Skip to content
Navigation Menu
Sign in
Appearance settings
Platform
AI CODE CREATION
GitHub Copilot
Write better code with AI
GitHub Copilot app
Direct agents from issue to merge
MCP Registry
Integrate external tools
DEVELOPER WORKFLOWS
Actions
Automate any workflow
Codespaces
Instant dev environments
Issues
Plan and track work
Code Review
Manage code changes
Code Quality
Enforce quality at merge
APPLICATION SECURITY
GitHub Advanced Security
Find and fix vulnerabilities
Code security
Secure your code as you build
Secret protection
Stop leaks before they start
EXPLORE
Why GitHub
Documentation
Blog
Changelog
Marketplace
View all features
Solutions
BY COMPANY SIZE
Enterprises
Small and medium teams
Startups
Nonprofits
BY USE CASE
App Modernization
DevSecOps
DevOps
CI/CD
View all use cases
BY INDUSTRY
Healthcare
Financial services
Manufacturing
Government
View all industries
View all solutions
Resources
EXPLORE BY TOPIC
AI
Software Development
DevOps
Security
View all topics
EXPLORE BY TYPE
Customer stories
Events & webinars
Ebooks & reports
Business insights
GitHub Skills
SUPPORT & SERVICES
Documentation
Customer support
Community forum
Trust center
Partners
View all resources
Open Source
COMMUNITY
GitHub Sponsors
Fund open source developers
PROGRAMS
Security Lab
Maintainer Community
GitHub Stars
Archive Program
REPOSITORIES
Topics
Trending
Collections
Enterprise
ENTERPRISE SOLUTIONS
Enterprise platform
AI-powered developer platform
AVAILABLE ADD-ONS
GitHub Advanced Security
Enterprise-grade security features
Copilot for Business
Enterprise-grade AI features
Premium Support
Enterprise-grade 24/7 support
Pricing
Search
/
Sign in
Sign up
Appearance settings
You signed in with another tab or window.
Reload
to refresh your session.
You signed out in another tab or window.
Reload
to refresh your session.
You switched accounts on another tab or window.
Reload
to refresh your session.
Dismiss alert
{{ message }}
ghzhost
/
omi
Public
forked from
BasedHardware/omi
Notifications
You must be signed in to change notification settings
Fork
0
Star
0
Code
Pull requests
0
Actions
Projects
Security and quality
0
Insights
Additional navigation options
Code
Pull requests
Actions
Projects
Security and quality
Insights
Commits
Breadcrumbs
History for
omi
backend
parakeet
gpu_worker.py
on
main
User selector
All users
Datepicker
All time
Commit history
Commits on Jul 23, 2026
fix(parakeet): recycle workers after fatal CUDA errors
Show description for 61ac0e2
kodjima33
committed
61ac0e2
View commit details
Copy full SHA for 61ac0e2
View code at this point
Browse repository at this point
Commits on Jul 4, 2026
chore(typecheck): enroll 5 subprojects + 6 scripts (424 total)
Show description for 25e47a7
Git-on-my-level
committed
25e47a7
View commit details
Copy full SHA for 25e47a7
View code at this point
Browse repository at this point
chore(typecheck): enroll production_like_model + agent-proxy + 11 scripts
Show description for 89f3676
Git-on-my-level
committed
89f3676
View commit details
Copy full SHA for 89f3676
View code at this point
Browse repository at this point
Commits on Jun 30, 2026
Add _submit_lock to GPUWorker to prevent stop/submit race
Show description for bdddec3
beastoin
and
claude
committed
bdddec3
View commit details
Copy full SHA for bdddec3
View code at this point
Browse repository at this point
VRAM-aware dynamic batch sizing for parakeet (#8535)
beastoin
authored
d2e9f9a
View commit details
Copy full SHA for d2e9f9a
View code at this point
Browse repository at this point
Commits on Jun 28, 2026
Add BF16 dtype tracking and lower auto-attention threshold to 300s
Show description for 7543bdb
beastoin
and
claude
committed
7543bdb
View commit details
Copy full SHA for 7543bdb
View code at this point
Browse repository at this point
Revert gpu_worker.py changes from PR #8428
Show description for 650057b
beastoin
and
claude
committed
650057b
View commit details
Copy full SHA for 650057b
View code at this point
Browse repository at this point
Lower auto attention threshold from 600s to 300s for batch VRAM safety
Show description for 0ea71c8
beastoin
and
claude
committed
0ea71c8
View commit details
Copy full SHA for 0ea71c8
View code at this point
Browse repository at this point
Fix BF16 dtype mismatch for local attention mode at load time
Show description for 5da5582
beastoin
and
claude
committed
5da5582
View commit details
Copy full SHA for 5da5582
View code at this point
Browse repository at this point
Fix BF16 dtype mismatch when auto-switching attention mode
Show description for 353ba7b
beastoin
and
claude
committed
353ba7b
View commit details
Copy full SHA for 353ba7b
View code at this point
Browse repository at this point
Add attention mode system and duration guard to GPU worker
Show description for 4c26a7c
beastoin
and
claude
committed
4c26a7c
View commit details
Copy full SHA for 4c26a7c
View code at this point
Browse repository at this point
Commits on Jun 23, 2026
Move in-function imports to module top level per AGENTS.md rules
Show description for 91d3cc1
beastoin
and
claude
committed
91d3cc1
View commit details
Copy full SHA for 91d3cc1
View code at this point
Browse repository at this point
Fix CUDA stream conflict: route embedding through GPU worker thread
Show description for 4115fda
beastoin
and
claude
committed
4115fda
View commit details
Copy full SHA for 4115fda
View code at this point
Browse repository at this point
Commits on Jun 15, 2026
Parakeet: add 6 observability metrics for ASR monitoring (#7960)
beastoin
authored
bf11c79
View commit details
Copy full SHA for bf11c79
View code at this point
Browse repository at this point
Commits on Jun 14, 2026
Disable torch.compile by default, add gc.disable for GPU thread
Show description for 94ec53f
beastoin
and
claude
committed
94ec53f
View commit details
Copy full SHA for 94ec53f
View code at this point
Browse repository at this point
Disable CUDA graphs by default to prevent batch inference crashes
Show description for a47f052
beastoin
and
claude
committed
a47f052
View commit details
Copy full SHA for a47f052
View code at this point
Browse repository at this point
Fix torch VRAM attribute: total_mem -> total_memory
Show description for e1e5898
beastoin
and
claude
committed
e1e5898
View commit details
Copy full SHA for e1e5898
View code at this point
Browse repository at this point
Restore BF16 model loading in GPU worker
Show description for f9e1510
beastoin
and
claude
committed
f9e1510
View commit details
Copy full SHA for f9e1510
View code at this point
Browse repository at this point
Add GPU worker + dynamic batching for parakeet batch transcription
Show description for 48c48ea
beastoin
and
claude
committed
48c48ea
View commit details
Copy full SHA for 48c48ea
View code at this point
Browse repository at this point
Loading
You can’t perform that action at this time.
