Add files using upload-large-folder tool

.gitattributes

@@ -1,35 +1,9 @@
-*.7z filter=lfs diff=lfs merge=lfs -text
-*.arrow filter=lfs diff=lfs merge=lfs -text
 *.bin filter=lfs diff=lfs merge=lfs -text
-*.bz2 filter=lfs diff=lfs merge=lfs -text
-*.ckpt filter=lfs diff=lfs merge=lfs -text
-*.ftz filter=lfs diff=lfs merge=lfs -text
-*.gz filter=lfs diff=lfs merge=lfs -text
-*.h5 filter=lfs diff=lfs merge=lfs -text
-*.joblib filter=lfs diff=lfs merge=lfs -text
-*.lfs.* filter=lfs diff=lfs merge=lfs -text
-*.mlmodel filter=lfs diff=lfs merge=lfs -text
-*.model filter=lfs diff=lfs merge=lfs -text
-*.msgpack filter=lfs diff=lfs merge=lfs -text
-*.npy filter=lfs diff=lfs merge=lfs -text
-*.npz filter=lfs diff=lfs merge=lfs -text
-*.onnx filter=lfs diff=lfs merge=lfs -text
-*.ot filter=lfs diff=lfs merge=lfs -text
-*.parquet filter=lfs diff=lfs merge=lfs -text
-*.pb filter=lfs diff=lfs merge=lfs -text
-*.pickle filter=lfs diff=lfs merge=lfs -text
-*.pkl filter=lfs diff=lfs merge=lfs -text
-*.pt filter=lfs diff=lfs merge=lfs -text
-*.pth filter=lfs diff=lfs merge=lfs -text
-*.rar filter=lfs diff=lfs merge=lfs -text
+*.xml filter=lfs diff=lfs merge=lfs -text
 *.safetensors filter=lfs diff=lfs merge=lfs -text
-saved_model/**/* filter=lfs diff=lfs merge=lfs -text
-*.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
-*.tflite filter=lfs diff=lfs merge=lfs -text
-*.tgz filter=lfs diff=lfs merge=lfs -text
-*.wasm filter=lfs diff=lfs merge=lfs -text
-*.xz filter=lfs diff=lfs merge=lfs -text
-*.zip filter=lfs diff=lfs merge=lfs -text
-*.zst filter=lfs diff=lfs merge=lfs -text
-*tfevents* filter=lfs diff=lfs merge=lfs -text
+*.json text
+*.md text
+*.py text
+*.txt text
+*.jinja text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

.ipynb_checkpoints/README-checkpoint.md

@@ -0,0 +1,90 @@
+---
+license: apache-2.0
+base_model: Qwen/Qwen3.6-35B-A3B
+library_name: openvino
+pipeline_tag: image-text-to-text
+tags:
+- openvino
+- optimum-intel
+- int4
+- vlm
+- vision-language-model
+- moe
+- qwen
+- qwen3.6
+- converted
+---
+
+# Qwen3.6 35B-A3B OpenVINO INT4
+
+This repository contains an unofficial OpenVINO™ IR conversion of [`Qwen/Qwen3.6-35B-A3B`](https://huggingface.co/Qwen/Qwen3.6-35B-A3B) with INT4 weight compression.
+
+The model is intended for local vision-language generation using OpenVINO and Optimum Intel. It can take an image plus a text prompt and generate a natural-language response.
+
+## Original model
+
+- Original model: `Qwen/Qwen3.6-35B-A3B`
+- Model family: Qwen3.6
+- Architecture: sparse Mixture-of-Experts vision-language model
+- Converted format: OpenVINO IR
+- Weight format: INT4
+- Task: image-text-to-text / vision-language generation
+
+This is an unofficial converted model repository. Please refer to the original model card for full model details, license terms, intended use, safety notes, and limitations.
+
+## Model summary
+
+Qwen3.6-35B-A3B is a sparse Mixture-of-Experts model from the Qwen3.6 family. The upstream model is distributed on Hugging Face and is intended for multimodal use cases including image understanding and text generation.
+
+This OpenVINO version is designed for local inference on Intel hardware through the Optimum Intel OpenVINO runtime path.
+
+## Conversion
+
+This model was converted with Optimum Intel using the OpenVINO export path.
+
+```bash
+optimum-cli export openvino \
+  --model Qwen/Qwen3.6-35B-A3B \
+  --weight-format int4 \
+  --trust-remote-code \
+  ov_Qwen3.6-35B-A3B_int4
+  
+Runtime note
+
+This model currently uses the Optimum Intel OVModelForVisualCausalLM path for inference.
+
+The tested environment used a Qwen3.5/Qwen3.6 support branch of Optimum Intel and nightly OpenVINO packages:
+pip install -U git+https://github.com/rkazants/optimum-intel.git@support_qwen3_5
+pip install --pre -U openvino openvino-tokenizers nncf --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly
+pip install transformers==5.2.0
+pip install requests torchvision opencv-python
+OpenVINO Model Server compatibility is not claimed for this model. In local testing, OVMS 2026.1 rejected this model type during VLM initialization with an unsupported model-type error, while direct Optimum Intel OpenVINO inference worked.
+
+Quantization
+Weight format: INT4
+Export tool: Optimum Intel
+Runtime: Optimum Intel OpenVINO
+Model format: OpenVINO IR
+
+INT4 compression is intended to reduce model size and memory usage compared with higher precision weights. As with any converted and quantized model, quality and numerical behavior may differ from the original model and should be validated for your use case.
+Usage
+Run from inside the model directory
+python examples/run_qwen36_vlm_openvino.py \
+  --model-dir . \
+  --device CPU \
+  --max-new-tokens 64 \
+  --prompt "What animal is on the candy? Answer in one sentence."
+  
+Notes
+This is a vision-language model, not a text-only model.
+Depending on package versions, you may see:
+The following generation flags are not valid and may be ignored: ['top_p', 'top_k']
+The demo script uses deterministic generation with do_sample=False.
+OpenVINO Model Server compatibility is not claimed unless separately validated.
+Limitations
+
+This repository inherits the limitations of the original Qwen/Qwen3.6-35B-A3B model. Additional differences may arise from OpenVINO conversion, INT4 compression, and runtime configuration.
+
+Attribution
+
+This is an unofficial OpenVINO conversion of the original Qwen3.6 model. All rights to the original model, training, and licensing remain with the original authors.

README.md

@@ -0,0 +1,90 @@
+---
+license: apache-2.0
+base_model: Qwen/Qwen3.6-35B-A3B
+library_name: openvino
+pipeline_tag: image-text-to-text
+tags:
+- openvino
+- optimum-intel
+- int4
+- vlm
+- vision-language-model
+- moe
+- qwen
+- qwen3.6
+- converted
+---
+
+# Qwen3.6 35B-A3B OpenVINO INT4
+
+This repository contains an unofficial OpenVINO™ IR conversion of [`Qwen/Qwen3.6-35B-A3B`](https://huggingface.co/Qwen/Qwen3.6-35B-A3B) with INT4 weight compression.
+
+The model is intended for local vision-language generation using OpenVINO and Optimum Intel. It can take an image plus a text prompt and generate a natural-language response.
+
+## Original model
+
+- Original model: `Qwen/Qwen3.6-35B-A3B`
+- Model family: Qwen3.6
+- Architecture: sparse Mixture-of-Experts vision-language model
+- Converted format: OpenVINO IR
+- Weight format: INT4
+- Task: image-text-to-text / vision-language generation
+
+This is an unofficial converted model repository. Please refer to the original model card for full model details, license terms, intended use, safety notes, and limitations.
+
+## Model summary
+
+Qwen3.6-35B-A3B is a sparse Mixture-of-Experts model from the Qwen3.6 family. The upstream model is distributed on Hugging Face and is intended for multimodal use cases including image understanding and text generation.
+
+This OpenVINO version is designed for local inference on Intel hardware through the Optimum Intel OpenVINO runtime path.
+
+## Conversion
+
+This model was converted with Optimum Intel using the OpenVINO export path.
+
+```bash
+optimum-cli export openvino \
+  --model Qwen/Qwen3.6-35B-A3B \
+  --weight-format int4 \
+  --trust-remote-code \
+  ov_Qwen3.6-35B-A3B_int4
+  
+Runtime note
+
+This model currently uses the Optimum Intel OVModelForVisualCausalLM path for inference.
+
+The tested environment used a Qwen3.5/Qwen3.6 support branch of Optimum Intel and nightly OpenVINO packages:
+pip install -U git+https://github.com/rkazants/optimum-intel.git@support_qwen3_5
+pip install --pre -U openvino openvino-tokenizers nncf --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly
+pip install transformers==5.2.0
+pip install requests torchvision opencv-python
+OpenVINO Model Server compatibility is not claimed for this model. In local testing, OVMS 2026.1 rejected this model type during VLM initialization with an unsupported model-type error, while direct Optimum Intel OpenVINO inference worked.
+
+Quantization
+Weight format: INT4
+Export tool: Optimum Intel
+Runtime: Optimum Intel OpenVINO
+Model format: OpenVINO IR
+
+INT4 compression is intended to reduce model size and memory usage compared with higher precision weights. As with any converted and quantized model, quality and numerical behavior may differ from the original model and should be validated for your use case.
+Usage
+Run from inside the model directory
+python examples/run_qwen36_vlm_openvino.py \
+  --model-dir . \
+  --device CPU \
+  --max-new-tokens 64 \
+  --prompt "What animal is on the candy? Answer in one sentence."
+  
+Notes
+This is a vision-language model, not a text-only model.
+Depending on package versions, you may see:
+The following generation flags are not valid and may be ignored: ['top_p', 'top_k']
+The demo script uses deterministic generation with do_sample=False.
+OpenVINO Model Server compatibility is not claimed unless separately validated.
+Limitations
+
+This repository inherits the limitations of the original Qwen/Qwen3.6-35B-A3B model. Additional differences may arise from OpenVINO conversion, INT4 compression, and runtime configuration.
+
+Attribution
+
+This is an unofficial OpenVINO conversion of the original Qwen3.6 model. All rights to the original model, training, and licensing remain with the original authors.

chat_template.jinja

@@ -0,0 +1,154 @@
+{%- set image_count = namespace(value=0) %}
+{%- set video_count = namespace(value=0) %}
+{%- macro render_content(content, do_vision_count, is_system_content=false) %}
+    {%- if content is string %}
+        {{- content }}
+    {%- elif content is iterable and content is not mapping %}
+        {%- for item in content %}
+            {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
+                {%- if is_system_content %}
+                    {{- raise_exception('System message cannot contain images.') }}
+                {%- endif %}
+                {%- if do_vision_count %}
+                    {%- set image_count.value = image_count.value + 1 %}
+                {%- endif %}
+                {%- if add_vision_id %}
+                    {{- 'Picture ' ~ image_count.value ~ ': ' }}
+                {%- endif %}
+                {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
+            {%- elif 'video' in item or item.type == 'video' %}
+                {%- if is_system_content %}
+                    {{- raise_exception('System message cannot contain videos.') }}
+                {%- endif %}
+                {%- if do_vision_count %}
+                    {%- set video_count.value = video_count.value + 1 %}
+                {%- endif %}
+                {%- if add_vision_id %}
+                    {{- 'Video ' ~ video_count.value ~ ': ' }}
+                {%- endif %}
+                {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
+            {%- elif 'text' in item %}
+                {{- item.text }}
+            {%- else %}
+                {{- raise_exception('Unexpected item type in content.') }}
+            {%- endif %}
+        {%- endfor %}
+    {%- elif content is none or content is undefined %}
+        {{- '' }}
+    {%- else %}
+        {{- raise_exception('Unexpected content type.') }}
+    {%- endif %}
+{%- endmacro %}
+{%- if not messages %}
+    {{- raise_exception('No messages provided.') }}
+{%- endif %}
+{%- if tools and tools is iterable and tools is not mapping %}
+    {{- '<|im_start|>system\n' }}
+    {{- "# Tools\n\nYou have access to the following functions:\n\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>" }}
+    {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n<tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>\nvalue_1\n</parameter>\n<parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n</parameter>\n</function>\n</tool_call>\n\n<IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner <function=...></function> block must be nested within <tool_call></tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n</IMPORTANT>' }}
+    {%- if messages[0].role == 'system' %}
+        {%- set content = render_content(messages[0].content, false, true)|trim %}
+        {%- if content %}
+            {{- '\n\n' + content }}
+        {%- endif %}
+    {%- endif %}
+    {{- '<|im_end|>\n' }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {%- set content = render_content(messages[0].content, false, true)|trim %}
+        {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" %}
+        {%- set content = render_content(message.content, false)|trim %}
+        {%- if not(content.startswith('<tool_response>') and content.endswith('</tool_response>')) %}
+            {%- set ns.multi_step_tool = false %}
+            {%- set ns.last_query_index = index %}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if ns.multi_step_tool %}
+    {{- raise_exception('No user query found in messages.') }}
+{%- endif %}
+{%- for message in messages %}
+    {%- set content = render_content(message.content, true)|trim %}
+    {%- if message.role == "system" %}
+        {%- if not loop.first %}
+            {{- raise_exception('System message must be at the beginning.') }}
+        {%- endif %}
+    {%- elif message.role == "user" %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- set reasoning_content = reasoning_content|trim %}
+        {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}
+            {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content + '\n</think>\n\n' + content }}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if tool_call.function is defined %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {%- if loop.first %}
+                    {%- if content|trim %}
+                        {{- '\n\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                    {%- else %}
+                        {{- '<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                    {%- endif %}
+                {%- else %}
+                    {{- '\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
+                {%- endif %}
+                {%- if tool_call.arguments is defined %}
+                    {%- for args_name, args_value in tool_call.arguments|items %}
+                        {{- '<parameter=' + args_name + '>\n' }}
+                        {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}
+                        {{- args_value }}
+                        {{- '\n</parameter>\n' }}
+                    {%- endfor %}
+                {%- endif %}
+                {{- '</function>\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.previtem and loop.previtem.role != "tool" %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- content }}
+        {{- '\n</tool_response>' }}
+        {%- if not loop.last and loop.nextitem.role != "tool" %}
+            {{- '<|im_end|>\n' }}
+        {%- elif loop.last %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- else %}
+        {{- raise_exception('Unexpected message role.') }}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- else %}
+        {{- '<think>\n' }}
+    {%- endif %}
+{%- endif %}

config.json

@@ -0,0 +1,121 @@
+{
+  "architectures": [
+    "Qwen3_5MoeForConditionalGeneration"
+  ],
+  "dtype": "bfloat16",
+  "image_token_id": 248056,
+  "model_type": "qwen3_5_moe",
+  "text_config": {
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "attn_output_gate": true,
+    "bos_token_id": 248044,
+    "dtype": "bfloat16",
+    "eos_token_id": 248044,
+    "full_attention_interval": 4,
+    "head_dim": 256,
+    "hidden_act": "silu",
+    "hidden_size": 2048,
+    "initializer_range": 0.02,
+    "layer_types": [
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention",
+      "linear_attention",
+      "linear_attention",
+      "linear_attention",
+      "full_attention"
+    ],
+    "linear_conv_kernel_dim": 4,
+    "linear_key_head_dim": 128,
+    "linear_num_key_heads": 16,
+    "linear_num_value_heads": 32,
+    "linear_value_head_dim": 128,
+    "mamba_ssm_dtype": "float32",
+    "max_position_embeddings": 262144,
+    "model_type": "qwen3_5_moe_text",
+    "moe_intermediate_size": 512,
+    "mtp_num_hidden_layers": 1,
+    "mtp_use_dedicated_embeddings": false,
+    "num_attention_heads": 16,
+    "num_experts": 256,
+    "num_experts_per_tok": 8,
+    "num_hidden_layers": 40,
+    "num_key_value_heads": 2,
+    "output_router_logits": false,
+    "pad_token_id": null,
+    "partial_rotary_factor": 0.25,
+    "rms_norm_eps": 1e-06,
+    "rope_parameters": {
+      "mrope_interleaved": true,
+      "mrope_section": [
+        11,
+        11,
+        10
+      ],
+      "partial_rotary_factor": 0.25,
+      "rope_theta": 10000000,
+      "rope_type": "default"
+    },
+    "router_aux_loss_coef": 0.001,
+    "shared_expert_intermediate_size": 512,
+    "tie_word_embeddings": false,
+    "use_cache": true,
+    "vocab_size": 248320
+  },
+  "tie_word_embeddings": false,
+  "transformers_version": "5.2.0",
+  "video_token_id": 248057,
+  "vision_config": {
+    "deepstack_visual_indexes": [],
+    "depth": 27,
+    "dtype": "bfloat16",
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 1152,
+    "in_channels": 3,
+    "initializer_range": 0.02,
+    "intermediate_size": 4304,
+    "model_type": "qwen3_5_moe",
+    "num_heads": 16,
+    "num_position_embeddings": 2304,
+    "out_hidden_size": 2048,
+    "patch_size": 16,
+    "spatial_merge_size": 2,
+    "temporal_patch_size": 2
+  },
+  "vision_end_token_id": 248054,
+  "vision_start_token_id": 248053
+}

examples/requirements.txt

@@ -0,0 +1,12 @@
+git+https://github.com/rkazants/optimum-intel.git@support_qwen3_5
+--extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly
+openvino
+openvino-tokenizers
+nncf
+transformers==5.2.0
+torch
+pillow
+requests
+torchvision
+opencv-python
+huggingface-hub

examples/run_qwen36_vlm_openvino.py

@@ -0,0 +1,122 @@
+from pathlib import Path
+from io import BytesIO
+import argparse
+import requests
+import torch
+from PIL import Image
+
+from transformers import AutoProcessor
+
+try:
+    from optimum.intel import OVModelForVisualCausalLM
+except ImportError:
+    from optimum.intel.openvino import OVModelForVisualCausalLM
+
+
+def load_image(path_or_url: str) -> Image.Image:
+    path_or_url = str(path_or_url)
+
+    if path_or_url.startswith(("http://", "https://")):
+        response = requests.get(path_or_url, timeout=30)
+        response.raise_for_status()
+        return Image.open(BytesIO(response.content)).convert("RGB")
+
+    return Image.open(path_or_url).convert("RGB")
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser()
+
+    parser.add_argument(
+        "--model-dir",
+        default=".",
+        help="Path to the OpenVINO-converted Qwen3.6 model directory.",
+    )
+
+    parser.add_argument(
+        "--device",
+        default="CPU",
+        help="OpenVINO device, for example CPU, GPU, GPU.0, or AUTO.",
+    )
+
+    parser.add_argument(
+        "--max-new-tokens",
+        type=int,
+        default=64,
+        help="Maximum number of newly generated tokens.",
+    )
+
+    parser.add_argument(
+        "--image",
+        default="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG",
+        help="Image path or image URL.",
+    )
+
+    parser.add_argument(
+        "--prompt",
+        default="What animal is on the candy? Answer in one sentence.",
+        help="Prompt for the image.",
+    )
+
+    args = parser.parse_args()
+
+    model_dir = Path(args.model_dir)
+
+    if not model_dir.exists():
+        raise FileNotFoundError(f"Model directory not found: {model_dir}")
+
+    print("Using model:", model_dir)
+    print("Using device:", args.device)
+
+    processor = AutoProcessor.from_pretrained(
+        model_dir,
+        trust_remote_code=True,
+    )
+
+    model = OVModelForVisualCausalLM.from_pretrained(
+        model_dir,
+        device=args.device,
+        trust_remote_code=True,
+    )
+
+    image = load_image(args.image)
+
+    messages = [
+        {
+            "role": "user",
+            "content": [
+                {"type": "image", "image": image},
+                {"type": "text", "text": args.prompt},
+            ],
+        }
+    ]
+
+    inputs = processor.apply_chat_template(
+        messages,
+        add_generation_prompt=True,
+        tokenize=True,
+        return_dict=True,
+        return_tensors="pt",
+    )
+
+    with torch.no_grad():
+        output_ids = model.generate(
+            **inputs,
+            max_new_tokens=args.max_new_tokens,
+            do_sample=False,
+        )
+
+    prompt_len = inputs["input_ids"].shape[-1]
+    generated_ids = output_ids[0][prompt_len:]
+
+    answer = processor.decode(
+        generated_ids,
+        skip_special_tokens=True,
+    ).strip()
+
+    print("\nGenerated answer:")
+    print(answer)
+
+
+if __name__ == "__main__":
+    main()

generation_config.json

@@ -0,0 +1,13 @@
+{
+  "bos_token_id": 248044,
+  "do_sample": true,
+  "eos_token_id": [
+    248046,
+    248044
+  ],
+  "pad_token_id": 248044,
+  "temperature": 1.0,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "5.2.0"
+}

openvino_config.json

@@ -0,0 +1,44 @@
+{
+  "dtype": "int4",
+  "input_info": null,
+  "optimum_version": "2.1.0.dev0",
+  "output_attentions": false,
+  "quantization_config": {
+    "_dataset_kwargs": {},
+    "dataset": null,
+    "default_config": {
+      "quant_method": "default"
+    },
+    "ignored_scope": null,
+    "num_samples": null,
+    "processor": "Qwen/Qwen3.6-35B-A3B",
+    "quantization_configs": {
+      "lm_model": {
+        "_dataset_kwargs": {},
+        "all_layers": null,
+        "backup_precision": null,
+        "bits": 4,
+        "dataset": null,
+        "dq_group_size": null,
+        "dtype": "int4",
+        "gptq": null,
+        "group_size": 128,
+        "group_size_fallback": null,
+        "ignored_scope": null,
+        "lora_correction": null,
+        "num_samples": null,
+        "processor": "Qwen/Qwen3.6-35B-A3B",
+        "quant_method": "default",
+        "ratio": 1.0,
+        "scale_estimation": null,
+        "sensitivity_metric": null,
+        "statistics_path": null,
+        "sym": false,
+        "tokenizer": "Qwen/Qwen3.6-35B-A3B"
+      }
+    },
+    "tokenizer": "Qwen/Qwen3.6-35B-A3B"
+  },
+  "save_onnx_model": false,
+  "transformers_version": "5.2.0"
+}

openvino_detokenizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48fc2c6a7a1056994982911eb9695826acef933c3d742364c56d16317a0b2a79
+size 3828184

openvino_detokenizer.xml

@@ -0,0 +1,221 @@
+<?xml version="1.0"?>
+<net name="detokenizer" version="11">
+	<layers>
+		<layer id="0" name="Parameter_489734" type="Parameter" version="opset1">
+			<data shape="?,?" element_type="i64" />
+			<output>
+				<port id="0" precision="I64" names="Parameter_489734">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="1" name="Convert_489769" type="Convert" version="opset1">
+			<data destination_type="i32" />
+			<input>
+				<port id="0" precision="I64">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="2" name="Constant_489736" type="Const" version="opset1">
+			<data element_type="i32" shape="248077" offset="0" size="992308" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>248077</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="3" name="Constant_489738" type="Const" version="opset1">
+			<data element_type="i32" shape="248077" offset="992308" size="992308" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>248077</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="4" name="Constant_489740" type="Const" version="opset1">
+			<data element_type="u8" shape="1843484" offset="1984616" size="1843484" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>1843484</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="5" name="Slice_489745" type="Const" version="opset1">
+			<data element_type="i32" shape="21" offset="3828100" size="84" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>21</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="6" name="VocabDecoder_489747" type="VocabDecoder" version="extension">
+			<data skip_tokens="" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>248077</dim>
+				</port>
+				<port id="2" precision="I32">
+					<dim>248077</dim>
+				</port>
+				<port id="3" precision="U8">
+					<dim>1843484</dim>
+				</port>
+				<port id="4" precision="I32">
+					<dim>21</dim>
+				</port>
+			</input>
+			<output>
+				<port id="5" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="6" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="7" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="8" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="9" precision="U8">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="7" name="FuzeRagged_489748" type="FuzeRagged" version="extension">
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="3" precision="I32">
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="4" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="5" precision="I32">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="8" name="UTF8Validate_489749" type="UTF8Validate" version="extension">
+			<data replace_mode="true" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="U8">
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="3" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="4" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="5" precision="U8">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="9" name="StringTensorPack_489750" type="StringTensorPack" version="opset15">
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="U8">
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="3" precision="STRING" names="Result_489751,string_output">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="10" name="Result_489751" type="Result" version="opset1" output_names="Result_489751,string_output">
+			<input>
+				<port id="0" precision="STRING">
+					<dim>-1</dim>
+				</port>
+			</input>
+		</layer>
+	</layers>
+	<edges>
+		<edge from-layer="0" from-port="0" to-layer="1" to-port="0" />
+		<edge from-layer="1" from-port="1" to-layer="6" to-port="0" />
+		<edge from-layer="2" from-port="0" to-layer="6" to-port="1" />
+		<edge from-layer="3" from-port="0" to-layer="6" to-port="2" />
+		<edge from-layer="4" from-port="0" to-layer="6" to-port="3" />
+		<edge from-layer="5" from-port="0" to-layer="6" to-port="4" />
+		<edge from-layer="6" from-port="7" to-layer="7" to-port="2" />
+		<edge from-layer="6" from-port="9" to-layer="8" to-port="2" />
+		<edge from-layer="6" from-port="8" to-layer="7" to-port="3" />
+		<edge from-layer="6" from-port="6" to-layer="7" to-port="1" />
+		<edge from-layer="6" from-port="5" to-layer="7" to-port="0" />
+		<edge from-layer="7" from-port="4" to-layer="8" to-port="0" />
+		<edge from-layer="7" from-port="5" to-layer="8" to-port="1" />
+		<edge from-layer="8" from-port="3" to-layer="9" to-port="0" />
+		<edge from-layer="8" from-port="4" to-layer="9" to-port="1" />
+		<edge from-layer="8" from-port="5" to-layer="9" to-port="2" />
+		<edge from-layer="9" from-port="3" to-layer="10" to-port="0" />
+	</edges>
+	<rt_info>
+		<info name="OpenVINO Runtime" value="2026.2.0-21860-7be1ffd093f" />
+		<add_attention_mask value="True" />
+		<add_prefix_space />
+		<add_special_tokens value="True" />
+		<chat_template value="{%- set image_count = namespace(value=0) %}&#10;{%- set video_count = namespace(value=0) %}&#10;{%- macro render_content(content, do_vision_count, is_system_content=false) %}&#10;    {%- if content is string %}&#10;        {{- content }}&#10;    {%- elif content is iterable and content is not mapping %}&#10;        {%- for item in content %}&#10;            {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}&#10;                {%- if is_system_content %}&#10;                    {{- raise_exception('System message cannot contain images.') }}&#10;                {%- endif %}&#10;                {%- if do_vision_count %}&#10;                    {%- set image_count.value = image_count.value + 1 %}&#10;                {%- endif %}&#10;                {%- if add_vision_id %}&#10;                    {{- 'Picture ' ~ image_count.value ~ ': ' }}&#10;                {%- endif %}&#10;                {{- '&lt;|vision_start|>&lt;|image_pad|>&lt;|vision_end|>' }}&#10;            {%- elif 'video' in item or item.type == 'video' %}&#10;                {%- if is_system_content %}&#10;                    {{- raise_exception('System message cannot contain videos.') }}&#10;                {%- endif %}&#10;                {%- if do_vision_count %}&#10;                    {%- set video_count.value = video_count.value + 1 %}&#10;                {%- endif %}&#10;                {%- if add_vision_id %}&#10;                    {{- 'Video ' ~ video_count.value ~ ': ' }}&#10;                {%- endif %}&#10;                {{- '&lt;|vision_start|>&lt;|video_pad|>&lt;|vision_end|>' }}&#10;            {%- elif 'text' in item %}&#10;                {{- item.text }}&#10;            {%- else %}&#10;                {{- raise_exception('Unexpected item type in content.') }}&#10;            {%- endif %}&#10;        {%- endfor %}&#10;    {%- elif content is none or content is undefined %}&#10;        {{- '' }}&#10;    {%- else %}&#10;        {{- raise_exception('Unexpected content type.') }}&#10;    {%- endif %}&#10;{%- endmacro %}&#10;{%- if not messages %}&#10;    {{- raise_exception('No messages provided.') }}&#10;{%- endif %}&#10;{%- if tools and tools is iterable and tools is not mapping %}&#10;    {{- '&lt;|im_start|>system\n' }}&#10;    {{- &quot;# Tools\n\nYou have access to the following functions:\n\n&lt;tools>&quot; }}&#10;    {%- for tool in tools %}&#10;        {{- &quot;\n&quot; }}&#10;        {{- tool | tojson }}&#10;    {%- endfor %}&#10;    {{- &quot;\n&lt;/tools>&quot; }}&#10;    {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n&lt;tool_call>\n&lt;function=example_function_name>\n&lt;parameter=example_parameter_1>\nvalue_1\n&lt;/parameter>\n&lt;parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n&lt;/parameter>\n&lt;/function>\n&lt;/tool_call>\n\n&lt;IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner &lt;function=...>&lt;/function> block must be nested within &lt;tool_call>&lt;/tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n&lt;/IMPORTANT>' }}&#10;    {%- if messages[0].role == 'system' %}&#10;        {%- set content = render_content(messages[0].content, false, true)|trim %}&#10;        {%- if content %}&#10;            {{- '\n\n' + content }}&#10;        {%- endif %}&#10;    {%- endif %}&#10;    {{- '&lt;|im_end|>\n' }}&#10;{%- else %}&#10;    {%- if messages[0].role == 'system' %}&#10;        {%- set content = render_content(messages[0].content, false, true)|trim %}&#10;        {{- '&lt;|im_start|>system\n' + content + '&lt;|im_end|>\n' }}&#10;    {%- endif %}&#10;{%- endif %}&#10;{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}&#10;{%- for message in messages[::-1] %}&#10;    {%- set index = (messages|length - 1) - loop.index0 %}&#10;    {%- if ns.multi_step_tool and message.role == &quot;user&quot; %}&#10;        {%- set content = render_content(message.content, false)|trim %}&#10;        {%- if not(content.startswith('&lt;tool_response>') and content.endswith('&lt;/tool_response>')) %}&#10;            {%- set ns.multi_step_tool = false %}&#10;            {%- set ns.last_query_index = index %}&#10;        {%- endif %}&#10;    {%- endif %}&#10;{%- endfor %}&#10;{%- if ns.multi_step_tool %}&#10;    {{- raise_exception('No user query found in messages.') }}&#10;{%- endif %}&#10;{%- for message in messages %}&#10;    {%- set content = render_content(message.content, true)|trim %}&#10;    {%- if message.role == &quot;system&quot; %}&#10;        {%- if not loop.first %}&#10;            {{- raise_exception('System message must be at the beginning.') }}&#10;        {%- endif %}&#10;    {%- elif message.role == &quot;user&quot; %}&#10;        {{- '&lt;|im_start|>' + message.role + '\n' + content + '&lt;|im_end|>' + '\n' }}&#10;    {%- elif message.role == &quot;assistant&quot; %}&#10;        {%- set reasoning_content = '' %}&#10;        {%- if message.reasoning_content is string %}&#10;            {%- set reasoning_content = message.reasoning_content %}&#10;        {%- else %}&#10;            {%- if '&lt;/think>' in content %}&#10;                {%- set reasoning_content = content.split('&lt;/think>')[0].rstrip('\n').split('&lt;think>')[-1].lstrip('\n') %}&#10;                {%- set content = content.split('&lt;/think>')[-1].lstrip('\n') %}&#10;            {%- endif %}&#10;        {%- endif %}&#10;        {%- set reasoning_content = reasoning_content|trim %}&#10;        {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}&#10;            {{- '&lt;|im_start|>' + message.role + '\n&lt;think>\n' + reasoning_content + '\n&lt;/think>\n\n' + content }}&#10;        {%- else %}&#10;            {{- '&lt;|im_start|>' + message.role + '\n' + content }}&#10;        {%- endif %}&#10;        {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}&#10;            {%- for tool_call in message.tool_calls %}&#10;                {%- if tool_call.function is defined %}&#10;                    {%- set tool_call = tool_call.function %}&#10;                {%- endif %}&#10;                {%- if loop.first %}&#10;                    {%- if content|trim %}&#10;                        {{- '\n\n&lt;tool_call>\n&lt;function=' + tool_call.name + '>\n' }}&#10;                    {%- else %}&#10;                        {{- '&lt;tool_call>\n&lt;function=' + tool_call.name + '>\n' }}&#10;                    {%- endif %}&#10;                {%- else %}&#10;                    {{- '\n&lt;tool_call>\n&lt;function=' + tool_call.name + '>\n' }}&#10;                {%- endif %}&#10;                {%- if tool_call.arguments is defined %}&#10;                    {%- for args_name, args_value in tool_call.arguments|items %}&#10;                        {{- '&lt;parameter=' + args_name + '>\n' }}&#10;                        {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}&#10;                        {{- args_value }}&#10;                        {{- '\n&lt;/parameter>\n' }}&#10;                    {%- endfor %}&#10;                {%- endif %}&#10;                {{- '&lt;/function>\n&lt;/tool_call>' }}&#10;            {%- endfor %}&#10;        {%- endif %}&#10;        {{- '&lt;|im_end|>\n' }}&#10;    {%- elif message.role == &quot;tool&quot; %}&#10;        {%- if loop.previtem and loop.previtem.role != &quot;tool&quot; %}&#10;            {{- '&lt;|im_start|>user' }}&#10;        {%- endif %}&#10;        {{- '\n&lt;tool_response>\n' }}&#10;        {{- content }}&#10;        {{- '\n&lt;/tool_response>' }}&#10;        {%- if not loop.last and loop.nextitem.role != &quot;tool&quot; %}&#10;            {{- '&lt;|im_end|>\n' }}&#10;        {%- elif loop.last %}&#10;            {{- '&lt;|im_end|>\n' }}&#10;        {%- endif %}&#10;    {%- else %}&#10;        {{- raise_exception('Unexpected message role.') }}&#10;    {%- endif %}&#10;{%- endfor %}&#10;{%- if add_generation_prompt %}&#10;    {{- '&lt;|im_start|>assistant\n' }}&#10;    {%- if enable_thinking is defined and enable_thinking is false %}&#10;        {{- '&lt;think>\n\n&lt;/think>\n\n' }}&#10;    {%- else %}&#10;        {{- '&lt;think>\n' }}&#10;    {%- endif %}&#10;{%- endif %}" />
+		<clean_up_tokenization_spaces />
+		<detokenizer_input_type value="i64" />
+		<eos_token_id value="248046" />
+		<handle_special_tokens_with_re />
+		<max_length />
+		<number_of_inputs value="1" />
+		<openvino_tokenizers_version value="2026.2.0.0-677-b5eb36f08b1" />
+		<openvino_version value="2026.2.0-21860-7be1ffd093f" />
+		<original_post_processor_template value="{&quot;type&quot;: &quot;ByteLevel&quot;, &quot;add_prefix_space&quot;: false, &quot;trim_offsets&quot;: false, &quot;use_regex&quot;: false}" />
+		<original_tokenizer_class value="&lt;class 'transformers.tokenization_utils_tokenizers.TokenizersBackend'>" />
+		<pad_token_id value="248044" />
+		<processed_post_processor_template value="{&quot;single&quot;: {&quot;ids&quot;: [-1], &quot;type_ids&quot;: [0]}, &quot;pair&quot;: {&quot;ids&quot;: [-1, -2], &quot;type_ids&quot;: [0, 0]}}" />
+		<sentencepiece_version value="0.2.1" />
+		<skip_special_tokens value="True" />
+		<streaming_detokenizer value="False" />
+		<tokenizer_output_type value="i64" />
+		<tokenizers_version value="0.22.2" />
+		<transformers_version value="5.2.0" />
+		<use_max_padding value="False" />
+		<use_sentencepiece_backend value="False" />
+		<utf8_replace_mode value="replace" />
+		<with_detokenizer value="True" />
+	</rt_info>
+</net>

openvino_language_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ffdcd60a3cb7332c5eabd3a4d9ec2c6c80498d42b62e722b5cb19d93fde9f29
+size 17989975554

openvino_text_embeddings_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9957db96d6aca2a8cb0e193cf970f801dbbea183cfd01bf04e649405d9b8b51
+size 509056004

openvino_text_embeddings_model.xml

@@ -0,0 +1,180 @@
+<?xml version="1.0"?>
+<net name="Model69" version="11">
+	<layers>
+		<layer id="0" name="input" type="Parameter" version="opset1">
+			<data shape="?,?" element_type="i64" />
+			<output>
+				<port id="0" precision="I64" names="input">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="1" name="self.weight" type="Const" version="opset1">
+			<data element_type="i8" shape="248320, 2048" offset="0" size="508559360" />
+			<output>
+				<port id="0" precision="I8">
+					<dim>248320</dim>
+					<dim>2048</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="2" name="Convert_3160785" type="Convert" version="opset1">
+			<data destination_type="f16" />
+			<input>
+				<port id="0" precision="I8">
+					<dim>248320</dim>
+					<dim>2048</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="FP16">
+					<dim>248320</dim>
+					<dim>2048</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="3" name="self.weight/scale" type="Const" version="opset1">
+			<data element_type="f16" shape="248320, 1" offset="508559360" size="496640" />
+			<output>
+				<port id="0" precision="FP16">
+					<dim>248320</dim>
+					<dim>1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="4" name="self.weight/fq_weights_0" type="Multiply" version="opset1">
+			<data auto_broadcast="numpy" />
+			<input>
+				<port id="0" precision="FP16">
+					<dim>248320</dim>
+					<dim>2048</dim>
+				</port>
+				<port id="1" precision="FP16">
+					<dim>248320</dim>
+					<dim>1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="2" precision="FP16">
+					<dim>248320</dim>
+					<dim>2048</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="5" name="ov_ext::embedding/Convert" type="Convert" version="opset1">
+			<data destination_type="f32" />
+			<rt_info>
+				<attribute name="decompression" version="0" />
+			</rt_info>
+			<input>
+				<port id="0" precision="FP16">
+					<dim>248320</dim>
+					<dim>2048</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="FP32">
+					<dim>248320</dim>
+					<dim>2048</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="6" name="ov_ext::embedding/Convert_1" type="Convert" version="opset1">
+			<data destination_type="i32" />
+			<input>
+				<port id="0" precision="I64">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="7" name="ov_ext::embedding/Constant" type="Const" version="opset1">
+			<data element_type="i32" shape="" offset="509056000" size="4" />
+			<output>
+				<port id="0" precision="I32" />
+			</output>
+		</layer>
+		<layer id="8" name="ov_ext::embedding/Gather" type="Gather" version="opset8">
+			<data batch_dims="0" />
+			<input>
+				<port id="0" precision="FP32">
+					<dim>248320</dim>
+					<dim>2048</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="I32" />
+			</input>
+			<output>
+				<port id="3" precision="FP32" names="inputs_embeds">
+					<dim>-1</dim>
+					<dim>-1</dim>
+					<dim>2048</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="9" name="Result_485002" type="Result" version="opset1" output_names="inputs_embeds">
+			<input>
+				<port id="0" precision="FP32">
+					<dim>-1</dim>
+					<dim>-1</dim>
+					<dim>2048</dim>
+				</port>
+			</input>
+		</layer>
+	</layers>
+	<edges>
+		<edge from-layer="0" from-port="0" to-layer="6" to-port="0" />
+		<edge from-layer="1" from-port="0" to-layer="2" to-port="0" />
+		<edge from-layer="2" from-port="1" to-layer="4" to-port="0" />
+		<edge from-layer="3" from-port="0" to-layer="4" to-port="1" />
+		<edge from-layer="4" from-port="2" to-layer="5" to-port="0" />
+		<edge from-layer="5" from-port="1" to-layer="8" to-port="0" />
+		<edge from-layer="6" from-port="1" to-layer="8" to-port="1" />
+		<edge from-layer="7" from-port="0" to-layer="8" to-port="2" />
+		<edge from-layer="8" from-port="3" to-layer="9" to-port="0" />
+	</edges>
+	<rt_info>
+		<info name="OpenVINO Runtime" value="2026.2.0-21860-7be1ffd093f" />
+		<Runtime_version value="2026.2.0-21860-7be1ffd093f" />
+		<conversion_parameters>
+			<framework value="pytorch" />
+			<is_python_object value="True" />
+		</conversion_parameters>
+		<nncf>
+			<friendly_names_were_updated value="True" />
+			<version value="3.1.0" />
+			<weight_compression>
+				<advanced_parameters value="{'statistics_path': None, 'lora_adapter_rank': 256, 'group_size_fallback_mode': 'error', 'min_adjusted_group_size': 32, 'awq_params': {'subset_size': 32, 'percent_to_apply': 0.002, 'alpha_min': 0.0, 'alpha_max': 1.0, 'steps': 100, 'prefer_data_aware_scaling': True}, 'scale_estimation_params': {'subset_size': 64, 'initial_steps': 5, 'scale_steps': 5, 'weight_penalty': -1.0}, 'gptq_params': {'damp_percent': 0.1, 'block_size': 128, 'subset_size': 128}, 'lora_correction_params': {'adapter_rank': 8, 'num_iterations': 3, 'apply_regularization': True, 'subset_size': 128, 'use_int8_adapters': True}, 'backend_params': {}, 'codebook': None, 'adaptive_codebook_params': {'value_type': 'f8e4m3', 'across_blocks': False, 'num_elements': 16}}" />
+				<all_layers value="False" />
+				<awq value="False" />
+				<backup_mode value="int8_asym" />
+				<compression_format value="dequantize" />
+				<gptq value="False" />
+				<group_size value="-1" />
+				<ignored_scope value="[]" />
+				<lora_correction value="False" />
+				<mode value="int8_sym" />
+				<ratio value="1.0" />
+				<scale_estimation value="False" />
+				<sensitivity_metric value="weight_quantization_error" />
+			</weight_compression>
+		</nncf>
+		<optimum>
+			<nncf_version value="3.1.0" />
+			<optimum_intel_version value="1.27.0.dev0+ff1fb6d" />
+			<optimum_version value="2.1.0.dev0" />
+			<pytorch_version value="2.11.0" />
+			<transformers_version value="5.2.0" />
+		</optimum>
+	</rt_info>
+</net>

openvino_tokenizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea3fcce2251902efffbc774cbc8ffff69e981412347a1e510e430c6557a63701
+size 9631154

openvino_tokenizer.xml

@@ -0,0 +1,772 @@
+<?xml version="1.0"?>
+<net name="tokenizer" version="11">
+	<layers>
+		<layer id="0" name="Parameter_489606" type="Parameter" version="opset1">
+			<data shape="?" element_type="string" />
+			<output>
+				<port id="0" precision="STRING" names="Parameter_489606">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="1" name="Constant_489612" type="Const" version="opset1">
+			<data element_type="i64" shape="" offset="0" size="8" />
+			<output>
+				<port id="0" precision="I64" />
+			</output>
+		</layer>
+		<layer id="2" name="StringTensorUnpack_489607" type="StringTensorUnpack" version="opset15">
+			<input>
+				<port id="0" precision="STRING">
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="3" precision="U8">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="3" name="ShapeOf_489608" type="ShapeOf" version="opset3">
+			<data output_type="i64" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="I64">
+					<dim>1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="4" name="Constant_489609" type="Const" version="opset1">
+			<data element_type="i64" shape="" offset="0" size="8" />
+			<output>
+				<port id="0" precision="I64" />
+			</output>
+		</layer>
+		<layer id="5" name="Constant_489610" type="Const" version="opset1">
+			<data element_type="i64" shape="" offset="0" size="8" />
+			<output>
+				<port id="0" precision="I64" />
+			</output>
+		</layer>
+		<layer id="6" name="Gather_489611" type="Gather" version="opset8">
+			<data batch_dims="0" />
+			<input>
+				<port id="0" precision="I64">
+					<dim>1</dim>
+				</port>
+				<port id="1" precision="I64" />
+				<port id="2" precision="I64" />
+			</input>
+			<output>
+				<port id="3" precision="I64" />
+			</output>
+		</layer>
+		<layer id="7" name="Constant_489613" type="Const" version="opset1">
+			<data element_type="i64" shape="" offset="8" size="8" />
+			<output>
+				<port id="0" precision="I64" />
+			</output>
+		</layer>
+		<layer id="8" name="Range_489614" type="Range" version="opset4">
+			<data output_type="i32" />
+			<input>
+				<port id="0" precision="I64" />
+				<port id="1" precision="I64" />
+				<port id="2" precision="I64" />
+			</input>
+			<output>
+				<port id="3" precision="I32">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="9" name="Constant_489615" type="Const" version="opset1">
+			<data element_type="i64" shape="" offset="8" size="8" />
+			<output>
+				<port id="0" precision="I64" />
+			</output>
+		</layer>
+		<layer id="10" name="Constant_489616" type="Const" version="opset1">
+			<data element_type="i64" shape="" offset="8" size="8" />
+			<output>
+				<port id="0" precision="I64" />
+			</output>
+		</layer>
+		<layer id="11" name="Add_489617" type="Add" version="opset1">
+			<data auto_broadcast="numpy" />
+			<input>
+				<port id="0" precision="I64" />
+				<port id="1" precision="I64" />
+			</input>
+			<output>
+				<port id="2" precision="I64" />
+			</output>
+		</layer>
+		<layer id="12" name="Constant_489618" type="Const" version="opset1">
+			<data element_type="i64" shape="" offset="8" size="8" />
+			<output>
+				<port id="0" precision="I64" />
+			</output>
+		</layer>
+		<layer id="13" name="Range_489619" type="Range" version="opset4">
+			<data output_type="i32" />
+			<input>
+				<port id="0" precision="I64" />
+				<port id="1" precision="I64" />
+				<port id="2" precision="I64" />
+			</input>
+			<output>
+				<port id="3" precision="I32">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="14" name="Constant_489683" type="Const" version="opset1">
+			<data element_type="u8" shape="588" offset="16" size="588" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>588</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="15" name="SpecialTokensSplit_489684" type="SpecialTokensSplit" version="extension">
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="3" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="4" precision="U8">
+					<dim>-1</dim>
+				</port>
+				<port id="5" precision="U8">
+					<dim>588</dim>
+				</port>
+			</input>
+			<output>
+				<port id="6" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="7" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="8" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="9" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="10" precision="U8">
+					<dim>-1</dim>
+				</port>
+				<port id="11" precision="BOOL">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="16" name="CharsMapNormalization_489685" type="CharsMapNormalization" version="extension">
+			<data add_dummy_prefix="false" remove_extra_whitespaces="false" escape_whitespaces="false" normalization_form="nfc" case_fold="false" nmt="false" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="U8">
+					<dim>-1</dim>
+				</port>
+				<port id="3" precision="BOOL">
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="4" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="5" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="6" precision="U8">
+					<dim>-1</dim>
+				</port>
+				<port id="7" precision="BOOL">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="17" name="Constant_489687" type="Const" version="opset1">
+			<data element_type="u8" shape="122" offset="604" size="122" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>122</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="18" name="RegexSplit_489688" type="RegexSplit" version="extension">
+			<data behaviour="isolate" invert="false" max_splits="-1" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="3" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="4" precision="U8">
+					<dim>-1</dim>
+				</port>
+				<port id="5" precision="BOOL">
+					<dim>-1</dim>
+				</port>
+				<port id="6" precision="U8">
+					<dim>122</dim>
+				</port>
+			</input>
+			<output>
+				<port id="7" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="8" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="9" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="10" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="11" precision="U8">
+					<dim>-1</dim>
+				</port>
+				<port id="12" precision="BOOL">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="19" name="Constant_489690" type="Const" version="opset1">
+			<data element_type="i32" shape="248077" offset="726" size="992308" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>248077</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="20" name="Constant_489692" type="Const" version="opset1">
+			<data element_type="i32" shape="248077" offset="993034" size="992308" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>248077</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="21" name="Constant_489694" type="Const" version="opset1">
+			<data element_type="u8" shape="1843484" offset="1985342" size="1843484" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>1843484</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="22" name="Constant_489702" type="Const" version="opset1">
+			<data element_type="i32" shape="247587" offset="3828826" size="990348" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>247587</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="23" name="Constant_489704" type="Const" version="opset1">
+			<data element_type="i32" shape="247587" offset="4819174" size="990348" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>247587</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="24" name="Constant_489706" type="Const" version="opset1">
+			<data element_type="u8" shape="938547" offset="5809522" size="938547" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>938547</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="25" name="Constant_489708" type="Const" version="opset1">
+			<data element_type="i32" shape="247587" offset="6748069" size="990348" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>247587</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="26" name="Constant_489710" type="Const" version="opset1">
+			<data element_type="i32" shape="247587" offset="7738417" size="990348" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>247587</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="27" name="Constant_489712" type="Const" version="opset1">
+			<data element_type="u8" shape="901525" offset="8728765" size="901525" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>901525</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="28" name="Constant_489696" type="Const" version="opset1">
+			<data element_type="i32" shape="33" offset="9630290" size="132" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>33</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="29" name="Constant_489698" type="Const" version="opset1">
+			<data element_type="i32" shape="33" offset="9630422" size="132" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>33</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="30" name="Constant_489700" type="Const" version="opset1">
+			<data element_type="u8" shape="439" offset="9630554" size="439" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>439</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="31" name="Constant_489713" type="Const" version="opset1">
+			<data element_type="i32" shape="33" offset="9630993" size="132" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>33</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="32" name="BPETokenizer_489714" type="BPETokenizer" version="extension">
+			<data unk_token="" fuse_unk="false" suffix_indicator="" end_suffix="" byte_fallback="false" cache_capacity="49608" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="3" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="4" precision="U8">
+					<dim>-1</dim>
+				</port>
+				<port id="5" precision="I32">
+					<dim>248077</dim>
+				</port>
+				<port id="6" precision="I32">
+					<dim>248077</dim>
+				</port>
+				<port id="7" precision="U8">
+					<dim>1843484</dim>
+				</port>
+				<port id="8" precision="I32">
+					<dim>247587</dim>
+				</port>
+				<port id="9" precision="I32">
+					<dim>247587</dim>
+				</port>
+				<port id="10" precision="U8">
+					<dim>938547</dim>
+				</port>
+				<port id="11" precision="I32">
+					<dim>247587</dim>
+				</port>
+				<port id="12" precision="I32">
+					<dim>247587</dim>
+				</port>
+				<port id="13" precision="U8">
+					<dim>901525</dim>
+				</port>
+				<port id="14" precision="I32">
+					<dim>33</dim>
+				</port>
+				<port id="15" precision="I32">
+					<dim>33</dim>
+				</port>
+				<port id="16" precision="U8">
+					<dim>439</dim>
+				</port>
+				<port id="17" precision="I32">
+					<dim>33</dim>
+				</port>
+			</input>
+			<output>
+				<port id="18" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="19" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="20" precision="I32">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="33" name="Constant_489715" type="Const" version="opset1">
+			<data element_type="i32" shape="" offset="9631125" size="4" />
+			<output>
+				<port id="0" precision="I32" />
+			</output>
+		</layer>
+		<layer id="34" name="Constant_489717" type="Const" version="opset1">
+			<data element_type="u8" shape="4" offset="9631129" size="4" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>4</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="35" name="Constant_489719" type="Const" version="opset1">
+			<data element_type="u8" shape="13" offset="9631133" size="13" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>13</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="36" name="Truncate_489720" type="Truncate" version="extension">
+			<data m_num_inputs="1" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="3" precision="I32" />
+				<port id="4" precision="U8">
+					<dim>4</dim>
+				</port>
+				<port id="5" precision="U8">
+					<dim>13</dim>
+				</port>
+			</input>
+			<output>
+				<port id="6" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="7" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="8" precision="I32">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="37" name="Constant_489721" type="Const" version="opset1">
+			<data element_type="i32" shape="1" offset="9631146" size="4" />
+			<output>
+				<port id="0" precision="I32">
+					<dim>1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="38" name="CombineSegments_489722" type="CombineSegments" version="extension">
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="3" precision="I32">
+					<dim>1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="4" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="5" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="6" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="7" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="8" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="9" precision="I32">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="39" name="Subtract_489723" type="Subtract" version="opset1">
+			<data auto_broadcast="numpy" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="2" precision="I32">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="40" name="Constant_489724" type="Const" version="opset1">
+			<data element_type="i32" shape="" offset="9631146" size="4" />
+			<output>
+				<port id="0" precision="I32" />
+			</output>
+		</layer>
+		<layer id="41" name="ReduceMax_489725" type="ReduceMax" version="opset1">
+			<data keep_dims="false" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32" />
+			</input>
+			<output>
+				<port id="2" precision="I32" />
+			</output>
+		</layer>
+		<layer id="42" name="Constant_489726" type="Const" version="opset1">
+			<data element_type="i32" shape="" offset="9631150" size="4" />
+			<output>
+				<port id="0" precision="I32" />
+			</output>
+		</layer>
+		<layer id="43" name="RaggedToDense_489727" type="RaggedToDense" version="extension">
+			<data pad_right="false" m_pad_max_length="false" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="3" precision="I32" />
+				<port id="4" precision="I32" />
+			</input>
+			<output>
+				<port id="5" precision="I32">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+				<port id="6" precision="BOOL">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="44" name="Convert_489728" type="Convert" version="opset1">
+			<data destination_type="i32" />
+			<input>
+				<port id="0" precision="BOOL">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="45" name="Convert_489728.0" type="Convert" version="opset1">
+			<data destination_type="i64" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="I64" names="attention_mask">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="46" name="RaggedToDense_489727.0" type="Convert" version="opset1">
+			<data destination_type="i64" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="I64" names="input_ids">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="47" name="Result_489731" type="Result" version="opset1" output_names="input_ids">
+			<input>
+				<port id="0" precision="I64">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</input>
+		</layer>
+		<layer id="48" name="Result_489733" type="Result" version="opset1" output_names="attention_mask">
+			<input>
+				<port id="0" precision="I64">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</input>
+		</layer>
+	</layers>
+	<edges>
+		<edge from-layer="0" from-port="0" to-layer="2" to-port="0" />
+		<edge from-layer="1" from-port="0" to-layer="8" to-port="0" />
+		<edge from-layer="2" from-port="1" to-layer="3" to-port="0" />
+		<edge from-layer="2" from-port="3" to-layer="15" to-port="4" />
+		<edge from-layer="2" from-port="2" to-layer="15" to-port="3" />
+		<edge from-layer="2" from-port="1" to-layer="15" to-port="2" />
+		<edge from-layer="3" from-port="1" to-layer="6" to-port="0" />
+		<edge from-layer="4" from-port="0" to-layer="6" to-port="1" />
+		<edge from-layer="5" from-port="0" to-layer="6" to-port="2" />
+		<edge from-layer="6" from-port="3" to-layer="11" to-port="0" />
+		<edge from-layer="6" from-port="3" to-layer="8" to-port="1" />
+		<edge from-layer="7" from-port="0" to-layer="8" to-port="2" />
+		<edge from-layer="8" from-port="3" to-layer="15" to-port="0" />
+		<edge from-layer="9" from-port="0" to-layer="13" to-port="0" />
+		<edge from-layer="10" from-port="0" to-layer="11" to-port="1" />
+		<edge from-layer="11" from-port="2" to-layer="13" to-port="1" />
+		<edge from-layer="12" from-port="0" to-layer="13" to-port="2" />
+		<edge from-layer="13" from-port="3" to-layer="15" to-port="1" />
+		<edge from-layer="14" from-port="0" to-layer="15" to-port="5" />
+		<edge from-layer="15" from-port="10" to-layer="16" to-port="2" />
+		<edge from-layer="15" from-port="7" to-layer="18" to-port="1" />
+		<edge from-layer="15" from-port="6" to-layer="18" to-port="0" />
+		<edge from-layer="15" from-port="11" to-layer="16" to-port="3" />
+		<edge from-layer="15" from-port="9" to-layer="16" to-port="1" />
+		<edge from-layer="15" from-port="8" to-layer="16" to-port="0" />
+		<edge from-layer="16" from-port="4" to-layer="18" to-port="2" />
+		<edge from-layer="16" from-port="5" to-layer="18" to-port="3" />
+		<edge from-layer="16" from-port="6" to-layer="18" to-port="4" />
+		<edge from-layer="16" from-port="7" to-layer="18" to-port="5" />
+		<edge from-layer="17" from-port="0" to-layer="18" to-port="6" />
+		<edge from-layer="18" from-port="7" to-layer="32" to-port="0" />
+		<edge from-layer="18" from-port="8" to-layer="32" to-port="1" />
+		<edge from-layer="18" from-port="9" to-layer="32" to-port="2" />
+		<edge from-layer="18" from-port="10" to-layer="32" to-port="3" />
+		<edge from-layer="18" from-port="11" to-layer="32" to-port="4" />
+		<edge from-layer="19" from-port="0" to-layer="32" to-port="5" />
+		<edge from-layer="20" from-port="0" to-layer="32" to-port="6" />
+		<edge from-layer="21" from-port="0" to-layer="32" to-port="7" />
+		<edge from-layer="22" from-port="0" to-layer="32" to-port="8" />
+		<edge from-layer="23" from-port="0" to-layer="32" to-port="9" />
+		<edge from-layer="24" from-port="0" to-layer="32" to-port="10" />
+		<edge from-layer="25" from-port="0" to-layer="32" to-port="11" />
+		<edge from-layer="26" from-port="0" to-layer="32" to-port="12" />
+		<edge from-layer="27" from-port="0" to-layer="32" to-port="13" />
+		<edge from-layer="28" from-port="0" to-layer="32" to-port="14" />
+		<edge from-layer="29" from-port="0" to-layer="32" to-port="15" />
+		<edge from-layer="30" from-port="0" to-layer="32" to-port="16" />
+		<edge from-layer="31" from-port="0" to-layer="32" to-port="17" />
+		<edge from-layer="32" from-port="18" to-layer="36" to-port="0" />
+		<edge from-layer="32" from-port="19" to-layer="36" to-port="1" />
+		<edge from-layer="32" from-port="20" to-layer="36" to-port="2" />
+		<edge from-layer="33" from-port="0" to-layer="36" to-port="3" />
+		<edge from-layer="34" from-port="0" to-layer="36" to-port="4" />
+		<edge from-layer="35" from-port="0" to-layer="36" to-port="5" />
+		<edge from-layer="36" from-port="6" to-layer="38" to-port="0" />
+		<edge from-layer="36" from-port="8" to-layer="38" to-port="2" />
+		<edge from-layer="36" from-port="7" to-layer="38" to-port="1" />
+		<edge from-layer="37" from-port="0" to-layer="38" to-port="3" />
+		<edge from-layer="38" from-port="5" to-layer="39" to-port="0" />
+		<edge from-layer="38" from-port="4" to-layer="39" to-port="1" />
+		<edge from-layer="38" from-port="4" to-layer="43" to-port="0" />
+		<edge from-layer="38" from-port="5" to-layer="43" to-port="1" />
+		<edge from-layer="38" from-port="6" to-layer="43" to-port="2" />
+		<edge from-layer="39" from-port="2" to-layer="41" to-port="0" />
+		<edge from-layer="40" from-port="0" to-layer="41" to-port="1" />
+		<edge from-layer="41" from-port="2" to-layer="43" to-port="3" />
+		<edge from-layer="42" from-port="0" to-layer="43" to-port="4" />
+		<edge from-layer="43" from-port="6" to-layer="44" to-port="0" />
+		<edge from-layer="43" from-port="5" to-layer="46" to-port="0" />
+		<edge from-layer="44" from-port="1" to-layer="45" to-port="0" />
+		<edge from-layer="45" from-port="1" to-layer="48" to-port="0" />
+		<edge from-layer="46" from-port="1" to-layer="47" to-port="0" />
+	</edges>
+	<rt_info>
+		<info name="OpenVINO Runtime" value="2026.2.0-21860-7be1ffd093f" />
+		<add_attention_mask value="True" />
+		<add_prefix_space />
+		<add_special_tokens value="True" />
+		<chat_template value="{%- set image_count = namespace(value=0) %}&#10;{%- set video_count = namespace(value=0) %}&#10;{%- macro render_content(content, do_vision_count, is_system_content=false) %}&#10;    {%- if content is string %}&#10;        {{- content }}&#10;    {%- elif content is iterable and content is not mapping %}&#10;        {%- for item in content %}&#10;            {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}&#10;                {%- if is_system_content %}&#10;                    {{- raise_exception('System message cannot contain images.') }}&#10;                {%- endif %}&#10;                {%- if do_vision_count %}&#10;                    {%- set image_count.value = image_count.value + 1 %}&#10;                {%- endif %}&#10;                {%- if add_vision_id %}&#10;                    {{- 'Picture ' ~ image_count.value ~ ': ' }}&#10;                {%- endif %}&#10;                {{- '&lt;|vision_start|>&lt;|image_pad|>&lt;|vision_end|>' }}&#10;            {%- elif 'video' in item or item.type == 'video' %}&#10;                {%- if is_system_content %}&#10;                    {{- raise_exception('System message cannot contain videos.') }}&#10;                {%- endif %}&#10;                {%- if do_vision_count %}&#10;                    {%- set video_count.value = video_count.value + 1 %}&#10;                {%- endif %}&#10;                {%- if add_vision_id %}&#10;                    {{- 'Video ' ~ video_count.value ~ ': ' }}&#10;                {%- endif %}&#10;                {{- '&lt;|vision_start|>&lt;|video_pad|>&lt;|vision_end|>' }}&#10;            {%- elif 'text' in item %}&#10;                {{- item.text }}&#10;            {%- else %}&#10;                {{- raise_exception('Unexpected item type in content.') }}&#10;            {%- endif %}&#10;        {%- endfor %}&#10;    {%- elif content is none or content is undefined %}&#10;        {{- '' }}&#10;    {%- else %}&#10;        {{- raise_exception('Unexpected content type.') }}&#10;    {%- endif %}&#10;{%- endmacro %}&#10;{%- if not messages %}&#10;    {{- raise_exception('No messages provided.') }}&#10;{%- endif %}&#10;{%- if tools and tools is iterable and tools is not mapping %}&#10;    {{- '&lt;|im_start|>system\n' }}&#10;    {{- &quot;# Tools\n\nYou have access to the following functions:\n\n&lt;tools>&quot; }}&#10;    {%- for tool in tools %}&#10;        {{- &quot;\n&quot; }}&#10;        {{- tool | tojson }}&#10;    {%- endfor %}&#10;    {{- &quot;\n&lt;/tools>&quot; }}&#10;    {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n&lt;tool_call>\n&lt;function=example_function_name>\n&lt;parameter=example_parameter_1>\nvalue_1\n&lt;/parameter>\n&lt;parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n&lt;/parameter>\n&lt;/function>\n&lt;/tool_call>\n\n&lt;IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner &lt;function=...>&lt;/function> block must be nested within &lt;tool_call>&lt;/tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n&lt;/IMPORTANT>' }}&#10;    {%- if messages[0].role == 'system' %}&#10;        {%- set content = render_content(messages[0].content, false, true)|trim %}&#10;        {%- if content %}&#10;            {{- '\n\n' + content }}&#10;        {%- endif %}&#10;    {%- endif %}&#10;    {{- '&lt;|im_end|>\n' }}&#10;{%- else %}&#10;    {%- if messages[0].role == 'system' %}&#10;        {%- set content = render_content(messages[0].content, false, true)|trim %}&#10;        {{- '&lt;|im_start|>system\n' + content + '&lt;|im_end|>\n' }}&#10;    {%- endif %}&#10;{%- endif %}&#10;{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}&#10;{%- for message in messages[::-1] %}&#10;    {%- set index = (messages|length - 1) - loop.index0 %}&#10;    {%- if ns.multi_step_tool and message.role == &quot;user&quot; %}&#10;        {%- set content = render_content(message.content, false)|trim %}&#10;        {%- if not(content.startswith('&lt;tool_response>') and content.endswith('&lt;/tool_response>')) %}&#10;            {%- set ns.multi_step_tool = false %}&#10;            {%- set ns.last_query_index = index %}&#10;        {%- endif %}&#10;    {%- endif %}&#10;{%- endfor %}&#10;{%- if ns.multi_step_tool %}&#10;    {{- raise_exception('No user query found in messages.') }}&#10;{%- endif %}&#10;{%- for message in messages %}&#10;    {%- set content = render_content(message.content, true)|trim %}&#10;    {%- if message.role == &quot;system&quot; %}&#10;        {%- if not loop.first %}&#10;            {{- raise_exception('System message must be at the beginning.') }}&#10;        {%- endif %}&#10;    {%- elif message.role == &quot;user&quot; %}&#10;        {{- '&lt;|im_start|>' + message.role + '\n' + content + '&lt;|im_end|>' + '\n' }}&#10;    {%- elif message.role == &quot;assistant&quot; %}&#10;        {%- set reasoning_content = '' %}&#10;        {%- if message.reasoning_content is string %}&#10;            {%- set reasoning_content = message.reasoning_content %}&#10;        {%- else %}&#10;            {%- if '&lt;/think>' in content %}&#10;                {%- set reasoning_content = content.split('&lt;/think>')[0].rstrip('\n').split('&lt;think>')[-1].lstrip('\n') %}&#10;                {%- set content = content.split('&lt;/think>')[-1].lstrip('\n') %}&#10;            {%- endif %}&#10;        {%- endif %}&#10;        {%- set reasoning_content = reasoning_content|trim %}&#10;        {%- if (preserve_thinking is defined and preserve_thinking is true) or (loop.index0 > ns.last_query_index) %}&#10;            {{- '&lt;|im_start|>' + message.role + '\n&lt;think>\n' + reasoning_content + '\n&lt;/think>\n\n' + content }}&#10;        {%- else %}&#10;            {{- '&lt;|im_start|>' + message.role + '\n' + content }}&#10;        {%- endif %}&#10;        {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}&#10;            {%- for tool_call in message.tool_calls %}&#10;                {%- if tool_call.function is defined %}&#10;                    {%- set tool_call = tool_call.function %}&#10;                {%- endif %}&#10;                {%- if loop.first %}&#10;                    {%- if content|trim %}&#10;                        {{- '\n\n&lt;tool_call>\n&lt;function=' + tool_call.name + '>\n' }}&#10;                    {%- else %}&#10;                        {{- '&lt;tool_call>\n&lt;function=' + tool_call.name + '>\n' }}&#10;                    {%- endif %}&#10;                {%- else %}&#10;                    {{- '\n&lt;tool_call>\n&lt;function=' + tool_call.name + '>\n' }}&#10;                {%- endif %}&#10;                {%- if tool_call.arguments is defined %}&#10;                    {%- for args_name, args_value in tool_call.arguments|items %}&#10;                        {{- '&lt;parameter=' + args_name + '>\n' }}&#10;                        {%- set args_value = args_value | string if args_value is string else args_value | tojson | safe %}&#10;                        {{- args_value }}&#10;                        {{- '\n&lt;/parameter>\n' }}&#10;                    {%- endfor %}&#10;                {%- endif %}&#10;                {{- '&lt;/function>\n&lt;/tool_call>' }}&#10;            {%- endfor %}&#10;        {%- endif %}&#10;        {{- '&lt;|im_end|>\n' }}&#10;    {%- elif message.role == &quot;tool&quot; %}&#10;        {%- if loop.previtem and loop.previtem.role != &quot;tool&quot; %}&#10;            {{- '&lt;|im_start|>user' }}&#10;        {%- endif %}&#10;        {{- '\n&lt;tool_response>\n' }}&#10;        {{- content }}&#10;        {{- '\n&lt;/tool_response>' }}&#10;        {%- if not loop.last and loop.nextitem.role != &quot;tool&quot; %}&#10;            {{- '&lt;|im_end|>\n' }}&#10;        {%- elif loop.last %}&#10;            {{- '&lt;|im_end|>\n' }}&#10;        {%- endif %}&#10;    {%- else %}&#10;        {{- raise_exception('Unexpected message role.') }}&#10;    {%- endif %}&#10;{%- endfor %}&#10;{%- if add_generation_prompt %}&#10;    {{- '&lt;|im_start|>assistant\n' }}&#10;    {%- if enable_thinking is defined and enable_thinking is false %}&#10;        {{- '&lt;think>\n\n&lt;/think>\n\n' }}&#10;    {%- else %}&#10;        {{- '&lt;think>\n' }}&#10;    {%- endif %}&#10;{%- endif %}" />
+		<clean_up_tokenization_spaces />
+		<detokenizer_input_type value="i64" />
+		<eos_token_id value="248046" />
+		<handle_special_tokens_with_re />
+		<max_length />
+		<number_of_inputs value="1" />
+		<openvino_tokenizers_version value="2026.2.0.0-677-b5eb36f08b1" />
+		<openvino_version value="2026.2.0-21860-7be1ffd093f" />
+		<original_post_processor_template value="{&quot;type&quot;: &quot;ByteLevel&quot;, &quot;add_prefix_space&quot;: false, &quot;trim_offsets&quot;: false, &quot;use_regex&quot;: false}" />
+		<original_tokenizer_class value="&lt;class 'transformers.tokenization_utils_tokenizers.TokenizersBackend'>" />
+		<pad_token_id value="248044" />
+		<processed_post_processor_template value="{&quot;single&quot;: {&quot;ids&quot;: [-1], &quot;type_ids&quot;: [0]}, &quot;pair&quot;: {&quot;ids&quot;: [-1, -2], &quot;type_ids&quot;: [0, 0]}}" />
+		<sentencepiece_version value="0.2.1" />
+		<skip_special_tokens value="True" />
+		<streaming_detokenizer value="False" />
+		<tokenizer_output_type value="i64" />
+		<tokenizers_version value="0.22.2" />
+		<transformers_version value="5.2.0" />
+		<use_max_padding value="False" />
+		<use_sentencepiece_backend value="False" />
+		<utf8_replace_mode value="replace" />
+		<with_detokenizer value="True" />
+	</rt_info>
+</net>

openvino_vision_embeddings_merger_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a995b0d1c23a3247d580f74b491e82fc6fc023238d1248bf9fd1baa12d41b578
+size 443361892

openvino_vision_embeddings_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22d5ef784a6235cb837763e77638e8ace72b2d6001dc6347a7300c8310bd0b48
+size 1776440

openvino_vision_embeddings_model.xml

@@ -0,0 +1,291 @@
+<?xml version="1.0"?>
+<net name="Model63" version="11">
+	<layers>
+		<layer id="0" name="hidden_states" type="Parameter" version="opset1">
+			<data shape="?,?" element_type="f32" />
+			<output>
+				<port id="0" precision="FP32" names="hidden_states">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="1" name="Constant_409129" type="Const" version="opset1">
+			<data element_type="i64" shape="5" offset="0" size="40" />
+			<output>
+				<port id="0" precision="I64">
+					<dim>5</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="2" name="aten::view/Reshape" type="Reshape" version="opset1">
+			<data special_zero="false" />
+			<input>
+				<port id="0" precision="FP32">
+					<dim>-1</dim>
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I64">
+					<dim>5</dim>
+				</port>
+			</input>
+			<output>
+				<port id="2" precision="FP32" names="14,9,hidden_states_1,input">
+					<dim>-1</dim>
+					<dim>3</dim>
+					<dim>2</dim>
+					<dim>16</dim>
+					<dim>16</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="3" name="self.proj.weight" type="Const" version="opset1">
+			<data element_type="i8" shape="1152, 3, 2, 16, 16" offset="40" size="1769472" />
+			<output>
+				<port id="0" precision="I8">
+					<dim>1152</dim>
+					<dim>3</dim>
+					<dim>2</dim>
+					<dim>16</dim>
+					<dim>16</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="4" name="Convert_3165957" type="Convert" version="opset1">
+			<data destination_type="f16" />
+			<input>
+				<port id="0" precision="I8">
+					<dim>1152</dim>
+					<dim>3</dim>
+					<dim>2</dim>
+					<dim>16</dim>
+					<dim>16</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="FP16">
+					<dim>1152</dim>
+					<dim>3</dim>
+					<dim>2</dim>
+					<dim>16</dim>
+					<dim>16</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="5" name="self.proj.weight/scale" type="Const" version="opset1">
+			<data element_type="f16" shape="1152, 1, 1, 1, 1" offset="1769512" size="2304" />
+			<output>
+				<port id="0" precision="FP16">
+					<dim>1152</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="6" name="self.proj.weight/fq_weights_1" type="Multiply" version="opset1">
+			<data auto_broadcast="numpy" />
+			<input>
+				<port id="0" precision="FP16">
+					<dim>1152</dim>
+					<dim>3</dim>
+					<dim>2</dim>
+					<dim>16</dim>
+					<dim>16</dim>
+				</port>
+				<port id="1" precision="FP16">
+					<dim>1152</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="2" precision="FP16">
+					<dim>1152</dim>
+					<dim>3</dim>
+					<dim>2</dim>
+					<dim>16</dim>
+					<dim>16</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="7" name="self.proj.weight/fq_weights_1/convert" type="Convert" version="opset1">
+			<data destination_type="f32" />
+			<input>
+				<port id="0" precision="FP16">
+					<dim>1152</dim>
+					<dim>3</dim>
+					<dim>2</dim>
+					<dim>16</dim>
+					<dim>16</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="FP32">
+					<dim>1152</dim>
+					<dim>3</dim>
+					<dim>2</dim>
+					<dim>16</dim>
+					<dim>16</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="8" name="__module.proj/aten::_convolution/Convolution" type="Convolution" version="opset1">
+			<data strides="2, 16, 16" dilations="1, 1, 1" pads_begin="0, 0, 0" pads_end="0, 0, 0" auto_pad="explicit" />
+			<input>
+				<port id="0" precision="FP32">
+					<dim>-1</dim>
+					<dim>3</dim>
+					<dim>2</dim>
+					<dim>16</dim>
+					<dim>16</dim>
+				</port>
+				<port id="1" precision="FP32">
+					<dim>1152</dim>
+					<dim>3</dim>
+					<dim>2</dim>
+					<dim>16</dim>
+					<dim>16</dim>
+				</port>
+			</input>
+			<output>
+				<port id="2" precision="FP32">
+					<dim>-1</dim>
+					<dim>1152</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="9" name="__module.proj/aten::_convolution/Reshape" type="Const" version="opset1">
+			<data element_type="f32" shape="1, 1152, 1, 1, 1" offset="1771816" size="4608" />
+			<output>
+				<port id="0" precision="FP32">
+					<dim>1</dim>
+					<dim>1152</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="10" name="__module.proj/aten::_convolution/Add" type="Add" version="opset1">
+			<data auto_broadcast="numpy" />
+			<input>
+				<port id="0" precision="FP32">
+					<dim>-1</dim>
+					<dim>1152</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+				</port>
+				<port id="1" precision="FP32">
+					<dim>1</dim>
+					<dim>1152</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="2" precision="FP32" names="33">
+					<dim>-1</dim>
+					<dim>1152</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="11" name="Constant_409202" type="Const" version="opset1">
+			<data element_type="i64" shape="2" offset="1776424" size="16" />
+			<output>
+				<port id="0" precision="I64">
+					<dim>2</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="12" name="aten::view/Reshape_1" type="Reshape" version="opset1">
+			<data special_zero="false" />
+			<input>
+				<port id="0" precision="FP32">
+					<dim>-1</dim>
+					<dim>1152</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+					<dim>1</dim>
+				</port>
+				<port id="1" precision="I64">
+					<dim>2</dim>
+				</port>
+			</input>
+			<output>
+				<port id="2" precision="FP32" names="last_hidden_state">
+					<dim>-1</dim>
+					<dim>1152</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="13" name="Result_409230" type="Result" version="opset1" output_names="last_hidden_state">
+			<input>
+				<port id="0" precision="FP32">
+					<dim>-1</dim>
+					<dim>1152</dim>
+				</port>
+			</input>
+		</layer>
+	</layers>
+	<edges>
+		<edge from-layer="0" from-port="0" to-layer="2" to-port="0" />
+		<edge from-layer="1" from-port="0" to-layer="2" to-port="1" />
+		<edge from-layer="2" from-port="2" to-layer="8" to-port="0" />
+		<edge from-layer="3" from-port="0" to-layer="4" to-port="0" />
+		<edge from-layer="4" from-port="1" to-layer="6" to-port="0" />
+		<edge from-layer="5" from-port="0" to-layer="6" to-port="1" />
+		<edge from-layer="6" from-port="2" to-layer="7" to-port="0" />
+		<edge from-layer="7" from-port="1" to-layer="8" to-port="1" />
+		<edge from-layer="8" from-port="2" to-layer="10" to-port="0" />
+		<edge from-layer="9" from-port="0" to-layer="10" to-port="1" />
+		<edge from-layer="10" from-port="2" to-layer="12" to-port="0" />
+		<edge from-layer="11" from-port="0" to-layer="12" to-port="1" />
+		<edge from-layer="12" from-port="2" to-layer="13" to-port="0" />
+	</edges>
+	<rt_info>
+		<info name="OpenVINO Runtime" value="2026.2.0-21860-7be1ffd093f" />
+		<Runtime_version value="2026.2.0-21860-7be1ffd093f" />
+		<conversion_parameters>
+			<framework value="pytorch" />
+			<is_python_object value="True" />
+		</conversion_parameters>
+		<nncf>
+			<friendly_names_were_updated value="True" />
+			<version value="3.1.0" />
+			<weight_compression>
+				<advanced_parameters value="{'statistics_path': None, 'lora_adapter_rank': 256, 'group_size_fallback_mode': 'error', 'min_adjusted_group_size': 32, 'awq_params': {'subset_size': 32, 'percent_to_apply': 0.002, 'alpha_min': 0.0, 'alpha_max': 1.0, 'steps': 100, 'prefer_data_aware_scaling': True}, 'scale_estimation_params': {'subset_size': 64, 'initial_steps': 5, 'scale_steps': 5, 'weight_penalty': -1.0}, 'gptq_params': {'damp_percent': 0.1, 'block_size': 128, 'subset_size': 128}, 'lora_correction_params': {'adapter_rank': 8, 'num_iterations': 3, 'apply_regularization': True, 'subset_size': 128, 'use_int8_adapters': True}, 'backend_params': {}, 'codebook': None, 'adaptive_codebook_params': {'value_type': 'f8e4m3', 'across_blocks': False, 'num_elements': 16}}" />
+				<all_layers value="False" />
+				<awq value="False" />
+				<backup_mode value="int8_asym" />
+				<compression_format value="dequantize" />
+				<gptq value="False" />
+				<group_size value="-1" />
+				<ignored_scope value="[]" />
+				<lora_correction value="False" />
+				<mode value="int8_sym" />
+				<ratio value="1.0" />
+				<scale_estimation value="False" />
+				<sensitivity_metric value="weight_quantization_error" />
+			</weight_compression>
+		</nncf>
+		<optimum>
+			<nncf_version value="3.1.0" />
+			<optimum_intel_version value="1.27.0.dev0+ff1fb6d" />
+			<optimum_version value="2.1.0.dev0" />
+			<pytorch_version value="2.11.0" />
+			<transformers_version value="5.2.0" />
+		</optimum>
+	</rt_info>
+</net>

openvino_vision_embeddings_pos_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:261db08f77e8fe70b1624f2472433e90cfa04fd7e579abfca7febdfa8e236799
+size 2658820

openvino_vision_embeddings_pos_model.xml

@@ -0,0 +1,180 @@
+<?xml version="1.0"?>
+<net name="Model72" version="11">
+	<layers>
+		<layer id="0" name="input" type="Parameter" version="opset1">
+			<data shape="4,?" element_type="i64" />
+			<output>
+				<port id="0" precision="I64" names="input">
+					<dim>4</dim>
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="1" name="self.weight" type="Const" version="opset1">
+			<data element_type="i8" shape="2304, 1152" offset="0" size="2654208" />
+			<output>
+				<port id="0" precision="I8">
+					<dim>2304</dim>
+					<dim>1152</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="2" name="Convert_3741987" type="Convert" version="opset1">
+			<data destination_type="f16" />
+			<input>
+				<port id="0" precision="I8">
+					<dim>2304</dim>
+					<dim>1152</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="FP16">
+					<dim>2304</dim>
+					<dim>1152</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="3" name="self.weight/scale" type="Const" version="opset1">
+			<data element_type="f16" shape="2304, 1" offset="2654208" size="4608" />
+			<output>
+				<port id="0" precision="FP16">
+					<dim>2304</dim>
+					<dim>1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="4" name="self.weight/fq_weights_0" type="Multiply" version="opset1">
+			<data auto_broadcast="numpy" />
+			<input>
+				<port id="0" precision="FP16">
+					<dim>2304</dim>
+					<dim>1152</dim>
+				</port>
+				<port id="1" precision="FP16">
+					<dim>2304</dim>
+					<dim>1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="2" precision="FP16">
+					<dim>2304</dim>
+					<dim>1152</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="5" name="ov_ext::embedding/Convert" type="Convert" version="opset1">
+			<data destination_type="f32" />
+			<rt_info>
+				<attribute name="decompression" version="0" />
+			</rt_info>
+			<input>
+				<port id="0" precision="FP16">
+					<dim>2304</dim>
+					<dim>1152</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="FP32">
+					<dim>2304</dim>
+					<dim>1152</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="6" name="ov_ext::embedding/Convert_1" type="Convert" version="opset1">
+			<data destination_type="i32" />
+			<input>
+				<port id="0" precision="I64">
+					<dim>4</dim>
+					<dim>-1</dim>
+				</port>
+			</input>
+			<output>
+				<port id="1" precision="I32">
+					<dim>4</dim>
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="7" name="ov_ext::embedding/Constant" type="Const" version="opset1">
+			<data element_type="i32" shape="" offset="2658816" size="4" />
+			<output>
+				<port id="0" precision="I32" />
+			</output>
+		</layer>
+		<layer id="8" name="ov_ext::embedding/Gather" type="Gather" version="opset8">
+			<data batch_dims="0" />
+			<input>
+				<port id="0" precision="FP32">
+					<dim>2304</dim>
+					<dim>1152</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>4</dim>
+					<dim>-1</dim>
+				</port>
+				<port id="2" precision="I32" />
+			</input>
+			<output>
+				<port id="3" precision="FP32" names="last_hidden_state">
+					<dim>4</dim>
+					<dim>-1</dim>
+					<dim>1152</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="9" name="Result_487321" type="Result" version="opset1" output_names="last_hidden_state">
+			<input>
+				<port id="0" precision="FP32">
+					<dim>4</dim>
+					<dim>-1</dim>
+					<dim>1152</dim>
+				</port>
+			</input>
+		</layer>
+	</layers>
+	<edges>
+		<edge from-layer="0" from-port="0" to-layer="6" to-port="0" />
+		<edge from-layer="1" from-port="0" to-layer="2" to-port="0" />
+		<edge from-layer="2" from-port="1" to-layer="4" to-port="0" />
+		<edge from-layer="3" from-port="0" to-layer="4" to-port="1" />
+		<edge from-layer="4" from-port="2" to-layer="5" to-port="0" />
+		<edge from-layer="5" from-port="1" to-layer="8" to-port="0" />
+		<edge from-layer="6" from-port="1" to-layer="8" to-port="1" />
+		<edge from-layer="7" from-port="0" to-layer="8" to-port="2" />
+		<edge from-layer="8" from-port="3" to-layer="9" to-port="0" />
+	</edges>
+	<rt_info>
+		<info name="OpenVINO Runtime" value="2026.2.0-21860-7be1ffd093f" />
+		<Runtime_version value="2026.2.0-21860-7be1ffd093f" />
+		<conversion_parameters>
+			<framework value="pytorch" />
+			<is_python_object value="True" />
+		</conversion_parameters>
+		<nncf>
+			<friendly_names_were_updated value="True" />
+			<version value="3.1.0" />
+			<weight_compression>
+				<advanced_parameters value="{'statistics_path': None, 'lora_adapter_rank': 256, 'group_size_fallback_mode': 'error', 'min_adjusted_group_size': 32, 'awq_params': {'subset_size': 32, 'percent_to_apply': 0.002, 'alpha_min': 0.0, 'alpha_max': 1.0, 'steps': 100, 'prefer_data_aware_scaling': True}, 'scale_estimation_params': {'subset_size': 64, 'initial_steps': 5, 'scale_steps': 5, 'weight_penalty': -1.0}, 'gptq_params': {'damp_percent': 0.1, 'block_size': 128, 'subset_size': 128}, 'lora_correction_params': {'adapter_rank': 8, 'num_iterations': 3, 'apply_regularization': True, 'subset_size': 128, 'use_int8_adapters': True}, 'backend_params': {}, 'codebook': None, 'adaptive_codebook_params': {'value_type': 'f8e4m3', 'across_blocks': False, 'num_elements': 16}}" />
+				<all_layers value="False" />
+				<awq value="False" />
+				<backup_mode value="int8_asym" />
+				<compression_format value="dequantize" />
+				<gptq value="False" />
+				<group_size value="-1" />
+				<ignored_scope value="[]" />
+				<lora_correction value="False" />
+				<mode value="int8_sym" />
+				<ratio value="1.0" />
+				<scale_estimation value="False" />
+				<sensitivity_metric value="weight_quantization_error" />
+			</weight_compression>
+		</nncf>
+		<optimum>
+			<nncf_version value="3.1.0" />
+			<optimum_intel_version value="1.27.0.dev0+ff1fb6d" />
+			<optimum_version value="2.1.0.dev0" />
+			<pytorch_version value="2.11.0" />
+			<transformers_version value="5.2.0" />
+		</optimum>
+	</rt_info>
+</net>

preprocessor_config.json

@@ -0,0 +1,27 @@
+{
+  "data_format": "channels_first",
+  "do_convert_rgb": true,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "Qwen2VLImageProcessorFast",
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "merge_size": 2,
+  "patch_size": 16,
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "longest_edge": 16777216,
+    "shortest_edge": 65536
+  },
+  "temporal_patch_size": 2
+}

processor_config.json

@@ -0,0 +1,63 @@
+{
+  "image_processor": {
+    "data_format": "channels_first",
+    "do_convert_rgb": true,
+    "do_normalize": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "image_mean": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "image_processor_type": "Qwen2VLImageProcessorFast",
+    "image_std": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "merge_size": 2,
+    "patch_size": 16,
+    "resample": 3,
+    "rescale_factor": 0.00392156862745098,
+    "size": {
+      "longest_edge": 16777216,
+      "shortest_edge": 65536
+    },
+    "temporal_patch_size": 2
+  },
+  "processor_class": "Qwen3VLProcessor",
+  "video_processor": {
+    "data_format": "channels_first",
+    "default_to_square": true,
+    "do_convert_rgb": true,
+    "do_normalize": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "do_sample_frames": true,
+    "fps": 2,
+    "image_mean": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "image_std": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "max_frames": 768,
+    "merge_size": 2,
+    "min_frames": 4,
+    "patch_size": 16,
+    "resample": 3,
+    "rescale_factor": 0.00392156862745098,
+    "return_metadata": false,
+    "size": {
+      "longest_edge": 25165824,
+      "shortest_edge": 4096
+    },
+    "temporal_patch_size": 2,
+    "video_processor_type": "Qwen3VLVideoProcessor"
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87a7830d63fcf43bf241c3c5242e96e62dd3fdc29224ca26fed8ea333db72de4
+size 19989343

tokenizer_config.json

@@ -0,0 +1,32 @@
+{
+  "add_prefix_space": false,
+  "audio_bos_token": "<|audio_start|>",
+  "audio_eos_token": "<|audio_end|>",
+  "audio_token": "<|audio_pad|>",
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "image_token": "<|image_pad|>",
+  "is_local": false,
+  "model_max_length": 262144,
+  "model_specific_special_tokens": {
+    "audio_bos_token": "<|audio_start|>",
+    "audio_eos_token": "<|audio_end|>",
+    "audio_token": "<|audio_pad|>",
+    "image_token": "<|image_pad|>",
+    "video_token": "<|video_pad|>",
+    "vision_bos_token": "<|vision_start|>",
+    "vision_eos_token": "<|vision_end|>"
+  },
+  "pad_token": "<|endoftext|>",
+  "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
+  "processor_class": "Qwen3VLProcessor",
+  "split_special_tokens": false,
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": null,
+  "video_token": "<|video_pad|>",
+  "vision_bos_token": "<|vision_start|>",
+  "vision_eos_token": "<|vision_end|>"
+}