diff --git a/tensorrt_llm/_torch/speculative/mtp_dynamic_tree.py b/tensorrt_llm/_torch/speculative/mtp_dynamic_tree.py index 5dca84abae8e..053d23362f98 100644 --- a/tensorrt_llm/_torch/speculative/mtp_dynamic_tree.py +++ b/tensorrt_llm/_torch/speculative/mtp_dynamic_tree.py @@ -636,7 +636,7 @@ def _relocate_kv_eagerly(self, attn_metadata, batch_size): # Top-level forward # # ------------------------------------------------------------------ # @nvtx_range("mtp_dyn.forward") - def forward( + def _forward_impl( self, input_ids, position_ids, diff --git a/tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_1k1k_con2048_ctx1_tp1_gen1_dep2_eplb0_mtp0_ccb-NIXL.yaml b/tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_1k1k_con2048_ctx1_tp1_gen1_dep2_eplb0_mtp0_ccb-NIXL.yaml index 31d9830a3cbf..e027c0919652 100644 --- a/tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_1k1k_con2048_ctx1_tp1_gen1_dep2_eplb0_mtp0_ccb-NIXL.yaml +++ b/tests/scripts/perf-sanity/disaggregated/gb200_gpt-oss-120b-fp4_1k1k_con2048_ctx1_tp1_gen1_dep2_eplb0_mtp0_ccb-NIXL.yaml @@ -66,7 +66,6 @@ worker_config: cache_transceiver_config: max_tokens_in_buffer: 1024 backend: NIXL - transceiver_runtime: PYTHON disable_overlap_scheduler: true num_postprocess_workers: 4 stream_interval: 20 @@ -91,5 +90,4 @@ worker_config: cache_transceiver_config: max_tokens_in_buffer: 1024 backend: NIXL - transceiver_runtime: PYTHON disable_overlap_scheduler: true