The full procedural name indicates architecture, extended name, tile size, and layout.

Save a model checkpoint.

Init method based on N(0, sigma/sqrt(2*num_layers).

Sets the predicate mask, overriding value stored in predicate iterator

Reads a tab separated value file.

Computes the accuracy over the k top predictions for the specified values of k

Adds a pointer offset in units of Element

Adds a tile offset

Adds a tile offset

Compute the number of contiguous elements needed to store a tensor with the given size

Remove new lines and multiple spaces and adjust end of sentence dot.

Clears the predicate set efficiently

Streams the result to global memory

Clears the predicate set efficiently

Returns a pointer

Retrieves a single item from the dataset with the option to only return a portion of the item. get(idx) is the same as [idx] but get

Maps a CUTLASS numeric type to a cuDNN cudnnDataType_t

Simple linear layer with weight initialization.

Calculates optimal swizzle width

Gets the mask

Gets the mask

Call this method to modify your model to take advantage of sparse matrix multiplication. Note that this call alone only augments the model wit

Loads a fragment from memory with additional logical offset

Loads a fragment from memory with additional logical offset

Loads a fragment from memory with additional logical offset

Loads a fragment from memory with additional logical offset

Loads a fragment from memory with additional logical offset

Loads a fragment from memory with additional logical offset

Note that this call will sync across all ranks.

Overrides the internal iteration index

Sets the predicate mask, overriding value stored in predicate iterator

Called after accumulators have been exchanged for each accumulator vector

Test Conv2d Operation based on configurations

Make sure the input variable is not None.

Adds a pointer offset in units of Element

Helper to invoke the output functor over each vector of output

Clears the predicate set efficiently

Clears the predicate set efficiently

Computes the determinant of a 2-by-2 matrix

Clears the predicate set efficiently

Clears the predicate set efficiently

Clears the predicate set efficiently

Forward step for passed-in model. If first stage, input tensor is obtained from data_iterator, otherwise passed-in input_tensor is used.

Instantiate a PreTrainedBertModel from a pre-trained model file. Download and cache the pre-trained model file if needed.

Returns a pointer

Get a yacs CfgNode object with default values.

Build language model and return along with the key to save.

Gets the mask

Gets the mask

Return my rank for the pipeline model parallel group.