Upload folder using huggingface_hub

README.md

@@ -14,8 +14,8 @@ This is [Phi-3-mini-4k-instruct](https://huggingface.co/microsoft/Phi-3-mini-4k-
 Weight compression was performed using `nncf.compress_weights` with the following parameters:
 
 * mode: **int4_asym**
-* group_size: **128**
-* ratio: **0.8**
+* ratio: **1**
+* group_size: **64**
 
 For more information on quantization, check the [OpenVINO model optimization guide](https://docs.openvino.ai/2024/openvino-workflow/model-optimization-guide/weight-compression.html).
 
@@ -24,11 +24,10 @@ For more information on quantization, check the [OpenVINO model optimization gui
 
 The provided OpenVINO™ IR model is compatible with:
 
-* OpenVINO version 2024.2.0 and higher
-* Optimum Intel 1.18.0 and higher
-
-## Running Model Inference with [Optimum Intel](https://huggingface.co/docs/optimum/intel/index)
+* OpenVINO version 2024.4.0 and higher
+* Optimum Intel 1.23.1 and higher
 
+## Running Model Inference
 
 1. Install packages required for using [Optimum Intel](https://huggingface.co/docs/optimum/intel/index) integration with the OpenVINO backend:
 
@@ -55,40 +54,9 @@ print(text)
 
 For more examples and possible optimizations, refer to the [OpenVINO Large Language Model Inference Guide](https://docs.openvino.ai/2024/learn-openvino/llm_inference_guide.html).
 
-## Running Model Inference with [OpenVINO GenAI](https://github.com/openvinotoolkit/openvino.genai)
-
-1. Install packages required for using OpenVINO GenAI.
-```
-pip install openvino-genai huggingface_hub
-```
-
-2. Download model from HuggingFace Hub
-   
-```
-import huggingface_hub as hf_hub
-
-model_id = "OpenVINO/Phi-3-mini-4k-instruct-int4-ov"
-model_path = "Phi-3-mini-4k-instruct-int4-ov"
-
-hf_hub.snapshot_download(model_id, local_dir=model_path)
-
-```
-
-3. Run model inference:
-
-```
-import openvino_genai as ov_genai
-
-device = "CPU"
-pipe = ov_genai.LLMPipeline(model_path, device)
-print(pipe.generate("What is OpenVINO?", max_length=200))
-```
-
-More GenAI usage examples can be found in OpenVINO GenAI library [docs](https://github.com/openvinotoolkit/openvino.genai/blob/master/src/README.md) and [samples](https://github.com/openvinotoolkit/openvino.genai?tab=readme-ov-file#openvino-genai-samples)
-
 ## Limitations
 
-Check the original model card for [limitations](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct#responsible-ai-considerations).
+Check the original model card for [original model card](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct) for limitations.
 
 ## Legal information
 
@@ -96,4 +64,4 @@ The original model is distributed under [mit](https://choosealicense.com/license
 
 ## Disclaimer
 
-Intel is committed to respecting human rights and avoiding causing or contributing to adverse impacts on human rights. See [Intel’s Global Human Rights Principles](https://www.intel.com/content/dam/www/central-libraries/us/en/documents/policy-human-rights.pdf). Intel’s products and software are intended only to be used in applications that do not cause or contribute to adverse impacts on human rights.
+Intel is committed to respecting human rights and avoiding causing or contributing to adverse impacts on human rights. See [Intel’s Global Human Rights Principles](https://www.intel.com/content/dam/www/central-libraries/us/en/documents/policy-human-rights.pdf). Intel’s products and software are intended only to be used in applications that do not cause or contribute to adverse impacts on human rights.

config.json

@@ -3,6 +3,7 @@
   "architectures": [
     "Phi3ForCausalLM"
   ],
+  "attention_bias": false,
   "attention_dropout": 0.0,
   "auto_map": {
     "AutoConfig": "configuration_phi3.Phi3Config",
@@ -28,7 +29,8 @@
   "rope_theta": 10000.0,
   "sliding_window": 2047,
   "tie_word_embeddings": false,
-  "transformers_version": "4.41.2",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.45.2",
   "use_cache": true,
   "vocab_size": 32064
 }

configuration_phi3.py

@@ -1,213 +1,227 @@
-# coding=utf-8
-# Copyright 2024 Microsoft and the HuggingFace Inc. team. All rights reserved.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#     http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-
-""" Phi-3 model configuration"""
-
-
-from transformers.configuration_utils import PretrainedConfig
-from transformers.utils import logging
-
-
-logger = logging.get_logger(__name__)
-
-PHI3_PRETRAINED_CONFIG_ARCHIVE_MAP = {
-    "microsoft/Phi-3-mini-4k-instruct": "https://huggingface.co/microsoft/Phi-3-mini-4k-instruct/resolve/main/config.json",
-    "microsoft/Phi-3-mini-128k-instruct": "https://huggingface.co/microsoft/Phi-3-mini-128k-instruct/resolve/main/config.json",
-}
-
-
-class Phi3Config(PretrainedConfig):
-    r"""
-    This is the configuration class to store the configuration of a [`Phi3Model`]. It is used to instantiate a Phi-3
-    model according to the specified arguments, defining the model architecture. Instantiating a configuration with the
-    defaults will yield a similar configuration to that of the
-    [microsoft/Phi-3-mini-4k-instruct](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct).
-
-    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
-    documentation from [`PretrainedConfig`] for more information.
-
-    Args:
-        vocab_size (`int`, *optional*, defaults to 32064):
-            Vocabulary size of the Phi-3 model. Defines the number of different tokens that can be represented by the
-            `inputs_ids` passed when calling [`Phi3Model`].
-        hidden_size (`int`, *optional*, defaults to 3072):
-            Dimension of the hidden representations.
-        intermediate_size (`int`, *optional*, defaults to 8192):
-            Dimension of the MLP representations.
-        num_hidden_layers (`int`, *optional*, defaults to 32):
-            Number of hidden layers in the Transformer decoder.
-        num_attention_heads (`int`, *optional*, defaults to 32):
-            Number of attention heads for each attention layer in the Transformer decoder.
-        num_key_value_heads (`int`, *optional*):
-            This is the number of key_value heads that should be used to implement Grouped Query Attention. If
-            `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if
-            `num_key_value_heads=1 the model will use Multi Query Attention (MQA) otherwise GQA is used. When
-            converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed
-            by meanpooling all the original heads within that group. For more details checkout [this
-            paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to
-            `num_attention_heads`.
-        resid_pdrop (`float`, *optional*, defaults to 0.0):
-            Dropout probability for mlp outputs.
-        embd_pdrop (`int`, *optional*, defaults to 0.0):
-            The dropout ratio for the embeddings.
-        attention_dropout (`float`, *optional*, defaults to 0.0):
-            The dropout ratio after computing the attention scores.
-        hidden_act (`str` or `function`, *optional*, defaults to `"silu"`):
-            The non-linear activation function (function or string) in the decoder.
-        max_position_embeddings (`int`, *optional*, defaults to 4096):
-            The maximum sequence length that this model might ever be used with.
-        original_max_position_embeddings (`int`, *optional*, defaults to 4096):
-            The maximum sequence length that this model was trained with. This is used to determine the size of the
-            original RoPE embeddings when using long scaling.
-        initializer_range (`float`, *optional*, defaults to 0.02):
-            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
-        rms_norm_eps (`float`, *optional*, defaults to 1e-05):
-            The epsilon value used for the RMSNorm.
-        use_cache (`bool`, *optional*, defaults to `True`):
-            Whether or not the model should return the last key/values attentions (not used by all models). Only
-            relevant if `config.is_decoder=True`. Whether to tie weight embeddings or not.
-        tie_word_embeddings (`bool`, *optional*, defaults to `False`):
-            Whether to tie weight embeddings
-        rope_theta (`float`, *optional*, defaults to 10000.0):
-            The base period of the RoPE embeddings.
-        rope_scaling (`dict`, *optional*):
-            The scaling strategy for the RoPE embeddings. If `None`, no scaling is applied. If a dictionary, it must
-            contain the following keys: `type`, `short_factor` and `long_factor`. The `type` must be either `su` or `yarn` and
-            the `short_factor` and `long_factor` must be lists of numbers with the same length as the hidden size
-            divided by the number of attention heads divided by 2.
-        bos_token_id (`int`, *optional*, defaults to 1):
-            The id of the "beginning-of-sequence" token.
-        eos_token_id (`int`, *optional*, defaults to 32000):
-            The id of the "end-of-sequence" token.
-        pad_token_id (`int`, *optional*, defaults to 32000):
-            The id of the padding token.
-        sliding_window (`int`, *optional*):
-            Sliding window attention window size. If `None`, no sliding window is applied.
-
-    Example:
-
-    ```python
-    >>> from transformers import Phi3Model, Phi3Config
-
-    >>> # Initializing a Phi-3 style configuration
-    >>> configuration = Phi3Config.from_pretrained("microsoft/Phi-3-mini-4k-instruct")
-
-    >>> # Initializing a model from the configuration
-    >>> model = Phi3Model(configuration)
-
-    >>> # Accessing the model configuration
-    >>> configuration = model.config
-    ```"""
-
-    model_type = "phi3"
-    keys_to_ignore_at_inference = ["past_key_values"]
-
-    def __init__(
-        self,
-        vocab_size=32064,
-        hidden_size=3072,
-        intermediate_size=8192,
-        num_hidden_layers=32,
-        num_attention_heads=32,
-        num_key_value_heads=None,
-        resid_pdrop=0.0,
-        embd_pdrop=0.0,
-        attention_dropout=0.0,
-        hidden_act="silu",
-        max_position_embeddings=4096,
-        original_max_position_embeddings=4096,
-        initializer_range=0.02,
-        rms_norm_eps=1e-5,
-        use_cache=True,
-        tie_word_embeddings=False,
-        rope_theta=10000.0,
-        rope_scaling=None,
-        bos_token_id=1,
-        eos_token_id=32000,
-        pad_token_id=32000,
-        sliding_window=None,
-        **kwargs,
-    ):
-        self.vocab_size = vocab_size
-        self.hidden_size = hidden_size
-        self.intermediate_size = intermediate_size
-        self.num_hidden_layers = num_hidden_layers
-        self.num_attention_heads = num_attention_heads
-
-        if num_key_value_heads is None:
-            num_key_value_heads = num_attention_heads
-
-        self.num_key_value_heads = num_key_value_heads
-        self.resid_pdrop = resid_pdrop
-        self.embd_pdrop = embd_pdrop
-        self.attention_dropout = attention_dropout
-        self.hidden_act = hidden_act
-        self.max_position_embeddings = max_position_embeddings
-        self.original_max_position_embeddings = original_max_position_embeddings
-        self.initializer_range = initializer_range
-        self.rms_norm_eps = rms_norm_eps
-        self.use_cache = use_cache
-        self.rope_theta = rope_theta
-        self.rope_scaling = rope_scaling
-        self._rope_scaling_validation()
-        self.sliding_window = sliding_window
-
-        super().__init__(
-            bos_token_id=bos_token_id,
-            eos_token_id=eos_token_id,
-            pad_token_id=pad_token_id,
-            tie_word_embeddings=tie_word_embeddings,
-            **kwargs,
-        )
-
-    def _rope_scaling_validation(self):
-        """
-        Validate the `rope_scaling` configuration.
-        """
-        if self.rope_scaling is None:
-            return
-
-        if not isinstance(self.rope_scaling, dict) or len(self.rope_scaling) != 3:
-            raise ValueError(
-                "`rope_scaling` must be a dictionary with three fields, `type`, `short_factor` and `long_factor`, "
-                f"got {self.rope_scaling}"
-            )
-        rope_scaling_type = self.rope_scaling.get("type", None)
-        rope_scaling_short_factor = self.rope_scaling.get("short_factor", None)
-        rope_scaling_long_factor = self.rope_scaling.get("long_factor", None)
-        if rope_scaling_type is None or rope_scaling_type not in ["su", "yarn"]:
-            raise ValueError(f"`rope_scaling`'s type field must be one of ['su', 'yarn'], got {rope_scaling_type}")
-        if not (
-            isinstance(rope_scaling_short_factor, list)
-            and all(isinstance(x, (int, float)) for x in rope_scaling_short_factor)
-        ):
-            raise ValueError(
-                f"`rope_scaling`'s short_factor field must be a list of numbers, got {rope_scaling_short_factor}"
-            )
-        if not len(rope_scaling_short_factor) == self.hidden_size // self.num_attention_heads // 2:
-            raise ValueError(
-                f"`rope_scaling`'s short_factor field must have length {self.hidden_size // self.num_attention_heads // 2}, got {len(rope_scaling_short_factor)}"
-            )
-        if not (
-            isinstance(rope_scaling_long_factor, list)
-            and all(isinstance(x, (int, float)) for x in rope_scaling_long_factor)
-        ):
-            raise ValueError(
-                f"`rope_scaling`'s long_factor field must be a list of numbers, got {rope_scaling_long_factor}"
-            )
-        if not len(rope_scaling_long_factor) == self.hidden_size // self.num_attention_heads // 2:
-            raise ValueError(
-                f"`rope_scaling`'s long_factor field must have length {self.hidden_size // self.num_attention_heads // 2}, got {len(rope_scaling_long_factor)}"
-            )
+# coding=utf-8
+# Copyright 2024 Microsoft and the HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+""" Phi-3 model configuration"""
+
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+PHI3_PRETRAINED_CONFIG_ARCHIVE_MAP = {
+    "microsoft/Phi-3-mini-4k-instruct": "https://huggingface.co/microsoft/Phi-3-mini-4k-instruct/resolve/main/config.json",
+    "microsoft/Phi-3-mini-128k-instruct": "https://huggingface.co/microsoft/Phi-3-mini-128k-instruct/resolve/main/config.json",
+}
+
+
+class Phi3Config(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`Phi3Model`]. It is used to instantiate a Phi-3
+    model according to the specified arguments, defining the model architecture. Instantiating a configuration with the
+    defaults will yield a similar configuration to that of the
+    [microsoft/Phi-3-mini-4k-instruct](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct).
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        vocab_size (`int`, *optional*, defaults to 32064):
+            Vocabulary size of the Phi-3 model. Defines the number of different tokens that can be represented by the
+            `inputs_ids` passed when calling [`Phi3Model`].
+        hidden_size (`int`, *optional*, defaults to 3072):
+            Dimension of the hidden representations.
+        intermediate_size (`int`, *optional*, defaults to 8192):
+            Dimension of the MLP representations.
+        num_hidden_layers (`int`, *optional*, defaults to 32):
+            Number of hidden layers in the Transformer decoder.
+        num_attention_heads (`int`, *optional*, defaults to 32):
+            Number of attention heads for each attention layer in the Transformer decoder.
+        num_key_value_heads (`int`, *optional*):
+            This is the number of key_value heads that should be used to implement Grouped Query Attention. If
+            `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if
+            `num_key_value_heads=1 the model will use Multi Query Attention (MQA) otherwise GQA is used. When
+            converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed
+            by meanpooling all the original heads within that group. For more details checkout [this
+            paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to
+            `num_attention_heads`.
+        resid_pdrop (`float`, *optional*, defaults to 0.0):
+            Dropout probability for mlp outputs.
+        embd_pdrop (`int`, *optional*, defaults to 0.0):
+            The dropout ratio for the embeddings.
+        attention_dropout (`float`, *optional*, defaults to 0.0):
+            The dropout ratio after computing the attention scores.
+        hidden_act (`str` or `function`, *optional*, defaults to `"silu"`):
+            The non-linear activation function (function or string) in the decoder.
+        max_position_embeddings (`int`, *optional*, defaults to 4096):
+            The maximum sequence length that this model might ever be used with.
+        original_max_position_embeddings (`int`, *optional*, defaults to 4096):
+            The maximum sequence length that this model was trained with. This is used to determine the size of the
+            original RoPE embeddings when using long scaling.
+        initializer_range (`float`, *optional*, defaults to 0.02):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+        rms_norm_eps (`float`, *optional*, defaults to 1e-05):
+            The epsilon value used for the RMSNorm.
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should return the last key/values attentions (not used by all models). Only
+            relevant if `config.is_decoder=True`. Whether to tie weight embeddings or not.
+        tie_word_embeddings (`bool`, *optional*, defaults to `False`):
+            Whether to tie weight embeddings
+        rope_theta (`float`, *optional*, defaults to 10000.0):
+            The base period of the RoPE embeddings.
+        rope_scaling (`dict`, *optional*):
+            The scaling strategy for the RoPE embeddings. If `None`, no scaling is applied. If a dictionary, it must
+            contain the following keys: `type`, `short_factor` and `long_factor`. The `type` must be `longrope` and 
+            the `short_factor` and `long_factor` must be lists of numbers with the same length as the hidden size 
+            divided by the number of attention heads divided by 2.
+        bos_token_id (`int`, *optional*, defaults to 1):
+            The id of the "beginning-of-sequence" token.
+        eos_token_id (`int`, *optional*, defaults to 32000):
+            The id of the "end-of-sequence" token.
+        pad_token_id (`int`, *optional*, defaults to 32000):
+            The id of the padding token.
+        sliding_window (`int`, *optional*):
+            Sliding window attention window size. If `None`, no sliding window is applied.
+
+    Example:
+
+    ```python
+    >>> from transformers import Phi3Model, Phi3Config
+
+    >>> # Initializing a Phi-3 style configuration
+    >>> configuration = Phi3Config.from_pretrained("microsoft/Phi-3-mini-4k-instruct")
+
+    >>> # Initializing a model from the configuration
+    >>> model = Phi3Model(configuration)
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "phi3"
+    keys_to_ignore_at_inference = ["past_key_values"]
+
+    def __init__(
+        self,
+        vocab_size=32064,
+        hidden_size=3072,
+        intermediate_size=8192,
+        num_hidden_layers=32,
+        num_attention_heads=32,
+        num_key_value_heads=None,
+        resid_pdrop=0.0,
+        embd_pdrop=0.0,
+        attention_dropout=0.0,
+        hidden_act="silu",
+        max_position_embeddings=4096,
+        original_max_position_embeddings=4096,
+        initializer_range=0.02,
+        rms_norm_eps=1e-5,
+        use_cache=True,
+        tie_word_embeddings=False,
+        rope_theta=10000.0,
+        rope_scaling=None,
+        bos_token_id=1,
+        eos_token_id=32000,
+        pad_token_id=32000,
+        sliding_window=None,
+        **kwargs,
+    ):
+        self.vocab_size = vocab_size
+        self.hidden_size = hidden_size
+        self.intermediate_size = intermediate_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+
+        if num_key_value_heads is None:
+            num_key_value_heads = num_attention_heads
+
+        self.num_key_value_heads = num_key_value_heads
+        self.resid_pdrop = resid_pdrop
+        self.embd_pdrop = embd_pdrop
+        self.attention_dropout = attention_dropout
+        self.hidden_act = hidden_act
+        self.max_position_embeddings = max_position_embeddings
+        self.original_max_position_embeddings = original_max_position_embeddings
+        self.initializer_range = initializer_range
+        self.rms_norm_eps = rms_norm_eps
+        self.use_cache = use_cache
+        self.rope_theta = rope_theta
+        self.rope_scaling = rope_scaling
+        self._rope_scaling_adjustment()
+        self._rope_scaling_validation()
+        self.sliding_window = sliding_window
+
+        super().__init__(
+            bos_token_id=bos_token_id,
+            eos_token_id=eos_token_id,
+            pad_token_id=pad_token_id,
+            tie_word_embeddings=tie_word_embeddings,
+            **kwargs,
+        )
+
+    def _rope_scaling_adjustment(self):
+        """
+        Adjust the `type` of the `rope_scaling` configuration for backward compatibility.
+        """
+        if self.rope_scaling is None:
+            return
+
+        rope_scaling_type = self.rope_scaling.get("type", None)
+
+        # For backward compatibility if previous version used "su" or "yarn"
+        if rope_scaling_type is not None and rope_scaling_type in ["su", "yarn"]:
+            self.rope_scaling["type"] = "longrope"
+
+    def _rope_scaling_validation(self):
+        """
+        Validate the `rope_scaling` configuration.
+        """
+        if self.rope_scaling is None:
+            return
+
+        if not isinstance(self.rope_scaling, dict) or len(self.rope_scaling) != 3:
+            raise ValueError(
+                "`rope_scaling` must be a dictionary with three fields, `type`, `short_factor` and `long_factor`, "
+                f"got {self.rope_scaling}"
+            )
+        rope_scaling_type = self.rope_scaling.get("type", None)
+        rope_scaling_short_factor = self.rope_scaling.get("short_factor", None)
+        rope_scaling_long_factor = self.rope_scaling.get("long_factor", None)
+        if rope_scaling_type is None or rope_scaling_type not in ["longrope"]:
+            raise ValueError(f"`rope_scaling`'s type field must be one of ['longrope'], got {rope_scaling_type}")
+        if not (
+            isinstance(rope_scaling_short_factor, list)
+            and all(isinstance(x, (int, float)) for x in rope_scaling_short_factor)
+        ):
+            raise ValueError(
+                f"`rope_scaling`'s short_factor field must be a list of numbers, got {rope_scaling_short_factor}"
+            )
+        if not len(rope_scaling_short_factor) == self.hidden_size // self.num_attention_heads // 2:
+            raise ValueError(
+                f"`rope_scaling`'s short_factor field must have length {self.hidden_size // self.num_attention_heads // 2}, got {len(rope_scaling_short_factor)}"
+            )
+        if not (
+            isinstance(rope_scaling_long_factor, list)
+            and all(isinstance(x, (int, float)) for x in rope_scaling_long_factor)
+        ):
+            raise ValueError(
+                f"`rope_scaling`'s long_factor field must be a list of numbers, got {rope_scaling_long_factor}"
+            )
+        if not len(rope_scaling_long_factor) == self.hidden_size // self.num_attention_heads // 2:
+            raise ValueError(
+                f"`rope_scaling`'s long_factor field must have length {self.hidden_size // self.num_attention_heads // 2}, got {len(rope_scaling_long_factor)}"
+            )

generation_config.json

@@ -7,5 +7,5 @@
     32007
   ],
   "pad_token_id": 32000,
-  "transformers_version": "4.41.2"
+  "transformers_version": "4.45.2"
 }

openvino_config.json

@@ -0,0 +1,25 @@
+{
+  "compression": null,
+  "dtype": "int4",
+  "input_info": null,
+  "optimum_version": "1.23.1",
+  "quantization_config": {
+    "all_layers": null,
+    "bits": 4,
+    "dataset": "wikitext2",
+    "gptq": null,
+    "group_size": 64,
+    "ignored_scope": null,
+    "num_samples": null,
+    "quant_method": "default",
+    "ratio": 1.0,
+    "scale_estimation": true,
+    "sensitivity_metric": null,
+    "sym": false,
+    "tokenizer": null,
+    "trust_remote_code": true,
+    "weight_format": "int4"
+  },
+  "save_onnx_model": false,
+  "transformers_version": "4.45.2"
+}

openvino_detokenizer.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:630b6806812464da49d8dc0907d303055c3fa69f10b1f3533f6945437ab55b59
-size 499991
+oid sha256:abf0a5ac7698c27f1f3a8573b76a628e4a6a2c7eaddc7dd549ee3607a34d4061
+size 339125

openvino_detokenizer.xml

@@ -1,61 +1,235 @@
 <?xml version="1.0"?>
 <net name="detokenizer" version="11">
 	<layers>
-		<layer id="0" name="Parameter_37" type="Parameter" version="opset1">
+		<layer id="0" name="Parameter_282016" type="Parameter" version="opset1">
 			<data shape="?,?" element_type="i64" />
 			<output>
-				<port id="0" precision="I64" names="Parameter_37">
+				<port id="0" precision="I64" names="Parameter_282016">
 					<dim>-1</dim>
 					<dim>-1</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="1" name="Constant_2" type="Const" version="opset1">
-			<data element_type="u8" shape="499991" offset="0" size="499991" />
+		<layer id="1" name="Convert_282038" 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_281986" type="Const" version="opset1">
+			<data element_type="u8" shape="339118" offset="0" size="339118" />
 			<output>
 				<port id="0" precision="U8">
-					<dim>499991</dim>
+					<dim>339118</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="2" name="Convert_47" type="Convert" version="opset1">
-			<data destination_type="i32" />
+		<layer id="3" name="StringTensorUnpack_281987" type="StringTensorUnpack" version="extension">
+			<data mode="begins_ends" />
 			<input>
-				<port id="0" precision="I64">
+				<port id="0" precision="U8">
+					<dim>339118</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="4" name="VocabDecoder_282017" type="VocabDecoder" version="extension">
+			<data skip_tokens="0, 1, 32000, 32001, 32002, 32003, 32004, 32005, 32006, 32007, 32008, 32009, 32010" />
+			<input>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+					<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="U8">
 					<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="U8">
+					<dim>-1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="5" name="Constant_282019" type="Const" version="opset1">
+			<data element_type="u8" shape="3" offset="339118" size="3" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>3</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="6" name="Constant_282021" type="Const" version="opset1">
+			<data element_type="u8" shape="1" offset="339121" size="1" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="7" name="RegexNormalization_282022" type="RegexNormalization" version="extension">
+			<data global_replace="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>
+				<port id="3" precision="U8">
+					<dim>3</dim>
+				</port>
+				<port id="4" precision="U8">
+					<dim>1</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="U8">
 					<dim>-1</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="3" name="SentencepieceDetokenizer_38" type="SentencepieceDetokenizer" version="extension">
+		<layer id="8" name="ByteFallback_282023" type="ByteFallback" version="extension">
 			<input>
-				<port id="0" precision="U8">
-					<dim>499991</dim>
+				<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="FuzeRagged_282024" 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="10" name="Constant_282026" type="Const" version="opset1">
+			<data element_type="u8" shape="2" offset="339122" size="2" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>2</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="11" name="Constant_282028" type="Const" version="opset1">
+			<data element_type="u8" shape="0" offset="339124" size="1" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>0</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="12" name="RegexNormalization_282029" type="RegexNormalization" version="extension">
+			<data global_replace="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>
+				<port id="3" precision="U8">
+					<dim>2</dim>
+				</port>
 				<port id="4" precision="U8">
+					<dim>0</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="U8">
 					<dim>-1</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="4" name="StringTensorPack_39" type="StringTensorPack" version="extension">
+		<layer id="13" name="StringTensorPack_282030" type="StringTensorPack" version="extension">
 			<data mode="begins_ends" />
 			<input>
 				<port id="0" precision="I32">
@@ -74,7 +248,7 @@
 				</port>
 			</output>
 		</layer>
-		<layer id="5" name="Result_40" type="Result" version="opset1">
+		<layer id="14" name="Result_282031" type="Result" version="opset1">
 			<input>
 				<port id="0" precision="STRING">
 					<dim>-1</dim>
@@ -83,13 +257,33 @@
 		</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="3" to-port="0" />
-		<edge from-layer="2" from-port="1" to-layer="3" to-port="1" />
-		<edge from-layer="3" from-port="2" to-layer="4" to-port="0" />
-		<edge from-layer="3" from-port="3" to-layer="4" to-port="1" />
-		<edge from-layer="3" from-port="4" to-layer="4" to-port="2" />
-		<edge from-layer="4" from-port="3" to-layer="5" to-port="0" />
+		<edge from-layer="0" from-port="0" to-layer="1" to-port="0" />
+		<edge from-layer="1" from-port="1" to-layer="4" to-port="0" />
+		<edge from-layer="2" from-port="0" to-layer="3" to-port="0" />
+		<edge from-layer="3" from-port="1" to-layer="4" to-port="1" />
+		<edge from-layer="3" from-port="2" to-layer="4" to-port="2" />
+		<edge from-layer="3" from-port="3" to-layer="4" to-port="3" />
+		<edge from-layer="4" from-port="6" to-layer="7" to-port="0" />
+		<edge from-layer="4" from-port="7" to-layer="7" to-port="1" />
+		<edge from-layer="4" from-port="8" to-layer="7" to-port="2" />
+		<edge from-layer="4" from-port="5" to-layer="9" to-port="1" />
+		<edge from-layer="4" from-port="4" to-layer="9" to-port="0" />
+		<edge from-layer="5" from-port="0" to-layer="7" to-port="3" />
+		<edge from-layer="6" from-port="0" to-layer="7" to-port="4" />
+		<edge from-layer="7" from-port="7" to-layer="8" to-port="2" />
+		<edge from-layer="7" from-port="6" to-layer="8" to-port="1" />
+		<edge from-layer="7" from-port="5" to-layer="8" to-port="0" />
+		<edge from-layer="8" from-port="3" to-layer="9" to-port="2" />
+		<edge from-layer="8" from-port="4" to-layer="9" to-port="3" />
+		<edge from-layer="8" from-port="5" to-layer="12" to-port="2" />
+		<edge from-layer="9" from-port="4" to-layer="12" to-port="0" />
+		<edge from-layer="9" from-port="5" to-layer="12" to-port="1" />
+		<edge from-layer="10" from-port="0" to-layer="12" to-port="3" />
+		<edge from-layer="11" from-port="0" to-layer="12" to-port="4" />
+		<edge from-layer="12" from-port="5" to-layer="13" to-port="0" />
+		<edge from-layer="12" from-port="6" to-layer="13" to-port="1" />
+		<edge from-layer="12" from-port="7" to-layer="13" to-port="2" />
+		<edge from-layer="13" from-port="3" to-layer="14" to-port="0" />
 	</edges>
 	<rt_info>
 		<bos_token_id value="1" />

openvino_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:22291c7526fdd7b67a17f415287c6c8b473fae8fcd7afa4887e3ee436d306ea2
-size 2450248240
+oid sha256:4966cb2c4f416cb20b934a7130d4aa5aefa5007b56c817b55dcfe53b1506c415
+size 2151489888

openvino_tokenizer.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:996b60b8de0ee7597bb22dadabc6bed436a4255e0b2e2cbc7a11ad7ffbad8613
-size 500278
+oid sha256:012a18fef2df281b8293a03448630081625d563d29d199b482a64751609a5698
+size 1262043

openvino_tokenizer.xml

@@ -1,64 +1,204 @@
 <?xml version="1.0"?>
 <net name="tokenizer" version="11">
 	<layers>
-		<layer id="0" name="string_input" type="Parameter" version="opset1">
+		<layer id="0" name="Parameter_281898" type="Parameter" version="opset1">
 			<data shape="?" element_type="string" />
 			<output>
-				<port id="0" precision="STRING" names="string_input">
+				<port id="0" precision="STRING" names="Parameter_281898">
 					<dim>-1</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="1" name="Constant_16" type="Const" version="opset1">
-			<data element_type="i32" shape="" offset="0" size="4" />
+		<layer id="1" name="Constant_281904" type="Const" version="opset1">
+			<data element_type="i64" shape="" offset="0" size="8" />
 			<output>
-				<port id="0" precision="I32" />
+				<port id="0" precision="I64" />
 			</output>
 		</layer>
-		<layer id="2" name="Constant_1" type="Const" version="opset1">
-			<data element_type="u8" shape="499969" offset="4" size="499969" />
+		<layer id="2" name="StringTensorUnpack_281899" type="StringTensorUnpack" version="extension">
+			<data mode="begins_ends" />
+			<input>
+				<port id="0" precision="STRING">
+					<dim>-1</dim>
+				</port>
+			</input>
 			<output>
-				<port id="0" precision="U8">
-					<dim>499969</dim>
+				<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="StringTensorUnpack_4" type="StringTensorUnpack" version="extension">
-			<data mode="begins_ends" />
+		<layer id="3" name="ShapeOf_281900" type="ShapeOf" version="opset3">
+			<data output_type="i64" />
 			<input>
-				<port id="0" precision="STRING">
+				<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_281901" 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_281902" 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_281903" 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_281905" 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_281906" 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_281907" 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_281908" 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_281909" 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_281910" 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_281911" 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_281973" type="Const" version="opset1">
+			<data element_type="u8" shape="328" offset="16" size="328" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>328</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="15" name="SpecialTokensSplit_281974" 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="U8">
+				<port id="3" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="4" precision="U8">
+					<dim>-1</dim>
+				</port>
+				<port id="5" precision="U8">
+					<dim>328</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="4" name="Constant_6" type="Const" version="opset1">
-			<data element_type="u8" shape="7" offset="499973" size="7" />
+		<layer id="16" name="Constant_281976" type="Const" version="opset1">
+			<data element_type="u8" shape="7" offset="344" size="7" />
 			<output>
 				<port id="0" precision="U8">
 					<dim>7</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="5" name="Constant_8" type="Const" version="opset1">
-			<data element_type="u8" shape="3" offset="499980" size="3" />
+		<layer id="17" name="Constant_281978" type="Const" version="opset1">
+			<data element_type="u8" shape="5" offset="351" size="5" />
 			<output>
 				<port id="0" precision="U8">
-					<dim>3</dim>
+					<dim>5</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="6" name="RegexNormalization_9" type="RegexNormalization" version="extension">
+		<layer id="18" name="RegexNormalization_281979" type="RegexNormalization" version="extension">
 			<data global_replace="true" />
 			<input>
 				<port id="0" precision="I32">
@@ -70,27 +210,49 @@
 				<port id="2" precision="U8">
 					<dim>-1</dim>
 				</port>
-				<port id="3" precision="U8">
-					<dim>7</dim>
+				<port id="3" precision="BOOL">
+					<dim>-1</dim>
 				</port>
 				<port id="4" precision="U8">
-					<dim>3</dim>
+					<dim>7</dim>
+				</port>
+				<port id="5" precision="U8">
+					<dim>5</dim>
 				</port>
 			</input>
 			<output>
-				<port id="5" precision="I32">
+				<port id="6" precision="I32">
 					<dim>-1</dim>
 				</port>
-				<port id="6" precision="I32">
+				<port id="7" precision="I32">
 					<dim>-1</dim>
 				</port>
-				<port id="7" precision="U8">
+				<port id="8" precision="U8">
+					<dim>-1</dim>
+				</port>
+				<port id="9" precision="BOOL">
 					<dim>-1</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="7" name="StringTensorPack_10" type="StringTensorPack" version="extension">
-			<data mode="begins_ends" />
+		<layer id="19" name="Constant_281981" type="Const" version="opset1">
+			<data element_type="u8" shape="1" offset="356" size="1" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>1</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="20" name="Constant_281983" type="Const" version="opset1">
+			<data element_type="u8" shape="3" offset="357" size="3" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>3</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="21" name="RegexNormalization_281984" type="RegexNormalization" version="extension">
+			<data global_replace="true" />
 			<input>
 				<port id="0" precision="I32">
 					<dim>-1</dim>
@@ -101,26 +263,44 @@
 				<port id="2" precision="U8">
 					<dim>-1</dim>
 				</port>
+				<port id="3" precision="BOOL">
+					<dim>-1</dim>
+				</port>
+				<port id="4" precision="U8">
+					<dim>1</dim>
+				</port>
+				<port id="5" precision="U8">
+					<dim>3</dim>
+				</port>
 			</input>
 			<output>
-				<port id="3" precision="STRING">
+				<port id="6" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="7" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="8" precision="U8">
+					<dim>-1</dim>
+				</port>
+				<port id="9" precision="BOOL">
 					<dim>-1</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="8" name="Constant_12" type="Const" version="opset1">
-			<data element_type="u8" shape="223" offset="499983" size="223" />
+		<layer id="22" name="Constant_281986" type="Const" version="opset1">
+			<data element_type="u8" shape="339118" offset="360" size="339118" />
 			<output>
 				<port id="0" precision="U8">
-					<dim>223</dim>
+					<dim>339118</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="9" name="StringTensorUnpack_13" type="StringTensorUnpack" version="extension">
+		<layer id="23" name="StringTensorUnpack_281987" type="StringTensorUnpack" version="extension">
 			<data mode="begins_ends" />
 			<input>
 				<port id="0" precision="U8">
-					<dim>223</dim>
+					<dim>339118</dim>
 				</port>
 			</input>
 			<output>
@@ -135,89 +315,173 @@
 				</port>
 			</output>
 		</layer>
-		<layer id="10" name="Constant_14" type="Const" version="opset1">
-			<data element_type="i32" shape="14" offset="500206" size="56" />
+		<layer id="24" name="Constant_281992" type="Const" version="opset1">
+			<data element_type="u8" shape="499127" offset="339478" size="499127" />
 			<output>
-				<port id="0" precision="I32">
-					<dim>14</dim>
+				<port id="0" precision="U8">
+					<dim>499127</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="11" name="SentencepieceTokenizer_15" type="SentencepieceTokenizer" version="extension">
-			<data nbest_size="0" alpha="0" add_bos="false" add_eos="false" reverse="true" />
+		<layer id="25" name="StringTensorUnpack_281993" type="StringTensorUnpack" version="extension">
+			<data mode="begins_ends" />
 			<input>
 				<port id="0" precision="U8">
-					<dim>499969</dim>
+					<dim>499127</dim>
 				</port>
-				<port id="1" precision="STRING">
+			</input>
+			<output>
+				<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="3" precision="U8">
 					<dim>-1</dim>
 				</port>
-				<port id="4" precision="U8">
-					<dim>-1</dim>
+			</output>
+		</layer>
+		<layer id="26" name="Constant_281995" type="Const" version="opset1">
+			<data element_type="u8" shape="412810" offset="838605" size="412810" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>412810</dim>
 				</port>
-				<port id="5" precision="I32">
-					<dim>14</dim>
+			</output>
+		</layer>
+		<layer id="27" name="StringTensorUnpack_281996" type="StringTensorUnpack" version="extension">
+			<data mode="begins_ends" />
+			<input>
+				<port id="0" precision="U8">
+					<dim>412810</dim>
 				</port>
 			</input>
 			<output>
-				<port id="6" precision="I64">
+				<port id="1" precision="I32">
 					<dim>-1</dim>
-					<dim>2</dim>
 				</port>
-				<port id="7" precision="I32">
+				<port id="2" precision="I32">
 					<dim>-1</dim>
 				</port>
-				<port id="8" precision="I64">
-					<dim>2</dim>
+				<port id="3" precision="U8">
+					<dim>-1</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="12" name="Broadcast_17" type="Broadcast" version="opset3">
-			<data mode="numpy" />
+		<layer id="28" name="Constant_281989" type="Const" version="opset1">
+			<data element_type="u8" shape="8716" offset="1251415" size="8716" />
+			<output>
+				<port id="0" precision="U8">
+					<dim>8716</dim>
+				</port>
+			</output>
+		</layer>
+		<layer id="29" name="StringTensorUnpack_281990" type="StringTensorUnpack" version="extension">
+			<data mode="begins_ends" />
 			<input>
-				<port id="0" precision="I32" />
-				<port id="1" precision="I64">
-					<dim>2</dim>
+				<port id="0" precision="U8">
+					<dim>8716</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="13" name="Constant_18" type="Const" version="opset1">
-			<data element_type="i32" shape="" offset="500262" size="4" />
+		<layer id="30" name="Constant_281997" type="Const" version="opset1">
+			<data element_type="i32" shape="475" offset="1260131" size="1900" />
 			<output>
-				<port id="0" precision="I32" />
+				<port id="0" precision="I32">
+					<dim>475</dim>
+				</port>
 			</output>
 		</layer>
-		<layer id="14" name="ShapeOf_19" type="ShapeOf" version="opset3">
-			<data output_type="i64" />
+		<layer id="31" name="BPETokenizer_281998" type="BPETokenizer" version="extension">
+			<data unk_token="&lt;unk>" fuse_unk="true" suffix_indicator="" end_suffix="" byte_fallback="true" cache_capacity="20000" />
 			<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>-1</dim>
+				</port>
+				<port id="6" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="7" precision="U8">
+					<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="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="12" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="13" precision="U8">
+					<dim>-1</dim>
+				</port>
+				<port id="14" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="15" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="16" precision="U8">
+					<dim>-1</dim>
+				</port>
+				<port id="17" precision="I32">
+					<dim>475</dim>
+				</port>
 			</input>
 			<output>
-				<port id="1" precision="I64">
-					<dim>1</dim>
+				<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="15" name="Broadcast_20" type="Broadcast" version="opset3">
-			<data mode="numpy" />
+		<layer id="32" name="Subtract_281999" type="Subtract" version="opset1">
+			<data auto_broadcast="numpy" />
 			<input>
-				<port id="0" precision="I32" />
-				<port id="1" precision="I64">
-					<dim>1</dim>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
 				</port>
 			</input>
 			<output>
@@ -226,137 +490,183 @@
 				</port>
 			</output>
 		</layer>
-		<layer id="16" name="ScatterNDUpdate_24" type="ScatterNDUpdate" version="opset4">
+		<layer id="33" name="Constant_282000" type="Const" version="opset1">
+			<data element_type="i32" shape="" offset="1262031" size="4" />
+			<output>
+				<port id="0" precision="I32" />
+			</output>
+		</layer>
+		<layer id="34" name="Minimum_282001" 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>
-				<port id="1" precision="I64">
+			</output>
+		</layer>
+		<layer id="35" name="Subtract_282002" type="Subtract" version="opset1">
+			<data auto_broadcast="numpy" />
+			<input>
+				<port id="0" precision="I32">
 					<dim>-1</dim>
-					<dim>2</dim>
 				</port>
-				<port id="2" precision="I32">
+				<port id="1" precision="I32">
 					<dim>-1</dim>
 				</port>
 			</input>
 			<output>
-				<port id="3" precision="I32">
-					<dim>-1</dim>
+				<port id="2" precision="I32">
 					<dim>-1</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="17" name="Constant_28" type="Const" version="opset1">
-			<data element_type="i64" shape="1" offset="500266" size="8" />
+		<layer id="36" name="Constant_282003" type="Const" version="opset1">
+			<data element_type="i32" shape="1" offset="1262035" size="4" />
 			<output>
-				<port id="0" precision="I64">
+				<port id="0" precision="I32">
 					<dim>1</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="18" name="Reverse_29" type="Reverse" version="opset1">
-			<data mode="index" />
+		<layer id="37" name="CombineSegments_282004" 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="1" precision="I64">
+				<port id="2" precision="I32">
+					<dim>-1</dim>
+				</port>
+				<port id="3" precision="I32">
 					<dim>1</dim>
 				</port>
 			</input>
 			<output>
-				<port id="2" precision="I32">
+				<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="19" name="Reverse_29" type="Convert" version="opset1">
-			<data destination_type="i64" />
+		<layer id="38" name="Subtract_282005" 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="1" precision="I64" names="attention_mask">
-					<dim>-1</dim>
+				<port id="2" precision="I32">
 					<dim>-1</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="21" name="Constant_25" type="Const" version="opset1">
-			<data element_type="i32" shape="" offset="500274" size="4" />
+		<layer id="39" name="Constant_282006" type="Const" version="opset1">
+			<data element_type="i32" shape="" offset="1262035" size="4" />
 			<output>
 				<port id="0" precision="I32" />
 			</output>
 		</layer>
-		<layer id="22" name="Broadcast_26" type="Broadcast" version="opset3">
-			<data mode="bidirectional" />
+		<layer id="40" name="ReduceMax_282007" type="ReduceMax" version="opset1">
+			<data keep_dims="false" />
 			<input>
-				<port id="0" precision="I32" />
-				<port id="1" precision="I64">
-					<dim>2</dim>
+				<port id="0" precision="I32">
+					<dim>-1</dim>
 				</port>
+				<port id="1" precision="I32" />
 			</input>
 			<output>
-				<port id="2" precision="I32">
-					<dim>-1</dim>
-					<dim>-1</dim>
-				</port>
+				<port id="2" precision="I32" />
+			</output>
+		</layer>
+		<layer id="41" name="Constant_282008" type="Const" version="opset1">
+			<data element_type="i32" shape="" offset="1262039" size="4" />
+			<output>
+				<port id="0" precision="I32" />
 			</output>
 		</layer>
-		<layer id="23" name="ScatterNDUpdate_27" type="ScatterNDUpdate" version="opset4">
+		<layer id="42" name="RaggedToDense_282009" type="RaggedToDense" version="extension">
+			<data pad_right="false" />
 			<input>
 				<port id="0" precision="I32">
 					<dim>-1</dim>
-					<dim>-1</dim>
 				</port>
-				<port id="1" precision="I64">
+				<port id="1" precision="I32">
 					<dim>-1</dim>
-					<dim>2</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="3" precision="I32">
+				<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="24" name="Constant_30" type="Const" version="opset1">
-			<data element_type="i64" shape="1" offset="500266" size="8" />
+		<layer id="43" name="Convert_282010" 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="0" precision="I64">
-					<dim>1</dim>
+				<port id="1" precision="I32">
+					<dim>-1</dim>
+					<dim>-1</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="25" name="Reverse_31" type="Reverse" version="opset1">
-			<data mode="index" />
+		<layer id="44" name="Convert_282010" type="Convert" version="opset1">
+			<data destination_type="i64" />
 			<input>
 				<port id="0" precision="I32">
 					<dim>-1</dim>
 					<dim>-1</dim>
 				</port>
-				<port id="1" precision="I64">
-					<dim>1</dim>
-				</port>
 			</input>
 			<output>
-				<port id="2" precision="I32">
+				<port id="1" precision="I64" names="attention_mask">
 					<dim>-1</dim>
 					<dim>-1</dim>
 				</port>
 			</output>
 		</layer>
-		<layer id="26" name="Reverse_31" type="Convert" version="opset1">
+		<layer id="46" name="RaggedToDense_282009.0" type="Convert" version="opset1">
 			<data destination_type="i64" />
 			<input>
 				<port id="0" precision="I32">
@@ -371,7 +681,7 @@
 				</port>
 			</output>
 		</layer>
-		<layer id="27" name="Result_32" type="Result" version="opset1">
+		<layer id="47" name="Result_282013" type="Result" version="opset1">
 			<input>
 				<port id="0" precision="I64">
 					<dim>-1</dim>
@@ -379,7 +689,7 @@
 				</port>
 			</input>
 		</layer>
-		<layer id="20" name="Result_33" type="Result" version="opset1">
+		<layer id="45" name="Result_282015" type="Result" version="opset1">
 			<input>
 				<port id="0" precision="I64">
 					<dim>-1</dim>
@@ -389,43 +699,83 @@
 		</layer>
 	</layers>
 	<edges>
-		<edge from-layer="0" from-port="0" to-layer="3" to-port="0" />
-		<edge from-layer="1" from-port="0" to-layer="12" to-port="0" />
-		<edge from-layer="2" from-port="0" to-layer="11" to-port="0" />
+		<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="3" from-port="2" to-layer="6" to-port="1" />
-		<edge from-layer="3" from-port="3" 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="5" to-layer="7" to-port="0" />
-		<edge from-layer="6" from-port="6" to-layer="7" to-port="1" />
-		<edge from-layer="6" from-port="7" to-layer="7" to-port="2" />
-		<edge from-layer="7" from-port="3" to-layer="11" to-port="1" />
-		<edge from-layer="8" from-port="0" to-layer="9" to-port="0" />
-		<edge from-layer="9" from-port="1" to-layer="11" to-port="2" />
-		<edge from-layer="9" from-port="2" to-layer="11" to-port="3" />
-		<edge from-layer="9" from-port="3" to-layer="11" to-port="4" />
-		<edge from-layer="10" from-port="0" to-layer="11" to-port="5" />
-		<edge from-layer="11" from-port="7" to-layer="14" to-port="0" />
-		<edge from-layer="11" from-port="7" to-layer="23" to-port="2" />
-		<edge from-layer="11" from-port="6" to-layer="23" to-port="1" />
-		<edge from-layer="11" from-port="8" to-layer="22" to-port="1" />
-		<edge from-layer="11" from-port="6" to-layer="16" to-port="1" />
-		<edge from-layer="11" from-port="8" to-layer="12" to-port="1" />
-		<edge from-layer="12" from-port="2" to-layer="16" to-port="0" />
-		<edge from-layer="13" from-port="0" to-layer="15" to-port="0" />
-		<edge from-layer="14" from-port="1" to-layer="15" to-port="1" />
-		<edge from-layer="15" from-port="2" to-layer="16" to-port="2" />
-		<edge from-layer="16" from-port="3" to-layer="18" to-port="0" />
-		<edge from-layer="17" from-port="0" to-layer="18" to-port="1" />
-		<edge from-layer="18" from-port="2" to-layer="19" to-port="0" />
-		<edge from-layer="19" from-port="1" to-layer="20" to-port="0" />
-		<edge from-layer="21" from-port="0" to-layer="22" to-port="0" />
-		<edge from-layer="22" from-port="2" to-layer="23" to-port="0" />
-		<edge from-layer="23" from-port="3" to-layer="25" to-port="0" />
-		<edge from-layer="24" from-port="0" to-layer="25" to-port="1" />
-		<edge from-layer="25" from-port="2" to-layer="26" to-port="0" />
-		<edge from-layer="26" from-port="1" to-layer="27" 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="9" to-layer="18" to-port="1" />
+		<edge from-layer="15" from-port="6" to-layer="31" to-port="0" />
+		<edge from-layer="15" from-port="7" to-layer="31" to-port="1" />
+		<edge from-layer="15" from-port="11" to-layer="18" to-port="3" />
+		<edge from-layer="15" from-port="10" to-layer="18" to-port="2" />
+		<edge from-layer="15" from-port="8" to-layer="18" to-port="0" />
+		<edge from-layer="16" from-port="0" to-layer="18" to-port="4" />
+		<edge from-layer="17" from-port="0" to-layer="18" to-port="5" />
+		<edge from-layer="18" from-port="6" to-layer="21" to-port="0" />
+		<edge from-layer="18" from-port="7" to-layer="21" to-port="1" />
+		<edge from-layer="18" from-port="8" to-layer="21" to-port="2" />
+		<edge from-layer="18" from-port="9" to-layer="21" to-port="3" />
+		<edge from-layer="19" from-port="0" to-layer="21" to-port="4" />
+		<edge from-layer="20" from-port="0" to-layer="21" to-port="5" />
+		<edge from-layer="21" from-port="8" to-layer="31" to-port="4" />
+		<edge from-layer="21" from-port="7" to-layer="31" to-port="3" />
+		<edge from-layer="21" from-port="6" to-layer="31" to-port="2" />
+		<edge from-layer="22" from-port="0" to-layer="23" to-port="0" />
+		<edge from-layer="23" from-port="1" to-layer="31" to-port="5" />
+		<edge from-layer="23" from-port="2" to-layer="31" to-port="6" />
+		<edge from-layer="23" from-port="3" to-layer="31" to-port="7" />
+		<edge from-layer="24" from-port="0" to-layer="25" to-port="0" />
+		<edge from-layer="25" from-port="1" to-layer="31" to-port="8" />
+		<edge from-layer="25" from-port="2" to-layer="31" to-port="9" />
+		<edge from-layer="25" from-port="3" to-layer="31" to-port="10" />
+		<edge from-layer="26" from-port="0" to-layer="27" to-port="0" />
+		<edge from-layer="27" from-port="3" to-layer="31" to-port="13" />
+		<edge from-layer="27" from-port="2" to-layer="31" to-port="12" />
+		<edge from-layer="27" from-port="1" to-layer="31" to-port="11" />
+		<edge from-layer="28" from-port="0" to-layer="29" to-port="0" />
+		<edge from-layer="29" from-port="1" to-layer="31" to-port="14" />
+		<edge from-layer="29" from-port="2" to-layer="31" to-port="15" />
+		<edge from-layer="29" from-port="3" 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="32" 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="35" to-port="0" />
+		<edge from-layer="31" from-port="20" to-layer="37" to-port="2" />
+		<edge from-layer="31" from-port="19" to-layer="37" to-port="1" />
+		<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="37" to-port="0" />
+		<edge from-layer="36" from-port="0" to-layer="37" to-port="3" />
+		<edge from-layer="37" from-port="5" to-layer="42" to-port="1" />
+		<edge from-layer="37" from-port="6" to-layer="42" to-port="2" />
+		<edge from-layer="37" from-port="4" to-layer="42" to-port="0" />
+		<edge from-layer="37" from-port="4" to-layer="38" to-port="1" />
+		<edge from-layer="37" from-port="5" to-layer="38" to-port="0" />
+		<edge from-layer="38" from-port="2" to-layer="40" to-port="0" />
+		<edge from-layer="39" from-port="0" to-layer="40" to-port="1" />
+		<edge from-layer="40" from-port="2" to-layer="42" to-port="3" />
+		<edge from-layer="41" from-port="0" to-layer="42" to-port="4" />
+		<edge from-layer="42" from-port="6" to-layer="43" to-port="0" />
+		<edge from-layer="42" from-port="5" to-layer="46" to-port="0" />
+		<edge from-layer="43" from-port="1" to-layer="44" to-port="0" />
+		<edge from-layer="44" from-port="1" to-layer="45" to-port="0" />
+		<edge from-layer="46" from-port="1" to-layer="47" to-port="0" />
 	</edges>
 	<rt_info>
 		<bos_token_id value="1" />