Code

[ci][gh200] dockerfile clean up#11351

Code

Merged

youkaichao merged 27 commits intomainvllm-project/vllm:mainfrom gh200_cleanvllm-project/vllm:gh200_cleanCopy head branch name to clipboardDec 20, 2024

+51-37

Commits

Commits on Nov 20, 2024

• Update Docker for aarch64 builds

  Show description for d2cbe42
  drikster80

  committed
  d2cbe42

  Copy full SHA for d2cbe42

• Update docs for arm64 docker builds & GH200 example

  Show description for 29ed358
  drikster80

  committed
  29ed358

  Copy full SHA for 29ed358

• Fix FLASHINFER not installing by default on x86_64

  Show description for f2635a4
  drikster80

  committed
  f2635a4

  Copy full SHA for f2635a4

Commits on Dec 4, 2024

• Removed build of flash-attn. Added libnccl2

  drikster80

  committed
  ccef455

  Copy full SHA for ccef455

Commits on Dec 15, 2024

• simplified Dockerfile build process for ARM64 systems

  cennn

  committed
  8b3c0b3

  Copy full SHA for 8b3c0b3

• bitsandbytes>=0.45.0

  cennn

  committed
  69bb567

  Copy full SHA for 69bb567

Commits on Dec 16, 2024

• Resolve merge conflict in Dockerfile

  cennn

  committed
  600b805

  Copy full SHA for 600b805

• rm needless kmod/libnccl2 && rm needless build from source for triton/causal-conv1d/mamba/flashinfer/bitsandbytes

  cennn

  committed
  d5d2911

  Copy full SHA for d5d2911

• update Memory usage && rename image && pinned version for torch torchvision

  cennn

  committed
  075ab9d

  Copy full SHA for 075ab9d

Commits on Dec 18, 2024

• Merge remote-tracking branch 'origin/main'

  Show description for 0ddc464
  cennn

  committed
  0ddc464

  Copy full SHA for 0ddc464

• rm requirements-cuda-arm64.txt && rm redundant torch install/uninstall && rm arm64 platform condition in requirements && add python3 use_existing_torch.py before build in run-gh200-test.sh

  cennn

  committed
  34e7140

  Copy full SHA for 34e7140

• fix docs --platform "linux/arm64"

  cennn

  committed
  c65bf72

  Copy full SHA for c65bf72

• rm docs blanks

  cennn

  committed
  6bddcd4

  Copy full SHA for 6bddcd4

• add torch nightly build in run-gh200-test.sh

  cennn

  committed
  6b80175

  Copy full SHA for 6b80175

• edit doc note && rm build-arg max_jobs nvcc_threads

  cennn

  committed
  93b82d7

  Copy full SHA for 93b82d7

• add args max_jobs nvcc_threads to speed up && install pytorch first && outlines == 0.1.11 # Requires pytorch && rm torch install before docker build

  cennn

  committed
  d026489

  Copy full SHA for d026489

• update vllm-gh200-openai Image size: 3.23GB

  cennn

  committed
  fb1759e

  Copy full SHA for fb1759e

Commits on Dec 19, 2024

• add debug "python3 -m pip list | grep torch || true" && compressed-tensors requires pytorch

  cennn

  committed
  62c8e29

  Copy full SHA for 62c8e29

• rm "python3 -m pip list | grep torch || true" && add cloudpickle && update image size

  cennn

  committed
  dd483d9

  Copy full SHA for dd483d9

• rm sudo in doc

  cennn

  committed
  022eaf9

  Copy full SHA for 022eaf9

Commits on Dec 20, 2024

• avoid installing python

  Show description for c61c579
  youkaichao

  committed
  c61c579

  Copy full SHA for c61c579

• add comments

  Show description for 5c056e6
  youkaichao

  committed
  5c056e6

  Copy full SHA for 5c056e6

• fix comments

  Show description for 383f799
  youkaichao

  committed
  383f799

  Copy full SHA for 383f799

• doc change

  Show description for fccb985
  youkaichao

  committed
  fccb985

  Copy full SHA for fccb985

• fix docs

  Show description for 67129a7
  youkaichao

  committed
  67129a7

  Copy full SHA for 67129a7

• fix docs

  Show description for f7f54f6
  youkaichao

  committed
  f7f54f6

  Copy full SHA for f7f54f6

• fix docs

  Show description for 7dbe69e
  youkaichao

  committed
  7dbe69e

  Copy full SHA for 7dbe69e