Hacktoberfest 2026: le issue che i maintainer hanno segnato per ottobre, aperte e adatte ai principianti. Sfoglia le issue Hacktoberfest

DPO manual and CLI wizard reverse dpo_beta: they say a lower beta keeps the model closer to the reference model, but in TRL a higher beta does (F-W1-DOCS-TRAIN-07)

Aperta Adatta ai principianti
  • Qualcosa non va?
#880 0 commenti 0 reazioni 0 assegnatari Vedi su GitHub

Nessuno ha ancora preso questa issue.

Valutazione

Difficoltà
2/5
Tempo stimato
1-3 ore
Idoneità per principianti
68/100
Tipo di issue
Documentazione
Chiarezza
Specificata chiaramente
Stato di attività
Attiva
Stack tecnologico
markdown, python
Ambito
cli, documentation

Direzione di ricerca

Inizia dal parametro DPO e dalle tabelle di tuning in docs/usermanuals/en/training/dpo.md (intorno alle righe 111 e 136-145), poi passa al corrispettivo turco in docs/usermanuals/tr/training/dpo.md (righe 102 e 125-134). Modifica inoltre la stringa del prompt della procedura guidata in forgelm/wizard/_collectors.py vicino alla riga 380. La riga 732 di forgelm/trainer.py si limita a passare il valore e deve restare invariata. Il lavoro è completato quando ogni descrizione di beta in entrambe le lingue e il prompt della procedura guidata dicono che un beta più alto mantiene la policy più vicina al riferimento, e quando il suggerimento consiglia di aumentare beta quando la valutazione peggiora.

Scritto dal modello di indicizzazione a partire dal testo della issue.

Descrizione

documentation good first issue severity: medium source: review-2026-09 wave: 3

Wave 3 · group W3 Docs · highest severity Medium · 1 finding(s) · effort docs fix

Summary

The DPO page tells users that a lower dpo_beta keeps the policy closer to the reference model and a higher one makes a more aggressive preference shift, and advises lowering beta when evaluation regresses. TRL, which receives the value unchanged, defines the opposite: a higher β means less deviation from the reference model. A user who follows the tip to fix over-fitting weakens the KL regularisation and moves the model further from the reference.

Primary finding

F-W1-DOCS-TRAIN-07 — DPO beta guidance reverses the regularisation direction

severity Medium · confidence high · checked against the installed TRL source and its reference docs

What happens. The parameter table (docs/usermanuals/en/training/dpo.md:111), the tuning table and the tip (docs/usermanuals/en/training/dpo.md:136-145) describe a lower β as staying close to the reference and a higher β as more aggressive; the Turkish mirror says the same (docs/usermanuals/tr/training/dpo.md:102, docs/usermanuals/tr/training/dpo.md:125-134). The CLI wizard prompt repeats it: dpo_beta — temperature / KL strength (lower → closer to reference model) (forgelm/wizard/_collectors.py:380). ForgeLM passes training.dpo_beta to TRL's DPOConfig unchanged (forgelm/trainer.py:732). The site's own wizard text already states the correct direction (site/js/translations.js:1651: "higher beta keeps it close to the reference"), so the site and the manual contradict each other.

Where.

Evidence (re-verified on 2026-10-09).

  • Installed TRL 1.7.1, trl/trainer/dpo_config.py: "Parameter controlling the deviation from the reference model. Higher β means less deviation from the reference model." The official reference says the same: https://huggingface.co/docs/trl/dpo_trainer#trl.DPOConfig
  • forgelm/trainer.py:732 sets kwargs["beta"] = self.config.training.dpo_beta with no transformation.

Recommended fix. Reverse the direction in the parameter table, the tuning table and the tip, in English and Turkish: a lower β weakens the KL penalty and lets the policy move further from the reference (a more aggressive preference shift); a higher β keeps it closer. When evaluation regresses while training loss keeps falling, advise a higher β (or fewer steps, or a lower learning rate) checked against held-out evaluation. Change the CLI wizard prompt to (higher → closer to reference model). Linking TRL's DPOConfig reference from the parameter row keeps the two from drifting again.

Acceptance criteria

  • The English and Turkish DPO pages describe β the way TRL's DPOConfig does, in the parameter table, the tuning table and the tip.
  • The CLI wizard prompt and the site wizard describe the same direction.
  • Documentation listed above is corrected in English and in its Turkish mirror in the same PR.
  • The finding is resolved or explicitly closed with a reason.

Found by the September 2026 independent review of ForgeLM against commit cab2947 and re-verified against development at f94595f on 2026-10-09. Code links point at cab2947; line numbers may have moved since. Finding IDs (F-W1-…) are stable references for this backlog.

Lingua principale
Python
Stelle
9
Fork
1
Merge medio
4h 3m
PR unite (30g)
3

Preparare l'ambiente

Come iniziare

  1. Leggi tutta la issue e poi la guida ai contributi del progetto.
  2. Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
  3. Fai un fork del repository e lavora su un branch.
  4. Apri una pull request che faccia riferimento al numero della issue.

Altre issue di HodeTech/ForgeLM

Tutte le issue di HodeTech/ForgeLM

Issue simili

Altre issue su Python

Ricevi le nuove issue nella tua casella

Un breve riepilogo di issue GitHub adatte ai principianti.