class Context(object):
def __init__(self, instructions):
self.registers = Registers()
self.flags = Flags()
self.instructions = instructions
self.heap = Memory(size=40, mode=HEAP, start_address=0x0000)
self.stack = Memory(size=40, mode=STACK, start_address=0xFFFF)
self.registers.set(SP, 0xFFFF)
def run(self):
"""
initialize the context and execute the first instruction
"""
self.registers.reset()
def step(self):
"""
execute the next instruction whose address in the EIP value
:return:
"""
self.registers.tick()
self.flags.tick()
self.heap.tick()
self.stack.tick()
next_address = self.registers.get(IP).value
if next_address < len(self.instructions):
next_instruction = self.instructions[next_address]
next_instruction.execute(self)
self.registers.set(IP, next_address + 1)
return True
else:
return False
def try_alf_word(self):
if self.get() == '"':
# exactly five mix-chars in inverted or
self.next()
if self.get() == '"':
s = ""
else:
s = self.get()
if self.look() != '"':
raise UnquotedStringError(self.line.argument)
self.next()
else:
# less than six mix-chars not in inverted
s = self.line.argument.rstrip('\n\r')
self.ct = len(self.tokens) - 1
if s is None:
s = ""
s = s[:5]
while len(s) < 5:
s += " "
# now s - string with len = 5
word = Memory.positive_zero()
for i in xrange(1, 6):
word[i] = charset.ord(s[i - 1])
if word[i] is None:
raise InvalidCharError(s[i - 1])
return Memory.mix2dec(word)
def try_w_exp(self):
"""This function DO SELF.NEXT()"""
word = Memory.positive_zero()
value = self.try_exp()
if value is None:
return None
if self.look() == "(":
self.next()
field = self.try_f_part()
else:
# it's property of w-exp that empty f-part means not default value
# but 0:5
field = 5
self.next()
if Memory.apply_to_word(value, word, field) is None:
raise InvalidFieldSpecError(field)
while True:
if self.get() != ",":
break
self.next()
value = self.try_exp()
if value is None:
raise ExpectedExpError(self.get_all_before_this())
if self.look() == "(":
self.next()
field = self.try_f_part()
else:
field = get_codes(self.line.operation)[1]
self.next()
if Memory.apply_to_word(value, word, field) is None:
raise InvalidFieldSpecError(field)
return Memory.mix2dec(word)
def __init__(self, parent, name, address, device_info=None, auto_update=False):
self.auto_update = auto_update
self.parent = parent
self.last_values = {}
Memory.__init__(self, name=name, width_bits=32, address=address, length_bytes=4)
self.process_info(device_info)
LOGGER.debug('New Register %s' % self)
def __init__(self, instructions):
self.registers = Registers()
self.flags = Flags()
self.instructions = instructions
self.heap = Memory(size=40, mode=HEAP, start_address=0x0000)
self.stack = Memory(size=40, mode=STACK, start_address=0xFFFF)
self.registers.set(SP, 0xFFFF)
def _determine_blockshape(self, outputSlot):
"""
Choose a blockshape using the slot metadata (if available) or an arbitrary guess otherwise.
"""
input_shape = outputSlot.meta.shape
ideal_blockshape = outputSlot.meta.ideal_blockshape
ram_usage_per_requested_pixel = outputSlot.meta.ram_usage_per_requested_pixel
num_channels = 1
tagged_shape = outputSlot.meta.getTaggedShape()
# Generally, we don't want to split requests across channels.
if 'c' in tagged_shape.keys():
num_channels = tagged_shape['c']
channel_index = tagged_shape.keys().index('c')
input_shape = input_shape[:channel_index] + input_shape[channel_index+1:]
if ideal_blockshape:
# Never enlarge 'ideal' in the channel dimension.
num_channels = ideal_blockshape[channel_index]
ideal_blockshape = ideal_blockshape[:channel_index] + ideal_blockshape[channel_index+1:]
max_blockshape = input_shape
available_ram = Memory.getAvailableRamComputation()
if ram_usage_per_requested_pixel is None:
# Make a conservative guess: 2*(bytes for dtype) * (num channels) + (fudge factor=4)
ram_usage_per_requested_pixel = 2*outputSlot.meta.dtype().nbytes*num_channels + 4
warnings.warn( "Unknown per-pixel RAM requirement. Making a guess." )
# Safety factor (fudge factor): Double the estimated RAM usage per pixel
safety_factor = 2.0
logger.info("Estimated RAM usage per pixel is {} * safety factor ({})"
.format( Memory.format(ram_usage_per_requested_pixel), safety_factor ) )
ram_usage_per_requested_pixel *= safety_factor
if ideal_blockshape is None:
blockshape = determineBlockShape( input_shape, available_ram/(self._num_threads*ram_usage_per_requested_pixel) )
if 'c' in outputSlot.meta.getAxisKeys():
blockshape = blockshape[:channel_index] + (num_channels,) + blockshape[channel_index:]
warnings.warn( "Chose an arbitrary request blockshape {}".format( blockshape ) )
else:
logger.info("determining blockshape assuming available_ram is {}"
", split between {} threads"
.format(Memory.format(available_ram), self._num_threads))
# By convention, ram_usage_per_requested_pixel refers to the ram used when requesting ALL channels of a 'pixel'
# Therefore, we do not include the channel dimension in the blockshapes here.
blockshape = determine_optimal_request_blockshape( max_blockshape,
ideal_blockshape,
ram_usage_per_requested_pixel,
self._num_threads,
available_ram )
if 'c' in outputSlot.meta.getAxisKeys():
blockshape = blockshape[:channel_index] + (num_channels,) + blockshape[channel_index:]
logger.info( "Chose blockshape: {}".format( blockshape ) )
fmt = Memory.format(ram_usage_per_requested_pixel *
numpy.prod(blockshape[:-1]))
logger.info("Estimated RAM usage per block is {}".format(fmt))
return blockshape
def __init__(self, env, model, epsilon=.9, min_epsilon=.1, epsilon_decay=1e-3): self.env = env self.model = model self.epsilon = epsilon self.min_epsilon = min_epsilon self.epsilon_decay = epsilon_decay self.episode = 0 self.positiveMemory = Memory(model=self.model, episode_max_size=20) self.negativeMemory = Memory(model=self.model, episode_max_size=10)
def memory():
"""
Create Memory instance for testing.
:return: new memory instance
"""
from gb import GB
from memory import Memory
gb = GB()
mem = Memory(gb)
mem.load_cartridge(cartridge_data=bytes.fromhex("00")*0x8000)
return mem
def _build_memory(screen):
""" Initialize CPC464 memory map """
memory = Memory(screen)
memory.add_chunk(0x0000, RomChunk(_get_data_from_file('6128L.rom')))
memory.add_chunk(0xC000, UpperRomChunk(0, _get_data_from_file('6128U-basic.rom')))
memory.apply_ram_map(0)
memory.dump_map()
return memory
def __init__(self, parent, name, address, length_bytes, device_info=None):
"""
:param parent: Parent object who owns this TenGbe instance
:param name: Unique name of the instance
:param address:
:param length_bytes:
:param device_info: Information about this device
:return:
"""
Memory.__init__(self, name, 32, address, length_bytes)
Gbe.__init__(self, parent, name, address, length_bytes, device_info)
def run(path, debug, max_cycles):
with open(path, "rb") as rom_file:
debug_title = debug == "TITLE"
debug_header = debug == "HEADER" or debug == "ALL"
debug_mem = debug == "MEMORY" or debug == "ALL"
debug_instructions = debug == "INSTRUCTIONS" or debug == "ALL"
debug_registers = debug == "REGISTERS" or debug == "ALL"
rom = [i for i in rom_file.read()]
header = Header(rom, debug_header)
mem = Memory(rom, header)
if debug_title:
print("Title: " + header.name)
if debug_instructions:
print("PC: Operation")
interrupts = Interrupts()
cpu = CPU(mem, interrupts, debug_instructions, debug_registers)
timer = Timer(interrupts)
sound = Sound()
link = Link()
joypad = Joypad()
lcdc = LCDC(mem, interrupts)
mem.setupIO(lcdc, interrupts, timer, sound, link, joypad)
total_cycles = 0
try:
pygame.init()
while cpu.run_state != "QUIT":
for event in pygame.event.get():
if event.type == pygame.QUIT:
cpu.run_state = "QUIT"
if event.type == pygame.KEYDOWN or event.type == pygame.KEYUP:
joypad.keyEvent(event)
interrupts.update()
if cpu.run_state == "RUN":
cpu.run()
else:
cpu.cycles += 1
timer.update(cpu.cycles)
lcdc.update(cpu.cycles)
total_cycles += cpu.popCycles()
if max_cycles >= 0 and total_cycles > max_cycles:
cpu.run_state = "QUIT"
except AssertionError as e:
if debug_mem:
mem.display()
traceback.print_tb(e.__traceback__)
except KeyboardInterrupt as e:
if debug_mem:
mem.display()
else:
if debug_mem:
mem.display()
def test(test_iter, folds, training_folds):
results = []
mem = Memory()
for i in range(test_iter):
print "iteration %d ..." % (i + 1)
ini_set = split_set2(folds, senseval.instances()[0:])
for j in range(folds):
print"...fold %d ..." % (j + 1)
sets = partition_set(training_folds, ini_set, j)
print "-$$Train time$$-"
mem.train(sets[0])
print "-$$results time$$-"
results.append(mem.test(sets[1]))
return results
def test_main():
mem = Memory()
print "loading data_set"
ini_set = split_set2(5, senseval.instances()[0:10000])
data_set = partition_set(4, ini_set, 0)
#Serializer.save("/tmp/portioned_data", data_set)
#data_set = Serializer.load("/tmp/portioned_data")
print "training data"
mem.train(data_set[0])
#print "saving data"
#mem.save_values("/tmp/mem_internals")
#mem.load_values("/tmp/mem_internals")
print "------*********testing**********------"
results = mem.test(data_set[1])
print "%3.1f %% accuracy" %(sum(results)/len(results) * 100)
def __init__(self, environment, inputs):
self.environment = environment
self.state_size = inputs
self.nr_actions = environment.action_space.n
self.memory = Memory(30000)
self.discountFactor = 0.975
self.predictionModels = []
def __init__(self, config):
self.config = config # 应用程序配置
self.connections = 0 # 连接客户端列表
if self.config['savetime'] != 0: # 不保存数据
self.thread = PeriodicCallback(self.save_db,
int(self.config['savetime']) * 1000)
self.thread.start() # 背景保存服务线程启动
self.workpool = ThreadPool(int(config['work_pool'])) # 工作线程池
self.status = Status(CacheServer.status_fields) # 服务器状态
self.slave_clients = {} # 同步客户端
self.is_sync = False # 是否在同步
if self.config['master'] is not None: # 从服务器启动
shutil.rmtree(config['db'])
self.master_server = PyCachedClient(self.config['master'][0],
self.config['master'][1])
self.master_server.sync(self.config['port'])
self.slavepool = None
self.slave = True # 是否为Slave模式
else: # 主服务器启动, 需要启动从命令工作线程
self.slavepool = ThreadPool(int(config['slave_pool'])) # slave Command send pools
self.slave = False # 是否为Slave模式
self.memory = Memory(config['db']) # 缓存服务类
super(CacheServer, self).__init__()
def __init__(self, obs_dim, action_dim, hiddens_actor, hiddens_critic, layer_norm=False, memory_size=50000):
self.obs_dim = obs_dim
self.action_dim = action_dim
self.noise_stddev = 1.
self.noise_stddev_decrease = 5e-4
self.noise_stddev_lower = 5e-2
actor_activations = [dy.tanh for _ in range(len(hiddens_actor))] + [dy.tanh]
critic_activations = [dy.tanh for _ in range(len(hiddens_critic))] + [None]
self.actor = MLP(inpt_shape=(obs_dim,), hiddens=hiddens_actor + [action_dim], activation=actor_activations,
layer_norm=layer_norm)
self.critic = MLP(inpt_shape=(obs_dim + action_dim,), hiddens=hiddens_critic + [1],
activation=critic_activations, layer_norm=layer_norm)
self.actor_target = MLP(inpt_shape=(obs_dim,), hiddens=hiddens_actor + [action_dim],
activation=actor_activations, layer_norm=layer_norm)
self.critic_target = MLP(inpt_shape=(obs_dim + action_dim,), hiddens=hiddens_critic + [1],
activation=critic_activations, layer_norm=layer_norm)
self.actor_target.update(self.actor, soft=False)
self.critic_target.update(self.critic, soft=False)
self.trainer_actor = dy.AdamTrainer(self.actor.pc)
self.trainer_critic = dy.AdamTrainer(self.critic.pc)
self.trainer_actor.set_learning_rate(1e-4)
self.trainer_critic.set_learning_rate(1e-3)
self.memory = Memory(memory_size)
def __init__(self, environment, inputs):
self.input_size = inputs
self.output_size = environment.action_space.n
self.memory = Memory(2000)
self.discountFactor = 0.975
self.learnStart = 36
self.models = [None] * 5
def __init__(self, size_state, nr_actions, memorySize, discountFactor, learningRate, learnStart):
self.input_size = size_state
self.output_size = nr_actions
self.memory = Memory(memorySize)
self.discountFactor = discountFactor
self.learnStart = learnStart
self.learningRate = learningRate
def __init__(self, queuetype, memtype, t_cs_val = 13):
self.t_cs = t_cs_val
self.t_slice = 80 # in ms
self.t_memmove = 10 # in ms
if queuetype == "FCFS":
self.process_queue = FCFSQueue()
elif queuetype == "SRT":
self.process_queue = SRTQueue()
elif queuetype == "PWA":
self.process_queue = PWAQueue()
elif queuetype == "RR":
self.process_queue = FCFSQueue()
self.mem = Memory(memtype)
self.queueType = queuetype
self.CPUIdle = True
self.n = 0
self.active_n = 0
self.maxID = 1
self.processInCPU = None #Note: processInCPU is the process in use of CPU, NOT in content switch
self.burstTimeSum = 0
self.waitTimeSum = 0
self.waitTimeNum = 0
self.turnaroundTimeSum = 0
self.contentSwitchSum = 0
self.processInCPU_tobe = None
self.defragtime = 0
self.p_list = []
def __init__(self, inputs, outputs, memorySize, discountFactor, learningRate, learnStart):
self.input_size = inputs
self.output_size = outputs
self.memory = Memory(memorySize)
self.discountFactor = discountFactor
self.learnStart = learnStart
self.learningRate = learningRate
def __init__(self):
self.memory = Memory()
self.OPERATORS = {":=" : self.opDefine, "<<" : self.opDefineKeyword, "++" : self.opIncrement}
self.COMMANDS = {"remember" : self.comRemember, "recall" : self.comFindQuote,
"evaluate" : self.comEvaluate, "count" : self.comCount, "findfactoid" : self.comFactoidSearch,
"findquote" : self.comQuoteSearch, "delete" : self.comDeleteFactoid }
self.PROCESSCOMMANDS = {"remember" : False, "recall" : False, "evaluate" : True, "count" : False,
"findfactoid" : False, "findquote" : False, "delete" : False }
def __init__(self, env, params):
self.env = env
params.actions = env.actions()
self.num_actions = env.actions()
self.episodes = params.episodes
self.steps = params.steps
self.train_steps = params.train_steps
self.update_freq = params.update_freq
self.save_weights = params.save_weights
self.history_length = params.history_length
self.discount = params.discount
self.eps = params.init_eps
self.eps_delta = (params.init_eps - params.final_eps) / params.final_eps_frame
self.replay_start_size = params.replay_start_size
self.eps_endt = params.final_eps_frame
self.random_starts = params.random_starts
self.batch_size = params.batch_size
self.ckpt_file = params.ckpt_dir+'/'+params.game
self.global_step = tf.Variable(0, trainable=False)
if params.lr_anneal:
self.lr = tf.train.exponential_decay(params.lr, self.global_step, params.lr_anneal, 0.96, staircase=True)
else:
self.lr = params.lr
self.buffer = Buffer(params)
self.memory = Memory(params.size, self.batch_size)
with tf.variable_scope("train") as self.train_scope:
self.train_net = ConvNet(params, trainable=True)
with tf.variable_scope("target") as self.target_scope:
self.target_net = ConvNet(params, trainable=False)
self.optimizer = tf.train.RMSPropOptimizer(self.lr, params.decay_rate, 0.0, self.eps)
self.actions = tf.placeholder(tf.float32, [None, self.num_actions])
self.q_target = tf.placeholder(tf.float32, [None])
self.q_train = tf.reduce_max(tf.mul(self.train_net.y, self.actions), reduction_indices=1)
self.diff = tf.sub(self.q_target, self.q_train)
half = tf.constant(0.5)
if params.clip_delta > 0:
abs_diff = tf.abs(self.diff)
clipped_diff = tf.clip_by_value(abs_diff, 0, 1)
linear_part = abs_diff - clipped_diff
quadratic_part = tf.square(clipped_diff)
self.diff_square = tf.mul(half, tf.add(quadratic_part, linear_part))
else:
self.diff_square = tf.mul(half, tf.square(self.diff))
if params.accumulator == 'sum':
self.loss = tf.reduce_sum(self.diff_square)
else:
self.loss = tf.reduce_mean(self.diff_square)
# backprop with RMS loss
self.task = self.optimizer.minimize(self.loss, global_step=self.global_step)
def __init__(self):
'''
Births a new Parrott,
with a Naive Bayes brain
and a Solr memory.
'''
self.brain = naive_bayes.NaiveBayes()
self.twitter = membrane.twitter.api()
self.memory = Memory()
def __init__(self):
self.control = Control()
self.register = Registers(32)
self.pcreg = PipelineReg()
self.if_id = PipelineReg()
self.id_exe = PipelineReg()
self.exe_mem = PipelineReg()
self.mem_wb = PipelineReg()
self.alu = Alu()
self.memory = Memory()
self.initializePipeline()
def read(self, **kwargs):
"""
Memory.read returns a list for all bitfields, so just put those
values into single values.
"""
memdata = Memory.read(self, **kwargs)
results = memdata['data']
timestamp = memdata['timestamp']
for k, v in results.iteritems():
results[k] = v[0]
self.last_values = results
return {'data': results, 'timestamp': timestamp}
def run_model(config_file, machine_id):
print "Running {}".format(machine_id)
machine = Machine(config_file)
machine.get_id(machine_id)
memory = Memory(machine)
networking = Networking(machine)
ai_features = dict(memory.ai_features().items() + networking.ai_features().items())
procfs_features = dict(memory.procfs_features().items() + networking.procfs_features().items())
sysfs_features = dict(memory.sysfs_features().items() + networking.sysfs_features().items())
machine.write_features('ai_features', ai_features)
machine.write_features('procfs', procfs_features)
machine.write_features('sysfs', sysfs_features)
machine.write_features('quick', {
'Security': 0,
'Stats': len(procfs_features) + len(sysfs_features),
})
def reset(self):
with open(self.fname) as f:
self.mem = Memory(f)
self.door_unlocked = False
self.step_count = 1
self.break_at_finish = -1
self.registers = Registers()
self.registers[PC] = self.mem[0xfffe]
self.call_targets = [self.registers[PC]]
self.callsites = []
self.current_block_start = self.registers[PC]
self.insn_count = 0
def __init__(self):
self.logger = Log()
# Create components
self.cpu = CPU(self)
self.memory = Memory(self)
self.interrupts = Interrupts(self)
self.screen = Screen(self)
self.gpu = GPU(self)
self.debug_mode = False
self.step_mode = False
class Stack(object):
def __init__(self, size):
self._stack = Memory(size)
self._sp = 0
def push(self, word):
self._sp -= 2
self._stack.write_word(self._sp, word)
def pop(self):
self._sp += 2
return self._stack.read_word(self._sp-2)
def set_position(self, position):
self._sp = position
def position(self):
return self._sp
def __repr__(self):
return str(self._stack)
def __init__(self, numIntRegisters = 32, numFloatRegisters = 32, timingModel = timingmodel.defaultTimingModel) :
self.memory = Memory()
self.registerFile = {}
self.numIntRegisters = numIntRegisters
self.numFloatRegisters = numFloatRegisters
self.__createRegisterFile()
self.memoryManager = MemoryManager(self.memory.heap[0], self.memory.heap[1] - self.memory.heap[0])
# print("Memory allocator " + str(self.memoryManager))
self.timingModel = timingModel()
# print(self.timingModel)
self.prog = None
self.pc = self.memory.text[0]
self.registerFile['sp'].write(self.memory.text[1] - 4) #initialize the stack pointer
def update_weights(self, memory: Memory):
gamma = 0.95
elements = memory.sample(200)
x = []
y = []
for element in elements:
x.append(element.old_state)
new_reward = element.reward + gamma * np.max(self._predict(element.new_state))
scores = self._predict(element.old_state)
scores[element.action] = new_reward
y.append(scores)
x = np.asarray(x)
y = np.asarray(y)
self._model.fit(x=x, y=y, verbose=0)
def main():
env = gym.make(env_name)
env.seed(500)
torch.manual_seed(500)
num_inputs = env.observation_space.shape[0]
num_actions = env.action_space.n
env.close()
global_target_model = Model(num_inputs, num_actions)
global_online_model = Model(num_inputs, num_actions)
global_target_model.train()
global_online_model.train()
global_target_model.load_state_dict(global_online_model.state_dict())
global_target_model.share_memory()
global_online_model.share_memory()
global_memory = Memory(replay_memory_capacity)
global_ep, global_ep_r, res_queue, global_memory_pipe = mp.Value('i', 0), mp.Value('d', 0.), mp.Queue(), mp.Queue()
writer = SummaryWriter('logs')
n = 2
epsilons = [(i * 0.05 + 0.1) for i in range(n)]
actors = [Actor(global_target_model, global_memory_pipe, global_ep, global_ep_r, epsilons[i], i) for i in range(n)]
[w.start() for w in actors]
learner = Learner(global_online_model, global_target_model, global_memory, global_memory_pipe, res_queue)
learner.start()
res = []
while True:
r = res_queue.get()
if r is not None:
res.append(r)
[ep, loss] = r
# writer.add_scalar('log/score', float(ep_r), ep)
writer.add_scalar('log/loss', float(loss), ep)
else:
break
[w.join() for w in actors]
def main():
# set up everything here
# Initilaise pygame
pygame.init()
nes_icon = pygame.image.load('icon.png')
pygame.display.set_icon(nes_icon)
screen = pygame.display.set_mode((256, 240))
pygame.display.set_caption('NESS')
done = False
# Emulator components initialised here
memory = Memory()
cpu = CPU(memory)
rom = ROM(memory)
while not done:
for event in pygame.event.get():
if event.type == pygame.QUIT:
done = True
cpu.execute()
def main():
env = Game()
if initial.INITIAL_MEMORY_VERSION == None:
memory = Memory(config.MEMORY_SIZE)
# load neural network
current_NN = Residual_CNN(config.REG_CONST,config.LEARNING_RATE,(2,) + env.grid_shape, env.action_size, config.HIDDEN_CNN_LAYERS) #????
best_NN = Residual_CNN(config.REG_CONST,config.LEARNING_RATE,(2,) + env.grid_shape, env.action_size, config.HIDDEN_CNN_LAYERS)
best_player_version = 0
best_NN.model.set_weights(current_NN.model.get_weights())
#create players
current_player =
best_player =
iter = 0
while 1:
pass
def get_value(self, memory: Memory, paddress: int):
rslt = {
"%s[%.8X-%.8X]" % (self.name, self.offset, paddress): self.name,
"content": None
}
print("get_value(%s) -->" % self.name)
print(paddress)
content = []
for field in self.fields:
if field.pointer:
address = memory.get_pointer_be(paddress + field.offset)
else:
address = paddress + field.offset
print("processing %s.%s[0x%.8X]" %
(self.name, field.name, address))
content.append(field.get_value(memory, address))
rslt["content"] = content
return rslt
class TestCache(unittest.TestCase):
def setUp(self):
self.sys = MemSys("Test Platform", 64)
self.memory = Memory(self.sys, 4, 4, 4, 2, 64, 1866, True, None)
self.l2c = Cache(self.sys, 1037000, 0, 1, 16, 8, 534288, 64, 3, True,
self.memory.get_component_id())
self.l1c0 = Cache(self.sys, 1037000, 0, 0, 16, 8, 16384, 64, 1, True,
self.l2c.get_component_id())
self.cu0 = ComputeUnit(self.sys, 1037000, 0, True,
self.l1c0.get_component_id())
self.sys.build_map()
def test_cache_line_addr_calc(self):
self.assertEqual(self.l1c0.get_cache_line(0x4000), 0x100)
self.assertEqual(self.l1c0.get_cache_line(0x4010), 0x100)
self.assertEqual(self.l1c0.get_cache_line(0x4110), 0x104)
self.assertEqual(self.l1c0.get_cache_line(0x4112), 0x104)
self.assertEqual(self.l1c0.get_cache_line(0x41100000), 0x1044000)
self.assertEqual(self.l1c0.get_cache_line(0x4110000000000000),
0x104400000000000)
def handle_create(self, cpu, clock):
base_task_id = cpu.memory.info[cpu.internal_state.PC + 1]
print(f"\n--- CREATING TASK FOR BASE PROGR {base_task_id} ---")
try:
new_task = Task()
new_task.initialize(len(self.tasks), self.base_tasks[base_task_id],
Memory(50), Internal_State(), IO())
new_task.task_state = "READY"
for i in range(len(new_task.progr)):
new_task.memory.write(i, new_task.progr[i], cpu.error)
self.tasks.append(new_task)
cpu.internal_state.PC = cpu.internal_state.PC + 1 + 1
except:
cpu.error.update(5)
cpu.internal_state.mode = "PARADA"
cpu.internal_state.mode_info = "RECEIVED PROGRAM " + \
str(base_task_id)
def __init__(self,
step_size=96,
types=1,
spread=10,
pip_cost=1000,
leverage=500,
min_lots=0.01,
assets=100000,
available_assets_rate=0.8):
self.step_size = step_size
spread /= pip_cost
self.data()
self.memory = Memory(5000000)
self.types = types
reward = Reward if types == 1 else Reward2
self.rewards = reward(spread, leverage, pip_cost, min_lots, assets,
available_assets_rate)
self.rewards.max_los_cut = -np.mean(self.atr) * pip_cost
def main(): human = HumanWASDPlayer() play(human, human) pvnet = PolicyValueNet(board_n, model_filename) mem = Memory() while True: mcts_player = MCTSPlayer(pvnet.get_pvnet_fn(), play_style = 3) bh, ph, vh = play(mcts_player, human, mcts_player) mem.save_data((bh, ph, vh)) mcts_player = MCTSPlayer(pvnet.get_pvnet_fn(), play_style = 3) bh, ph, vh = play(human, mcts_player, mcts_player) mem.save_data((bh, ph, vh))
def run(self):
# Constants
game = snakegame(gui=False, seed=1)
MAX_MEMORY_SIZE = 100000
NEURONS = 6 #6
LAYERS = 2 #2
NUM_INPUTS = 8 #208#8
NUM_OUTPUTS = 3
LEARNING_RATE = 0.005
MAX_GAME_STEPS = 300000
LAMBDA = 0
# Setup trainer, memory and model architecture
trainer = Trainer()
memory = Memory(max_memory_size=MAX_MEMORY_SIZE)
model = Network(neurons=NEURONS,
layers=LAYERS,
num_inputs=NUM_INPUTS,
num_outputs=NUM_OUTPUTS,
learning_rate=LEARNING_RATE,
l=LAMBDA)
# Create random agent to fill memory and train it
rand_agent = RandomAgent(memory=memory, model=model)
trainer.train(rand_agent,
game,
max_steps=MAX_MEMORY_SIZE,
save_model=False)
# Create smart agent and train it, use memory from random agent
smart_agent = DDQNAgent(memory=rand_agent.memory,
model=model,
prioritized_replay=True,
epsilon_min=10,
stop_explore=30000)
trainer.train(smart_agent,
game,
max_steps=MAX_GAME_STEPS,
save_model=True)
def main(args):
model_store_sprefix = "snapshot"
# NormalizedEnv
env = gym.make(args.env)
env.seed(args.seed)
torch.manual_seed(args.seed)
env, generator, model, cont = get_functions(env, args)
optimizer = optim.Adam(model.parameters(), lr=args.rllr)
memory = Memory(args)
agent = PPOAgent(args, model, optimizer, env, generator, memory, cont)
if args.resume:
agent.load_model(model_store_sprefix)
agent.train(model_store_sprefix, args.save_interval)
def __init__(self, config_module=None):
"""
param:config_module: a scenario/robot specific module to prepare setup,
that has the following members:
get_all_conditions() -> return a list of conditions
get_all_actions() -> return a list of actions
"""
self.memory = Memory()
self.worldstate = WorldState()
self.actions = set()
if config_module is not None:
for condition in config_module.get_all_conditions(self.memory):
Condition.add(condition)
for action in config_module.get_all_actions(self.memory):
self.actions.add(action)
self.planner = Planner(self.actions, self.worldstate, None)
self._last_goal = None
self._preempt_requested = False # preemption mechanism
def main():
memory = Memory()
math = Math(memory)
memory.set_math(math)
while 42:
command = input("> ")
if command == "exit":
break
command = remove.whitespace(command)
command = command.lower()
validate.equation(command)
if command.endswith('?'):
math.calc(command[:-1])
else:
memory.add_command(command)
print(memory.array)
def run(env_id, seed, noise_type, **kwargs):
# Create envs.
env = gym.make(env_id)
# Parse noise_type
action_noise = None
nb_actions = env.action_space.shape[-1]
for current_noise_type in noise_type.split(','):
current_noise_type = current_noise_type.strip()
if current_noise_type == 'none':
pass
elif 'normal' in current_noise_type:
_, stddev = current_noise_type.split('_')
action_noise = NormalActionNoise(mu=np.zeros(nb_actions),
sigma=float(stddev) *
np.ones(nb_actions))
elif 'ou' in current_noise_type:
_, stddev = current_noise_type.split('_')
action_noise = OrnsteinUhlenbeckActionNoise(
mu=np.zeros(nb_actions),
sigma=float(stddev) * np.ones(nb_actions))
# Configure components.
memory = Memory(limit=int(1e6),
action_shape=env.action_space.shape,
observation_shape=env.observation_space.shape)
# Seed everything to make things reproducible.
logger.info('seed={}, logdir={}'.format(seed, logger.get_dir()))
if seed > 0:
np.random.seed(seed)
torch.manual_seed(seed)
random.seed(seed)
env.seed(seed)
torch.cuda.manual_seed(seed)
start_time = time.time()
train(env=env, action_noise=action_noise, memory=memory, **kwargs)
env.close()
logger.info('total runtime: {}s'.format(time.time() - start_time))
def test_play_matches_neural_network(self):
memory = Memory(config.MEMORY_SIZE)
# At the beginning, we set a random model. It will be similar to an untrained CNN, and quicker.
# We also set config.MCTS_SIMS, which is rather low, and will produce poor estimations from the MCTS.
# The idea is encourage exploration and generate a lot of boards in memory, even if the probabilities
# associated to their possible actions are wrong.
# Memory is completed at the end of the game according to the final winner, in order to correct the values
# of each move. All the moves of the winner receive value=1 and all the moves of the loser receive value=-1
# The neural network will learn to predict the probabilities and the values.
# It will learn wrong probas and values at the beginning, but after some time, the CNN and the neural network
# will improve from eachother and converge.
player1 = Agent('cnn_agent_1', config.GRID_SHAPE[0] * config.GRID_SHAPE[1], config.GRID_SHAPE[1], config.MCTS_SIMS, config.CPUCT, GenRandomModel())
player2 = Agent('cnn_agent_2', config.GRID_SHAPE[0] * config.GRID_SHAPE[1], config.GRID_SHAPE[1], config.MCTS_SIMS, config.CPUCT, GenRandomModel())
scores, memory, points, sp_scores = play_matches.playMatches(player1, player2, config.EPISODES, lg.logger_main, turns_until_tau0 = config.TURNS_UNTIL_TAU0, memory = memory)
# play_matches.playMatches() has copied stmemory to ltmemory, so we can clear stmemory safely
memory.clear_stmemory()
cnn1 = Residual_CNN(config.REG_CONST, config.LEARNING_RATE, (1,) + config.GRID_SHAPE, config.GRID_SHAPE[1], config.HIDDEN_CNN_LAYERS)
cnn2 = Residual_CNN(config.REG_CONST, config.LEARNING_RATE, (1,) + config.GRID_SHAPE, config.GRID_SHAPE[1], config.HIDDEN_CNN_LAYERS)
cnn2.model.set_weights(cnn1.model.get_weights())
cnn1.plot_model()
player1.model = cnn1
######## RETRAINING ########
player1.replay(memory.ltmemory)
for _ in range(1):
scores, memory, points, sp_scores = play_matches.playMatches(player1, player2, config.EPISODES, lg.logger_main, turns_until_tau0 = config.TURNS_UNTIL_TAU0, memory = memory)
# play_matches.playMatches() has copied stmemory to ltmemory, so we can clear stmemory safely
memory.clear_stmemory()
player1.replay(memory.ltmemory)
print('TOURNAMENT...')
scores, _, points, sp_scores = play_matches.playMatches(player1, player2, config.EVAL_EPISODES, lg.logger_main, turns_until_tau0 = 0, memory = None)
print('\nSCORES')
print(scores)
print('\nSTARTING PLAYER / NON-STARTING PLAYER SCORES')
print(sp_scores)
def main():
parser = argparse.ArgumentParser(description='Emulates the 6502 chip.')
parser.add_argument('--hex_file',
required=True,
help='Path to the hex file to be loaded in memory')
parser.add_argument('--verbose', '-v', action='count', default=0)
args = parser.parse_args()
memory = Memory()
cpu = CPU(memory)
cpu.reset()
memory.load_file(args.hex_file)
while True:
cpu.print_cpu_stats()
command = raw_input("\nPress (R) to reset, (N) to execute next "\
"instruction, (M_[NUMBER]) to print a memory"\
" page or (Q) to quit: ")
if command.upper() == 'R':
cpu.reset()
memory.load_file(args.hex_file)
elif command.upper() == 'N':
cpu.execute(1)
elif command.upper() == 'Q':
print('Good bye...')
exit(0)
elif command.upper().startswith('M_'):
page_number = int(command.split('_')[1])
memory.dump_page(page_number)
else:
print('Command [%s] not recognized' % command)
def run(self):
while self.global_ep.value < max_episode:
self.local_model.pull_from_global_model(self.global_model)
done = False
score = 0
steps = 0
state = self.env.reset()
state = torch.Tensor(state)
state = state.unsqueeze(0)
memory = Memory(n_step)
while True:
policy, value = self.local_model(state)
action = self.get_action(policy, self.num_actions)
next_state, reward, done, _ = self.env.step(action)
next_state = torch.Tensor(next_state)
next_state = next_state.unsqueeze(0)
mask = 0 if done else 1
reward = reward if not done or score == 499 else -1
action_one_hot = torch.zeros(2)
action_one_hot[action] = 1
memory.push(state, next_state, action_one_hot, reward, mask)
score += reward
state = next_state
if len(memory) == n_step or done:
batch = memory.sample()
loss = self.local_model.push_to_global_model(
batch, self.global_model, self.global_optimizer)
self.local_model.pull_from_global_model(self.global_model)
memory = Memory(n_step)
if done:
running_score = self.record(score, loss)
break
self.res_queue.put(None)
def open_app(app_name, cells, audio, arduino):
current_app = None
app_name = app_name.replace(" ", "")
if app_name == 'riddles':
current_app = Riddles("Riddles", cells, audio, arduino)
elif app_name == 'learn':
current_app = Learn("Learn", cells, audio, arduino)
elif app_name == 'tutor':
current_app = Tutor("Tutor", cells, audio, arduino)
elif app_name == 'headlines':
current_app = Headlines("Headlines", cells, audio, arduino)
elif app_name == 'memory':
current_app = Memory("Memory", cells, audio, arduino)
if current_app is not None:
audio.speak("Opening the application " + app_name)
current_app.on_start()
else:
audio.speak(
"I did not recognize the app. Could you try to open the app again?"
)
def __init__(self):
args = sys.argv
self.inst_file = args[1]
self.data_file = args[2]
self.config_file = args[3]
self.res_file = args[4]
if len(args) != 5:
self.print_help()
sys.exit(1)
self.instructions = []
self.scoreboards = []
self.clock_mgr = ClockMgr()
self.memory_bus = MemoryBus()
self.memory = Memory(self.data_file)
self.reg_pc = 0
self.dcache = DCache(self.clock_mgr, self.memory_bus, self.memory)
self.cpu = CPU(self.config_file)
self.icache = ICache(self.cpu.icache_config[0],
self.cpu.icache_config[1], self.clock_mgr,
self.memory_bus)
def main():
environment_id, directory = argument_parsing()
directory = os.path.join(directory, environment_id)
environment = Environment(environment_id, directory)
long_term_memory = LongTermMemory(directory)
memory = long_term_memory.restoring_memory_object()
if memory is None:
memory = Memory(environment.observation_space,
number_of_actions=environment.action_space.n)
exploration_rate = long_term_memory.restoring_exploration_object()
if exploration_rate is None:
exploration_rate = 1
decision_maker = DecisionMaker(
state_space=environment.observation_space,
number_of_actions=environment.action_space.n,
model_dir=directory,
exploration_rate=exploration_rate)
agent = Agent(environment, decision_maker, memory)
agent.play()
long_term_memory.saving_memory(memory)
long_term_memory.saving_exploration_rate(
decision_maker.get_exploration_rate())
def load_file(filename):
global s_filename, again, memo, processes, times_to_compact_memory, n_processes, n_accesses
s_filename = filename
again = False
processes = []
times_to_compact_memory = []
with open(filename) as f:
lines = [line.strip() for line in f.readlines()]
memo = Memory(lines[0].split())
n_accesses = 0
pid = 0
for line in lines[1:]:
line = line.split()
if (len(line) == 2): # COMPACTA
times_to_compact_memory.append(int(line[0]))
else: # Processo
new_process = Process(pid, line, memo)
processes.append(new_process)
n_accesses += len(new_process.accesses)
pid = (pid + 1) % 128 # Cada pid tem 8 bits
times_to_compact_memory.sort()
n_processes = len(processes)
return True
class CPU(InstructionSet6502):
register = CpuRegister()
memory = Memory()
def execute_instruction(self, op_code):
pass
def execute_pc(self, pc):
pass
def reset(self):
pass
def get_pc(self):
pass
def nmi(self, vector):
# nmi中断
pass
def irq(self, vector):
# irq中断
pass
def open_app(self, app_name):
current_app = None
app_name = app_name.replace(" ", "")
if app_name == 'riddles':
current_app = Riddles("Riddles")
elif app_name == 'learn':
current_app = Learn("Learn")
elif app_name == 'tutor':
current_app = Tutor("Tutor")
elif app_name == 'headlines':
current_app = Headlines("Headlines")
elif app_name == 'memory':
current_app = Memory("Memory")
if current_app is not None:
glob.mainApp.audio.speak("Opening the application")
glob.mainApp.audio.speak(current_app.name)
current_app.on_start()
else: # shouldn't occur
glob.mainApp.audio.speak(
"I did not recognize the app. Could you try to open the app again?"
)
def setup():
## Start Environment
state_size = reduce(lambda x, y: x * y, env.observation_space.shape)
# initialize memory
mem_size = 2 * state_size + 1 + 1 + 1 # 1=action, 1=reward, 1 = done
memory = Memory(mem_size, 10000) # default size = 10000
# get networks
with tf.variable_scope('net') as scope:
net = QNet(4)
net.append(DenseLayer((4, 64)))
net.append(ActivationLayer('tanh'))
net.append(DenseLayer((64, 2)))
net.setup()
with tf.variable_scope('target') as scope:
target_net = QNet(4)
target_net.append(DenseLayer((4, 64)))
target_net.append(ActivationLayer('tanh'))
target_net.append(DenseLayer((64, 2)))
target_net.setup()
return net, target_net, memory
def run(env_id, exp_prefix, lr_p, lr_v, run_num, seed=1, gamma=.97):
current_time = datetime.now().strftime('%b%d_%H-%M-%S')
try:
os.makedirs(F"./runs/{exp_prefix}")
except OSError:
pass
logdir = F"./runs/{exp_prefix}/lr_p{lr_p}-lr_v{lr_v}-run_num{run_num}-{current_time}"
writer = SummaryWriter(logdir)
env = gym.make(env_id)
memory = Memory(limit=int(1e6),
action_shape=env.action_space.shape,
observation_shape=env.observation_space.shape)
critic = Critic(observation_shape=env.observation_space.shape)
nactions = env.action_space.shape[-1]
actor = Actor(nactions, observation_shape=env.observation_space.shape)
torch.manual_seed(seed)
train(env=env, memory=memory, actor=actor, critic=critic, writer=writer)
writer.close()
def __init__(self,
index,
memory_size,
batch_size,
gamma,
state_global,
action_global,
local=False):
self.hidden = 200
self.memory = Memory(memory_size)
self.state_dim = state_global[index]
self.action_dim = action_global[index]
self.Actor = Actor(self.state_dim, self.action_dim, self.hidden,
self.hidden).to(device)
# local决定是用局部信息还是全局信息，也决定是DDPG算法还是MADDPG算法
if not local:
self.Critic = Critic(sum(state_global), sum(action_global),
self.hidden, self.hidden).to(device)
else:
self.Critic = Critic(self.state_dim, self.action_dim, self.hidden,
self.hidden).to(device)
self.Actor_target = Actor(self.state_dim, self.action_dim, self.hidden,
self.hidden).to(device)
if not local:
self.Critic_target = Critic(sum(state_global), sum(action_global),
self.hidden, self.hidden).to(device)
else:
self.Critic_target = Critic(self.state_dim, self.action_dim,
self.hidden, self.hidden).to(device)
self.critic_train = torch.optim.Adam(self.Critic.parameters(), lr=LR_C)
self.actor_train = torch.optim.Adam(self.Actor.parameters(), lr=LR_A)
self.loss_td = nn.MSELoss()
self.batch_size = batch_size
self.gamma = gamma
self.tau = 0.5
self.local = local
def __init__(self, global_scope="", current_scope=""):
self.global_scope = global_scope
self.current_scope = current_scope
self.function_directory = FunctionDirectory()
self.semantic_cube = SemanticCube()
self.memory = Memory()
self.temporal_variables = []
self.temporal_parameters_names = []
self.temporal_parameters_types = []
self.temporal_arguments_types = []
self.operand_stack = []
self.type_stack = []
self.operator_stack = []
self.quadruple_list = []
self.jump_list = []
self.return_list = []
self.temporal_variable_counter = 0
self.quadruple_number = 1
self.relational_operations = ['>', '<', '>=', '<=', '==', '!=']
self.return_flag = False
self.current_dimensioned_varible = {}
self.dimensioned_varible_stack = []
self.dimensioned_varible_flag = False
self.negation_stack = []
def __init__(self, environment, load_memory=True):
self.action_space = Parameters.GAMES.get_action_space(Parameters.GAME)
self.environment = environment
if load_memory:
self.memory = PrioritizedMemory() if Parameters.USE_PRIORITIZATION else Memory()
self.step = 0
# select the type of DQN based on Parameters
dqn_type = DDDQN if Parameters.USE_DDDQN else DQN
# initialize the DQN and target DQN (with respective placeholders)
self.dqn_input = tf.placeholder(tf.float32,
[None,
Parameters.IMAGE_HEIGHT,
Parameters.IMAGE_WIDTH,
Parameters.AGENT_HISTORY_LENGTH],
name="DQN_input")
self.dqn = dqn_type(self.dqn_input, self.action_space)
self.target_dqn_input = tf.placeholder(
tf.float32,
[None,
Parameters.IMAGE_HEIGHT,
Parameters.IMAGE_WIDTH,
Parameters.AGENT_HISTORY_LENGTH],
name="target_DQN_input")
self.target_dqn = dqn_type(self.target_dqn_input, self.action_space)
# initialize the tensorflow session and variables
self.tf_session = tf.Session()
self.tf_saver = tf.train.Saver()
self.load_session()
self.initial_time = self.last_time = time.time()
self.initial_step = self.step
self.last_action = randint(0, self.action_space)
def main():
parser = argparse.ArgumentParser(description='NES Emulator.')
parser.add_argument('rom_path',
metavar='R',
type=str,
help='path to the nes rom')
args = parser.parse_args()
print(args.rom_path)
with open(args.rom_path, 'rb') as file:
rom_bytes = file.read()
rom = ROM(rom_bytes)
memory = Memory()
ppu = PPU()
cpu = CPU(memory, ppu)
cpu.initialization()
cpu.load_rom(rom)
def __init__(self):
"""Model of a standard contoller unit demonstrating the Texas 4-step (fetch,decode,execute,store) with methods for each."""
self.R1 = 0 # General purpose register to store final result.
self.R2 = 0 # General purpose register to store file length.
self.R3 = "" # Storage register to store mnemonic from memory.
self.IR = 0 # Instruction register
self.PC = 0 # Program counter/accumulator
self.running = False # Machine state
self.clock = time.time() # Start system clock
self.ALU = ALU() # Arithmetic-logic units
self.MEM = Memory() # System memory
