Issues
- 1
Importing DistributedSamplerWrapper will invalidate the setting CUDA_VISIBLE_DEVICE.
#1451 opened by zezhishao - 2
Bug in catalyst/callbacks/backward.py if the grad_clip_fn value is set.
#1444 opened by AleksandrMinin - 2
- 6
KWArg fp16 does not exist
#1440 opened by ppenlite - 1
Replace getters with properties
#1441 opened by YodaEmbedding - 2
- 1
Bug in catalyst/callbacks/backward.py if the grad_clip_fn value is set.
#1446 opened by AleksandrMinin - 1
Bug in catalyst/callbacks/backward.py if the grad_clip_fn value is set.
#1445 opened by AleksandrMinin - 4
Multi Criterion Training
#1435 opened by GirinChutia - 3
No utils.initialization file
#1436 opened by Klins101 - 2
utils.process_model_params
#1434 opened by Chris-Ran - 3
Crashes on 2xT4 GPUs
#1433 opened by Philmod - 1
DynamicBalanceClassSampler example does not use sampler
#1438 opened by I8dNLo - 1
Custom loader stages
#1432 opened by YodaEmbedding - 2
Idea: offload all launch-related code Accelerate
#1430 opened by tadejsv - 2
- 3
comprehensive classification example
#1421 opened by ashkanaev - 4
- 3
- 2
- 10
Tensorboard Logger; Log weights and distributions.
#1408 opened by alfieroddan - 5
- 4
Quick start 101 seems to need additional dependency
#1415 opened by filipmu - 2
imread( grayscale: bool=True) gives error
#1412 opened by Saharsh1005 - 2
- 2
- 3
MultilabelPrecisionRecallF1SupportCallback do not working
#1391 opened by Inkorak - 1
Monotonically increasing total epoch counter
#1393 opened by YodaEmbedding - 2
Segmentation metric callback outputs don't contain per-class values in current version
#1392 opened by chekhovana - 3
Metrics calculation is going much slower compared with version 21.07 in semantic segmentation task
#1351 opened by penguinflys - 1
ENUMs for inner catalyst convention parameters
#1315 opened by Nimrais - 3
- 3
A question in CV - MNIST segmentation example
#1389 opened by fjd330621 - 13
WandB doesn't work with DDP
#1316 opened by ivan-chai - 3
Unscale gradients before gradient clipping
#1319 opened by ivan-chai - 5
Multi-node support to Catalyst distributed training
#1276 opened by smos1 - 5
Wrong loader length if overfit=True
#1301 opened by Nimrais - 3
FocalLossMultiClass does not work on TPU's
#1347 opened by SupreethRao99 - 20
WandB batch metrics logging error
#1290 opened by ivan-chai - 4
Best checkpoint can't be loaded
#1328 opened by ivan-chai - 8
ControlFlowCallback error in DDP
#1317 opened by ivan-chai - 2
Zero seed error
#1320 opened by ivan-chai - 3
Does Catalyst use eval() during evaluation
#1323 opened by MathieuHaller - 6
Support for Comet Logger
#1281 opened by DN6 - 3
`runner.batch_metrics` UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach()
#1222 opened by km4sh - 6
Incorrect md5 sums in Imagenette and Imagewoof from `catalyst.contrib.datasets`
#1265 opened by NetherQuartz - 7
- 2
Segmentation tutorial
#1227 opened by sayakpaul - 2
TracingCallback fails in ConfigAPI
#1228 opened by y-ksenia - 3
Accessing model/Runner inside catalyst.dl.Logger
#1233 opened by benihime91