[Feature] support new sampler backend with triton (#7639)

* support new sampler backend with triton

* test case

* fix the test case

Author: zeroRains

.flake8

@@ -5,3 +5,4 @@ max-line-length = 119
 # E402: module level import not at top of file
 per-file-ignores =
     __init__.py:F401,F403,E402
+    fastdeploy/model_executor/layers/sample/ops/top_k_top_p_triton.py:E241,E121,E131,E266

fastdeploy/envs.py

@@ -66,7 +66,7 @@ def _validate_split_kv_size(value: int) -> int:
     # Set attention backend. "NATIVE_ATTN", "APPEND_ATTN"
     # and "MLA_ATTN" can be set currently.
     "FD_ATTENTION_BACKEND": lambda: os.getenv("FD_ATTENTION_BACKEND", "APPEND_ATTN"),
-    # Set sampling class. "base", "base_non_truncated", "air" and "rejection" can be set currently.
+    # Set sampling class. "base", "base_non_truncated", "air", "rejection" and "triton" can be set currently.
     "FD_SAMPLING_CLASS": lambda: os.getenv("FD_SAMPLING_CLASS", "base"),
     # Set moe backend."cutlass","marlin", "triton", "flashinfer-cutlass", "flashinfer-cutedsl" and "flashinfer-trtllm" can be set currently.
     "FD_MOE_BACKEND": lambda: os.getenv("FD_MOE_BACKEND", "cutlass"),