[LLM] First commit the llm deployment code

test/operators/test_air_topp_sampling.py

@@ -0,0 +1,51 @@
+# Copyright (c) 2024 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+""" UT for air_topp_sampling kernel """
+
+import paddle
+import unittest
+import numpy as np
+import fastdeploy.model_executor.ops.gpu
+
+
+class Test(unittest.TestCase):
+    def setUp(self):
+        """
+        Initialize.
+        """
+        paddle.seed(2024)
+        np.random.seed(42)
+        print(paddle.device.cuda.get_device_properties())
+        print(paddle.__git_commit__)
+
+    def test_air_topp_sampling(self):
+        """
+        Check air_topp_sampling output with paddle.tensor.top_p_sampling.
+        """
+        prop = paddle.device.cuda.get_device_properties()
+        cc = prop.major * 10 + prop.minor
+        if cc < 89:
+            self.skipTest("air_topp_sampling only support sm89+")
+        x = paddle.randn([1, 100])
+        x = paddle.nn.functional.softmax(x)
+        x = paddle.cast(x, "float32")
+        top_ps = paddle.to_tensor(np.random.uniform(0, 1, [1]).astype(np.float32))
+        out = fastdeploy.model_executor.ops.gpu.air_topp_sampling(
+            x.cuda(), top_ps.cuda(), None, None, seed=0, k=1, mode="truncated"
+        )
+
+
+if __name__ == "__main__":
+    unittest.main()