pytorch/botorch

Issues

• Load Custom Pre-Trained PyTorch MLP as Surrogate Model

  #2566 opened 2 months ago by vinkyc
  4

• [Bug] `KroneckerMultiTaskGP` incompatible with `batch_cross_validation` or batched models

  #2553 opened 2 months ago by slishak-PX
  2

• [Bug] Unable to fit Heteroskedastic GP with input warping

  #2551 opened 2 months ago by SaiAakash
  5

• [Bug] Parameter Constraints Do Not Work

  #2542 opened 4 months ago by kayween
  9

• [Feature Request] Enable type checkers

  #2569 opened 3 months ago by AdrianSosic
  4

• Modifying the vae-mnist example to incorporate SVDKL model from gpytorch.

  #2567 opened 3 months ago by utkarshp1161
  0

• [Question] How can I see the debugging logs from `fit.py` ?

  #2506 opened 3 months ago by rashisht1
  1

• How to get mean and covariance matrix from GP model?

  #2559 opened 3 months ago by ricvolpi
  5

• [Bug] The fit_gpytorch_mll() will crash after a period of normal fit.

  #2545 opened 4 months ago by lixiangru123
  6

• [Bug] GP Output Standardization Without Specified `output_transform`

  #2547 opened 4 months ago by stefanpricopie
  1

• [Bug]Analytic Acquisition Functions not working with SingleTaskVariationalGP for multi-output data

  #2530 opened 4 months ago by SaiAakash
  2

• [Style] Use `typing_extensions.Self` pre-3.11 Python

  #2487 opened 4 months ago by eddiebergman
  2

• [Bug] Inputs are not transformed when using `GPyTorchModel.condition_on_observations`

  #2533 opened 4 months ago by mikkelbue
  2

• [Feature Request] T-batch acquisition optimization

  #2528 opened 4 months ago by yunshengtian
  2

• [Bug] Possibly inconsistent likelihood calls with input transforms

  #2515 opened 4 months ago by ruicoelhopedro
  1

• [Bug] Memory leak when using a DataLoader for SingleTaskVariationalGP

  #2526 opened 4 months ago by mikkelbue
  3

• [Bug] `AffineInputTransform` does not check dimensions after training

  #2509 opened 4 months ago by slishak-PX
  2

• [Question] Anyway to ensure non-duplicated inclusion of pending configs when wrapping acquisitions?

  #2493 opened 4 months ago by eddiebergman
  1

• Consistent use of torch device in ‎HitAndRunPolytopeSampler

  #2500 opened 4 months ago by mhrmsn
  2

• [Bug] Handler added to logger twice

  #2503 opened 4 months ago by slishak-PX
  1

• [Documentation] Thompson sampling notebook runs incorrect experiments

  #2483 opened 4 months ago by slishak-PX
  2

• [Feature Request] Sample Shape as Constructor Parameter in MCSamplerMixin and Children

  #2482 opened 5 months ago by julianStreibel
  1

• [Bug] `get_polytope_samples` fails for inequalities over `q`-batches and the actual dimension

  #2468 opened 5 months ago by AVHopp
  3

• Chained Objectives

  #2418 opened 6 months ago by kstone40
  3

• When to use model(X) vs. model.posterior(X) ?

  #2455 opened 5 months ago by lukas-frank
  2

• [Documentation] Documentation in botorch.models.gpytorch [Bug]

  #2446 opened 6 months ago by brunzema
  1

• [Bug] Documentation not rendering for classes that have no class docstring (but have an __init__ docstring)

  #2419 opened 6 months ago by esantorella
  2

• [Bug] _get_noiseless_fantasy_model for LogNoisyExpectedImprovement and NoisyExpectedImprovement only works for default GP with default Matern & no input or outcome transforms

  #2412 opened 6 months ago by 71c
  4

• Choosing between maximization & minimization with BoTorch acquisition functions

  #2439 opened 6 months ago by AdrianSosic
  1

• Documentation, especially on the botorch.org landing page, should reflect current best practices

  #2421 opened 6 months ago by esantorella
  0

• [Documentation] Incomplete documentation for implementing custom models

  #2427 opened 6 months ago by slishak-PX
  3

• [Bug] Inconsistent behavior: `standardize` vs. `Standardize` with n < 2

  #2394 opened 6 months ago by AdrianSosic
  9

• How to make qEHVI or qNEHVI work with MultitaskGP?

  #2416 opened 6 months ago by sheikhahnaf
  5

• GPyTorch version bump

  #2417 opened 6 months ago by apaleyes
  2

• [Bug] fit_gpytorch_mll gives backward pass runtime exception on second model fit attempt with HeteroskedasticSingleTaskGP

  #2370 opened 6 months ago by Tim-Infl
  12

• [Bug] Inconsistent Gradient Values for N-Dimensional Integral of Multidimensional Normal Distribution

  #2411 opened 6 months ago by simplifier89
  1

• [Request] Upgrade to GPytorch 1.12

  #2407 opened 6 months ago by M-Colley
  0

• `ModelListGP.fantasize` does not support `MultiTaskGP`s, but other `ModelListGP` methods do

  #2398 opened 6 months ago by esantorella
  1

• [Bug] optimize_acqf erroring out with SingleTaskMultiFidelityGP

  #2402 opened 6 months ago by esantorella
  1

• [Documentation] Inaccurate and incomplete documentation for `SingleTaskMultiFidelityGP`

  #2403 opened 6 months ago by esantorella
  0

• [Bug] Lengthscale change with and without normalization

  #2392 opened 7 months ago by Talheus
  2

• [Bug] Feasible method of PolytopeSampler class returns False in spite of the sample being feasible

  #2383 opened 7 months ago by dsanyal
  2

• [Bug] Unexpected `BotorchTensorDimensionWarning`

  #2376 opened 7 months ago by AdrianSosic
  4

• Can MultiTaskGP be trained with batch data?

  #2369 opened 7 months ago by ybwang75
  5

• [Bug] get_polytope_samples provides non-random samples when there is a fixed value constraint (x = 1)

  #2351 opened 7 months ago by Waschenbacher
  5

• [Feature Request] MultiTaskGP with multiple task features

  #2359 opened 7 months ago by AdrianSosic
  3

• [Bug] MultiTaskGP without data

  #2360 opened 7 months ago by AdrianSosic
  2

• [Bug] Input normalization makes GP fail to fit data

  #2340 opened 8 months ago by martinakaduc
  1

• Custom Objective/ Acquistion function

  #2317 opened 8 months ago by Yash-Pisat
  2

• qKnowledgeGradient CPU & GPU usage

  #2318 opened 8 months ago by ruhanaazam
  3