Commits · 4d302ab6c969871e0b7f7b83d3777c5946f1cb9f · OpenDAS / dynamo

20 Mar, 2026 1 commit

feat: [vLLM] add remote embedding loader to prefill/decode worker. Port E/P/D usage (#7507) · 4d302ab6

GuanLuo authored Mar 19, 2026


Signed-off-by: Guan Luo <41310872+GuanLuo@users.noreply.github.com>
Signed-off-by: GuanLuo <41310872+GuanLuo@users.noreply.github.com>

4d302ab6

18 Mar, 2026 1 commit

feat: GPU VRAM profiler via memory fraction injection + profiled test markers... · 0b20745e

Keiven C authored Mar 18, 2026


feat: GPU VRAM profiler via memory fraction injection + profiled test markers (part 2 - vLLM only) (#6719)
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>

0b20745e

16 Mar, 2026 1 commit
- feat: vllm omni cli (#6788) · f81f5277
  Ayush Agarwal authored Mar 16, 2026
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  f81f5277
12 Mar, 2026 1 commit
- feat: vllm omni image to video support (#6530) · 1182e207
  Ayush Agarwal authored Mar 12, 2026
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  1182e207
11 Mar, 2026 1 commit
- feat: refactor launch scripts with shared launch_utils.sh for consistent failure handling (#7008) · 34ccc0b1
  Keiven C authored Mar 11, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  34ccc0b1
10 Mar, 2026 2 commits
- docs: fix disaggregated deployment example in tracing.md (#6999) · fa474d36
  dagil-nvidia authored Mar 10, 2026
```
Signed-off-by: Dan Gil <dagil@nvidia.com>
```
  fa474d36
- feat: add an auto GPU VRAM estimator for disagg-same-GPU (#6868) · 423e4b45
  Keiven C authored Mar 10, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  423e4b45
05 Mar, 2026 1 commit
- perf: multimodal benchmark sweep (#6795) · 9780bf3a
  Qi Wang authored Mar 04, 2026
  
  9780bf3a
03 Mar, 2026 5 commits

fix: properly setup and register vLLM worker for external / hybrid load... · 90d74637

GuanLuo authored Mar 03, 2026


fix: properly setup and register vLLM worker for external / hybrid load balancing. Update launch script (#6695)
Signed-off-by: Guan Luo <41310872+GuanLuo@users.noreply.github.com>

90d74637

docs: delete script and add instructions (#6763) · 552ae186
Qi Wang authored Mar 03, 2026

552ae186
docs: vLLM README container instructions and KV offloading page (#6793) · 9fe03dd8
Alec authored Mar 02, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
```
9fe03dd8

feat: FlexKV integration in Dynamo (#5858) · 4d7c9845

Yue Yu authored Mar 03, 2026


Signed-off-by: zhuofan1123 <zhuofanl@nvidia.com>
Co-authored-by: zhuofan1123 <zhuofanl@nvidia.com>

4d7c9845

docs: restructure vLLM docs and add startup banners to launch scripts (#6698) · 75bf1e09

Alec authored Mar 02, 2026


Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: ishandhanani <ishandhanani@gmail.com>

75bf1e09

02 Mar, 2026 1 commit
- fix(perf): add embedding transfer implementation with NIXL WRITE initiation (#6651) · 2cab0f7f
  GuanLuo authored Mar 02, 2026
```
Signed-off-by: Guan Luo <41310872+GuanLuo@users.noreply.github.com>
```
  2cab0f7f
27 Feb, 2026 1 commit
- chore: migrate vllm e/p/d test from gpu_2 -> gpu_1 (#6638) · b1930a61
  KrishnanPrash authored Feb 27, 2026
```
Signed-off-by: Krishnan Prashanth <kprashanth@nvidia.com>
```
  b1930a61
25 Feb, 2026 3 commits
- docs: embedding cache in vLLM and TRT-LLM (#6555) · d5add7ff
  Qi Wang authored Feb 25, 2026
  
  d5add7ff
- fix: vllm launch script errors for disagg and spec decoding (#6562) · 35323da2
  Alec authored Feb 24, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
```
  35323da2
- feat: add --kv-transfer-config NixlConnector to disagg scripts and recipes (#6560) · eac94322
  Alec authored Feb 24, 2026
```
Signed-off-by: alec-flowers <aflowers@nvidia.com>
```
  eac94322
24 Feb, 2026 3 commits

fix: restore E/P/D multimodal disagg serving and add Qwen3-VL-30B-A3B support (#6533) · 8e236375

KrishnanPrash authored Feb 24, 2026


Signed-off-by: Krishnan Prashanth <kprashanth@nvidia.com>
Signed-off-by: KrishnanPrash <140860868+KrishnanPrash@users.noreply.github.com>

8e236375

feat: remove --connector flag for vLLM backend (LLM-90) (#6450) · 6d3b92f0

Alec authored Feb 24, 2026


Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

6d3b92f0

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

22 Feb, 2026 1 commit

feat: add --headless mode for multi-node TP/PP in dynamo.vllm (#6204) · 4ebb244b

Alec authored Feb 22, 2026


Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

4ebb244b

20 Feb, 2026 4 commits
- feat: add embedding cache to pd worker (#6061) · c82fe888
  Qi Wang authored Feb 20, 2026
  
  c82fe888
- 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
- chore: unify filesystem usage across all frameworks and workers (#6391) · d6c49779
  Ayush Agarwal authored Feb 19, 2026
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  d6c49779
- feat: vllm omni text to video generation pipeline (#6104) · 6dd3ce2e
  Ayush Agarwal authored Feb 19, 2026
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  6dd3ce2e
18 Feb, 2026 2 commits
- refactor: introduce worker factory in vLLM multimodal (#6060) · 4d0380d5
  Qi Wang authored Feb 18, 2026
  
  4d0380d5
- refactor: delete handlers and disagg EC producer/consumer (#6051) · d2a57839
  Qi Wang authored Feb 17, 2026
  
  d2a57839
10 Feb, 2026 1 commit
- fix: llama4 vllm agg multimodal script (#6103) · 120ae649
  Indrajit Bhosale authored Feb 10, 2026
  
  120ae649
09 Feb, 2026 1 commit
- feat: text to image vLLM Omni (#5912) · 9f76d060
  Ayush Agarwal authored Feb 09, 2026
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  9f76d060
08 Feb, 2026 1 commit

chore: enable local indexers by default, and use normal event plane by default... · 6783bdca

Yan Ru Pei authored Feb 08, 2026


chore: enable local indexers by default, and use normal event plane by default (not jetstream) (#5941)
Signed-off-by: PeaBrane <yanrpei@gmail.com>

6783bdca

05 Feb, 2026 1 commit
- fix: Update vllm multimodal PD Disagg launch script (#5951) · a72f41f6
  Indrajit Bhosale authored Feb 05, 2026
  
  a72f41f6
04 Feb, 2026 1 commit
- feat: basic vllm omni pipeline support (#5608) · 76e0e207
  Ayush Agarwal authored Feb 04, 2026
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  76e0e207
03 Feb, 2026 1 commit
- fix: Add --is-decode-worker flag to vllm examples (#5899) · 3b9b3f31
  Sean SH Choi authored Feb 02, 2026
```
Signed-off-by: Sean Choi <sechoi@nvidia.com>
```
  3b9b3f31
29 Jan, 2026 1 commit
- fix: VLLM Multimodal minor fixes (#5748) · 842f0f15
  Indrajit Bhosale authored Jan 28, 2026
```
Signed-off-by: Indrajit Bhosale <iamindrajitb@gmail.com>
```
  842f0f15
28 Jan, 2026 2 commits
- chore: adjust gpu-memory-utilization to accommodate vLLM's runtime GPU memory requirement (#5755) · 77aadb72
  GuanLuo authored Jan 28, 2026
```
Signed-off-by: Guan Luo <gluo@nvidia.com>
```
  77aadb72
- 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
16 Jan, 2026 1 commit
- chore: support for muliple MM requests for vLLM Encoder + Rust Based Preprocessor (#5463) · f9918f61
  Ayush Agarwal authored Jan 16, 2026
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  f9918f61
15 Jan, 2026 1 commit
- feat: [router] serve worker KV query over dynamo endpoint instead of nats (#5451) · 7ebd5f82
  Biswa Panda authored Jan 15, 2026
  
  7ebd5f82
14 Jan, 2026 1 commit

refactor: vllm EPD refactor (#4994) · 334cbd9b

GuanLuo authored Jan 14, 2026


Signed-off-by: Guan Luo <gluo@nvidia.com>
Signed-off-by: Guan Luo <41310872+GuanLuo@users.noreply.github.com>

334cbd9b