Commits · da40db4045080b80d31e4824f359bf407754db38 · OpenDAS / dynamo · GitLab

14 Apr, 2026 4 commits
- feat: add unified backend architecture with DynamoBackend (#8003) · da40db40
  Tanmay Verma authored Apr 14, 2026
```
Signed-off-by: Tanmay Verma <tanmayv@nvidia.com>
```
  da40db40
- feat: lmcache on CUDA 13 and ARM (#7534) · 21b44473
  Dmitry Tokarev authored Apr 14, 2026
```
Signed-off-by: Dmitry Tokarev <dtokarev@nvidia.com>
```
  21b44473
- chore: remove deprecated examples/multimodal directory (#8141) · 617d55c0
  Ryan McCormick authored Apr 14, 2026
```
Signed-off-by: Ryan McCormick <rmccormick@nvidia.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
```
  617d55c0
- fix(tests): re-enable vLLM LoRA tests and fix gpu_1 test flakiness (#8094) · 326a702d
  Keiven C authored Apr 14, 2026
```
Signed-off-by: Keiven Chang <keivenc@nvidia.com>
```
  326a702d
13 Apr, 2026 2 commits
- fix(ci): move test_multimodal_b64 to gpu_1 and force tool calls in... · 6bfc6d1f
  knarangN authored Apr 13, 2026
```
fix(ci): move test_multimodal_b64 to gpu_1 and force tool calls in aggregated_toolcalling (DYN-2662) (#8126)
Signed-off-by: Kavita Narang <knarang@nvidia.com>
```
  6bfc6d1f
- fix: disagg_same_gpu - profiling markers, GPU pinning, and memory args (#7996) · e3e728a8
  Keiven C authored Apr 13, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  e3e728a8
10 Apr, 2026 2 commits
- fix(ci): multimodal vLLM nightly — predownload, audio KV, 7B toolcalling (DYN-2569) (#7964) · eb772381
  knarangN authored Apr 10, 2026
```
Signed-off-by: Kavita Narang <knarang@nvidia.com>
```
  eb772381
- test: skip persistent nightly CI failures and provide associated tickets (#8006) · bcd0fc06
  Tushar Sharma authored Apr 10, 2026
```
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
```
  bcd0fc06
09 Apr, 2026 1 commit
- feat(test): add generalized multimodal model coverage framework (#7975) · f45a6985
  Qi Wang authored Apr 08, 2026
```
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
```
  f45a6985
08 Apr, 2026 1 commit
- feat(vllm): compute Qwen VL grid_thw from PIL images for P/D disagg (#7885) · d517fb80
  Qi Wang authored Apr 08, 2026
```
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
```
  d517fb80
07 Apr, 2026 2 commits
- refactor: split build_gpu_mem_args into engine-specific functions (#7916) · 4cdc49c2
  Keiven C authored Apr 07, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  4cdc49c2
- fix: broken vLLM nightly gpu_2 test and vLLM API compatibility errors (#7865) · 3ad7b7c8
  Keiven C authored Apr 07, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  3ad7b7c8
06 Apr, 2026 1 commit
- perf(multimodal): strip inline base64 from TCP transport in frontend decoding (#7895) · 6b8545fc
  Qi Wang authored Apr 06, 2026
```
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
```
  6b8545fc
03 Apr, 2026 1 commit
- chore(multimodal): Cleanup multimodal docs and consolidate launch scripts (#7845) · dacb2980
  Ryan McCormick authored Apr 03, 2026
  
  dacb2980
02 Apr, 2026 2 commits
- refactor(vLLM): Move video support from example to backend (#7663) · 4791aaaa
  Ryan McCormick authored Apr 02, 2026
  
  4791aaaa
- feat: GPU-parallel test runner with VRAM-aware scheduling (#7560) · 6dc85fbc
  Keiven C authored Apr 02, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  6dc85fbc
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

19 Mar, 2026 3 commits
- fix: unskip multimodal_disagg_qwen EPD test (#7523) · fa92958a
  KrishnanPrash authored Mar 19, 2026
```
Signed-off-by: Krishnan Prashanth <kprashanth@nvidia.com>
```
  fa92958a
- fix: increase `/dev/shm` for CI pytest containers to fix NIXL UCX segfault (#7515) · 4fc2b5e8
  KrishnanPrash authored Mar 19, 2026
  
  4fc2b5e8
- fix: restore serve test timeouts for slower CI GPUs (#7510) · 934b0669
  Keiven C authored Mar 19, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
```
  934b0669
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

17 Mar, 2026 1 commit
- chore: unskip several multi-gpu tests and enable sglang multi-gpu tests (#7443) · 9681225a
  Dmitry Tokarev authored Mar 17, 2026
```
Signed-off-by: Dmitry Tokarev <dtokarev@nvidia.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
```
  9681225a
15 Mar, 2026 1 commit
- ci: enable multi-GPU tests in PR pipeline (#7355) · 4751f769
  Anant Sharma authored Mar 15, 2026
```
Signed-off-by: Anant Sharma <anants@nvidia.com>
```
  4751f769
12 Mar, 2026 1 commit
- feat(kv-router): add worker_type to router selection log (#7258) · dcbccbcd
  Yan Ru Pei authored Mar 12, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  dcbccbcd
11 Mar, 2026 2 commits
- test: pytests able to run locally now (#7219) · 4bd6299b
  Graham King authored Mar 11, 2026
```
Signed-off-by: Graham King <grahamk@nvidia.com>
```
  4bd6299b
- chore: skip lmcache tests on cuda13 env (#7196) · 9eec94b2
  Dmitry Tokarev authored Mar 10, 2026
```
Signed-off-by: Dmitry Tokarev <dtokarev@nvidia.com>
```
  9eec94b2
01 Mar, 2026 1 commit
- chore: fix post merge (#6704) · 1412e44b
  Dmitry Tokarev authored Mar 01, 2026
```
Signed-off-by: Dmitry Tokarev <dtokarev@nvidia.com>
```
  1412e44b
28 Feb, 2026 1 commit
- feat(kv-router): add ActiveSequences benchmark and extract common bench utils (#6633) · e5850e23
  Yan Ru Pei authored Feb 27, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
```
  e5850e23
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
24 Feb, 2026 1 commit

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

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 1 commit
- feat: add embedding cache to pd worker (#6061) · c82fe888
  Qi Wang authored Feb 20, 2026
  
  c82fe888
19 Feb, 2026 1 commit
- test: add vllm audio tests to nightly ci pipeline (#6392) · 6f4b33f7
  knarangN authored Feb 19, 2026
```
Signed-off-by: Kavita Narang <knarang@nvidia.com>
```
  6f4b33f7
18 Feb, 2026 2 commits
- test: Add multimodal video tests to nightly CI pipeline (#6023) · 638d8e68
  knarangN authored Feb 18, 2026
```
Signed-off-by: Kavita Narang <knarang@nvidia.com>
```
  638d8e68
- refactor: delete handlers and disagg EC producer/consumer (#6051) · d2a57839
  Qi Wang authored Feb 17, 2026
  
  d2a57839
03 Feb, 2026 1 commit
- feat: vLLM backend with frontend media decoding (#5781) · 9bff03f2
  milesial authored Feb 03, 2026
```
Signed-off-by: Alexandre Milesi <milesial@users.noreply.github.com>
```
  9bff03f2
27 Jan, 2026 1 commit
- test: consolidate guided decoding vLLM configs (#5670) · 2dbfed89
  Alec authored Jan 26, 2026
  
  2dbfed89
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

09 Jan, 2026 2 commits
- feat: vllm sglang cuda13 main (#5218) · 4a2235d1
  Dmitry Tokarev authored Jan 09, 2026
```
Signed-off-by: Anant Sharma <anants@nvidia.com>
Signed-off-by: Dmitry Tokarev <dtokarev@nvidia.com>
Co-authored-by: Anant Sharma <anants@nvidia.com>
Co-authored-by: Dillon Cullinan <dcullinan@nvidia.com>
```
  4a2235d1
- fix: enable toggling kv events pub/sub (currently nats based) with --no-kv-events flag (#5237) · 0d3ff440
  Biswa Panda authored Jan 08, 2026
  
  0d3ff440