PR Template Check

Workflow runs for PR Template Check

• feat: log raw importance ratios and fraction of truncation/masking in vLLM importance sampling correction

  skipped:
  PR Template Check#2477:Pull request#7778openedbymuupan
  Oct 17, 20261s
  More actions for workflow run 2477

• Only cast QLoRA adapter weights to bf16 when training in bf16

  failed:
  PR Template Check#2476:Pull request#7777openedbySailaja-ASB
  Oct 16, 202612s
  More actions for workflow run 2476

• Load the tokenizer for models that take no images

  skipped:
  PR Template Check#2475:Pull request#7774openedbyqgallouedec
  Oct 16, 20261s
  More actions for workflow run 2475

• Place vLLM image features at the model config's image token

  skipped:
  PR Template Check#2474:Pull request#7773openedbyalbertvillanova
  Oct 16, 20261s
  More actions for workflow run 2474

• Warn when LoRA skips the dense and shared expert layers of MoE models on transformers v5 with peft 0.19 to 0.21.2

  skipped:
  PR Template Check#2473:Pull request#7771openedbyalbertvillanova
  Oct 16, 20267s
  More actions for workflow run 2473

• Docs: stop pointing Async GRPO at hybrid Qwen3.5 models

  failed:
  PR Template Check#2472:Pull request#7770openedbyEryx-Z
  Oct 16, 202614s
  More actions for workflow run 2472

• Raise when LoRA would skip the routed experts of MoE models on transformers v5 with peft<0.19

  skipped:
  PR Template Check#2471:Pull request#7769openedbyalbertvillanova
  Oct 16, 20262s
  More actions for workflow run 2471

• Test that LoRA on an MoE model also trains its dense and shared expert layers

  skipped:
  PR Template Check#2470:Pull request#7768openedbyalbertvillanova
  Oct 16, 20262s
  More actions for workflow run 2470

• Read and write text files as UTF-8 in tests

  skipped:
  PR Template Check#2469:Pull request#7764openedbyalbertvillanova
  Oct 16, 20261s
  More actions for workflow run 2469

• Test Qwen3.5 loaded as a text-only causal LM

  skipped:
  PR Template Check#2468:Pull request#7763openedbyqgallouedec
  Oct 16, 20261s
  More actions for workflow run 2468

• Allow FlashAttention 2 in Async GRPO tree attention

  skipped:
  PR Template Check#2467:Pull request#7762openedbyDimensionSTP
  Oct 16, 20262s
  More actions for workflow run 2467

• Asyncgrpo score centering

  skipped:
  PR Template Check#2466:Pull request#7760openedbyAmineDiro
  Oct 15, 20262s
  More actions for workflow run 2466

• Fix empty-target training in SDPO and SDFT

  skipped:
  PR Template Check#2465:Pull request#7758openedbyDimensionSTP
  Oct 15, 20261s
  More actions for workflow run 2465

• ⬆️ Bump dev version

  skipped:
  PR Template Check#2464:Pull request#7757openedbyqgallouedec
  Oct 15, 20261s
  More actions for workflow run 2464

• Release: v1.16

  skipped:
  PR Template Check#2463:Pull request#7756openedbyqgallouedec
  Oct 15, 202611s
  More actions for workflow run 2463

• Read and write text files as UTF-8

  skipped:
  PR Template Check#2462:Pull request#7755openedbyalbertvillanova
  Oct 15, 20269s
  More actions for workflow run 2462

• Add an LFM2.5-8B-A1B research agent example (SFT + GRPO)

  skipped:
  PR Template Check#2461:Pull request#7754openedbyqgallouedec
  Oct 15, 20261s
  More actions for workflow run 2461

• Add LFM2.5-24B-A2B support and testing

  skipped:
  PR Template Check#2460:Pull request#7753openedbyqgallouedec
  Oct 15, 20262s
  More actions for workflow run 2460

• Require torch>=2.7.0

  skipped:
  PR Template Check#2459:Pull request#7752openedbyalbertvillanova
  Oct 15, 20262s
  More actions for workflow run 2459

• Fix the default IS mode in the GRPOConfig docstring

  skipped:
  PR Template Check#2458:Pull request#7751openedbysergiopaniego
  Oct 15, 20262s
  More actions for workflow run 2458

• Skip the async GRPO tests on transformers < 5.2.0

  skipped:
  PR Template Check#2457:Pull request#7750openedbyalbertvillanova
  Oct 15, 20262s
  More actions for workflow run 2457

• Example / Experimental: Single GPU LoRA + single harness

  skipped:
  PR Template Check#2456:Pull request#7749openedbyyounesbelkada
  Oct 15, 20262s
  More actions for workflow run 2456

• Fix activation offloading on torch < 2.4

  skipped:
  PR Template Check#2455:Pull request#7748openedbyalbertvillanova
  Oct 15, 20261s
  More actions for workflow run 2455

• Pin bitsandbytes to its floor in the min-versions CI job

  skipped:
  PR Template Check#2454:Pull request#7746openedbyalbertvillanova
  Oct 15, 20261s
  More actions for workflow run 2454

• Document that the minimum-versions CI job pins every dependency at its floor

  skipped:
  PR Template Check#2453:Pull request#7745openedbyalbertvillanova
  Oct 15, 20262s
  More actions for workflow run 2453