Commits · 93208162753986f9449d3671d6a263dfc4f4381e · OpenDAS / dynamo · GitLab

08 Sep, 2025 10 commits
- refactor: standardize e2e tests across 3 frameworks (#2827) · 93208162
  Alec authored Sep 08, 2025
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
```
  93208162
- build: OPS-597: restructure sglang to follow container strategy structure (#2803) · f0cea269
  Tushar Sharma authored Sep 08, 2025
```
Signed-off-by: Tushar Sharma <tusharma@nvidia.com>
```
  f0cea269
- Update README.md (#2938) · 96064614
  Harry Kim authored Sep 08, 2025
```
Signed-off-by: Harry Kim <harry_kim@live.com>
```
  96064614
- chore: add preference logic for using tool-call and reasoning parsers for sglang (#2922) · 4a73b3c6
  Ayush Agarwal authored Sep 08, 2025
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  4a73b3c6
- chore: added utility to detect possible tool call start for a chunk (#2923) · b929e134
  Ayush Agarwal authored Sep 08, 2025
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  b929e134
- ci: Add concurrency check to auto cancel running actions. (#2438) · 5ea6b8d7
  Pavithra Vijayakrishnan authored Sep 08, 2025
```
Signed-off-by: Pavithra Vijayakrishnan <160681768+pvijayakrish@users.noreply.github.com>
```
  5ea6b8d7
- fix: fix typo in multinode example (#2931) · 6441ed36
  julienmancuso authored Sep 08, 2025
```
Signed-off-by: Julien Mancuso <jmancuso@nvidia.com>
```
  6441ed36
- fix: CI is broken with a deprecated dependency on pynvml (#2926) · fda1efe4
  Harrison Saturley-Hall authored Sep 08, 2025
```
Signed-off-by: Harrison Saturley-Hall <hsaturleyhal@nvidia.com>
Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: alec-flowers <aflowers@nvidia.com>
```
  fda1efe4
- ci: OPS-724: Move to ARC runners (#2904) · 41dacce7
  Dillon Cullinan authored Sep 08, 2025
```
Signed-off-by: Dillon Cullinan <dcullinan@nvidia.com>
```
  41dacce7
- fix: fix hermes tool call config (#2915) · f93a7cad
  Ayush Agarwal authored Sep 08, 2025
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  f93a7cad
05 Sep, 2025 12 commits
- chore: update CODEOWNERS for multimodal examples (#2878) · 1477f6ef
  Biswa Panda authored Sep 05, 2025
```
Signed-off-by: Biswa Panda <biswa.panda@gmail.com>
```
  1477f6ef
- fix: Load the tokenizer JSON once for chat and completions. (#2910) · cb5a657a
  Graham King authored Sep 05, 2025
```
Signed-off-by: Graham King <grahamk@nvidia.com>
```
  cb5a657a
- fix: use correct prompt template in agg_qwen.yaml (#2909) · 9ef13289
  GuanLuo authored Sep 05, 2025
```
Signed-off-by: GuanLuo <41310872+GuanLuo@users.noreply.github.com>
```
  9ef13289
- chore: add python codeowners team (#2908) · e0bafce1
  Anant Sharma authored Sep 05, 2025
```
Signed-off-by: Anant Sharma <anants@nvidia.com>
```
  e0bafce1
- feat: add trtllm example with config (#2895) · 0d9c8994
  julienmancuso authored Sep 05, 2025
```
Signed-off-by: Julien Mancuso <jmancuso@nvidia.com>
```
  0d9c8994
- docs: Update README.md (#2906) · 357efee3
  Harry Kim authored Sep 05, 2025
```
Signed-off-by: Harry Kim <harry_kim@live.com>
```
  357efee3
- build: OPS-757 Install a newer version of nsight-systems in devel container (#2891) · abf4cafc
  Tushar Sharma authored Sep 05, 2025
```
Signed-off-by: Tushar Sharma <tusharma@nvidia.com>
```
  abf4cafc
- refactor: Refactor discovery ModelManager to use `parking_lot::RwLock` (#2902) · 8065fe12
  Paul Hendricks authored Sep 05, 2025
```
Signed-off-by: Paul Hendricks <phendricks@nvidia.com>
```
  8065fe12
- feat: kvbm transfer context v2 (dis-598) (#2873) · 85b3dd44
  Ryan Olson authored Sep 04, 2025
```
Signed-off-by: Ryan Olson <ryanolson@users.noreply.github.com>
```
  85b3dd44
- revert: restore Dockerfile.vllm and build/run scripts to August 28 vers (#2892) · b6b3a767
  Keiven C authored Sep 04, 2025
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  b6b3a767
- feat: Router benchmarking (#2828) · 94f100bc
  Yan Ru Pei authored Sep 04, 2025
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  94f100bc
- docs: change docs to default port 8000 (#2876) · 1995ef9a
  Yan Ru Pei authored Sep 04, 2025
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  1995ef9a
04 Sep, 2025 14 commits
- fix: reduce nats stats query frequency (#2847) · 4df2e2d6
  Keiven C authored Sep 04, 2025
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  4df2e2d6
- test: add test for pre-deployment script (#2857) · f72dc01d
  Hongkuan Zhou authored Sep 04, 2025
```
Signed-off-by: hongkuanz <hongkuanz@nvidia.com>
```
  f72dc01d
- chore: add planner code owner (#2871) · 23ede721
  Hongkuan Zhou authored Sep 04, 2025
```
Signed-off-by: hongkuanz <hongkuanz@nvidia.com>
```
  23ede721
- fix: NIXL wheel missing from system packages in dev env (#2879) · 1fc4050a
  Harrison Saturley-Hall authored Sep 04, 2025
```
Signed-off-by: Harrison King Saturley-Hall <hsaturleyhal@nvidia.com>
```
  1fc4050a
- fix: router slot manager needs force expire requests (#2840) · 307d4039
  Yan Ru Pei authored Sep 04, 2025
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  307d4039
- fix: fix broken link in docs (#2870) · 432e8290
  julienmancuso authored Sep 04, 2025
```
Signed-off-by: Julien Mancuso <jmancuso@nvidia.com>
Co-authored-by: Dmitry Tokarev <dtokarev@nvidia.com>
```
  432e8290
- chore: add agg_qwen.yaml to multimodal deploy (#2872) · c403d18a
  GuanLuo authored Sep 04, 2025
```
Signed-off-by: GuanLuo <41310872+GuanLuo@users.noreply.github.com>
```
  c403d18a
- docs: note that DCO check is required to merge (#2747) · a4f35a24
  Harrison Saturley-Hall authored Sep 04, 2025
```
Signed-off-by: Harrison King Saturley-Hall <hsaturleyhal@nvidia.com>
```
  a4f35a24
- bug: In non-epp dynamo integration we do not provide the config file (#2862) · 4581fccb
  atchernych authored Sep 04, 2025
```
Signed-off-by: Anna Tchernych <atchernych@nvidia.com>
```
  4581fccb
- fix: always enable graceful shutdown as migration is not set in sglang (#2865) · eb1c9f24
  Hongkuan Zhou authored Sep 04, 2025
```
Signed-off-by: hongkuanz <hongkuanz@nvidia.com>
```
  eb1c9f24
- fix: Adding back runtime_config so that tool_parser work properly on … (#2855) · 9524a464
  zhongdaor-nv authored Sep 03, 2025
```
Co-authored-by: Zhongdao Ren <zhongdaor@ipp1-3309.ipp1a1.colossus.nvidia.com>
```
  9524a464
- refactor: refactor to be tree-based code + feat: framework check (#2799) · ab9d8f7b
  Keiven C authored Sep 03, 2025
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  ab9d8f7b
- chore: pass in mocker engine args directly in python cli + default frontend port to 8000 (#2853) · 3c7c1d64
  Yan Ru Pei authored Sep 03, 2025
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  3c7c1d64
- chore: frontend API changes for thinking budget (#2848) · 8d30753f
  Bhuvan Agrawal authored Sep 03, 2025
```
Signed-off-by: Bhuvan Agrawal <11240550+bhuvan002@users.noreply.github.com>
```
  8d30753f
03 Sep, 2025 4 commits

fix: dev target needs special permission, now VLLM version works for run.sh... · eb317cdf

Keiven C authored Sep 03, 2025

fix: dev target needs special permission, now VLLM version works for run.sh and Dev Container (#2822)
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>

eb317cdf

refactor: Split ModelType to ModelInput for request and response type;... · 27fad26f

Olga Andreeva authored Sep 03, 2025

refactor: Split ModelType to ModelInput for request and response type; ModelType for the supported workloads (#2714)
Signed-off-by: Guan Luo <gluo@nvidia.com>
Signed-off-by: GuanLuo <41310872+GuanLuo@users.noreply.github.com>
Co-authored-by: Guan Luo <gluo@nvidia.com>
Co-authored-by: GuanLuo <41310872+GuanLuo@users.noreply.github.com>

27fad26f

feat: workaround for bitnami brownout (#2843) · b97db875
julienmancuso authored Sep 03, 2025
```
Signed-off-by: Julien Mancuso <jmancuso@nvidia.com>
```
b97db875
feat: Add --custom-jinja-template argument to pass a custom chat template for vLLM (#2829) · c920cbd9
KrishnanPrash authored Sep 03, 2025
```
Signed-off-by: Krishnan Prashanth <kprashanth@nvidia.com>
```
c920cbd9