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
Accelerator
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 }}
Leolty
/
verl
Public
forked from
verl-project/verl
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
verl
verl
on
main
User selector
All users
All time
Commit history
Commits on Apr 1, 2025
docs: improve installation and ulysses docs (#854)
eric-haibin-lin
authored
776b0a9
View commit details
Copy full SHA for 776b0a9
View code at this point
Browse repository at this point
feat: support no reference model; fix KL issues (#644)
Show description for 072fc9f
dirtyDan0
authored
072fc9f
View commit details
Copy full SHA for 072fc9f
View code at this point
Browse repository at this point
Commits on Mar 31, 2025
[ulysses] fix: repeat kv heads by sp_size//nheads_k if nheads_k is less than sp_size (#850)
wuxibin89
authored
c0621e1
View commit details
Copy full SHA for c0621e1
View code at this point
Browse repository at this point
[BREAKING] feat: support custom datasets for SFT trainer (#832)
Show description for 77babf1
eric-haibin-lin
authored
77babf1
View commit details
Copy full SHA for 77babf1
View code at this point
Browse repository at this point
[doc] feat: doc for val_before_train (#840)
tongyx361
authored
816dacc
View commit details
Copy full SHA for 816dacc
View code at this point
Browse repository at this point
[fix] Add param to resolve custom model loading failure (#845)
BearBiscuit05
authored
1f78e8b
View commit details
Copy full SHA for 1f78e8b
View code at this point
Browse repository at this point
[BREAKING config] fix: move val_before_train to config yaml. Using trainer.val_before_train instead of +trainer.val_before_train going forward (#820)
tongyx361
authored
64bddb6
View commit details
Copy full SHA for 64bddb6
View code at this point
Browse repository at this point
Commits on Mar 30, 2025
[BREAKING config] feat: add mlflow val generation log and uri config (#822)
Show description for 7fbf609
ccclyu
authored
7fbf609
View commit details
Copy full SHA for 7fbf609
View code at this point
Browse repository at this point
[sglang] fix: add memory saver support to sglang rollout to avoid OOMs (#756)
Show description for 5138a22
SwordFaith
and
ocss884
authored
5138a22
View commit details
Copy full SHA for 5138a22
View code at this point
Browse repository at this point
Megatron checkpoint default not save hf_models, and provide model merge tool. (#780)
Show description for ccab836
ETOgaosion
authored
ccab836
View commit details
Copy full SHA for ccab836
View code at this point
Browse repository at this point
Commits on Mar 29, 2025
[misc] add deepseek v3 flops compute func (#814)
BearBiscuit05
authored
0cf4ca4
View commit details
Copy full SHA for 0cf4ca4
View code at this point
Browse repository at this point
[megatron] fix: remove redundant return value for hf_config (#722)
caaatch22
authored
f3913d0
View commit details
Copy full SHA for f3913d0
View code at this point
Browse repository at this point
docs: update checkpoint doc (#800)
Show description for 50cba4a
ETOgaosion
authored
50cba4a
View commit details
Copy full SHA for 50cba4a
View code at this point
Browse repository at this point
Commits on Mar 28, 2025
[trainer] fix: skip the update step when encountering gradient overflow (#789)
Show description for 093e959
zhao9797
authored
093e959
View commit details
Copy full SHA for 093e959
View code at this point
Browse repository at this point
Fix padding length for sglang rollout in veRL (#773)
Show description for 52e80fc
GeLee-Q
and
GeLee-Q
authored
52e80fc
View commit details
Copy full SHA for 52e80fc
View code at this point
Browse repository at this point
Fix: Undefined variable name in the print_rank_0 statement (#788)
Show description for 44ac51f
jinqinn
authored
44ac51f
View commit details
Copy full SHA for 44ac51f
View code at this point
Browse repository at this point
Commits on Mar 27, 2025
refactor: remove redundant response_mask computation (#792)
tongyx361
authored
f09e365
View commit details
Copy full SHA for f09e365
View code at this point
Browse repository at this point
doc: upgrade to vllm 0.8.2 (#769)
hiyouga
authored
1b8b0e8
View commit details
Copy full SHA for 1b8b0e8
View code at this point
Browse repository at this point
prevent division by zero in SFT (#786)
cpchenpi
authored
076acdf
View commit details
Copy full SHA for 076acdf
View code at this point
Browse repository at this point
fix: prevent NaN when all items are intentionally masked by adding epsilon to masked_mean (#782)
The-Hierophant
authored
e9bcd2b
View commit details
Copy full SHA for e9bcd2b
View code at this point
Browse repository at this point
fix: set torch dtype to auto (#749)
Show description for a1dd922
ccclyu
authored
a1dd922
View commit details
Copy full SHA for a1dd922
View code at this point
Browse repository at this point
Commits on Mar 26, 2025
fix: prompt_token_ids should be list[int] instead of np.array (#772)
Show description for fc27caf
0x404
and
hiyouga
authored
fc27caf
View commit details
Copy full SHA for fc27caf
View code at this point
Browse repository at this point
[Bug fix] change max_model_len to be configurable in vllm_rollout_spmd (#677)
Show description for 7f10e57
Tian Wang
authored
7f10e57
View commit details
Copy full SHA for 7f10e57
View code at this point
Browse repository at this point
[Feat] add max_ckpt_to_keep for old ckpt removal (#724)
Show description for afb9f9f
guanning03
authored
afb9f9f
View commit details
Copy full SHA for afb9f9f
View code at this point
Browse repository at this point
Commits on Mar 25, 2025
fix: prime refactor ignores extra_info (#717)
ganler
authored
8594627
View commit details
Copy full SHA for 8594627
View code at this point
Browse repository at this point
refactor: unify ulysses flash attention patch to avoid single model patches (#735)
Show description for ab7b941
wuxibin89
authored
ab7b941
View commit details
Copy full SHA for ab7b941
View code at this point
Browse repository at this point
Add tqdm progress bar to RayPPOTrainer to visualize training progress (#615)
Show description for 36c10bf
BeSkyer
and
hiyouga
authored
36c10bf
View commit details
Copy full SHA for 36c10bf
View code at this point
Browse repository at this point
Commits on Mar 23, 2025
fix: slicing returns DataProto not DataProtoItem (#718)
ganler
authored
44a65f9
View commit details
Copy full SHA for 44a65f9
View code at this point
Browse repository at this point
[feat] Megatron checkpoint support for current Llama and Qwen models (#687)
Show description for 5d0a7ea
4 people
authored
5d0a7ea
View commit details
Copy full SHA for 5d0a7ea
View code at this point
Browse repository at this point
skip special tokens (#715)
Show description for a339f6f
gameofdimension
authored
a339f6f
View commit details
Copy full SHA for a339f6f
View code at this point
Browse repository at this point
Fix checkpoint loading in fsdp_checkpoint_manager.py and ray_trainer.py (#712)
haoy-zzz
authored
c523a31
View commit details
Copy full SHA for c523a31
View code at this point
Browse repository at this point
Commits on Mar 22, 2025
fix: support transformers==4.50.0 (#704)
Show description for 3f6d45d
dirtyDan0
authored
3f6d45d
View commit details
Copy full SHA for 3f6d45d
View code at this point
Browse repository at this point
Commits on Mar 21, 2025
[Bug Fix] Fix SGLang rollout error under multi node (#652)
ocss884
authored
612823a
View commit details
Copy full SHA for 612823a
View code at this point
Browse repository at this point
[misc] Add Ulysses parallel config precheck (#674)
Show description for e67dea6
uygnef
authored
e67dea6
View commit details
Copy full SHA for e67dea6
View code at this point
Browse repository at this point
Commits on Mar 20, 2025
[tracking] swanlab add `verl` config (#663)
Show description for 9478885
Zeyi-Lin
authored
9478885
View commit details
Copy full SHA for 9478885
View code at this point
Browse repository at this point
Previous
Next
You can’t perform that action at this time.