-
The bleu/chrf/ce-mean-words_stalled metrics are supposed to be on the online training dashboards but they are missing.
https://wandb.ai/moz-translations/zh-en/runs/backwards_E8S5g?nw=nwuserepavlov
…
-
Hi,
I'm trying to train a multi output nn and I need to change the weight of each loss component depending on the epoch number. In some previous versions of keras I implemented this mechanism by defi…
-
When using `during_training_metric`, the respective quantity is calculated in every epoch, which may be costly if `during_training_metric="total_energy"`.
When using shuffled snapshots, adding the …
-
Airflow metrics don't get reset after a restart, however, the metrics did not get initialized. This lead to some unexpected PromQL responses when querying with missing data.
For example, a task st…
-
I am experiencing an issue with Filebeat in our Kubernetes cluster and would like to know if others have encountered a similar situation or have any insights into this behavior.
Issue: I have obser…
-
We had installed the helm chart to collect metrics from postgres. We can Collect very nice metrics, but we can´t find Metrics with 'pg_stat_statements'.
We had activate the Extenstion and also in Con…
-
First of all, I love the new dashboard stuff!
However, would it be possible to view longer time periods? The limit seems to be 24 weeks - at [UBC Launch Pad](https://github.com/ubclaunchpad), as a …
-
-
We have enabled Kafka Exporter metrics for our cluster.
When I query kafka topic metrics like `kafka_topic_partitions`, it shows the respective data for all topics.
But when I query kafka producer …
-
### Component(s)
datadogexporter
### What happened?
## Description
We are currently experiencing an issue with the OpenTelemetry Collector running in our Kubernetes cluster, which is managed…