262647 -0.03136726 -0.79207429\n",
"C -15.10503508 0.29261834 6.90345486 -2.22105290 2.34675963 -6.38519060\n",
"C -4.49987240 0.25155170 -7.21023863 0.27927988 0.68979268 1.56306647\n"
]
}
],
"source": [
"!head train-cgap17.xyz"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Configuration\n",
"\n",
"Now that we've saved our labelled structures to suitable files, we're ready to train a model.\n",
"\n",
"To do this, we have specified the following in the ``quickstart-cgap17.yaml`` file:\n",
"\n",
"* the model architecture to instantiate and train, here [NequIP](https://jla-gardner.github.io/graph-pes/models/many-body/nequip.html). Note that we also include a [FixedOffset](https://jla-gardner.github.io/graph-pes/models/offsets.html#graph_pes.models.FixedOffset) component to account for the fact that the C-GAP-17 labels have an arbitrary offset.\n",
"* the data to train on, here a random split of the [C-GAP-17](https://jla-gardner.github.io/load-atoms/datasets/C-GAP-17.html) dataset we just downloaded\n",
"* the loss function to use, here a combination of a per-atom energy loss and a per-atom force loss\n",
"* and various other training hyperparameters (e.g. the learning rate, batch size, etc.)\n",
"\n"
]
},
{
"cell_type": "raw",
"metadata": {
"raw_mimetype": "text/restructuredtext",
"vscode": {
"languageId": "raw"
}
},
"source": [
".. dropdown:: ``quickstart-cgap17.yaml``\n",
"\n",
" .. literalinclude:: quickstart-cgap17.yaml\n",
" :language: yaml"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can download [this config file](https://raw.githubusercontent.com/jla-gardner/graph-pes/refs/heads/main/docs/source/quickstart/quickstart-cgap17.yaml) using wget:"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"%%bash\n",
"\n",
"if [ ! -f quickstart-cgap17.yaml ]; then\n",
" wget https://tinyurl.com/graph-pes-quickstart-cgap17 -O quickstart-cgap17.yaml\n",
"fi"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Training\n",
"\n",
"To train the model, we use the [graph-pes-train](https://jla-gardner.github.io/graph-pes/cli/graph-pes-train/root.html) command.\n",
"\n",
"You can see the output of the original training run I ran in this [Weights and Biases dashboard](https://wandb.ai/jla-gardner/graph-pes-quickstart/runs/train-nequip).\n"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[graph-pes INFO]: Started `graph-pes-train` at 2025-04-11 11:21:47.393\n",
"[graph-pes INFO]: Successfully parsed config.\n",
"[graph-pes INFO]: Logging to graph-pes-results/train-nequip/rank-0.log\n",
"[graph-pes INFO]: ID for this training run: train-nequip\n",
"[graph-pes INFO]: \n",
"Output for this training run can be found at:\n",
" └─ graph-pes-results/train-nequip\n",
" ├─ rank-0.log # find a verbose log here\n",
" ├─ model.pt # the best model (according to valid/loss/total)\n",
" ├─ lammps_model.pt # the best model deployed to LAMMPS\n",
" ├─ train-config.yaml # the complete config used for this run\n",
" └─ summary.yaml # the summary of the training run\n",
"\n",
"GPU available: True (cuda), used: True\n",
"TPU available: False, using: 0 TPU cores\n",
"HPU available: False, using: 0 HPUs\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: Using wandb-core as the SDK backend. Please refer to https://wandb.me/wandb-core for more information.\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: Currently logged in as: \u001b[33mjla-gardner\u001b[0m to \u001b[32mhttps://api.wandb.ai\u001b[0m. Use \u001b[1m`wandb login --relogin`\u001b[0m to force relogin\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: Tracking run with wandb version 0.19.9\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: Run data is saved locally in \u001b[35m\u001b[1mgraph-pes-results/wandb/run-20250411_112150-train-nequip\u001b[0m\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: Run \u001b[1m`wandb offline`\u001b[0m to turn off syncing.\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: Syncing run \u001b[33mtrain-nequip\u001b[0m\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: ⭐️ View project at \u001b[34m\u001b[4mhttps://wandb.ai/jla-gardner/graph-pes-quickstart\u001b[0m\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 🚀 View run at \u001b[34m\u001b[4mhttps://wandb.ai/jla-gardner/graph-pes-quickstart/runs/train-nequip\u001b[0m\n",
"[graph-pes INFO]: Preparing data\n",
"[graph-pes INFO]: Setting up datasets\n",
"[graph-pes INFO]: Pre-fitting the model on 1,280 samples\n",
"[graph-pes INFO]: \n",
"Number of learnable params:\n",
" offset (FixedOffset): 0\n",
" many-body (NequIP) : 4,233\n",
"\n",
"[graph-pes INFO]: Sanity checking the model...\n",
"[graph-pes INFO]: Starting fit...\n",
"LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n",
"/home/calcite/vld/jesu2890/graph-pes/.venv/lib/python3.9/site-packages/pytorch_lightning/trainer/connectors/data_connector.py:425: The 'val_dataloader' does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` to `num_workers=35` in the `DataLoader` to improve performance.\n",
"/home/calcite/vld/jesu2890/graph-pes/.venv/lib/python3.9/site-packages/pytorch_lightning/trainer/connectors/data_connector.py:425: The 'train_dataloader' does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` to `num_workers=35` in the `DataLoader` to improve performance.\n",
"/home/calcite/vld/jesu2890/graph-pes/.venv/lib/python3.9/site-packages/pytorch_lightning/loops/fit_loop.py:310: The number of training batches (20) is smaller than the logging interval Trainer(log_every_n_steps=50). Set a lower value for log_every_n_steps if you want to see logs for the training epoch.\n",
" valid/metrics valid/metrics valid/metrics valid/metrics valid/metrics valid/metrics timer/its_per_s timer/its_per_s\n",
" epoch time per_atom_energy_rmse per_atom_energy_mae energy_rmse energy_mae forces_rmse forces_mae train valid\n",
" 5 8.7 0.62315 0.48031 35.21251 29.65179 1.39054 1.04576 21.27660 36.14532\n",
" 10 17.3 0.25256 0.12514 8.46061 6.64184 1.06623 0.81321 21.27660 36.29926\n",
" 15 25.8 0.30826 0.27898 20.66504 16.95788 0.97669 0.75126 21.27660 35.42593\n",
" 20 34.1 0.26312 0.23339 17.40298 14.02269 0.94271 0.72719 21.27660 36.31066\n",
" 25 42.7 0.13665 0.10314 8.13669 6.32626 0.91978 0.70968 21.27660 36.44294\n",
" 30 51.2 0.16764 0.12787 9.88240 7.46267 0.90267 0.69606 21.73913 36.60828\n",
" 35 59.3 0.16015 0.13001 10.08410 7.73408 0.89099 0.68656 21.27660 36.47427\n",
" 40 67.8 0.16035 0.13706 10.79822 8.60235 0.87526 0.67691 21.27660 36.45434\n",
" 45 76.4 0.11800 0.08278 6.48016 4.68341 0.86985 0.67295 21.27660 36.45434\n",
" 50 84.9 0.09155 0.05939 4.56686 3.27326 0.86002 0.66365 20.83333 36.28899\n",
" 55 93.4 0.13949 0.11225 8.63337 6.53863 0.85629 0.66144 20.40816 36.28900\n",
" 60 101.5 0.10444 0.07786 6.25070 4.86074 0.84143 0.64941 20.83333 36.60828\n",
" 65 110.0 0.11025 0.08216 6.76819 5.04506 0.84349 0.65043 21.27660 36.44294\n",
" 70 118.2 0.12537 0.10013 7.94580 5.97333 0.83898 0.64568 21.27660 36.63748\n",
" 75 126.4 0.08905 0.05865 4.85177 3.39582 0.83235 0.64291 20.83333 36.00164\n",
" 80 135.0 0.08428 0.05420 4.29199 3.05238 0.81563 0.62996 20.83333 36.14532\n",
" 85 143.6 0.08458 0.05217 4.20698 2.94109 0.83121 0.64083 20.83333 36.43267\n",
" 90 151.8 0.15399 0.13354 10.28086 8.12393 0.80666 0.62302 20.40816 36.31066\n",
" 95 160.0 0.08045 0.04999 4.00663 2.77370 0.80932 0.62424 20.00000 36.19618\n",
" 100 168.6 0.14926 0.13008 10.68001 8.59175 0.80941 0.62616 20.83333 36.00164\n",
" 105 176.8 0.17366 0.15842 12.33288 10.00910 0.81932 0.63167 20.83333 36.44294\n",
" 110 185.0 0.12503 0.10650 8.62543 6.94882 0.79976 0.61861 20.83333 36.14532\n",
" 115 193.1 0.11280 0.09268 7.43654 5.89931 0.79611 0.61464 21.73913 36.76162\n",
" 120 201.3 0.10204 0.07421 6.05022 4.22249 0.79113 0.61078 20.40816 36.33059\n",
" 125 209.5 0.08550 0.05395 4.18860 2.93031 0.79094 0.61093 20.83333 36.01091\n",
" 130 218.1 0.09214 0.06968 5.29362 4.10699 0.80209 0.61774 20.83333 36.48453\n",
" 135 226.3 0.09654 0.07404 6.03257 4.73452 0.78720 0.60859 20.40816 36.15672\n",
" 140 234.5 0.13333 0.11478 8.86486 7.14934 0.79077 0.61130 21.73913 36.33059\n",
" 145 242.7 0.10346 0.08377 6.58328 5.23925 0.78829 0.60864 21.73913 36.49593\n",
" 150 250.9 0.08593 0.05443 4.48703 3.11985 0.78729 0.60804 21.73913 36.31066\n",
" 155 259.4 0.07854 0.05421 4.18323 3.20169 0.78762 0.60745 21.27660 36.14532\n",
" 160 268.0 0.13054 0.11253 8.83321 7.05205 0.78856 0.61020 21.27660 36.61795\n",
" 165 276.1 0.13436 0.11566 9.05795 7.12900 0.78077 0.60236 21.27660 36.47427\n",
" 170 284.3 0.10105 0.07397 6.03266 4.18178 0.78658 0.60791 20.83333 36.61795\n",
" 175 292.5 0.09442 0.07199 6.01171 4.44474 0.78398 0.60551 21.73913 36.00164\n",
" 180 300.7 0.08973 0.07091 5.71718 4.47134 0.77670 0.60024 21.27660 36.00164\n",
" 185 308.8 0.16145 0.14754 11.38972 9.28238 0.77806 0.60138 21.27660 36.63961\n",
" 190 317.0 0.07035 0.04360 3.59079 2.42807 0.78680 0.60764 20.83333 36.44294\n",
" 195 325.5 0.11411 0.09597 7.54478 5.83939 0.77606 0.59926 20.83333 36.28899\n",
" 200 333.7 0.07146 0.04575 3.67843 2.56994 0.78213 0.60312 20.83333 36.44294\n",
" 205 342.2 0.10034 0.07853 6.16118 4.51715 0.77907 0.60105 20.83333 36.02117\n",
" 210 350.4 0.09016 0.06550 5.56907 3.87992 0.77324 0.59818 21.27660 36.45434\n",
" 215 358.6 0.06793 0.04424 3.41206 2.43981 0.77539 0.59888 21.27660 36.78329\n",
" 220 367.1 0.22132 0.21056 16.07953 13.36178 0.80770 0.62159 20.83333 36.44294\n",
" 225 375.3 0.07922 0.05494 4.71128 3.32506 0.77587 0.59897 20.00000 36.00164\n",
" 230 383.5 0.06517 0.04025 3.21447 2.26121 0.77772 0.60106 20.40816 36.28900\n",
" 235 392.0 0.07309 0.04843 4.15934 2.81432 0.77288 0.59675 21.73913 36.14532\n",
" 240 400.1 0.07428 0.05531 4.23787 3.27787 0.77770 0.59988 21.27660 36.31066\n",
" 245 408.3 0.12987 0.11495 8.94932 7.16290 0.77176 0.59641 20.40816 36.29926\n",
" 250 416.5 0.07047 0.05124 3.92862 3.03030 0.77647 0.60081 20.83333 36.48453\n",
"`Trainer.fit` stopped: `max_epochs=250` reached.\n",
"[graph-pes INFO]: Loading best weights from \"/home/calcite/vld/jesu2890/graph-pes/docs/source/quickstart/graph-pes-results/train-nequip/checkpoints/best.ckpt\"\n",
"[graph-pes INFO]: Training complete.\n",
"[graph-pes INFO]: Testing best model...\n",
"You are using the plain ModelCheckpoint callback. Consider using LitModelCheckpoint which with seamless uploading to Model registry.\n",
"GPU available: True (cuda), used: True\n",
"TPU available: False, using: 0 TPU cores\n",
"HPU available: False, using: 0 HPUs\n",
"LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n",
"/home/calcite/vld/jesu2890/graph-pes/.venv/lib/python3.9/site-packages/pytorch_lightning/trainer/connectors/data_connector.py:425: The 'test_dataloader' does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` to `num_workers=35` in the `DataLoader` to improve performance.\n",
"Testing DataLoader 2: 100%|███████████████████████| 8/8 [00:00<00:00, 11.64it/s]\n",
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓\n",
"┃\u001b[1m \u001b[0m\u001b[1m Test metric \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m DataLoader 0 \u001b[0m\u001b[1m \u001b[0m┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/forces_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/forces_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/test/forces_rmse_batchwise \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/per_atom_energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/per_atom_energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 2.2048792839050293 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 3.148214817047119 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/forces_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.6043804883956909 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/forces_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.7820170521736145 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/train/forces_rmse_batchwise\u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.7818971872329712 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/per_atom_energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.036605119705200195 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/train/per_atom_energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.049800578504800797 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/forces_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/forces_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/valid/forces_rmse_batchwise\u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/per_atom_energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/valid/per_atom_energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"└──────────────────────────────────┴──────────────────────────────────┘\n",
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓\n",
"┃\u001b[1m \u001b[0m\u001b[1m Test metric \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m DataLoader 1 \u001b[0m\u001b[1m \u001b[0m┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/forces_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/forces_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/test/forces_rmse_batchwise \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/per_atom_energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/per_atom_energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/forces_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/forces_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/train/forces_rmse_batchwise\u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/per_atom_energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/train/per_atom_energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 2.2612078189849854 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 3.16567063331604 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/forces_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.6010645627975464 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/forces_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.7808666825294495 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/valid/forces_rmse_batchwise\u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.7776868939399719 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/per_atom_energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.04025261849164963 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/valid/per_atom_energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.05424034968018532 \u001b[0m\u001b[35m \u001b[0m│\n",
"└──────────────────────────────────┴──────────────────────────────────┘\n",
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓\n",
"┃\u001b[1m \u001b[0m\u001b[1m Test metric \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m DataLoader 2 \u001b[0m\u001b[1m \u001b[0m┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 2.414318561553955 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 3.2008798122406006 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/forces_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.6022810935974121 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/forces_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.780879020690918 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/test/forces_rmse_batchwise \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.7811077237129211 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/per_atom_energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.041916195303201675 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/test/per_atom_energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.05537675321102142 \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/forces_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/forces_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/train/forces_rmse_batchwise\u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/train/per_atom_energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/train/per_atom_energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/forces_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/forces_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/valid/forces_rmse_batchwise\u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36m test/valid/per_atom_energy_mae \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"│\u001b[36m \u001b[0m\u001b[36mtest/valid/per_atom_energy_rmse \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m \u001b[0m\u001b[35m \u001b[0m│\n",
"└──────────────────────────────────┴──────────────────────────────────┘\n",
"[graph-pes INFO]: Testing complete.\n",
"[graph-pes INFO]: Awaiting final Lightning and W&B shutdown...\n",
"\u001b[1;34mwandb\u001b[0m: \n",
"\u001b[1;34mwandb\u001b[0m: 🚀 View run \u001b[33mtrain-nequip\u001b[0m at: \u001b[34mhttps://wandb.ai/jla-gardner/graph-pes-quickstart/runs/train-nequip\u001b[0m\n"
]
}
],
"source": [
"!graph-pes-train quickstart-cgap17.yaml general/run_id=train-nequip"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Model analysis\n",
"\n",
"As part of the ``graph-pes-train`` run, the model was tested on the test set we specified in the config file (see the final section of the logs above).\n",
"\n",
"To analyse the model in more detail, we first need to load it from disk.\n",
"You can see from the command we used, and the training logs above, that the best model from the training run (i.e. the set of weights that gave the lowest validation loss) has been saved as ``graph-pes-results/train-nequip/model.pt``.\n",
"\n",
"Let's load that best model, put it on the GPU for accelerated inference if available, and get it ready for evaluation:"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"import torch\n",
"from graph_pes.models import load_model\n",
"\n",
"device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
"best_model = (\n",
" load_model(\"graph-pes-results/train-nequip/model.pt\") # load the model\n",
" .to(device) # move to GPU if available\n",
" .eval() # set to evaluation mode\n",
")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The easiest way to use our model is to use the [GraphPESCalculator](https://jla-gardner.github.io/graph-pes/tools/ase.html#graph_pes.utils.calculator.GraphPESCalculator) to act directly on [ase.Atoms](https://wiki.fysik.dtu.dk/ase/ase/atoms.html#module-ase.atoms) objects:\n",
"\n",
""
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"{'energy': -9994.06640625,\n",
" 'forces': array([[-4.2254949e+00, 5.9772301e-01, 8.1164581e-01],\n",
" [-9.8559284e-01, 2.4811971e+00, -8.7177944e+00],\n",
" [ 2.1818477e-01, -3.9657693e+00, 5.4294224e+00],\n",
" [-2.2534198e-01, -6.7757750e-01, 3.9687052e-01],\n",
" [-1.6062340e-01, 1.6424814e+00, 1.6815267e+00],\n",
" [ 3.8409348e+00, -2.3497558e+00, 1.0593975e+00],\n",
" [ 2.6760058e+00, 1.4250646e+00, 1.5218904e+00],\n",
" [ 1.8543882e+00, -6.4903003e-01, -1.2432039e+00],\n",
" [ 1.2891605e+00, 1.8805935e+00, -1.1635600e+00],\n",
" [ 6.2873564e+00, 6.9245615e+00, 2.0475407e+00],\n",
" [ 1.1643579e+00, 9.7333103e-01, 2.1005793e+00],\n",
" [-4.5520332e-01, -1.0353167e+00, 9.0359229e-01],\n",
" [ 2.0876718e+00, -2.0918713e+00, -1.7296244e+00],\n",
" [ 7.9885268e-01, 2.8175316e+00, 4.0605450e+00],\n",
" [ 7.1246386e-01, -3.1887600e+00, -2.6775827e+00],\n",
" [-6.5839797e-02, -1.8803604e+00, -2.6095929e+00],\n",
" [-2.1252718e+00, 1.7640973e+00, -8.7072477e-02],\n",
" [-5.2741468e-02, -9.8783600e-01, -4.0110202e+00],\n",
" [-2.6717019e+00, -3.9906960e+00, -3.7461739e+00],\n",
" [ 5.5837898e+00, 4.5657673e+00, -1.3618302e+00],\n",
" [-8.3528572e-01, 4.5085421e+00, -2.4682379e+00],\n",
" [-9.1291595e-01, 4.3814039e+00, 3.4088683e+00],\n",
" [-1.9985964e+00, -6.2825119e-01, -1.0749267e+00],\n",
" [-3.2094419e-03, -2.0321414e+00, 7.1312678e-01],\n",
" [ 6.3161063e-01, -2.6924348e+00, 2.3123736e+00],\n",
" [ 9.2511237e-01, -1.5875727e+01, 3.2328601e+00],\n",
" [ 1.1260903e-01, 8.7551790e-01, -2.2891724e-01],\n",
" [ 4.2413688e+00, 3.1473446e+00, -1.8300122e+00],\n",
" [ 4.2376418e+00, 3.6069365e+00, -1.5691175e+00],\n",
" [-2.3413777e+00, 3.6433258e+00, 6.4318693e-01],\n",
" [-2.1698081e-01, -7.6708636e+00, -1.4212118e+00],\n",
" [-4.6958515e-01, -6.1527020e-01, -2.7724439e-01],\n",
" [ 6.1276513e-01, 3.1687143e+00, 2.5931001e+00],\n",
" [-4.0827327e+00, 2.9075354e-01, 2.4999285e+00],\n",
" [ 2.6901686e-01, -1.8677192e+00, -1.0254226e+00],\n",
" [-6.8144751e+00, -4.8341813e+00, 4.1638579e+00],\n",
" [ 7.2648329e-01, -4.7186742e+00, -1.0133797e+00],\n",
" [ 5.0572991e-02, 8.6126399e-01, 1.5870973e+00],\n",
" [ 9.3245506e-02, -3.3464322e+00, -1.3165364e+00],\n",
" [ 4.5235795e-01, 2.3301392e+00, 1.1348567e+00],\n",
" [-1.8452766e+00, 1.3446496e+00, -7.6116276e-01],\n",
" [ 9.4428259e-01, 1.2193685e+00, -6.4569908e-01],\n",
" [-7.1592855e-01, 6.9566975e+00, 1.0641152e-01],\n",
" [ 1.4557087e+00, 1.4954034e+01, -2.1514454e+00],\n",
" [ 1.3845201e+00, 1.4602650e+00, -2.5141425e+00],\n",
" [ 4.8026240e-01, -1.2326572e+00, -1.6421114e+00],\n",
" [-9.1919702e-01, 1.5869403e-01, 2.8427663e+00],\n",
" [-1.3124061e+00, -4.3377948e+00, -3.8554204e-01],\n",
" [ 5.0649233e+00, 2.3024335e+00, -1.2069061e+00],\n",
" [-2.9632771e+00, -3.0417490e+00, 3.1358151e+00],\n",
" [-2.2530427e+00, -5.0966077e+00, 2.2911510e+00],\n",
" [-5.0742316e+00, 7.7338238e+00, 6.0399427e+00],\n",
" [-1.0427495e+00, -6.5423548e-02, 3.4518700e+00],\n",
" [-5.3656592e+00, 1.3809166e+00, 1.3118589e+00],\n",
" [ 2.0248129e+00, 1.1432605e+00, -6.3636169e+00],\n",
" [ 1.3965229e+00, -5.2572565e+00, -4.1274199e+00],\n",
" [ 7.5580758e-01, -2.7472277e+00, 3.2230120e+00],\n",
" [-2.8447154e+00, -3.5290098e-01, -1.6035125e+00],\n",
" [-9.3414867e-01, -1.0672436e+00, -1.5498080e+00],\n",
" [-1.9216528e+00, 2.6706436e+00, -1.8475140e+00],\n",
" [ 1.3133743e+00, -3.7629669e+00, 4.7236176e+00],\n",
" [-1.2435791e+00, -2.0665624e+00, -3.4463222e+00],\n",
" [ 7.7818692e-01, -9.4866943e-01, 1.5642781e+00],\n",
" [ 2.6144829e+00, 1.8646499e+00, -3.1753240e+00]], dtype=float32),\n",
" 'stress': array([ 0.03880393, 0.03565932, 0.05275575, -0.00330898, -0.00880773,\n",
" -0.00247441], dtype=float32)}"
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"calculator = best_model.ase_calculator()\n",
"calculator.calculate(test[0], properties=[\"energy\", \"forces\", \"stress\"])\n",
"calculator.results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can see from a single data point that our model has done a reasonable job of learning the PES:"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(-9994.06640625, np.float64(-9998.70784))"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"calculator.get_potential_energy(test[0]), test[0].info[\"energy\"]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"``graph-pes`` provides a few utility functions for visualising model performance:"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {
"image/png": {
"height": 308,
"width": 385
}
},
"output_type": "display_data"
}
],
"source": [
"import matplotlib.pyplot as plt\n",
"from graph_pes.utils.analysis import parity_plot\n",
"\n",
"%config InlineBackend.figure_format = 'retina'\n",
"\n",
"parity_plot(\n",
" best_model,\n",
" test,\n",
" property=\"energy_per_atom\",\n",
" units=\"eV / atom\",\n",
" lw=0,\n",
" s=12,\n",
" color=\"crimson\",\n",
")\n",
"plt.xlim(-158.5, -155)\n",
"plt.ylim(-158.5, -155);"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {
"image/png": {
"height": 309,
"width": 365
}
},
"output_type": "display_data"
}
],
"source": [
"parity_plot(\n",
" best_model,\n",
" test,\n",
" property=\"forces\",\n",
" units=\"eV / Å\",\n",
" lw=0,\n",
" s=2,\n",
" alpha=0.5,\n",
" color=\"crimson\",\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {
"image/png": {
"height": 309,
"width": 363
}
},
"output_type": "display_data"
}
],
"source": [
"from graph_pes.utils.analysis import dimer_curve\n",
"\n",
"dimer_curve(best_model, system=\"CC\", units=\"eV\", rmin=0.85, rmax=4.0);"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Beyond static evaluations, there are many more use cases for these models - head over to e.g. our [ASE examples](https://jla-gardner.github.io/graph-pes/tools/ase.html) notebook for more details"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "graph-pes",
"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": "3.9.22"
}
},
"nbformat": 4,
"nbformat_minor": 2
}