transformers
CI: add vllm-test-init and vllm-test-transformers jobs on dedicated runners
#47934
Merged

Commits
  • Test vLLM CI on larger runner with test_init and test_transformers enabled
    ydshieh committed 99 days ago
  • Trigger tests_processors CI run
    ydshieh committed 99 days ago
  • Split test_initialization into 8-shard matrix job; add HF_HOME env
    ydshieh committed 99 days ago
  • Trigger vLLM CI run
    ydshieh committed 99 days ago
  • CI: isolate JambaForCausalLM test (1 shard) to debug mamba_mixer IndexError
    ydshieh committed 99 days ago
  • CI: switch isolated test to FalconMambaForCausalLM (same mamba_mixer IndexError)
    ydshieh committed 99 days ago
  • CI: patch mamba_mixer conv1d weight.view to handle 2D dummy-loaded weight
    ydshieh committed 99 days ago
  • CI: patch mamba_mixer out_proj transpose contiguous for CPU onednn_mm
    ydshieh committed 99 days ago
  • CI: fix mamba_mixer sed patch (preserve closing paren of out_proj call)
    ydshieh committed 99 days ago
  • CI: add detailed vLLM upstream bug comments for mamba_mixer CPU patches
    ydshieh committed 99 days ago
  • CI: add GPU job to verify mamba_mixer bugs are CPU-specific
    ydshieh committed 99 days ago
  • CI: fix GPU test-init job (no curl, shell bash, ipc host, pip not uv)
    ydshieh committed 99 days ago
  • CI: install vLLM from CUDA 13.0 nightly wheel for GPU test job
    ydshieh committed 99 days ago
  • CI: unset VLLM_TARGET_DEVICE for GPU job, add shell bash to pytest step
    ydshieh committed 99 days ago
  • CI: use PyPI vllm for GPU job (nightly wheel has broken _C_stable_libtorch)
    ydshieh committed 99 days ago
  • CI: disable GPU test-init job (vllm._C_stable_libtorch ABI mismatch)
    ydshieh committed 99 days ago
  • CI: run test_can_initialize_small_subset (8 shards), comment out mamba patches
    ydshieh committed 99 days ago
  • CI: reduce vllm-test-init to 4 shards (small subset has only 12 tests)
    ydshieh committed 99 days ago
  • CI: skip Gemma3nForCausalLM in vllm-test-init (hangs on shm_broadcast)
    ydshieh committed 99 days ago
  • CI: disable vllm-test-init job (Gemma3nForCausalLM hangs on shm_broadcast)
    ydshieh committed 99 days ago
  • CI: add vllm-test-transformers job on dedicated aws-m8i-8xl-cache runner
    ydshieh committed 99 days ago
  • CI: add sharding (4 shards) to vllm-test-transformers job
    ydshieh committed 99 days ago
  • CI: add sentence-transformers to vllm-test-transformers install
    ydshieh committed 99 days ago
  • CI: add comment to vllm-test-transformers gpu_memory patch step
    ydshieh committed 99 days ago
  • CI: re-enable vllm-test-init, vllm-multimodal-processing, and inline test steps
    ydshieh committed 99 days ago
  • CI: revert vllm job to 2xl runner, drop gpu_memory patches, restore if: always()
    ydshieh committed 99 days ago
  • CI: move vllm and multimodal jobs to 8xl runner, restore gpu_memory patches
    ydshieh committed 99 days ago
  • CI: document why all vllm jobs use aws-m8i-8xl-cache runner
    ydshieh committed 99 days ago
  • CI: revert qwen3_asr test change, clean up comments
    ydshieh committed 99 days ago
  • CI: clarify why vllm-test-init uses 8xl runner
    ydshieh committed 99 days ago
  • CI: remove stale comments from gpu_memory patch steps
    ydshieh committed 99 days ago
Loading