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 15, 07:15 UTC
[Torch Nightly]add missing dependency (#18770)
Signed-off-by: Yang Wang <elainewy@meta.com>
Yang Wang committed
Jun 4, 2025 at 21:56 UTC
25b918eee6438d4d014ad7c900c43325a6908fba
2 files changed
+10
-2
docker/Dockerfile.nightly_torch
+3
@@ -312,4 +312,7 @@ RUN --mount=type=cache,target=/root/.cache/uv \
312
# Logging to confirm the torch versions
313
RUN pip freeze | grep -E 'torch|xformers|vllm|flashinfer'
314
315
+# Logging to confirm all the packages are installed
316
+RUN pip freeze
317
+
318
#################### UNITTEST IMAGE #############################
requirements/nightly_torch_test.txt
+7
-2
@@ -9,7 +9,9 @@ pytest-shard
9
pytest-timeout
10
11
librosa # required by audio tests in entrypoints/openai
12
-sentence-transformers
12
+sentence-transformers # required for embedding tests
13
+transformers==4.51.3
14
+transformers_stream_generator # required for qwen-vl test
15
numba == 0.61.2; python_version > '3.9'
16
# testing utils
17
boto3
@@ -38,4 +40,7 @@ matplotlib # required for qwen-vl test
40
# required for Multi-Modal Models Test (Standard)
41
num2words # required for smolvlm test
42
pqdm
41
-timm # required for internvl test
\ No newline at end of file
43
+timm # required for internvl test
44
+
45
+schemathesis>=3.39.15 # Required for openai schema test.
46
+mteb>=1.38.11, <2 # required for mteb test