From ae7374c32cb3a53c1ece433ed706102434bc8fa3 Mon Sep 17 00:00:00 2001 From: dhuangnm Date: Tue, 28 May 2024 18:40:04 -0400 Subject: [PATCH 1/4] bump up version and gate magic-wand version --- setup.py | 11 +++++------ vllm/__init__.py | 2 +- 2 files changed, 6 insertions(+), 7 deletions(-) diff --git a/setup.py b/setup.py index 7f87ee806057e..71538c1f94430 100644 --- a/setup.py +++ b/setup.py @@ -411,12 +411,6 @@ def _read_requirements(filename: str) -> List[str]: if not _is_neuron(): ext_modules.append(CMakeExtension(name="vllm._C")) -# UPSTREAM SYNC: needed for sparsity -_sparsity_deps = ["nm-magic-wand-nightly"] -nm_release_type = os.getenv(NM_RELEASE_TYPE) -if nm_release_type == 'RELEASE': - _sparsity_deps = ["nm-magic-wand"] - package_data = { "vllm": ["py.typed", "model_executor/layers/fused_moe/configs/*.json"] } @@ -458,6 +452,11 @@ def _read_requirements(filename: str) -> List[str]: python_requires=">=3.8", install_requires=get_requirements(), ext_modules=ext_modules, + # UPSTREAM SYNC: needed for sparsity + _sparsity_deps = [f"nm-magic-wand-nightly~={version}"] + nm_release_type = os.getenv(NM_RELEASE_TYPE) + if nm_release_type == 'RELEASE': + _sparsity_deps = [f"nm-magic-wand~={version}"] extras_require={ "tensorizer": ["tensorizer==2.9.0"], # UPSTREAM SYNC: required for sparsity diff --git a/vllm/__init__.py b/vllm/__init__.py index 0f658058f5c3f..5ba4845b4e6cc 100644 --- a/vllm/__init__.py +++ b/vllm/__init__.py @@ -10,7 +10,7 @@ from vllm.sampling_params import SamplingParams # UPSTREAM SYNC: use the current downstream. -__version__ = "0.3.0" +__version__ = "0.4.0" __all__ = [ "LLM", From 94c89364b30710094e69c3bfbe9ce0c1030659b2 Mon Sep 17 00:00:00 2001 From: dhuangnm Date: Wed, 29 May 2024 09:51:00 -0400 Subject: [PATCH 2/4] fix version and only gate for the release --- setup.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/setup.py b/setup.py index 71538c1f94430..b43d8d4615091 100644 --- a/setup.py +++ b/setup.py @@ -453,10 +453,12 @@ def _read_requirements(filename: str) -> List[str]: install_requires=get_requirements(), ext_modules=ext_modules, # UPSTREAM SYNC: needed for sparsity - _sparsity_deps = [f"nm-magic-wand-nightly~={version}"] + _sparsity_deps = [f"nm-magic-wand-nightly"] nm_release_type = os.getenv(NM_RELEASE_TYPE) if nm_release_type == 'RELEASE': - _sparsity_deps = [f"nm-magic-wand~={version}"] + # gate magic-wand version in nm-vllm for release; for nightly, we always install the latest + magic_wand_version_dep = "0.2.2" + _sparsity_deps = [f"nm-magic-wand~={magic_wand_version_dep}"] extras_require={ "tensorizer": ["tensorizer==2.9.0"], # UPSTREAM SYNC: required for sparsity From 22845dc2ffeb54955d1188849c50c79afade0d3e Mon Sep 17 00:00:00 2001 From: dhuangnm Date: Wed, 29 May 2024 10:43:08 -0400 Subject: [PATCH 3/4] fix a typo --- setup.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/setup.py b/setup.py index b43d8d4615091..f3c68528305d1 100644 --- a/setup.py +++ b/setup.py @@ -453,7 +453,7 @@ def _read_requirements(filename: str) -> List[str]: install_requires=get_requirements(), ext_modules=ext_modules, # UPSTREAM SYNC: needed for sparsity - _sparsity_deps = [f"nm-magic-wand-nightly"] + _sparsity_deps = ["nm-magic-wand-nightly"] nm_release_type = os.getenv(NM_RELEASE_TYPE) if nm_release_type == 'RELEASE': # gate magic-wand version in nm-vllm for release; for nightly, we always install the latest From 817f66a91facd22f8ee14baa9173666896fa980c Mon Sep 17 00:00:00 2001 From: dhuangnm Date: Wed, 29 May 2024 10:48:41 -0400 Subject: [PATCH 4/4] move back --- setup.py | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/setup.py b/setup.py index f3c68528305d1..bb563e6483710 100644 --- a/setup.py +++ b/setup.py @@ -411,6 +411,14 @@ def _read_requirements(filename: str) -> List[str]: if not _is_neuron(): ext_modules.append(CMakeExtension(name="vllm._C")) +# UPSTREAM SYNC: needed for sparsity +_sparsity_deps = ["nm-magic-wand-nightly"] +nm_release_type = os.getenv(NM_RELEASE_TYPE) +if nm_release_type == 'RELEASE': + # gate magic-wand version in nm-vllm for release; for nightly, we always install the latest + magic_wand_version_dep = "0.2.2" + _sparsity_deps = [f"nm-magic-wand~={magic_wand_version_dep}"] + package_data = { "vllm": ["py.typed", "model_executor/layers/fused_moe/configs/*.json"] } @@ -452,13 +460,6 @@ def _read_requirements(filename: str) -> List[str]: python_requires=">=3.8", install_requires=get_requirements(), ext_modules=ext_modules, - # UPSTREAM SYNC: needed for sparsity - _sparsity_deps = ["nm-magic-wand-nightly"] - nm_release_type = os.getenv(NM_RELEASE_TYPE) - if nm_release_type == 'RELEASE': - # gate magic-wand version in nm-vllm for release; for nightly, we always install the latest - magic_wand_version_dep = "0.2.2" - _sparsity_deps = [f"nm-magic-wand~={magic_wand_version_dep}"] extras_require={ "tensorizer": ["tensorizer==2.9.0"], # UPSTREAM SYNC: required for sparsity