Commits · 0a517dc05a29aca7a1af1157ba244475607676e3 · OpenDAS / dynamo

10 Dec, 2025 9 commits

fix: remove pod normalization logic (#4853) · 0a517dc0
Julien Mancuso authored Dec 10, 2025
```
Signed-off-by: Julien Mancuso <jmancuso@nvidia.com>
```
0a517dc0
ci: re enable sgl unit tests and remove pyproject.toml from runtime image (#4803) · cf269dba
Anant Sharma authored Dec 10, 2025
```
Signed-off-by: Anant Sharma <anants@nvidia.com>
```
cf269dba

fix: sccache compiler launcher errors when sccache is disabled (#4846) · cc2ee1ed

Keiven C authored Dec 10, 2025


Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Signed-off-by: Dillon Cullinan <dcullinan@nvidia.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Dillon Cullinan <dcullinan@nvidia.com>

cc2ee1ed

feat: non blocking radix snapshot upload (#4839) · c7ad78f3
Yan Ru Pei authored Dec 10, 2025
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
c7ad78f3
feat: vLLM pre/postprocessing in-framework (#4529) · fe10dbfd
jthomson04 authored Dec 09, 2025
```
Signed-off-by: jthomson04 <jwillthomson19@gmail.com>
```
fe10dbfd
feat: KVBM V2 optimized bounce buffer transfer + benchmark (#3947) · 01819b87
jthomson04 authored Dec 09, 2025
```
Signed-off-by: jthomson04 <jwillthomson19@gmail.com>
```
01819b87

feat: Add logprobs support to vLLM backend (#4683) (#4697) · f57fd72f

Aryan Bagade authored Dec 09, 2025


Signed-off-by: Aryan Bagade <aryan@aryanbagade.com>
Signed-off-by: Aryan Bagade <73382554+AryanBagade@users.noreply.github.com>
Co-authored-by: Ryan McCormick <rmccormick@nvidia.com>

f57fd72f

feat: DeepSeek V3.2 tool calling support (#4822) · 6d091bf3

Vladislav Nosivskoy authored Dec 10, 2025


Signed-off-by: Vladislav Nosivskoy <vladnosiv@gmail.com>
Co-authored-by: Ayush Agarwal <ayushag@nvidia.com>

6d091bf3

ci: OPS-2136: Setup caching for docker build (#4535) · c9e445a9

Dillon Cullinan authored Dec 09, 2025

Signed-off-by: Dillon Cullinan <dcullinan@nvidia.com>
Signed-off-by: Dillon Cullinan <dillon-cullinan@users.noreply.github.com>
Co-authored-by: Dillon Cullinan <dillon-cullinan@users.noreply.github.com>

c9e445a9

09 Dec, 2025 23 commits
- feat: nixl_connect: Improve Concurrency Support (#4433) · 96fe63fe
  J Wyman authored Dec 09, 2025
```
Signed-off-by: J Wyman <jwyman@nvidia.com>
```
  96fe63fe
- chore: fix cache flush template (#4809) · 0ce7280d
  Neal Vaidya authored Dec 09, 2025
```
Signed-off-by: Neal Vaidya <nealv@nvidia.com>
```
  0ce7280d
- feat: make planner use DGD Scaling Adapters (#4825) · 9573d34c
  Julien Mancuso authored Dec 09, 2025
```
Signed-off-by: Julien Mancuso <jmancuso@nvidia.com>
```
  9573d34c
- feat: Add logprobs support to TRTLLM backend (#4759) · 01bfbea1
  Elijah Soba authored Dec 09, 2025
```
Signed-off-by: Elijah Soba <esoba@nvidia.com>
```
  01bfbea1
- chore: cleanups of passing around prefill and decode worker ids (#4829) · 1e5b20b2
  Yan Ru Pei authored Dec 09, 2025
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  1e5b20b2
- feat: add a grafana dashboard for planner (#4815) · 14321c8f
  Hongkuan Zhou authored Dec 09, 2025
```
Signed-off-by: hongkuanz <hongkuanz@nvidia.com>
```
  14321c8f
- fix: backend error serialization and add HTTP status code detection (#4786) · 065f466e
  Tzu-Ling Kan authored Dec 09, 2025
```
Signed-off-by: tzulingk@nvidia.com <tzulingk@nvidia.com>
```
  065f466e
- feat: unified tracing across dynamo + sglang (#4248) · fc161924
  ishandhanani authored Dec 09, 2025
  
  fc161924
- feat: KV aware LoRA request routing for vllm (#4810) · 8559c2bd
  Biswa Panda authored Dec 09, 2025
  
  8559c2bd
- fix: Add model registration to SGLang multimodal workers fixing bug #4486 (#4512) · ad5afb7b
  nancya-nv authored Dec 09, 2025
```
Signed-off-by: Nancy Agarwal <nancya@nvidia.com>
Co-authored-by: Kris Hung <krish@nvidia.com>
Co-authored-by: Ryan McCormick <rmccormick@nvidia.com>
```
  ad5afb7b
- chore(pipeline): Move migration outside of backend (#4823) · d5f425ab
  Graham King authored Dec 09, 2025
```
Signed-off-by: Graham King <grahamk@nvidia.com>
```
  d5f425ab
- feat: Disable health checks by default; auto-enable in K8s via operator (#4804) · 7c15166d
  Tzu-Ling Kan authored Dec 09, 2025
```
Signed-off-by: tzulingk@nvidia.com <tzulingk@nvidia.com>
```
  7c15166d
- feat: fault tolerance rolling upgrade test scenarios (#4558) · a473402f
  Thomas Montfort authored Dec 09, 2025
  
  a473402f
- feat: add tool_choice support (#4722) · 5585f803
  Vladislav Nosivskoy authored Dec 09, 2025
```
Signed-off-by: Vladislav Nosivskoy <vladnosiv@gmail.com>
```
  5585f803
- docs: Add multimodal documentation vllm, sglang, and trtllm backends (#4510) · 94d145a9
  Indrajit Bhosale authored Dec 09, 2025
```
Signed-off-by: Indrajit Bhosale <iamindrajitb@gmail.com>
Co-authored-by: krishung5 <krish@nvidia.com>
```
  94d145a9
- feat: add scaling adapter (#4699) · 09f2314d
  Julien Mancuso authored Dec 09, 2025
```
Signed-off-by: Julien Mancuso <jmancuso@nvidia.com>
```
  09f2314d
- feat: Change health check log level from info to debug (#4818) · 1f9b69b0
  Tzu-Ling Kan authored Dec 09, 2025
```
Signed-off-by: tzulingk@nvidia.com <tzulingk@nvidia.com>
```
  1f9b69b0
- fix: add timeout to tests before lora to unblock (#4808) · 1e37c103
  Keiven C authored Dec 08, 2025
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  1e37c103
- test: Pre-download models before tests are ran (#4811) · 4ca1679c
  Jacky authored Dec 08, 2025
```
Signed-off-by: Jacky <18255193+kthui@users.noreply.github.com>
```
  4ca1679c
- fix: have frontend populate "stop" field in OpenAI request (#4782) · e6de33f8
  Page authored Dec 08, 2025
  
  e6de33f8
- fix: Fix warning messages with healthcheck (#4793) · bd4366d0
  jthomson04 authored Dec 08, 2025
```
Signed-off-by: jthomson04 <jwillthomson19@gmail.com>
```
  bd4366d0
- fix: guided decoding params handling in vLLM (#4770) · 3e4b480b
  Vladislav Nosivskoy authored Dec 09, 2025
```
Signed-off-by: Vladislav Nosivskoy <vladnosiv@gmail.com>
Co-authored-by: Karen Chung <karenc@nvidia.com>
```
  3e4b480b
- feat: migrate planner metrics to backend (1/3 vLLM) (#4134) · 197e0227
  daiyaanarfeen authored Dec 08, 2025
```
Signed-off-by: Daiyaan <darfeen@nvidia.com>
Co-authored-by: tmontfort <tmontfort@nvidia.com>
```
  197e0227
08 Dec, 2025 8 commits
- refactor: update Dockerfiles to use group writes for development, saving build time (#4657) · 0dd51694
  Keiven C authored Dec 08, 2025
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  0dd51694
- feat: slurm jobs added fp4 and 8k1k (#4747) · 80dfb82c
  ishandhanani authored Dec 08, 2025
  
  80dfb82c
- test: router + sglang in pre-merge CI (#4767) · 3fea2e10
  Yan Ru Pei authored Dec 08, 2025
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  3fea2e10
- chore: de-spaghetti mocker scheduler (#4789) · 377711d3
  Yan Ru Pei authored Dec 08, 2025
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  377711d3
- feat: add runtime sanity checks for runtime images (#4788) · 8f0ac731
  Keiven C authored Dec 08, 2025
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  8f0ac731
- feat: update examples to use public lora (#4807) · 3bad9666
  Biswa Panda authored Dec 08, 2025
  
  3bad9666
- fix: remove workflow metric job (#4806) · d13c017d
  Nate Mailhot authored Dec 08, 2025
  
  d13c017d
- chore: jamba based parsers (#4776) · c9fad85c
  Ayush Agarwal authored Dec 08, 2025
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  c9fad85c