Complete reductions if full pipeline is not selected or overlap is not allowed.

given a query and all galleries, compute its ap and cmc

3x3 convolution with padding

Converts a module's immediate parameters and buffers to dtype.

Computes cosine distance. Args: input1 (torch.Tensor): 2-D feature matrix. input2 (torch.Tensor): 2-D feature matrix. Return

Create a temporary table and index it to speed up repeated SQL quesries. The table is an INNER JOIN of CUPTI_ACTIVITY_KIND_MARKER with itself.

Demangle a C++ string

detach_hidden()

Downloads file from a url to a destination. Args: url (str): url to download file. dst (str): destination path.

Given process id (pid) and thread id (tid), return the object id. object id = pid (little endian 4 bytes) + tid (little endian 8 bytes)

Args: x: pytorch Variable, with shape [m, d] y: pytorch Variable, with shape [n, d] Returns: dist: pytorch Variable, with s

Computes euclidean squared distance. Args: input1 (torch.Tensor): 2-D feature matrix. input2 (torch.Tensor): 2-D feature matrix.

Evaluation with cuhk03 metric Key: one image for each gallery identity is randomly sampled for each query identity. Random sampling is perform

Evaluation with market1501 metric Key: for each query identity, its gallery images from the same camera view are discarded.

Given the correlation id, get CPU start, end, thread id, process id. The information can be in the runtime table or the driver table.

Get GPU kernel info

This function first finds all NVTX markers encapsulating a runtime / driver kernel launch. It then splits the markers into many lists. layer

Get the profile start time

Returns a shorter kernel name

Returns the number of datasets included. Each tuple in data contains (img_path(s), pid, camid, dsetid).

gets params of reparameterization based on known attribute names

self.img('input_img',t.Tensor(64,64))

Turning a batch of images to a grid e.g. input shape: (36, 64, 64) Will be a grid of 6x6, each grid is an image size

Turning a batch of images to a grid e.g. input shape: (36, 64, 64) Will be a grid of 6x6, each grid is an image size

init_inference()

Loads a state_dict created by an earlier call to state_dict(). If ``fp16_optimizer_instance`` was constructed from some ``init_optim

Compute knowledge-distillation (KD) loss given [scores] and [target_scores]. Both [scores] and [target_scores] should be tensors, although [t