# tensorplay.autograd.gradcheck.gradgradcheck

Source: https://www.tensorplay.cn/docs/generated/tensorplay.autograd.gradcheck.gradgradcheck.html

```python
tensorplay.autograd.gradcheck.gradgradcheck(func, inputs, grad_outputs=None, *, eps: float = 1e-06, atol: float = 1e-05, rtol: float = 0.001, gen_non_contig_grad_outputs: bool = False, raise_exception: bool = True, nondet_tol: float = 0.0, check_undefined_grad: bool = True, check_grad_dtypes: bool = False, check_batched_grad: bool = False, check_fwd_over_rev: bool = False, check_rev_over_rev: bool = True, fast_mode: bool = False, masked: bool = False) → bool
```

Check gradients of gradients computed via small finite differences
against analytical gradients wrt tensors in inputs and
grad_outputs that are of floating point or complex type and with
requires_grad=True.

This function checks that backpropagating through the gradients computed
to the given grad_outputs are correct.

The check between numerical and analytical gradients uses allclose().

> **Note**
>
> The default values are designed for input and grad_outputs of double precision. This check will likely fail if they are of less precision, e.g., FloatTensor.

> **Warning**
>
> If any checked tensor in input and grad_outputs has overlapping memory, i.e., different indices pointing to the same memory address (e.g., from expand), this check will likely fail because the numerical gradients computed by point perturbation at such indices will change values at all other indices that share the same memory address.

Parameters:

- func (function) – a Python function that takes Tensor inputs and returns a Tensor or a tuple of Tensors

- inputs ([tuple](https://docs.python.org/3/builtins/stdtypes.html#tuple) of [Tensor](/docs/generated/tensorplay.Tensor.html#tensorplay.Tensor) or [Tensor](/docs/generated/tensorplay.Tensor.html#tensorplay.Tensor)) – inputs to the function

- grad_outputs ([tuple](https://docs.python.org/3/builtins/stdtypes.html#tuple) of [[Tensor](/docs/generated/tensorplay.Tensor.html#tensorplay.Tensor) or None] or [Tensor](/docs/generated/tensorplay.Tensor.html#tensorplay.Tensor), optional) – The gradients with respect to the function’s outputs.

- eps ([float](https://docs.python.org/3/builtins/functions.html#float), optional) – perturbation for finite differences

- atol ([float](https://docs.python.org/3/builtins/functions.html#float), optional) – absolute tolerance

- rtol ([float](https://docs.python.org/3/builtins/functions.html#float), optional) – relative tolerance

- gen_non_contig_grad_outputs ([bool](https://docs.python.org/3/builtins/functions.html#bool), optional) – if grad_outputs is None and gen_non_contig_grad_outputs is True, the randomly generated gradient outputs are made to be noncontiguous

- raise_exception ([bool](https://docs.python.org/3/builtins/functions.html#bool), optional) – indicating whether to raise an exception if the check fails. The exception gives more information about the exact nature of the failure. This is helpful when debugging gradchecks.

- nondet_tol ([float](https://docs.python.org/3/builtins/functions.html#float), optional) – tolerance for non-determinism. When running identical inputs through the differentiation, the results must either match exactly (default, 0.0) or be within this tolerance. Note that a small amount of nondeterminism in the gradient will lead to larger inaccuracies in the second derivative.

- check_undefined_grad ([bool](https://docs.python.org/3/builtins/functions.html#bool), optional) – if True, check if undefined output grads are supported and treated as zeros

- check_batched_grad ([bool](https://docs.python.org/3/builtins/functions.html#bool), optional) – if True, check if we can compute batched gradients using tensorplay.vmap(). Defaults to False.

- check_fwd_over_rev ([bool](https://docs.python.org/3/builtins/functions.html#bool), optional) – if True, check the second derivative with forward-mode AD over the first-order backward. Defaults to False.

- check_rev_over_rev ([bool](https://docs.python.org/3/builtins/functions.html#bool), optional) – if True, check the second derivative with reverse-mode AD over the first-order backward. Defaults to True.

- fast_mode ([bool](https://docs.python.org/3/builtins/functions.html#bool), optional) – if True, run a faster implementation of gradgradcheck that no longer computes the entire jacobian.

- masked ([bool](https://docs.python.org/3/builtins/functions.html#bool), optional) – if True, the gradients of unspecified elements of sparse tensors are ignored (default, False).

Returns:

True if all differences satisfy allclose condition