Commit ·
31ed992
1
Parent(s): 4ae4ae8
llama-cpp-python: prebuilt CPU wheel (fix build OOM)
Browse files- requirements.txt +2 -0
requirements.txt
CHANGED
|
@@ -4,6 +4,8 @@ gradio>=6.0
|
|
| 4 |
httpx>=0.27.0
|
| 5 |
huggingface-hub>=0.23.0
|
| 6 |
spaces
|
|
|
|
|
|
|
| 7 |
llama-cpp-python
|
| 8 |
transformers>=5.4.0
|
| 9 |
torch==2.11.0
|
|
|
|
| 4 |
httpx>=0.27.0
|
| 5 |
huggingface-hub>=0.23.0
|
| 6 |
spaces
|
| 7 |
+
# prebuilt CPU wheel for llama-cpp-python — avoids the slow from-source CMake compile on the Space:
|
| 8 |
+
--extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cpu
|
| 9 |
llama-cpp-python
|
| 10 |
transformers>=5.4.0
|
| 11 |
torch==2.11.0
|