API Gluon reference Parameters ParameterDict ParameterDict.zero_grad

Quick search

Show Source

D2L Book GitHub

Table Of Contents

Guide
- Getting started with NP on MXNet
- What is NP on MXNet
API
- NP on MXNet reference
  - Array objects
    - The N-dimensional array (ndarray)
    - Indexing
  - Routines
    - Array creation routines
      - eye
      - empty
      - full
      - identity
      - ones
      - ones_like
      - zeros
      - zeros_like
      - array
      - copy
      - arange
      - linspace
      - logspace
      - meshgrid
      - tril
    - Array manipulation routines
      - reshape
      - ravel
      - ndarray.flatten
      - swapaxes
      - ndarray.T
      - transpose
      - moveaxis
      - expand_dims
      - squeeze
      - broadcast_to
      - broadcast_arrays
      - concatenate
      - stack
      - dstack
      - vstack
      - split
      - hsplit
      - vsplit
      - tile
      - repeat
      - unique
      - reshape
      - flip
      - roll
      - rot90
    - Input and output
      - genfromtxt
    - Linear algebra (numpy.linalg)
      - dot
      - vdot
      - inner
      - outer
      - tensordot
      - einsum
      - svd
      - norm
      - trace
    - Mathematical functions
      - sin
      - cos
      - tan
      - arcsin
      - arccos
      - arctan
      - degrees
      - radians
      - hypot
      - arctan2
      - deg2rad
      - rad2deg
      - sinh
      - cosh
      - tanh
      - arcsinh
      - arccosh
      - arctanh
      - rint
      - fix
      - floor
      - ceil
      - trunc
      - around
      - sum
      - prod
      - cumsum
      - exp
      - expm1
      - log
      - log10
      - log2
      - log1p
      - ldexp
      - lcm
      - add
      - reciprocal
      - negative
      - divide
      - power
      - subtract
      - mod
      - multiply
      - true_divide
      - remainder
      - clip
      - sqrt
      - cbrt
      - square
      - absolute
      - sign
      - maximum
      - minimum
    - np.random
      - choice
      - shuffle
      - normal
      - uniform
      - rand
      - randint
    - Sorting, searching, and counting
      - argmax
      - argmin
    - Statistics
      - min
      - max
      - mean
      - std
      - var
      - histogram
  - NP Extensions
    - set_np
    - reset_np
    - cpu
    - cpu_pinned
    - gpu
    - gpu_memory_info
    - current_context
    - num_gpus
    - activation
    - batch_norm
    - convolution
    - dropout
    - embedding
    - fully_connected
    - layer_norm
    - pooling
    - rnn
    - leaky_relu
    - multibox_detection
    - multibox_prior
    - multibox_target
    - roi_pooling
    - sigmoid
    - smooth_l1
    - softmax
    - mxnet.npx.threading
    - topk
    - waitall
    - load
    - save
    - one_hot
    - pick
    - reshape_like
    - batch_flatten
    - batch_dot
    - gamma
    - sequence_mask
- Gluon reference

Table Of Contents

Guide
- Getting started with NP on MXNet
- What is NP on MXNet
API
- NP on MXNet reference
  - Array objects
    - The N-dimensional array (ndarray)
    - Indexing
  - Routines
    - Array creation routines
      - eye
      - empty
      - full
      - identity
      - ones
      - ones_like
      - zeros
      - zeros_like
      - array
      - copy
      - arange
      - linspace
      - logspace
      - meshgrid
      - tril
    - Array manipulation routines
      - reshape
      - ravel
      - ndarray.flatten
      - swapaxes
      - ndarray.T
      - transpose
      - moveaxis
      - expand_dims
      - squeeze
      - broadcast_to
      - broadcast_arrays
      - concatenate
      - stack
      - dstack
      - vstack
      - split
      - hsplit
      - vsplit
      - tile
      - repeat
      - unique
      - reshape
      - flip
      - roll
      - rot90
    - Input and output
      - genfromtxt
    - Linear algebra (numpy.linalg)
      - dot
      - vdot
      - inner
      - outer
      - tensordot
      - einsum
      - svd
      - norm
      - trace
    - Mathematical functions
      - sin
      - cos
      - tan
      - arcsin
      - arccos
      - arctan
      - degrees
      - radians
      - hypot
      - arctan2
      - deg2rad
      - rad2deg
      - sinh
      - cosh
      - tanh
      - arcsinh
      - arccosh
      - arctanh
      - rint
      - fix
      - floor
      - ceil
      - trunc
      - around
      - sum
      - prod
      - cumsum
      - exp
      - expm1
      - log
      - log10
      - log2
      - log1p
      - ldexp
      - lcm
      - add
      - reciprocal
      - negative
      - divide
      - power
      - subtract
      - mod
      - multiply
      - true_divide
      - remainder
      - clip
      - sqrt
      - cbrt
      - square
      - absolute
      - sign
      - maximum
      - minimum
    - np.random
      - choice
      - shuffle
      - normal
      - uniform
      - rand
      - randint
    - Sorting, searching, and counting
      - argmax
      - argmin
    - Statistics
      - min
      - max
      - mean
      - std
      - var
      - histogram
  - NP Extensions
    - set_np
    - reset_np
    - cpu
    - cpu_pinned
    - gpu
    - gpu_memory_info
    - current_context
    - num_gpus
    - activation
    - batch_norm
    - convolution
    - dropout
    - embedding
    - fully_connected
    - layer_norm
    - pooling
    - rnn
    - leaky_relu
    - multibox_detection
    - multibox_prior
    - multibox_target
    - roi_pooling
    - sigmoid
    - smooth_l1
    - softmax
    - mxnet.npx.threading
    - topk
    - waitall
    - load
    - save
    - one_hot
    - pick
    - reshape_like
    - batch_flatten
    - batch_dot
    - gamma
    - sequence_mask
- Gluon reference

ParameterDict.zero_grad¶

ParameterDict.zero_grad(self)¶: Sets all Parameters’ gradient buffer to 0.

ParameterDict.reset_ctx

ParameterDict.prefix