fix

Author: co63oc

tests/operators/test_eagle_get_hidden_states.py

@@ -60,7 +60,7 @@ def test_eagle_get_hidden_states(self):
             base_model_seq_lens_encoder_tensor,
             actual_draft_token_num,
         )
-        out_ref = np.array([6, 4, 3, 3], dtype=np.int32)
+        out_ref = np.array([6, 4, 3, 3], dtype=np.float16)
         np.testing.assert_allclose(gpu_out.numpy()[0][0:4], out_ref)
 
 

tests/operators/test_eagle_get_self_hidden_states.py

@@ -0,0 +1,97 @@
+# Copyright (c) 2025 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.
+
+import unittest
+
+import numpy as np
+import paddle
+
+from fastdeploy.model_executor.ops.gpu import eagle_get_self_hidden_states
+
+
+def computeOrder(last_seq_lens_this_time, seq_lens_this_time, step_idx, src_map, bsz):
+    in_offset = 0
+    out_offset = 0
+    for i in range(bsz):
+        cur_seq_lens_this_time = seq_lens_this_time[i]
+        cur_last_seq_lens_this_time = last_seq_lens_this_time[i]
+        if step_idx[i] == 1 and cur_seq_lens_this_time > 0:
+            in_offset += 1
+            src_map[out_offset] = in_offset - 1
+            out_offset += 1
+        elif cur_seq_lens_this_time > 0:
+            in_offset += cur_last_seq_lens_this_time
+            src_map[out_offset] = in_offset - 1
+            out_offset += 1
+        else:
+            if step_idx[i] == 1:
+                in_offset += 1 if cur_last_seq_lens_this_time > 0 else 0
+            else:
+                in_offset += cur_last_seq_lens_this_time
+
+    return (out_offset, src_map)
+
+
+def rebuildSelfHiddenStatesKernel(input, src_map, out, dim_embed, elem_cnt):
+    for elem_id in range(elem_cnt):
+        output_token_idx = elem_id // dim_embed
+        input_token_idx = src_map[output_token_idx]
+        offset = elem_id % dim_embed
+        out[output_token_idx * dim_embed + offset] = input[input_token_idx * dim_embed + offset]
+    return out
+
+
+def ref_eagle_get_self_hidden_states(input, last_seq_lens_this_time, seq_lens_this_time, step_idx):
+    input_token_num = input.shape[0]
+    dim_embed = input.shape[1]
+    bsz = seq_lens_this_time.shape[0]
+    src_map = np.full(input_token_num, -1, seq_lens_this_time.dtype)
+    output_token_num, src_map = computeOrder(last_seq_lens_this_time, seq_lens_this_time, step_idx, src_map, bsz)
+    out = np.full([output_token_num * dim_embed], -1, input.dtype)
+    elem_cnt = output_token_num * dim_embed
+    out = rebuildSelfHiddenStatesKernel(input, src_map, out, dim_embed, elem_cnt)
+    out = out.reshape([output_token_num, dim_embed])
+    return out
+
+
+class TestEagleGetSelfHiddenStates(unittest.TestCase):
+    def test_eagle_get_self_hidden_states(self):
+        paddle.seed(2023)
+        np.random.seed(2023)
+        bs = np.random.randint(1, 8 + 1, dtype=np.int32)
+        input_token_num = np.random.randint(2 * 1024, 4 * 1024 + 1, dtype=np.int32)
+        dim_embed = np.array(1024, dtype=np.int32)
+
+        last_seq_lens_this_time = np.random.randint(0, input_token_num // bs, bs, dtype=np.int32)
+        seq_lens_this_time = np.random.randint(0, input_token_num // bs, bs, dtype=np.int32)
+        step_idx = np.arange(0, bs, dtype=np.int32)
+
+        last_seq_lens_this_time_tensor = paddle.to_tensor(last_seq_lens_this_time, dtype=paddle.int32)
+        seq_lens_this_time_tensor = paddle.to_tensor(seq_lens_this_time, dtype=paddle.int32)
+        step_idx_tensor = paddle.to_tensor(step_idx, dtype=paddle.int64)
+
+        input = np.random.randint(0, 10, (input_token_num, dim_embed), dtype=np.int32)
+        input_tensor = paddle.to_tensor(input, dtype=paddle.float16)
+        gpu_out = eagle_get_self_hidden_states(
+            input_tensor,
+            last_seq_lens_this_time_tensor,
+            seq_lens_this_time_tensor,
+            step_idx_tensor,
+        )
+        out_ref = np.array([5, 4, 2, 8], dtype=np.float16)
+        np.testing.assert_allclose(gpu_out.numpy()[0][0:4], out_ref)
+
+
+if __name__ == "__main__":
+    unittest.main()