anodyine/vllm-omni
1
0 Fork 0
Code Issues 4 Pull Requests 2 Actions Packages Projects Releases Wiki Activity

[Bugfix][NPU] Add _model_forward for ModelRunner (#505)

Browse Source 
Signed-off-by: gcanlin <canlinguosdu@gmail.com>
This commit is contained in:
Canlin Guo authored and GitHub committed 2025-12-28 04:09:34 +08:00
1 parent df4909630c
commit be81443c8a
3 files changed
+33 -14
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

vllm_omni/model_executor/stage_configs/npu/qwen3_omni_moe.yaml
+1 -1
View File
@@ -79,7 +79,7 @@ stage_args:
engine_output_type: audio # Final output: audio waveform
gpu_memory_utilization: 0.1
distributed_executor_backend: "mp"
max_num_batched_tokens: 4096
max_num_batched_tokens: 1000000
hf_config_name: thinker_config
engine_input_source: [1]
custom_process_input_func: vllm_omni.model_executor.stage_input_processors.qwen3_omni.talker2code2wav