Skip to content

Pull requests: fw-ai/cookbook

New pull request New
59 Open 340 Closed
59 Open 340 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

feat(training): pass grad_accum to trainer and clarify batch_size docs
#406 opened Apr 30, 2026 by hershalb Contributor Loading…
1 of 4 tasks
feat(verifier): empirical renderer probe (Phase 0) + spinup helper
#405 opened Apr 29, 2026 by Hecate0821 Collaborator Loading…
3 of 4 tasks
1
Fix dataloader cursor accounting (RL, IGPO, DPO, SFT)
#404 opened Apr 29, 2026 by Hecate0821 Collaborator Loading…
2
fix(infra): strip pinned /versions/ suffix from shape IDs and enrich 400 error messages
#392 opened Apr 27, 2026 by andrefoo • Draft
[codex] add sampled-token OPD recipe
#391 opened Apr 25, 2026 by renfeichen-fw Contributor Loading…
feat(opd): add on-policy distillation metrics and recipe (arXiv:2604.13016)
#390 opened Apr 24, 2026 by renfeichen-fw Contributor • Draft
14 of 28 tasks
feat(rl): gate-native async RL loop + flat Rollout contract
#382 opened Apr 23, 2026 by Hecate0821 Collaborator • Draft
docs(training): LR API surface — market research across 9 frameworks
#361 opened Apr 21, 2026 by benjibc Contributor • Draft
1 of 6 tasks
feat(recipes): add per-step learning rate support to all training recipes
#359 opened Apr 21, 2026 by renfeichen-fw Contributor Loading…
15 of 28 tasks
1
perf(training): parallelize SFT dataset rendering with multiprocessing
#358 opened Apr 21, 2026 by xiaoyifan Contributor Loading…
5 of 11 tasks
fix(client): close gateway-route race with data-plane warmup + verified-404-retry
#348 opened Apr 17, 2026 by websterbei Contributor • Draft
Retry transient 404 (NotFoundError) in ReconnectableClient
#344 opened Apr 17, 2026 by websterbei Contributor Loading…
fix(training_shapes): add fallback passes and diagnostic logging for shape selection
#338 opened Apr 15, 2026 by benjibc Contributor • Draft
fix(training_shapes): handle 403/non-success model fetch gracefully in auto-select
#337 opened Apr 15, 2026 by websterbei Contributor • Draft
feat: add RLSD (Self-Distilled RLVR) credit assignment
#333 opened Apr 14, 2026 by morgendave Contributor Loading…
fix(training): retry with backoff on 429 when creating RLOR trainer job
#331 opened Apr 14, 2026 by hershalb Contributor Loading…
1 of 2 tasks
Preflight training smoke secret setup
#320 opened Apr 11, 2026 by benjibc Contributor • Draft
feat(checkpoint): validate checkpoint entries before resume
#314 opened Apr 9, 2026 by hershalb Contributor Loading…
14 of 23 tasks
feat(training): add accelerator_type filtering to training shape selection
#308 opened Apr 7, 2026 by hershalb Contributor Loading…
Add cookbook release workflows and runbook
#307 opened Apr 7, 2026 by benjibc Contributor Loading…
1
Add cookbook release process and release blocker checklist
#305 opened Apr 6, 2026 by benjibc Contributor • Draft
Add KV-cache key compression reproducible comparison
#303 opened Apr 6, 2026 by yi-fireworks Loading…
4 of 5 tasks
feat: LoRA self-reference RL — use policy trainer as KL reference
#299 opened Apr 6, 2026 by mayinghan Contributor Loading…
use parsed content as output
#296 opened Apr 3, 2026 by hershalb Contributor Loading…
[codex] Fix training SDK smoke compatibility
#292 opened Apr 2, 2026 by benjibc Contributor Loading…
1
ProTip! Exclude everything labeled bug with -label:bug.