function

benchmarks.hota.train_learned.train_step

def train_step(propagator: Propagator, fuser: Fuser, clip: Clip, optimizer: torch.optim.Optimizer, temperature: float = TEMPERATURE) -> float

Run one contrastive multi-frame optimization step on a single clip.

Parameters

NameTypeDescription
propagatorPropagatorThe learned modules being trained.
fuserPropagatorThe learned modules being trained.
clipClipA list of frames ``[(embeddings (n_f, D), gt_ids (n_f,)), ...]`` for one clip in frame order. Detection embeddings are L2-normalized internally so training matches inference (where the Fuser emits unit norm and the cosine cost normalizes).
optimizertorch.optim.OptimizerAn optimizer over both modules' parameters.
temperature = TEMPERATUREfloatInfoNCE temperature; smaller demands a sharper margin.

Returns

float — The scalar loss before the gradient step (``0.0`` if the clip has no multi-frame GT track to roll out).

Source: unitrack/benchmarks/hota/train_learned.py:84