RepoJournal
Alignment & Post-training Alignment & Post-training
47 wires and counting

$ follow Alignment & Post-training

Keep up with Alignment & Post-training in about 3 minutes: what actually shipped — the commits, pull requests, releases, and security advisories that matter.

or

fair warning: these emails are deeply technical. diffs, version numbers, CVEs, benchmark deltas. if that's not your idea of a good read, this isn't your newsletter.

Folds into your digest — weekly by default, monthly if you prefer. Unsubscribe in one click.

$ status

wire 2026-11-11
stories 24

© 2026 RepoJournal Home Showcase How it works Privacy

$ the-wire · showcase

verl adds batch-invariant actor forward, MXFP4 MoE recipe lands

By RepoJournal · Filed · About Alignment & Post-training · Composed from the cited sources · methodology

verl shipped the training-side half of exact 0-diff train-rollout log-probs and an MXFP4 recipe for MoE routed experts, while open-instruct fixed a reward verifier that was zeroing correct answers.

[fsdp, trainer, cfg, doc, tests] feat: batch-invariant actor forward and sampler log-prob formula verl-project/verl

by cr-gao

An opt-in FSDP engine option, `batch_invariant` (default off), installs vLLM's batch-invariant kernel overrides (matmul, softmax, mean, RMSNorm plus the cuBLAS and TF32 settings) and computes log-probs with the vLLM sampler formula: fp32 `log_softmax` over the full vocabulary, then gather. Turn it on if you want train and rollout log-probs to agree; leave it off and nothing about your current r...

Take the last declared multiple-choice answer and read row-requested formats allenai/open-instruct

by no0p

`mcqa_final_choice` used to zero out correct answers that repeated the letter, wrapped it in bold, parentheses or `\text{}`, or glossed it on the same line, and could not read the per-row answer formats a prompt actually requests. It now takes the last explicit declaration in the final answer and pulls the row's one-capture regex from `template_metadata.output_regex`, so reward scores on prepar...

[megatron, doc] feat: MXFP4 MoE recipe and Qwen3-30B-A3B example verl-project/verl

by 0z5a

`examples/mxfp4/run_qwen3_30b_a3b_megatron.sh` runs GRPO on Qwen3-30B-A3B with DAPO-Math-17k on a single 4-GPU Blackwell node, tested on 4×B300 (4×B200 is untested). It reuses the data, hyper-parameters and layout of the MXFP8 recipe: training TP2 × EP2, full recompute, optimizer on CPU.

[algo, megatron, cfg, doc, tests] feat: score centering on the Megatron actor verl-project/verl

by cr-gao

Score centering, previously FSDP-side, now runs on the Megatron actor: sampler head log-probs are normalized over the vocab-parallel shards, with the row max and normalizer all-reduced over the TP group and the heads split by context parallel the same way top-k distillation does. If you run Megatron and were blocked from enabling score centering, that restriction is gone.

Prepare Olmo 3.5 release data fixes allenai/open-instruct

by abhishekraok

On the long tail: open-instruct added CPU preparation tooling for the Olmo 3.5 release-data fixes (24,906 audited exclusion IDs, 58 replacement identity seeds, a knowledge-cutoff check on the release tokenizer), and V1 sync PPO picked up a first proof-of-concept for custom dataloaders.

Quick answers

What shipped in Alignment & Post-training on November 11, 2026?
verl shipped the training-side half of exact 0-diff train-rollout log-probs and an MXFP4 recipe for MoE routed experts, while open-instruct fixed a reward verifier that was zeroing correct answers. In total, 12 commits and 12 pull requests landed.
Who contributed to Alignment & Post-training on November 11, 2026?
6 developers shipped this update, including cr-gao, 0z5a, Mecoli1219, renjunying2, no0p, and abhishekraok.
What were the notable Alignment & Post-training updates?
[fsdp, trainer, cfg, doc, tests] feat: batch-invariant actor forward and sampler log-prob formula, Take the last declared multiple-choice answer and read row-requested formats, and [megatron, doc] feat: MXFP4 MoE recipe and Qwen3-30B-A3B example.