4ba48b7993
feat: add model field to multimodal protocol for LoRA identity ( #6382 )
2026-02-19 12:48:00 -08:00
b6824ae095
fix: restrict Image.open() formats to prevent PSD parsing (workaround) ( #6212 )
...
Signed-off-by: Dan Gil <dagil@nvidia.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-11 15:40:20 -08:00
842f0f15ec
fix: VLLM Multimodal minor fixes ( #5748 )
...
Signed-off-by: Indrajit Bhosale <iamindrajitb@gmail.com>
2026-01-29 02:45:41 +00:00
e22bb0376f
fix: Properly handle multiple text components from request ( #5196 )
2026-01-14 11:13:22 -05:00
a1333a8ddb
fix: update vLLM to 0.13.0 with API compatibility fixes ( #5222 )
...
Signed-off-by: Vasilis Vagias <vvagias@nvidia.com>
Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: Vasilis Vagias <vvagias@nvidia.com>
2026-01-06 15:08:08 -08:00
cf433e6825
chore: update all copyright headers in repo to 2026 ( #5130 )
...
Signed-off-by: Tushar Sharma <tusharma@nvidia.com>
2026-01-02 22:08:23 +00:00
0173d5e6fb
fix: Fix multimodal EPD examples for vllm version bump ( #4849 )
2025-12-10 20:24:32 +00:00
db5687f535
fix: adjust usage of vLLM for deprecation of 'disable_log_requests' ( #4659 )
...
Co-authored-by: Ryan McCormick <rmccormick@nvidia.com>
2025-12-03 08:52:41 +00:00
0f5dd2b780
feat: Add vLLM multimodal audio support ( #2760 )
...
Signed-off-by: Yuekai Zhang <zhangyuekai@foxmail.com>
Signed-off-by: krishung5 <krish@nvidia.com>
Co-authored-by: Kris Hung <krish@nvidia.com>
2025-11-13 13:55:48 -08:00
6f8fd8655c
refactor: align multimodal example port allocation with vLLM components ( #4163 )
...
Co-authored-by: Claude <noreply@anthropic.com>
2025-11-07 13:24:10 -08:00
90dc758938
chore: bump vllm to 0.11.0 ( #3422 )
...
Signed-off-by: alec-flowers <aflowers@nvidia.com>
2025-10-12 14:26:45 -07:00
5bb7490448
chore: bump vllm version to 0.10.2 ( #3180 )
...
Signed-off-by: Alec <aflowers@nvidia.com>
Signed-off-by: Alec <35311602+alec-flowers@users.noreply.github.com>
Signed-off-by: krishung5 <krish@nvidia.com>
Co-authored-by: Kris Hung <krish@nvidia.com>
2025-09-26 04:01:57 +00:00
3865a94148
feat: Port vllm port allocator to Rust in bindings ( #3125 )
...
Signed-off-by: Graham King <grahamk@nvidia.com>
2025-09-19 18:34:52 +00:00
960dc8967e
fix: replace hard coded dynamo namespace with env var ( #3048 )
...
Signed-off-by: Biswa Panda <biswa.panda@gmail.com>
2025-09-16 14:56:51 -07:00
78a1107441
feat: Add vLLM multimodal video support ( #2738 )
...
Signed-off-by: krishung5 <krish@nvidia.com>
2025-08-28 03:41:56 +00:00
50cd81f320
feat: Add vllm multimodal qwen aggregated support ( #2694 )
2025-08-26 14:18:39 -07:00
0a71aea62b
feat: Remove Duplicate Multimodel Nixl Connect Example ( #2622 )
2025-08-22 10:33:41 -07:00
844f8819a1
docs: Bring back some missed release/0.4.0 doc changes, fix broken links, add lychee link checker github action ( #2482 )
2025-08-18 10:26:15 -07:00
4dc529a1cc
chore: remove vLLM v0 multimodal example ( #2099 )
2025-07-25 15:22:33 -07:00
e1ae0f1507
feat: add multimodal k8s deployment manifest ( #1836 )
2025-07-10 00:34:54 -07:00
75503dae0b
feat: Video support with Dynamo ( #1443 )
...
Added support in examples for multimodal video with aggregated and disaggregated architecture using LLaVA-NeXT-Video-7B
2025-06-13 12:33:48 -07:00
e924a7c7c0
feat: generalize VLM embedding extraction ( #1388 )
...
Signed-off-by: hhzhang16 <54051230+hhzhang16@users.noreply.github.com>
Co-authored-by: Kris Hung <krish@nvidia.com>
2025-06-11 14:42:14 -07:00
e61f1c8a40
chore: Remove nats-py dependency ( #1387 )
2025-06-05 15:50:36 -07:00
b80482a4b2
feat: set model specific prompt templates in the multimodal config files, add documentation for multimodal example deployment ( #1366 )
2025-06-04 11:47:20 -07:00
92e33b86d8
perf: Create default sampling params only once during initialization ( #1294 )
2025-05-30 15:02:18 -07:00
5a30923f08
feat: Support OAI frontend format and add async image handing for multimodal ( #1214 )
...
Co-authored-by: J Wyman <jwyman@nvidia.com>
2025-05-28 16:38:01 -07:00
75e774d431
feat: NIXL Based RDMA Support w/ Multimodal Example ( #1060 )
2025-05-27 12:29:53 -07:00
aeb79e6277
feat: Support multiple models on single ingress node ( #1127 )
...
We can now do this:
- Node 1:
```
dynamo-run in=http out=dyn
```
- Node 2 and 3, two instances of component 'backend' in the nemotron_ultra pipeline:
```
dynamo-run in=dyn://nemotron_ultra.backend.generate out=vllm /data/models/NemotronUltra
```
- Node 4 and 5, two instances of the 'backend' component in nemotron_super pipeline:
```
dynamo-run in=dyn://nemotron_super.backend.generate out=vllm /data/models/NemotronSuper
```
The ingress node will discover all four instances and route correctly. We have been planning for this for a long time now.
As part of this auto-discovery is now always `out=dyn`, with no extra URL parts. Previously it could only route to a single pipeline.
Also:
- Refactor endpoint / instance naming now that I understand them
- Fix removing models when their instance stops.
2025-05-19 18:41:07 -04:00
10e912647a
feat: Add multimodal example with disaggregated serving ( #811 )
2025-05-07 16:21:44 -07:00
58df5aca4c
feat: Add multimodal example with aggregated serving ( #709 )
2025-05-02 12:45:31 -07:00
Biswa Panda