minor: update flashinfer nightly

[zhyncs]

Motivation

Modifications

Checklist

• Format your code according to the Contributor Guide.
• Add unit tests as outlined in the Contributor Guide.
• Update documentation as needed, including docstrings or example tutorials.

[zhyncs]

There are two scenarios: one is a stable release, and the other is a nightly release. The first scenario uses cache because the script defaults to installing the stable version, making it very fast. For the second scenario, it depends on whether it's the first installation or not. It only triggers when users manually select a nightly release during PR testing, and if it's not the first time, it also uses the nightly cache, which is also very fast.

[zhyncs]

https://github.com/flashinfer-ai/flashinfer/tree/zhyncs/main
https://github.com/flashinfer-ai/flashinfer-nightly/releases/tag/0.1.6%2B51236c9

2024-12-16 14:55:59,006 - INFO - flashinfer.jit: Loading JIT ops: batch_decode_with_kv_cache_dtype_q_bf16_dtype_kv_bf16_dtype_o_bf16_dtype_idx_i32_head_dim_128_posenc_0_use_swa_False_use_logits_cap_False
/usr/local/lib/python3.10/dist-packages/torch/utils/cpp_extension.py:1964: UserWarning: TORCH_CUDA_ARCH_LIST is not set, all archs for visible cards are included for compilation. 
If this is not desired, please set os.environ['TORCH_CUDA_ARCH_LIST'].
  warnings.warn(
[rank0]: Traceback (most recent call last):
[rank0]:   File "/actions-runner/_work/sglang/sglang/python/sglang/srt/model_executor/cuda_graph_runner.py", line 207, in __init__
[rank0]:     self.capture()
[rank0]:   File "/actions-runner/_work/sglang/sglang/python/sglang/srt/model_executor/cuda_graph_runner.py", line [26](https://github.com/sgl-project/sglang/actions/runs/12355218604/job/34478328251#step:4:27)8, in capture
[rank0]:     ) = self.capture_one_batch_size(bs, forward)
[rank0]:   File "/actions-runner/_work/sglang/sglang/python/sglang/srt/model_executor/cuda_graph_runner.py", line 297, in capture_one_batch_size
[rank0]:     self.model_runner.attn_backend.init_forward_metadata_capture_cuda_graph(
[rank0]:   File "/actions-runner/_work/sglang/sglang/python/sglang/srt/layers/attention/flashinfer_backend.py", line 194, in init_forward_metadata_capture_cuda_graph
[rank0]:     self.indices_updater_decode.update(
[rank0]:   File "/actions-runner/_work/sglang/sglang/python/sglang/srt/layers/attention/flashinfer_backend.py", line 378, in update_single_wrapper
[rank0]:     self.call_begin_forward(
[rank0]:   File "/actions-runner/_work/sglang/sglang/python/sglang/srt/layers/attention/flashinfer_backend.py", line 478, in call_begin_forward
[rank0]:     wrapper.begin_forward(
[rank0]:   File "/usr/local/lib/python3.10/dist-packages/flashinfer/decode.py", line 788, in plan
[rank0]:     self._cached_module = get_batch_decode_module(
[rank0]:   File "/usr/local/lib/python3.10/dist-packages/flashinfer/decode.py", line 148, in get_batch_decode_module
[rank0]:     mod = gen_batch_decode_module(*args)
[rank0]:   File "/usr/local/lib/python3.10/dist-packages/flashinfer/jit/attention.py", line 173, in gen_batch_decode_module
[rank0]:     return load_cuda_ops(uri, source_paths)
[rank0]:   File "/usr/local/lib/python3.10/dist-packages/flashinfer/jit/core.py", line 110, in load_cuda_ops
[rank0]:     module = torch_cpp_ext.load(
[rank0]:   File "/usr/local/lib/python3.10/dist-packages/torch/utils/cpp_extension.py", line 1[31](https://github.com/sgl-project/sglang/actions/runs/12355218604/job/34478328251#step:4:32)4, in load
[rank0]:     return _jit_compile(
[rank0]:   File "/usr/local/lib/python3.10/dist-packages/torch/utils/cpp_extension.py", line 1721, in _jit_compile
[rank0]:     _write_ninja_file_and_build_library(
[rank0]:   File "/usr/local/lib/python3.10/dist-packages/torch/utils/cpp_extension.py", line 1803, in _write_ninja_file_and_build_library
[rank0]:     verify_ninja_availability()
[rank0]:   File "/usr/local/lib/python3.10/dist-packages/torch/utils/cpp_extension.py", line 1852, in verify_ninja_availability
[rank0]:     raise RuntimeError("Ninja is required to load C++ extensions")
[rank0]: RuntimeError: Ninja is required to load C++ extensions

[rank0]: During handling of the above exception, another exception occurred:

[rank0]: Traceback (most recent call last):
[rank0]:   File "/usr/lib/python3.10/runpy.py", line 196, in _run_module_as_main
[rank0]:     return _run_code(code, main_globals, None,
[rank0]:   File "/usr/lib/python3.10/runpy.py", line 86, in _run_code