Instructions to use OpenVINO/phi-4-int4-ov with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use OpenVINO/phi-4-int4-ov with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="OpenVINO/phi-4-int4-ov")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("OpenVINO/phi-4-int4-ov")
model = AutoModelForCausalLM.from_pretrained("OpenVINO/phi-4-int4-ov", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use OpenVINO/phi-4-int4-ov with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "OpenVINO/phi-4-int4-ov"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "OpenVINO/phi-4-int4-ov",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/OpenVINO/phi-4-int4-ov

SGLang

How to use OpenVINO/phi-4-int4-ov with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "OpenVINO/phi-4-int4-ov" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "OpenVINO/phi-4-int4-ov",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "OpenVINO/phi-4-int4-ov" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "OpenVINO/phi-4-int4-ov",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use OpenVINO/phi-4-int4-ov with Docker Model Runner:
```
docker model run hf.co/OpenVINO/phi-4-int4-ov
```

phi-4-int4-ov / openvino_tokenizer.xml

openvino-ci

Upload folder using huggingface_hub

129eafd verified over 1 year ago

Raw

History Blame Contribute Delete

21.5 kB

	<?xml version="1.0"?>
	<net name="tokenizer" version="11">
	<layers>
	<layer id="0" name="Parameter_210429" type="Parameter" version="opset1">
	<data shape="?" element_type="string" />
	<output>
	<port id="0" precision="STRING" names="Parameter_210429">
	<dim>-1</dim>
	</port>
	</output>
	</layer>
	<layer id="1" name="Constant_210435" 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_210430" 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_210431" 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_210432" 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_210433" 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_210434" 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_210436" 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_210437" 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_210438" 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_210439" 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_210440" 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_210441" 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_210442" 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_210504" type="Const" version="opset1">
	<data element_type="u8" shape="3163" offset="16" size="3163" />
	<output>
	<port id="0" precision="U8">
	<dim>3163</dim>
	</port>
	</output>
	</layer>
	<layer id="15" name="SpecialTokensSplit_210505" 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>3163</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="Constant_210507" type="Const" version="opset1">
	<data element_type="u8" shape="115" offset="3179" size="115" />
	<output>
	<port id="0" precision="U8">
	<dim>115</dim>
	</port>
	</output>
	</layer>
	<layer id="17" name="RegexSplit_210508" type="RegexSplit" version="extension">
	<data behaviour="remove" invert="true" 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>115</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="18" name="Constant_210510" type="Const" version="opset1">
	<data element_type="i32" shape="100352" offset="3294" size="401408" />
	<output>
	<port id="0" precision="I32">
	<dim>100352</dim>
	</port>
	</output>
	</layer>
	<layer id="19" name="Constant_210512" type="Const" version="opset1">
	<data element_type="i32" shape="100352" offset="404702" size="401408" />
	<output>
	<port id="0" precision="I32">
	<dim>100352</dim>
	</port>
	</output>
	</layer>
	<layer id="20" name="Constant_210514" type="Const" version="opset1">
	<data element_type="u8" shape="645170" offset="806110" size="645170" />
	<output>
	<port id="0" precision="U8">
	<dim>645170</dim>
	</port>
	</output>
	</layer>
	<layer id="21" name="Constant_210522" type="Const" version="opset1">
	<data element_type="i32" shape="100000" offset="1451280" size="400000" />
	<output>
	<port id="0" precision="I32">
	<dim>100000</dim>
	</port>
	</output>
	</layer>
	<layer id="22" name="Constant_210524" type="Const" version="opset1">
	<data element_type="i32" shape="100000" offset="1851280" size="400000" />
	<output>
	<port id="0" precision="I32">
	<dim>100000</dim>
	</port>
	</output>
	</layer>
	<layer id="23" name="Constant_210526" type="Const" version="opset1">
	<data element_type="u8" shape="309844" offset="2251280" size="309844" />
	<output>
	<port id="0" precision="U8">
	<dim>309844</dim>
	</port>
	</output>
	</layer>
	<layer id="24" name="Constant_210528" type="Const" version="opset1">
	<data element_type="i32" shape="100000" offset="2561124" size="400000" />
	<output>
	<port id="0" precision="I32">
	<dim>100000</dim>
	</port>
	</output>
	</layer>
	<layer id="25" name="Constant_210530" type="Const" version="opset1">
	<data element_type="i32" shape="100000" offset="2961124" size="400000" />
	<output>
	<port id="0" precision="I32">
	<dim>100000</dim>
	</port>
	</output>
	</layer>
	<layer id="26" name="Constant_210532" type="Const" version="opset1">
	<data element_type="u8" shape="333730" offset="3361124" size="333730" />
	<output>
	<port id="0" precision="U8">
	<dim>333730</dim>
	</port>
	</output>
	</layer>
	<layer id="27" name="Constant_210516" type="Const" version="opset1">
	<data element_type="i32" shape="4128" offset="3694854" size="16512" />
	<output>
	<port id="0" precision="I32">
	<dim>4128</dim>
	</port>
	</output>
	</layer>
	<layer id="28" name="Constant_210518" type="Const" version="opset1">
	<data element_type="i32" shape="4128" offset="3711366" size="16512" />
	<output>
	<port id="0" precision="I32">
	<dim>4128</dim>
	</port>
	</output>
	</layer>
	<layer id="29" name="Constant_210520" type="Const" version="opset1">
	<data element_type="u8" shape="56852" offset="3727878" size="56852" />
	<output>
	<port id="0" precision="U8">
	<dim>56852</dim>
	</port>
	</output>
	</layer>
	<layer id="30" name="Constant_210533" type="Const" version="opset1">
	<data element_type="i32" shape="4128" offset="3784730" size="16512" />
	<output>
	<port id="0" precision="I32">
	<dim>4128</dim>
	</port>
	</output>
	</layer>
	<layer id="31" name="BPETokenizer_210534" type="BPETokenizer" version="extension">
	<data unk_token="" fuse_unk="false" suffix_indicator="" end_suffix="" byte_fallback="false" cache_capacity="20070" />
	<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>100352</dim>
	</port>
	<port id="6" precision="I32">
	<dim>100352</dim>
	</port>
	<port id="7" precision="U8">
	<dim>645170</dim>
	</port>
	<port id="8" precision="I32">
	<dim>100000</dim>
	</port>
	<port id="9" precision="I32">
	<dim>100000</dim>
	</port>
	<port id="10" precision="U8">
	<dim>309844</dim>
	</port>
	<port id="11" precision="I32">
	<dim>100000</dim>
	</port>
	<port id="12" precision="I32">
	<dim>100000</dim>
	</port>
	<port id="13" precision="U8">
	<dim>333730</dim>
	</port>
	<port id="14" precision="I32">
	<dim>4128</dim>
	</port>
	<port id="15" precision="I32">
	<dim>4128</dim>
	</port>
	<port id="16" precision="U8">
	<dim>56852</dim>
	</port>
	<port id="17" precision="I32">
	<dim>4128</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="32" name="Subtract_210535" 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="33" name="Constant_210536" type="Const" version="opset1">
	<data element_type="i32" shape="" offset="3801242" size="4" />
	<output>
	<port id="0" precision="I32" />
	</output>
	</layer>
	<layer id="34" name="Minimum_210537" type="Minimum" version="opset1">
	<data auto_broadcast="numpy" />
	<input>
	<port id="0" precision="I32">
	<dim>-1</dim>
	</port>
	<port id="1" precision="I32" />
	</input>
	<output>
	<port id="2" precision="I32">
	<dim>-1</dim>
	</port>
	</output>
	</layer>
	<layer id="35" name="Subtract_210538" 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="36" name="Subtract_210539" 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="37" name="Constant_210540" type="Const" version="opset1">
	<data element_type="i32" shape="" offset="3801246" size="4" />
	<output>
	<port id="0" precision="I32" />
	</output>
	</layer>
	<layer id="38" name="ReduceMax_210541" 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="39" name="Constant_210542" type="Const" version="opset1">
	<data element_type="i32" shape="" offset="3801250" size="4" />
	<output>
	<port id="0" precision="I32" />
	</output>
	</layer>
	<layer id="40" name="RaggedToDense_210543" 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="41" name="Convert_210544" 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="42" name="Convert_210544.0" 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="44" name="RaggedToDense_210543.0" 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="45" name="Result_210545" 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="43" name="Result_210546" 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="2" to-layer="15" to-port="3" />
	<edge from-layer="2" from-port="3" to-layer="15" to-port="4" />
	<edge from-layer="2" from-port="1" to-layer="15" to-port="2" />
	<edge from-layer="2" from-port="1" to-layer="3" to-port="0" />
	<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="8" to-port="1" />
	<edge from-layer="6" from-port="3" to-layer="11" to-port="0" />
	<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="6" to-layer="17" to-port="0" />
	<edge from-layer="15" from-port="7" to-layer="17" to-port="1" />
	<edge from-layer="15" from-port="8" to-layer="17" to-port="2" />
	<edge from-layer="15" from-port="9" to-layer="17" to-port="3" />
	<edge from-layer="15" from-port="10" to-layer="17" to-port="4" />
	<edge from-layer="15" from-port="11" to-layer="17" to-port="5" />
	<edge from-layer="16" from-port="0" to-layer="17" to-port="6" />
	<edge from-layer="17" from-port="7" to-layer="31" to-port="0" />
	<edge from-layer="17" from-port="8" to-layer="31" to-port="1" />
	<edge from-layer="17" from-port="9" to-layer="31" to-port="2" />
	<edge from-layer="17" from-port="10" to-layer="31" to-port="3" />
	<edge from-layer="17" from-port="11" to-layer="31" to-port="4" />
	<edge from-layer="18" from-port="0" to-layer="31" to-port="5" />
	<edge from-layer="19" from-port="0" to-layer="31" to-port="6" />
	<edge from-layer="20" from-port="0" to-layer="31" to-port="7" />
	<edge from-layer="21" from-port="0" to-layer="31" to-port="8" />
	<edge from-layer="22" from-port="0" to-layer="31" to-port="9" />
	<edge from-layer="23" from-port="0" to-layer="31" to-port="10" />
	<edge from-layer="24" from-port="0" to-layer="31" to-port="11" />
	<edge from-layer="25" from-port="0" to-layer="31" to-port="12" />
	<edge from-layer="26" from-port="0" to-layer="31" to-port="13" />
	<edge from-layer="27" from-port="0" to-layer="31" to-port="14" />
	<edge from-layer="28" from-port="0" to-layer="31" to-port="15" />
	<edge from-layer="29" from-port="0" to-layer="31" to-port="16" />
	<edge from-layer="30" from-port="0" to-layer="31" to-port="17" />
	<edge from-layer="31" from-port="19" to-layer="35" to-port="0" />
	<edge from-layer="31" from-port="20" to-layer="40" to-port="2" />
	<edge from-layer="31" from-port="19" to-layer="40" to-port="1" />
	<edge from-layer="31" from-port="19" to-layer="36" to-port="0" />
	<edge from-layer="31" from-port="18" to-layer="32" to-port="1" />
	<edge from-layer="31" from-port="19" to-layer="32" to-port="0" />
	<edge from-layer="32" from-port="2" to-layer="34" to-port="0" />
	<edge from-layer="33" from-port="0" to-layer="34" to-port="1" />
	<edge from-layer="34" from-port="2" to-layer="35" to-port="1" />
	<edge from-layer="35" from-port="2" to-layer="36" to-port="1" />
	<edge from-layer="35" from-port="2" to-layer="40" to-port="0" />
	<edge from-layer="36" from-port="2" to-layer="38" to-port="0" />
	<edge from-layer="37" from-port="0" to-layer="38" to-port="1" />
	<edge from-layer="38" from-port="2" to-layer="40" to-port="3" />
	<edge from-layer="39" from-port="0" to-layer="40" to-port="4" />
	<edge from-layer="40" from-port="6" to-layer="41" to-port="0" />
	<edge from-layer="40" from-port="5" to-layer="44" to-port="0" />
	<edge from-layer="41" from-port="1" to-layer="42" to-port="0" />
	<edge from-layer="42" from-port="1" to-layer="43" to-port="0" />
	<edge from-layer="44" from-port="1" to-layer="45" to-port="0" />
	</edges>
	<rt_info>
	<add_attention_mask value="True" />
	<add_prefix_space />
	<add_special_tokens value="True" />
	<bos_token_id value="100257" />
	<chat_template value="{% for message in messages %}{% if (message['role'] == 'system') %}{{'<\|im_start\|>system<\|im_sep\|>' + message['content'] + '<\|im_end\|>'}}{% elif (message['role'] == 'user') %}{{'<\|im_start\|>user<\|im_sep\|>' + message['content'] + '<\|im_end\|>'}}{% elif (message['role'] == 'assistant') %}{{'<\|im_start\|>assistant<\|im_sep\|>' + message['content'] + '<\|im_end\|>'}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<\|im_start\|>assistant<\|im_sep\|>' }}{% endif %}" />
	<clean_up_tokenization_spaces />
	<detokenizer_input_type value="i64" />
	<eos_token_id value="100265" />
	<handle_special_tokens_with_re />
	<max_length />
	<number_of_inputs value="1" />
	<openvino_tokenizers_version value="2025.1.0.0-523-710ddf14de8" />
	<openvino_version value="2025.1.0-18503-6fec06580ab-releases/2025/1" />
	<original_tokenizer_class value="<class 'transformers.models.gpt2.tokenization_gpt2_fast.GPT2TokenizerFast'>" />
	<pad_token_id value="100349" />
	<sentencepiece_version value="0.2.0" />
	<skip_special_tokens value="True" />
	<streaming_detokenizer value="False" />
	<tiktoken_version value="0.9.0" />
	<tokenizer_output_type value="i64" />
	<tokenizers_version value="0.21.1" />
	<transformers_version value="4.51.3" />
	<use_max_padding value="False" />
	<use_sentencepiece_backend value="False" />
	<utf8_replace_mode value="replace" />
	<with_detokenizer value="True" />
	</rt_info>
	</net>