# Latest **URL:** https://discuss.bayesflow.org/latest.md [Latest](https://discuss.bayesflow.org/latest.md) · [Categories](https://discuss.bayesflow.org/categories.md) · [Tags](https://discuss.bayesflow.org/tags.md) --- ## [Welcome to BayesFlow Forums! 👋](https://discuss.bayesflow.org/t/welcome-to-bayesflow-forums/5)
**Author:** [@system](https://discuss.bayesflow.org/u/system)\ **Replies:** 0\ **Last updated:** [November 30, 2023, 11:44pm UTC](https://discuss.bayesflow.org/t/welcome-to-bayesflow-forums/5 "2023-11-30T23:44:13Z")
We are so glad you joined us. BayesFlow Forums The BayesFlow Forums provide a community for asking and answering questions about all aspects of BayesFlow and amortized Bayesian workflows. Here are some things you can… --- ## [Cross attention for conditioning image generation](https://discuss.bayesflow.org/t/cross-attention-for-conditioning-image-generation/280)
**Author:** [@corte](https://discuss.bayesflow.org/u/corte)\ **Replies:** 1\ **Last updated:** [September 8, 2026, 4:34pm UTC](https://discuss.bayesflow.org/t/cross-attention-for-conditioning-image-generation/280 "2026-09-08T16:34:07Z")
Hi, I have been working in a project in which I have to predict an image from an input with shape (timesteps, quantity). “timesteps” has a size of about 2000, while “quantity” about 500. In order to compress the input,… --- ## [Model portability issue across machines](https://discuss.bayesflow.org/t/model-portability-issue-across-machines/279)
**Author:** [@alenzi](https://discuss.bayesflow.org/u/alenzi)\ **Replies:** 2\ **Last updated:** [August 20, 2026, 10:17am UTC](https://discuss.bayesflow.org/t/model-portability-issue-across-machines/279 "2026-08-20T10:17:53Z")
Hello, Together with a collaborator, we are using BayesFlow for amortised posterior estimation and have run into a model portability issue that we have not been able to resolve. We would be very grateful for any guidanc… --- ## [Training instability with spline coupling flows in BayesFlow 2.0.7](https://discuss.bayesflow.org/t/training-instability-with-spline-coupling-flows-in-bayesflow-2-0-7/278)
**Author:** [@Simon](https://discuss.bayesflow.org/u/Simon)\ **Replies:** 3\ **Last updated:** [July 29, 2026, 3:20pm UTC](https://discuss.bayesflow.org/t/training-instability-with-spline-coupling-flows-in-bayesflow-2-0-7/278 "2026-07-29T15:20:49Z")
Hi, I am using BayesFlow 2.0.7 with the PyTorch backend and have repeatedly encountered training instabilities with spline coupling flows. Training initially proceeds normally, but after a variable number of epochs, the… --- ## [What is the recommended Bayesian workflow in BayesFlow when we have many candidate parameters but little prior knowledge about which ones are truly important?](https://discuss.bayesflow.org/t/what-is-the-recommended-bayesian-workflow-in-bayesflow-when-we-have-many-candidate-parameters-but-little-prior-knowledge-about-which-ones-are-truly-important/276)
**Author:** [@vivian20021116](https://discuss.bayesflow.org/u/vivian20021116)\ **Replies:** 3\ **Last updated:** [July 28, 2026, 4:18pm UTC](https://discuss.bayesflow.org/t/what-is-the-recommended-bayesian-workflow-in-bayesflow-when-we-have-many-candidate-parameters-but-little-prior-knowledge-about-which-ones-are-truly-important/276 "2026-07-28T16:18:10Z")
Hi, Suppose we begin with little prior knowledge and therefore assign priors to all candidate coefficients. After observing data, what is currently considered the best Bayesian strategy in BayesFlow for identifying whi… --- ## [Parameter-dependent prior](https://discuss.bayesflow.org/t/parameter-dependent-prior/275)
**Author:** [@MasterODF](https://discuss.bayesflow.org/u/MasterODF)\ **Replies:** 6\ **Last updated:** [July 16, 2026, 7:30am UTC](https://discuss.bayesflow.org/t/parameter-dependent-prior/275 "2026-07-16T07:30:05Z")
Hello! I’m working with BayesFlow on a birth-death process experiment, but I’ve hit a technical problem I can’t quite figure out. My model uses a modified Gaussian prior that is not entirely independent. Specifically, i… --- ## [Discussion of Normalizing flow](https://discuss.bayesflow.org/t/discussion-of-normalizing-flow/96)
**Author:** [@Jice](https://discuss.bayesflow.org/u/Jice)\ **Replies:** 7\ **Last updated:** [May 26, 2026, 2:20pm UTC](https://discuss.bayesflow.org/t/discussion-of-normalizing-flow/96 "2026-05-26T14:20:53Z")
Hi, I have applied the BayesFlow framework to my area, and the performance has been fairly good so far. The mechanism of BayesFlow relies on the use of normalizing flow (NF). Here, I would like to discuss some aspects o… --- ## [Poor parameter recovery for hybrid reinforcement learning models in a two step task](https://discuss.bayesflow.org/t/poor-parameter-recovery-for-hybrid-reinforcement-learning-models-in-a-two-step-task/252)
**Author:** [@pavgreen](https://discuss.bayesflow.org/u/pavgreen)\ **Replies:** 3\ **Last updated:** [May 12, 2026, 4:52pm UTC](https://discuss.bayesflow.org/t/poor-parameter-recovery-for-hybrid-reinforcement-learning-models-in-a-two-step-task/252 "2026-05-12T16:52:48Z")
Hello, I am currently attempting to use Bayesflow (v1) for a hybrid reinforcement learning model for a two-step RL task but the parameter recovery is quite poor. To briefly describe the model, we assume that first stag… --- ## [Keras ValueError re-appearing in newest package version?](https://discuss.bayesflow.org/t/keras-valueerror-re-appearing-in-newest-package-version/249)
**Author:** [@floralynth](https://discuss.bayesflow.org/u/floralynth)\ **Replies:** 8\ **Last updated:** [May 6, 2026, 4:04am UTC](https://discuss.bayesflow.org/t/keras-valueerror-re-appearing-in-newest-package-version/249 "2026-05-06T04:04:10Z")
Hi all, I seem to have bumped into an issue previously seen as resolved: As discussed in that post, using SetTransformer or DeepSet results in an error once workflow.fit() is called: ValueError: For a \`build()\` metho… --- ## [Hierarchical modelling](https://discuss.bayesflow.org/t/hierarchical-modelling/248)
**Author:** [@Will](https://discuss.bayesflow.org/u/Will)\ **Replies:** 3\ **Last updated:** [May 5, 2026, 7:32pm UTC](https://discuss.bayesflow.org/t/hierarchical-modelling/248 "2026-05-05T19:32:37Z")
Hello! Thanks to your advice and help I’ve now got models with good diagnostics up and running which has been great. While these models are good I always intended to use a hierarchical model due to the low trial count i… --- ## [Use of JAX-based generative models - explicit RNG management](https://discuss.bayesflow.org/t/use-of-jax-based-generative-models-explicit-rng-management/241)
**Author:** [@floralynth](https://discuss.bayesflow.org/u/floralynth)\ **Replies:** 2\ **Last updated:** [May 5, 2026, 6:25am UTC](https://discuss.bayesflow.org/t/use-of-jax-based-generative-models-explicit-rng-management/241 "2026-05-05T06:25:11Z")
Hi all, First time here and aiming to use Bayesflow to speed up some existing MCMC based workflows for cognitive models. I already have a series of models written using JAX-numpy and numpyro which can be used for likel… --- ## [Can BayesFlow diffusion models infer high-dimensional sparse binary parameters?](https://discuss.bayesflow.org/t/can-bayesflow-diffusion-models-infer-high-dimensional-sparse-binary-parameters/240)
**Author:** [@Jice](https://discuss.bayesflow.org/u/Jice)\ **Replies:** 5\ **Last updated:** [April 30, 2026, 5:37pm UTC](https://discuss.bayesflow.org/t/can-bayesflow-diffusion-models-infer-high-dimensional-sparse-binary-parameters/240 "2026-04-30T17:37:21Z")
Hi everyone, I am applying BayesFlow to a posterior inference problem of the form p(\\theta|y), where y is the observed data and \\theta is the target parameter. In my case, \\theta is high-dimensional and binary. For exa… --- ## [Any help appreciated!](https://discuss.bayesflow.org/t/any-help-appreciated/237)
**Author:** [@Will](https://discuss.bayesflow.org/u/Will)\ **Replies:** 5\ **Last updated:** [April 15, 2026, 9:57pm UTC](https://discuss.bayesflow.org/t/any-help-appreciated/237 "2026-04-15T21:57:50Z")
Hello! I’m in the process of trying to make a nice basic 3 parameter DDM and I hope to be able to nail the basic version so I can branch out in the future. I had some initial success but there were some flaws in the logi… --- ## [ONNX Probabilistic Programming Working Group – Collaboration with the BayesFlow Community](https://discuss.bayesflow.org/t/onnx-probabilistic-programming-working-group-collaboration-with-the-bayesflow-community/238)
**Author:** [@bparbhu](https://discuss.bayesflow.org/u/bparbhu)\ **Replies:** 0\ **Last updated:** [April 15, 2026, 12:16pm UTC](https://discuss.bayesflow.org/t/onnx-probabilistic-programming-working-group-collaboration-with-the-bayesflow-community/238 "2026-04-15T12:16:18Z")
Hi everyone, I wanted to share that a new ONNX Working Group focused on Probabilistic Programming has recently been formed and invite participation from the BayesFlow community. The goal of this working group is to bri… --- ## [Calibration Loss](https://discuss.bayesflow.org/t/calibration-loss/232)
**Author:** [@matthias.kloft](https://discuss.bayesflow.org/u/matthias.kloft)\ **Replies:** 2\ **Last updated:** [March 3, 2026, 12:47pm UTC](https://discuss.bayesflow.org/t/calibration-loss/232 "2026-03-03T12:47:48Z")
Hello fellow Bayesians, were/are there any attempts towards integrating a calibration loss, e.g., as proposed in \[2310.13402\] Calibrating Neural Simulation-Based Inference with Differentiable Coverage Probability , into… --- ## [Multimodal data with multiple summary nets possible?](https://discuss.bayesflow.org/t/multimodal-data-with-multiple-summary-nets-possible/231)
**Author:** [@Simon](https://discuss.bayesflow.org/u/Simon)\ **Replies:** 2\ **Last updated:** [February 12, 2026, 8:57am UTC](https://discuss.bayesflow.org/t/multimodal-data-with-multiple-summary-nets-possible/231 "2026-02-12T08:57:57Z")
Hi BayesFlow Team, I’m currently training an NPE model on time series data using the TimeSeriesNetwork as my summary network. In addition to the time series, I have a second, low-dimensional data modality. What I would… --- ## [Release 2.0.8 - The Dawn of Diffusion Models](https://discuss.bayesflow.org/t/release-2-0-8-the-dawn-of-diffusion-models/230)
**Author:** [@KLDivergence](https://discuss.bayesflow.org/u/KLDivergence)\ **Replies:** 0\ **Last updated:** [January 29, 2026, 3:16pm UTC](https://discuss.bayesflow.org/t/release-2-0-8-the-dawn-of-diffusion-models/230 "2026-01-29T15:16:32Z")
Release notes :sparkles: The next BayesFlow release adds substantial new functionality, largely centered on diffusion models (score-based, flow-matching, consistency) for simulation-based inference (see \[2512.20685\] Dif… --- ## [BayesFlow incompatible with keras==3.13.1](https://discuss.bayesflow.org/t/bayesflow-incompatible-with-keras-3-13-1/229)
**Author:** [@nbrawand](https://discuss.bayesflow.org/u/nbrawand)\ **Replies:** 2\ **Last updated:** [January 29, 2026, 3:15pm UTC](https://discuss.bayesflow.org/t/bayesflow-incompatible-with-keras-3-13-1/229 "2026-01-29T15:15:17Z")
Hey BF Team, I am a first time user of BF and I followed the install section on the readme and tried to run the first example but it failed with a Keras error: python3.11/site-packages/keras/src/layers/core/dense.py", … --- ## [units=None set in EquivariantLayer()?](https://discuss.bayesflow.org/t/units-none-set-in-equivariantlayer/220)
**Author:** [@m-schuettler](https://discuss.bayesflow.org/u/m-schuettler)\ **Replies:** 5\ **Last updated:** [January 13, 2026, 3:45pm UTC](https://discuss.bayesflow.org/t/units-none-set-in-equivariantlayer/220 "2026-01-13T15:45:58Z")
Hi everyone! I’m new to BayesFlow and new-ish to ML in general, so maybe this is a stupid question. I’m trying to run someone else’s code, but I’m always getting an error message when trying to run EquivariantLayer(). T… --- ## [Multilevel models](https://discuss.bayesflow.org/t/multilevel-models/210)
**Author:** [@Username1](https://discuss.bayesflow.org/u/Username1)\ **Replies:** 3\ **Last updated:** [November 21, 2025, 7:39pm UTC](https://discuss.bayesflow.org/t/multilevel-models/210 "2025-11-21T19:39:32Z")
Hi, I have recently been reading up on the amortized bayesian inference litterature and find it an interesting fit to my nonlinear mixed effects model. I was reading your paper “amortized bayesian multilevel models“ an… --- ## [Different ways of conditioned on data](https://discuss.bayesflow.org/t/different-ways-of-conditioned-on-data/212)
**Author:** [@Jice](https://discuss.bayesflow.org/u/Jice)\ **Replies:** 1\ **Last updated:** [November 21, 2025, 7:34pm UTC](https://discuss.bayesflow.org/t/different-ways-of-conditioned-on-data/212 "2025-11-21T19:34:42Z")
Hi everyone, I have been using BayesFlow 1.0 and 2.0 in my work, and they have performed very well for my studies. Recently, I started revisiting how generative models actually incorporate conditioning on data/observati… --- ## [Enabling multi‐GPU (2× GPUs) training in BayesFlow](https://discuss.bayesflow.org/t/enabling-multi-gpu-2x-gpus-training-in-bayesflow/171)
**Author:** [@Noura](https://discuss.bayesflow.org/u/Noura)\ **Replies:** 7\ **Last updated:** [November 21, 2025, 7:30pm UTC](https://discuss.bayesflow.org/t/enabling-multi-gpu-2x-gpus-training-in-bayesflow/171 "2025-11-21T19:30:16Z")
Hi BayesFlow team, I’m running BayesFlow v1’s train\_online on a single GPU but would like to leverage two GPUs (for both simulation-heavy and neural network phases). Does the Trainer offer built in data, or model paral… --- ## [Additional training of an already-trained network](https://discuss.bayesflow.org/t/additional-training-of-an-already-trained-network/214)
**Author:** [@ErikStuchly](https://discuss.bayesflow.org/u/ErikStuchly)\ **Replies:** 1\ **Last updated:** [November 21, 2025, 7:28pm UTC](https://discuss.bayesflow.org/t/additional-training-of-an-already-trained-network/214 "2025-11-21T19:28:40Z")
Hi there, I was wondering whether it is possible to load a network that has been trained online for, say, 100 epochs, and give it additional training over a few more epochs (for instance, because the loss function has n… --- ## [Dependency conflict and NumPy 2.0 support](https://discuss.bayesflow.org/t/dependency-conflict-and-numpy-2-0-support/209)
**Author:** [@freund](https://discuss.bayesflow.org/u/freund)\ **Replies:** 1\ **Last updated:** [November 6, 2025, 8:43pm UTC](https://discuss.bayesflow.org/t/dependency-conflict-and-numpy-2-0-support/209 "2025-11-06T20:43:47Z")
Hello, I’m running into an installation conflict and wanted to ask about potential plans for numpy \>=2.0 support. I have a project where I’d like to use both bayesflow and a recent version of another package, ssm-simul… --- ## [Model comparison for models with different configurators and summary networks (stable-legacy version)](https://discuss.bayesflow.org/t/model-comparison-for-models-with-different-configurators-and-summary-networks-stable-legacy-version/207)
**Author:** [@Robin](https://discuss.bayesflow.org/u/Robin)\ **Replies:** 4\ **Last updated:** [November 6, 2025, 1:40pm UTC](https://discuss.bayesflow.org/t/model-comparison-for-models-with-different-configurators-and-summary-networks-stable-legacy-version/207 "2025-11-06T13:40:52Z")
Hi, I’d like to compare two models to determine which fits my empirical data best. Model 1: A standard drift diffusion model (DDM) with four parameters. Model 2: An extended DDM in which some parameters fluctuate … --- ## [Keras ValueError: For a \`build()\` method with more than one argument, all arguments should have a \`\_shape\` suffix and match an argument from \`call()\`](https://discuss.bayesflow.org/t/keras-valueerror-for-a-build-method-with-more-than-one-argument-all-arguments-should-have-a-shape-suffix-and-match-an-argument-from-call/206)
**Author:** [@freund](https://discuss.bayesflow.org/u/freund)\ **Replies:** 10\ **Last updated:** [November 6, 2025, 7:41am UTC](https://discuss.bayesflow.org/t/keras-valueerror-for-a-build-method-with-more-than-one-argument-all-arguments-should-have-a-shape-suffix-and-match-an-argument-from-call/206 "2025-11-06T07:41:45Z")
Hello, I’m encountering this error when attempting to run the minimal working example on current github readme. For troubleshooting, I’ve attempted bayesflow 2.0.7 and 2.0.6, jax GPU, jax CPU, tensorflow CPU backends, … --- ## [What does \`epoch\` mean when using an online data generating process?](https://discuss.bayesflow.org/t/what-does-epoch-mean-when-using-an-online-data-generating-process/201)
**Author:** [@benjamin](https://discuss.bayesflow.org/u/benjamin)\ **Replies:** 3\ **Last updated:** [October 28, 2025, 4:27pm UTC](https://discuss.bayesflow.org/t/what-does-epoch-mean-when-using-an-online-data-generating-process/201 "2025-10-28T16:27:46Z")
Disclaimer: I am just getting started with learning BayesFlow and I am following the Linear regression example here. The example uses BasicWorkflow.fit\_online with epochs=50, batch\_size=64, and num\_batches\_per\_epoch=200.… --- ## [Difficulty estimating multilevel ordinal model](https://discuss.bayesflow.org/t/difficulty-estimating-multilevel-ordinal-model/200)
**Author:** [@CourtneyBHilton](https://discuss.bayesflow.org/u/CourtneyBHilton)\ **Replies:** 4\ **Last updated:** [October 6, 2025, 11:59am UTC](https://discuss.bayesflow.org/t/difficulty-estimating-multilevel-ordinal-model/200 "2025-10-06T11:59:41Z")
Hi all, First time posting in the forum so want to start by saying thanks for the amazing tool! I’ve already used it for one project where it enabled some really cool things we wouldn’t have been able to do otherwise. I… --- ## [Question about ordered model output and context variables](https://discuss.bayesflow.org/t/question-about-ordered-model-output-and-context-variables/198)
**Author:** [@deakj](https://discuss.bayesflow.org/u/deakj)\ **Replies:** 11\ **Last updated:** [September 16, 2025, 3:37pm UTC](https://discuss.bayesflow.org/t/question-about-ordered-model-output-and-context-variables/198 "2025-09-16T15:37:39Z")
Hi all, Thanks in advance for any help. This may be long. My set up is quite similar to the one defined here Fitting a value-based decision model with trial-level context variables . The model I am using predicts fixat… --- ## [Fitting a value-based decision model with trial-level context variables](https://discuss.bayesflow.org/t/fitting-a-value-based-decision-model-with-trial-level-context-variables/194)
**Author:** [@ErikStuchly](https://discuss.bayesflow.org/u/ErikStuchly)\ **Replies:** 4\ **Last updated:** [September 15, 2025, 9:33am UTC](https://discuss.bayesflow.org/t/fitting-a-value-based-decision-model-with-trial-level-context-variables/194 "2025-09-15T09:33:13Z")
Hi there, I am very new to amortised bayesian inference, so apologies in advance if my questions seem uninformed or contain some misconceptions! Anyway, I am hoping to create and fit a decision-making model, not quite … [Next page](https://discuss.bayesflow.org/latest.md?page=1)