omnijev: add the native eager worker

README.md

@@ -162,7 +162,7 @@ runs as a Python worker or as a native worker on CUDA. Open-Jev-27B-v1.1
 and Open-Jev-9B run on the same native Rust/CUDA worker. Cua-S1 also has a
 Python screenshot worker, Valen-Preview-0923 has a Python reference worker for
 text and one-image state, CLM has a stub-encoder contract recipe, and OmniJev-4B
-v1.1 has its checkpoint export and CPU contract, with no worker yet. D3
+v1.1 runs on a native Rust/CUDA worker. D3
 (Decision 3.0) has a Python reference worker that runs the checkpoint's own
 pinned runtime for text, image and video state:
 
@@ -179,7 +179,7 @@ pinned runtime for text, image and video state:
 | autotrust/JEV-27B-VL | [Experimental Rust/CUDA worker](recipe/jev_vl/README.md); single-question text, online inline PNG/JPEG and offline-preencoded image inputs; [validation scope and limits](recipe/jev_vl/validation.md) |
 | CLM-v0.1-8B | [External worker with a CPU stub encoder](recipe/clm/README.md); contract checks only, real Qwen3-8B decisions unverified by this recipe |
 | LFM2.5-350M | [Transformers worker](recipe/lfm2/README.md); text state and independent choice questions |
-| OmniJev-4B v1.1 | [Checkpoint export and CPU contract](recipe/omnijev/README.md) only; the native worker is in progress ([#114](https://github.com/ThinkFlowLab/system1-omni/issues/114)) |
+| OmniJev-4B v1.1 | [Native Rust/CUDA worker](recipe/omnijev/native.md), eager: one image, Choice, Noul and Score; [reference comparison on sm_89](recipe/omnijev/native.md#comparison-with-the-reference) |
 | D3 (Decision 3.0) | [Python reference worker](recipe/d3/reference.md) on the checkpoint's pinned runtime; `choice`/`noul`/`score` over text, image and video state; author-run on A100 for text and media serving (see the integration and media PRs) |
 
 [Supported models and hardware](docs/supported-models.md) lists the devices

docs/supported-models.md

@@ -25,7 +25,7 @@ Models that are being added are also tracked in issues labeled [new model](https
 | CLM-v0.1-8B | [External `clm-serve` recipe](../recipe/clm/README.md) with a CPU stub embeddings server | **Stub-encoder contract checks only** ([#23](https://github.com/ThinkFlowLab/system1-omni/pull/23)); not real Qwen3-8B decisions | Real encoder unverified by the merged recipe | Unverified | Python, upstream CLM and head checkpoint; a real encoder requires a separate embeddings server |
 | LFM2.5-350M | [Transformers choice worker](../recipe/lfm2/README.md) | Tiny-model regression tests; full checkpoint unverified | L40S FP16 [recorded execution checks](../recipe/lfm2/RESULTS.md) | Unverified | Python 3.12, [pinned dependencies](../src/models/lfm2/requirements.lock.txt), explicit candidate batch size |
 | CLM-v0.1-8B | [Native Rust engine](../src/models/clm/README.md), `clm-run`, in front of a Qwen3-8B `/v1/embeddings` server | Engine-only agreement **4.5e-06** on identical vectors ([#29](https://github.com/ThinkFlowLab/system1-omni/pull/29), [validation](../recipe/clm/native/VALIDATION.md)); contract checks documented ([#23](https://github.com/ThinkFlowLab/system1-omni/pull/23)) | **Validated on compute capability 8.9** ([#29](https://github.com/ThinkFlowLab/system1-omni/pull/29), [validation](../recipe/clm/native/VALIDATION.md)) | Unverified | Rust to build and a separate Qwen3-8B `/v1/embeddings` server for the encoder; the engine itself has no device-specific code |
-| OmniJev-4B v1.1 | None yet: [checkpoint export and CPU contract](../recipe/omnijev/README.md); the native worker is in progress ([#114](https://github.com/ThinkFlowLab/system1-omni/issues/114)) | Request contract, preparation, heads and finishing checked against the reference ([#120](https://github.com/ThinkFlowLab/system1-omni/pull/120)); no inference | Not yet | Not supported | Python 3.10 reference environment for the export and fixtures |
+| OmniJev-4B v1.1 | [Native Rust/CUDA worker](../recipe/omnijev/native.md) on the shared Qwen3.5 vision and language path, eager | Contract, preparation, pixels, heads and finishing checked against the reference; no CPU inference | Compared with the reference on compute capability 8.9 for [140 answers](../recipe/omnijev/native.md#comparison-with-the-reference) | Not supported | Compute capability 8.0 or newer, CUDA toolkit to build, the export from `export.py` |
 | D3 (Decision 3.0) | [Python reference worker](../recipe/d3/reference.md) on the checkpoint's pinned runtime — text, image and video state over `choice`/`noul`/`score`, with the deliberate transport differences documented in the recipe; the native worker is in progress ([#136](https://github.com/ThinkFlowLab/system1-omni/issues/136)) | Unverified | Author-run on NVIDIA A100 (sm_80): serving smoke and reference parity for text and media (exact decisions, max delta 0.0); evidence in the integration and media PRs. Owner: @MrDongsls. | Unverified | Python 3.12, the [pinned runtime requirements](../recipe/d3/requirements.txt), the [pinned package](https://huggingface.co/vllm-sr/d3) (~52 GB); [media budgets and the 64 MiB total request cap](../recipe/d3/reference.md) |
 
 - **Validated:** covered by the recipe on `main` or by the checks in the linked merged pull request.

mkdocs.yml

@@ -78,6 +78,7 @@ nav:
       - JEV-27B-VL experimental worker: recipe/jev_vl/README.md
       - JEV-27B-VL validation status: recipe/jev_vl/validation.md
       - OmniJev preparation: recipe/omnijev/README.md
+      - OmniJev native worker: recipe/omnijev/native.md
       - CLM stub-encoder contract: recipe/clm/README.md
       - LFM2.5-350M choice worker: recipe/lfm2/README.md
       - D3 reference worker: recipe/d3/reference.md

recipe/README.md

@@ -31,7 +31,8 @@ For a first real decision, follow the [complete CPU walkthrough](../docs/getting
 - [JEV-27B-VL experimental worker](jev_vl/README.md): native text and preencoded
   image decisions; bounded H800 validation with explicit deployment limits.
 - [OmniJev-4B v1.1 preparation](omnijev/README.md): export the checkpoint and
-  regenerate the CPU fixtures; the native worker is not available yet.
+  regenerate the CPU fixtures; [native worker](omnijev/native.md): serve it on CUDA
+  and compare it with the reference.
 - [CLM behind the frontend](clm/README.md): run CLM's own server behind the frontend on
   CPU with a stub encoder, and what the response comparison has to allow for.
 - [D3 reference worker](d3/reference.md): download the pinned vllm-sr/d3 package,

recipe/omnijev/README.md

@@ -1,10 +1,10 @@
 # OmniJev-4B v1.1 preparation
 
 This recipe prepares OmniJev-4B v1.1 ([#114](https://github.com/ThinkFlowLab/system1-omni/issues/114))
-for the native worker: it exports the checkpoint and regenerates the CPU fixtures
-that the contract and preparation tests use. The native worker itself is not
-available yet; [the model README](../../src/models/omnijev/README.md) describes what
-is implemented. Run the commands from the repository root.
+for [the native worker](native.md): it exports the checkpoint and regenerates the CPU
+fixtures that the contract, preparation and pixel tests use. [The model
+README](../../src/models/omnijev/README.md) describes what is implemented. Run the
+commands from the repository root.
 
 ## Reference environment
 
@@ -65,8 +65,8 @@ OMNIJEV_CHECKPOINT=$PWD/omnijev-v1.1 OMNIJEV_EXPORT=$PWD/omnijev-export \
     cargo test --locked -p omni-omnijev-native -- --include-ignored
 ```
 
-The generator runs the reference's preparation, layout, rotary positions, heads and
-finishing without loading the backbone. The ordinary tests need only the committed
+The generator runs the reference's image loading and processor, preparation, layout,
+rotary positions, heads and finishing without loading the backbone. The ordinary tests need only the committed
 fixtures; the opt-in ones compare token ids with the checkpoint's tokenizer and the
 heads with the export. Cargo runs the tests from the crate's directory, so the two
 paths must be absolute.

recipe/omnijev/gate_questions.json

@@ -0,0 +1,242 @@
+{
+ "one_token": {
+  "type": "choice",
+  "instructions": "Which button should be clicked next?",
+  "options": [
+   {
+    "key": "click OK",
+    "text": "click OK"
+   },
+   {
+    "key": "click Cancel",
+    "text": "click Cancel"
+   },
+   {
+    "key": "click Close",
+    "text": "click Close"
+   },
+   {
+    "key": "click Help",
+    "text": "click Help"
+   },
+   {
+    "abstain": true
+   }
+  ]
+ },
+ "long_start": {
+  "type": "choice",
+  "instructions": "What should the user do next?",
+  "options": [
+   {
+    "key": "first",
+    "text": "open the settings menu at the top of the screen and then select the first item"
+   },
+   {
+    "key": "second",
+    "text": "open the settings menu at the top of the screen and then select the second item"
+   },
+   {
+    "key": "third",
+    "text": "open the settings menu at the top of the screen and then select the third item"
+   },
+   {
+    "abstain": true
+   }
+  ]
+ },
+ "no_fit": {
+  "type": "choice",
+  "instructions": "Which animal is shown in the picture?",
+  "options": [
+   {
+    "key": "a giraffe",
+    "text": "a giraffe"
+   },
+   {
+    "key": "a whale",
+    "text": "a whale"
+   },
+   {
+    "key": "a penguin",
+    "text": "a penguin"
+   },
+   {
+    "abstain": true
+   }
+  ]
+ },
+ "score10": {
+  "type": "score",
+  "instructions": "How bright is the image overall?",
+  "levels": [
+   "pitch black",
+   "very dark",
+   "dark",
+   "dim",
+   "slightly dim",
+   "neutral",
+   "slightly bright",
+   "bright",
+   "very bright",
+   "blinding"
+  ]
+ },
+ "choice32": {
+  "type": "choice",
+  "instructions": "Which word best describes the image?",
+  "options": [
+   {
+    "key": "game",
+    "text": "game"
+   },
+   {
+    "key": "chart",
+    "text": "chart"
+   },
+   {
+    "key": "map",
+    "text": "map"
+   },
+   {
+    "key": "menu",
+    "text": "menu"
+   },
+   {
+    "key": "photo",
+    "text": "photo"
+   },
+   {
+    "key": "diagram",
+    "text": "diagram"
+   },
+   {
+    "key": "table",
+    "text": "table"
+   },
+   {
+    "key": "form",
+    "text": "form"
+   },
+   {
+    "key": "chat",
+    "text": "chat"
+   },
+   {
+    "key": "email",
+    "text": "email"
+   },
+   {
+    "key": "video",
+    "text": "video"
+   },
+   {
+    "key": "code",
+    "text": "code"
+   },
+   {
+    "key": "terminal",
+    "text": "terminal"
+   },
+   {
+    "key": "calendar",
+    "text": "calendar"
+   },
+   {
+    "key": "document",
+    "text": "document"
+   },
+   {
+    "key": "spreadsheet",
+    "text": "spreadsheet"
+   },
+   {
+    "key": "slide",
+    "text": "slide"
+   },
+   {
+    "key": "browser",
+    "text": "browser"
+   },
+   {
+    "key": "robot",
+    "text": "robot"
+   },
+   {
+    "key": "board",
+    "text": "board"
+   },
+   {
+    "key": "phone",
+    "text": "phone"
+   },
+   {
+    "key": "street",
+    "text": "street"
+   },
+   {
+    "key": "kitchen",
+    "text": "kitchen"
+   },
+   {
+    "key": "office",
+    "text": "office"
+   },
+   {
+    "key": "forest",
+    "text": "forest"
+   },
+   {
+    "key": "beach",
+    "text": "beach"
+   },
+   {
+    "key": "city",
+    "text": "city"
+   },
+   {
+    "key": "animal",
+    "text": "animal"
+   },
+   {
+    "key": "person",
+    "text": "person"
+   },
+   {
+    "key": "car",
+    "text": "car"
+   },
+   {
+    "key": "screen",
+    "text": "screen"
+   },
+   {
+    "key": "error",
+    "text": "error"
+   },
+   {
+    "abstain": true
+   }
+  ]
+ },
+ "noul_day": {
+  "type": "noul",
+  "instructions": "The image was taken during the day."
+ },
+ "noul_people": {
+  "type": "noul",
+  "instructions": "There are more than three people in the image."
+ },
+ "noul_region": {
+  "type": "noul",
+  "instructions": "There is text in this region.",
+  "region": {
+   "box": [
+    100,
+    100,
+    700,
+    400
+   ]
+  }
+ }
+}