Python OrderedDict.iteritems方法代码示例

# 需要导入模块: from theano.compat import OrderedDict [as 别名]
# 或者: from theano.compat.OrderedDict import iteritems [as 别名]

#.........这里部分代码省略.........
        if (isinstance(arg.variable, SharedVariable) and not arg.update)
    ]
    other_shared_inner_args = [
        safe_new(arg.variable, "_copy")
        for arg in dummy_f.maker.expanded_inputs
        if (isinstance(arg.variable, SharedVariable) and not arg.update)
    ]
    givens.update(dict(zip(other_shared_scan_args, other_shared_inner_args)))

    ##
    # Step 6. Re-order the outputs and clone them replacing things
    # using the givens
    ##
    inner_inputs = (
        inner_seqs
        + mit_mot_inner_inputs
        + mit_sot_inner_inputs
        + sit_sot_inner_inputs
        + shared_inner_inputs
        + other_shared_inner_args
        + other_inner_args
    )

    inner_outs = (
        mit_mot_inner_outputs
        + mit_sot_inner_outputs
        + sit_sot_inner_outputs
        + nit_sot_inner_outputs
        + shared_inner_outputs
    )
    if condition is not None:
        inner_outs.append(condition)
    new_givens = OrderedDict()
    for w, w_copy in givens.iteritems():
        new_givens[w] = w.type.filter_variable(w_copy)

    new_outs = scan_utils.clone(inner_outs, replace=new_givens)

    ##
    # Step 7. Create the Scan Op
    ##

    tap_array = mit_sot_tap_array + [[-1] for x in xrange(n_sit_sot)]
    info = OrderedDict()

    info["tap_array"] = tap_array
    info["n_seqs"] = n_seqs
    info["n_mit_mot"] = n_mit_mot
    info["n_mit_mot_outs"] = n_mit_mot_outs
    info["mit_mot_out_slices"] = mit_mot_out_slices
    info["n_mit_sot"] = n_mit_sot
    info["n_sit_sot"] = n_sit_sot
    info["n_shared_outs"] = n_shared_outs
    info["n_nit_sot"] = n_nit_sot
    info["truncate_gradient"] = -1
    info["name"] = name
    info["mode"] = mode
    info["destroy_map"] = OrderedDict()
    info["inplace"] = False
    info["gpu"] = False
    info["as_while"] = as_while
    info["profile"] = profile
    info["_scan_savemem_visited"] = True
    info["allow_gc"] = allow_gc

    local_op = scan_op.Scan(inner_inputs, new_outs, info)

# 需要导入模块: from theano.compat import OrderedDict [as 别名]
# 或者: from theano.compat.OrderedDict import iteritems [as 别名]

#.........这里部分代码省略.........
                           other_shared_inner_args)))

    ##
    # Step 6. Re-order the outputs and clone them replacing things
    # using the givens
    ##
    inner_inputs = (inner_seqs +
                    mit_mot_inner_inputs +
                    mit_sot_inner_inputs +
                    sit_sot_inner_inputs +
                    shared_inner_inputs +
                    other_shared_inner_args +
                    other_inner_args)

    inner_outs = (mit_mot_inner_outputs +
                  mit_sot_inner_outputs +
                  sit_sot_inner_outputs +
                  nit_sot_inner_outputs +
                  shared_inner_outputs)
    if condition is not None:
        inner_outs.append(condition)
    # Cuda is imported here, instead of being imported on top of the file
    # because forces on the user some dependencies that we might do not want
    # to. Currently we are working on removing the dependencies on sandbox
    # code completeley.
    from theano.sandbox import cuda
    if cuda.cuda_available:
        # very often we end up in this situation when we want to
        # replace w with w_copy, where w is CudaNdarray
        # and w_copy is TensorType. This is caused because shared
        # variables are put on GPU right aways >:| ,
        new_givens = OrderedDict()

        for w, w_copy in givens.iteritems():
            if (isinstance(w.type, cuda.CudaNdarrayType)
                and isinstance(w_copy.type, tensor.TensorType)):
                for o in inner_outs:
                    new_givens = traverse(o, w, w_copy, new_givens)
            else:
                new_givens[w] = w_copy
    else:
        new_givens = givens

    new_outs = scan_utils.clone(inner_outs, replace=new_givens)

    ##
    # Step 7. Create the Scan Op
    ##

    tap_array = mit_sot_tap_array + [[-1] for x in xrange(n_sit_sot)]
    if allow_gc is None:
        allow_gc = config.scan.allow_gc
    info = OrderedDict()

    info['tap_array'] = tap_array
    info['n_seqs'] = n_seqs
    info['n_mit_mot'] = n_mit_mot
    info['n_mit_mot_outs'] = n_mit_mot_outs
    info['mit_mot_out_slices'] = mit_mot_out_slices
    info['n_mit_sot'] = n_mit_sot
    info['n_sit_sot'] = n_sit_sot
    info['n_shared_outs'] = n_shared_outs
    info['n_nit_sot'] = n_nit_sot
    info['truncate_gradient'] = truncate_gradient
    info['name'] = name
    info['mode'] = mode

# 需要导入模块: from theano.compat import OrderedDict [as 别名]
# 或者: from theano.compat.OrderedDict import iteritems [as 别名]
                  shared_inner_outputs)
    if condition is not None:
        inner_outs.append(condition)
    # Cuda is imported here, instead of being imported on top of the file
    # because forces on the user some dependencies that we might do not want
    # to. Currently we are working on removing the dependencies on sandbox
    # code completeley.
    from theano.sandbox import cuda
    if cuda.cuda_available:
        # very often we end up in this situation when we want to
        # replace w with w_copy, where w is CudaNdarray
        # and w_copy is TensorType. This is caused because shared
        # variables are put on GPU right aways >:| ,
        new_givens = OrderedDict()

        for w, w_copy in givens.iteritems():
            if (isinstance(w.type, cuda.CudaNdarrayType)
                and isinstance(w_copy.type, tensor.TensorType)):
                for o in inner_outs:
                    new_givens = traverse(o, w, w_copy, new_givens)
            else:
                new_givens[w] = w_copy
    else:
        new_givens = givens

    new_outs = scan_utils.clone(inner_outs, replace=new_givens)

    ##
    # Step 7. Create the Scan Op
    ##