issues Search Results · language:Dune language:Python language:JavaScript language:Java linked:pr language:Java
Filter by
5.8M results
When invalid number entered Change loop so not all info lines are printed
bug
Prerequisites
- [x] I searched open issues and this has not already been proposed.
- [x] I searched discussions and this is not already being debated there.
- [x] This is a concrete, actionable ...
ready for review
lsi_model.py:130 imports rag_config via from ranking.ranking import rag_config inside a function. All other modules
import directly from rag.config at top-level. Should be: from rag.config import rag_config ...
bug
Refs #1882
Follow-up after PR #2329.
Scope:
- Document that disabled rate-limit storage scaffold outcomes are currently covered by the existing dependency-adapter
unknown storage-code safe-fail ...
documentation
product
Task
Build and fully host-test the host-side half of the minimal ReID gallery (the implementable-now part of #21, split out
so it can proceed ahead of the #17/#8 blocker):
- A concrete Gallery implementation ...
area:inference
prio:P1
status:ready
type:chore
v1
v2-fwd
Task
Add a committed, device/gpu-marked test stub for the on-demand ReID dispatch path:
tests/device/test_reid_dispatch_device.py. It is skipped on the host (markers) and becomes the on-device acceptance ...
area:inference
area:infra
needs:on-device
prio:P1
status:ready
type:chore
v1
🐛 Describe the bug
Problem
When building PyTorch on Windows with:
set Debug=1
linking torch_python.dll fails with:
LINK : fatal error LNK1104: cannot open file python310_d.lib
Root Cause
With ...
bot-triaged
module: build
module: debug-build
module: pybind
module: windows
triaged
Bug Description
When using an OpenAI-compatible Realtime API (e.g. xAI), the response.done event sometimes returns status_details as a
plain string (e.g. incomplete) instead of a structured object { type ...
bug
Describe the bug
When serving Gemma 4 models (specifically google/gemma-4-E2B-it-qat-q4_0-unquantized) that leverage Key-Value (KV) cache
sharing (num_kv_shared_layers 0), the vLLM engine initialization ...
Paper Title
Simultaneous Multi-Topology Multi-Objective Sizing Across Thousands of Analog Circuit Topologies
Paper URL (PDF / arXiv / DOI)
https://ieeexplore.ieee.org/document/4261319
Venue / Journal ...
new-paper

Learn how you can use GitHub Issues to plan and track your work.
Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub IssuesProTip! Restrict your search to the title by using the in:title qualifier.