-
- [x] test local with MongoDB Atlas
- [x] Error checking for the connection
- [x] test prod with MongoDB Atlas
- [ ] Routine clearing of collections
-
We removed subset functionality because it did not fit into the NIST testing framework, and because it is not straightforward in the TA2TA3 API. However, this needs to be brought back into the main r…
-
- To run django channels in docker, the daphne server is needed
- An interim method would be to run the dev server
-
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
### Bug General Description
Wallet inconsistently handles authentication with complex passwords (loss of access to funds)
### Description:
Af…
-
The dsbox.template.runtime.Runtime.produce() method executes all primitives of the pipeline graph. However, some steps should not be ran for test datasets, i.e. datasets without target column. The Ext…
-
When testing on the dataset (e.g. No.38), it seems "d3m.primitives.sklearn_wrap.SKLinearSVC" primitive will cost a very very long time to fit the model and produce the results. This error is especiall…
-
Like this
`
{
"name": "extract_target_step",
"primitives": [{
"primitive": "d3m.primitives.data.ExtractColumnsBySemanticTypes",
"h…
-
```
Process Process-1:
Traceback (most recent call last):
File "/anaconda3/envs/d3m-devel-710/lib/python3.6/multiprocessing/process.py", line 258, in _bootstrap
self.run()
File "/anaconda…
-
We've found a few issues that have only been present when the server is proxied. We should add an NGINX instance to our internal and D3M test environments to better replicate deployment conditions.
-
(adjustment for Thurs. 11pm config updates)
https://datadrivendiscovery.org/wiki/pages/viewpage.action?pageId=11275766
### directory layout
```
├── search_config.json
└── TRAIN
└── data…