VI using deprecated Aesara MRG sampler #4523

twiecki · 2021-03-10T06:48:11Z

The current usage of the mrng sampler in ADVI does not support JAX (pymc-devs/pytensor#322). It should be fairly easy to instead make it use the new RandomVariable Op pymc-devs/pytensor#296 to get JAX support for ADVI.

CC @ferrine

The text was updated successfully, but these errors were encountered:

ferrine · 2021-03-13T19:11:28Z

@brandonwillard should I use the master branch or V4 for development?

brandonwillard · 2021-03-13T19:17:11Z

We've been putting in PRs to the v4 branch.

ferrine · 2021-03-13T19:18:56Z

So yes, I'll open a PR to v4 branch:tada:

brandonwillard · 2021-03-13T19:20:59Z

Oh, wait, the change requested here could go toward v3 or v4.

I would prioritize v4, but we need to port the VI code to v4 first and foremost, and that work is independent of this exact issue. Both could be done simultaneously, though.

ferrine · 2021-03-14T07:45:27Z

@brandonwillard, running pymc3/tests/test_variational_inference.py I have errors unrelataed to VI, is it expected?

ERROR pymc3/tests/test_variational_inference.py::test_init_groups[raises0-grouping0]
ERROR pymc3/tests/test_variational_inference.py::test_init_groups[raises1-grouping1]
ERROR pymc3/tests/test_variational_inference.py::test_init_groups[raises2-grouping2]
ERROR pymc3/tests/test_variational_inference.py::test_init_groups[raises3-grouping3]
ERROR pymc3/tests/test_variational_inference.py::test_init_groups[raises4-grouping4]
ERROR pymc3/tests/test_variational_inference.py::test_init_groups[raises5-grouping5]
ERROR pymc3/tests/test_variational_inference.py::test_sample_aevb[MeanFieldGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_sample_aevb[FullRankGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_sample_aevb[NormalizingFlowGroup: {'flow': 'scale'}]
ERROR pymc3/tests/test_variational_inference.py::test_sample_aevb[NormalizingFlowGroup: {'flow': 'loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_sample_aevb[NormalizingFlowGroup: {'flow': 'hh'}]
ERROR pymc3/tests/test_variational_inference.py::test_sample_aevb[NormalizingFlowGroup: {'flow': 'planar'}]
ERROR pymc3/tests/test_variational_inference.py::test_sample_aevb[NormalizingFlowGroup: {'flow': 'radial'}]
ERROR pymc3/tests/test_variational_inference.py::test_sample_aevb[NormalizingFlowGroup: {'flow': 'radial-loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_replacements_in_sample_node_aevb[MeanFieldGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_replacements_in_sample_node_aevb[FullRankGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_replacements_in_sample_node_aevb[NormalizingFlowGroup: {'flow': 'scale'}]
ERROR pymc3/tests/test_variational_inference.py::test_replacements_in_sample_node_aevb[NormalizingFlowGroup: {'flow': 'loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_replacements_in_sample_node_aevb[NormalizingFlowGroup: {'flow': 'hh'}]
ERROR pymc3/tests/test_variational_inference.py::test_replacements_in_sample_node_aevb[NormalizingFlowGroup: {'flow': 'planar'}]
ERROR pymc3/tests/test_variational_inference.py::test_replacements_in_sample_node_aevb[NormalizingFlowGroup: {'flow': 'radial'}]
ERROR pymc3/tests/test_variational_inference.py::test_replacements_in_sample_node_aevb[NormalizingFlowGroup: {'flow': 'radial-loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_1_sample_1_var[MeanFieldGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_1_sample_1_var[FullRankGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_1_sample_1_var[NormalizingFlowGroup: {'flow': 'scale'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_1_sample_1_var[NormalizingFlowGroup: {'flow': 'loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_1_sample_1_var[NormalizingFlowGroup: {'flow': 'hh'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_1_sample_1_var[NormalizingFlowGroup: {'flow': 'planar'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_1_sample_1_var[NormalizingFlowGroup: {'flow': 'radial'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_1_sample_1_var[NormalizingFlowGroup: {'flow': 'radial-loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_2_sample_2_var[MeanFieldGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_2_sample_2_var[FullRankGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_2_sample_2_var[NormalizingFlowGroup: {'flow': 'scale'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_2_sample_2_var[NormalizingFlowGroup: {'flow': 'loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_2_sample_2_var[NormalizingFlowGroup: {'flow': 'hh'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_2_sample_2_var[NormalizingFlowGroup: {'flow': 'planar'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_2_sample_2_var[NormalizingFlowGroup: {'flow': 'radial'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_2_sample_2_var[NormalizingFlowGroup: {'flow': 'radial-loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_sample_aevb[MeanFieldGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_sample_aevb[FullRankGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_sample_aevb[NormalizingFlowGroup: {'flow': 'scale'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_sample_aevb[NormalizingFlowGroup: {'flow': 'loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_sample_aevb[NormalizingFlowGroup: {'flow': 'hh'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_sample_aevb[NormalizingFlowGroup: {'flow': 'planar'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_sample_aevb[NormalizingFlowGroup: {'flow': 'radial'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_mini_sample_aevb[NormalizingFlowGroup: {'flow': 'radial-loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_aevb[MeanFieldGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_aevb[FullRankGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_aevb[NormalizingFlowGroup: {'flow': 'scale'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_aevb[NormalizingFlowGroup: {'flow': 'loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_aevb[NormalizingFlowGroup: {'flow': 'hh'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_aevb[NormalizingFlowGroup: {'flow': 'planar'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_aevb[NormalizingFlowGroup: {'flow': 'radial'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_aevb[NormalizingFlowGroup: {'flow': 'radial-loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_logq_globals[MeanFieldGroup: None]
ERROR pymc3/tests/test_variational_inference.py::test_logq_globals[FullRankGroup: None, MeanFieldGroup: ['one']]
ERROR pymc3/tests/test_variational_inference.py::test_logq_globals[MeanFieldGroup: ['one'], FullRankGroup: ['two'], NormalizingFlowGroup: ['three']]
ERROR pymc3/tests/test_variational_inference.py::test_logq_globals[MeanFieldGroup: ['one'], FullRankGroup: ['two', 'three']]
ERROR pymc3/tests/test_variational_inference.py::test_logq_globals[MeanFieldGroup: ['one'], EmpiricalGroup: ['two', 'three']]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises0-mean_field-MeanFieldGroup-kw0]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises1-mf-MeanFieldGroup-kw1]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises2-full_rank-FullRankGroup-kw2]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises3-fr-FullRankGroup-kw3]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises4-FR-FullRankGroup-kw4]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises5-loc-NormalizingFlowGroup-kw5]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises6-scale-NormalizingFlowGroup-kw6]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises7-hh-NormalizingFlowGroup-kw7]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises8-planar-NormalizingFlowGroup-kw8]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises9-radial-NormalizingFlowGroup-kw9]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises10-scale-loc-NormalizingFlowGroup-kw10]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises11-empirical-EmpiricalGroup-kw11]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_vfam[raises12-empirical-EmpiricalGroup-kw12]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_params[raises0-params0-MeanFieldGroup-kw0-None]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_params[raises1-params1-FullRankGroup-kw1-None]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_params[raises2-params2-NormalizingFlowGroup-kw2-loc]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_params[raises3-params3-NormalizingFlowGroup-kw3-scale]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_params[raises4-params4-NormalizingFlowGroup-kw4-hh]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_params[raises5-params5-NormalizingFlowGroup-kw5-planar]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_params[raises6-params6-NormalizingFlowGroup-kw6-radial]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_params[raises7-params7-NormalizingFlowGroup-kw7-scale-loc]
ERROR pymc3/tests/test_variational_inference.py::test_group_api_params[raises8-params8-EmpiricalGroup-kw8-None]
ERROR pymc3/tests/test_variational_inference.py::test_single_group_shortcuts[MeanFieldGroup-MeanField-kw0]
ERROR pymc3/tests/test_variational_inference.py::test_single_group_shortcuts[FullRankGroup-FullRank-kw1]
ERROR pymc3/tests/test_variational_inference.py::test_single_group_shortcuts[EmpiricalGroup-Empirical-kw2]
ERROR pymc3/tests/test_variational_inference.py::test_single_group_shortcuts[NormalizingFlowGroup-NormalizingFlow-kw3]
ERROR pymc3/tests/test_variational_inference.py::test_single_group_shortcuts[NormalizingFlowGroup-NormalizingFlow-kw4]
ERROR pymc3/tests/test_variational_inference.py::test_single_group_shortcuts[NormalizingFlowGroup-NormalizingFlow-kw5]
ERROR pymc3/tests/test_variational_inference.py::test_fit_oo[NFVI=scale-loc-mini]
ERROR pymc3/tests/test_variational_inference.py::test_profile[NFVI=scale-loc-mini]
ERROR pymc3/tests/test_variational_inference.py::test_fit_oo[NFVI=scale-loc-full]
ERROR pymc3/tests/test_variational_inference.py::test_profile[NFVI=scale-loc-full]
ERROR pymc3/tests/test_variational_inference.py::test_fit_oo[ADVI-full] - Att...
ERROR pymc3/tests/test_variational_inference.py::test_profile[ADVI-full] - At...
ERROR pymc3/tests/test_variational_inference.py::test_fit_oo[ADVI-mini] - Att...
ERROR pymc3/tests/test_variational_inference.py::test_profile[ADVI-mini] - At...
ERROR pymc3/tests/test_variational_inference.py::test_aevb[ADVI] - Deprecatio...
ERROR pymc3/tests/test_variational_inference.py::test_replacements[ADVI] - Ty...
ERROR pymc3/tests/test_variational_inference.py::test_sample_replacements[ADVI]
ERROR pymc3/tests/test_variational_inference.py::test_fit_oo[FullRankADVI-full]
ERROR pymc3/tests/test_variational_inference.py::test_profile[FullRankADVI-full]
ERROR pymc3/tests/test_variational_inference.py::test_fit_oo[FullRankADVI-mini]
ERROR pymc3/tests/test_variational_inference.py::test_profile[FullRankADVI-mini]
ERROR pymc3/tests/test_variational_inference.py::test_aevb[FullRankADVI] - De...
ERROR pymc3/tests/test_variational_inference.py::test_replacements[FullRankADVI]
ERROR pymc3/tests/test_variational_inference.py::test_sample_replacements[FullRankADVI]
ERROR pymc3/tests/test_variational_inference.py::test_fit_oo[SVGD-full] - Att...
ERROR pymc3/tests/test_variational_inference.py::test_profile[SVGD-full] - At...
ERROR pymc3/tests/test_variational_inference.py::test_fit_oo[SVGD-mini] - Att...
ERROR pymc3/tests/test_variational_inference.py::test_profile[SVGD-mini] - At...
ERROR pymc3/tests/test_variational_inference.py::test_aevb[SVGD] - Deprecatio...
ERROR pymc3/tests/test_variational_inference.py::test_replacements[SVGD] - Ty...
ERROR pymc3/tests/test_variational_inference.py::test_sample_replacements[SVGD]
ERROR pymc3/tests/test_variational_inference.py::test_fit_oo[ASVGD-full] - At...
ERROR pymc3/tests/test_variational_inference.py::test_profile[ASVGD-full] - A...
ERROR pymc3/tests/test_variational_inference.py::test_fit_oo[ASVGD-mini] - At...
ERROR pymc3/tests/test_variational_inference.py::test_profile[ASVGD-mini] - A...
ERROR pymc3/tests/test_variational_inference.py::test_aevb[ASVGD] - Deprecati...
ERROR pymc3/tests/test_variational_inference.py::test_replacements[ASVGD] - T...
ERROR pymc3/tests/test_variational_inference.py::test_sample_replacements[ASVGD]
ERROR pymc3/tests/test_variational_inference.py::test_aevb[NFVI=scale-loc] - ...
ERROR pymc3/tests/test_variational_inference.py::test_replacements[NFVI=scale-loc]
ERROR pymc3/tests/test_variational_inference.py::test_sample_replacements[NFVI=scale-loc]
ERROR pymc3/tests/test_variational_inference.py::test_rowwise_approx[MeanFieldGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_rowwise_approx[FullRankGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_rowwise_approx[NormalizingFlowGroup: {'flow': 'scale'}]
ERROR pymc3/tests/test_variational_inference.py::test_rowwise_approx[NormalizingFlowGroup: {'flow': 'loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_rowwise_approx[NormalizingFlowGroup: {'flow': 'hh'}]
ERROR pymc3/tests/test_variational_inference.py::test_rowwise_approx[NormalizingFlowGroup: {'flow': 'planar'}]
ERROR pymc3/tests/test_variational_inference.py::test_rowwise_approx[NormalizingFlowGroup: {'flow': 'radial'}]
ERROR pymc3/tests/test_variational_inference.py::test_rowwise_approx[NormalizingFlowGroup: {'flow': 'radial-loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx[MeanFieldGroup: None]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx[FullRankGroup: None, MeanFieldGroup: ['one']]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx[MeanFieldGroup: ['one'], FullRankGroup: ['two'], NormalizingFlowGroup: ['three']]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx[MeanFieldGroup: ['one'], FullRankGroup: ['two', 'three']]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx[MeanFieldGroup: ['one'], EmpiricalGroup: ['two', 'three']]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_single_group - D...
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx_aevb[MeanFieldGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx_aevb[FullRankGroup: {}]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx_aevb[NormalizingFlowGroup: {'flow': 'scale'}]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx_aevb[NormalizingFlowGroup: {'flow': 'loc'}]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx_aevb[NormalizingFlowGroup: {'flow': 'hh'}]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx_aevb[NormalizingFlowGroup: {'flow': 'planar'}]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx_aevb[NormalizingFlowGroup: {'flow': 'radial'}]
ERROR pymc3/tests/test_variational_inference.py::test_pickle_approx_aevb[NormalizingFlowGroup: {'flow': 'radial-loc'}]

ferrine · 2021-03-14T07:46:50Z

I have aesara v2.0.2 there

ferrine · 2021-03-14T07:49:45Z

using aesara master results in the same

ferrine · 2021-03-14T07:51:31Z

I have errors like this

self = <pymc3.step_methods.metropolis.Metropolis object at 0x7f44b6c0efd0>, vars = [x], S = None, proposal_dist = None, scaling = 1.0, tune = True, tune_interval = 100
model = <pymc3.model.Model object at 0x7f44b64c5430>, mode = None, kwargs = {}

    def __init__(
        self,
        vars=None,
        S=None,
        proposal_dist=None,
        scaling=1.0,
        tune=True,
        tune_interval=100,
        model=None,
        mode=None,
        **kwargs
    ):
        """Create an instance of a Metropolis stepper

        Parameters
        ----------
        vars: list
            List of variables for sampler
        S: standard deviation or covariance matrix
            Some measure of variance to parameterize proposal distribution
        proposal_dist: function
            Function that returns zero-mean deviates when parameterized with
            S (and n). Defaults to normal.
        scaling: scalar or array
            Initial scale factor for proposal. Defaults to 1.
        tune: bool
            Flag for tuning. Defaults to True.
        tune_interval: int
            The frequency of tuning. Defaults to 100 iterations.
        model: PyMC Model
            Optional model for sampling step. Defaults to None (taken from context).
        mode: string or `Mode` instance.
            compilation mode passed to Aesara functions
        """

        model = pm.modelcontext(model)

        if vars is None:
            vars = model.vars
        vars = pm.inputvars(vars)

        if S is None:
            # XXX: This needs to be refactored
            S = None  # np.ones(sum(v.dsize for v in vars))

        if proposal_dist is not None:
            self.proposal_dist = proposal_dist(S)
>       elif S.ndim == 1:
E       AttributeError: 'NoneType' object has no attribute 'ndim'

brandonwillard · 2021-03-14T14:22:15Z

Yes, you should expect all sorts of errors.

fonnesbeck · 2021-07-02T14:27:22Z

Does any of this change with the Numba backend?

brandonwillard · 2021-07-28T17:55:42Z

Does any of this change with the Numba backend?

It shouldn't; in general, if we use Aesara, its backends should handle everything without requiring backend-specific logic at this level.

bukson · 2022-08-26T14:15:43Z

It is not working with pymc 4:

import pymc

with pm.Model():
    x = pm.Normal("x")
    pm.fit()

WARNING (aesara.tensor.blas): Using NumPy C-API based implementation for BLAS functions.
Traceback (most recent call last):
File "/usr/local/lib/python3.10/site-packages/IPython/core/interactiveshell.py", line 3398, in run_code
exec(code_obj, self.user_global_ns, self.user_ns)
File "", line 5, in <cell line: 3>
pm.fit()
File "/usr/local/lib/python3.10/site-packages/pymc/variational/inference.py", line 744, in fit
return inference.fit(n, **kwargs)
File "/usr/local/lib/python3.10/site-packages/pymc/variational/inference.py", line 138, in fit
step_func = self.objective.step_function(score=score, **kwargs)
File "/usr/local/lib/python3.10/site-packages/aesara/configparser.py", line 47, in res
return f(*args, **kwargs)
File "/usr/local/lib/python3.10/site-packages/pymc/variational/opvi.py", line 355, in step_function
updates = self.updates(
File "/usr/local/lib/python3.10/site-packages/pymc/variational/opvi.py", line 244, in updates
self.add_obj_updates(
File "/usr/local/lib/python3.10/site-packages/pymc/variational/opvi.py", line 289, in add_obj_updates
obj_target = self(
File "/usr/local/lib/python3.10/site-packages/aesara/configparser.py", line 47, in res
return f(*args, **kwargs)
File "/usr/local/lib/python3.10/site-packages/pymc/variational/opvi.py", line 407, in call
a = self.approx.set_size_and_deterministic(a, nmc, 0, kwargs.get("more_replacements"))
File "/usr/local/lib/python3.10/site-packages/aesara/configparser.py", line 47, in res
return f(*args, **kwargs)
File "/usr/local/lib/python3.10/site-packages/pymc/variational/opvi.py", line 1359, in set_size_and_deterministic
flat2rand = self.make_size_and_deterministic_replacements(s, d, more_replacements)
File "/usr/local/lib/python3.10/site-packages/pymc/variational/opvi.py", line 1333, in make_size_and_deterministic_replacements
flat2rand.update(g.make_size_and_deterministic_replacements(s, d, more_replacements))
File "/usr/local/lib/python3.10/site-packages/pymc/variational/opvi.py", line 1067, in make_size_and_deterministic_replacements
initial = self._new_initial(s, d, more_replacements)
File "/usr/local/lib/python3.10/site-packages/pymc/variational/opvi.py", line 978, in _new_initial
return getattr(self._rng, dist_name)(size=shape)
File "/usr/local/lib/python3.10/site-packages/aesara/sandbox/rng_mrg.py", line 1184, in normal
uniform = self.uniform(
File "/usr/local/lib/python3.10/site-packages/aesara/sandbox/rng_mrg.py", line 914, in uniform
rstates = self.get_substream_rstates(nstreams, dtype)
File "/usr/local/lib/python3.10/site-packages/aesara/configparser.py", line 47, in res
return f(*args, **kwargs)
File "/usr/local/lib/python3.10/site-packages/aesara/sandbox/rng_mrg.py", line 818, in get_substream_rstates
multMatVect(rval[0], A1p72, M1, A2p72, M2)
File "/usr/local/lib/python3.10/site-packages/aesara/sandbox/rng_mrg.py", line 66, in multMatVect
multMatVect.dot_modulo = function(
File "/usr/local/lib/python3.10/site-packages/aesara/compile/function/init.py", line 317, in function
fn = pfunc(
File "/usr/local/lib/python3.10/site-packages/aesara/compile/function/pfunc.py", line 374, in pfunc
return orig_function(
File "/usr/local/lib/python3.10/site-packages/aesara/compile/function/types.py", line 1763, in orig_function
fn = m.create(defaults)
File "/usr/local/lib/python3.10/site-packages/aesara/compile/function/types.py", line 1656, in create
_fn, _i, _o = self.linker.make_thunk(
File "/usr/local/lib/python3.10/site-packages/aesara/link/basic.py", line 254, in make_thunk
return self.make_all(
File "/usr/local/lib/python3.10/site-packages/aesara/link/basic.py", line 697, in make_all
thunks, nodes, jit_fn = self.create_jitable_thunk(
File "/usr/local/lib/python3.10/site-packages/aesara/link/basic.py", line 646, in create_jitable_thunk
converted_fgraph = self.fgraph_convert(
File "/usr/local/lib/python3.10/site-packages/aesara/link/jax/linker.py", line 13, in fgraph_convert
return jax_funcify(fgraph, **kwargs)
File "/usr/local/lib/python3.10/functools.py", line 889, in wrapper
return dispatch(args[0].class)(*args, **kw)
File "/usr/local/lib/python3.10/site-packages/aesara/link/jax/dispatch.py", line 670, in jax_funcify_FunctionGraph
return fgraph_to_python(
File "/usr/local/lib/python3.10/site-packages/aesara/link/utils.py", line 741, in fgraph_to_python
compiled_func = op_conversion_fn(
File "/usr/local/lib/python3.10/functools.py", line 889, in wrapper
return dispatch(args[0].class)(*args, **kw)
File "/usr/local/lib/python3.10/site-packages/aesara/link/jax/dispatch.py", line 143, in jax_funcify
raise NotImplementedError(f"No JAX conversion for the given Op: {op}")
NotImplementedError: No JAX conversion for the given Op: DotModulo

ricardoV94 · 2022-08-26T14:39:36Z

@bukson please open an issue on our discourse at: https://discourse.pymc.io/

twiecki · 2022-09-15T08:42:41Z

No this is a legit issue, we should port VI to use random variable: aesara-devs/aesara#322 (comment). CC @ferrine

Closes #4523 Closes #6277

twiecki added the jax label Mar 10, 2021

ferrine added the v4 label Mar 14, 2021

brandonwillard added the help wanted label Mar 24, 2021

ferrine mentioned this issue Mar 29, 2021

Make VI work on v4 #4582

Merged

twiecki assigned ferrine Apr 5, 2021

twiecki removed the help wanted label Apr 5, 2021

ricardoV94 modified the milestones: vNext (4.0.0), v4.0.1 Jun 27, 2021

ricardoV94 removed the jax label Nov 18, 2021

michaelosthege modified the milestones: v4.0.0b2, v4.0.0b3 Jan 7, 2022

ricardoV94 modified the milestones: v4.0.0b3, v4.0.0 Feb 7, 2022

ricardoV94 closed this as completed Mar 7, 2022

twiecki reopened this Sep 15, 2022

ricardoV94 changed the title ~~Port VI to use RandomVariable~~ VI fails in JAX mode Nov 5, 2022

ricardoV94 added the VI Variational Inference label Nov 5, 2022

ricardoV94 modified the milestones: v4.0.0, v4.4.0 Nov 5, 2022

ricardoV94 pinned this issue Nov 5, 2022

ricardoV94 changed the title ~~VI fails in JAX mode~~ VI using deprecated Aesara MRG sampler Nov 7, 2022

This was referenced Nov 7, 2022

Use default Aesara RandomStream for VI #6273

Closed

Use RandomVariables for Minibatch #6277

Closed

ricardoV94 removed this from the v4.4.0 milestone Nov 9, 2022

ferrine mentioned this issue Nov 15, 2022

Refactor MiniBatch and stop using deprecated MRG sampler #6304

Merged

michaelosthege pushed a commit that referenced this issue Dec 12, 2022

Refactor Minibatch and stop using MRG_RandomStream in VI

82dfbf6

Closes #4523 Closes #6277

michaelosthege closed this as completed in #6304 Dec 12, 2022

michaelosthege pushed a commit that referenced this issue Dec 12, 2022

Refactor Minibatch and stop using MRG_RandomStream in VI

adf6fff

Closes #4523 Closes #6277

ricardoV94 unpinned this issue Dec 14, 2022

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

VI using deprecated Aesara MRG sampler #4523

VI using deprecated Aesara MRG sampler #4523

twiecki commented Mar 10, 2021

ferrine commented Mar 13, 2021

brandonwillard commented Mar 13, 2021

ferrine commented Mar 13, 2021

brandonwillard commented Mar 13, 2021 •

edited

Loading

ferrine commented Mar 14, 2021

ferrine commented Mar 14, 2021

ferrine commented Mar 14, 2021

ferrine commented Mar 14, 2021

brandonwillard commented Mar 14, 2021

fonnesbeck commented Jul 2, 2021

brandonwillard commented Jul 28, 2021

bukson commented Aug 26, 2022 •

edited

Loading

ricardoV94 commented Aug 26, 2022

twiecki commented Sep 15, 2022 •

edited

Loading

VI using deprecated Aesara MRG sampler #4523

VI using deprecated Aesara MRG sampler #4523

Comments

twiecki commented Mar 10, 2021

ferrine commented Mar 13, 2021

brandonwillard commented Mar 13, 2021

ferrine commented Mar 13, 2021

brandonwillard commented Mar 13, 2021 • edited Loading

ferrine commented Mar 14, 2021

ferrine commented Mar 14, 2021

ferrine commented Mar 14, 2021

ferrine commented Mar 14, 2021

brandonwillard commented Mar 14, 2021

fonnesbeck commented Jul 2, 2021

brandonwillard commented Jul 28, 2021

bukson commented Aug 26, 2022 • edited Loading

ricardoV94 commented Aug 26, 2022

twiecki commented Sep 15, 2022 • edited Loading

brandonwillard commented Mar 13, 2021 •

edited

Loading

bukson commented Aug 26, 2022 •

edited

Loading

twiecki commented Sep 15, 2022 •

edited

Loading