Commits · 4c648b11e911f2bfdf7cd687f90e706a92296e18 · OpenDAS / dynamo

26 Feb, 2026 1 commit

refactor: move core logics of DPP -> AIC and support static profiling (#6285) · 4c648b11

Hongkuan Zhou authored Feb 25, 2026


Signed-off-by: hongkuanz <hongkuanz@nvidia.com>
Signed-off-by: Hannah Zhang <hannahz@nvidia.com>
Co-authored-by: hhzhang16 <54051230+hhzhang16@users.noreply.github.com>

4c648b11

25 Feb, 2026 3 commits
- chore: Upgrade to tensorrt-llm==1.3.0rc5 (#6579) · 8483e4a0
  Yuewei Na authored Feb 25, 2026
```
Signed-off-by: Yuewei Na <nv-yna@users.noreply.github.com>
Co-authored-by: Yuewei Na <nv-yna@users.noreply.github.com>
```
  8483e4a0
- fix: guarantee RouterRequestMetrics availability & documentation updates (#6558) · ff06b17e
  Keiven C authored Feb 25, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
```
  ff06b17e
- test: introduce multimodal benchmark toolkit (#6330) · 5a4c96db
  Qi Wang authored Feb 24, 2026
  
  5a4c96db
24 Feb, 2026 3 commits
- chore: use aic release/0.7.0 (#6494) · 4ea80e95
  Jason Zhou authored Feb 24, 2026
  
  4ea80e95
- feat: add --disaggregation-mode enum to vLLM backend (#6483) · 7893f268
  Alec authored Feb 23, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
```
  7893f268
- fix: fold scheduling into queue so backpressure actually works (#6470) · 0cb1d733
  Yan Ru Pei authored Feb 23, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
Signed-off-by: Yan Ru Pei <yanrpei@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
```
  0cb1d733
20 Feb, 2026 1 commit
- feat: default kv-events-config to empty (align with vLLM defaults) (#6404) · 7bbacce1
  Alec authored Feb 19, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
```
  7bbacce1
19 Feb, 2026 2 commits

feat: use embedding transfer classes for EPD (#6223) · a2a6917f
GuanLuo authored Feb 19, 2026
```
Signed-off-by: Guan Luo <41310872+GuanLuo@users.noreply.github.com>
```
a2a6917f

chore: Upgrade to tensorrt-llm==1.3.0rc3 (#6402) · 9a15730a

Yuewei Na authored Feb 19, 2026


Signed-off-by: Yuewei Na <nv-yna@users.noreply.github.com>
Co-authored-by: Yuewei Na <nv-yna@users.noreply.github.com>
Co-authored-by: dagil-nvidia <dagil@nvidia.com>

9a15730a

15 Feb, 2026 1 commit

feat: speculative prefill (#6230) · 4ede59a2

Yan Ru Pei authored Feb 15, 2026


Signed-off-by: PeaBrane <yanrpei@gmail.com>
Co-authored-by: Janelle Cai <jcai18@mit.edu>

4ede59a2

13 Feb, 2026 1 commit

feat: router priority queue (#6010) · bc514fbe

Yan Ru Pei authored Feb 12, 2026


Signed-off-by: PeaBrane <yanrpei@gmail.com>
Co-authored-by: ishandhanani <82981111+ishandhanani@users.noreply.github.com>

bc514fbe

12 Feb, 2026 2 commits
- feat: support AIC DGD gen call (WILL BREAK DGDR) (#6216) · a04b5631
  Hongkuan Zhou authored Feb 12, 2026
```
Signed-off-by: hongkuanz <hongkuanz@nvidia.com>
```
  a04b5631
- docs: migrate Fern docs from fern/ into docs/ (#6206) · 39d645e5
  Jonathan Tong authored Feb 11, 2026
```
Signed-off-by: Jont828 <jt572@cornell.edu>
```
  39d645e5
10 Feb, 2026 2 commits
- fix: scale synthesized data length correctly for expected cache hit stats (#6117) · 8707dc2c
  Karen Chung authored Feb 10, 2026
  
  8707dc2c
- fix: use actual service names for profiler logs and handle FileNotFoundError correctly (#6112) · 1aab7f6b
  hhzhang16 authored Feb 10, 2026
```
Signed-off-by: Hannah Zhang <hannahz@nvidia.com>
```
  1aab7f6b
09 Feb, 2026 1 commit

fix: profiler deployment timeout handling for MoE models (#6086) · 67329d10

MatejKosec authored Feb 09, 2026

Wrap wait_for_deployment_ready() in try/except TimeoutError for both prefill and decode profiling sweeps
On timeout: log error, record via add_profiling_error(), clean up the timed-out deployment, and continue to the next parallelization mapping
Previously, a single deployment timeout would crash the entire profiler job

67329d10

06 Feb, 2026 3 commits

feat: NAT telemetry conversion (#6022) · 5035447f
Yan Ru Pei authored Feb 05, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
5035447f
docs: cleanup of docs refactor for components, integrations, and features (#6019) · b19de4ed
dagil-nvidia authored Feb 05, 2026
```
Signed-off-by: Dan Gil <dagil@nvidia.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
```
b19de4ed

docs: Migrate router documentation to three-tier structure (#5979) · 80e7bafd

akshatha-k authored Feb 05, 2026


Signed-off-by: akshatha-k <akshutk@gmail.com>
Signed-off-by: dagil-nvidia <dagil@nvidia.com>
Signed-off-by: Dan Gil <dagil@nvidia.com>
Co-authored-by: dagil-nvidia <dagil@nvidia.com>
Co-authored-by: Cursor <cursoragent@cursor.com>

80e7bafd

05 Feb, 2026 3 commits
- fix: ignore benchmarks pytest for now (#5983) · 50e17783
  Yan Ru Pei authored Feb 05, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  50e17783
- fix: e2e aiperf profiling on NAT dataset (#5990) · 88efa735
  Yan Ru Pei authored Feb 04, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  88efa735
- feat: remove default model name in Profiler; validate for one of served model... · 7e4bc716
  hhzhang16 authored Feb 04, 2026
```
feat: remove default model name in Profiler; validate for one of served model name and model path in Profiler (#5950)
Signed-off-by: Hannah Zhang <hannahz@nvidia.com>
```
  7e4bc716
04 Feb, 2026 2 commits
- chore: use aiperf utils in prefix synthesizer (#5906) · 4fcee92f
  Yan Ru Pei authored Feb 04, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  4fcee92f
- fix: Add status file to prevent output-copier hang on failures (#5898) · 0268aea4
  hhzhang16 authored Feb 03, 2026
```
Signed-off-by: Hannah Zhang <hannahz@nvidia.com>
```
  0268aea4
03 Feb, 2026 2 commits
- chore: applying rolling hasher in prefix synthesizer (#5903) · cad453f2
  Yan Ru Pei authored Feb 02, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  cad453f2
- fix: check --served-model-name first before --model/--model-path (#5881) · 2b199540
  hhzhang16 authored Feb 02, 2026
```
Signed-off-by: Hannah Zhang <hannahz@nvidia.com>
```
  2b199540
29 Jan, 2026 1 commit
- chore: Upgrade to Tensorrt-LLM 1.3.0rc1 (#5700) · ba711cc1
  Tanmay Verma authored Jan 29, 2026
```
Co-authored-by: Pavithra Vijayakrishnan <160681768+pvijayakrish@users.noreply.github.com>
```
  ba711cc1
28 Jan, 2026 1 commit
- chore: add multimodal image benchmark scripts for performance evaluation (#5509) · 903f8184
  GuanLuo authored Jan 27, 2026
```
Signed-off-by: Guan Luo <41310872+GuanLuo@users.noreply.github.com>
```
  903f8184
27 Jan, 2026 2 commits
- feat: convert NAT trace to mooncake-style trace (#5675) · e00ada20
  Yan Ru Pei authored Jan 26, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  e00ada20
- fix: sglang version in build.sh and docs (#5665) · 546f1bbc
  Dmitry Tokarev authored Jan 26, 2026
```
Signed-off-by: Dmitry Tokarev <dtokarev@nvidia.com>
```
  546f1bbc
26 Jan, 2026 1 commit
- chore: use aic release/0.6.0 (#5600) · aecf0051
  Jason Zhou authored Jan 26, 2026
  
  aecf0051
23 Jan, 2026 1 commit
- chore: Bump TRTLLM to 1.2.0rc6.post2 (#5580) · 64ba7dd0
  jthomson04 authored Jan 22, 2026
```
Signed-off-by: jthomson04 <jwillthomson19@gmail.com>
```
  64ba7dd0
21 Jan, 2026 1 commit
- feat: warmup dataset for planner load predictor (#5529) · b0959cfd
  Hongkuan Zhou authored Jan 21, 2026
```
Signed-off-by: hongkuanz <hongkuanz@nvidia.com>
```
  b0959cfd
18 Jan, 2026 1 commit
- feat: track output tokens / blocks in the Router (optional) (#5452) · 050906b5
  Yan Ru Pei authored Jan 18, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  050906b5
16 Jan, 2026 3 commits
- fix: make it clear what features are only available 0.8.1 and on (#5492) · a2c7d0f9
  hhzhang16 authored Jan 16, 2026
```
Signed-off-by: Hannah Zhang <hannahz@nvidia.com>
```
  a2c7d0f9
- fix: include missing reqs in benchmark (#5477) · abe9d127
  Yan Ru Pei authored Jan 16, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  abe9d127
- chore: make precentile plots in router benchmark (#5476) · d83ab662
  Yan Ru Pei authored Jan 16, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  d83ab662
15 Jan, 2026 1 commit
- fix: use correct argument name max_isl in synthesize_requests (#5441) · ef09f718
  dagil-nvidia authored Jan 15, 2026
```
Signed-off-by: Dan Gil <dagil@nvidia.com>
```
  ef09f718
14 Jan, 2026 1 commit

feat: SLA Planner configuration support for DEP and TEP VLLM (#4783) · 7cfd966e

MatejKosec authored Jan 14, 2026

Extends the MOE planner profiler to support  TEP (tensor expert parallel) and DEP (data expert parallel) configs with the vllm backend

7cfd966e