Skip to content

docs: add aggregated example turning on MTP with DeepSeek R1 #1421

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
richardhuo-nv merged 3 commits into from
Jun 7, 2025
Merged

docs: add aggregated example turning on MTP with DeepSeek R1 #1421

Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f50d990
add example for turning on MTP in aggregated serving
richardhuo-nv Jun 6, 2025
7f41cdb
fix the command and wording
richardhuo-nv Jun 6, 2025
4f115f4
add more notes
richardhuo-nv Jun 6, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
fix the command and wording
  • Loading branch information
@richardhuo-nv
richardhuo-nv committed Jun 6, 2025
commit 7f41cdb7d807c4aea1bca8feaa8e7ab72c6209f2
4 changes: 2 additions & 2 deletions examples/tensorrt_llm/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -121,12 +121,12 @@ dynamo serve graphs.disagg_router:Frontend -f ./configs/disagg_router.yaml
#### Aggregated serving with Multi-Token Prediction(MTP) and DeepSeek R1
```bash
cd /workspace/examples/tensorrt_llm
dynamo serve graphs.disagg_router:Frontend -f configs/deepseek_r1/mtp/mtp_agg.yaml
dynamo serve graphs.agg:Frontend -f configs/deepseek_r1/mtp/mtp_agg.yaml
```
Notes:
- There is a noticeable latency for the first two inference requests. Please send warm-up requests before starting the benchmark.
- Please keep the `cuda_graph_padding_enabled` setting as `false` in the model engine's configuration. There is a known bug, and the fix will be included in the next release of TensorRT-LLM.
- Disaggregated support for MTP in Dynamo + TensorRT-LLM is coming soon.
- MTP support for Disaggregation in Dynamo + TensorRT-LLM is coming soon.

#### Multi-Node Disaggregated Serving

Expand Down
Loading