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, 22:15 UTC
[TPU] use torch2.6 with whl package (#13860)
Signed-off-by: Chenyaaang <llccyy1212@gmail.com>
Chenyaaang committed
Feb 26, 2025 at 05:18 UTC
b27122acc29f800e833048311ace650257db20b7
1 file changed
+3
-1
requirements-tpu.txt
+3
-1
@@ -17,7 +17,9 @@ ray[default]
17
--find-links https://storage.googleapis.com/libtpu-releases/index.html
18
--find-links https://storage.googleapis.com/jax-releases/jax_nightly_releases.html
19
--find-links https://storage.googleapis.com/jax-releases/jaxlib_nightly_releases.html
20
-torch==2.6.0.dev20241216+cpu
20
+torch @ https://download.pytorch.org/whl/nightly/cpu/torch-2.6.0.dev20241216%2Bcpu-cp39-cp39-linux_x86_64.whl ; python_version == "3.9"
21
+torch @ https://download.pytorch.org/whl/nightly/cpu/torch-2.6.0.dev20241216%2Bcpu-cp310-cp310-linux_x86_64.whl ; python_version == "3.10"
22
+torch @ https://download.pytorch.org/whl/nightly/cpu/torch-2.6.0.dev20241216%2Bcpu-cp311-cp311-linux_x86_64.whl ; python_version == "3.11"
23
torch_xla[tpu, pallas] @ https://storage.googleapis.com/pytorch-xla-releases/wheels/tpuvm/torch_xla-2.7.0.dev20250124-cp39-cp39-linux_x86_64.whl ; python_version == "3.9"
24
torch_xla[tpu, pallas] @ https://storage.googleapis.com/pytorch-xla-releases/wheels/tpuvm/torch_xla-2.7.0.dev20250124-cp310-cp310-linux_x86_64.whl ; python_version == "3.10"
25
torch_xla[tpu, pallas] @ https://storage.googleapis.com/pytorch-xla-releases/wheels/tpuvm/torch_xla-2.7.0.dev20250124-cp311-cp311-linux_x86_64.whl ; python_version == "3.11"