mirror of
https://github.com/vale981/ray
synced 2025-03-06 18:41:40 -05:00
9e0a59d94a
Co-authored-by: brettskymind <brett@pathmind.com> Co-authored-by: Max Pumperla <max.pumperla@googlemail.com>
412 lines
11 KiB
Text
412 lines
11 KiB
Text
{
|
||
"cells": [
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "58fc50bc",
|
||
"metadata": {},
|
||
"source": [
|
||
"# Running Tune experiments with HyperOpt\n",
|
||
"\n",
|
||
"In this tutorial we introduce HyperOpt, while running a simple Ray Tune experiment. Tune’s Search Algorithms integrate with HyperOpt and, as a result, allow you to seamlessly scale up a Hyperopt optimization process - without sacrificing performance.\n",
|
||
"\n",
|
||
"HyperOpt provides gradient/derivative-free optimization able to handle noise over the objective landscape, including evolutionary, bandit, and Bayesian optimization algorithms. Nevergrad internally supports search spaces which are continuous, discrete or a mixture of thereof. It also provides a library of functions on which to test the optimization algorithms and compare with other benchmarks.\n",
|
||
"\n",
|
||
"In this example we minimize a simple objective to briefly demonstrate the usage of HyperOpt with Ray Tune via `HyperOptSearch`. It's useful to keep in mind that despite the emphasis on machine learning experiments, Ray Tune optimizes any implicit or explicit objective. Here we assume `hyperopt==0.2.5` library is installed. To learn more, please refer to [HyperOpt website](http://hyperopt.github.io/hyperopt).\n",
|
||
"\n",
|
||
"We include a important example on conditional search spaces (stringing together relationships among hyperparameters)."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "e4586d28",
|
||
"metadata": {},
|
||
"source": [
|
||
"Background information:\n",
|
||
"- [HyperOpt website](http://hyperopt.github.io/hyperopt)\n",
|
||
"\n",
|
||
"Necessary requirements:\n",
|
||
"- `pip install ray[tune]`\n",
|
||
"- `pip install hyperopt==0.2.5`"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "6567f2dc",
|
||
"metadata": {
|
||
"tags": [
|
||
"remove-cell"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"# !pip install ray[tune]\n",
|
||
"!pip install hyperopt==0.2.5"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "b8e9e0cd",
|
||
"metadata": {},
|
||
"source": [
|
||
"Click below to see all the imports we need for this example.\n",
|
||
"You can also launch directly into a Binder instance to run this notebook yourself.\n",
|
||
"Just click on the rocket symbol at the top of the navigation."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "6592315e",
|
||
"metadata": {
|
||
"tags": [
|
||
"hide-input"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"import time\n",
|
||
"\n",
|
||
"import ray\n",
|
||
"from ray import tune\n",
|
||
"from ray.tune.suggest import ConcurrencyLimiter\n",
|
||
"from ray.tune.suggest.hyperopt import HyperOptSearch\n",
|
||
"from hyperopt import hp"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "d4b6d1d5",
|
||
"metadata": {},
|
||
"source": [
|
||
"Let's start by defining a simple evaluation function.\n",
|
||
"We artificially sleep for a bit (`0.1` seconds) to simulate a long-running ML experiment.\n",
|
||
"This setup assumes that we're running multiple `step`s of an experiment and try to tune two hyperparameters,\n",
|
||
"namely `width` and `height`."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "12d4efc8",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"def evaluate(step, width, height):\n",
|
||
" time.sleep(0.1)\n",
|
||
" return (0.1 + width * step / 100) ** (-1) + height * 0.1"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "4f4f5aa2",
|
||
"metadata": {},
|
||
"source": [
|
||
"Next, our ``objective`` function takes a Tune ``config``, evaluates the `score` of your experiment in a training loop,\n",
|
||
"and uses `tune.report` to report the `score` back to Tune."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "c9818009",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"def objective(config):\n",
|
||
" for step in range(config[\"steps\"]):\n",
|
||
" score = evaluate(step, config[\"width\"], config[\"height\"])\n",
|
||
" tune.report(iterations=step, mean_loss=score)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "33eddcb9",
|
||
"metadata": {
|
||
"tags": [
|
||
"remove-cell"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"ray.init(configure_logging=False)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "5be35d5e",
|
||
"metadata": {},
|
||
"source": [
|
||
"While defining the search algorithm, we may choose to provide an initial set of hyperparameters that we believe are especially promising or informative, and\n",
|
||
"pass this information as a helpful starting point for the `HyperOptSearch` object.\n",
|
||
"\n",
|
||
"We also set the maximum concurrent trials to `4` with a `ConcurrencyLimiter`."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "d4615bed",
|
||
"metadata": {
|
||
"lines_to_next_cell": 0
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"initial_params = [\n",
|
||
" {\"width\": 1, \"height\": 2},\n",
|
||
" {\"width\": 4, \"height\": 2},\n",
|
||
"]\n",
|
||
"algo = HyperOptSearch(points_to_evaluate=initial_params)\n",
|
||
"algo = ConcurrencyLimiter(algo, max_concurrent=4)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "2a51e7c1",
|
||
"metadata": {},
|
||
"source": [
|
||
"The number of samples is the number of hyperparameter combinations that will be tried out. This Tune run is set to `1000` samples.\n",
|
||
"(you can decrease this if it takes too long on your machine)."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "2dbb2be0",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"num_samples = 1000"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "950558ed",
|
||
"metadata": {
|
||
"tags": [
|
||
"remove-cell"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"# If 1000 samples take too long, you can reduce this number.\n",
|
||
"# We override this number here for our smoke tests.\n",
|
||
"num_samples = 10"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "6e3629cb",
|
||
"metadata": {},
|
||
"source": [
|
||
"Next we define a search space. The critical assumption is that the optimal hyperparamters live within this space. Yet, if the space is very large, then those hyperparameters may be difficult to find in a short amount of time."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "65189946",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"search_config = {\n",
|
||
" \"steps\": 100,\n",
|
||
" \"width\": tune.uniform(0, 20),\n",
|
||
" \"height\": tune.uniform(-100, 100),\n",
|
||
" \"activation\": tune.choice([\"relu, tanh\"])\n",
|
||
"}"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "1b94c93b",
|
||
"metadata": {},
|
||
"source": [
|
||
"Finally, we run the experiment to `\"min\"`imize the \"mean_loss\" of the `objective` by searching `search_config` via `algo`, `num_samples` times. This previous sentence is fully characterizes the search problem we aim to solve. With this in mind, notice how efficient it is to execute `tune.run()`."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "9a99a3a7",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"analysis = tune.run(\n",
|
||
" objective,\n",
|
||
" search_alg=algo,\n",
|
||
" metric=\"mean_loss\",\n",
|
||
" mode=\"min\",\n",
|
||
" name=\"hyperopt_exp\",\n",
|
||
" num_samples=num_samples,\n",
|
||
" config=search_space,\n",
|
||
")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "49be6f01",
|
||
"metadata": {},
|
||
"source": [
|
||
"Here are the hyperparamters found to minimize the mean loss of the defined objective."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "7036798c",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"print(\"Best hyperparameters found were: \", analysis.best_config)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "504e9d2a",
|
||
"metadata": {},
|
||
"source": [
|
||
"## Conditional search spaces\n",
|
||
"\n",
|
||
"Sometimes we may want to build a more complicated search space that has conditional dependencies on other hyperparameters. In this case, we pass a nested dictionary to `objective_two`, which has been slightly adjusted from `objective` to deal with the conditional search space."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "2f7b5449",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"def evaluation_fn(step, width, height, mult=1):\n",
|
||
" return (0.1 + width * step / 100) ** (-1) + height * 0.1 * mult"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "4b83b81c",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"def objective_two(config):\n",
|
||
" width, height = config[\"width\"], config[\"height\"]\n",
|
||
" sub_dict = config[\"activation\"]\n",
|
||
" mult = sub_dict.get(\"mult\", 1)\n",
|
||
" \n",
|
||
" for step in range(config[\"steps\"]):\n",
|
||
" intermediate_score = evaluation_fn(step, width, height, mult)\n",
|
||
" tune.report(iterations=step, mean_loss=intermediate_score)\n",
|
||
" time.sleep(0.1)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "75cea99e",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"conditional_space = {\n",
|
||
" \"activation\": hp.choice(\n",
|
||
" \"activation\",\n",
|
||
" [\n",
|
||
" {\"activation\": \"relu\", \"mult\": hp.uniform(\"mult\", 1, 2)},\n",
|
||
" {\"activation\": \"tanh\"},\n",
|
||
" ],\n",
|
||
" ),\n",
|
||
" \"width\": hp.uniform(\"width\", 0, 20),\n",
|
||
" \"height\": hp.uniform(\"height\", -100, 100),\n",
|
||
" \"steps\": 100,\n",
|
||
"}"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "7df282c1",
|
||
"metadata": {},
|
||
"source": [
|
||
"Now we the define the search algorithm built from `HyperOptSearch` constrained by `ConcurrencyLimiter`. When the hyperparameter search space is conditional, we pass it (`conditional_space`) into `HyperOptSearch`."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "ea2c71a6",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"algo = HyperOptSearch(space=conditional_space, metric=\"mean_loss\", mode=\"min\")\n",
|
||
"algo = ConcurrencyLimiter(algo, max_concurrent=4)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "630f84ab",
|
||
"metadata": {},
|
||
"source": [
|
||
"Now we run the experiment, this time with an empty `config` because we instead provided `space` to the `HyperOptSearch` `search_alg`."
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "14111e9e",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"analysis = tune.run(\n",
|
||
" objective_two,\n",
|
||
" metric=\"mean_loss\",\n",
|
||
" mode=\"min\",\n",
|
||
" search_alg=algo,\n",
|
||
" num_samples=num_samples\n",
|
||
")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "e6172afa",
|
||
"metadata": {},
|
||
"source": [
|
||
"Finally, we again show the hyperparameters that minimize the mean loss defined by the score of the objective function above. "
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "03c3fc49",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"print(\"Best hyperparameters found were: \", analysis.best_config)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": null,
|
||
"id": "2f7b72d3",
|
||
"metadata": {
|
||
"tags": [
|
||
"remove-cell"
|
||
]
|
||
},
|
||
"outputs": [],
|
||
"source": [
|
||
"ray.shutdown()"
|
||
]
|
||
}
|
||
],
|
||
"metadata": {
|
||
"kernelspec": {
|
||
"display_name": "Python 3 (ipykernel)",
|
||
"language": "python",
|
||
"name": "python3"
|
||
},
|
||
"orphan": true
|
||
},
|
||
"nbformat": 4,
|
||
"nbformat_minor": 5
|
||
}
|