tony/tensorneat-mend
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

remove create_func....

Browse Source
This commit is contained in:
wls2002
2023-08-02 13:26:01 +08:00
parent 85318f98f3
commit 1499e062fe
34 changed files with 558 additions and 1022 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
examples/test.py Normal file
View File

@@ -0,0 +1,24 @@
from functools import partial
import jax
class A:
def __init__(self):
self.a = 1
self.b = 2
self.isTrue = False
@partial(jax.jit, static_argnums=(0,))
def step(self):
if self.isTrue:
return self.a + 1
else:
return self.b + 1
AA = A()
print(AA.step(), hash(AA))
print(AA.step(), hash(AA))
print(AA.step(), hash(AA))
AA.a = (2, 3, 4)
print(AA.step(), hash(AA))

11
examples/xor.py
View File

@@ -3,7 +3,8 @@ import numpy as np
from config import Config, BasicConfig, NeatConfig
from pipeline import Pipeline
from algorithm import NEAT, NormalGene, NormalGeneConfig
from algorithm import NEAT
from algorithm.neat.gene import NormalGene, NormalGeneConfig
xor_inputs = np.array([[0, 0], [0, 1], [1, 0], [1, 1]], dtype=np.float32)
xor_outputs = np.array([[0], [1], [1], [0]], dtype=np.float32)
@@ -23,15 +24,15 @@ def evaluate(forward_func):
if __name__ == '__main__':
config = Config(
basic=BasicConfig(
fitness_target=3.99999,
fitness_target=3.9999999,
pop_size=10000
),
neat=NeatConfig(
maximum_nodes=20,
maximum_conns=50,
),
gene=NormalGeneConfig()
)
)
algorithm = NEAT(config, NormalGene)
normal_gene = NormalGene(NormalGeneConfig())
algorithm = NEAT(config, normal_gene)
pipeline = Pipeline(config, algorithm)
pipeline.auto_run(evaluate)

49
examples/xor_hyperNEAT.py
View File

@@ -1,49 +0,0 @@
import jax
import numpy as np
from config import Config, BasicConfig, NeatConfig
from pipeline import Pipeline
from algorithm import NEAT, RecurrentGene, RecurrentGeneConfig
from algorithm import HyperNEAT, NormalSubstrate, NormalSubstrateConfig
xor_inputs = np.array([[0, 0], [0, 1], [1, 0], [1, 1]], dtype=np.float32)
xor_outputs = np.array([[0], [1], [1], [0]], dtype=np.float32)
def evaluate(forward_func):
"""
:param forward_func: (4: batch, 2: input size) -> (pop_size, 4: batch, 1: output size)
:return:
"""
outs = forward_func(xor_inputs)
outs = jax.device_get(outs)
fitnesses = 4 - np.sum((outs - xor_outputs) ** 2, axis=(1, 2))
return fitnesses
if __name__ == '__main__':
config = Config(
basic=BasicConfig(
fitness_target=3.99999,
pop_size=100
),
neat=NeatConfig(
network_type="recurrent",
maximum_nodes=50,
maximum_conns=100,
inputs=4,
outputs=1
),
gene=RecurrentGeneConfig(
activation_default="tanh",
activation_options=("tanh", ),
),
substrate=NormalSubstrateConfig(),
)
neat = NEAT(config, RecurrentGene)
hyperNEAT = HyperNEAT(config, neat, NormalSubstrate)
pipeline = Pipeline(config, hyperNEAT)
pipeline.auto_run(evaluate)

39
examples/xor_recurrent.py
View File

@@ -1,39 +0,0 @@
import jax
import numpy as np
from config import Config, BasicConfig, NeatConfig
from pipeline import Pipeline
from algorithm import NEAT, RecurrentGene, RecurrentGeneConfig
xor_inputs = np.array([[0, 0], [0, 1], [1, 0], [1, 1]], dtype=np.float32)
xor_outputs = np.array([[0], [1], [1], [0]], dtype=np.float32)
def evaluate(forward_func):
"""
:param forward_func: (4: batch, 2: input size) -> (pop_size, 4: batch, 1: output size)
:return:
"""
outs = forward_func(xor_inputs)
outs = jax.device_get(outs)
fitnesses = 4 - np.sum((outs - xor_outputs) ** 2, axis=(1, 2))
return fitnesses
if __name__ == '__main__':
config = Config(
basic=BasicConfig(
fitness_target=3.99999,
pop_size=10000
),
neat=NeatConfig(
network_type="recurrent",
maximum_nodes=50,
maximum_conns=100
),
gene=RecurrentGeneConfig()
)
algorithm = NEAT(config, RecurrentGene)
pipeline = Pipeline(config, algorithm)
pipeline.auto_run(evaluate)