Reference implementation for Token-level Direct Preference Optimization(TDPO)
active 2024-04-16 → 2026-05-23 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 768 days from 2024-04-16 to 2026-05-23. Pushes: 13 total, peak 2 in a day. Pull requests: 2 total, peak 2 in a day. Issues: 11 total, peak 3 in a day. Comments: 8 total, peak 3 in a day. Stars: 139 total, peak 6 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
Recent activity
Latest issues, pull requests and releases
- Issue#7Vance01242024-11-13 02:33Some questions about the paper
- Issue#8junkangwu2024-09-26 07:03Inquiry About KL Divergence in ICML 2024 Paper
- Issue comment#8junkangwu2024-09-26 07:03Inquiry About KL Divergence in ICML 2024 Paper
- Issue comment#8Vance01242024-09-26 04:25Inquiry About KL Divergence in ICML 2024 Paper
- Issue comment#7Vance01242024-09-26 04:15Some questions about the paper
- Issue#7robin0872024-09-24 07:33Some questions about the paper
- Issue#6tcxia2024-09-11 08:01Can you train DPO directly? Using open-source base models.
- Issue#5wangxu08202024-07-15 03:17How do you go about evaluating ALIGNMENT(accuracy) and DIVERSITY(entropy)? Is there a code available?
- Issue#4Vance01242024-07-06 08:38How to eval the models
- Issue#3Vance01242024-07-06 08:37How about the loss curve? especially when converge
- Issue#2Vance01242024-07-06 08:37Some questions about the code
- Issue comment#4Vance01242024-06-27 14:33How to eval the models
- Issue comment#3Vance01242024-06-27 14:13How about the loss curve? especially when converge
- Issue comment#2Vance01242024-06-27 14:08Some questions about the code
- Issue#4Yeeesir2024-06-26 12:38How to eval the models
- Issue#3LuckerYi2024-06-13 09:18How about the loss curve? especially when converge
- Issue comment#2yuchen8142024-06-05 14:39Some questions about the code
- Issue comment#2Vance01242024-05-28 09:23Some questions about the code
- Issue#2yuchen8142024-05-18 14:38Some questions about the code
- Pull request#1Vance01242024-05-06 13:09
- Pull request#1fiberleif2024-05-06 13:05
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 139 stars here means stars gained during the window, not the repo's star count.