Python utils.conv_to_fc方法代码示例

# 需要导入模块: from baselines.a2c import utils [as 别名]
# 或者: from baselines.a2c.utils import conv_to_fc [as 别名]
def __init__(self, params, ob_space, ac_space, nbatch, nsteps): #pylint: disable=W0613
        nenv = nbatch // nsteps
        ob_shape = (nbatch,) + ob_space.shape
        nact = ac_space.n
        X = tf.placeholder(tf.float32, ob_shape, name='Ob') #obs
        with tf.name_scope('policy_new'):
            activ = tf.nn.relu
            h1 = activ(tf.nn.conv2d(X/255.0, params['policy/c1/w:0'], [1, 4, 4, 1], 'VALID') + params['policy/c1/b:0'])
            h2 = activ(tf.nn.conv2d(h1, params['policy/c2/w:0'], [1, 2, 2, 1], 'VALID') + params['policy/c2/b:0'])
            h3 = activ(tf.nn.conv2d(h2, params['policy/c3/w:0'], [1, 1, 1, 1], 'VALID') + params['policy/c3/b:0'])
            h3 = conv_to_fc(h3)
            h4 = activ(tf.nn.xw_plus_b(h3, params['policy/fc1/w:0'], params['policy/fc1/b:0']))
            pi = tf.nn.xw_plus_b(h4, params['policy/pi/w:0'], params['policy/pi/b:0'])

        self.pdtype = make_pdtype(ac_space)
        self.pd = self.pdtype.pdfromflat(pi)

        self.X = X 

# 需要导入模块: from baselines.a2c import utils [as 别名]
# 或者: from baselines.a2c.utils import conv_to_fc [as 别名]
def nature_cnn(unscaled_images):
    """
    CNN from Nature paper.
    """
    scaled_images = tf.cast(unscaled_images, tf.float32) / 255.
    activ = tf.nn.relu
    h = activ(conv(scaled_images, 'c1', nf=32, rf=8, stride=4, init_scale=np.sqrt(2)))
    h2 = activ(conv(h, 'c2', nf=64, rf=4, stride=2, init_scale=np.sqrt(2)))
    h3 = activ(conv(h2, 'c3', nf=64, rf=3, stride=1, init_scale=np.sqrt(2)))
    h3 = conv_to_fc(h3)
    return activ(fc(h3, 'fc1', nh=512, init_scale=np.sqrt(2))) 

# 需要导入模块: from baselines.a2c import utils [as 别名]
# 或者: from baselines.a2c.utils import conv_to_fc [as 别名]
def nature_cnn(unscaled_images, **conv_kwargs):
    """
    CNN from Nature paper.
    """
    scaled_images = tf.cast(unscaled_images, tf.float32) / 255.
    activ = tf.nn.relu
    h = activ(conv(scaled_images, 'c1', nf=32, rf=8, stride=4, init_scale=np.sqrt(2),
                   **conv_kwargs))
    h2 = activ(conv(h, 'c2', nf=64, rf=4, stride=2, init_scale=np.sqrt(2), **conv_kwargs))
    h3 = activ(conv(h2, 'c3', nf=64, rf=3, stride=1, init_scale=np.sqrt(2), **conv_kwargs))
    h3 = conv_to_fc(h3)
    return activ(fc(h3, 'fc1', nh=512, init_scale=np.sqrt(2))) 

# 需要导入模块: from baselines.a2c import utils [as 别名]
# 或者: from baselines.a2c.utils import conv_to_fc [as 别名]
def cnn_small(**conv_kwargs):
    def network_fn(X):
        h = tf.cast(X, tf.float32) / 255.
        
        activ = tf.nn.relu
        h = activ(conv(h, 'c1', nf=8, rf=8, stride=4, init_scale=np.sqrt(2), **conv_kwargs))
        h = activ(conv(h, 'c2', nf=16, rf=4, stride=2, init_scale=np.sqrt(2), **conv_kwargs))
        h = conv_to_fc(h)
        h = activ(fc(h, 'fc1', nh=128, init_scale=np.sqrt(2)))
        return h, None
    return network_fn 

# 需要导入模块: from baselines.a2c import utils [as 别名]
# 或者: from baselines.a2c.utils import conv_to_fc [as 别名]
def cnn_small(**conv_kwargs):
    def network_fn(X):
        h = tf.cast(X, tf.float32) / 255.

        activ = tf.nn.relu
        h = activ(conv(h, 'c1', nf=8, rf=8, stride=4, init_scale=np.sqrt(2), **conv_kwargs))
        h = activ(conv(h, 'c2', nf=16, rf=4, stride=2, init_scale=np.sqrt(2), **conv_kwargs))
        h = conv_to_fc(h)
        h = activ(fc(h, 'fc1', nh=128, init_scale=np.sqrt(2)))
        return h
    return network_fn 

# 需要导入模块: from baselines.a2c import utils [as 别名]
# 或者: from baselines.a2c.utils import conv_to_fc [as 别名]
def nature_cnn(unscaled_images, first_layer_mode='', trainable=True):
    h3, _h3 = nature_cnn_h3(unscaled_images, first_layer_mode, trainable)
    h3 = conv_to_fc(h3)
    return tf.nn.relu(fc(h3, 'fc1', nh=512, init_scale=np.sqrt(2), trainable=trainable)) 

# 需要导入模块: from baselines.a2c import utils [as 别名]
# 或者: from baselines.a2c.utils import conv_to_fc [as 别名]
def __init__(self, sess, ob_space, ac_space, nenv, nsteps, nstack, nlstm=256, reuse=False):
        nbatch = nenv*nsteps
        nh, nw, nc = ob_space.shape
        ob_shape = (nbatch, nh, nw, nc*nstack)
        nact = ac_space.n
        X = tf.placeholder(tf.uint8, ob_shape) #obs
        M = tf.placeholder(tf.float32, [nbatch]) #mask (done t-1)
        S = tf.placeholder(tf.float32, [nenv, nlstm*2]) #states
        with tf.variable_scope("model", reuse=reuse):
            h = conv(tf.cast(X, tf.float32)/255., 'c1', nf=32, rf=8, stride=4, init_scale=np.sqrt(2))
            h2 = conv(h, 'c2', nf=64, rf=4, stride=2, init_scale=np.sqrt(2))
            h3 = conv(h2, 'c3', nf=64, rf=3, stride=1, init_scale=np.sqrt(2))
            h3 = conv_to_fc(h3)
            h4 = fc(h3, 'fc1', nh=512, init_scale=np.sqrt(2))
            xs = batch_to_seq(h4, nenv, nsteps)
            ms = batch_to_seq(M, nenv, nsteps)
            h5, snew = lnlstm(xs, ms, S, 'lstm1', nh=nlstm)
            h5 = seq_to_batch(h5)
            pi = fc(h5, 'pi', nact, act=lambda x:x)
            vf = fc(h5, 'v', 1, act=lambda x:x)

        v0 = vf[:, 0]
        a0 = sample(pi)
        self.initial_state = np.zeros((nenv, nlstm*2), dtype=np.float32)

        def step(ob, state, mask):
            a, v, s = sess.run([a0, v0, snew], {X:ob, S:state, M:mask})
            return a, v, s

        def value(ob, state, mask):
            return sess.run(v0, {X:ob, S:state, M:mask})

        self.X = X
        self.M = M
        self.S = S
        self.pi = pi
        self.vf = vf
        self.step = step
        self.value = value 

# 需要导入模块: from baselines.a2c import utils [as 别名]
# 或者: from baselines.a2c.utils import conv_to_fc [as 别名]
def __init__(self, sess, ob_space, ac_space, nenv, nsteps, nstack, reuse=False):
        nbatch = nenv * nsteps
        nh, nw, nc = ob_space.shape
        ob_shape = (nbatch, nh, nw, nc * nstack)
        nact = ac_space.n
        X = tf.placeholder(tf.uint8, ob_shape)  # obs
        with tf.variable_scope("model", reuse=reuse):
            h = conv(tf.cast(X, tf.float32) / 255., 'c1', nf=32, rf=8, stride=4, init_scale=np.sqrt(2))
            h2 = conv(h, 'c2', nf=64, rf=4, stride=2, init_scale=np.sqrt(2))
            h3 = conv(h2, 'c3', nf=64, rf=3, stride=1, init_scale=np.sqrt(2))
            h3 = conv_to_fc(h3)
            h4 = fc(h3, 'fc1', nh=512, init_scale=np.sqrt(2))
            pi_logits = fc(h4, 'pi', nact, act=lambda x: x, init_scale=0.01)
            pi = tf.nn.softmax(pi_logits)
            q = fc(h4, 'q', nact, act=lambda x: x)

        a = sample(pi_logits)  # could change this to use self.pi instead
        self.initial_state = []  # not stateful
        self.X = X
        self.pi = pi  # actual policy params now
        self.q = q

        def step(ob, *args, **kwargs):
            # returns actions, mus, states
            a0, pi0 = sess.run([a, pi], {X: ob})
            return a0, pi0, []  # dummy state

        def out(ob, *args, **kwargs):
            pi0, q0 = sess.run([pi, q], {X: ob})
            return pi0, q0

        def act(ob, *args, **kwargs):
            return sess.run(a, {X: ob})

        self.step = step
        self.out = out
        self.act = act