Commits · 525030324e3583f2485dd8f0fdb28dd3988549ff · OpenDAS / dynamo · GitLab

10 Dec, 2025 1 commit
- fix: Don't set min_tokens when ignore_eos is true (#4872) · 664b458a
  jthomson04 authored Dec 10, 2025
```
Signed-off-by: jthomson04 <jwillthomson19@gmail.com>
```
  664b458a
09 Dec, 2025 3 commits
- 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 tool_choice support (#4722) · 5585f803
  Vladislav Nosivskoy authored Dec 09, 2025
```
Signed-off-by: Vladislav Nosivskoy <vladnosiv@gmail.com>
```
  5585f803
- fix: have frontend populate "stop" field in OpenAI request (#4782) · e6de33f8
  Page authored Dec 08, 2025
  
  e6de33f8
04 Dec, 2025 1 commit
- fix: parallel tool call indices in streaming (#4723) · 8601ccdb
  Vladislav Nosivskoy authored Dec 04, 2025
```
Signed-off-by: Vladislav Nosivskoy <vladnosiv@gmail.com>
```
  8601ccdb
02 Dec, 2025 1 commit

refactor: New config types for tool calls (#4575) · 44e8600a

William Zhang authored Dec 02, 2025

* Why?

We would like the ability to configure different parser types. Prior to
this commit, only the JSON parser could be configured.

* What?

This commit refactors the tool parser config in the following ways:
- the `format` and `json` fields of `ToolParserConfig` are merged into
  a single `config` field that is a "discriminated union" type. Each
  parser type can declare its own configuration options.
- a `XmlParserConfig` is defined with a default factory method that
  corresponds to the Qwen3 coder configuration.
- affected calls and tests are adjusted.

44e8600a

25 Nov, 2025 1 commit

feat: decoded media via NIXL (#3988) · 0a065918

milesial authored Nov 25, 2025


Signed-off-by: Alexandre Milesi <milesial@users.noreply.github.com>
Signed-off-by: Ayush Agarwal <ayushag@nvidia.com>
Signed-off-by: Krishnan Prashanth <kprashanth@nvidia.com>
Co-authored-by: Ayush Agarwal <ayushag@nvidia.com>
Co-authored-by: Krishnan Prashanth <kprashanth@nvidia.com>
Co-authored-by: KrishnanPrash <140860868+KrishnanPrash@users.noreply.github.com>

0a065918

21 Nov, 2025 1 commit
- fix: missing tool calling usage (#4516) · 179ee38b
  Ayush Agarwal authored Nov 20, 2025
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  179ee38b
19 Nov, 2025 1 commit
- feat: nvext field to OpenAI APIs and add worker_id reporting (vLLM) (#4372) · 263f99dc
  Zhongxuan (Daniel) Wang authored Nov 18, 2025
```
Signed-off-by: Zhongxuan Wang <daniewang@nvidia.com>
```
  263f99dc
18 Nov, 2025 1 commit
- feat: support prompt_tokens_details in usage (#4239) · be9d6b2b
  Vladislav Nosivskoy authored Nov 19, 2025
```
Signed-off-by: Vladislav Nosivskoy <vladnosiv@gmail.com>
```
  be9d6b2b
08 Nov, 2025 2 commits
- fix: no more multiple finish reasons in stream (#4154) · 04f7579b
  Ayush Agarwal authored Nov 07, 2025
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  04f7579b
- feat: Add support for skip_special_tokens parameter in v1/completions and... · 441473c3
  Ryan McCormick authored Nov 07, 2025
```
feat: Add support for skip_special_tokens parameter in v1/completions and v1/chat/completions endpoints (#4175)
```
  441473c3
07 Nov, 2025 2 commits
- feat: enable HTTP completion endpoint to accept arrays of prompts and generate... · 93ada899
  zhongdaor-nv authored Nov 07, 2025
```
feat: enable HTTP completion endpoint to accept arrays of prompts and generate multiple completions per prompt (#3953)
Signed-off-by: zhongdaor <zhongdaor@nvidia.com>
```
  93ada899
- fix: Don't trim whitespace from end of each chat completion delta in aggregator (#4152) · f5094935
  Ryan McCormick authored Nov 06, 2025
  
  f5094935
06 Nov, 2025 2 commits
- fix: Remove constant gpt-2 token_id upper bound check (#3760) · 6336eea4
  Nathan Barry authored Nov 05, 2025
```
Co-authored-by: Ryan McCormick <rmccormick@nvidia.com>
```
  6336eea4
- feat: Reject unsupported parameters with 400 Bad Request (`/v1/completions`) (#4140) · 3d036fc4
  KrishnanPrash authored Nov 05, 2025
```
Signed-off-by: Krishnan Prashanth <kprashanth@nvidia.com>
```
  3d036fc4
03 Nov, 2025 1 commit
- feat: Reject unsupported parameters with 400 Bad Request (#4021) · c837b5ba
  KrishnanPrash authored Nov 03, 2025
```
Signed-off-by: Krishnan Prashanth <kprashanth@nvidia.com>
```
  c837b5ba
30 Oct, 2025 1 commit

feat: allow Triton model config specification in TensorModelConfig (#3874) · eb3a486d

GuanLuo authored Oct 31, 2025


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

eb3a486d

29 Oct, 2025 1 commit
- chore: add usage field to non-streaming responses by default (#3922) · 3fe5653b
  Ayush Agarwal authored Oct 29, 2025
```
Signed-off-by: ayushag <ayushag@nvidia.com>
Co-authored-by: Ryan McCormick <rmccormick@nvidia.com>
```
  3fe5653b
27 Oct, 2025 2 commits
- feat: Media URL passthrough in OAI preprocessor (#3733) · a79122c6
  milesial authored Oct 27, 2025
```
Signed-off-by: Alexandre Milesi <30204471+milesial@users.noreply.github.com>
```
  a79122c6
- chore: add request validation and better error message for n > 1 and temperature=0 (#3914) · 980bae03
  zhongdaor-nv authored Oct 27, 2025
```
Signed-off-by: zhongdaor <zhongdaor@nvidia.com>
```
  980bae03
24 Oct, 2025 1 commit

feat: Extend TensorRequest and TensorResponse to contains extra parameters (#3761) · 90caf3ea

KrishnanPrash authored Oct 24, 2025

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

90caf3ea

23 Oct, 2025 1 commit
- chore: Use KeyValueStoreManager instead of etcd::Client (#3822) · 7731b024
  Graham King authored Oct 23, 2025
```
Signed-off-by: Graham King <grahamk@nvidia.com>
```
  7731b024
16 Oct, 2025 1 commit
- feat: dp rank routing (#3597) · f978f4d1
  Yan Ru Pei authored Oct 15, 2025
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  f978f4d1
14 Oct, 2025 1 commit
- fix: Add support for single element arrays for chat and completions prompts (#3482) · 3dc25129
  zhongdaor-nv authored Oct 14, 2025
```
Signed-off-by: zhongdaor <zhongdaor@nvidia.com>
```
  3dc25129
10 Oct, 2025 1 commit
- chore: remove deprecated nvext parameters for 6.0 (#3551) · b954a249
  ryan-lempka authored Oct 10, 2025
  
  b954a249
08 Oct, 2025 1 commit
- feat: prefill routing for trtllm (#3471) · f3987f6c
  Yan Ru Pei authored Oct 08, 2025
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  f3987f6c
04 Oct, 2025 1 commit
- feat: use KvPushRouter for prefill router (#3401) · 30610e73
  Yan Ru Pei authored Oct 03, 2025
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  30610e73
01 Oct, 2025 1 commit
- feat: reasoning parser transformation (#3295) · 2a2bf58a
  Ayush Agarwal authored Oct 01, 2025
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  2a2bf58a
29 Sep, 2025 2 commits
- chore: relaxing constraints on metadata field, adding metadata field to completions API (#3240) · 13156361
  nv-nedelman-1 authored Sep 29, 2025
```
Signed-off-by: Nicholas Edelman <nedelman@nvidia.com>
```
  13156361
- feat: enable parallel tool calling and add testing (#3188) · 67e1f6ee
  Elyas Mehtabuddin authored Sep 29, 2025
```
Signed-off-by: Elyas Mehtabuddin <emehtabuddin@nvidia.com>
Co-authored-by: Ryan McCormick <rmccormick@nvidia.com>
```
  67e1f6ee
26 Sep, 2025 1 commit
- chore: added more api error code validations (#3231) · d4f0d2bc
  Ayush Agarwal authored Sep 26, 2025
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  d4f0d2bc
24 Sep, 2025 2 commits

chore: jail stream optimizations (v1) (#3195) · 2ae20102
Ayush Agarwal authored Sep 24, 2025
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
2ae20102

feat: tensor type for generic inference. (#2746) · 6ba64c31

GuanLuo authored Sep 24, 2025


Signed-off-by: Guan Luo <gluo@nvidia.com>
Signed-off-by: GuanLuo <41310872+GuanLuo@users.noreply.github.com>
Co-authored-by: Olga Andreeva <124622579+oandreeva-nv@users.noreply.github.com>
Co-authored-by: Ryan McCormick <rmccormick@nvidia.com>

6ba64c31

23 Sep, 2025 1 commit
- feat: JailedStream (#3034) · c63cceaa
  Ryan Olson authored Sep 23, 2025
```
Signed-off-by: ayushag <ayushag@nvidia.com>
```
  c63cceaa
18 Sep, 2025 2 commits
- feat: enhance GPT OSS frontend with improved harmony tool calling parser and... · 6675bfc8
  zhongdaor-nv authored Sep 18, 2025
```
feat: enhance GPT OSS frontend with improved harmony tool calling parser and reasoning parser (#2999)
Signed-off-by: zhongdaor <zhongdaor@nvidia.com>
```
  6675bfc8
- feat: add finish reason = tool_calls for stream=False and phi-4 detect token start fix (#3087) · 31f5ed3c
  Elyas Mehtabuddin authored Sep 18, 2025
```
Signed-off-by: Elyas Mehtabuddin <emehtabuddin@nvidia.com>
```
  31f5ed3c
17 Sep, 2025 2 commits
- chore: fillout sampling params (seed, n, best_of, min_p) (#3055) · 3de04dd9
  Greg Clark authored Sep 17, 2025
```
Signed-off-by: Greg Clark <grclark@nvidia.com>
```
  3de04dd9
- feat: add chat_template_kwargs param to v1/chat/completion (#3016) · eb6722e3
  Chi McIsaac authored Sep 17, 2025
```
Signed-off-by: Chi McIsaac <chixie.mcisaac@gmail.com>
```
  eb6722e3
16 Sep, 2025 1 commit
- chore(llm): Remove extra license headers (#3065) · bc0a7633
  Graham King authored Sep 16, 2025
```
Signed-off-by: Graham King <grahamk@nvidia.com>
```
  bc0a7633