Issues

• investigate: weights load at ~1.87x their on-disk size, inflating GPU sizing

  feature/granite-turboctc-defaultinvestigationprio:lowvllm
  generative-computing/granite-switch#128 · Filter by author aviv1ron1 opened on Sep 8, 2026

• tests: invocation-tokenization tests depend on a live HF tokenizer download

  feature/granite-turboctc-defaultflaky-testprio:mediumtests
  generative-computing/granite-switch#127 · Filter by author aviv1ron1 opened on Sep 8, 2026

• composer: composed checkpoints emit torch_dtype: null, which can make vLLM resolve float32

  bugcomposerfeature/granite-turboctc-defaultprio:medium
  generative-computing/granite-switch#126 · Filter by author aviv1ron1 opened on Sep 8, 2026

• vllm: HasInnerState is declared but unhonoured, and excludes the model from the V2 model runner

  bugfeature/granite-turboctc-defaultprio:lowvllm
  generative-computing/granite-switch#125 · Filter by author aviv1ron1 opened on Sep 8, 2026

• composer: a mamba/hybrid base is accepted and its layers silently relabelled as attention

  bugcomposerfeature/granite-turboctc-defaultprio:medium
  generative-computing/granite-switch#124 · Filter by author aviv1ron1 opened on Sep 8, 2026

• tests: a vLLM engine process survives its test and starves later GPU tests

  bugfeature/granite-turboctc-defaultprio:hightestsvllm
  generative-computing/granite-switch#123 · Filter by author aviv1ron1 opened on Sep 8, 2026

• Control token swapping fail for transformers 5.9.0 for user message activations

  bug
  generative-computing/granite-switch#108 · Filter by author antonpibm opened on Jul 27, 2026
  

• Build MultiSwitch: model-triggered, mid-sequence adapter invoke/return (revisit ADR-018 Pattern-1)

  generative-computing/granite-switch#107 · Filter by author AlonMalach opened on Jul 21, 2026

• REQ-002 "Satisfied" overstates model-autonomous invocation — criterion verifies only the plumbing

  generative-computing/granite-switch#105 · Filter by author AlonMalach opened on Jul 21, 2026

• Bug: Install errors in Google colab in tutorials

  generative-computing/granite-switch#92 · Filter by author serjikibm opened on Jun 11, 2026
  1

• Fix missing colab link in tutorial

  generative-computing/granite-switch#90 · Filter by author lastras opened on Jun 3, 2026

• Add CI/CD pipeline for code quality and compliance

  generative-computing/granite-switch#44 · Filter by author aviv1ron1 opened on May 19, 2026

• downloading the model for the first time is very slow

  generative-computing/granite-switch#43 · Filter by author aviv1ron1 opened on May 19, 2026
  4

• Test suite standardization and cleanup

  generative-computing/granite-switch#42 · Filter by author antonpibm opened on May 19, 2026
  1

• Incremental Adapter Composition (Add / Remove / Update)

  enhancement
  generative-computing/granite-switch#21 · Filter by author antonpibm opened on May 12, 2026

• Verify and add tests for quantization support

  generative-computing/granite-switch#16 · Filter by author yairallouche opened on May 12, 2026
  1

  

• Add support for llama.cpp

  generative-computing/granite-switch#15 · Filter by author yairallouche opened on May 12, 2026
  8

• Path to upstream vLLM and HF support for granite_switch architecture

  generative-computing/granite-switch#14 · Filter by author yairallouche opened on May 12, 2026
  1

• Create a Hugging Face Spaces playground

  generative-computing/granite-switch#13 · Filter by author yairallouche opened on May 12, 2026
  1

• vllm and vllm20 dependency resolving issues

  generative-computing/granite-switch#12 · Filter by author yairallouche opened on May 12, 2026

• Add IVR/IVS tutorial notebooks

  generative-computing/granite-switch#10 · Filter by author yairallouche opened on May 12, 2026

• clarify_query intrinsic enters repetition loop on embedded ALORA switch model

  generative-computing/granite-switch#9 · Filter by author yairallouche opened on May 12, 2026
  1