Training Parameters

General

AlphaZero.ParamsType
Params

The AlphaZero training hyperparameters.

ParameterTypeDefault
self_playSelfPlayParams-
learningLearningParams-
arena[Union{Nothing, ArenaParams}]-
memory_analysisUnion{Nothing, MemAnalysisParams}nothing
num_itersInt-
use_symmetriesBoolfalse
ternary_rewardsBoolfalse
mem_buffer_sizePLSchedule{Int}-

Explanation

The AlphaZero training process consists in num_iters iterations. Each iteration can be decomposed into a self-play phase (see SelfPlayParams) and a learning phase (see LearningParams).

  • ternary_rewards: set to true if the rewards issued by the game environment always belong to $\{-1, 0, 1\}$ so that the logging and profiling tools can take advantage of this property.
  • use_symmetries: if set to true, board symmetries are used for data augmentation before learning.
  • mem_buffer_size: size schedule of the memory buffer, in terms of number of samples. It is typical to start with a small memory buffer that is grown progressively so as to wash out the initial low-quality self-play data more quickly.
  • memory_analysis: parameters for the memory analysis step that is performed at each iteration (see MemAnalysisParams), or nothing if no analysis is to be performed.

AlphaGo Zero Parameters

In the original AlphaGo Zero paper:

  • About 5 millions games of self-play are played across 200 iterations.
  • The memory buffer contains 500K games, which makes about 100M samples as an average game of Go lasts about 200 turns.
source

Self-Play

AlphaZero.SelfPlayParamsType
SelfPlayParams

Parameters governing self-play.

ParameterTypeDefault
mctsMctsParams-
num_gamesInt-
num_workersInt-
use_gpuBoolfalse
reset_mcts_everyUnion{Int, Nothing}1

Explanation

  • The MCTS tree is reset every reset_mcts_every games (or never if nothing is passed).

AlphaGo Zero Parameters

In the original AlphaGo Zero paper, num_games=25_000 (5 millions games of self-play across 200 iterations).

source

Learning

AlphaZero.LearningParamsType
LearningParams

Parameters governing the learning phase of a training iteration, where the neural network is updated to fit the data in the memory buffer.

ParameterTypeDefault
use_gpuBoolfalse
use_position_averagingBooltrue
samples_weighing_policySamplesWeighingPolicy-
optimiserOptimiserSpec-
l2_regularizationFloat32-
nonvalidity_penaltyFloat321f0
batch_sizeInt-
loss_computation_batch_sizeInt-
min_checkpoints_per_epochFloat64-
max_batches_per_checkpointInt-
num_checkpointsInt-

Description

The neural network goes through num_checkpoints series of n updates using batches of size batch_size drawn from memory, where n is defined as follows:

n = min(max_batches_per_checkpoint, ntotal ÷ min_checkpoints_per_epoch)

with ntotal the total number of batches in memory. Between each series, the current network is evaluated against the best network so far (see ArenaParams).

  • nonvalidity_penalty is the multiplicative constant of a loss term that corresponds to the average probability weight that the network puts on invalid actions.
  • batch_size is the batch size used for gradient descent.
  • loss_computation_batch_size is the batch size that is used to compute the loss between each epochs.
  • If use_position_averaging is set to true, samples in memory that correspond to the same board position are averaged together. The merged sample is reweighted according to samples_weighing_policy.

AlphaGo Zero Parameters

In the original AlphaGo Zero paper:

  • The batch size for gradient updates is $2048$.
  • The L2 regularization parameter is set to $10^{-4}$.
  • Checkpoints are produced every 1000 training steps, which corresponds to seeing about 20% of the samples in the memory buffer: $(1000 × 2048) / 10^7 ≈ 0.2$.
  • It is unclear how many checkpoints are taken or how many training steps are performed in total.
source
AlphaZero.SamplesWeighingPolicyType
SamplesWeighingPolicy

During self-play, early board positions are possibly encountered many times across several games. The corresponding samples can be merged together and given a weight $W$ that is a nondecreasing function of the number $n$ of merged samples:

  • CONSTANT_WEIGHT: $W(n) = 1$
  • LOG_WEIGHT: $W(n) = \log_2(n) + 1$
  • LINEAR_WEIGHT: $W(n) = n$
source

Arena

AlphaZero.ArenaParamsType
ArenaParams

Parameters that govern the evaluation process that compares the current neural network with the best one seen so far (which is used to generate data).

ParameterTypeDefault
mctsMctsParams-
num_gamesInt-
num_workersInt-
flip_probabilityFloat640.
reset_mcts_everyUnion{Nothing, Int}1
update_thresholdFloat64-

Explanation

  • The two competing networks are instantiated into two MCTS players of parameter mcts and then play num_games games, switching color after each game.
  • The evaluated network is to replace the current best if its average collected reward is greater or equal than update_threshold.
  • The MCTS trees of both players are reset every reset_mcts_every games (or never if nothing is passed).
  • To add randomization and before every game turn, the game board is "flipped" according to a symmetric transformation with probability flip_probability.

Remarks

AlphaGo Zero Parameters

In the original AlphaGo Zero paper, 400 games are played to evaluate a network and the update_threshold parameter is set to a value that corresponds to a 55% win rate.

source

Memory Analysis

AlphaZero.MemAnalysisParamsType
MemAnalysisParams

Parameters governing the analysis of the memory buffer (for debugging and profiling purposes).

ParameterTypeDefault
num_game_stagesInt-

Explanation

The memory analysis consists in partitioning the memory buffer in num_game_stages parts of equal size, according to the number of remaining moves until the end of the game for each sample. Then, the quality of the predictions of the current neural network is evaluated on each subset (see Report.Memory).

This is useful to get an idea of how the neural network performance varies depending on the game stage (typically, good value estimates for endgame board positions are available earlier in the training process than good values for middlegame positions).

source

MCTS

AlphaZero.MctsParamsType

Parameters of an MCTS player.

ParameterTypeDefault
num_iters_per_turnInt-
gammaFloat641.
cpuctFloat641.
temperatureAbstractSchedule{Float64}ConstSchedule(1.)
dirichlet_noise_ϵFloat64-
dirichlet_noise_αFloat64-
prior_temperatureFloat641.

Explanation

An MCTS player picks an action as follows. Given a game state, it launches num_iters_per_turn MCTS iterations, with UCT exploration constant cpuct. Rewards are discounted using the gamma factor.

Then, an action is picked according to the distribution $π$ where $π_i ∝ n_i^τ$ with $n_i$ the number of times that the $i^{\text{th}}$ action was visited and $τ$ the temperature parameter.

It is typical to use a high value of the temperature parameter $τ$ during the first moves of a game to increase exploration and then switch to a small value. Therefore, temperature is am AbstractSchedule.

For information on parameters cpuct, dirichlet_noise_ϵ, dirichlet_noise_α and prior_temperature, see MCTS.Env.

AlphaGo Zero Parameters

In the original AlphaGo Zero paper:

  • The discount factor gamma is set to 1.
  • The number of MCTS iterations per move is 1600, which corresponds to 0.4s of computation time.
  • The temperature is set to 1 for the 30 first moves and then to an infinitesimal value.
  • The $ϵ$ parameter for the Dirichlet noise is set to $0.25$ and the $α$ parameter to $0.03$, which is consistent with the heuristic of using $α = 10/n$ with $n$ the maximum number of possibles moves, which is $19 × 19 + 1 = 362$ in the case of Go.
source

Utilities

AlphaZero.necessary_samplesFunction
necessary_samples(ϵ, β) = log(1 / β) / (2 * ϵ^2)

Compute the number of times $N$ that a random variable $X \sim \text{Ber}(p)$ has to be sampled so that if the empirical average of $X$ is greather than $1/2 + ϵ$, then $p > 1/2$ with probability at least $1-β$.

This bound is based on Hoeffding's inequality .

source
AlphaZero.AbstractScheduleType
AbstractSchedule{R}

Abstract type for a parameter schedule, which represents a function from nonnegative integers to numbers of type R. Subtypes must implement the getindex(s::AbstractSchedule, i::Int) operator.

source
AlphaZero.StepScheduleType
StepSchedule{R} <: AbstractSchedule{R}

Type for step function schedules.

Constructor

StepSchedule(;start, change_at, values)

Return a schedule that has initial value start. For all i, the schedule takes value values[i] at step change_at[i].

source
AlphaZero.PLScheduleType
PLSchedule{R} <: AbstractSchedule{R}

Type for piecewise linear schedules.

Constructors

PLSchedule(cst)

Return a schedule with a constant value cst.

PLSchedule(xs, ys)

Return a piecewise linear schedule such that:

  • For all i, (xs[i], ys[i]) belongs to the schedule's graph.
  • Before xs[1], the schedule has value ys[1].
  • After xs[end], the schedule has value ys[end].
source