Add missing frontend functions to compile popular libraries

[xoiga123]

Below is the list of missing frontend functions needed to compile popular libraries, ranked from high to low priority. This list should be updated once per month.

Torch

• torch.nn.functional.scaled_dot_product_attention
  torch.histc #14474
  torch.Tensor.stride
  torch.Tensor.scatter_add_
  torch.nn.functional.grid_sample
  torch.Tensor.scatter_
  torch.Tensor.scatter_reduce_
  torch.nn.functional.adaptive_max_pool2d
  torch.Tensor.triu_
  torch.fft.rfft #22107
  torch.Tensor.as_subclass
  torch.Tensor.fill_diagonal_
  torch.nn.functional.dropout2d
  torch.grid_sampler
  torch.fft.irfft
  torch.Tensor.zero_
  torch.linalg.lu_factor_ex
  torch.linalg.lu_solve
  torch.diag_embed
  torch.Tensor.clamp_min
  torch.Tensor.masked_scatter_
  torch.Tensor.new
  torch.Tensor.bernoulli_
  torch.bucketize
  torch.Tensor.uniform_
  torch.sparse_coo_tensor
  torch.Tensor.coalesce
  torch.Tensor.index_put
  torch.linalg.norm
  torch.Tensor.split_with_sizes
  torch.Tensor.erf
  torch.Tensor.erf #27788
  torch.fft.fft #22110
  torch.Tensor.masked_scatter
  torch.nn.functional.adaptive_max_pool3d
  torch.Tensor.erfinv_
  torch.Tensor.unique
  torch.nn.functional.max_pool3d
  torch.Tensor.copy_
  torch.nn.functional.adaptive_max_pool1d
  torch.Tensor.to_sparse_csc
  torch.Tensor.frac #27406
  torch.Tensor.index_copy_
  torch.Tensor.indices
  torch.Tensor.log_softmax
  torch.Tensor.masked_select
  torch.cdist
  torch.Tensor.scatter
  torch.Tensor.values
  torch.Tensor.to_sparse_csr
  torch.histc #14474
  torch.nn.functional.scaled_dot_product_attention #18492
  torch.Tensor.as_strided #16267
  torch.Tensor.var #16269
  torch.nn.functional.multi_head_attention_forward #17447
  torch.nn.functional.affine_grid #17382
  torch.Tensor.narrow #16273
  torch.Tensor.log1p #16997
  torch.Tensor.floor_ #16999
  torch.Tensor.baddbmm #17001
  torch.Tensor.gather #17005
  torch.Tensor.addcdiv_ #17195
  torch.Tensor.diag #17077
  Tensorflow

• tensorflow.scatter_nd #17466
  Jax

Numpy

• numpy.ndarray.sum #19208
  numpy.asanyarray
  numpy.gradient #27737
  numpy.dot
  numpy.ndarray.dot
  numpy.ascontiguousarray
  numpy.max #27392
  numpy.min #27387
  numpy.pad #22194
  numpy.arctan2
  numpy.asfortranarray
  numpy.round
  numpy.apply_over_axes
  numpy.einsum #21398
  numpy.hypot
  numpy.histogram
  numpy.setdiff1d
  numpy.apply_along_axis
  numpy.take #27460
  numpy.isin #27409
  numpy.choose #22777
  numpy.ndarray.setflags #27414
  numpy.ndarray.trace
  numpy.insert #27415
  numpy.roots #22147

• numpy.column_stack #27755
  numpy.in1d

• numpy.ndarray.dot #19382
  numpy.dot #18122

[20481A05K0]

Hello there
Once check that there is no transpose method available in Numpy Framwork

[20481A05K0]

Can I open an issue for this?

[Anuhya30110]

Okay i will check

[20481A05K0]

• numpy.roots #22147

[AnnaTz]

@alvin-98 please take care to name your issues correctly. If the name of the issue is not exactly the same as the item on the list, the issue will not get linked to the ToDo. E.g. your issues for numpy's min and max should be named "numpy.min", "numpy.max".

[alvin-98]

@AnnaTz thanks for letting me know. I've fixed the issue names now. Will follow the naming convention going forward.