-
I couldn't find it with npm, so I used
`npm install --prefix ~/.webppl https://github.com/mhtess/webppl-csv`
-
(This is a meta-issue for discussing the overall architecture and goals of the new test infrastructure, making a roadmap, and for tracking progress on its constituent issues. I'll continue filling it…
-
Consider the [vectorized](https://github.com/hawkrobe/hbnn/blob/master/models/regression/hierarchical_vectorized.py) and [non-vectorized](https://github.com/hawkrobe/hbnn/blob/master/models/regression…
-
Along the lines of #16 we should follow [webPPL's example](https://github.com/probmods/webppl/blob/dev/tests/test-inference.js) closely and make the inference test interface more generic, automate tes…
-
Deep time-series models such as [DMM](https://arxiv.org/abs/1609.09869) and [DVBF](https://arxiv.org/abs/1605.06432) will be anchor models for the first release. We need to implement them in pyro, imp…
-
`KL_QP`-like algorithms currently have the args `model_fixed=False` and `guide_fixed=False` to control whether their parameters are fixed during inference (`KL_QP.step()`). This is restrictive for com…
-
Curious why all the params are auto-magically global with a shared namespace. Seems like a very different design choice than the nice cleanly scoped pytorch variables.
srush updated
6 years ago
-
On L2:C21 of `src/inference/dream/gradient.js`
`var ad = require('..../ad');`
is missing a forward slash.
-
The unit test on the recent dev branch failed because our tests for `TestNormalChol.test_mean_and_var` and `TestNormal.test_mean_and_var` were just outside the tolerance threshold. We can fix this by …
-
currently, in contexts where independence information can be useful (e.g. for doing elbo with non-reparameterizable rvs), our only construct for marking things as independent is `map_data`. so, for ex…