Commits · 535528a5a110401a7d28931331a1da7d5f02d53e · OpenDAS / dynamo · GitLab

07 Jan, 2026 5 commits
- fix: Optimize TRTLLM multimodal request processing by reusing the tokenizer (#5217) · 535528a5
  Indrajit Bhosale authored Jan 07, 2026
  
  535528a5
- fix: return None for hidden stop tokens to prevent EOS in output (#5238) · 50af4cdc
  Chaitanya Malakala authored Jan 07, 2026
```
Signed-off-by: Chaitanya Malakala <106723863+chay1045@users.noreply.github.com>
```
  50af4cdc
- fix: KvCacheConfig Settings Lost When Publishing Events (#5198) · 66dfc494
  Indrajit Bhosale authored Jan 07, 2026
  
  66dfc494
- ci: Run the same tests across Github and Gitlab CI for premerge (#5082) · 996077f4
  Pavithra Vijayakrishnan authored Jan 07, 2026
```
Signed-off-by: pvijayakrish <pvijayakrish@nvidia.com>
```
  996077f4
- docs: update LoRA model name in docs (#5210) · 975870d5
  Biswa Panda authored Jan 06, 2026
  
  975870d5
06 Jan, 2026 20 commits
- fix: Fix consolidator tests by computing the number of requests needed to... · f7f2fb26
  Kyle McGill authored Jan 06, 2026
```
fix: Fix consolidator tests by computing the number of requests needed to trigger an expected pass (#5220)
```
  f7f2fb26
- docs: kube-based service discovery (#5211) · c92a422b
  mohammedabdulwahhab authored Jan 06, 2026
```
Signed-off-by: mohammedabdulwahhab <furkhan324@berkeley.edu>
```
  c92a422b
- fix: update vLLM to 0.13.0 with API compatibility fixes (#5222) · a1333a8d
  Alec authored Jan 06, 2026
```
Signed-off-by: Vasilis Vagias <vvagias@nvidia.com>
Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: Vasilis Vagias <vvagias@nvidia.com>
```
  a1333a8d
- feat(operator): have Scaling Adapter disabled by default (#5180) · 110cd568
  Thomas Montfort authored Jan 06, 2026
  
  110cd568
- fix: add sweep range for moe dgdr example (#5225) · a8213787
  Hongkuan Zhou authored Jan 06, 2026
```
Signed-off-by: hongkuanz <hongkuanz@nvidia.com>
```
  a8213787
- refactor: extract changed-files logic into reusable composite action (#5216) · 17528f81
  Alec authored Jan 06, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
```
  17528f81
- fix: assign unique ports per vLLM worker to avoid ZMQ bind conflicts (#5224) · 19b5917c
  Alec authored Jan 06, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
```
  19b5917c
- ci: add uncovered files check and improve filter coverage (#5214) · 15a2e8de
  Alec authored Jan 06, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
```
  15a2e8de
- docs: fix tracing doc ZMQ port conflict (#5200) · 194128ec
  Keiven C authored Jan 06, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  194128ec
- docs: Update TensorRT-LLM install prerequisites for 1.2.0rc5 (#5194) · e8c87f17
  Neelay Shah authored Jan 06, 2026
```
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
```
  e8c87f17
- fix: add endpointslices RBAC permission for Planner service account (#5195) · c90b87e2
  hhzhang16 authored Jan 06, 2026
```
Signed-off-by: Hannah Zhang <hannahz@nvidia.com>
```
  c90b87e2
- chore: Remove deprecated Earthfile (#5206) · 11d7d3f5
  Neelay Shah authored Jan 06, 2026
```
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
```
  11d7d3f5
- build: build.sh enhancement (#5053) · 96e42f72
  Ran Rubin authored Jan 06, 2026
  
  96e42f72
- fix: handle concurrent load lora calls (#5184) · 07721d1c
  Biswa Panda authored Jan 05, 2026
  
  07721d1c
- feat: expected output tokens plumbing (#5181) · d9cc6f6b
  Yan Ru Pei authored Jan 05, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  d9cc6f6b
- fix: prevent deadlock in etcd when flushing initial keys (#5091) · f56483cb
  mohammedabdulwahhab authored Jan 05, 2026
```
Signed-off-by: mohammedabdulwahhab <furkhan324@berkeley.edu>
```
  f56483cb
- feat: add --no-framework-check option to sanity_check (#5065) · 651569ff
  Keiven C authored Jan 05, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  651569ff
- ci: switch from dorny/paths-filter to tj-actions/changed-files (#5160) · c5760f6a
  Alec authored Jan 05, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
```
  c5760f6a
- fix: increase gpu-memory-utilization to 0.95 for dsr1_dep (#5144) · e8953558
  Alec authored Jan 05, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
```
  e8953558
- fix: block on notification of at least one runtime config (#5191) · cdeda221
  Yan Ru Pei authored Jan 05, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  cdeda221
05 Jan, 2026 11 commits
- fix: record cached_tokens metric for non-streaming requests (#5193) · e7918716
  Keiven C authored Jan 05, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  e7918716
- fix: the error when encountering pure text conversations (#5088) · ea950355
  shpgy-shpgy authored Jan 06, 2026
```
Signed-off-by: shpgy-shpgy <875664365@qq.com>
```
  ea950355
- fix: Add docker syntax flags for trtllm and sgl containers (#5182) · 5c225145
  jthomson04 authored Jan 05, 2026
```
Signed-off-by: jthomson04 <jwillthomson19@gmail.com>
```
  5c225145
- test: Resolving wrong model for testing (#5135) · 99ab75b8
  Olga Andreeva authored Jan 05, 2026
```
Signed-off-by: Olga Andreeva <oandreeva@nvidia.com>
Signed-off-by: Olga Andreeva <124622579+oandreeva-nv@users.noreply.github.com>
Co-authored-by: GuanLuo <41310872+GuanLuo@users.noreply.github.com>
```
  99ab75b8
- fix: Allow requests to include audio contents before text (#5143) · 2b157d89
  Tanmay Verma authored Jan 05, 2026
  
  2b157d89
- fix: trtllm has moved to version 1.2+ to enable cuda graph testing; fixed path... · 6be9c9a4
  Kyle McGill authored Jan 05, 2026
```
fix: trtllm has moved to version 1.2+ to enable cuda graph testing; fixed path to config.yaml (#5133)
```
  6be9c9a4
- fix: remove old docs and unify model paths (#5179) · 5a158552
  ishandhanani authored Jan 05, 2026
  
  5a158552
- fix: fix lora script name in docs and update setup_minio.sh script to work... · 7b5cdc42
  Biswa Panda authored Jan 05, 2026
```
fix: fix lora script name in docs and update setup_minio.sh script to work across HF cli package versions (#5153)
```
  7b5cdc42
- fix(DYN-1644): Add launch screen to sglang runtime (#5173) · fcddc352
  Dillon Cullinan authored Jan 05, 2026
```
Signed-off-by: Dillon Cullinan <dcullinan@nvidia.com>
```
  fcddc352
- fix: Pin GNU gcc version to 14 to build CUDA (#5145) · 4cfd603c
  jh-nv authored Jan 05, 2026
```
Signed-off-by: Jie Hao <jihao@nvidia.com>
Co-authored-by: Yan Ru Pei <yanrpei@gmail.com>
```
  4cfd603c
- feat: support PVC model cache in profiler (#5124) · fbe6bb0a
  Hongkuan Zhou authored Jan 05, 2026
```
Signed-off-by: hongkuanz <hongkuanz@nvidia.com>
```
  fbe6bb0a
04 Jan, 2026 3 commits
- fix: sglang metrics and prefill router fix (#5147) · 007c5b60
  ishandhanani authored Jan 04, 2026
  
  007c5b60
- chore: remove unused codes (#5151) · e5502e78
  Kante Yin authored Jan 04, 2026
```
Signed-off-by: kerthcet <kerthcet@gmail.com>
Co-authored-by: Yan Ru Pei <yanrpei@gmail.com>
```
  e5502e78
- fix: block on KV subscriber initialization (#5149) · 43ba25bc
  Yan Ru Pei authored Jan 03, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  43ba25bc
03 Jan, 2026 1 commit
- feat: add continuous_usage_stats option for per-chunk usage (#5139) · 6bfb41de
  GuanLuo authored Jan 03, 2026
```
Signed-off-by: Guan Luo <gluo@nvidia.com>
```
  6bfb41de