A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 8, 14:13 UTC
Bump `mamba-ssm` version in CI for Transformers v5 compatibility (#34233)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Harry Mellor committed
Feb 10, 2026 at 14:46 UTC
d0bc52056915e108c347aa4b5520e163e5c5b726
2 files changed
+4
-4
.buildkite/test-pipeline.yaml
+2
-2
@@ -862,7 +862,7 @@ steps:
862
commands:
863
# Install fast path packages for testing against transformers
864
# Note: also needed to run plamo2 model in vLLM
865
- - uv pip install --system --no-build-isolation 'git+https://github.com/state-spaces/mamba@v2.2.5'
865
+ - uv pip install --system --no-build-isolation 'git+https://github.com/state-spaces/mamba@v2.3.0'
866
- uv pip install --system --no-build-isolation 'git+https://github.com/Dao-AILab/causal-conv1d@v1.5.2'
867
# Shard hybrid language model tests
868
- pytest -v -s models/language/generation \
@@ -881,7 +881,7 @@ steps:
881
commands:
882
# Install fast path packages for testing against transformers
883
# Note: also needed to run plamo2 model in vLLM
884
- - uv pip install --system --no-build-isolation 'git+https://github.com/state-spaces/mamba@v2.2.5'
884
+ - uv pip install --system --no-build-isolation 'git+https://github.com/state-spaces/mamba@v2.3.0'
885
- uv pip install --system --no-build-isolation 'git+https://github.com/Dao-AILab/causal-conv1d@v1.5.2'
886
- pytest -v -s models/language/generation -m '(not core_model) and (not hybrid_model)'
887
.buildkite/test_areas/models_language.yaml
+2
-2
@@ -40,7 +40,7 @@ steps:
40
commands:
41
# Install fast path packages for testing against transformers
42
# Note: also needed to run plamo2 model in vLLM
43
- - uv pip install --system --no-build-isolation 'git+https://github.com/state-spaces/mamba@v2.2.5'
43
+ - uv pip install --system --no-build-isolation 'git+https://github.com/state-spaces/mamba@v2.3.0'
44
- uv pip install --system --no-build-isolation 'git+https://github.com/Dao-AILab/causal-conv1d@v1.5.2'
45
# Shard hybrid language model tests
46
- pytest -v -s models/language/generation -m hybrid_model --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB
@@ -56,7 +56,7 @@ steps:
56
commands:
57
# Install fast path packages for testing against transformers
58
# Note: also needed to run plamo2 model in vLLM
59
- - uv pip install --system --no-build-isolation 'git+https://github.com/state-spaces/mamba@v2.2.5'
59
+ - uv pip install --system --no-build-isolation 'git+https://github.com/state-spaces/mamba@v2.3.0'
60
- uv pip install --system --no-build-isolation 'git+https://github.com/Dao-AILab/causal-conv1d@v1.5.2'
61
- pytest -v -s models/language/generation -m '(not core_model) and (not hybrid_model)'
62