-
I deployed a Vespa cluster in Azure Kubernetes with HTTP/2 enabled using the latest Vespa build. My setup is similar to this one: https://github.com/vespa-engine/sample-apps/tree/master/basic-search-o…
-
- think the data are short cadence but fpp.ini has `cadence = 0.02`
- where do star.ini priors on Teff/logg/feh come from? we may want to use only the Teff and feh priors from spectroscopy
-
I opened this feature-request on the basis of this [slack thread](https://vespatalk.slack.com/archives/C01QNBPPNT1/p1630055029072900) and @jobergum 's [comment](https://vespatalk.slack.com/archives/C0…
-
### Motivation
For ONNX models we would like to be able to control number of intra and inter threads, and possible execution mode on a *per model* basis. The motivation behind this feature request …
-
Given the following query:
```
{
"ranking": {
"profile": "top_news"
},
"hits": 0,
"select": "all(group(group_doc_id) max(15) order(-avg(relevance())) each(output(count…
-
as used by @qgroom Quentin Groom and Christina Preda in https://github.com/qgroom/Vespa-velutina .
-
**Describe the bug**
I noticed that the throughput was great towards the beginning of the feed session but as it got to the tail of the feed, the speed very heavily declined. For some reference here …
-
@jrhillae (als trekker), @SanderDevisscher (als ondersteuning) -- Dag Jasmijn, gezien jouw beschikbaarheid (vanaf 4 juli) en het veldwerk van Sander, laat ik jou hier de start maken.
De (eerste) dr…
-
related to #703 and https://github.com/qgroom/Vespa-velutina/issues/3 .
Currently, when mapping for http://purl.obolibrary.org/obo/RO_0002574 is specified in interaction_types_mapping.csv , the …
-
It would be nice if user can specify number of slices (e.g 25) when performing a large visitation operation in order to use parallel clients where each client will visit a subset of the total corpus s…