DataPredict [Release 1.21] - General Purpose Machine Learning And Deep Learning Library (Learning AIs, Generative AIs, and more!)

MYOriginsWorkshop · June 18, 2023, 9:09am

Update 1.6.0: Recurrent Neural Network added. Go ahead and try it out!

MYOriginsWorkshop · June 22, 2023, 8:32am

Update 1.8.0:

LSTM model added!
Adjustable input, hidden, output layers for both RNN and LSTM
Made some changes to OnlineLearning so that it accepts both sequential and non-sequential inputs/outputs.

Note: You many need to update your Matrix Library.

CCTVStudios · June 22, 2023, 3:17pm

Woah woah, you are telling me with this I can make those AI’s that pathfind and become better each time they train?

If this is stable, performant and good might actually use it to make a pathfinding NPC, Ive heard you can use raycasts and shapecasts to create a view for it, and the AI can learn every simulation more.

MYOriginsWorkshop · June 22, 2023, 6:40pm

Absolutely! What’s even better about this library, you can:

Use Online Learning to do real-time training if you were to collect data real-time and train the model when you are not around.
Retrieve and store the model parameters values into datastores.

It’s about to reach the final development stage. I’m thinking whether or not to add feature that enables the models avoid exhausting script execution time. Not only that, I’m trying to add some more kernels to support vector machine right now.

CCTVStudios · June 22, 2023, 7:05pm

I have no idea about machine learning AI, so if I decide to dive into it ill deffinetely use this, it just sounds too hard for me atm. but good job on this

MYOriginsWorkshop · June 24, 2023, 11:06am

Update 1.9.0:

Added new kernels for support vector machine.
Added internal model wait component to avoid exhaustion time.
RNN and LSTM can take in multiple token sequences.
Changed the parameters for online learning.
Some fixes.

MYOriginsWorkshop · July 4, 2023, 5:48am

Update 1.10.0:

You can now change any of neural network’s layers’s properties using setLayer() function.
Some calculation fixes. This leads to training improvement to certain models.
And some more other stuff.

PS: I recommend you to update both the Machine Learning Library and Matrix Library.

MYOriginsWorkshop · July 13, 2023, 9:20am

Release Version 1.0 (Beta Version 1.11.0) is here! Get the ModuleScript from the original post now!

Fixed a couple of stuff.
Added new optimizers.
Many more!

UFSKFS · July 13, 2023, 4:15pm

Greetings. I read the full version of the documentation, but I couldn’t find anywhere to save the model.

MYOriginsWorkshop · July 13, 2023, 4:19pm

Every model is inherited from a BaseModel class. The BaseModel class contains functions getModelParameters() and setModelParameters(). You can call those functions from any models that inherits it.

Then you can save the model parameters to Roblox’s DataStore or print out the values.

UFSKFS · July 13, 2023, 4:22pm

Oh, I understand you, thank you very much!

MYOriginsWorkshop · July 31, 2023, 11:39pm

Alright guys, it seems like I will not updating this for a long time. Let m know if there are any issues with the codes or any feature requests you want me to add to this library.

Akanizs · August 1, 2023, 12:14am

I’m really interested in this and it seems incredible, i would only like to see some tutorial of this being applied in a NPC

MYOriginsWorkshop · August 1, 2023, 12:18am

I’ll probably make a tutorial video next month once I am done with my work contract.

You can give me some suggestions and I’ll see if I can fit any of the models for those suggestions.

LambHubGoBrrrrr · August 1, 2023, 9:11am

Suggestion : make a pathfinding ai in the tutorial

Akanizs · August 2, 2023, 7:06pm

Alright! Enemy NPCs(Follow and Attack the player) but for that i would like to see if these kind of Enemy NPCs could be trained to dodge projectiles, block attacks, maybe even using objects around him to his advantage like to avoid being hit by a projectile, hope this made sense!

Cffex · August 5, 2023, 7:41am

Hello, I am grateful for this library, however, I am still confused on how to intergrate it into gameplay. Suppose I have an NPC, I want it to train with players, so I have this pseudo-code:

local NPC = workspace.NPC

local function moveRight()
end

local function moveLeft()
end

local function moveFoward()
end

local function moveBackward()
end

local function moveJump()
end

local function attack()
end

local function getInputFromEyeWithRays()
end

How would I implement your module with this pseudo-code?

MYOriginsWorkshop · August 5, 2023, 1:29pm

Sure! Something like this with an additional functions added. Unfortunately, I am not good at pseudo code, so I’ll give the partially-completed Lua code instead (not tested).

NPC trained using player data and online learning.

local NPC = workspace.NPC

local NeuralNetwork = DataPredict.Models.NeuralNetwork

local OnlineLearning = DataPredict.Others.OnlineLearning

local function moveRight()
end

local function moveLeft()
end

local function moveFoward()
end

local function moveBackward()
end

local function moveJump()
end

local function attack()
end

local function checkIfIsCharacter(HittedPart)

	local ParentPart = HittedPart.ParentPart
	
	local Humanoid = ParentPart:FindFirstChild("Humanoid")
	
	if Humanoid then
	
		return 1
	
	else
	
		return 0
	
	end

end

local function performAction(actionNumber)

	if (actionNumber == 1) then
	
		attack()
		
	if (actionNumber == 2) then
	
		jump()
		
	else -- You can add more actions here.
		
		
		
	end	

end

local function getInputFromEyeWithRays()

	local inputVector
	
	local materialEnumValue
	
	local distance
	
	local isPlayer
	
	local ForwardRaycast = workspace:Raycast()
	
	if ForwardRaycast then
	
		materialEnumValue = ForwardRaycast.Material
		
		distance = ForwardRaycast.Distance
		
		isCharacter = checkIfIsCharacter(ForwardRaycast.Instance)
	
	else
		
		materialEnumValue = 0
		
		distance = inf
		
		isCharacter = 0
	
	end
	
	inputVector = {{1, materialEnumValue, distance, isCharacter}} -- 1 is added for bias 
	
	return inputVector

end

local function convertControlToActionNumber(ReceivedControls)

	-- Assign each integer number to each control here
	
	local actionNumber = 0
	
	if (ReceivedControls == "Control1") then
	
		actionNumber = 1
		
	end	
	
	return actionNumber

end

local function startTrainingFromPlayer(Player)

	local inputVector
	
	local NeuralNetworkForThisPlayer = NeuralNetwork.new()
	
	NeuralNetworkForThisPlayer:addLayer(3, true) -- First input layer. Add bias too.
	
	NeuralNetworkForThisPlayer:addLayer(5, true) -- Second input layer. Add bias too.
	
	NeuralNetworkForThisPlayer:addLayer(6, false) -- Final output layer. Value is six because of 6 actions.
	
	NeuralNetworkForThisPlayer:setClassesList({1, 2, 3, 4, 5, 6}) -- 6 different actions, so six classes.
	
	local OnlineLearningForThisPlayer = OnlineLearning.new(NeuralNetworkForThisPlayer)
	
	local DataHarvestRemoteEvent
	
	DataHarvestRemoteEvent.OnServerEvent:Connect(function(ReceivedPlayer, ReceivedControls)
	
		if (ReceivedPlayer == Player) then
		
			inputVector = getInputFromEyeWithRays()
		
			actionNumber = convertControlToActionNumber(ReceivedControls)
		
			OnlineLearningForThisPlayer:addInputToOnlineLearningQueue(inputVector)
		
			OnlineLearningForThisPlayer:addOutputToOnlineLearningQueue(actionNumber)
		
		end
	
	end)
	
	OnlineLearningForThisPlayer:startOnlineLearning()

end

local function runTrainedNPC()
	
	local inputVector
	
	local predictedActionNumber
	
	local NeuralNetworkForThisNPC = NeuralNetwork.new()
	
	-- The layers are the same to our previous neural network
	
	NeuralNetworkForThisNPC:addLayer(3, true) -- First input layer. Add bias too.
	
	NeuralNetworkForThisNPC:addLayer(5, true) -- Second input layer. Add bias too.
	
	NeuralNetworkForThisNPC:addLayer(6, false) -- Final output layer. Value is six because of 6 actions.
	
	NeuralNetworkForThisNPC:setClassesList({1, 2, 3, 4, 5, 6}) -- 6 different actions, so six classes.
	
	while true do
	
		inputVector = getInputFromEyeWithRays()
	
		predictedActionNumber = NeuralNetworkForThisNPC:predict(inputVector)
		
		performAction(actionNumber)
	
	end

end

NPC trained using reinforcement (accidentally misread the question, but I’ll just leave it here).

local NPC = workspace.NPC

local NeuralNetwork = DataPredict.Models.NeuralNetwork.new()

NeuralNetwork:addLayer(3, true) -- First input layer. Add bias too.
	
NeuralNetwork:addLayer(5, true) -- Second input layer. Add bias too.
	
NeuralNetwork:addLayer(6, false) -- Final output layer. Value is six because of 6 actions.

NeuralNetwork:setClassesList({1, 2, 3, 4, 5, 6}) -- 6 different actions, so six classes.

local function moveRight()
end

local function moveLeft()
end

local function moveFoward()
end

local function moveBackward()
end

local function moveJump()
end

local function attack()
end

local function checkIfIsCharacter(HittedPart)

	local ParentPart = HittedPart.ParentPart
	
	local Humanoid = ParentPart:FindFirstChild("Humanoid")
	
	if Humanoid then
	
		return 1
	
	else
	
		return 0
	
	end

end

local function getInputFromEyeWithRays()

	local inputVector
	
	local materialEnumValue
	
	local distance
	
	local isPlayer
	
	local ForwardRaycast = workspace:Raycast()
	
	if ForwardRaycast then
	
		materialEnumValue = ForwardRaycast.Material
		
		distance = ForwardRaycast.Distance
		
		isCharacter = checkIfIsCharacter(ForwardRaycast.Instance)
	
	else
		
		materialEnumValue = 0
		
		distance = inf
		
		isCharacter = 0
	
	end
	
	inputVector = {{1, materialEnumValue, distance, isCharacter}} -- 1 is added for bias 
	
	return inputVector

end

local function performAction(actionNumber)

	if (actionNumber == 1) then
	
		attack()
		
	if (actionNumber == 2) then
	
		jump()
		
	else -- You can add more actions here.
		
		
		
	end	

end

local function evaluateRealActionNumber(predictedActionNumber)

	local realActionNumber = 0

	--[[ 
	
		Put your own conditions if certain conditions are met.
		For example, if the NPC takes a hit while fighting, then move back.
		Another one is that if the NPC doesnt move forward because of something blocking it, then it needs to jump.
	
	--]]
	
	return realActionNumber
	
end

local function run()

	local realActionNumber = 0
	
	local inputVector
	
	local predictedActionNumber
	
	local rewardValue = 0.3
	
	local punishValue = 0.5
	
	while true do
	
		inputVector = getInputFromEyeWithRays()
	
		predictedActionNumber = NeuralNetwork:reinforce(inputVector, realActionNumber, rewardValue, punishValue)
		
		performAction(actionNumber)
		
		realActionNumber = evaluateRealActionNumber()
	
	end

end

Alternatively, you can use the QueuedReinforcementNeuralNetwork in AqwamCustomModels. However, you need the while loop.

qurxts · August 6, 2023, 5:58am

Look’s very interesting i’m gonna check it out.

MYOriginsWorkshop · August 6, 2023, 6:34am

Don’t forget to leave a like on the first post if you find this useful! I’d appreciate it.