A high-throughput and memory-efficient inference and serving engine for LLMs
[Doc] Fix uv dependency resolution failure for setuptools during CPU source builds (x86 & ARM) (#45412)
Signed-off-by: midas <the.anon.github@gmail.com>
midas committed
Jun 13, 2026 at 11:48 UTC
0d29612292c6b1e312af42ac00cf649af16a438b
2 files changed
+4
-4
docs/getting_started/installation/cpu.arm.inc.md
+2
-2
@@ -96,8 +96,8 @@ cd vllm_source
96
Third, install required dependencies:
97
98
```bash
99
-uv pip install -r requirements/build/cpu.txt --torch-backend cpu
100
-uv pip install -r requirements/cpu.txt --torch-backend cpu
99
+uv pip install -r requirements/build/cpu.txt --torch-backend cpu --index-strategy unsafe-best-match
100
+uv pip install -r requirements/cpu.txt --torch-backend cpu --index-strategy unsafe-best-match
101
```
102
103
??? console "pip"
docs/getting_started/installation/cpu.x86.inc.md
+2
-2
@@ -88,8 +88,8 @@ cd vllm_source
88
Install the required dependencies:
89
90
```bash
91
-uv pip install -r requirements/build/cpu.txt --torch-backend cpu
92
-uv pip install -r requirements/cpu.txt --torch-backend cpu
91
+uv pip install -r requirements/build/cpu.txt --torch-backend cpu --index-strategy unsafe-best-match
92
+uv pip install -r requirements/cpu.txt --torch-backend cpu --index-strategy unsafe-best-match
93
```
94
95
??? console "pip"