a  d@sddlZddlmZmZmZmZmZmZmZddl Z ddl m Z ddl m Z ddl mZddlmZmZddlmZdZee ee fe d d d Ze e d d d Ze e d ddZe e d ddZe e d ddZeed ddZeeeefd ddZd2e eee dddZd3e eee dddZ d4e ee dd d!Z!e e d d"d#Z"eed$d%d&Z#d5e eee d'd(d)Z$d6e eeee j%e d*d+d,Z&e e d d-d.Z'e e ed/d0d1Z(dS)7N)AnyDictListOptionalSequenceTupleUnion)apply_to_collection)Tensor)TorchMetricsUserWarning)_TORCH_GREATER_EQUAL_1_12_XLA_AVAILABLE)rank_zero_warngư>)xreturncCs8t|tjr|Sdd|D}|s*tdtj|ddS)z'Concatenation along the zero dimension.cSs0g|](}|dkr(|jdkr(|dn|qS)r)numelndim unsqueeze).0yrd/var/www/html/stable-diffusion-webui/venv/lib/python3.9/site-packages/torchmetrics/utilities/data.py z dim_zero_cat..zNo samples to concatenaterdim) isinstancetorchr ValueErrorcatrrrr dim_zero_cats  r"cCstj|ddS)z#Summation along the zero dimension.rr)rsumr!rrr dim_zero_sum&sr$cCstj|ddS)z!Average along the zero dimension.rr)rmeanr!rrr dim_zero_mean+sr&cCstj|ddjS)zMax along the zero dimension.rr)rmaxvaluesr!rrr dim_zero_max0sr)cCstj|ddjS)zMin along the zero dimension.rr)rminr(r!rrr dim_zero_min5sr+cCsdd|DS)z&Flatten list of list into single list.cSsg|]}|D]}|q qSrr)rZsublistitemrrrr<rz_flatten..rr!rrr_flatten:sr-cCshi}d}|D]N\}}t|trJ|D]\}}||vr>d}|||<q*q||vrVd}|||<q||fS)zYFlatten dict of dicts into single dict and checking for duplicates in keys along the way.FT)itemsrdict)rZnew_dict duplicateskeyvaluekvrrr _flatten_dict?s   r5) label_tensor num_classesrcCsr|dur t|d}tj|jd|g|jddR|j|jd}| d |}| d|dS)aConvert a dense label tensor to one-hot format. Args: label_tensor: dense label tensor, with shape [N, d1, d2, ...] num_classes: number of classes C Returns: A sparse label tensor with shape [N, C, d1, d2, ...] Example: >>> x = torch.tensor([1, 2, 3]) >>> to_onehot(x) tensor([[0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]]) Nrr)dtypedevice?) intr'detachr,rzerosshaper8r9longrZ expand_asZscatter_)r6r7Z tensor_onehotindexrrr to_onehotPs rAr) prob_tensortopkrrcCsNt|}|dkr,|||j|ddd}n|||j||djd}|S)awConvert a probability tensor to binary by selecting top-k the highest entries. Args: prob_tensor: dense tensor of shape ``[..., C, ...]``, where ``C`` is in the position defined by the ``dim`` argument topk: number of the highest entries to turn into 1s dim: dimension on which to compare entries Returns: A binary tensor of the same shape as the input tensor of type ``torch.int32`` Example: >>> x = torch.tensor([[1.1, 2.0, 3.0], [2.0, 1.0, 0.5]]) >>> select_topk(x, topk=2) tensor([[0, 1, 1], [1, 1, 0]], dtype=torch.int32) rT)rZkeepdimr:)r3r)rZ zeros_likeZscatterargmaxrCindicesr;)rBrCrr=Z topk_tensorrrr select_topkss  rF)r argmax_dimrcCstj||dS)awConvert a tensor of probabilities to a dense label tensor. Args: x: probabilities to get the categorical label [N, d1, d2, ...] argmax_dim: dimension to apply Return: A tensor with categorical labels [N, d2, ...] Example: >>> x = torch.tensor([[0.2, 0.5], [0.9, 0.1]]) >>> to_categorical(x) tensor([1, 0]) r)rrD)rrGrrrto_categoricalsrHcCs|dkr|S|S)Nr)rZsqueezer!rrr_squeeze_scalar_element_tensorsrI)datarcCs t|ttS)N)r r rI)rJrrr_squeeze_if_scalarsrK)r minlengthrcCsp|durtt|}ts,ts,trb|jrbtj||jtj d}t |D]}||k ||<qH|Stj ||dS)a Implement custom bincount. PyTorch currently does not support ``torch.bincount`` for: - deterministic mode on GPU. - MPS devices This implementation fallback to a for-loop counting occurrences in that case. Args: x: tensor to count minlength: minimum length to count Returns: Number of occurrences for each unique element in x Example: >>> x = torch.tensor([0,0,0,1,1,2,2,2,2]) >>> _bincount(x, minlength=3) tensor([3, 2, 4]) N)r9r8rL) lenrunique$are_deterministic_algorithms_enabledr r Zis_mpsr=r9r?ranger#Zbincount)rrLoutputirrr _bincounts rT)rrr8rcCsPtr@|jr@|r@tjdkr@tdt|j ||d Stj |||dS)Nwin32zYou are trying to use a metric in deterministic mode on GPU that uses `torch.cumsum`, which is currently not supported. The tensor will be copied to the CPU memory to compute it and then copied back to GPU. Expect some slowdowns.)rr8) rrPZis_cudaZis_floating_pointsysplatformrr cpuZcumsumZcuda)rrr8rrr_cumsums rYcCs4||}t|}t|t|dd}||S)zSimilar to `_bincount`, but works also with tensor that do not contain continuous values. Args: x: tensor to count Returns: Number of occurrences for each unique element in x rrM)r*rrOrTr')rZunique_xrRrrr_flexible_bincounts  rZ)tensor1tensor2rcCs&|j|jkr|j|jd}t||S)z:Wrap torch.allclose to be robust towards dtype difference.)r8)r8torallclose)r[r\rrrr^s r^)N)rr)r)N)rN))rVtypingrrrrrrrrZlightning_utilitiesr r Z!torchmetrics.utilities.exceptionsr Ztorchmetrics.utilities.importsr r Ztorchmetrics.utilities.printsrZ METRIC_EPSr"r$r&r)r+listr-boolr5r;rArFrHrIrKrTr8rYrZr^rrrrs:$      #!