https://github.com/GPflow/GPflow
Revision f5b94dd14d551e781f24d266f3beb2bc3f5dae36 authored by James Hensman on 26 October 2016, 19:10:25 UTC, committed by James Hensman on 26 October 2016, 19:10:25 UTC
The Autoflow function now makes its own graph for each autoflow in the model. Some extra tests in test_AutoFlow are added to make sure that it works with dataholders. The dataHolder class has changed a little: it's included in make_tf_array now, so that the placeholders can be made inside the Autoflow graph: before, all placeholders were on the default graph.
1 parent b811d4f
Tip revision: f5b94dd14d551e781f24d266f3beb2bc3f5dae36 authored by James Hensman on 26 October 2016, 19:10:25 UTC
Keep AutoFlow computations off the default graph
Keep AutoFlow computations off the default graph
Tip revision: f5b94dd
| File | Mode | Size |
|---|---|---|
| tfops | ||
| __init__.py | -rw-r--r-- | 826 bytes |
| _settings.py | -rw-r--r-- | 3.9 KB |
| _version.py | -rw-r--r-- | 600 bytes |
| conditionals.py | -rw-r--r-- | 5.1 KB |
| densities.py | -rw-r--r-- | 2.7 KB |
| gpflowrc | -rw-r--r-- | 218 bytes |
| gpmc.py | -rw-r--r-- | 3.2 KB |
| gpr.py | -rw-r--r-- | 3.1 KB |
| hmc.py | -rw-r--r-- | 4.6 KB |
| kernels.py | -rw-r--r-- | 15.2 KB |
| kullback_leiblers.py | -rw-r--r-- | 4.8 KB |
| likelihoods.py | -rw-r--r-- | 17.7 KB |
| mean_functions.py | -rw-r--r-- | 3.0 KB |
| model.py | -rwxr-xr-x | 14.6 KB |
| param.py | -rwxr-xr-x | 31.7 KB |
| priors.py | -rw-r--r-- | 3.2 KB |
| scoping.py | -rw-r--r-- | 1.1 KB |
| sgpmc.py | -rw-r--r-- | 3.1 KB |
| sgpr.py | -rw-r--r-- | 10.2 KB |
| svgp.py | -rw-r--r-- | 5.7 KB |
| tf_hacks.py | -rw-r--r-- | 1.2 KB |
| tf_wraps.py | -rw-r--r-- | 1.4 KB |
| transforms.py | -rw-r--r-- | 7.8 KB |
| vgp.py | -rwxr-xr-x | 5.7 KB |
Computing file changes ...
