diff --git a/bayesflow/experimental/__init__.py b/bayesflow/experimental/__init__.py new file mode 100644 index 000000000..e69de29bb diff --git a/bayesflow/networks/consistency_models/continuous_consistency_model.py b/bayesflow/experimental/continuous_time_consistency_model.py similarity index 98% rename from bayesflow/networks/consistency_models/continuous_consistency_model.py rename to bayesflow/experimental/continuous_time_consistency_model.py index a4ca2941a..e0f9a262c 100644 --- a/bayesflow/networks/consistency_models/continuous_consistency_model.py +++ b/bayesflow/experimental/continuous_time_consistency_model.py @@ -19,12 +19,12 @@ ) -from ..inference_network import InferenceNetwork -from ..embeddings import FourierEmbedding +from bayesflow.networks import InferenceNetwork +from bayesflow.networks.embeddings import FourierEmbedding @register_keras_serializable(package="bayesflow.networks") -class ContinuousConsistencyModel(InferenceNetwork): +class ContinuousTimeConsistencyModel(InferenceNetwork): """Implements an sCM (simple, stable, and scalable Consistency Model) with continous-time Consistency Training (CT) as described in [1]. The sampling procedure is taken from [2]. diff --git a/bayesflow/networks/__init__.py b/bayesflow/networks/__init__.py index 9a915572b..7ac7fcc89 100644 --- a/bayesflow/networks/__init__.py +++ b/bayesflow/networks/__init__.py @@ -1,5 +1,5 @@ from .cif import CIF -from .consistency_models import ConsistencyModel, ContinuousConsistencyModel +from .consistency_models import ConsistencyModel from .coupling_flow import CouplingFlow from .deep_set import DeepSet from .flow_matching import FlowMatching diff --git a/bayesflow/networks/consistency_models/__init__.py b/bayesflow/networks/consistency_models/__init__.py index 33592d62b..b40725ced 100644 --- a/bayesflow/networks/consistency_models/__init__.py +++ b/bayesflow/networks/consistency_models/__init__.py @@ -1,2 +1 @@ from .consistency_model import ConsistencyModel -from .continuous_consistency_model import ContinuousConsistencyModel diff --git a/examples/Continuous_Consistency_Model_Playground.ipynb b/examples/Continuous_Consistency_Model_Playground.ipynb deleted file mode 100644 index 015a31de0..000000000 --- a/examples/Continuous_Consistency_Model_Playground.ipynb +++ /dev/null @@ -1,679 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "id": "c3036e98", - "metadata": {}, - "source": [ - "# Amortization with Continuous Consistency Models\n", - "\n", - "_Author: Valentin Pratz_" - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "id": "d5f88a59", - "metadata": { - "ExecuteTime": { - "end_time": "2024-10-24T08:36:22.149034Z", - "start_time": "2024-10-24T08:36:20.807192Z" - } - }, - "outputs": [], - "source": [ - "%load_ext autoreload\n", - "%autoreload 2\n", - "import matplotlib.pyplot as plt\n", - "import numpy as np\n", - "import seaborn as sns\n", - "\n", - "# Ensure the backend is set\n", - "import os\n", - "if \"KERAS_BACKEND\" not in os.environ:\n", - " # set this to \"torch\", \"tensorflow\", or \"jax\"\n", - " os.environ[\"KERAS_BACKEND\"] = \"jax\"\n", - "\n", - "import keras\n", - "\n", - "# For BayesFlow devs: this ensures that the latest dev version can be found\n", - "import sys\n", - "sys.path.append('../')\n", - "\n", - "import bayesflow as bf" - ] - }, - { - "cell_type": "markdown", - "id": "315dcf39-c29f-40dc-ad52-69252f9514e4", - "metadata": {}, - "source": [ - "This notebook serves as a playground for testing continuous-time consistency models. Later on, it will probably evolve into a full tutorial notebook. For now, please refer to the starter notebook if you encounter concepts that are not explained." - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "id": "2425f2a2-5aec-4eca-882e-4d93bc82a80c", - "metadata": {}, - "outputs": [], - "source": [ - "simulator = bf.simulators.TwoMoons()" - ] - }, - { - "cell_type": "markdown", - "id": "f6e1eb5777c59eba", - "metadata": {}, - "source": [ - "Let's generate some data to see what the simulator does:" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "id": "e6218e61d529e357", - "metadata": { - "ExecuteTime": { - "end_time": "2024-10-24T08:36:22.350483Z", - "start_time": "2024-10-24T08:36:22.345161Z" - } - }, - "outputs": [], - "source": [ - "# generate 3 random draws from the joint distribution p(r, alpha, theta, x)\n", - "sample_data = simulator.sample((20,))" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "id": "46174ccb0167026c", - "metadata": { - "ExecuteTime": { - "end_time": "2024-10-24T08:36:22.470435Z", - "start_time": "2024-10-24T08:36:22.464836Z" - } - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Type of sample_data:\n", - "\t \n", - "Keys of sample_data:\n", - "\t dict_keys(['r', 'alpha', 'theta', 'x'])\n", - "Types of sample_data values:\n", - "\t {'r': , 'alpha': , 'theta': , 'x': }\n", - "Shapes of sample_data values:\n", - "\t {'r': (20, 1), 'alpha': (20, 1), 'theta': (20, 2), 'x': (20, 2)}\n" - ] - } - ], - "source": [ - "print(\"Type of sample_data:\\n\\t\", type(sample_data))\n", - "print(\"Keys of sample_data:\\n\\t\", sample_data.keys())\n", - "print(\"Types of sample_data values:\\n\\t\", {k: type(v) for k, v in sample_data.items()})\n", - "print(\"Shapes of sample_data values:\\n\\t\", {k: v.shape for k, v in sample_data.items()})" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "id": "5c9c2dc70f53d103", - "metadata": { - "ExecuteTime": { - "end_time": "2024-10-24T08:36:26.618926Z", - "start_time": "2024-10-24T08:36:26.614443Z" - } - }, - "outputs": [ - { - "data": { - "text/plain": [ - "Adapter([Keep(['theta', 'x']) -> ToArray -> ConvertDType -> Standardize -> Rename('theta' -> 'inference_variables') -> Rename('x' -> 'inference_conditions')])" - ] - }, - "execution_count": 5, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "adapter = (\n", - " bf.adapters.Adapter()\n", - "\n", - " # drop data that we do not need\n", - " .keep((\"theta\", \"x\"))\n", - " \n", - " # convert any non-arrays to numpy arrays\n", - " .to_array()\n", - " \n", - " # convert from numpy's default float64 to deep learning friendly float32\n", - " .convert_dtype(\"float64\", \"float32\")\n", - " \n", - " # standardize all variables to zero mean and unit variance\n", - " .standardize(momentum=None) # standardization with momentum is currently not working\n", - " \n", - " # rename the variables to match the required approximator inputs\n", - " .rename(\"theta\", \"inference_variables\")\n", - " .rename(\"x\", \"inference_conditions\")\n", - ")\n", - "adapter" - ] - }, - { - "cell_type": "markdown", - "id": "254e287b2bccdad", - "metadata": {}, - "source": [ - "## Dataset\n", - "\n", - "For this example, we will sample our training data ahead of time and use offline training with a `bf.datasets.OfflineDataset`.\n", - "\n", - "This makes the training process faster, since we avoid repeated sampling. If you want to use online training, you can use an `OnlineDataset` analogously, or just pass your simulator directly to `approximator.fit()`!" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "id": "39cb5a1c9824246f", - "metadata": { - "ExecuteTime": { - "end_time": "2024-09-23T14:39:46.950573Z", - "start_time": "2024-09-23T14:39:46.948624Z" - } - }, - "outputs": [], - "source": [ - "num_training_batches = 512\n", - "num_validation_batches = 128\n", - "batch_size = 64\n", - "epochs = 30\n", - "total_steps = num_training_batches * epochs" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "id": "9dee7252ef99affa", - "metadata": { - "ExecuteTime": { - "end_time": "2024-09-23T14:39:53.268860Z", - "start_time": "2024-09-23T14:39:46.994697Z" - } - }, - "outputs": [], - "source": [ - "training_samples = simulator.sample((num_training_batches * batch_size,))\n", - "validation_samples = simulator.sample((num_validation_batches * batch_size,))" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "id": "51045bbed88cb5c2", - "metadata": { - "ExecuteTime": { - "end_time": "2024-09-23T14:39:53.281170Z", - "start_time": "2024-09-23T14:39:53.275921Z" - } - }, - "outputs": [], - "source": [ - "training_dataset = bf.datasets.OfflineDataset(\n", - " data=training_samples, \n", - " batch_size=batch_size, \n", - " adapter=adapter\n", - ")\n", - "\n", - "validation_dataset = bf.datasets.OfflineDataset(\n", - " data=validation_samples, \n", - " batch_size=batch_size, \n", - " adapter=adapter\n", - ")" - ] - }, - { - "cell_type": "markdown", - "id": "2d4c6eb0", - "metadata": {}, - "source": [ - "## Training a neural network to approximate all posteriors" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "id": "09206e6f", - "metadata": { - "ExecuteTime": { - "end_time": "2024-09-23T14:39:53.339590Z", - "start_time": "2024-09-23T14:39:53.319852Z" - } - }, - "outputs": [], - "source": [ - "inference_network = bf.networks.ContinuousConsistencyModel(\n", - " subnet=\"mlp\",\n", - " sigma_data=1.0, # as we have standardized our parameters, the standard deviation is 1.0\n", - " subnet_kwargs={\"widths\": (256,)*6, \"dropout\": 0.0}, # use an inner network with 6 hidden layers of 256 units\n", - " embedding_kwargs={\"embed_dim\": 2}\n", - ")" - ] - }, - { - "cell_type": "markdown", - "id": "851e522f", - "metadata": {}, - "source": [ - "This inference network is just a general Flow Matching backbone, not yet adapted to the specific inference task at hand (i.e., posterior appproximation). To achieve this adaptation, we combine the network with our data adapter, which together form an `approximator`. In this case, we need a `ContinuousApproximator` since the target we want to approximate is the posterior of the *continuous* parameter vector $\\theta$." - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "id": "96ca6ffa", - "metadata": { - "ExecuteTime": { - "end_time": "2024-09-23T14:39:53.371691Z", - "start_time": "2024-09-23T14:39:53.369375Z" - } - }, - "outputs": [], - "source": [ - "cm_approximator = bf.ContinuousApproximator(\n", - " inference_network=inference_network,\n", - " adapter=adapter,\n", - ")" - ] - }, - { - "cell_type": "markdown", - "id": "566264eadc76c2c", - "metadata": {}, - "source": [ - "### Optimizer and Learning Rate\n", - "We find learning rate schedules, such as [cosine decay](https://keras.io/api/optimizers/learning_rate_schedules/cosine_decay/), work well for a wide variety of approximation tasks." - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "id": "e8d7e053", - "metadata": { - "ExecuteTime": { - "end_time": "2024-09-23T14:39:53.433012Z", - "start_time": "2024-09-23T14:39:53.415903Z" - } - }, - "outputs": [], - "source": [ - "initial_learning_rate = 5e-4\n", - "scheduled_lr = keras.optimizers.schedules.CosineDecay(\n", - " initial_learning_rate=initial_learning_rate,\n", - " decay_steps=total_steps,\n", - " alpha=1e-8\n", - ")\n", - "\n", - "optimizer = keras.optimizers.Adam(learning_rate=scheduled_lr)" - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "id": "51808fcd560489ac", - "metadata": { - "ExecuteTime": { - "end_time": "2024-09-23T14:39:53.476089Z", - "start_time": "2024-09-23T14:39:53.466001Z" - } - }, - "outputs": [], - "source": [ - "cm_approximator.compile(optimizer=optimizer)" - ] - }, - { - "cell_type": "markdown", - "id": "708b1303", - "metadata": {}, - "source": [ - "### Training\n", - "\n", - "We are ready to train our deep posterior approximator on the two moons example. We pass the dataset object to the `fit` method and watch as Bayesflow trains." - ] - }, - { - "cell_type": "code", - "execution_count": 13, - "id": "0f496bda", - "metadata": { - "ExecuteTime": { - "end_time": "2024-09-23T14:42:36.067393Z", - "start_time": "2024-09-23T14:39:53.513436Z" - } - }, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "INFO:bayesflow:Fitting on dataset instance of OfflineDataset.\n", - "INFO:bayesflow:Building on a test batch.\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Epoch 1/30\n", - "512/512 - 4s - 8ms/step - loss: -9.0581e-01 - loss/inference_loss: -9.0581e-01 - val_loss: -9.1429e-01 - val_loss/inference_loss: -9.1429e-01\n", - "Epoch 2/30\n", - "512/512 - 3s - 6ms/step - loss: -9.7834e-01 - loss/inference_loss: -9.7834e-01 - val_loss: -9.6997e-01 - val_loss/inference_loss: -9.6997e-01\n", - "Epoch 3/30\n", - "512/512 - 3s - 5ms/step - loss: -9.8114e-01 - loss/inference_loss: -9.8114e-01 - val_loss: -9.8932e-01 - val_loss/inference_loss: -9.8932e-01\n", - "Epoch 4/30\n", - "512/512 - 3s - 5ms/step - loss: -9.3032e-01 - loss/inference_loss: -9.3032e-01 - val_loss: -9.3277e-01 - val_loss/inference_loss: -9.3277e-01\n", - "Epoch 5/30\n", - "512/512 - 3s - 5ms/step - loss: -1.0659e+00 - loss/inference_loss: -1.0659e+00 - val_loss: -1.0488e+00 - val_loss/inference_loss: -1.0488e+00\n", - "Epoch 6/30\n", - "512/512 - 3s - 6ms/step - loss: -1.0973e+00 - loss/inference_loss: -1.0973e+00 - val_loss: -9.5333e-01 - val_loss/inference_loss: -9.5333e-01\n", - "Epoch 7/30\n", - "512/512 - 3s - 5ms/step - loss: -9.8796e-01 - loss/inference_loss: -9.8796e-01 - val_loss: -9.1071e-01 - val_loss/inference_loss: -9.1071e-01\n", - "Epoch 8/30\n", - "512/512 - 3s - 5ms/step - loss: -1.0167e+00 - loss/inference_loss: -1.0167e+00 - val_loss: -1.0523e+00 - val_loss/inference_loss: -1.0523e+00\n", - "Epoch 9/30\n", - "512/512 - 3s - 5ms/step - loss: -1.0387e+00 - loss/inference_loss: -1.0387e+00 - val_loss: -9.8802e-01 - val_loss/inference_loss: -9.8802e-01\n", - "Epoch 10/30\n", - "512/512 - 3s - 5ms/step - loss: -9.8820e-01 - loss/inference_loss: -9.8820e-01 - val_loss: -9.0955e-01 - val_loss/inference_loss: -9.0955e-01\n", - "Epoch 11/30\n", - "512/512 - 3s - 5ms/step - loss: -9.9296e-01 - loss/inference_loss: -9.9296e-01 - val_loss: -8.9993e-01 - val_loss/inference_loss: -8.9993e-01\n", - "Epoch 12/30\n", - "512/512 - 3s - 5ms/step - loss: -1.0739e+00 - loss/inference_loss: -1.0739e+00 - val_loss: -9.5941e-01 - val_loss/inference_loss: -9.5941e-01\n", - "Epoch 13/30\n", - "512/512 - 3s - 5ms/step - loss: -1.1442e+00 - loss/inference_loss: -1.1442e+00 - val_loss: -1.0661e+00 - val_loss/inference_loss: -1.0661e+00\n", - "Epoch 14/30\n", - "512/512 - 3s - 5ms/step - loss: -1.1030e+00 - loss/inference_loss: -1.1030e+00 - val_loss: -1.0260e+00 - val_loss/inference_loss: -1.0260e+00\n", - "Epoch 15/30\n", - "512/512 - 3s - 6ms/step - loss: -1.0170e+00 - loss/inference_loss: -1.0170e+00 - val_loss: -1.0567e+00 - val_loss/inference_loss: -1.0567e+00\n", - "Epoch 16/30\n", - "512/512 - 3s - 5ms/step - loss: -1.0415e+00 - loss/inference_loss: -1.0415e+00 - val_loss: -1.0732e+00 - val_loss/inference_loss: -1.0732e+00\n", - "Epoch 17/30\n", - "512/512 - 3s - 5ms/step - loss: -1.0504e+00 - loss/inference_loss: -1.0504e+00 - val_loss: -1.1912e+00 - val_loss/inference_loss: -1.1912e+00\n", - "Epoch 18/30\n", - "512/512 - 3s - 5ms/step - loss: -1.0053e+00 - loss/inference_loss: -1.0053e+00 - val_loss: -1.0339e+00 - val_loss/inference_loss: -1.0339e+00\n", - "Epoch 19/30\n", - "512/512 - 3s - 5ms/step - loss: -1.1080e+00 - loss/inference_loss: -1.1080e+00 - val_loss: -1.0603e+00 - val_loss/inference_loss: -1.0603e+00\n", - "Epoch 20/30\n", - "512/512 - 3s - 5ms/step - loss: -1.1058e+00 - loss/inference_loss: -1.1058e+00 - val_loss: -9.7252e-01 - val_loss/inference_loss: -9.7252e-01\n", - "Epoch 21/30\n", - "512/512 - 3s - 5ms/step - loss: -1.0851e+00 - loss/inference_loss: -1.0851e+00 - val_loss: -1.0735e+00 - val_loss/inference_loss: -1.0735e+00\n", - "Epoch 22/30\n", - "512/512 - 3s - 5ms/step - loss: -1.0744e+00 - loss/inference_loss: -1.0744e+00 - val_loss: -1.1786e+00 - val_loss/inference_loss: -1.1786e+00\n", - "Epoch 23/30\n", - "512/512 - 3s - 6ms/step - loss: -1.0129e+00 - loss/inference_loss: -1.0129e+00 - val_loss: -1.1164e+00 - val_loss/inference_loss: -1.1164e+00\n", - "Epoch 24/30\n", - "512/512 - 3s - 6ms/step - loss: -1.1158e+00 - loss/inference_loss: -1.1158e+00 - val_loss: -1.0653e+00 - val_loss/inference_loss: -1.0653e+00\n", - "Epoch 25/30\n", - "512/512 - 3s - 6ms/step - loss: -1.0603e+00 - loss/inference_loss: -1.0603e+00 - val_loss: -1.0243e+00 - val_loss/inference_loss: -1.0243e+00\n", - "Epoch 26/30\n", - "512/512 - 3s - 5ms/step - loss: -1.1274e+00 - loss/inference_loss: -1.1274e+00 - val_loss: -1.1840e+00 - val_loss/inference_loss: -1.1840e+00\n", - "Epoch 27/30\n", - "512/512 - 3s - 5ms/step - loss: -1.0597e+00 - loss/inference_loss: -1.0597e+00 - val_loss: -1.0407e+00 - val_loss/inference_loss: -1.0407e+00\n", - "Epoch 28/30\n", - "512/512 - 3s - 5ms/step - loss: -1.0995e+00 - loss/inference_loss: -1.0995e+00 - val_loss: -1.1431e+00 - val_loss/inference_loss: -1.1431e+00\n", - "Epoch 29/30\n", - "512/512 - 3s - 5ms/step - loss: -1.1468e+00 - loss/inference_loss: -1.1468e+00 - val_loss: -1.0729e+00 - val_loss/inference_loss: -1.0729e+00\n", - "Epoch 30/30\n", - "512/512 - 3s - 6ms/step - loss: -1.0368e+00 - loss/inference_loss: -1.0368e+00 - val_loss: -1.1233e+00 - val_loss/inference_loss: -1.1233e+00\n", - "CPU times: user 2min 57s, sys: 12.4 s, total: 3min 9s\n", - "Wall time: 1min 25s\n" - ] - } - ], - "source": [ - "%%time\n", - "fm_history = cm_approximator.fit(\n", - " epochs=epochs,\n", - " dataset=training_dataset,\n", - " validation_data=validation_dataset,\n", - " verbose=2, # set verbose=2 to avoid flooding the notebook\n", - ")" - ] - }, - { - "cell_type": "markdown", - "id": "f4785f35-794e-40c7-b863-f5100a90ef13", - "metadata": {}, - "source": [ - "Note that after a certain time, the loss is no longer indicative of training performance." - ] - }, - { - "cell_type": "code", - "execution_count": 14, - "id": "209e4bbd-9d4e-4639-82b7-0e974f7258ca", - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "# Set the number of posterior draws you want to get\n", - "num_samples = 3000\n", - "steps = 15\n", - "rho = 3.5\n", - "\n", - "# Obtain samples from amortized posterior\n", - "\n", - "# conditions = {\"x\": np.array([[0.0, 0.0]]).astype(\"float32\")}\n", - "# samples_0 = cm_approximator.sample(conditions=conditions, batch_size=1, num_samples=num_samples)[\"theta\"][0]\n", - "\n", - "# manually sample using _inverse to have access to sampling parameters\n", - "# (will not be necessary anymore when .sample forwards those arguments.\n", - "# Take care to correctly apply the data adapter.\n", - "samples_0 = adapter.inverse({\n", - " \"inference_variables\": keras.ops.convert_to_numpy(\n", - " cm_approximator.inference_network._inverse(\n", - " keras.random.normal((num_samples, 2)),\n", - " conditions=adapter.forward({\"x\": np.zeros((num_samples, 2))}, strict=False)[\"inference_conditions\"],\n", - " steps=steps, rho=rho)\n", - " ),\n", - "}, strict=False)[\"theta\"]\n", - "\n", - "# Prepare figure\n", - "f, axes = plt.subplots(1, 2, figsize=(12, 6))\n", - "\n", - "# Plot samples (once without limits to see outliers/problems\n", - "samples = [samples_0, samples_0]\n", - "names = [\"Continuous-time CM\", \"Without Axis Limits\"]\n", - "colors = [\"#153c7a\", \"#7a1515\"]\n", - "\n", - "for ax, thetas, name, color in zip(axes, samples, names, colors):\n", - "\n", - " # Plot samples\n", - " ax.scatter(thetas[:, 0], thetas[:, 1], color=color, alpha=0.75, s=0.5)\n", - " sns.despine(ax=ax)\n", - " ax.set_title(f\"{name}\", fontsize=16)\n", - " ax.grid(alpha=0.3)\n", - " ax.set_aspect(\"equal\", adjustable=\"box\")\n", - " if not name.lower().startswith(\"without\"):\n", - " ax.set_xlim([-0.5, 0.5])\n", - " ax.set_ylim([-0.5, 0.5])\n", - " ax.set_xlabel(r\"$\\theta_1$\", fontsize=15)\n", - " ax.set_ylabel(r\"$\\theta_2$\", fontsize=15)\n", - "\n", - "f.tight_layout()" - ] - }, - { - "cell_type": "markdown", - "id": "1e4f4062-a535-41e5-808b-321346038993", - "metadata": {}, - "source": [ - "Plot the discretization schedule:" - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "id": "e728c992-e91f-4d51-ad6b-33d7f759201b", - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "discretized_time = np.flip(keras.ops.convert_to_numpy(\n", - " cm_approximator.inference_network._discretize_time(steps, rho=rho)\n", - "))\n", - "plt.plot(discretized_time, marker=\"x\", linestyle=\"dotted\")\n", - "plt.ylabel(\"t\")\n", - "plt.xlabel(\"step\")\n", - "plt.ylim(0.0, np.pi/2)\n", - "plt.xlim(0, len(discretized_time) - 1)\n", - "plt.yticks([0.0, np.pi/8, np.pi/4, 3*np.pi/8, np.pi/2], labels=[\"0\", \"$\\pi/8$\", \"$\\pi/4$\", \"$3\\pi/8$\", \"$\\pi/2$\"])\n", - "plt.grid()\n", - "_ = plt.title(\"Discretization schedule\")" - ] - }, - { - "cell_type": "markdown", - "id": "7f8532f2-bbe1-4690-b74f-285d94960ab5", - "metadata": {}, - "source": [ - "Plot the time embedding:" - ] - }, - { - "cell_type": "code", - "execution_count": 16, - "id": "ffbecec3-b297-48db-b99d-9095b3e9f7a9", - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "t = keras.ops.linspace(0.001, np.pi/2, 500)[:, None]\n", - "emb = inference_network.time_emb(t)\n", - "plt.plot(keras.ops.convert_to_numpy(t)[:,0], keras.ops.convert_to_numpy(emb))\n", - "plt.ylabel(\"emb(t)\")\n", - "plt.xlabel(\"t\")\n", - "plt.xticks([0.0, np.pi/4, np.pi/2], labels=[\"0\", \"$\\pi/4$\", \"$\\pi/2$\"])\n", - "_ = plt.title(\"Time embedding\")" - ] - }, - { - "cell_type": "markdown", - "id": "acce03bb-a802-40a9-ab4f-4a2c3cc0f3de", - "metadata": {}, - "source": [ - "Plot the learned adaptive weighting function:" - ] - }, - { - "cell_type": "code", - "execution_count": 17, - "id": "e233f634-3f34-4f0e-baa8-6204f41de971", - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "plt.plot(t, inference_network.weight_fn_projector(inference_network.weight_fn(t)))\n", - "plt.plot(t, 1/(inference_network.sigma_data*np.tan(t)))\n", - "plt.plot(t, inference_network.weight_fn_projector(inference_network.weight_fn(t))/(inference_network.sigma_data*np.tan(t)))\n", - "plt.ylabel(\"$w_\\phi(t)$\")\n", - "plt.xlabel(\"t\")\n", - "plt.yscale(\"log\")\n", - "plt.xticks([0.0, np.pi/4, np.pi/2], labels=[\"0\", \"$\\pi/4$\", \"$\\pi/2$\"])\n", - "_ = plt.title(\"Learned adaptive weighting function\")" - ] - }, - { - "cell_type": "code", - "execution_count": 18, - "id": "2afa35bf", - "metadata": {}, - "outputs": [], - "source": [ - "metric = bf.metrics.MaximumMeanDiscrepancy()" - ] - } - ], - "metadata": { - "kernelspec": { - "display_name": "Python 3 (ipykernel)", - "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": "3.11.10" - }, - "toc": { - "base_numbering": 1, - "nav_menu": {}, - "number_sections": true, - "sideBar": true, - "skip_h1_title": true, - "title_cell": "Table of Contents", - "title_sidebar": "Contents", - "toc_cell": true, - "toc_position": { - "height": "calc(100% - 180px)", - "left": "10px", - "top": "150px", - "width": "165px" - }, - "toc_section_display": true, - "toc_window_display": true - } - }, - "nbformat": 4, - "nbformat_minor": 5 -} diff --git a/examples/experimental/Continuous_Consistency_Model_Playground.ipynb b/examples/experimental/Continuous_Consistency_Model_Playground.ipynb new file mode 100644 index 000000000..75446639e --- /dev/null +++ b/examples/experimental/Continuous_Consistency_Model_Playground.ipynb @@ -0,0 +1,674 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "c3036e98", + "metadata": {}, + "source": [ + "# Amortization with Continuous Consistency Models\n", + "\n", + "_Author: Valentin Pratz_" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "id": "d5f88a59", + "metadata": { + "ExecuteTime": { + "end_time": "2024-10-24T08:36:22.149034Z", + "start_time": "2024-10-24T08:36:20.807192Z" + } + }, + "outputs": [], + "source": [ + "%load_ext autoreload\n", + "%autoreload 2\n", + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "import seaborn as sns\n", + "\n", + "# Ensure the backend is set\n", + "import os\n", + "\n", + "if \"KERAS_BACKEND\" not in os.environ:\n", + " # set this to \"torch\", \"tensorflow\", or \"jax\"\n", + " os.environ[\"KERAS_BACKEND\"] = \"jax\"\n", + "\n", + "import keras\n", + "\n", + "# For BayesFlow devs: this ensures that the latest dev version can be found\n", + "import sys\n", + "\n", + "sys.path.append(\"../..\")\n", + "\n", + "import bayesflow as bf\n", + "from bayesflow.experimental.continuous_time_consistency_model import (\n", + " ContinuousTimeConsistencyModel,\n", + ")" + ] + }, + { + "cell_type": "markdown", + "id": "315dcf39-c29f-40dc-ad52-69252f9514e4", + "metadata": {}, + "source": [ + "This notebook serves as a playground for testing continuous-time consistency models. They are still experimental, as they are not fully tested yet, so please use them at your own discretion and share potential successes/failures. This will help us to judge their usefulness in the context of SBI. Later on, this notebook might evolve into a full tutorial notebook. For now, please refer to the starter notebook if you encounter concepts that are not explained." + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "id": "2425f2a2-5aec-4eca-882e-4d93bc82a80c", + "metadata": {}, + "outputs": [], + "source": [ + "simulator = bf.simulators.TwoMoons()" + ] + }, + { + "cell_type": "markdown", + "id": "f6e1eb5777c59eba", + "metadata": {}, + "source": [ + "Let's generate some data to see what the simulator does:" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "e6218e61d529e357", + "metadata": { + "ExecuteTime": { + "end_time": "2024-10-24T08:36:22.350483Z", + "start_time": "2024-10-24T08:36:22.345161Z" + } + }, + "outputs": [], + "source": [ + "# generate 3 random draws from the joint distribution p(r, alpha, theta, x)\n", + "sample_data = simulator.sample((20,))" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "id": "46174ccb0167026c", + "metadata": { + "ExecuteTime": { + "end_time": "2024-10-24T08:36:22.470435Z", + "start_time": "2024-10-24T08:36:22.464836Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Type of sample_data:\n", + "\t \n", + "Keys of sample_data:\n", + "\t dict_keys(['r', 'alpha', 'theta', 'x'])\n", + "Types of sample_data values:\n", + "\t {'r': , 'alpha': , 'theta': , 'x': }\n", + "Shapes of sample_data values:\n", + "\t {'r': (20, 1), 'alpha': (20, 1), 'theta': (20, 2), 'x': (20, 2)}\n" + ] + } + ], + "source": [ + "print(\"Type of sample_data:\\n\\t\", type(sample_data))\n", + "print(\"Keys of sample_data:\\n\\t\", sample_data.keys())\n", + "print(\"Types of sample_data values:\\n\\t\", {k: type(v) for k, v in sample_data.items()})\n", + "print(\"Shapes of sample_data values:\\n\\t\", {k: v.shape for k, v in sample_data.items()})" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "id": "5c9c2dc70f53d103", + "metadata": { + "ExecuteTime": { + "end_time": "2024-10-24T08:36:26.618926Z", + "start_time": "2024-10-24T08:36:26.614443Z" + } + }, + "outputs": [ + { + "data": { + "text/plain": [ + "Adapter([0: Keep(['theta', 'x']) -> 1: ToArray -> 2: ConvertDType -> 3: Standardize -> 4: Rename('theta' -> 'inference_variables') -> 5: Rename('x' -> 'inference_conditions')])" + ] + }, + "execution_count": 5, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "adapter = (\n", + " bf.adapters.Adapter()\n", + " # drop data that we do not need\n", + " .keep((\"theta\", \"x\"))\n", + " # convert any non-arrays to numpy arrays\n", + " .to_array()\n", + " # convert from numpy's default float64 to deep learning friendly float32\n", + " .convert_dtype(\"float64\", \"float32\")\n", + " # standardize all variables to zero mean and unit variance\n", + " .standardize(momentum=None)\n", + " # rename the variables to match the required approximator inputs\n", + " .rename(\"theta\", \"inference_variables\").rename(\"x\", \"inference_conditions\")\n", + ")\n", + "adapter" + ] + }, + { + "cell_type": "markdown", + "id": "254e287b2bccdad", + "metadata": {}, + "source": [ + "## Dataset\n", + "\n", + "For this example, we will sample our training data ahead of time and use offline training with a `bf.datasets.OfflineDataset`.\n", + "\n", + "This makes the training process faster, since we avoid repeated sampling. If you want to use online training, you can use an `OnlineDataset` analogously, or just pass your simulator directly to `approximator.fit()`!" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "id": "39cb5a1c9824246f", + "metadata": { + "ExecuteTime": { + "end_time": "2024-09-23T14:39:46.950573Z", + "start_time": "2024-09-23T14:39:46.948624Z" + } + }, + "outputs": [], + "source": [ + "num_training_batches = 512\n", + "num_validation_batches = 128\n", + "batch_size = 64\n", + "epochs = 30\n", + "total_steps = num_training_batches * epochs" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "id": "9dee7252ef99affa", + "metadata": { + "ExecuteTime": { + "end_time": "2024-09-23T14:39:53.268860Z", + "start_time": "2024-09-23T14:39:46.994697Z" + } + }, + "outputs": [], + "source": [ + "training_samples = simulator.sample((num_training_batches * batch_size,))\n", + "validation_samples = simulator.sample((num_validation_batches * batch_size,))" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "id": "51045bbed88cb5c2", + "metadata": { + "ExecuteTime": { + "end_time": "2024-09-23T14:39:53.281170Z", + "start_time": "2024-09-23T14:39:53.275921Z" + } + }, + "outputs": [], + "source": [ + "training_dataset = bf.datasets.OfflineDataset(\n", + " data=training_samples, batch_size=batch_size, adapter=adapter\n", + ")\n", + "\n", + "validation_dataset = bf.datasets.OfflineDataset(\n", + " data=validation_samples, batch_size=batch_size, adapter=adapter\n", + ")" + ] + }, + { + "cell_type": "markdown", + "id": "2d4c6eb0", + "metadata": {}, + "source": [ + "## Training a neural network to approximate all posteriors" + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "id": "09206e6f", + "metadata": { + "ExecuteTime": { + "end_time": "2024-09-23T14:39:53.339590Z", + "start_time": "2024-09-23T14:39:53.319852Z" + } + }, + "outputs": [], + "source": [ + "inference_network = ContinuousTimeConsistencyModel(\n", + " subnet=\"mlp\",\n", + " sigma_data=1.0, # as we have standardized our parameters, the standard deviation is 1.0\n", + " subnet_kwargs={\n", + " \"widths\": (256,) * 6,\n", + " \"dropout\": 0.0,\n", + " }, # use an inner network with 6 hidden layers of 256 units\n", + " embedding_kwargs={\"embed_dim\": 2},\n", + ")" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "id": "8b818296-70f4-45b2-afc2-9a1b6cd96849", + "metadata": {}, + "outputs": [], + "source": [ + "from bayesflow import experimental" + ] + }, + { + "cell_type": "markdown", + "id": "851e522f", + "metadata": {}, + "source": [ + "This inference network is just a general Flow Matching backbone, not yet adapted to the specific inference task at hand (i.e., posterior appproximation). To achieve this adaptation, we combine the network with our data adapter, which together form an `approximator`. In this case, we need a `ContinuousApproximator` since the target we want to approximate is the posterior of the *continuous* parameter vector $\\theta$." + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "id": "96ca6ffa", + "metadata": { + "ExecuteTime": { + "end_time": "2024-09-23T14:39:53.371691Z", + "start_time": "2024-09-23T14:39:53.369375Z" + } + }, + "outputs": [], + "source": [ + "cm_approximator = bf.ContinuousApproximator(\n", + " inference_network=inference_network,\n", + " adapter=adapter,\n", + ")" + ] + }, + { + "cell_type": "markdown", + "id": "566264eadc76c2c", + "metadata": {}, + "source": [ + "### Optimizer and Learning Rate\n", + "We find learning rate schedules, such as [cosine decay](https://keras.io/api/optimizers/learning_rate_schedules/cosine_decay/), work well for a wide variety of approximation tasks." + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "id": "e8d7e053", + "metadata": { + "ExecuteTime": { + "end_time": "2024-09-23T14:39:53.433012Z", + "start_time": "2024-09-23T14:39:53.415903Z" + } + }, + "outputs": [], + "source": [ + "initial_learning_rate = 5e-4\n", + "scheduled_lr = keras.optimizers.schedules.CosineDecay(\n", + " initial_learning_rate=initial_learning_rate, decay_steps=total_steps, alpha=1e-8\n", + ")\n", + "\n", + "optimizer = keras.optimizers.Adam(learning_rate=scheduled_lr)" + ] + }, + { + "cell_type": "code", + "execution_count": 13, + "id": "51808fcd560489ac", + "metadata": { + "ExecuteTime": { + "end_time": "2024-09-23T14:39:53.476089Z", + "start_time": "2024-09-23T14:39:53.466001Z" + } + }, + "outputs": [], + "source": [ + "cm_approximator.compile(optimizer=optimizer)" + ] + }, + { + "cell_type": "markdown", + "id": "708b1303", + "metadata": {}, + "source": [ + "### Training\n", + "\n", + "We are ready to train our deep posterior approximator on the two moons example. We pass the dataset object to the `fit` method and watch as Bayesflow trains." + ] + }, + { + "cell_type": "code", + "execution_count": 14, + "id": "0f496bda", + "metadata": { + "ExecuteTime": { + "end_time": "2024-09-23T14:42:36.067393Z", + "start_time": "2024-09-23T14:39:53.513436Z" + } + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "INFO:bayesflow:Fitting on dataset instance of OfflineDataset.\n", + "INFO:bayesflow:Building on a test batch.\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch 1/30\n", + "512/512 - 4s - 7ms/step - loss: -9.3843e-01 - loss/inference_loss: -9.3843e-01 - val_loss: -9.1318e-01 - val_loss/inference_loss: -9.1318e-01\n", + "Epoch 2/30\n", + "512/512 - 3s - 5ms/step - loss: -9.9757e-01 - loss/inference_loss: -9.9757e-01 - val_loss: -1.0789e+00 - val_loss/inference_loss: -1.0789e+00\n", + "Epoch 3/30\n", + "512/512 - 2s - 5ms/step - loss: -9.7956e-01 - loss/inference_loss: -9.7956e-01 - val_loss: -1.0510e+00 - val_loss/inference_loss: -1.0510e+00\n", + "Epoch 4/30\n", + "512/512 - 3s - 5ms/step - loss: -1.0268e+00 - loss/inference_loss: -1.0268e+00 - val_loss: -1.0437e+00 - val_loss/inference_loss: -1.0437e+00\n", + "Epoch 5/30\n", + "512/512 - 3s - 5ms/step - loss: -1.0449e+00 - loss/inference_loss: -1.0449e+00 - val_loss: -9.4702e-01 - val_loss/inference_loss: -9.4702e-01\n", + "Epoch 6/30\n", + "512/512 - 2s - 5ms/step - loss: -9.5895e-01 - loss/inference_loss: -9.5895e-01 - val_loss: -9.5736e-01 - val_loss/inference_loss: -9.5736e-01\n", + "Epoch 7/30\n", + "512/512 - 2s - 5ms/step - loss: -9.3300e-01 - loss/inference_loss: -9.3300e-01 - val_loss: -1.1226e+00 - val_loss/inference_loss: -1.1226e+00\n", + "Epoch 8/30\n", + "512/512 - 2s - 5ms/step - loss: -9.3088e-01 - loss/inference_loss: -9.3088e-01 - val_loss: -1.1311e+00 - val_loss/inference_loss: -1.1311e+00\n", + "Epoch 9/30\n", + "512/512 - 2s - 5ms/step - loss: -9.3398e-01 - loss/inference_loss: -9.3398e-01 - val_loss: -1.0012e+00 - val_loss/inference_loss: -1.0012e+00\n", + "Epoch 10/30\n", + "512/512 - 2s - 5ms/step - loss: -9.5127e-01 - loss/inference_loss: -9.5127e-01 - val_loss: -1.1021e+00 - val_loss/inference_loss: -1.1021e+00\n", + "Epoch 11/30\n", + "512/512 - 2s - 5ms/step - loss: -1.0212e+00 - loss/inference_loss: -1.0212e+00 - val_loss: -9.9398e-01 - val_loss/inference_loss: -9.9398e-01\n", + "Epoch 12/30\n", + "512/512 - 3s - 5ms/step - loss: -1.1326e+00 - loss/inference_loss: -1.1326e+00 - val_loss: -9.5806e-01 - val_loss/inference_loss: -9.5806e-01\n", + "Epoch 13/30\n", + "512/512 - 3s - 5ms/step - loss: -1.0533e+00 - loss/inference_loss: -1.0533e+00 - val_loss: -9.8955e-01 - val_loss/inference_loss: -9.8955e-01\n", + "Epoch 14/30\n", + "512/512 - 3s - 5ms/step - loss: -1.0783e+00 - loss/inference_loss: -1.0783e+00 - val_loss: -1.1011e+00 - val_loss/inference_loss: -1.1011e+00\n", + "Epoch 15/30\n", + "512/512 - 3s - 5ms/step - loss: -1.0251e+00 - loss/inference_loss: -1.0251e+00 - val_loss: -1.1740e+00 - val_loss/inference_loss: -1.1740e+00\n", + "Epoch 16/30\n", + "512/512 - 3s - 5ms/step - loss: -1.0389e+00 - loss/inference_loss: -1.0389e+00 - val_loss: -1.1353e+00 - val_loss/inference_loss: -1.1353e+00\n", + "Epoch 17/30\n", + "512/512 - 3s - 5ms/step - loss: -1.0906e+00 - loss/inference_loss: -1.0906e+00 - val_loss: -1.1350e+00 - val_loss/inference_loss: -1.1350e+00\n", + "Epoch 18/30\n", + "512/512 - 2s - 5ms/step - loss: -1.1508e+00 - loss/inference_loss: -1.1508e+00 - val_loss: -1.0663e+00 - val_loss/inference_loss: -1.0663e+00\n", + "Epoch 19/30\n", + "512/512 - 2s - 5ms/step - loss: -1.0334e+00 - loss/inference_loss: -1.0334e+00 - val_loss: -1.1079e+00 - val_loss/inference_loss: -1.1079e+00\n", + "Epoch 20/30\n", + "512/512 - 3s - 5ms/step - loss: -1.1104e+00 - loss/inference_loss: -1.1104e+00 - val_loss: -1.1154e+00 - val_loss/inference_loss: -1.1154e+00\n", + "Epoch 21/30\n", + "512/512 - 3s - 5ms/step - loss: -1.0883e+00 - loss/inference_loss: -1.0883e+00 - val_loss: -1.1937e+00 - val_loss/inference_loss: -1.1937e+00\n", + "Epoch 22/30\n", + "512/512 - 2s - 5ms/step - loss: -1.1893e+00 - loss/inference_loss: -1.1893e+00 - val_loss: -1.1405e+00 - val_loss/inference_loss: -1.1405e+00\n", + "Epoch 23/30\n", + "512/512 - 2s - 5ms/step - loss: -1.2122e+00 - loss/inference_loss: -1.2122e+00 - val_loss: -1.1151e+00 - val_loss/inference_loss: -1.1151e+00\n", + "Epoch 24/30\n", + "512/512 - 2s - 5ms/step - loss: -1.0478e+00 - loss/inference_loss: -1.0478e+00 - val_loss: -1.1214e+00 - val_loss/inference_loss: -1.1214e+00\n", + "Epoch 25/30\n", + "512/512 - 2s - 5ms/step - loss: -1.1838e+00 - loss/inference_loss: -1.1838e+00 - val_loss: -1.1838e+00 - val_loss/inference_loss: -1.1838e+00\n", + "Epoch 26/30\n", + "512/512 - 2s - 5ms/step - loss: -1.1238e+00 - loss/inference_loss: -1.1238e+00 - val_loss: -1.1740e+00 - val_loss/inference_loss: -1.1740e+00\n", + "Epoch 27/30\n", + "512/512 - 2s - 5ms/step - loss: -1.1162e+00 - loss/inference_loss: -1.1162e+00 - val_loss: -1.0619e+00 - val_loss/inference_loss: -1.0619e+00\n", + "Epoch 28/30\n", + "512/512 - 2s - 5ms/step - loss: -1.2349e+00 - loss/inference_loss: -1.2349e+00 - val_loss: -1.1863e+00 - val_loss/inference_loss: -1.1863e+00\n", + "Epoch 29/30\n", + "512/512 - 3s - 5ms/step - loss: -1.3770e+00 - loss/inference_loss: -1.3770e+00 - val_loss: -1.2584e+00 - val_loss/inference_loss: -1.2584e+00\n", + "Epoch 30/30\n", + "512/512 - 3s - 5ms/step - loss: -1.1379e+00 - loss/inference_loss: -1.1379e+00 - val_loss: -1.1435e+00 - val_loss/inference_loss: -1.1435e+00\n", + "CPU times: user 2min 54s, sys: 10.6 s, total: 3min 4s\n", + "Wall time: 1min 17s\n" + ] + } + ], + "source": [ + "%%time\n", + "fm_history = cm_approximator.fit(\n", + " epochs=epochs,\n", + " dataset=training_dataset,\n", + " validation_data=validation_dataset,\n", + " verbose=2, # set verbose=2 to avoid flooding the notebook\n", + ")" + ] + }, + { + "cell_type": "markdown", + "id": "f4785f35-794e-40c7-b863-f5100a90ef13", + "metadata": {}, + "source": [ + "Note that after a certain time, the loss is no longer indicative of training performance." + ] + }, + { + "cell_type": "code", + "execution_count": 15, + "id": "209e4bbd-9d4e-4639-82b7-0e974f7258ca", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "# Set the number of posterior draws you want to get\n", + "num_samples = 3000\n", + "steps = 15\n", + "rho = 3.5\n", + "\n", + "# Obtain samples from amortized posterior\n", + "\n", + "conditions = {\"x\": np.array([[0.0, 0.0]]).astype(\"float32\")}\n", + "samples_0 = cm_approximator.sample(\n", + " conditions=conditions, num_samples=num_samples, steps=steps, rho=rho\n", + ")[\"theta\"][0]\n", + "\n", + "# Prepare figure\n", + "f, axes = plt.subplots(1, 2, figsize=(12, 6))\n", + "\n", + "# Plot samples (once without limits to see outliers/problems\n", + "samples = [samples_0, samples_0]\n", + "names = [\"Continuous-time CM\", \"Without Axis Limits\"]\n", + "colors = [\"#153c7a\", \"#7a1515\"]\n", + "\n", + "for ax, thetas, name, color in zip(axes, samples, names, colors):\n", + "\n", + " # Plot samples\n", + " ax.scatter(thetas[:, 0], thetas[:, 1], color=color, alpha=0.75, s=0.5)\n", + " sns.despine(ax=ax)\n", + " ax.set_title(f\"{name}\", fontsize=16)\n", + " ax.grid(alpha=0.3)\n", + " ax.set_aspect(\"equal\", adjustable=\"box\")\n", + " if not name.lower().startswith(\"without\"):\n", + " ax.set_xlim([-0.5, 0.5])\n", + " ax.set_ylim([-0.5, 0.5])\n", + " ax.set_xlabel(r\"$\\theta_1$\", fontsize=15)\n", + " ax.set_ylabel(r\"$\\theta_2$\", fontsize=15)\n", + "\n", + "f.tight_layout()" + ] + }, + { + "cell_type": "markdown", + "id": "1e4f4062-a535-41e5-808b-321346038993", + "metadata": {}, + "source": [ + "Plot the discretization schedule:" + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "id": "e728c992-e91f-4d51-ad6b-33d7f759201b", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "discretized_time = np.flip(\n", + " keras.ops.convert_to_numpy(\n", + " cm_approximator.inference_network._discretize_time(steps, rho=rho)\n", + " )\n", + ")\n", + "plt.plot(discretized_time, marker=\"x\", linestyle=\"dotted\")\n", + "plt.ylabel(\"t\")\n", + "plt.xlabel(\"step\")\n", + "plt.ylim(0.0, np.pi / 2)\n", + "plt.xlim(0, len(discretized_time) - 1)\n", + "plt.yticks(\n", + " [0.0, np.pi / 8, np.pi / 4, 3 * np.pi / 8, np.pi / 2],\n", + " labels=[\"0\", r\"$\\pi/8$\", r\"$\\pi/4$\", r\"$3\\pi/8$\", r\"$\\pi/2$\"],\n", + ")\n", + "plt.grid()\n", + "_ = plt.title(\"Discretization schedule\")" + ] + }, + { + "cell_type": "markdown", + "id": "7f8532f2-bbe1-4690-b74f-285d94960ab5", + "metadata": {}, + "source": [ + "Plot the time embedding:" + ] + }, + { + "cell_type": "code", + "execution_count": 17, + "id": "ffbecec3-b297-48db-b99d-9095b3e9f7a9", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "t = keras.ops.linspace(0.001, np.pi / 2, 500)[:, None]\n", + "emb = inference_network.time_emb(t)\n", + "plt.plot(keras.ops.convert_to_numpy(t)[:, 0], keras.ops.convert_to_numpy(emb))\n", + "plt.ylabel(\"emb(t)\")\n", + "plt.xlabel(\"t\")\n", + "plt.xticks([0.0, np.pi / 4, np.pi / 2], labels=[\"0\", r\"$\\pi/4$\", r\"$\\pi/2$\"])\n", + "_ = plt.title(\"Time embedding\")" + ] + }, + { + "cell_type": "markdown", + "id": "acce03bb-a802-40a9-ab4f-4a2c3cc0f3de", + "metadata": {}, + "source": [ + "Plot the learned adaptive weighting function:" + ] + }, + { + "cell_type": "code", + "execution_count": 18, + "id": "e233f634-3f34-4f0e-baa8-6204f41de971", + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "plt.plot(t, inference_network.weight_fn_projector(inference_network.weight_fn(t)))\n", + "plt.plot(t, 1 / (inference_network.sigma_data * np.tan(t)))\n", + "plt.plot(\n", + " t,\n", + " inference_network.weight_fn_projector(inference_network.weight_fn(t))\n", + " / (inference_network.sigma_data * np.tan(t)),\n", + ")\n", + "plt.ylabel(r\"$w_\\phi(t)$\")\n", + "plt.xlabel(\"t\")\n", + "plt.yscale(\"log\")\n", + "plt.xticks([0.0, np.pi / 4, np.pi / 2], labels=[\"0\", r\"$\\pi/4$\", r\"$\\pi/2$\"])\n", + "_ = plt.title(\"Learned adaptive weighting function\")" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3 (ipykernel)", + "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": "3.11.11" + }, + "toc": { + "base_numbering": 1, + "nav_menu": {}, + "number_sections": true, + "sideBar": true, + "skip_h1_title": true, + "title_cell": "Table of Contents", + "title_sidebar": "Contents", + "toc_cell": true, + "toc_position": { + "height": "calc(100% - 180px)", + "left": "10px", + "top": "150px", + "width": "165px" + }, + "toc_section_display": true, + "toc_window_display": true + } + }, + "nbformat": 4, + "nbformat_minor": 5 +}