-
We are trying to test out feed performance with 12 container nodes and 10 content nodes, running on on-prem hosts with 8 cores. We are running 12 feed clients concurrently, with a maximum of 4 conn…
-
В понс предлагат:
недостиг [o. дефицит]
-
There's a noticeable speed issue in the current implementation. Investigate the root cause of this performance bottleneck and implement optimizations to improve overall system speed. Likely this can b…
-
We've done quite well at cutting down the compute demands of the pipeline, but there are still several steps that are limiting on runtime (e.g. [1](https://github.com/naobservatory/mgs-workflow/issues…
-
Hi experts:
I git the openpiton_dev branch, and changed the code reference the second last Metro-MPI commit (https://github.com/metro-mpi/metro-mpi/commits/metro-mpi/ commit https://github.com/Prin…
-
I did the same installation steps for Linux and Windows and used the same generation parameters
# OS: Linux
**Model: Flux dev nf4
UI: latest
Cuda: 12.1
Python: 3.10**
**console output during g…
-
**Bottleneck Delay**: The delay is simply the sum of the individual PeMS segment delays for the entire duration and spatial extent of the bottleneck. The delay is calculated with respect to a threshol…
-
There are roughly 2000 possible moves that the E-type models are predicting. Your `d_model` parameter is 512 for the CT-E-20 model, which is smaller than 2000. You have the following line in your mode…
-
We could make improvements to schemafeed's `pauseOrResumePolling` loop by avoiding the use of errors for checking.
Context: https://cockroachlabs.slack.com/archives/C065X5307U3/p1724335407360119?th…
-
Recent profiling [profvis_arduous.dingo_20240815.html.zip](https://github.com/user-attachments/files/16641071/profvis_arduous.dingo_20240815.html.zip) from exploration-tooling uncovered a couple slow …