Skip to content

Commit 82b6f95

Browse files
committed
Bump vllm to 0.11.2, torch to 2.9, transformers to 4.57.1
Signed-off-by: Yi-Fu Wu <[email protected]>
1 parent 4a59436 commit 82b6f95

File tree

3 files changed

+775
-311
lines changed

3 files changed

+775
-311
lines changed
Submodule Automodel updated 426 files

pyproject.toml

Lines changed: 6 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -18,11 +18,11 @@ license = { text = "Apache 2.0" }
1818
dependencies = [
1919
"setuptools",
2020
"ninja", # for flash-attn parallel build
21-
"torch==2.8.0",
21+
"torch==2.9.0",
2222
"triton; sys_platform == 'linux' and (platform_machine == 'x86_64' or platform_machine == 'aarch64')",
2323
"colored==2.2.3",
2424
"ray[default]==2.49.2",
25-
"transformers>=4.55.4",
25+
"transformers>=4.57.1",
2626
"wandb",
2727
"numpy",
2828
"datasets>=4.0.0",
@@ -57,7 +57,7 @@ automodel = [
5757
# Flash-attn version should be selected to satisfy both TE + vLLM requirements (xformers in particular)
5858
# https://github.com/NVIDIA/TransformerEngine/blob/v2.3/transformer_engine/pytorch/attention/dot_product_attention/utils.py#L108
5959
# https://github.com/facebookresearch/xformers/blob/8354497deb2c04c67fbb2e2ad911e86530da0e90/xformers/ops/fmha/flash.py#L76
60-
"vllm==0.11.0", # Remove this once https://github.com/NVIDIA-NeMo/RL/issues/811 resolved
60+
"vllm==0.11.2", # Remove this once https://github.com/NVIDIA-NeMo/RL/issues/811 resolved
6161
"flash-attn==2.8.1",
6262
"mamba-ssm",
6363
"causal-conv1d",
@@ -69,7 +69,7 @@ vllm = [
6969
# sudo apt-get update
7070
# sudo apt-get install libibverbs-dev
7171
"deep_ep @ git+https://github.com/deepseek-ai/DeepEP.git@e3908bf5bd0cc6265bcb225d15cd8c996d4759ef",
72-
"vllm==0.11.0",
72+
"vllm==0.11.2",
7373
"num2words>=0.5.14",
7474
# Remove this once https://github.com/NVIDIA-NeMo/RL/issues/501 resolved
7575
"flash-attn==2.8.1",
@@ -92,7 +92,7 @@ mcore = [
9292
"megatron-core",
9393
"megatron-bridge",
9494
# Remove this once https://github.com/NVIDIA-NeMo/RL/issues/501 resolved
95-
"vllm==0.11.0",
95+
"vllm==0.11.2",
9696
# Flash-attn version should be selected to satisfy both TE + vLLM requirements (xformers in particular)
9797
# https://github.com/NVIDIA/TransformerEngine/blob/v2.3/transformer_engine/pytorch/attention/dot_product_attention/utils.py#L108
9898
# https://github.com/facebookresearch/xformers/blob/8354497deb2c04c67fbb2e2ad911e86530da0e90/xformers/ops/fmha/flash.py#L76
@@ -105,7 +105,7 @@ penguin = ["penguin"]
105105
# This is a default group so that we install these even with bare `uv sync`
106106
build = [
107107
# Build requirement for TE
108-
"torch==2.8.0",
108+
"torch==2.9.0",
109109
# Build requirement for TE
110110
"setuptools",
111111
"packaging",

0 commit comments

Comments
 (0)