Upload model

config.json

@@ -0,0 +1,209 @@
+{
+  "_commit_hash": null,
+  "_name_or_path": "/data/kch/output/v5-20250118-162439/checkpoint-852",
+  "architectures": [
+    "InternVLChatModel"
+  ],
+  "auto_map": {
+    "AutoConfig": "configuration_internvl_chat.InternVLChatConfig",
+    "AutoModel": "modeling_internvl_chat.InternVLChatModel",
+    "AutoModelForCausalLM": "modeling_internvl_chat.InternVLChatModel"
+  },
+  "downsample_ratio": 0.5,
+  "dynamic_image_size": true,
+  "force_image_size": 448,
+  "hidden_size": 4096,
+  "keys_to_ignore_at_inference": [
+    "past_key_values"
+  ],
+  "llm_config": {
+    "_attn_implementation_autoset": true,
+    "_name_or_path": "internlm/internlm2_5-7b-chat",
+    "add_cross_attention": false,
+    "architectures": [
+      "InternLM2ForCausalLM"
+    ],
+    "attn_implementation": "flash_attention_2",
+    "auto_map": {
+      "AutoConfig": "configuration_internlm2.InternLM2Config",
+      "AutoModel": "modeling_internlm2.InternLM2ForCausalLM",
+      "AutoModelForCausalLM": "modeling_internlm2.InternLM2ForCausalLM",
+      "AutoModelForSequenceClassification": "modeling_internlm2.InternLM2ForSequenceClassification"
+    },
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bias": false,
+    "bos_token_id": 1,
+    "chunk_size_feed_forward": 0,
+    "cross_attention_hidden_size": null,
+    "decoder_start_token_id": null,
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "early_stopping": false,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": 2,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "hidden_act": "silu",
+    "hidden_size": 4096,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "initializer_range": 0.02,
+    "intermediate_size": 14336,
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "max_position_embeddings": 32768,
+    "min_length": 0,
+    "model_type": "internlm2",
+    "no_repeat_ngram_size": 0,
+    "num_attention_heads": 32,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_hidden_layers": 32,
+    "num_key_value_heads": 8,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": 2,
+    "prefix": null,
+    "pretraining_tp": 1,
+    "problem_type": null,
+    "pruned_heads": {},
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "rms_norm_eps": 1e-05,
+    "rope_scaling": {
+      "factor": 2.0,
+      "type": "dynamic"
+    },
+    "rope_theta": 1000000,
+    "sep_token_id": null,
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "tf_legacy_loss": false,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": false,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torch_dtype": "bfloat16",
+    "torchscript": false,
+    "transformers_version": "4.46.3",
+    "typical_p": 1.0,
+    "use_bfloat16": true,
+    "use_cache": true,
+    "vocab_size": 92553
+  },
+  "max_dynamic_patch": 12,
+  "min_dynamic_patch": 1,
+  "model_type": "internvl_chat",
+  "pad_token_id": 2,
+  "ps_version": "v2",
+  "select_layer": -1,
+  "template": "internvl2_5",
+  "torch_dtype": "bfloat16",
+  "transformers_version": null,
+  "use_backbone_lora": 0,
+  "use_cache": false,
+  "use_llm_lora": 0,
+  "use_thumbnail": true,
+  "vision_config": {
+    "_attn_implementation_autoset": true,
+    "_name_or_path": "",
+    "add_cross_attention": false,
+    "architectures": [
+      "InternVisionModel"
+    ],
+    "attention_dropout": 0.0,
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": null,
+    "chunk_size_feed_forward": 0,
+    "cross_attention_hidden_size": null,
+    "decoder_start_token_id": null,
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "drop_path_rate": 0.0,
+    "dropout": 0.0,
+    "early_stopping": false,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": null,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "hidden_act": "gelu",
+    "hidden_size": 1024,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "image_size": 448,
+    "initializer_factor": 1.0,
+    "initializer_range": 0.02,
+    "intermediate_size": 4096,
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "layer_norm_eps": 1e-06,
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "min_length": 0,
+    "model_type": "intern_vit_6b",
+    "no_repeat_ngram_size": 0,
+    "norm_type": "layer_norm",
+    "num_attention_heads": 16,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_channels": 3,
+    "num_hidden_layers": 24,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": null,
+    "patch_size": 14,
+    "prefix": null,
+    "problem_type": null,
+    "pruned_heads": {},
+    "qk_normalization": false,
+    "qkv_bias": true,
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "sep_token_id": null,
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "tf_legacy_loss": false,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": true,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torch_dtype": "bfloat16",
+    "torchscript": false,
+    "transformers_version": "4.46.3",
+    "typical_p": 1.0,
+    "use_bfloat16": true,
+    "use_flash_attn": true
+  }
+}

configuration_intern_vit.py

@@ -0,0 +1,120 @@
+# --------------------------------------------------------
+# InternVL
+# Copyright (c) 2024 OpenGVLab
+# Licensed under The MIT License [see LICENSE for details]
+# --------------------------------------------------------
+
+import os
+from typing import Union
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+
+logger = logging.get_logger(__name__)
+
+
+class InternVisionConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`InternVisionModel`]. It is used to
+    instantiate a vision encoder according to the specified arguments, defining the model architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        num_channels (`int`, *optional*, defaults to 3):
+            Number of color channels in the input images (e.g., 3 for RGB).
+        patch_size (`int`, *optional*, defaults to 14):
+            The size (resolution) of each patch.
+        image_size (`int`, *optional*, defaults to 224):
+            The size (resolution) of each image.
+        qkv_bias (`bool`, *optional*, defaults to `False`):
+            Whether to add a bias to the queries and values in the self-attention layers.
+        hidden_size (`int`, *optional*, defaults to 3200):
+            Dimensionality of the encoder layers and the pooler layer.
+        num_attention_heads (`int`, *optional*, defaults to 25):
+            Number of attention heads for each attention layer in the Transformer encoder.
+        intermediate_size (`int`, *optional*, defaults to 12800):
+            Dimensionality of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder.
+        qk_normalization (`bool`, *optional*, defaults to `True`):
+            Whether to normalize the queries and keys in the self-attention layers.
+        num_hidden_layers (`int`, *optional*, defaults to 48):
+            Number of hidden layers in the Transformer encoder.
+        use_flash_attn (`bool`, *optional*, defaults to `True`):
+            Whether to use flash attention mechanism.
+        hidden_act (`str` or `function`, *optional*, defaults to `"gelu"`):
+            The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
+            `"relu"`, `"selu"` and `"gelu_new"` ``"gelu"` are supported.
+        layer_norm_eps (`float`, *optional*, defaults to 1e-6):
+            The epsilon used by the layer normalization layers.
+        dropout (`float`, *optional*, defaults to 0.0):
+            The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
+        drop_path_rate (`float`, *optional*, defaults to 0.0):
+            Dropout rate for stochastic depth.
+        attention_dropout (`float`, *optional*, defaults to 0.0):
+            The dropout ratio for the attention probabilities.
+        initializer_range (`float`, *optional*, defaults to 0.02):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+        initializer_factor (`float`, *optional*, defaults to 0.1):
+            A factor for layer scale.
+    """
+
+    model_type = 'intern_vit_6b'
+
+    def __init__(
+            self,
+            num_channels=3,
+            patch_size=14,
+            image_size=224,
+            qkv_bias=False,
+            hidden_size=3200,
+            num_attention_heads=25,
+            intermediate_size=12800,
+            qk_normalization=True,
+            num_hidden_layers=48,
+            use_flash_attn=True,
+            hidden_act='gelu',
+            norm_type='rms_norm',
+            layer_norm_eps=1e-6,
+            dropout=0.0,
+            drop_path_rate=0.0,
+            attention_dropout=0.0,
+            initializer_range=0.02,
+            initializer_factor=0.1,
+            **kwargs,
+    ):
+        super().__init__(**kwargs)
+
+        self.hidden_size = hidden_size
+        self.intermediate_size = intermediate_size
+        self.dropout = dropout
+        self.drop_path_rate = drop_path_rate
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+        self.num_channels = num_channels
+        self.patch_size = patch_size
+        self.image_size = image_size
+        self.initializer_range = initializer_range
+        self.initializer_factor = initializer_factor
+        self.attention_dropout = attention_dropout
+        self.layer_norm_eps = layer_norm_eps
+        self.hidden_act = hidden_act
+        self.norm_type = norm_type
+        self.qkv_bias = qkv_bias
+        self.qk_normalization = qk_normalization
+        self.use_flash_attn = use_flash_attn
+
+    @classmethod
+    def from_pretrained(cls, pretrained_model_name_or_path: Union[str, os.PathLike], **kwargs) -> 'PretrainedConfig':
+        config_dict, kwargs = cls.get_config_dict(pretrained_model_name_or_path, **kwargs)
+
+        if 'vision_config' in config_dict:
+            config_dict = config_dict['vision_config']
+
+        if 'model_type' in config_dict and hasattr(cls, 'model_type') and config_dict['model_type'] != cls.model_type:
+            logger.warning(
+                f"You are using a model of type {config_dict['model_type']} to instantiate a model of type "
+                f'{cls.model_type}. This is not supported for all configurations of models and can yield errors.'
+            )
+
+        return cls.from_dict(config_dict, **kwargs)

configuration_internlm2.py

@@ -0,0 +1,150 @@
+# Copyright (c) The InternLM team and The HuggingFace Inc. team. All rights reserved.
+#
+# This code is based on transformers/src/transformers/models/llama/configuration_llama.py
+#
+# 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.
+""" InternLM2 model configuration"""
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+
+logger = logging.get_logger(__name__)
+
+INTERNLM2_PRETRAINED_CONFIG_ARCHIVE_MAP = {}
+
+
+# Modified from transformers.model.llama.configuration_llama.LlamaConfig
+class InternLM2Config(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`InternLM2Model`]. It is used to instantiate
+    an InternLM2 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 InternLM2-7B.
+
+    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 32000):
+            Vocabulary size of the InternLM2 model. Defines the number of different tokens that can be represented by the
+            `inputs_ids` passed when calling [`InternLM2Model`]
+        hidden_size (`int`, *optional*, defaults to 4096):
+            Dimension of the hidden representations.
+        intermediate_size (`int`, *optional*, defaults to 11008):
+            Dimension of the MLP representations.
+        num_hidden_layers (`int`, *optional*, defaults to 32):
+            Number of hidden layers in the Transformer encoder.
+        num_attention_heads (`int`, *optional*, defaults to 32):
+            Number of attention heads for each attention layer in the Transformer encoder.
+        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`.
+        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 2048):
+            The maximum sequence length that this model might ever be used with. Typically set this to something large
+            just in case (e.g., 512 or 1024 or 2048).
+        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-12):
+            The epsilon used by the rms normalization layers.
+        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`.
+        tie_word_embeddings(`bool`, *optional*, defaults to `False`):
+            Whether to tie weight embeddings
+        Example:
+
+    """
+    model_type = 'internlm2'
+    _auto_class = 'AutoConfig'
+
+    def __init__(  # pylint: disable=W0102
+        self,
+        vocab_size=103168,
+        hidden_size=4096,
+        intermediate_size=11008,
+        num_hidden_layers=32,
+        num_attention_heads=32,
+        num_key_value_heads=None,
+        hidden_act='silu',
+        max_position_embeddings=2048,
+        initializer_range=0.02,
+        rms_norm_eps=1e-6,
+        use_cache=True,
+        pad_token_id=0,
+        bos_token_id=1,
+        eos_token_id=2,
+        tie_word_embeddings=False,
+        bias=True,
+        rope_theta=10000,
+        rope_scaling=None,
+        attn_implementation='eager',
+        **kwargs,
+    ):
+        self.vocab_size = vocab_size
+        self.max_position_embeddings = max_position_embeddings
+        self.hidden_size = hidden_size
+        self.intermediate_size = intermediate_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+        self.bias = bias
+
+        if num_key_value_heads is None:
+            num_key_value_heads = num_attention_heads
+        self.num_key_value_heads = num_key_value_heads
+
+        self.hidden_act = hidden_act
+        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.attn_implementation = attn_implementation
+        if self.attn_implementation is None:
+            self.attn_implementation = 'eager'
+        super().__init__(
+            pad_token_id=pad_token_id,
+            bos_token_id=bos_token_id,
+            eos_token_id=eos_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) != 2:
+            raise ValueError(
+                '`rope_scaling` must be a dictionary with with two fields, `type` and `factor`, '
+                f'got {self.rope_scaling}'
+            )
+        rope_scaling_type = self.rope_scaling.get('type', None)
+        rope_scaling_factor = self.rope_scaling.get('factor', None)
+        if rope_scaling_type is None or rope_scaling_type not in ['linear', 'dynamic']:
+            raise ValueError(
+                f"`rope_scaling`'s type field must be one of ['linear', 'dynamic'], got {rope_scaling_type}"
+            )
+        if rope_scaling_factor is None or not isinstance(rope_scaling_factor, float) or rope_scaling_factor < 1.0:
+            raise ValueError(f"`rope_scaling`'s factor field must be a float >= 1, got {rope_scaling_factor}")

configuration_internvl_chat.py

@@ -0,0 +1,96 @@
+# --------------------------------------------------------
+# InternVL
+# Copyright (c) 2024 OpenGVLab
+# Licensed under The MIT License [see LICENSE for details]
+# --------------------------------------------------------
+
+import copy
+
+from transformers import AutoConfig, LlamaConfig
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+
+from .configuration_intern_vit import InternVisionConfig
+from .configuration_internlm2 import InternLM2Config
+
+logger = logging.get_logger(__name__)
+
+
+class InternVLChatConfig(PretrainedConfig):
+    model_type = 'internvl_chat'
+    is_composition = True
+
+    def __init__(
+            self,
+            vision_config=None,
+            llm_config=None,
+            use_backbone_lora=0,
+            use_llm_lora=0,
+            select_layer=-1,
+            force_image_size=None,
+            downsample_ratio=0.5,
+            template=None,
+            dynamic_image_size=False,
+            use_thumbnail=False,
+            ps_version='v1',
+            min_dynamic_patch=1,
+            max_dynamic_patch=6,
+            **kwargs):
+        super().__init__(**kwargs)
+
+        if vision_config is None:
+            vision_config = {'architectures': ['InternVisionModel']}
+            logger.info('vision_config is None. Initializing the InternVisionConfig with default values.')
+
+        if llm_config is None:
+            llm_config = {'architectures': ['InternLM2ForCausalLM']}
+            logger.info('llm_config is None. Initializing the LlamaConfig config with default values (`LlamaConfig`).')
+
+        self.vision_config = InternVisionConfig(**vision_config)
+        if llm_config.get('architectures')[0] == 'LlamaForCausalLM':
+            self.llm_config = LlamaConfig(**llm_config)
+        elif llm_config.get('architectures')[0] == 'InternLM2ForCausalLM':
+            self.llm_config = InternLM2Config(**llm_config)
+        else:
+            raise ValueError('Unsupported architecture: {}'.format(llm_config.get('architectures')[0]))
+        self.use_backbone_lora = use_backbone_lora
+        self.use_llm_lora = use_llm_lora
+        self.select_layer = select_layer
+        self.force_image_size = force_image_size
+        self.downsample_ratio = downsample_ratio
+        self.template = template
+        self.dynamic_image_size = dynamic_image_size
+        self.use_thumbnail = use_thumbnail
+        self.ps_version = ps_version  # pixel shuffle version
+        self.min_dynamic_patch = min_dynamic_patch
+        self.max_dynamic_patch = max_dynamic_patch
+
+        logger.info(f'vision_select_layer: {self.select_layer}')
+        logger.info(f'ps_version: {self.ps_version}')
+        logger.info(f'min_dynamic_patch: {self.min_dynamic_patch}')
+        logger.info(f'max_dynamic_patch: {self.max_dynamic_patch}')
+
+    def to_dict(self):
+        """
+        Serializes this instance to a Python dictionary. Override the default [`~PretrainedConfig.to_dict`].
+
+        Returns:
+            `Dict[str, any]`: Dictionary of all the attributes that make up this configuration instance,
+        """
+        output = copy.deepcopy(self.__dict__)
+        output['vision_config'] = self.vision_config.to_dict()
+        output['llm_config'] = self.llm_config.to_dict()
+        output['model_type'] = self.__class__.model_type
+        output['use_backbone_lora'] = self.use_backbone_lora
+        output['use_llm_lora'] = self.use_llm_lora
+        output['select_layer'] = self.select_layer
+        output['force_image_size'] = self.force_image_size
+        output['downsample_ratio'] = self.downsample_ratio
+        output['template'] = self.template
+        output['dynamic_image_size'] = self.dynamic_image_size
+        output['use_thumbnail'] = self.use_thumbnail
+        output['ps_version'] = self.ps_version
+        output['min_dynamic_patch'] = self.min_dynamic_patch
+        output['max_dynamic_patch'] = self.max_dynamic_patch
+
+        return output

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "eos_token_id": [
+    92542,
+    92543
+  ],
+  "max_new_tokens": 64,
+  "pad_token_id": 2,
+  "transformers_version": "4.46.3"
+}

model-00001-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66dc1cdc2bab8daa2932aa98ca3a2b7dfada00a0c809895935a3d6873a7c5ca3
+size 4939944336

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0610aa70b5de4249870cb5e13c0bb9338a06badc909c97a45fca046fdef27d73
+size 4915914584

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c4e7b181874aa08f1714b6f6bdb103b57d4c93395c112d616edcdc9136ccef6
+size 4915914592

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23d1868e75eeca13ac0d740ba1d3154ec55e0493cfac6695e38442be6f2006ff
+size 1379026920

model.safetensors.index.json

@@ -0,0 +1,580 @@
+{
+  "metadata": {
+    "total_size": 16150730752
+  },
+  "weight_map": {
+    "language_model.model.layers.0.attention.wo.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.0.attention.wqkv.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.0.attention_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.0.feed_forward.w1.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.0.feed_forward.w2.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.0.feed_forward.w3.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.0.ffn_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.1.attention.wo.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.1.attention.wqkv.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.1.attention_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.1.feed_forward.w1.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.1.feed_forward.w2.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.1.feed_forward.w3.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.1.ffn_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.10.attention.wo.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.10.attention.wqkv.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.10.attention_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.10.feed_forward.w1.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.10.feed_forward.w2.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.10.feed_forward.w3.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.10.ffn_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.11.attention.wo.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.11.attention.wqkv.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.11.attention_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.11.feed_forward.w1.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.11.feed_forward.w2.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.11.feed_forward.w3.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.11.ffn_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.12.attention.wo.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.12.attention.wqkv.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.12.attention_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.12.feed_forward.w1.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.12.feed_forward.w2.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.12.feed_forward.w3.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.12.ffn_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.13.attention.wo.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.13.attention.wqkv.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.13.attention_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.13.feed_forward.w1.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.13.feed_forward.w2.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.13.feed_forward.w3.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.13.ffn_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.14.attention.wo.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.14.attention.wqkv.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.14.attention_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.14.feed_forward.w1.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.14.feed_forward.w2.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.14.feed_forward.w3.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.14.ffn_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.15.attention.wo.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.15.attention.wqkv.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.15.attention_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.15.feed_forward.w1.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.15.feed_forward.w2.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.15.feed_forward.w3.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.15.ffn_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.16.attention.wo.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.16.attention.wqkv.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.16.attention_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.16.feed_forward.w1.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.16.feed_forward.w2.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.16.feed_forward.w3.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.16.ffn_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.17.attention.wo.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.17.attention.wqkv.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.17.attention_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.17.feed_forward.w1.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.17.feed_forward.w2.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.17.feed_forward.w3.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.17.ffn_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.18.attention.wo.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.18.attention.wqkv.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.18.attention_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.18.feed_forward.w1.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.18.feed_forward.w2.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.18.feed_forward.w3.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.18.ffn_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.19.attention.wo.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.19.attention.wqkv.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.19.attention_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.19.feed_forward.w1.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.19.feed_forward.w2.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.19.feed_forward.w3.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.19.ffn_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.2.attention.wo.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.2.attention.wqkv.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.2.attention_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.2.feed_forward.w1.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.2.feed_forward.w2.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.2.feed_forward.w3.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.2.ffn_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.20.attention.wo.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.20.attention.wqkv.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.20.attention_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.20.feed_forward.w1.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.20.feed_forward.w2.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.20.feed_forward.w3.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.20.ffn_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.21.attention.wo.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.21.attention.wqkv.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.21.attention_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.21.feed_forward.w1.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.21.feed_forward.w2.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.21.feed_forward.w3.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.21.ffn_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.22.attention.wo.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.22.attention.wqkv.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.22.attention_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.22.feed_forward.w1.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.22.feed_forward.w2.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.22.feed_forward.w3.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.22.ffn_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.23.attention.wo.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.23.attention.wqkv.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.23.attention_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.23.feed_forward.w1.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.23.feed_forward.w2.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.23.feed_forward.w3.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.23.ffn_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.24.attention.wo.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.24.attention.wqkv.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.24.attention_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.24.feed_forward.w1.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.24.feed_forward.w2.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.24.feed_forward.w3.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.24.ffn_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.25.attention.wo.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.25.attention.wqkv.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.25.attention_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.25.feed_forward.w1.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.25.feed_forward.w2.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.25.feed_forward.w3.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.25.ffn_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.26.attention.wo.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.26.attention.wqkv.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.26.attention_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.26.feed_forward.w1.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.26.feed_forward.w2.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.26.feed_forward.w3.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.26.ffn_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.27.attention.wo.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.27.attention.wqkv.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.27.attention_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.27.feed_forward.w1.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.27.feed_forward.w2.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.27.feed_forward.w3.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.27.ffn_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.28.attention.wo.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.28.attention.wqkv.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.28.attention_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.28.feed_forward.w1.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.28.feed_forward.w2.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.28.feed_forward.w3.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.28.ffn_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.29.attention.wo.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.29.attention.wqkv.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.29.attention_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.29.feed_forward.w1.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.29.feed_forward.w2.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.29.feed_forward.w3.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.29.ffn_norm.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.3.attention.wo.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.3.attention.wqkv.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.3.attention_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.3.feed_forward.w1.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.3.feed_forward.w2.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.3.feed_forward.w3.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.3.ffn_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.30.attention.wo.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.30.attention.wqkv.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.30.attention_norm.weight": "model-00004-of-00004.safetensors",
+    "language_model.model.layers.30.feed_forward.w1.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.30.feed_forward.w2.weight": "model-00004-of-00004.safetensors",
+    "language_model.model.layers.30.feed_forward.w3.weight": "model-00003-of-00004.safetensors",
+    "language_model.model.layers.30.ffn_norm.weight": "model-00004-of-00004.safetensors",
+    "language_model.model.layers.31.attention.wo.weight": "model-00004-of-00004.safetensors",
+    "language_model.model.layers.31.attention.wqkv.weight": "model-00004-of-00004.safetensors",
+    "language_model.model.layers.31.attention_norm.weight": "model-00004-of-00004.safetensors",
+    "language_model.model.layers.31.feed_forward.w1.weight": "model-00004-of-00004.safetensors",
+    "language_model.model.layers.31.feed_forward.w2.weight": "model-00004-of-00004.safetensors",
+    "language_model.model.layers.31.feed_forward.w3.weight": "model-00004-of-00004.safetensors",
+    "language_model.model.layers.31.ffn_norm.weight": "model-00004-of-00004.safetensors",
+    "language_model.model.layers.4.attention.wo.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.4.attention.wqkv.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.4.attention_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.4.feed_forward.w1.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.4.feed_forward.w2.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.4.feed_forward.w3.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.4.ffn_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.5.attention.wo.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.5.attention.wqkv.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.5.attention_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.5.feed_forward.w1.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.5.feed_forward.w2.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.5.feed_forward.w3.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.5.ffn_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.6.attention.wo.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.6.attention.wqkv.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.6.attention_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.6.feed_forward.w1.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.6.feed_forward.w2.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.6.feed_forward.w3.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.6.ffn_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.7.attention.wo.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.7.attention.wqkv.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.7.attention_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.7.feed_forward.w1.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.7.feed_forward.w2.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.7.feed_forward.w3.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.7.ffn_norm.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.8.attention.wo.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.8.attention.wqkv.weight": "model-00001-of-00004.safetensors",
+    "language_model.model.layers.8.attention_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.8.feed_forward.w1.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.8.feed_forward.w2.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.8.feed_forward.w3.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.8.ffn_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.9.attention.wo.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.9.attention.wqkv.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.9.attention_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.9.feed_forward.w1.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.9.feed_forward.w2.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.9.feed_forward.w3.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.layers.9.ffn_norm.weight": "model-00002-of-00004.safetensors",
+    "language_model.model.norm.weight": "model-00004-of-00004.safetensors",
+    "language_model.model.tok_embeddings.weight": "model-00001-of-00004.safetensors",
+    "language_model.output.weight": "model-00004-of-00004.safetensors",
+    "mlp1.0.bias": "model-00004-of-00004.safetensors",
+    "mlp1.0.weight": "model-00004-of-00004.safetensors",
+    "mlp1.1.bias": "model-00004-of-00004.safetensors",
+    "mlp1.1.weight": "model-00004-of-00004.safetensors",
+    "mlp1.3.bias": "model-00004-of-00004.safetensors",
+    "mlp1.3.weight": "model-00004-of-00004.safetensors",
+    "vision_model.embeddings.class_embedding": "model-00001-of-00004.safetensors",
+    "vision_model.embeddings.patch_embedding.bias": "model-00001-of-00004.safetensors",
+    "vision_model.embeddings.patch_embedding.weight": "model-00001-of-00004.safetensors",
+    "vision_model.embeddings.position_embedding": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.0.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.1.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.10.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.11.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.12.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.13.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.14.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.15.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.16.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.17.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.18.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.19.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.2.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.20.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.21.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.22.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.23.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.3.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.4.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.5.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.6.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.7.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.8.norm2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.attn.proj.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.attn.proj.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.attn.qkv.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.attn.qkv.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.ls1": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.ls2": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.mlp.fc1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.mlp.fc1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.mlp.fc2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.mlp.fc2.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.norm1.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.norm1.weight": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.norm2.bias": "model-00001-of-00004.safetensors",
+    "vision_model.encoder.layers.9.norm2.weight": "model-00001-of-00004.safetensors"
+  }
+}