-
on 6.2.1 of kafka connect it was all fine but in 7.2.0 i see 3 problems:
1. it is doing DEBUG level logs (printing sensitive token)
```
{"log":"0 [main] DEBUG io.confluent.admin.utils.cli.KafkaRe…
-
## Project Incubating Request
### Redis on TiKV (Tidis)
#### Summary
Tidis is the service layer for TiKV, aims to provide multiple model distributed storage service.
Tidis has been complet…
-
### Is there an existing issue for this?
- [X] I have searched the existing issues
### Environment
```markdown
- Milvus version: 2.4.6
- Deployment mode(standalone or cluster): cluster
- MQ type(…
-
Give me a shirt or a bag
-
## 🐛 Bug
When running fairseq-generate for task translation_multi_simple_epoch, I get the hydra config error:
hydra.errors.ConfigCompositionException: Error merging override distributed_training.d…
-
**Issue summary**
CGGMP paper requires that the ring-Pedersen parameters is a product of safe primes. For security we need to generate 1024 bit safe primes, which is quite time-consuming. One issue…
-
One large limitation on spark/hive stuff over traditional datawarehouse is to provide auto-increment fields.
I wonder if delta could achieve this. One idea would be to store the actual max_value wi…
-
### 🚀 The feature, motivation and pitch
When using `TORCH_DISTRIBUTED_DEBUG=DETAIL` we collect collectives fingerprints and those are quite helpful when troubleshooting issues like stragglers.
One…
-
For
- TensorFlow
- PyTorch
- XGBoost
we need to consider the following questions
1. How to combine the in-train-loop data transformation and the training iteration, and how to make the com…
-
I found that if you spin up a multi-node cluster that uses the AWS Data Protection Provider and all nodes are starting at the exact same time, then you might encounter an issue where more than one "/M…