Add files using upload-large-folder tool

vlm-siglip2-sllm_210/opt_step-10000/resume_run_infos.json

@@ -0,0 +1,91 @@
+{
+  "train_logs": {
+    "lr": 0.0001,
+    "num_opt_steps": 10000,
+    "num_epochs": 0,
+    "per_token_loss": {
+      "sft": 0.9180383384227753,
+      "all": 0.9180383384227753
+    },
+    "z_loss": {
+      "sft": 0.0,
+      "all": 0.0
+    },
+    "watt/s": {
+      "sft": 276.72318082969514,
+      "all": 276.72318082969514
+    },
+    "tflops": {
+      "sft": 9.964916582778226,
+      "all": 9.964916582778226
+    },
+    "tflop_counter": {
+      "sft": 9138625.206542969,
+      "all": 9138625.206542969
+    },
+    "fwd_bwd_time": {
+      "sft": 771777.3967046738,
+      "all": 771777.3967046738
+    },
+    "tflops_acc": {
+      "sft": 11.841011728981654,
+      "all": 11.841011728981654
+    },
+    "num_per_device_batches": {
+      "sft": 640000,
+      "all": 640000
+    },
+    "num_images": {
+      "sft": 36943306,
+      "all": 36943306
+    },
+    "num_image_tokens": {
+      "sft": 2364371584,
+      "all": 2364371584
+    },
+    "num_tokens": {
+      "sft": 1177031055,
+      "all": 1177031055
+    },
+    "image_to_text_ratio": {
+      "sft": 0.04172288626432419,
+      "all": 0.04172288626432419
+    },
+    "pixel_values_sum": {
+      "sft": 9614803855744.0,
+      "all": 9614803855744.0
+    },
+    "num_padding": {
+      "sft": 639181809,
+      "all": 639181809
+    },
+    "num_per_device_batches_in_curr_epoch": {
+      "sft": 640000,
+      "all": 640000
+    },
+    "num_batches": {
+      "all": 640000
+    },
+    "num_batches_in_curr_epoch": {
+      "all": 640000
+    },
+    "per_token_loss_acc": {},
+    "z_loss_acc": {},
+    "num_batches_since_training_logged": {},
+    "num_per_device_batches_since_training_logged": {},
+    "tflop_counter_since_training_logged": {},
+    "total_energy_delta_since_training_logged": {},
+    "fwd_bwd_time_since_training_logged": {},
+    "global_batch_size_current": 256
+  },
+  "wandb_run_id": "hp0rcdxj",
+  "seed": 42,
+  "resume_opt_step": 10000,
+  "resume_epoch": 0,
+  "gbs_running": {
+    "global_seen_samples": 2560000,
+    "global_batch_size_current": 256,
+    "next_goal_samples": 0,
+    "grad_acc_size_current": 4
+  }
+}

vlm-siglip2-sllm_210/opt_step-10000__merged/chat_template.jinja

@@ -0,0 +1,2 @@
+<|begin_of_text|>{% for message in messages %}{{message['role'] | capitalize}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>
+{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}

vlm-siglip2-sllm_210/opt_step-10000__merged/chat_template.json

@@ -0,0 +1,3 @@
+{
+    "chat_template": "<|begin_of_text|>{% for message in messages %}{{message['role'] | capitalize}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>\n{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}"
+}

vlm-siglip2-sllm_210/opt_step-10000__merged/config.json

@@ -0,0 +1,51 @@
+{
+  "additional_vocab_size": 40,
+  "architectures": [
+    "VLlamaForCausalLM"
+  ],
+  "auto_map": {
+    "AutoConfig": "configuration_vllama.VLlamaConfig",
+    "AutoModel": "modeling_vllama.VLlamaModel",
+    "AutoModelForCausalLM": "modeling_vllama.VLlamaForCausalLM",
+    "AutoModelForVision2Seq": "modeling_vllama.VLlamaForVision2Seq"
+  },
+  "freeze_config": {
+    "freeze_lm_head": true,
+    "freeze_text_layers": true,
+    "freeze_vision_layers": true
+  },
+  "hidden_size": 768,
+  "image_token_id": 128295,
+  "initializer_range": 0.02,
+  "max_position_embeddings": 8192,
+  "model_type": "VLlama",
+  "neftune_noise_alpha": 0.0,
+  "output_attentions": false,
+  "pixel_shuffle_factor": 4,
+  "qk_layer_norms": false,
+  "scale_factor": 4,
+  "text_config": {
+    "hidden_size": 768,
+    "intermediate_size": 3072,
+    "mlp_bias": false,
+    "model_type": "VLlama",
+    "num_hidden_layers": 12,
+    "text_model_name": "SmolVEncoder/decoder-210m",
+    "vocab_size": 128256
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": null,
+  "use_cache": true,
+  "use_resampler": false,
+  "vision_config": {
+    "embed_dim": 768,
+    "image_size": 512,
+    "intermediate_size": 3072,
+    "model_type": "VLlama",
+    "num_hidden_layers": 12,
+    "patch_size": 16,
+    "vision_model_name": "google/siglip2-base-patch16-512"
+  },
+  "vocab_size": 128256
+}

vlm-siglip2-sllm_210/opt_step-10000__merged/configuration_vllama.py

@@ -0,0 +1,233 @@
+import copy
+import os
+
+from typing import Union, Any, Dict
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+from transformers import AutoConfig
+
+logger = logging.get_logger(__name__)
+
+def collect_arg_in_candidates(config, candidates, default = None) -> Any:
+    """ Gets the argument in a config given a list of candidates """
+    for c in candidates:
+        if hasattr(config, c):
+            return getattr(config, c)
+        elif c in config:
+            return config[c]
+    if default is not None:
+        return default
+    raise ValueError("No matching arguments found in candidates. Candidates: {}, Config: {}".format(candidates, config))
+
+class VLlamaTextConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`LlamaModel`]. It is used to instantiate an LLaMA
+    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 LLaMA-7B.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        embed_dim (`int`, *optional*, defaults to 1152):
+            Dimensionality of the encoder layers and the pooler layer. (elsewhere referred to as `embed_dim`)
+        image_size (`int`, *optional*, defaults to 384):
+            The size (resolution) of each image.
+    """
+    model_type = "VLlama"
+
+    def __init__(
+        self,
+        # Case for when vllama3 is from the hub with no vision_model_name
+        text_model_name="HuggingFaceTB/SmolLM2-135M-Instruct",
+        **kwargs,
+    ):
+        self.text_model_name = text_model_name
+        text_config = AutoConfig.from_pretrained(text_model_name, trust_remote_code=True)
+        if hasattr(text_config, "text_config"):
+            text_config = text_config.text_config
+        
+        self.hidden_size = collect_arg_in_candidates(text_config, ["hidden_size", "embed_dim"])
+        self.num_hidden_layers = collect_arg_in_candidates(text_config, ["num_hidden_layers", "num_hidden_blocks"])
+        self.intermediate_size = collect_arg_in_candidates(text_config, ["intermediate_size", "mlp_dim"])
+        self.mlp_bias = collect_arg_in_candidates(text_config, ["mlp_bias", "mlp_hidden_bias"], default = False)
+        self.vocab_size = collect_arg_in_candidates(text_config, ["vocab_size"])
+    
+        super().__init__(text_model_name=text_model_name, **kwargs)
+
+class VLlamaVisionConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`LlamaModel`]. It is used to instantiate an LLaMA
+    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 LLaMA-7B.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        embed_dim (`int`, *optional*, defaults to 1152):
+            Dimensionality of the encoder layers and the pooler layer. (elsewhere referred to as `embed_dim`)
+        image_size (`int`, *optional*, defaults to 384):
+            The size (resolution) of each image.
+    """
+    model_type = "VLlama"
+    attribute_map = {
+        "hidden_size": "embed_dim",
+    }
+
+    def __init__(
+        self,
+        # Case for when vllama3 is from the hub with no vision_model_name
+        vision_model_name="google/siglip2-base-patch16-512",
+        **kwargs,
+    ):
+        self.vision_model_name = vision_model_name
+        vision_config = AutoConfig.from_pretrained(vision_model_name, trust_remote_code=True)
+        if hasattr(vision_config, "vision_config"):
+            vision_config = vision_config.vision_config
+
+        self.embed_dim = collect_arg_in_candidates(vision_config, ["embed_dim", "hidden_size"])
+        self.image_size = collect_arg_in_candidates(vision_config, ["image_size", "img_size"])
+        self.patch_size = collect_arg_in_candidates(vision_config, ["patch_size"])
+        self.num_hidden_layers = collect_arg_in_candidates(vision_config, ["num_hidden_layers", "num_hidden_blocks"])
+        self.intermediate_size = collect_arg_in_candidates(vision_config, ["intermediate_size", "mlp_dim"])
+
+        super().__init__(vision_model_name=vision_model_name, **kwargs)
+
+class VLlamaConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`SmolVLMModel`]. It is used to instantiate a
+    SmolVLM 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 model of the SmolVLM
+    [HuggingFaceTB/SmolVLM2-2.2B-Instruct](https://huggingface.co/HuggingFaceTB/SmolVLM2-2.2B-Instruct) architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should cache the key/value pairs of the attention mechanism. Only
+            relevant if `config.is_decoder=True`.
+        image_token_id (`int`, *optional*, defaults to 128257):
+            The id of the "image" token.
+        tie_word_embeddings (`bool`, *optional*, defaults to `False`):
+            Whether or not to tie the word embeddings with the token embeddings.
+        vision_config (`IdeficsVisionConfig` or `dict`, *optional*, defaults to `IdeficsVisionConfig`):
+            Custom vision config or dict for the vision tower
+        text_config (`PretrainedConfig` or `dict`, *optional*, defaults to `LlamaConfig`):
+            Custom text config or dict for the text model
+        scale_factor (`int`, *optional*, defaults to 2):
+            The scale factor for the image encoder.
+        pad_token_id (`int`, *optional*, defaults to 128002):
+            The id of the padding token.
+
+    Example:
+    ```python
+    >>> from transformers import SmolVLMModel, SmolVLMConfig
+    >>> # Initializing configuration
+    >>> configuration = SmolVLMConfig()
+    >>> # Initializing a model from the configuration
+    >>> model = SmolVLMModel(configuration)
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "VLlama"
+    is_composition = True
+    # sub_configs = {"text_config": VLlamaTextConfig, "vision_config": VLlamaVisionConfig}
+
+    DEFAULT_TEXT_MODEL_NAME = "EuroBERT/EuroBERT-210m"
+    DEFAULT_VISION_MODEL_NAME = "google/siglip2-base-patch16-512"
+
+    def __init__(
+        self,
+        text_config: Union[PretrainedConfig, Dict[str, Any]] = None,
+        vision_config: Union[PretrainedConfig, Dict[str, Any]] = None,
+        image_token_id: int = 128_257,
+        vocab_size=128_256,
+        use_cache = True,
+        tie_word_embeddings = False,
+        freeze_config = None,
+        pad_token_id = None,
+        initializer_range = 0.02,
+        pixel_shuffle_factor = 4,
+        use_resampler = False,
+        additional_vocab_size = 0,
+        neftune_noise_alpha = 0.0,
+        **kwargs,
+    ):        
+        self.image_token_id = image_token_id
+        self.use_cache = use_cache
+        self.tie_word_embeddings = tie_word_embeddings
+        self.scale_factor = pixel_shuffle_factor
+        self.additional_vocab_size = additional_vocab_size
+
+        if text_config is None:
+            text_config = AutoConfig.from_pretrained(self.DEFAULT_TEXT_MODEL_NAME, trust_remote_code=True)
+        elif isinstance(text_config, dict):
+            text_config = VLlamaTextConfig(text_config["text_model_name"])
+        self.text_config = text_config
+        
+        if vision_config is None:
+            vision_config = AutoConfig.from_pretrained(self.DEFAULT_VISION_MODEL_NAME, trust_remote_code=True)
+        elif isinstance(vision_config, dict):
+            vision_config = VLlamaVisionConfig(vision_config["vision_model_name"])
+        self.vision_config = vision_config
+
+        self.freeze_config = freeze_config
+
+        # Pixel shuffle factor
+        self.pixel_shuffle_factor = pixel_shuffle_factor
+        self.use_resampler = use_resampler
+
+        self.neftune_noise_alpha = neftune_noise_alpha
+
+        self.initializer_range = initializer_range
+
+        hidden_size = kwargs.pop("hidden_size", self.text_config.hidden_size)
+
+        super().__init__(
+            **kwargs, 
+            pad_token_id=pad_token_id, 
+            tie_word_embeddings=tie_word_embeddings,
+            vocab_size=vocab_size,
+            hidden_size=hidden_size,
+        )
+
+    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["model_type"] = self.__class__.model_type
+        output["vision_config"] = self.vision_config.to_dict()
+        output["text_config"] = self.text_config.to_dict()
+        # output["freeze_config"] = self.freeze_config.to_dict()
+
+        return output
+    
+    # @classmethod
+    # def from_pretrained(cls, pretrained_model_name_or_path: Union[str, os.PathLike], **kwargs) -> "PretrainedConfig":
+    #     outputs = super(VLlamaConfig, cls).from_pretrained(pretrained_model_name_or_path, **kwargs)
+    #     return outputs
+
+    @classmethod
+    def from_pretrained_models(
+        cls, 
+        text_model_name: Union[str, os.PathLike],
+        vision_model_name: Union[str, os.PathLike],
+        **kwargs
+    ) -> "PretrainedConfig":
+        # text_model_config = AutoConfig.from_pretrained(text_model_name, trust_remote_code=True)
+        # vision_model_config = AutoConfig.from_pretrained(vision_model_name, trust_remote_code=True)
+        text_model_config = VLlamaTextConfig(text_model_name)
+        vision_model_config = VLlamaVisionConfig(vision_model_name)
+        return cls(
+            text_config=text_model_config,
+            vision_config=vision_model_config,
+            **kwargs
+        )

vlm-siglip2-sllm_210/opt_step-10000__merged/modeling_vllama.py

@@ -0,0 +1,895 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from torch.nn import CrossEntropyLoss
+from typing import Optional, Tuple, Union, List
+
+# from transformers.models.smolvlm import SmolVLMModel, SmolVLMPreTrainedModel
+
+from .configuration_vllama import VLlamaConfig
+
+from transformers import AutoModel, AutoConfig, AutoModelForMaskedLM, GenerationMixin
+from transformers.cache_utils import Cache
+from transformers.modeling_utils import PreTrainedModel
+from transformers.modeling_outputs import BaseModelOutput
+from transformers.models.bert.modeling_bert import BaseModelOutputWithPoolingAndCrossAttentions, MaskedLMOutput
+from transformers.models.idefics3.modeling_idefics3 import Idefics3VisionTransformer
+from transformers.processing_utils import Unpack
+from transformers.modeling_flash_attention_utils import FlashAttentionKwargs
+from transformers.utils import LossKwargs
+
+from typing import List, Optional, Tuple, Union
+
+import torch
+import torch.utils.checkpoint
+
+from dataclasses import dataclass
+
+from transformers import logging 
+from transformers.utils import ContextManagers
+
+logger = logging.get_logger(__name__)
+
+class KwargsForCausalLM(FlashAttentionKwargs, LossKwargs): ...
+
+class DecoupledEmbedding(nn.Embedding):
+    # Derived from https://pytorch.org/docs/stable/_modules/torch/nn/modules/sparse.html#Embedding
+    """
+    Implements a decoupling of parameters to allow freezing (or not) a subset of the embeddings.
+    In practise, the regular `weight` can be trained or frozen (i.e. `partially_freeze=True`), and if `num_additional_embeddings` > 0, then it will create `num_additional_embeddings` additional parameters that are always trained.
+    If `num_additional_embeddings=0`, then the module defaults back to the regular behavior of `nn.Embedding`.
+    """
+
+    def __init__(
+        self,
+        num_embeddings,
+        num_additional_embeddings,
+        embedding_dim,
+        partially_freeze=False,
+        device=None,
+        dtype=None,
+        padding_idx=None,
+        **kwargs,
+    ) -> None:
+        """
+        num_additional_embeddings: int. Number of additional embeddings. Only useful when you `partially_freeze=True`.
+        partially_freeze: bool. If True, the regular `weight` will be frozen. `additional_weight` is never frozen.
+        Note: there are a lot of other parameters to initialize a standard `nn.Embedding` such as `padding_idx`, `max_norm` or `norm_type`. We are not supporting these.
+        """
+        if padding_idx is not None and padding_idx > num_embeddings:
+            raise ValueError(f"padding_idx must be within num_embeddings. Got {padding_idx} and {num_embeddings}")
+        super().__init__(
+            num_embeddings=num_embeddings,
+            embedding_dim=embedding_dim,
+            device=device,
+            dtype=dtype,
+            padding_idx=padding_idx,
+            **kwargs,
+        )
+        self.num_embeddings = num_embeddings
+        self.padding_idx = padding_idx
+        self.num_additional_embeddings = num_additional_embeddings
+        self.partially_freeze = partially_freeze
+
+        if partially_freeze:
+            self.weight.requires_grad_(False)
+
+        if self.num_additional_embeddings > 0:
+            self.additional_embedding = nn.Embedding(
+                num_embeddings=self.num_additional_embeddings,
+                embedding_dim=embedding_dim,
+                device=device,
+                dtype=dtype,
+            )
+
+    def forward(self, input_ids):
+        """
+        we have 2 embeddings, with different indices - one pretrained self.weight and another
+        self.additional_embedding.weight that is being trained.
+        in order to make a lookup of the input ids, we:
+        1. find out the indices of the entries belonging to the 2nd embedding
+        2. extract those values while subtracting the size of the first embedding (num_embeddings),
+           since the 2nd embedding starts from 0 and not num_embeddings
+        3. perform the 2nd embedding lookup
+        4. now we handle the 1st embedding, we overwrite indices belonging to the 2nd embedding with a padding index
+        5. perform the 1st embedding lookup
+        6. now we overwrite the values in the 1st embedding lookup with the values of the 2nd embedding lookup
+        note: for the 1st embedding lookup we could have looked up only the low indices and not do
+        the padding, but then we have to create a new tensor and populate it with 2 tensors that are
+        spread out across various indices - i.e. not a simple concat - I haven't benchmarked the
+        complex case if it's any faster, given that seqlens are usually relatively short it's
+        probably not faster or if faster not by much - but might be a good idea to measure.
+        """
+        if self.num_additional_embeddings == 0:
+            return self.additional_embedding(input_ids)
+
+        # Clone so that we don't modify the original input_ids later on
+        input_ids = input_ids.clone()
+        additional_vocab_indices = torch.where(input_ids >= self.num_embeddings)
+        input_ids_additional_vocab = input_ids[additional_vocab_indices]
+        additional_embeddings = self.additional_embedding(input_ids_additional_vocab - self.num_embeddings)
+
+        # for successful lookup replace input_ids with 0, the results of these will be discarded anyway
+        input_ids[additional_vocab_indices] = 0
+        full_vector = F.embedding(input_ids, self.weight)
+
+        # overwrite the records with high indices
+        full_vector[additional_vocab_indices] = additional_embeddings
+
+        return full_vector
+
+    def extra_repr(self) -> str:
+        return "num_embeddings={}, num_additional_embeddings={}, embedding_dim={}, partially_freeze={}".format(
+            self.num_embeddings,
+            self.num_additional_embeddings,
+            self.embedding_dim,
+            self.partially_freeze,
+        )
+
+@dataclass
+class VLlamaBaseModelOutputWithPast(BaseModelOutput):
+    """
+    Base class for VLlama3 model's outputs that may also contain a past key/values (to speed up sequential decoding).
+    Args:
+        last_hidden_state (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
+            Sequence of hidden-states at the output of the last layer of the model.
+            If `past_key_values` is used only the last hidden-state of the sequences of shape `(batch_size, 1,
+            hidden_size)` is output.
+        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+            Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
+            `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and optionally if
+            `config.is_encoder_decoder=True` 2 additional tensors of shape `(batch_size, num_heads,
+            encoder_sequence_length, embed_size_per_head)`.
+            Contains pre-computed hidden-states (key and values in the self-attention blocks and optionally if
+            `config.is_encoder_decoder=True` in the cross-attention blocks) that can be used (see `past_key_values`
+            input) to speed up sequential decoding.
+        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
+            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.
+            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
+        attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
+            Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
+            sequence_length)`.
+            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
+            heads.
+        image_hidden_states (`tuple(torch.FloatTensor)`, *optional*):
+            Tuple of `torch.FloatTensor` (one for the output of the image embeddings, `(batch_size, num_images,
+            sequence_length, hidden_size)`.
+            image_hidden_states of the model produced by the vision encoder, and optionally by the perceiver
+    """
+
+    last_hidden_state: torch.FloatTensor = None
+    past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None
+    hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None
+    attentions: Optional[Tuple[torch.FloatTensor, ...]] = None
+    image_hidden_states: Optional[Tuple[torch.FloatTensor]] = None
+
+@dataclass
+class VLlamaCausalLMOutputWithPast(BaseModelOutput):
+    """
+    Base class for VLlama3 causal language model (or autoregressive) outputs.
+    Args:
+        loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
+            Language modeling loss (for next-token prediction).
+        logits (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`):
+            Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
+        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+            Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
+            `(batch_size, num_heads, sequence_length, embed_size_per_head)`)
+            Contains pre-computed hidden-states (key and values in the self-attention blocks) that can be used (see
+            `past_key_values` input) to speed up sequential decoding.
+        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
+            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.
+            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
+        attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
+            Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
+            sequence_length)`.
+            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
+            heads.
+        image_hidden_states (`tuple(torch.FloatTensor)`, *optional*):
+            Tuple of `torch.FloatTensor` (one for the output of the image embeddings, `(batch_size, num_images,
+            sequence_length, hidden_size)`.
+            image_hidden_states of the model produced by the vision encoder, and optionally by the perceiver
+    """
+
+    loss: Optional[torch.FloatTensor] = None
+    logits: torch.FloatTensor = None
+    past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None
+    hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None
+    attentions: Optional[Tuple[torch.FloatTensor, ...]] = None
+    image_hidden_states: Optional[Tuple[torch.FloatTensor]] = None
+
+
+class VLlamaSimpleMLP(nn.Module):
+    def __init__(self, input_size, output_size):
+        super().__init__()
+        self.proj = nn.Linear(input_size, output_size, bias=False)
+
+    def forward(self, x):
+        return self.proj(x)
+
+class VLlamaConnector(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.scale_factor = config.pixel_shuffle_factor
+        self.modality_projection = VLlamaSimpleMLP(
+            input_size=config.vision_config.hidden_size * (config.scale_factor**2),
+            output_size=config.text_config.hidden_size
+        )
+
+    def pixel_shuffle(self, x, scale_factor):
+        bsz, seq, embed_dim = x.size()
+        height = width = int(seq**0.5)
+        x = x.view(bsz, height, width, embed_dim)
+        x = x.view(bsz, height, int(width / scale_factor), embed_dim * scale_factor)
+        x = x.permute(0, 2, 1, 3)
+        x = x.reshape(bsz, int(width / scale_factor), int(height / scale_factor), embed_dim * (scale_factor**2))
+        x = x.permute(0, 2, 1, 3)
+        x = x.reshape(bsz, int(seq / (scale_factor**2)), embed_dim * (scale_factor**2))
+        return x
+
+    def forward(self, image_hidden_states):
+        image_hidden_states = self.pixel_shuffle(image_hidden_states, self.scale_factor)
+        image_hidden_states = self.modality_projection(image_hidden_states)
+        return image_hidden_states
+
+class VLlamaPreTrainedModel(PreTrainedModel):
+    config_class = VLlamaConfig
+    base_model_prefix = "model"
+    supports_gradient_checkpointing = True
+    _no_split_modules = ["VLlamaDecoderLayer"]
+    _skip_keys_device_placement = "past_key_values"
+    _supports_flash_attn_2 = True
+    _supports_sdpa = True
+    _supports_cache_class = True
+
+    def _init_weights(self, module):
+        """Initialize the weights."""
+
+        std = (
+            self.config.initializer_range
+            if hasattr(self.config, "initializer_range")
+            else self.config.text_config.initializer_range
+        )
+
+        if hasattr(module, "class_embedding"):
+            module.class_embedding.data.normal_(mean=0.0, std=std)
+
+        if isinstance(module, (nn.Linear, nn.Conv2d)):
+            module.weight.data.normal_(mean=0.0, std=std)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.Embedding):
+            module.weight.data.normal_(mean=0.0, std=std)
+            if module.padding_idx is not None:
+                module.weight.data[module.padding_idx].zero_()
+
+class VLlamaModel(VLlamaPreTrainedModel):
+    """
+    A subclass of Idefics3Model. We do *not* remove or block the call to inputs_merger
+    in forward. Instead, we override inputs_merger here with custom logic.
+    """
+
+    def __init__(self, config: VLlamaConfig, **kwargs):
+        super().__init__(config)
+
+        self.vision_model = VLlamaModel.init_vision_model(config, **kwargs)
+        self.connector = VLlamaConnector(config)
+        self.text_model = VLlamaModel.init_language_model(config, **kwargs)
+
+        self.image_seq_len = int(
+            ((config.vision_config.image_size // config.vision_config.patch_size) ** 2) / (config.scale_factor**2)
+        )
+        self.image_token_id = self.config.image_token_id
+
+        self.post_init()
+
+    @staticmethod
+    def init_vision_model(config: VLlamaConfig, **kwargs):
+        vision_model_config = AutoConfig.from_pretrained(
+            config.vision_config.vision_model_name,
+            trust_remote_code=True,
+            **kwargs,
+        )
+
+        vision_model = AutoModel.from_config(vision_model_config, trust_remote_code=True, **kwargs)
+
+        if hasattr(vision_model, "vision_model"):
+            # If the model has a vision_model attribute, it means it's a wrapper around another model
+            vision_model = vision_model.vision_model
+
+        return vision_model
+
+    @staticmethod
+    def init_language_model(config: VLlamaConfig, **kwargs):
+        text_model_config = AutoConfig.from_pretrained(
+            config.text_config.text_model_name,
+            trust_remote_code=True,
+            **kwargs,
+        )
+
+        text_model = AutoModel.from_config(text_model_config, trust_remote_code=True, **kwargs)
+        # extractor = regex_lookup(language_model_name, language_model_name2model)
+
+        embed_layer = DecoupledEmbedding(
+            num_embeddings=text_model_config.vocab_size,
+            num_additional_embeddings=config.additional_vocab_size,
+            embedding_dim=config.hidden_size,
+            partially_freeze=config.freeze_config["freeze_text_layers"],
+            padding_idx=config.pad_token_id,
+        )
+
+        text_model.set_input_embeddings(embed_layer)
+
+        return text_model
+
+    def enable_input_require_grads(self):
+        """
+        Enables the gradients for the input embeddings.
+        This is useful for lora when using gradient checkpointing.
+        c.f. https://github.com/huggingface/peft/issues/1402#issuecomment-1913675032
+        Override to set output.requires_grad = True for both the decoder's and vision model's embeddings.
+        """
+
+        def get_lowest_module(module):
+            if len(list(module.children())) == 0:
+                # If the module has no children, it is a leaf module (e.g., Linear, Conv2d, etc.)
+                return module
+            else:
+                # Recursively call the function on each child module
+                return get_lowest_module(list(module.children())[0])
+
+        def make_inputs_require_grads(module, input, output):
+            output.requires_grad_(True)
+
+        self._text_require_grads_hook = self.get_input_embeddings().register_forward_hook(make_inputs_require_grads)
+        self._vision_require_grads_hook = get_lowest_module(self.vision_model).register_forward_hook(
+            make_inputs_require_grads
+        )
+
+    def disable_input_require_grads(self):
+        self._text_require_grads_hook.remove()
+        self._vision_require_grads_hook.remove()
+
+    def get_input_embeddings(self):
+        return self.text_model.get_input_embeddings()
+
+    def set_input_embeddings(self, value):
+        self.text_model.set_input_embeddings(value)
+
+    def inputs_merger(
+        self, input_ids: torch.LongTensor, inputs_embeds: torch.Tensor, image_hidden_states: torch.Tensor
+    ):
+        """
+        This method aims at merging the token embeddings with the image hidden states into one single sequence of vectors that are fed to the transformer LM.
+        The merging happens as follows:
+        - The text token sequence is: `tok_1 tok_2 tok_3 <fake_token_around_image> <image> <image> ... <image> <fake_token_around_image> tok_4`.
+        - We get the image hidden states for the image through the vision encoder and that hidden state, after a pixel shuffle operation, is then projected into the text embedding space.
+        We thus have a sequence of image hidden states of size (1, image_seq_len, hidden_dim), where 1 is for batch_size of 1 image and hidden_dim is the hidden_dim of the LM transformer.
+        - The merging happens so that we obtain the following sequence: `vector_tok_1 vector_tok_2 vector_tok_3 vector_fake_tok_around_image {sequence of image_seq_len image hidden states} vector_fake_toke_around_image vector_tok_4`. That sequence is fed to the LM.
+        - To fit the format of that sequence, `input_ids`, `input_embeds`, `attention_mask` are all 3 adapted to insert the image hidden states.
+        """
+        _, patch_size, _ = image_hidden_states.shape
+
+        image_mask = input_ids == self.image_token_id
+        num_image_tokens = image_mask.sum(dim=1)
+        if not torch.all(num_image_tokens % patch_size == 0):
+            raise ValueError("At least one sample has <image> tokens not divisible by patch_size.")
+
+        blocks_per_sample = num_image_tokens // patch_size
+
+        offsets = torch.nn.functional.pad(blocks_per_sample.cumsum(dim=0), (1, 0), value=0)
+        block_offset = offsets[:-1]
+        row_cum = image_mask.cumsum(dim=-1)
+        chunk_idx = (row_cum - 1) // patch_size
+        local_idx = (row_cum - 1) % patch_size
+        block_idx = block_offset.unsqueeze(1) + chunk_idx
+
+        image_embeds = torch.zeros_like(inputs_embeds)
+        image_embeds[image_mask] = image_hidden_states[block_idx[image_mask], local_idx[image_mask], :]
+
+        merged_embeds = torch.where(image_mask.unsqueeze(-1), image_embeds, inputs_embeds)
+        return merged_embeds
+    
+    def embed_tokens(self, input_ids: torch.LongTensor) -> torch.FloatTensor:
+        """
+        Override the embed_tokens method to use the text model's input embeddings.
+        This is necessary to ensure that the image token ID is correctly handled.
+        """
+        if self.text_model.get_input_embeddings() is None:
+            raise ValueError("The text model does not have input embeddings.")
+        
+        return self.text_model.get_input_embeddings()(input_ids).to(input_ids.device)
+
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[List[torch.FloatTensor]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        pixel_values: Optional[torch.FloatTensor] = None,
+        pixel_attention_mask: Optional[torch.BoolTensor] = None,
+        image_hidden_states: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        cache_position: Optional[torch.LongTensor] = None,
+    ) -> Union[Tuple, VLlamaBaseModelOutputWithPast]:    
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        use_cache = use_cache if use_cache is not None else self.config.use_cache
+
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        if (input_ids is None) ^ (inputs_embeds is not None):
+            raise ValueError(
+                "You cannot specify both input_ids and inputs_embeds at the same time, and must specify either one"
+            )
+
+        if self.training and self.text_model.gradient_checkpointing and use_cache:
+            logger.warning_once(
+                "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`."
+            )
+            use_cache = False
+
+        if inputs_embeds is None:
+            inputs_embeds = self.embed_tokens(input_ids)
+
+        if inputs_embeds is not None and input_ids is None:
+            raise ValueError("When first calling the model, if input_embeds are passed, input_ids should not be None.")
+
+        # START VISUAL INPUTS INTEGRATION
+        if pixel_values is not None and image_hidden_states is not None:
+            raise ValueError("You cannot specify both pixel_values and image_hidden_states at the same time")
+        elif pixel_values is not None:
+            batch_size, num_images, num_channels, height, width = pixel_values.shape
+            pixel_values = pixel_values
+            pixel_values = pixel_values.view(batch_size * num_images, *pixel_values.shape[2:])
+
+            # Remove padding images - padding images are full 0.
+            nb_values_per_image = pixel_values.shape[1:].numel()
+            real_images_inds = (pixel_values == 0.0).sum(dim=(-1, -2, -3)) != nb_values_per_image
+
+            if not any(real_images_inds):
+                # no images, leave one empty image.
+                real_images_inds[0] = True
+
+            pixel_values = pixel_values[real_images_inds].contiguous()
+
+            # Handle the vision attention mask
+            if pixel_attention_mask is None:
+                pixel_attention_mask = torch.ones(
+                    size=[pixel_values.shape[i] for i in (0, 2, 3)],
+                    dtype=torch.bool,
+                    device=pixel_values.device,
+                )
+            else:
+                # Remove padding images from the mask
+                pixel_attention_mask = pixel_attention_mask.view(
+                    batch_size * num_images, *pixel_attention_mask.shape[2:]
+                )
+                pixel_attention_mask = pixel_attention_mask[real_images_inds].contiguous()
+
+            # patch_size = self.config.vision_config.patch_size
+            # patches_subgrid = pixel_attention_mask.unfold(dimension=1, size=patch_size, step=patch_size)
+            # patches_subgrid = patches_subgrid.unfold(dimension=2, size=patch_size, step=patch_size)
+            # patch_attention_mask = (patches_subgrid.sum(dim=(-1, -2)) > 0).bool()
+
+            # Get sequence from the vision encoder
+            image_hidden_states = self.vision_model(
+                pixel_values=pixel_values,
+                # patch_attention_mask=patch_attention_mask,
+            ).last_hidden_state
+
+            # Modality projection & resampling
+            image_hidden_states = self.connector(image_hidden_states)
+
+        elif image_hidden_states is not None:
+            image_hidden_states = image_hidden_states.to(dtype=self.dtype, device=input_ids.device)
+
+        if inputs_embeds is not None and image_hidden_states is not None:
+            # When we embed, we don't want to replace the potential image_token_id that we generated by images
+            # that simply don't exist
+            inputs_embeds = self.inputs_merger(
+                input_ids=input_ids,
+                inputs_embeds=inputs_embeds,
+                image_hidden_states=image_hidden_states,
+            )
+
+        outputs = self.text_model(
+            inputs_embeds=inputs_embeds,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+            past_key_values=past_key_values,
+            use_cache=use_cache,
+            cache_position=cache_position,
+        )
+
+        if not return_dict:
+            return tuple(v for v in [*outputs, image_hidden_states] if v is not None)
+
+        return VLlamaBaseModelOutputWithPast(
+            last_hidden_state=outputs.last_hidden_state,
+            past_key_values=past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            image_hidden_states=image_hidden_states,
+        )
+
+class VLlamaForCausalLM(VLlamaPreTrainedModel):
+    # _tied_weights_keys = ["predictions.decoder.bias", "cls.predictions.decoder.weight"]
+
+    def __init__(self, config, **kwargs):
+        super().__init__(config)
+
+        self.image_token_id = config.image_token_id
+        self.in_features = config.hidden_size
+        self.out_additional_features = config.additional_vocab_size
+        self.vocab_size = config.vocab_size
+
+        self.model = VLlamaModel(config, **kwargs)
+        self.lm_head = VLlamaForCausalLM.init_lm_head(config, **kwargs)
+        if self.out_additional_features > 0:
+            self.additional_fc = nn.Linear(
+                in_features=self.in_features,
+                out_features=self.out_additional_features,
+                bias=False,
+            )
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    @staticmethod
+    def init_lm_head(config, **kwargs):
+        # Get the pretrained model config
+        text_model_config = AutoConfig.from_pretrained(
+            config.text_config.text_model_name,
+            trust_remote_code=True,
+            **kwargs,
+        )
+        model = AutoModelForMaskedLM.from_config(text_model_config, trust_remote_code=True, **kwargs)
+        # Get the lm head
+        lm_head = model.lm_head if hasattr(model, "lm_head") else model.decoder if hasattr(model, "decoder") else None
+        if lm_head is None:
+            logger.warning(f"No lm head was found for {config.text_config.text_model_name}, initializing a new one.")
+            lm_head = nn.Linear(config.hidden_size, config.vocab_size, False)
+        return lm_head
+
+    def forward(
+            self,
+            input_ids: torch.LongTensor = None,
+            attention_mask: Optional[torch.Tensor] = None,
+            position_ids: Optional[torch.LongTensor] = None,
+            past_key_values: Optional[List[torch.FloatTensor]] = None,
+            inputs_embeds: Optional[torch.FloatTensor] = None,
+            pixel_values: Optional[torch.FloatTensor] = None,
+            pixel_attention_mask: Optional[torch.BoolTensor] = None,
+            image_hidden_states: Optional[torch.FloatTensor] = None,
+            labels: Optional[torch.LongTensor] = None,
+            use_cache: Optional[bool] = None,
+            output_attentions: Optional[bool] = None,
+            output_hidden_states: Optional[bool] = None,
+            return_dict: Optional[bool] = None,
+            cache_position: Optional[torch.LongTensor] = None,
+    ) -> Union[Tuple, VLlamaCausalLMOutputWithPast]:
+        r"""
+        Args:
+            labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+                Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
+                config.vocab_size]` or `model.image_token_id` (where `model` is your instance of `Idefics3ForConditionalGeneration`).
+                Tokens with indices set to `model.image_token_id` are ignored (masked), the loss is only
+                computed for the tokens with labels in `[0, ..., config.vocab_size]`.
+        ```"""
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+
+        # Pass the inputs to VLlamaModel
+        outputs = self.model(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            pixel_values=pixel_values,
+            pixel_attention_mask=pixel_attention_mask,
+            image_hidden_states=image_hidden_states,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+            cache_position=cache_position,
+        )
+
+        # Pass the outputs to the MLM head
+        hidden_states = outputs[0]
+
+        logits = self.lm_head(hidden_states)
+        if self.out_additional_features > 0:
+            additional_features = self.additional_fc(hidden_states)
+            logits = torch.cat((logits, additional_features), -1)
+        logits = logits.float()
+
+        loss = None
+        if labels is not None:
+            # Shift so that tokens < n predict n
+            if attention_mask is not None:
+                shift_attention_mask = attention_mask[..., 1:]
+                shift_logits = logits[..., :-1, :][shift_attention_mask != 0].contiguous()
+                shift_labels = labels[..., 1:][shift_attention_mask != 0].contiguous()
+            else:
+                shift_logits = logits[..., :-1, :].contiguous()
+                shift_labels = labels[..., 1:].contiguous()
+            # Flatten the tokens
+            loss_fct = CrossEntropyLoss()
+            loss = loss_fct(shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1))
+
+        if not return_dict:
+            output = (logits,) + outputs[1:]
+            return (loss,) + output if loss is not None else output
+
+        return VLlamaCausalLMOutputWithPast(
+            loss=loss,
+            logits=logits,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            image_hidden_states=outputs.image_hidden_states,
+        )
+
+class VLlamaForVision2Seq(VLlamaPreTrainedModel, GenerationMixin):
+    def __init__(self, config, **kwargs):
+        super().__init__(config)
+
+        self.image_token_id = config.image_token_id
+        self.in_features = config.hidden_size
+        self.out_additional_features = config.additional_vocab_size
+        self.vocab_size = config.vocab_size
+
+        self.model = VLlamaModel(config, **kwargs)
+        self.lm_head = VLlamaForVision2Seq.init_lm_head(config, **kwargs)
+        if self.out_additional_features > 0:
+            self.additional_fc = nn.Linear(
+                in_features=self.in_features,
+                out_features=self.out_additional_features,
+                bias=False,
+            )
+
+        self.loss_fct = CrossEntropyLoss()
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    @staticmethod
+    def init_lm_head(config, **kwargs):
+        # Get the pretrained model config
+        text_model_config = AutoConfig.from_pretrained(
+            config.text_config.text_model_name,
+            trust_remote_code=True,
+            **kwargs,
+        )
+        model = AutoModelForMaskedLM.from_config(text_model_config, trust_remote_code=True, **kwargs)
+        # Get the lm head
+        lm_head = model.lm_head if hasattr(model, "lm_head") else model.decoder if hasattr(model, "decoder") else None
+        if lm_head is None:
+            logger.warning(f"No lm head was found for {config.text_config.text_model_name}, initializing a new one.")
+            lm_head = nn.Linear(config.hidden_size, config.vocab_size, False)
+        return lm_head
+
+    def enable_input_require_grads(self):
+        """
+        Enables the gradients for the input embeddings. This is useful for fine-tuning adapter weights while keeping
+        the model weights fixed.
+        """
+
+        def make_inputs_require_grads(module, input, output):
+            output.requires_grad_(True)
+
+        self._text_require_grads_hook = self.get_input_embeddings().register_forward_hook(make_inputs_require_grads)
+        self._vision_require_grads_hook = self.model.vision_model.get_input_embeddings().register_forward_hook(
+            make_inputs_require_grads
+        )
+
+    def disable_input_require_grads(self):
+        self._text_require_grads_hook.remove()
+        self._vision_require_grads_hook.remove()
+
+    def get_input_embeddings(self):
+        return self.model.text_model.get_input_embeddings()
+
+    def set_input_embeddings(self, value):
+        self.model.text_model.set_input_embeddings(value)
+
+    def get_output_embeddings(self):
+        return self.lm_head
+
+    def set_output_embeddings(self, new_embeddings):
+        self.lm_head = new_embeddings
+
+    def forward(
+        self,
+        input_ids: Optional[torch.LongTensor] = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[List[torch.FloatTensor]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        pixel_values: Optional[torch.FloatTensor] = None,
+        pixel_attention_mask: Optional[torch.BoolTensor] = None,
+        image_hidden_states: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        cache_position: Optional[torch.LongTensor] = None,
+        return_dict: Optional[bool] = None,
+        logits_to_keep: Union[int, torch.Tensor] = 0,
+        **kwargs: Unpack[KwargsForCausalLM],
+    ) -> Union[Tuple, VLlamaCausalLMOutputWithPast]:
+        r"""
+        pixel_attention_mask (`torch.Tensor` of shape `(batch_size, image_size, image_size)`, *optional*):
+            Mask to avoid performing attention on padding pixel indices.
+        image_hidden_states (`torch.FloatTensor` of shape `(batch_size, num_channels, image_size, image_size)`):
+            The hidden states of the image encoder after modality projection.
+        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
+            config.vocab_size]` or `model.image_token_id` (where `model` is your instance of `SmolVLMForConditionalGeneration`).
+            Tokens with indices set to `model.image_token_id` are ignored (masked), the loss is only
+            computed for the tokens with labels in `[0, ..., config.vocab_size]`.
+
+        Example:
+
+        ```python
+        >>> import requests
+        >>> import torch
+        >>> from PIL import Image
+        >>> from io import BytesIO
+
+        >>> from transformers import AutoProcessor, AutoModelForImageTextToText
+        >>> from transformers.image_utils import load_image
+
+        >>> # Note that passing the image urls (instead of the actual pil images) to the processor is also possible
+        >>> image1 = load_image("https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg")
+        >>> image2 = load_image("https://cdn.britannica.com/59/94459-050-DBA42467/Skyline-Chicago.jpg")
+        >>> image3 = load_image("https://cdn.britannica.com/68/170868-050-8DDE8263/Golden-Gate-Bridge-San-Francisco.jpg")
+
+        >>> processor = AutoProcessor.from_pretrained("HuggingFaceTB/SmolVLM2-2.2B-Instruct")
+        >>> model = AutoModelForImageTextToText.from_pretrained("HuggingFaceTB/SmolVLM2-2.2B-Instruct", torch_dtype=torch.bfloat16, device_map="auto")
+
+        >>> # Create inputs
+        >>> messages = [
+        ...     {
+        ...         "role": "user",
+        ...         "content": [
+        ...             {"type": "video", "path": path/to/video},
+        ...             {"type": "text", "text": "What is happening in this video?"},
+        ...         ]
+        ...     }
+        ... ]
+
+        >>> inputs = processor.apply_chat_template([messages], add_generation_prompt=True)
+
+        >>> # Generate
+        >>> generated_ids = model.generate(**inputs, max_new_tokens=256)
+        >>> generated_texts = processor.batch_decode(generated_ids, skip_special_tokens=True)
+
+        >>> print(generated_texts)
+        ```"""
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        # decoder outputs consists of (dec_features, layer_state, dec_hidden, dec_attn)
+        outputs = self.model(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            pixel_values=pixel_values,
+            pixel_attention_mask=pixel_attention_mask,
+            image_hidden_states=image_hidden_states,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            cache_position=cache_position,
+            return_dict=True,
+            **kwargs,
+        )
+
+        hidden_states = outputs[0]
+        slice_indices = slice(-logits_to_keep, None) if isinstance(logits_to_keep, int) else logits_to_keep
+        # Only compute necessary logits, and do not upcast them to float if we are not computing the loss
+        hidden_states = hidden_states[:, slice_indices, :]
+        logits = self.lm_head(hidden_states)
+        if self.out_additional_features > 0:
+            additional_features = self.additional_fc(hidden_states)
+            logits = torch.cat((logits, additional_features), -1)
+        logits = logits.float()
+
+        loss = None
+        if labels is not None:
+            loss = self.loss_fct(
+                logits=logits, labels=labels, vocab_size=self.config.text_config.vocab_size, **kwargs
+            )
+
+        return VLlamaCausalLMOutputWithPast(
+            loss=loss,
+            logits=logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            image_hidden_states=outputs.image_hidden_states,
+        )
+
+    def prepare_inputs_for_generation(
+        self,
+        input_ids,
+        past_key_values=None,
+        attention_mask=None,
+        inputs_embeds=None,
+        cache_position=None,
+        pixel_values=None,
+        pixel_attention_mask=None,
+        image_hidden_states=None,
+        logits_to_keep=None,
+        **kwargs,
+    ):
+        # Overwritten -- there are mutually exclusive inputs (if the logic to make `image_hidden_states` take
+        # precedence is moved to the model, we can remove this fn)
+
+        model_inputs = super().prepare_inputs_for_generation(
+            input_ids,
+            past_key_values=past_key_values,
+            attention_mask=attention_mask,
+            inputs_embeds=inputs_embeds,
+            cache_position=cache_position,
+            pixel_values=pixel_values,
+            pixel_attention_mask=pixel_attention_mask,
+            image_hidden_states=image_hidden_states,
+            logits_to_keep=logits_to_keep,
+            **kwargs,
+        )
+
+        # if `inputs_embeds` are passed, we only want to use them in the 1st generation step
+        # but IDEFICS requires both ids and embeds to be present
+        if inputs_embeds is not None and cache_position[0] == 0:
+            model_inputs["input_ids"] = input_ids
+
+        if image_hidden_states is not None:
+            model_inputs["pixel_values"] = None
+            model_inputs["pixel_attention_mask"] = None
+
+        return model_inputs
+
+    def _update_model_kwargs_for_generation(self, outputs, model_kwargs, is_encoder_decoder, **kwargs):
+        model_kwargs = super()._update_model_kwargs_for_generation(
+            outputs=outputs,
+            model_kwargs=model_kwargs,
+            is_encoder_decoder=is_encoder_decoder,
+            **kwargs,
+        )
+        # Get the precomputed image_hidden_states
+        model_kwargs["image_hidden_states"] = outputs.image_hidden_states
+        return model_kwargs
+    
+    @staticmethod
+    # Copied from transformers.models.llama.modeling_llama.LlamaForCausalLM._reorder_cache
+    def _reorder_cache(past_key_values, beam_idx):
+        reordered_past = ()
+        for layer_past in past_key_values:
+            reordered_past += (
+                tuple(past_state.index_select(0, beam_idx.to(past_state.device)) for past_state in layer_past),
+            )
+        return reordered_past

vlm-siglip2-sllm_210/opt_step-10000__merged/preprocessor_config.json

@@ -0,0 +1,28 @@
+{
+    "do_convert_rgb": true,
+    "do_image_splitting": true,
+    "do_normalize": true,
+    "do_pad": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "image_mean": [
+        0.5,
+        0.5,
+        0.5
+    ],
+    "image_processor_type": "Idefics3ImageProcessor",
+    "image_std": [
+        0.5,
+        0.5,
+        0.5
+    ],
+    "max_image_size": {
+        "longest_edge": 512
+    },
+    "processor_class": "Idefics3Processor",
+    "resample": 1,
+    "rescale_factor": 0.00392156862745098,
+    "size": {
+        "longest_edge": 2048
+    }
+}

vlm-siglip2-sllm_210/opt_step-10000__merged/processor_config.json

@@ -0,0 +1,4 @@
+{
+    "image_seq_len": 64,
+    "processor_class": "Idefics3Processor"
+}

vlm-siglip2-sllm_210/opt_step-10000__merged/special_tokens_map.json

@@ -0,0 +1,72 @@
+{
+  "additional_special_tokens": [
+    "<global-img>",
+    "<row_1_col_1>",
+    "<row_1_col_2>",
+    "<row_1_col_3>",
+    "<row_1_col_4>",
+    "<row_1_col_5>",
+    "<row_1_col_6>",
+    "<row_2_col_1>",
+    "<row_2_col_2>",
+    "<row_2_col_3>",
+    "<row_2_col_4>",
+    "<row_2_col_5>",
+    "<row_2_col_6>",
+    "<row_3_col_1>",
+    "<row_3_col_2>",
+    "<row_3_col_3>",
+    "<row_3_col_4>",
+    "<row_3_col_5>",
+    "<row_3_col_6>",
+    "<row_4_col_1>",
+    "<row_4_col_2>",
+    "<row_4_col_3>",
+    "<row_4_col_4>",
+    "<row_4_col_5>",
+    "<row_4_col_6>",
+    "<row_5_col_1>",
+    "<row_5_col_2>",
+    "<row_5_col_3>",
+    "<row_5_col_4>",
+    "<row_5_col_5>",
+    "<row_5_col_6>",
+    "<row_6_col_1>",
+    "<row_6_col_2>",
+    "<row_6_col_3>",
+    "<row_6_col_4>",
+    "<row_6_col_5>",
+    "<row_6_col_6>",
+    "<end_of_utterance>",
+    "<fake_token_around_image>",
+    "<image>"
+  ],
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|end_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "<|reserved_special_token_0|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|end_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

vlm-siglip2-sllm_210/opt_step-10000__merged/tokenizer_config.json

@@ -0,0 +1,2429 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|finetune_right_pad_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|eom_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|python_tag|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128256": {
+      "content": "<global-img>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128257": {
+      "content": "<row_1_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128258": {
+      "content": "<row_1_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128259": {
+      "content": "<row_1_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128260": {
+      "content": "<row_1_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128261": {
+      "content": "<row_1_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128262": {
+      "content": "<row_1_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128263": {
+      "content": "<row_2_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128264": {
+      "content": "<row_2_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128265": {
+      "content": "<row_2_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128266": {
+      "content": "<row_2_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128267": {
+      "content": "<row_2_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128268": {
+      "content": "<row_2_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128269": {
+      "content": "<row_3_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128270": {
+      "content": "<row_3_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128271": {
+      "content": "<row_3_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128272": {
+      "content": "<row_3_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128273": {
+      "content": "<row_3_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128274": {
+      "content": "<row_3_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128275": {
+      "content": "<row_4_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128276": {
+      "content": "<row_4_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128277": {
+      "content": "<row_4_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128278": {
+      "content": "<row_4_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128279": {
+      "content": "<row_4_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128280": {
+      "content": "<row_4_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128281": {
+      "content": "<row_5_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128282": {
+      "content": "<row_5_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128283": {
+      "content": "<row_5_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128284": {
+      "content": "<row_5_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128285": {
+      "content": "<row_5_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128286": {
+      "content": "<row_5_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128287": {
+      "content": "<row_6_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128288": {
+      "content": "<row_6_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128289": {
+      "content": "<row_6_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128290": {
+      "content": "<row_6_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128291": {
+      "content": "<row_6_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128292": {
+      "content": "<row_6_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128293": {
+      "content": "<end_of_utterance>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128294": {
+      "content": "<fake_token_around_image>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128295": {
+      "content": "<image>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<global-img>",
+    "<row_1_col_1>",
+    "<row_1_col_2>",
+    "<row_1_col_3>",
+    "<row_1_col_4>",
+    "<row_1_col_5>",
+    "<row_1_col_6>",
+    "<row_2_col_1>",
+    "<row_2_col_2>",
+    "<row_2_col_3>",
+    "<row_2_col_4>",
+    "<row_2_col_5>",
+    "<row_2_col_6>",
+    "<row_3_col_1>",
+    "<row_3_col_2>",
+    "<row_3_col_3>",
+    "<row_3_col_4>",
+    "<row_3_col_5>",
+    "<row_3_col_6>",
+    "<row_4_col_1>",
+    "<row_4_col_2>",
+    "<row_4_col_3>",
+    "<row_4_col_4>",
+    "<row_4_col_5>",
+    "<row_4_col_6>",
+    "<row_5_col_1>",
+    "<row_5_col_2>",
+    "<row_5_col_3>",
+    "<row_5_col_4>",
+    "<row_5_col_5>",
+    "<row_5_col_6>",
+    "<row_6_col_1>",
+    "<row_6_col_2>",
+    "<row_6_col_3>",
+    "<row_6_col_4>",
+    "<row_6_col_5>",
+    "<row_6_col_6>",
+    "<end_of_utterance>",
+    "<fake_token_around_image>",
+    "<image>"
+  ],
+  "bos_token": "<|begin_of_text|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|end_of_text|>",
+  "extra_special_tokens": {},
+  "legacy": false,
+  "mask_token": "<|reserved_special_token_0|>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask",
+    "pixel_values",
+    "pixel_attention_mask"
+  ],
+  "model_max_length": 8192,
+  "pad_token": "<|end_of_text|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}

vlm-siglip2-sllm_210/opt_step-12000/resume_run_infos.json

@@ -0,0 +1,91 @@
+{
+  "train_logs": {
+    "lr": 0.0001,
+    "num_opt_steps": 12000,
+    "num_epochs": 0,
+    "per_token_loss": {
+      "sft": 0.853016659617424,
+      "all": 0.853016659617424
+    },
+    "z_loss": {
+      "sft": 0.0,
+      "all": 0.0
+    },
+    "watt/s": {
+      "sft": 316.8570018391261,
+      "all": 316.8570018391261
+    },
+    "tflops": {
+      "sft": 11.26828050176133,
+      "all": 11.26828050176133
+    },
+    "tflop_counter": {
+      "sft": 10962013.012542725,
+      "all": 10962013.012542725
+    },
+    "fwd_bwd_time": {
+      "sft": 926257.0768032074,
+      "all": 926257.0768032074
+    },
+    "tflops_acc": {
+      "sft": 11.834741441734446,
+      "all": 11.834741441734446
+    },
+    "num_per_device_batches": {
+      "sft": 768000,
+      "all": 768000
+    },
+    "num_images": {
+      "sft": 44330752,
+      "all": 44330752
+    },
+    "num_image_tokens": {
+      "sft": 2837168128,
+      "all": 2837168128
+    },
+    "num_tokens": {
+      "sft": 1411562162,
+      "all": 1411562162
+    },
+    "image_to_text_ratio": {
+      "sft": 0.04141807556152344,
+      "all": 0.04141807556152344
+    },
+    "pixel_values_sum": {
+      "sft": 11536196737784.0,
+      "all": 11536196737784.0
+    },
+    "num_padding": {
+      "sft": 766141803,
+      "all": 766141803
+    },
+    "num_per_device_batches_in_curr_epoch": {
+      "sft": 768000,
+      "all": 768000
+    },
+    "num_batches": {
+      "all": 768000
+    },
+    "num_batches_in_curr_epoch": {
+      "all": 768000
+    },
+    "per_token_loss_acc": {},
+    "z_loss_acc": {},
+    "num_batches_since_training_logged": {},
+    "num_per_device_batches_since_training_logged": {},
+    "tflop_counter_since_training_logged": {},
+    "total_energy_delta_since_training_logged": {},
+    "fwd_bwd_time_since_training_logged": {},
+    "global_batch_size_current": 256
+  },
+  "wandb_run_id": "hp0rcdxj",
+  "seed": 42,
+  "resume_opt_step": 12000,
+  "resume_epoch": 0,
+  "gbs_running": {
+    "global_seen_samples": 3072000,
+    "global_batch_size_current": 256,
+    "next_goal_samples": 0,
+    "grad_acc_size_current": 4
+  }
+}

vlm-siglip2-sllm_210/opt_step-14000__merged/chat_template.jinja

@@ -0,0 +1,2 @@
+<|begin_of_text|>{% for message in messages %}{{message['role'] | capitalize}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>
+{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}

vlm-siglip2-sllm_210/opt_step-14000__merged/preprocessor_config.json

@@ -0,0 +1,28 @@
+{
+    "do_convert_rgb": true,
+    "do_image_splitting": true,
+    "do_normalize": true,
+    "do_pad": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "image_mean": [
+        0.5,
+        0.5,
+        0.5
+    ],
+    "image_processor_type": "Idefics3ImageProcessor",
+    "image_std": [
+        0.5,
+        0.5,
+        0.5
+    ],
+    "max_image_size": {
+        "longest_edge": 512
+    },
+    "processor_class": "Idefics3Processor",
+    "resample": 1,
+    "rescale_factor": 0.00392156862745098,
+    "size": {
+        "longest_edge": 2048
+    }
+}

vlm-siglip2-sllm_210/opt_step-16000/resume_run_infos.json

@@ -0,0 +1,91 @@
+{
+  "train_logs": {
+    "lr": 0.0001,
+    "num_opt_steps": 16000,
+    "num_epochs": 0,
+    "per_token_loss": {
+      "sft": 0.9875863194465637,
+      "all": 0.9875863194465637
+    },
+    "z_loss": {
+      "sft": 0.0,
+      "all": 0.0
+    },
+    "watt/s": {
+      "sft": 369.12963209640463,
+      "all": 369.12963209640463
+    },
+    "tflops": {
+      "sft": 14.106507837441953,
+      "all": 14.106507837441953
+    },
+    "tflop_counter": {
+      "sft": 14613755.036727905,
+      "all": 14613755.036727905
+    },
+    "fwd_bwd_time": {
+      "sft": 1233785.3667488098,
+      "all": 1233785.3667488098
+    },
+    "tflops_acc": {
+      "sft": 11.844649345483091,
+      "all": 11.844649345483091
+    },
+    "num_per_device_batches": {
+      "sft": 1024000,
+      "all": 1024000
+    },
+    "num_images": {
+      "sft": 59109381,
+      "all": 59109381
+    },
+    "num_image_tokens": {
+      "sft": 3783000384,
+      "all": 3783000384
+    },
+    "num_tokens": {
+      "sft": 1881469744,
+      "all": 1881469744
+    },
+    "image_to_text_ratio": {
+      "sft": 0.04391550272703171,
+      "all": 0.04391550272703171
+    },
+    "pixel_values_sum": {
+      "sft": 15381735685432.0,
+      "all": 15381735685432.0
+    },
+    "num_padding": {
+      "sft": 1020871450,
+      "all": 1020871450
+    },
+    "num_per_device_batches_in_curr_epoch": {
+      "sft": 1024000,
+      "all": 1024000
+    },
+    "num_batches": {
+      "all": 1024000
+    },
+    "num_batches_in_curr_epoch": {
+      "all": 1024000
+    },
+    "per_token_loss_acc": {},
+    "z_loss_acc": {},
+    "num_batches_since_training_logged": {},
+    "num_per_device_batches_since_training_logged": {},
+    "tflop_counter_since_training_logged": {},
+    "total_energy_delta_since_training_logged": {},
+    "fwd_bwd_time_since_training_logged": {},
+    "global_batch_size_current": 256
+  },
+  "wandb_run_id": "hp0rcdxj",
+  "seed": 42,
+  "resume_opt_step": 16000,
+  "resume_epoch": 0,
+  "gbs_running": {
+    "global_seen_samples": 4096000,
+    "global_batch_size_current": 256,
+    "next_goal_samples": 0,
+    "grad_acc_size_current": 4
+  }
+}

vlm-siglip2-sllm_210/opt_step-16000__merged/chat_template.jinja

@@ -0,0 +1,2 @@
+<|begin_of_text|>{% for message in messages %}{{message['role'] | capitalize}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>
+{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}

vlm-siglip2-sllm_210/opt_step-16000__merged/chat_template.json

@@ -0,0 +1,3 @@
+{
+    "chat_template": "<|begin_of_text|>{% for message in messages %}{{message['role'] | capitalize}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>\n{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}"
+}

vlm-siglip2-sllm_210/opt_step-16000__merged/configuration_vllama.py

@@ -0,0 +1,233 @@
+import copy
+import os
+
+from typing import Union, Any, Dict
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+from transformers import AutoConfig
+
+logger = logging.get_logger(__name__)
+
+def collect_arg_in_candidates(config, candidates, default = None) -> Any:
+    """ Gets the argument in a config given a list of candidates """
+    for c in candidates:
+        if hasattr(config, c):
+            return getattr(config, c)
+        elif c in config:
+            return config[c]
+    if default is not None:
+        return default
+    raise ValueError("No matching arguments found in candidates. Candidates: {}, Config: {}".format(candidates, config))
+
+class VLlamaTextConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`LlamaModel`]. It is used to instantiate an LLaMA
+    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 LLaMA-7B.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        embed_dim (`int`, *optional*, defaults to 1152):
+            Dimensionality of the encoder layers and the pooler layer. (elsewhere referred to as `embed_dim`)
+        image_size (`int`, *optional*, defaults to 384):
+            The size (resolution) of each image.
+    """
+    model_type = "VLlama"
+
+    def __init__(
+        self,
+        # Case for when vllama3 is from the hub with no vision_model_name
+        text_model_name="HuggingFaceTB/SmolLM2-135M-Instruct",
+        **kwargs,
+    ):
+        self.text_model_name = text_model_name
+        text_config = AutoConfig.from_pretrained(text_model_name, trust_remote_code=True)
+        if hasattr(text_config, "text_config"):
+            text_config = text_config.text_config
+        
+        self.hidden_size = collect_arg_in_candidates(text_config, ["hidden_size", "embed_dim"])
+        self.num_hidden_layers = collect_arg_in_candidates(text_config, ["num_hidden_layers", "num_hidden_blocks"])
+        self.intermediate_size = collect_arg_in_candidates(text_config, ["intermediate_size", "mlp_dim"])
+        self.mlp_bias = collect_arg_in_candidates(text_config, ["mlp_bias", "mlp_hidden_bias"], default = False)
+        self.vocab_size = collect_arg_in_candidates(text_config, ["vocab_size"])
+    
+        super().__init__(text_model_name=text_model_name, **kwargs)
+
+class VLlamaVisionConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`LlamaModel`]. It is used to instantiate an LLaMA
+    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 LLaMA-7B.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        embed_dim (`int`, *optional*, defaults to 1152):
+            Dimensionality of the encoder layers and the pooler layer. (elsewhere referred to as `embed_dim`)
+        image_size (`int`, *optional*, defaults to 384):
+            The size (resolution) of each image.
+    """
+    model_type = "VLlama"
+    attribute_map = {
+        "hidden_size": "embed_dim",
+    }
+
+    def __init__(
+        self,
+        # Case for when vllama3 is from the hub with no vision_model_name
+        vision_model_name="google/siglip2-base-patch16-512",
+        **kwargs,
+    ):
+        self.vision_model_name = vision_model_name
+        vision_config = AutoConfig.from_pretrained(vision_model_name, trust_remote_code=True)
+        if hasattr(vision_config, "vision_config"):
+            vision_config = vision_config.vision_config
+
+        self.embed_dim = collect_arg_in_candidates(vision_config, ["embed_dim", "hidden_size"])
+        self.image_size = collect_arg_in_candidates(vision_config, ["image_size", "img_size"])
+        self.patch_size = collect_arg_in_candidates(vision_config, ["patch_size"])
+        self.num_hidden_layers = collect_arg_in_candidates(vision_config, ["num_hidden_layers", "num_hidden_blocks"])
+        self.intermediate_size = collect_arg_in_candidates(vision_config, ["intermediate_size", "mlp_dim"])
+
+        super().__init__(vision_model_name=vision_model_name, **kwargs)
+
+class VLlamaConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`SmolVLMModel`]. It is used to instantiate a
+    SmolVLM 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 model of the SmolVLM
+    [HuggingFaceTB/SmolVLM2-2.2B-Instruct](https://huggingface.co/HuggingFaceTB/SmolVLM2-2.2B-Instruct) architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should cache the key/value pairs of the attention mechanism. Only
+            relevant if `config.is_decoder=True`.
+        image_token_id (`int`, *optional*, defaults to 128257):
+            The id of the "image" token.
+        tie_word_embeddings (`bool`, *optional*, defaults to `False`):
+            Whether or not to tie the word embeddings with the token embeddings.
+        vision_config (`IdeficsVisionConfig` or `dict`, *optional*, defaults to `IdeficsVisionConfig`):
+            Custom vision config or dict for the vision tower
+        text_config (`PretrainedConfig` or `dict`, *optional*, defaults to `LlamaConfig`):
+            Custom text config or dict for the text model
+        scale_factor (`int`, *optional*, defaults to 2):
+            The scale factor for the image encoder.
+        pad_token_id (`int`, *optional*, defaults to 128002):
+            The id of the padding token.
+
+    Example:
+    ```python
+    >>> from transformers import SmolVLMModel, SmolVLMConfig
+    >>> # Initializing configuration
+    >>> configuration = SmolVLMConfig()
+    >>> # Initializing a model from the configuration
+    >>> model = SmolVLMModel(configuration)
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "VLlama"
+    is_composition = True
+    # sub_configs = {"text_config": VLlamaTextConfig, "vision_config": VLlamaVisionConfig}
+
+    DEFAULT_TEXT_MODEL_NAME = "EuroBERT/EuroBERT-210m"
+    DEFAULT_VISION_MODEL_NAME = "google/siglip2-base-patch16-512"
+
+    def __init__(
+        self,
+        text_config: Union[PretrainedConfig, Dict[str, Any]] = None,
+        vision_config: Union[PretrainedConfig, Dict[str, Any]] = None,
+        image_token_id: int = 128_257,
+        vocab_size=128_256,
+        use_cache = True,
+        tie_word_embeddings = False,
+        freeze_config = None,
+        pad_token_id = None,
+        initializer_range = 0.02,
+        pixel_shuffle_factor = 4,
+        use_resampler = False,
+        additional_vocab_size = 0,
+        neftune_noise_alpha = 0.0,
+        **kwargs,
+    ):        
+        self.image_token_id = image_token_id
+        self.use_cache = use_cache
+        self.tie_word_embeddings = tie_word_embeddings
+        self.scale_factor = pixel_shuffle_factor
+        self.additional_vocab_size = additional_vocab_size
+
+        if text_config is None:
+            text_config = AutoConfig.from_pretrained(self.DEFAULT_TEXT_MODEL_NAME, trust_remote_code=True)
+        elif isinstance(text_config, dict):
+            text_config = VLlamaTextConfig(text_config["text_model_name"])
+        self.text_config = text_config
+        
+        if vision_config is None:
+            vision_config = AutoConfig.from_pretrained(self.DEFAULT_VISION_MODEL_NAME, trust_remote_code=True)
+        elif isinstance(vision_config, dict):
+            vision_config = VLlamaVisionConfig(vision_config["vision_model_name"])
+        self.vision_config = vision_config
+
+        self.freeze_config = freeze_config
+
+        # Pixel shuffle factor
+        self.pixel_shuffle_factor = pixel_shuffle_factor
+        self.use_resampler = use_resampler
+
+        self.neftune_noise_alpha = neftune_noise_alpha
+
+        self.initializer_range = initializer_range
+
+        hidden_size = kwargs.pop("hidden_size", self.text_config.hidden_size)
+
+        super().__init__(
+            **kwargs, 
+            pad_token_id=pad_token_id, 
+            tie_word_embeddings=tie_word_embeddings,
+            vocab_size=vocab_size,
+            hidden_size=hidden_size,
+        )
+
+    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["model_type"] = self.__class__.model_type
+        output["vision_config"] = self.vision_config.to_dict()
+        output["text_config"] = self.text_config.to_dict()
+        # output["freeze_config"] = self.freeze_config.to_dict()
+
+        return output
+    
+    # @classmethod
+    # def from_pretrained(cls, pretrained_model_name_or_path: Union[str, os.PathLike], **kwargs) -> "PretrainedConfig":
+    #     outputs = super(VLlamaConfig, cls).from_pretrained(pretrained_model_name_or_path, **kwargs)
+    #     return outputs
+
+    @classmethod
+    def from_pretrained_models(
+        cls, 
+        text_model_name: Union[str, os.PathLike],
+        vision_model_name: Union[str, os.PathLike],
+        **kwargs
+    ) -> "PretrainedConfig":
+        # text_model_config = AutoConfig.from_pretrained(text_model_name, trust_remote_code=True)
+        # vision_model_config = AutoConfig.from_pretrained(vision_model_name, trust_remote_code=True)
+        text_model_config = VLlamaTextConfig(text_model_name)
+        vision_model_config = VLlamaVisionConfig(vision_model_name)
+        return cls(
+            text_config=text_model_config,
+            vision_config=vision_model_config,
+            **kwargs
+        )

vlm-siglip2-sllm_210/opt_step-16000__merged/preprocessor_config.json

@@ -0,0 +1,28 @@
+{
+    "do_convert_rgb": true,
+    "do_image_splitting": true,
+    "do_normalize": true,
+    "do_pad": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "image_mean": [
+        0.5,
+        0.5,
+        0.5
+    ],
+    "image_processor_type": "Idefics3ImageProcessor",
+    "image_std": [
+        0.5,
+        0.5,
+        0.5
+    ],
+    "max_image_size": {
+        "longest_edge": 512
+    },
+    "processor_class": "Idefics3Processor",
+    "resample": 1,
+    "rescale_factor": 0.00392156862745098,
+    "size": {
+        "longest_edge": 2048
+    }
+}

vlm-siglip2-sllm_210/opt_step-16000__merged/processor_config.json

@@ -0,0 +1,4 @@
+{
+    "image_seq_len": 64,
+    "processor_class": "Idefics3Processor"
+}

vlm-siglip2-sllm_210/opt_step-16000__merged/special_tokens_map.json

@@ -0,0 +1,72 @@
+{
+  "additional_special_tokens": [
+    "<global-img>",
+    "<row_1_col_1>",
+    "<row_1_col_2>",
+    "<row_1_col_3>",
+    "<row_1_col_4>",
+    "<row_1_col_5>",
+    "<row_1_col_6>",
+    "<row_2_col_1>",
+    "<row_2_col_2>",
+    "<row_2_col_3>",
+    "<row_2_col_4>",
+    "<row_2_col_5>",
+    "<row_2_col_6>",
+    "<row_3_col_1>",
+    "<row_3_col_2>",
+    "<row_3_col_3>",
+    "<row_3_col_4>",
+    "<row_3_col_5>",
+    "<row_3_col_6>",
+    "<row_4_col_1>",
+    "<row_4_col_2>",
+    "<row_4_col_3>",
+    "<row_4_col_4>",
+    "<row_4_col_5>",
+    "<row_4_col_6>",
+    "<row_5_col_1>",
+    "<row_5_col_2>",
+    "<row_5_col_3>",
+    "<row_5_col_4>",
+    "<row_5_col_5>",
+    "<row_5_col_6>",
+    "<row_6_col_1>",
+    "<row_6_col_2>",
+    "<row_6_col_3>",
+    "<row_6_col_4>",
+    "<row_6_col_5>",
+    "<row_6_col_6>",
+    "<end_of_utterance>",
+    "<fake_token_around_image>",
+    "<image>"
+  ],
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|end_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "<|reserved_special_token_0|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|end_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

vlm-siglip2-sllm_210/opt_step-18000/resume_run_infos.json

@@ -0,0 +1,91 @@
+{
+  "train_logs": {
+    "lr": 0.0001,
+    "num_opt_steps": 18000,
+    "num_epochs": 0,
+    "per_token_loss": {
+      "sft": 0.889340803027153,
+      "all": 0.889340803027153
+    },
+    "z_loss": {
+      "sft": 0.0,
+      "all": 0.0
+    },
+    "watt/s": {
+      "sft": 426.92646766881666,
+      "all": 426.92646766881666
+    },
+    "tflops": {
+      "sft": 15.52232696171255,
+      "all": 15.52232696171255
+    },
+    "tflop_counter": {
+      "sft": 16440704.562393188,
+      "all": 16440704.562393188
+    },
+    "fwd_bwd_time": {
+      "sft": 1387798.465812683,
+      "all": 1387798.465812683
+    },
+    "tflops_acc": {
+      "sft": 11.846608111622064,
+      "all": 11.846608111622064
+    },
+    "num_per_device_batches": {
+      "sft": 1152000,
+      "all": 1152000
+    },
+    "num_images": {
+      "sft": 66499626,
+      "all": 66499626
+    },
+    "num_image_tokens": {
+      "sft": 4255976064,
+      "all": 4255976064
+    },
+    "num_tokens": {
+      "sft": 2116546501,
+      "all": 2116546501
+    },
+    "image_to_text_ratio": {
+      "sft": 0.03476424515247345,
+      "all": 0.03476424515247345
+    },
+    "pixel_values_sum": {
+      "sft": 17304773620424.0,
+      "all": 17304773620424.0
+    },
+    "num_padding": {
+      "sft": 1148525875,
+      "all": 1148525875
+    },
+    "num_per_device_batches_in_curr_epoch": {
+      "sft": 1152000,
+      "all": 1152000
+    },
+    "num_batches": {
+      "all": 1152000
+    },
+    "num_batches_in_curr_epoch": {
+      "all": 1152000
+    },
+    "per_token_loss_acc": {},
+    "z_loss_acc": {},
+    "num_batches_since_training_logged": {},
+    "num_per_device_batches_since_training_logged": {},
+    "tflop_counter_since_training_logged": {},
+    "total_energy_delta_since_training_logged": {},
+    "fwd_bwd_time_since_training_logged": {},
+    "global_batch_size_current": 256
+  },
+  "wandb_run_id": "hp0rcdxj",
+  "seed": 42,
+  "resume_opt_step": 18000,
+  "resume_epoch": 0,
+  "gbs_running": {
+    "global_seen_samples": 4608000,
+    "global_batch_size_current": 256,
+    "next_goal_samples": 0,
+    "grad_acc_size_current": 4
+  }
+}

vlm-siglip2-sllm_210/opt_step-18000__merged/chat_template.jinja

@@ -0,0 +1,2 @@
+<|begin_of_text|>{% for message in messages %}{{message['role'] | capitalize}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>
+{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}

vlm-siglip2-sllm_210/opt_step-18000__merged/chat_template.json

@@ -0,0 +1,3 @@
+{
+    "chat_template": "<|begin_of_text|>{% for message in messages %}{{message['role'] | capitalize}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>\n{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}"
+}

vlm-siglip2-sllm_210/opt_step-18000__merged/config.json

@@ -0,0 +1,51 @@
+{
+  "additional_vocab_size": 40,
+  "architectures": [
+    "VLlamaForCausalLM"
+  ],
+  "auto_map": {
+    "AutoConfig": "configuration_vllama.VLlamaConfig",
+    "AutoModel": "modeling_vllama.VLlamaModel",
+    "AutoModelForCausalLM": "modeling_vllama.VLlamaForCausalLM",
+    "AutoModelForVision2Seq": "modeling_vllama.VLlamaForVision2Seq"
+  },
+  "freeze_config": {
+    "freeze_lm_head": true,
+    "freeze_text_layers": true,
+    "freeze_vision_layers": true
+  },
+  "hidden_size": 768,
+  "image_token_id": 128295,
+  "initializer_range": 0.02,
+  "max_position_embeddings": 8192,
+  "model_type": "VLlama",
+  "neftune_noise_alpha": 0.0,
+  "output_attentions": false,
+  "pixel_shuffle_factor": 4,
+  "qk_layer_norms": false,
+  "scale_factor": 4,
+  "text_config": {
+    "hidden_size": 768,
+    "intermediate_size": 3072,
+    "mlp_bias": false,
+    "model_type": "VLlama",
+    "num_hidden_layers": 12,
+    "text_model_name": "SmolVEncoder/decoder-210m",
+    "vocab_size": 128256
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": null,
+  "use_cache": true,
+  "use_resampler": false,
+  "vision_config": {
+    "embed_dim": 768,
+    "image_size": 512,
+    "intermediate_size": 3072,
+    "model_type": "VLlama",
+    "num_hidden_layers": 12,
+    "patch_size": 16,
+    "vision_model_name": "google/siglip2-base-patch16-512"
+  },
+  "vocab_size": 128256
+}

vlm-siglip2-sllm_210/opt_step-18000__merged/configuration_vllama.py

@@ -0,0 +1,233 @@
+import copy
+import os
+
+from typing import Union, Any, Dict
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+from transformers import AutoConfig
+
+logger = logging.get_logger(__name__)
+
+def collect_arg_in_candidates(config, candidates, default = None) -> Any:
+    """ Gets the argument in a config given a list of candidates """
+    for c in candidates:
+        if hasattr(config, c):
+            return getattr(config, c)
+        elif c in config:
+            return config[c]
+    if default is not None:
+        return default
+    raise ValueError("No matching arguments found in candidates. Candidates: {}, Config: {}".format(candidates, config))
+
+class VLlamaTextConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`LlamaModel`]. It is used to instantiate an LLaMA
+    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 LLaMA-7B.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        embed_dim (`int`, *optional*, defaults to 1152):
+            Dimensionality of the encoder layers and the pooler layer. (elsewhere referred to as `embed_dim`)
+        image_size (`int`, *optional*, defaults to 384):
+            The size (resolution) of each image.
+    """
+    model_type = "VLlama"
+
+    def __init__(
+        self,
+        # Case for when vllama3 is from the hub with no vision_model_name
+        text_model_name="HuggingFaceTB/SmolLM2-135M-Instruct",
+        **kwargs,
+    ):
+        self.text_model_name = text_model_name
+        text_config = AutoConfig.from_pretrained(text_model_name, trust_remote_code=True)
+        if hasattr(text_config, "text_config"):
+            text_config = text_config.text_config
+        
+        self.hidden_size = collect_arg_in_candidates(text_config, ["hidden_size", "embed_dim"])
+        self.num_hidden_layers = collect_arg_in_candidates(text_config, ["num_hidden_layers", "num_hidden_blocks"])
+        self.intermediate_size = collect_arg_in_candidates(text_config, ["intermediate_size", "mlp_dim"])
+        self.mlp_bias = collect_arg_in_candidates(text_config, ["mlp_bias", "mlp_hidden_bias"], default = False)
+        self.vocab_size = collect_arg_in_candidates(text_config, ["vocab_size"])
+    
+        super().__init__(text_model_name=text_model_name, **kwargs)
+
+class VLlamaVisionConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`LlamaModel`]. It is used to instantiate an LLaMA
+    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 LLaMA-7B.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        embed_dim (`int`, *optional*, defaults to 1152):
+            Dimensionality of the encoder layers and the pooler layer. (elsewhere referred to as `embed_dim`)
+        image_size (`int`, *optional*, defaults to 384):
+            The size (resolution) of each image.
+    """
+    model_type = "VLlama"
+    attribute_map = {
+        "hidden_size": "embed_dim",
+    }
+
+    def __init__(
+        self,
+        # Case for when vllama3 is from the hub with no vision_model_name
+        vision_model_name="google/siglip2-base-patch16-512",
+        **kwargs,
+    ):
+        self.vision_model_name = vision_model_name
+        vision_config = AutoConfig.from_pretrained(vision_model_name, trust_remote_code=True)
+        if hasattr(vision_config, "vision_config"):
+            vision_config = vision_config.vision_config
+
+        self.embed_dim = collect_arg_in_candidates(vision_config, ["embed_dim", "hidden_size"])
+        self.image_size = collect_arg_in_candidates(vision_config, ["image_size", "img_size"])
+        self.patch_size = collect_arg_in_candidates(vision_config, ["patch_size"])
+        self.num_hidden_layers = collect_arg_in_candidates(vision_config, ["num_hidden_layers", "num_hidden_blocks"])
+        self.intermediate_size = collect_arg_in_candidates(vision_config, ["intermediate_size", "mlp_dim"])
+
+        super().__init__(vision_model_name=vision_model_name, **kwargs)
+
+class VLlamaConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`SmolVLMModel`]. It is used to instantiate a
+    SmolVLM 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 model of the SmolVLM
+    [HuggingFaceTB/SmolVLM2-2.2B-Instruct](https://huggingface.co/HuggingFaceTB/SmolVLM2-2.2B-Instruct) architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should cache the key/value pairs of the attention mechanism. Only
+            relevant if `config.is_decoder=True`.
+        image_token_id (`int`, *optional*, defaults to 128257):
+            The id of the "image" token.
+        tie_word_embeddings (`bool`, *optional*, defaults to `False`):
+            Whether or not to tie the word embeddings with the token embeddings.
+        vision_config (`IdeficsVisionConfig` or `dict`, *optional*, defaults to `IdeficsVisionConfig`):
+            Custom vision config or dict for the vision tower
+        text_config (`PretrainedConfig` or `dict`, *optional*, defaults to `LlamaConfig`):
+            Custom text config or dict for the text model
+        scale_factor (`int`, *optional*, defaults to 2):
+            The scale factor for the image encoder.
+        pad_token_id (`int`, *optional*, defaults to 128002):
+            The id of the padding token.
+
+    Example:
+    ```python
+    >>> from transformers import SmolVLMModel, SmolVLMConfig
+    >>> # Initializing configuration
+    >>> configuration = SmolVLMConfig()
+    >>> # Initializing a model from the configuration
+    >>> model = SmolVLMModel(configuration)
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "VLlama"
+    is_composition = True
+    # sub_configs = {"text_config": VLlamaTextConfig, "vision_config": VLlamaVisionConfig}
+
+    DEFAULT_TEXT_MODEL_NAME = "EuroBERT/EuroBERT-210m"
+    DEFAULT_VISION_MODEL_NAME = "google/siglip2-base-patch16-512"
+
+    def __init__(
+        self,
+        text_config: Union[PretrainedConfig, Dict[str, Any]] = None,
+        vision_config: Union[PretrainedConfig, Dict[str, Any]] = None,
+        image_token_id: int = 128_257,
+        vocab_size=128_256,
+        use_cache = True,
+        tie_word_embeddings = False,
+        freeze_config = None,
+        pad_token_id = None,
+        initializer_range = 0.02,
+        pixel_shuffle_factor = 4,
+        use_resampler = False,
+        additional_vocab_size = 0,
+        neftune_noise_alpha = 0.0,
+        **kwargs,
+    ):        
+        self.image_token_id = image_token_id
+        self.use_cache = use_cache
+        self.tie_word_embeddings = tie_word_embeddings
+        self.scale_factor = pixel_shuffle_factor
+        self.additional_vocab_size = additional_vocab_size
+
+        if text_config is None:
+            text_config = AutoConfig.from_pretrained(self.DEFAULT_TEXT_MODEL_NAME, trust_remote_code=True)
+        elif isinstance(text_config, dict):
+            text_config = VLlamaTextConfig(text_config["text_model_name"])
+        self.text_config = text_config
+        
+        if vision_config is None:
+            vision_config = AutoConfig.from_pretrained(self.DEFAULT_VISION_MODEL_NAME, trust_remote_code=True)
+        elif isinstance(vision_config, dict):
+            vision_config = VLlamaVisionConfig(vision_config["vision_model_name"])
+        self.vision_config = vision_config
+
+        self.freeze_config = freeze_config
+
+        # Pixel shuffle factor
+        self.pixel_shuffle_factor = pixel_shuffle_factor
+        self.use_resampler = use_resampler
+
+        self.neftune_noise_alpha = neftune_noise_alpha
+
+        self.initializer_range = initializer_range
+
+        hidden_size = kwargs.pop("hidden_size", self.text_config.hidden_size)
+
+        super().__init__(
+            **kwargs, 
+            pad_token_id=pad_token_id, 
+            tie_word_embeddings=tie_word_embeddings,
+            vocab_size=vocab_size,
+            hidden_size=hidden_size,
+        )
+
+    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["model_type"] = self.__class__.model_type
+        output["vision_config"] = self.vision_config.to_dict()
+        output["text_config"] = self.text_config.to_dict()
+        # output["freeze_config"] = self.freeze_config.to_dict()
+
+        return output
+    
+    # @classmethod
+    # def from_pretrained(cls, pretrained_model_name_or_path: Union[str, os.PathLike], **kwargs) -> "PretrainedConfig":
+    #     outputs = super(VLlamaConfig, cls).from_pretrained(pretrained_model_name_or_path, **kwargs)
+    #     return outputs
+
+    @classmethod
+    def from_pretrained_models(
+        cls, 
+        text_model_name: Union[str, os.PathLike],
+        vision_model_name: Union[str, os.PathLike],
+        **kwargs
+    ) -> "PretrainedConfig":
+        # text_model_config = AutoConfig.from_pretrained(text_model_name, trust_remote_code=True)
+        # vision_model_config = AutoConfig.from_pretrained(vision_model_name, trust_remote_code=True)
+        text_model_config = VLlamaTextConfig(text_model_name)
+        vision_model_config = VLlamaVisionConfig(vision_model_name)
+        return cls(
+            text_config=text_model_config,
+            vision_config=vision_model_config,
+            **kwargs
+        )

vlm-siglip2-sllm_210/opt_step-18000__merged/modeling_vllama.py

@@ -0,0 +1,895 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from torch.nn import CrossEntropyLoss
+from typing import Optional, Tuple, Union, List
+
+# from transformers.models.smolvlm import SmolVLMModel, SmolVLMPreTrainedModel
+
+from .configuration_vllama import VLlamaConfig
+
+from transformers import AutoModel, AutoConfig, AutoModelForMaskedLM, GenerationMixin
+from transformers.cache_utils import Cache
+from transformers.modeling_utils import PreTrainedModel
+from transformers.modeling_outputs import BaseModelOutput
+from transformers.models.bert.modeling_bert import BaseModelOutputWithPoolingAndCrossAttentions, MaskedLMOutput
+from transformers.models.idefics3.modeling_idefics3 import Idefics3VisionTransformer
+from transformers.processing_utils import Unpack
+from transformers.modeling_flash_attention_utils import FlashAttentionKwargs
+from transformers.utils import LossKwargs
+
+from typing import List, Optional, Tuple, Union
+
+import torch
+import torch.utils.checkpoint
+
+from dataclasses import dataclass
+
+from transformers import logging 
+from transformers.utils import ContextManagers
+
+logger = logging.get_logger(__name__)
+
+class KwargsForCausalLM(FlashAttentionKwargs, LossKwargs): ...
+
+class DecoupledEmbedding(nn.Embedding):
+    # Derived from https://pytorch.org/docs/stable/_modules/torch/nn/modules/sparse.html#Embedding
+    """
+    Implements a decoupling of parameters to allow freezing (or not) a subset of the embeddings.
+    In practise, the regular `weight` can be trained or frozen (i.e. `partially_freeze=True`), and if `num_additional_embeddings` > 0, then it will create `num_additional_embeddings` additional parameters that are always trained.
+    If `num_additional_embeddings=0`, then the module defaults back to the regular behavior of `nn.Embedding`.
+    """
+
+    def __init__(
+        self,
+        num_embeddings,
+        num_additional_embeddings,
+        embedding_dim,
+        partially_freeze=False,
+        device=None,
+        dtype=None,
+        padding_idx=None,
+        **kwargs,
+    ) -> None:
+        """
+        num_additional_embeddings: int. Number of additional embeddings. Only useful when you `partially_freeze=True`.
+        partially_freeze: bool. If True, the regular `weight` will be frozen. `additional_weight` is never frozen.
+        Note: there are a lot of other parameters to initialize a standard `nn.Embedding` such as `padding_idx`, `max_norm` or `norm_type`. We are not supporting these.
+        """
+        if padding_idx is not None and padding_idx > num_embeddings:
+            raise ValueError(f"padding_idx must be within num_embeddings. Got {padding_idx} and {num_embeddings}")
+        super().__init__(
+            num_embeddings=num_embeddings,
+            embedding_dim=embedding_dim,
+            device=device,
+            dtype=dtype,
+            padding_idx=padding_idx,
+            **kwargs,
+        )
+        self.num_embeddings = num_embeddings
+        self.padding_idx = padding_idx
+        self.num_additional_embeddings = num_additional_embeddings
+        self.partially_freeze = partially_freeze
+
+        if partially_freeze:
+            self.weight.requires_grad_(False)
+
+        if self.num_additional_embeddings > 0:
+            self.additional_embedding = nn.Embedding(
+                num_embeddings=self.num_additional_embeddings,
+                embedding_dim=embedding_dim,
+                device=device,
+                dtype=dtype,
+            )
+
+    def forward(self, input_ids):
+        """
+        we have 2 embeddings, with different indices - one pretrained self.weight and another
+        self.additional_embedding.weight that is being trained.
+        in order to make a lookup of the input ids, we:
+        1. find out the indices of the entries belonging to the 2nd embedding
+        2. extract those values while subtracting the size of the first embedding (num_embeddings),
+           since the 2nd embedding starts from 0 and not num_embeddings
+        3. perform the 2nd embedding lookup
+        4. now we handle the 1st embedding, we overwrite indices belonging to the 2nd embedding with a padding index
+        5. perform the 1st embedding lookup
+        6. now we overwrite the values in the 1st embedding lookup with the values of the 2nd embedding lookup
+        note: for the 1st embedding lookup we could have looked up only the low indices and not do
+        the padding, but then we have to create a new tensor and populate it with 2 tensors that are
+        spread out across various indices - i.e. not a simple concat - I haven't benchmarked the
+        complex case if it's any faster, given that seqlens are usually relatively short it's
+        probably not faster or if faster not by much - but might be a good idea to measure.
+        """
+        if self.num_additional_embeddings == 0:
+            return self.additional_embedding(input_ids)
+
+        # Clone so that we don't modify the original input_ids later on
+        input_ids = input_ids.clone()
+        additional_vocab_indices = torch.where(input_ids >= self.num_embeddings)
+        input_ids_additional_vocab = input_ids[additional_vocab_indices]
+        additional_embeddings = self.additional_embedding(input_ids_additional_vocab - self.num_embeddings)
+
+        # for successful lookup replace input_ids with 0, the results of these will be discarded anyway
+        input_ids[additional_vocab_indices] = 0
+        full_vector = F.embedding(input_ids, self.weight)
+
+        # overwrite the records with high indices
+        full_vector[additional_vocab_indices] = additional_embeddings
+
+        return full_vector
+
+    def extra_repr(self) -> str:
+        return "num_embeddings={}, num_additional_embeddings={}, embedding_dim={}, partially_freeze={}".format(
+            self.num_embeddings,
+            self.num_additional_embeddings,
+            self.embedding_dim,
+            self.partially_freeze,
+        )
+
+@dataclass
+class VLlamaBaseModelOutputWithPast(BaseModelOutput):
+    """
+    Base class for VLlama3 model's outputs that may also contain a past key/values (to speed up sequential decoding).
+    Args:
+        last_hidden_state (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
+            Sequence of hidden-states at the output of the last layer of the model.
+            If `past_key_values` is used only the last hidden-state of the sequences of shape `(batch_size, 1,
+            hidden_size)` is output.
+        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+            Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
+            `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and optionally if
+            `config.is_encoder_decoder=True` 2 additional tensors of shape `(batch_size, num_heads,
+            encoder_sequence_length, embed_size_per_head)`.
+            Contains pre-computed hidden-states (key and values in the self-attention blocks and optionally if
+            `config.is_encoder_decoder=True` in the cross-attention blocks) that can be used (see `past_key_values`
+            input) to speed up sequential decoding.
+        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
+            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.
+            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
+        attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
+            Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
+            sequence_length)`.
+            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
+            heads.
+        image_hidden_states (`tuple(torch.FloatTensor)`, *optional*):
+            Tuple of `torch.FloatTensor` (one for the output of the image embeddings, `(batch_size, num_images,
+            sequence_length, hidden_size)`.
+            image_hidden_states of the model produced by the vision encoder, and optionally by the perceiver
+    """
+
+    last_hidden_state: torch.FloatTensor = None
+    past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None
+    hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None
+    attentions: Optional[Tuple[torch.FloatTensor, ...]] = None
+    image_hidden_states: Optional[Tuple[torch.FloatTensor]] = None
+
+@dataclass
+class VLlamaCausalLMOutputWithPast(BaseModelOutput):
+    """
+    Base class for VLlama3 causal language model (or autoregressive) outputs.
+    Args:
+        loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
+            Language modeling loss (for next-token prediction).
+        logits (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`):
+            Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
+        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+            Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
+            `(batch_size, num_heads, sequence_length, embed_size_per_head)`)
+            Contains pre-computed hidden-states (key and values in the self-attention blocks) that can be used (see
+            `past_key_values` input) to speed up sequential decoding.
+        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
+            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.
+            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
+        attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
+            Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
+            sequence_length)`.
+            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
+            heads.
+        image_hidden_states (`tuple(torch.FloatTensor)`, *optional*):
+            Tuple of `torch.FloatTensor` (one for the output of the image embeddings, `(batch_size, num_images,
+            sequence_length, hidden_size)`.
+            image_hidden_states of the model produced by the vision encoder, and optionally by the perceiver
+    """
+
+    loss: Optional[torch.FloatTensor] = None
+    logits: torch.FloatTensor = None
+    past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None
+    hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None
+    attentions: Optional[Tuple[torch.FloatTensor, ...]] = None
+    image_hidden_states: Optional[Tuple[torch.FloatTensor]] = None
+
+
+class VLlamaSimpleMLP(nn.Module):
+    def __init__(self, input_size, output_size):
+        super().__init__()
+        self.proj = nn.Linear(input_size, output_size, bias=False)
+
+    def forward(self, x):
+        return self.proj(x)
+
+class VLlamaConnector(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.scale_factor = config.pixel_shuffle_factor
+        self.modality_projection = VLlamaSimpleMLP(
+            input_size=config.vision_config.hidden_size * (config.scale_factor**2),
+            output_size=config.text_config.hidden_size
+        )
+
+    def pixel_shuffle(self, x, scale_factor):
+        bsz, seq, embed_dim = x.size()
+        height = width = int(seq**0.5)
+        x = x.view(bsz, height, width, embed_dim)
+        x = x.view(bsz, height, int(width / scale_factor), embed_dim * scale_factor)
+        x = x.permute(0, 2, 1, 3)
+        x = x.reshape(bsz, int(width / scale_factor), int(height / scale_factor), embed_dim * (scale_factor**2))
+        x = x.permute(0, 2, 1, 3)
+        x = x.reshape(bsz, int(seq / (scale_factor**2)), embed_dim * (scale_factor**2))
+        return x
+
+    def forward(self, image_hidden_states):
+        image_hidden_states = self.pixel_shuffle(image_hidden_states, self.scale_factor)
+        image_hidden_states = self.modality_projection(image_hidden_states)
+        return image_hidden_states
+
+class VLlamaPreTrainedModel(PreTrainedModel):
+    config_class = VLlamaConfig
+    base_model_prefix = "model"
+    supports_gradient_checkpointing = True
+    _no_split_modules = ["VLlamaDecoderLayer"]
+    _skip_keys_device_placement = "past_key_values"
+    _supports_flash_attn_2 = True
+    _supports_sdpa = True
+    _supports_cache_class = True
+
+    def _init_weights(self, module):
+        """Initialize the weights."""
+
+        std = (
+            self.config.initializer_range
+            if hasattr(self.config, "initializer_range")
+            else self.config.text_config.initializer_range
+        )
+
+        if hasattr(module, "class_embedding"):
+            module.class_embedding.data.normal_(mean=0.0, std=std)
+
+        if isinstance(module, (nn.Linear, nn.Conv2d)):
+            module.weight.data.normal_(mean=0.0, std=std)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.Embedding):
+            module.weight.data.normal_(mean=0.0, std=std)
+            if module.padding_idx is not None:
+                module.weight.data[module.padding_idx].zero_()
+
+class VLlamaModel(VLlamaPreTrainedModel):
+    """
+    A subclass of Idefics3Model. We do *not* remove or block the call to inputs_merger
+    in forward. Instead, we override inputs_merger here with custom logic.
+    """
+
+    def __init__(self, config: VLlamaConfig, **kwargs):
+        super().__init__(config)
+
+        self.vision_model = VLlamaModel.init_vision_model(config, **kwargs)
+        self.connector = VLlamaConnector(config)
+        self.text_model = VLlamaModel.init_language_model(config, **kwargs)
+
+        self.image_seq_len = int(
+            ((config.vision_config.image_size // config.vision_config.patch_size) ** 2) / (config.scale_factor**2)
+        )
+        self.image_token_id = self.config.image_token_id
+
+        self.post_init()
+
+    @staticmethod
+    def init_vision_model(config: VLlamaConfig, **kwargs):
+        vision_model_config = AutoConfig.from_pretrained(
+            config.vision_config.vision_model_name,
+            trust_remote_code=True,
+            **kwargs,
+        )
+
+        vision_model = AutoModel.from_config(vision_model_config, trust_remote_code=True, **kwargs)
+
+        if hasattr(vision_model, "vision_model"):
+            # If the model has a vision_model attribute, it means it's a wrapper around another model
+            vision_model = vision_model.vision_model
+
+        return vision_model
+
+    @staticmethod
+    def init_language_model(config: VLlamaConfig, **kwargs):
+        text_model_config = AutoConfig.from_pretrained(
+            config.text_config.text_model_name,
+            trust_remote_code=True,
+            **kwargs,
+        )
+
+        text_model = AutoModel.from_config(text_model_config, trust_remote_code=True, **kwargs)
+        # extractor = regex_lookup(language_model_name, language_model_name2model)
+
+        embed_layer = DecoupledEmbedding(
+            num_embeddings=text_model_config.vocab_size,
+            num_additional_embeddings=config.additional_vocab_size,
+            embedding_dim=config.hidden_size,
+            partially_freeze=config.freeze_config["freeze_text_layers"],
+            padding_idx=config.pad_token_id,
+        )
+
+        text_model.set_input_embeddings(embed_layer)
+
+        return text_model
+
+    def enable_input_require_grads(self):
+        """
+        Enables the gradients for the input embeddings.
+        This is useful for lora when using gradient checkpointing.
+        c.f. https://github.com/huggingface/peft/issues/1402#issuecomment-1913675032
+        Override to set output.requires_grad = True for both the decoder's and vision model's embeddings.
+        """
+
+        def get_lowest_module(module):
+            if len(list(module.children())) == 0:
+                # If the module has no children, it is a leaf module (e.g., Linear, Conv2d, etc.)
+                return module
+            else:
+                # Recursively call the function on each child module
+                return get_lowest_module(list(module.children())[0])
+
+        def make_inputs_require_grads(module, input, output):
+            output.requires_grad_(True)
+
+        self._text_require_grads_hook = self.get_input_embeddings().register_forward_hook(make_inputs_require_grads)
+        self._vision_require_grads_hook = get_lowest_module(self.vision_model).register_forward_hook(
+            make_inputs_require_grads
+        )
+
+    def disable_input_require_grads(self):
+        self._text_require_grads_hook.remove()
+        self._vision_require_grads_hook.remove()
+
+    def get_input_embeddings(self):
+        return self.text_model.get_input_embeddings()
+
+    def set_input_embeddings(self, value):
+        self.text_model.set_input_embeddings(value)
+
+    def inputs_merger(
+        self, input_ids: torch.LongTensor, inputs_embeds: torch.Tensor, image_hidden_states: torch.Tensor
+    ):
+        """
+        This method aims at merging the token embeddings with the image hidden states into one single sequence of vectors that are fed to the transformer LM.
+        The merging happens as follows:
+        - The text token sequence is: `tok_1 tok_2 tok_3 <fake_token_around_image> <image> <image> ... <image> <fake_token_around_image> tok_4`.
+        - We get the image hidden states for the image through the vision encoder and that hidden state, after a pixel shuffle operation, is then projected into the text embedding space.
+        We thus have a sequence of image hidden states of size (1, image_seq_len, hidden_dim), where 1 is for batch_size of 1 image and hidden_dim is the hidden_dim of the LM transformer.
+        - The merging happens so that we obtain the following sequence: `vector_tok_1 vector_tok_2 vector_tok_3 vector_fake_tok_around_image {sequence of image_seq_len image hidden states} vector_fake_toke_around_image vector_tok_4`. That sequence is fed to the LM.
+        - To fit the format of that sequence, `input_ids`, `input_embeds`, `attention_mask` are all 3 adapted to insert the image hidden states.
+        """
+        _, patch_size, _ = image_hidden_states.shape
+
+        image_mask = input_ids == self.image_token_id
+        num_image_tokens = image_mask.sum(dim=1)
+        if not torch.all(num_image_tokens % patch_size == 0):
+            raise ValueError("At least one sample has <image> tokens not divisible by patch_size.")
+
+        blocks_per_sample = num_image_tokens // patch_size
+
+        offsets = torch.nn.functional.pad(blocks_per_sample.cumsum(dim=0), (1, 0), value=0)
+        block_offset = offsets[:-1]
+        row_cum = image_mask.cumsum(dim=-1)
+        chunk_idx = (row_cum - 1) // patch_size
+        local_idx = (row_cum - 1) % patch_size
+        block_idx = block_offset.unsqueeze(1) + chunk_idx
+
+        image_embeds = torch.zeros_like(inputs_embeds)
+        image_embeds[image_mask] = image_hidden_states[block_idx[image_mask], local_idx[image_mask], :]
+
+        merged_embeds = torch.where(image_mask.unsqueeze(-1), image_embeds, inputs_embeds)
+        return merged_embeds
+    
+    def embed_tokens(self, input_ids: torch.LongTensor) -> torch.FloatTensor:
+        """
+        Override the embed_tokens method to use the text model's input embeddings.
+        This is necessary to ensure that the image token ID is correctly handled.
+        """
+        if self.text_model.get_input_embeddings() is None:
+            raise ValueError("The text model does not have input embeddings.")
+        
+        return self.text_model.get_input_embeddings()(input_ids).to(input_ids.device)
+
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[List[torch.FloatTensor]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        pixel_values: Optional[torch.FloatTensor] = None,
+        pixel_attention_mask: Optional[torch.BoolTensor] = None,
+        image_hidden_states: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        cache_position: Optional[torch.LongTensor] = None,
+    ) -> Union[Tuple, VLlamaBaseModelOutputWithPast]:    
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        use_cache = use_cache if use_cache is not None else self.config.use_cache
+
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        if (input_ids is None) ^ (inputs_embeds is not None):
+            raise ValueError(
+                "You cannot specify both input_ids and inputs_embeds at the same time, and must specify either one"
+            )
+
+        if self.training and self.text_model.gradient_checkpointing and use_cache:
+            logger.warning_once(
+                "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`."
+            )
+            use_cache = False
+
+        if inputs_embeds is None:
+            inputs_embeds = self.embed_tokens(input_ids)
+
+        if inputs_embeds is not None and input_ids is None:
+            raise ValueError("When first calling the model, if input_embeds are passed, input_ids should not be None.")
+
+        # START VISUAL INPUTS INTEGRATION
+        if pixel_values is not None and image_hidden_states is not None:
+            raise ValueError("You cannot specify both pixel_values and image_hidden_states at the same time")
+        elif pixel_values is not None:
+            batch_size, num_images, num_channels, height, width = pixel_values.shape
+            pixel_values = pixel_values
+            pixel_values = pixel_values.view(batch_size * num_images, *pixel_values.shape[2:])
+
+            # Remove padding images - padding images are full 0.
+            nb_values_per_image = pixel_values.shape[1:].numel()
+            real_images_inds = (pixel_values == 0.0).sum(dim=(-1, -2, -3)) != nb_values_per_image
+
+            if not any(real_images_inds):
+                # no images, leave one empty image.
+                real_images_inds[0] = True
+
+            pixel_values = pixel_values[real_images_inds].contiguous()
+
+            # Handle the vision attention mask
+            if pixel_attention_mask is None:
+                pixel_attention_mask = torch.ones(
+                    size=[pixel_values.shape[i] for i in (0, 2, 3)],
+                    dtype=torch.bool,
+                    device=pixel_values.device,
+                )
+            else:
+                # Remove padding images from the mask
+                pixel_attention_mask = pixel_attention_mask.view(
+                    batch_size * num_images, *pixel_attention_mask.shape[2:]
+                )
+                pixel_attention_mask = pixel_attention_mask[real_images_inds].contiguous()
+
+            # patch_size = self.config.vision_config.patch_size
+            # patches_subgrid = pixel_attention_mask.unfold(dimension=1, size=patch_size, step=patch_size)
+            # patches_subgrid = patches_subgrid.unfold(dimension=2, size=patch_size, step=patch_size)
+            # patch_attention_mask = (patches_subgrid.sum(dim=(-1, -2)) > 0).bool()
+
+            # Get sequence from the vision encoder
+            image_hidden_states = self.vision_model(
+                pixel_values=pixel_values,
+                # patch_attention_mask=patch_attention_mask,
+            ).last_hidden_state
+
+            # Modality projection & resampling
+            image_hidden_states = self.connector(image_hidden_states)
+
+        elif image_hidden_states is not None:
+            image_hidden_states = image_hidden_states.to(dtype=self.dtype, device=input_ids.device)
+
+        if inputs_embeds is not None and image_hidden_states is not None:
+            # When we embed, we don't want to replace the potential image_token_id that we generated by images
+            # that simply don't exist
+            inputs_embeds = self.inputs_merger(
+                input_ids=input_ids,
+                inputs_embeds=inputs_embeds,
+                image_hidden_states=image_hidden_states,
+            )
+
+        outputs = self.text_model(
+            inputs_embeds=inputs_embeds,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+            past_key_values=past_key_values,
+            use_cache=use_cache,
+            cache_position=cache_position,
+        )
+
+        if not return_dict:
+            return tuple(v for v in [*outputs, image_hidden_states] if v is not None)
+
+        return VLlamaBaseModelOutputWithPast(
+            last_hidden_state=outputs.last_hidden_state,
+            past_key_values=past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            image_hidden_states=image_hidden_states,
+        )
+
+class VLlamaForCausalLM(VLlamaPreTrainedModel):
+    # _tied_weights_keys = ["predictions.decoder.bias", "cls.predictions.decoder.weight"]
+
+    def __init__(self, config, **kwargs):
+        super().__init__(config)
+
+        self.image_token_id = config.image_token_id
+        self.in_features = config.hidden_size
+        self.out_additional_features = config.additional_vocab_size
+        self.vocab_size = config.vocab_size
+
+        self.model = VLlamaModel(config, **kwargs)
+        self.lm_head = VLlamaForCausalLM.init_lm_head(config, **kwargs)
+        if self.out_additional_features > 0:
+            self.additional_fc = nn.Linear(
+                in_features=self.in_features,
+                out_features=self.out_additional_features,
+                bias=False,
+            )
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    @staticmethod
+    def init_lm_head(config, **kwargs):
+        # Get the pretrained model config
+        text_model_config = AutoConfig.from_pretrained(
+            config.text_config.text_model_name,
+            trust_remote_code=True,
+            **kwargs,
+        )
+        model = AutoModelForMaskedLM.from_config(text_model_config, trust_remote_code=True, **kwargs)
+        # Get the lm head
+        lm_head = model.lm_head if hasattr(model, "lm_head") else model.decoder if hasattr(model, "decoder") else None
+        if lm_head is None:
+            logger.warning(f"No lm head was found for {config.text_config.text_model_name}, initializing a new one.")
+            lm_head = nn.Linear(config.hidden_size, config.vocab_size, False)
+        return lm_head
+
+    def forward(
+            self,
+            input_ids: torch.LongTensor = None,
+            attention_mask: Optional[torch.Tensor] = None,
+            position_ids: Optional[torch.LongTensor] = None,
+            past_key_values: Optional[List[torch.FloatTensor]] = None,
+            inputs_embeds: Optional[torch.FloatTensor] = None,
+            pixel_values: Optional[torch.FloatTensor] = None,
+            pixel_attention_mask: Optional[torch.BoolTensor] = None,
+            image_hidden_states: Optional[torch.FloatTensor] = None,
+            labels: Optional[torch.LongTensor] = None,
+            use_cache: Optional[bool] = None,
+            output_attentions: Optional[bool] = None,
+            output_hidden_states: Optional[bool] = None,
+            return_dict: Optional[bool] = None,
+            cache_position: Optional[torch.LongTensor] = None,
+    ) -> Union[Tuple, VLlamaCausalLMOutputWithPast]:
+        r"""
+        Args:
+            labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+                Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
+                config.vocab_size]` or `model.image_token_id` (where `model` is your instance of `Idefics3ForConditionalGeneration`).
+                Tokens with indices set to `model.image_token_id` are ignored (masked), the loss is only
+                computed for the tokens with labels in `[0, ..., config.vocab_size]`.
+        ```"""
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+
+        # Pass the inputs to VLlamaModel
+        outputs = self.model(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            pixel_values=pixel_values,
+            pixel_attention_mask=pixel_attention_mask,
+            image_hidden_states=image_hidden_states,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+            cache_position=cache_position,
+        )
+
+        # Pass the outputs to the MLM head
+        hidden_states = outputs[0]
+
+        logits = self.lm_head(hidden_states)
+        if self.out_additional_features > 0:
+            additional_features = self.additional_fc(hidden_states)
+            logits = torch.cat((logits, additional_features), -1)
+        logits = logits.float()
+
+        loss = None
+        if labels is not None:
+            # Shift so that tokens < n predict n
+            if attention_mask is not None:
+                shift_attention_mask = attention_mask[..., 1:]
+                shift_logits = logits[..., :-1, :][shift_attention_mask != 0].contiguous()
+                shift_labels = labels[..., 1:][shift_attention_mask != 0].contiguous()
+            else:
+                shift_logits = logits[..., :-1, :].contiguous()
+                shift_labels = labels[..., 1:].contiguous()
+            # Flatten the tokens
+            loss_fct = CrossEntropyLoss()
+            loss = loss_fct(shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1))
+
+        if not return_dict:
+            output = (logits,) + outputs[1:]
+            return (loss,) + output if loss is not None else output
+
+        return VLlamaCausalLMOutputWithPast(
+            loss=loss,
+            logits=logits,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            image_hidden_states=outputs.image_hidden_states,
+        )
+
+class VLlamaForVision2Seq(VLlamaPreTrainedModel, GenerationMixin):
+    def __init__(self, config, **kwargs):
+        super().__init__(config)
+
+        self.image_token_id = config.image_token_id
+        self.in_features = config.hidden_size
+        self.out_additional_features = config.additional_vocab_size
+        self.vocab_size = config.vocab_size
+
+        self.model = VLlamaModel(config, **kwargs)
+        self.lm_head = VLlamaForVision2Seq.init_lm_head(config, **kwargs)
+        if self.out_additional_features > 0:
+            self.additional_fc = nn.Linear(
+                in_features=self.in_features,
+                out_features=self.out_additional_features,
+                bias=False,
+            )
+
+        self.loss_fct = CrossEntropyLoss()
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    @staticmethod
+    def init_lm_head(config, **kwargs):
+        # Get the pretrained model config
+        text_model_config = AutoConfig.from_pretrained(
+            config.text_config.text_model_name,
+            trust_remote_code=True,
+            **kwargs,
+        )
+        model = AutoModelForMaskedLM.from_config(text_model_config, trust_remote_code=True, **kwargs)
+        # Get the lm head
+        lm_head = model.lm_head if hasattr(model, "lm_head") else model.decoder if hasattr(model, "decoder") else None
+        if lm_head is None:
+            logger.warning(f"No lm head was found for {config.text_config.text_model_name}, initializing a new one.")
+            lm_head = nn.Linear(config.hidden_size, config.vocab_size, False)
+        return lm_head
+
+    def enable_input_require_grads(self):
+        """
+        Enables the gradients for the input embeddings. This is useful for fine-tuning adapter weights while keeping
+        the model weights fixed.
+        """
+
+        def make_inputs_require_grads(module, input, output):
+            output.requires_grad_(True)
+
+        self._text_require_grads_hook = self.get_input_embeddings().register_forward_hook(make_inputs_require_grads)
+        self._vision_require_grads_hook = self.model.vision_model.get_input_embeddings().register_forward_hook(
+            make_inputs_require_grads
+        )
+
+    def disable_input_require_grads(self):
+        self._text_require_grads_hook.remove()
+        self._vision_require_grads_hook.remove()
+
+    def get_input_embeddings(self):
+        return self.model.text_model.get_input_embeddings()
+
+    def set_input_embeddings(self, value):
+        self.model.text_model.set_input_embeddings(value)
+
+    def get_output_embeddings(self):
+        return self.lm_head
+
+    def set_output_embeddings(self, new_embeddings):
+        self.lm_head = new_embeddings
+
+    def forward(
+        self,
+        input_ids: Optional[torch.LongTensor] = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[List[torch.FloatTensor]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        pixel_values: Optional[torch.FloatTensor] = None,
+        pixel_attention_mask: Optional[torch.BoolTensor] = None,
+        image_hidden_states: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        cache_position: Optional[torch.LongTensor] = None,
+        return_dict: Optional[bool] = None,
+        logits_to_keep: Union[int, torch.Tensor] = 0,
+        **kwargs: Unpack[KwargsForCausalLM],
+    ) -> Union[Tuple, VLlamaCausalLMOutputWithPast]:
+        r"""
+        pixel_attention_mask (`torch.Tensor` of shape `(batch_size, image_size, image_size)`, *optional*):
+            Mask to avoid performing attention on padding pixel indices.
+        image_hidden_states (`torch.FloatTensor` of shape `(batch_size, num_channels, image_size, image_size)`):
+            The hidden states of the image encoder after modality projection.
+        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
+            config.vocab_size]` or `model.image_token_id` (where `model` is your instance of `SmolVLMForConditionalGeneration`).
+            Tokens with indices set to `model.image_token_id` are ignored (masked), the loss is only
+            computed for the tokens with labels in `[0, ..., config.vocab_size]`.
+
+        Example:
+
+        ```python
+        >>> import requests
+        >>> import torch
+        >>> from PIL import Image
+        >>> from io import BytesIO
+
+        >>> from transformers import AutoProcessor, AutoModelForImageTextToText
+        >>> from transformers.image_utils import load_image
+
+        >>> # Note that passing the image urls (instead of the actual pil images) to the processor is also possible
+        >>> image1 = load_image("https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg")
+        >>> image2 = load_image("https://cdn.britannica.com/59/94459-050-DBA42467/Skyline-Chicago.jpg")
+        >>> image3 = load_image("https://cdn.britannica.com/68/170868-050-8DDE8263/Golden-Gate-Bridge-San-Francisco.jpg")
+
+        >>> processor = AutoProcessor.from_pretrained("HuggingFaceTB/SmolVLM2-2.2B-Instruct")
+        >>> model = AutoModelForImageTextToText.from_pretrained("HuggingFaceTB/SmolVLM2-2.2B-Instruct", torch_dtype=torch.bfloat16, device_map="auto")
+
+        >>> # Create inputs
+        >>> messages = [
+        ...     {
+        ...         "role": "user",
+        ...         "content": [
+        ...             {"type": "video", "path": path/to/video},
+        ...             {"type": "text", "text": "What is happening in this video?"},
+        ...         ]
+        ...     }
+        ... ]
+
+        >>> inputs = processor.apply_chat_template([messages], add_generation_prompt=True)
+
+        >>> # Generate
+        >>> generated_ids = model.generate(**inputs, max_new_tokens=256)
+        >>> generated_texts = processor.batch_decode(generated_ids, skip_special_tokens=True)
+
+        >>> print(generated_texts)
+        ```"""
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        # decoder outputs consists of (dec_features, layer_state, dec_hidden, dec_attn)
+        outputs = self.model(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            pixel_values=pixel_values,
+            pixel_attention_mask=pixel_attention_mask,
+            image_hidden_states=image_hidden_states,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            cache_position=cache_position,
+            return_dict=True,
+            **kwargs,
+        )
+
+        hidden_states = outputs[0]
+        slice_indices = slice(-logits_to_keep, None) if isinstance(logits_to_keep, int) else logits_to_keep
+        # Only compute necessary logits, and do not upcast them to float if we are not computing the loss
+        hidden_states = hidden_states[:, slice_indices, :]
+        logits = self.lm_head(hidden_states)
+        if self.out_additional_features > 0:
+            additional_features = self.additional_fc(hidden_states)
+            logits = torch.cat((logits, additional_features), -1)
+        logits = logits.float()
+
+        loss = None
+        if labels is not None:
+            loss = self.loss_fct(
+                logits=logits, labels=labels, vocab_size=self.config.text_config.vocab_size, **kwargs
+            )
+
+        return VLlamaCausalLMOutputWithPast(
+            loss=loss,
+            logits=logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            image_hidden_states=outputs.image_hidden_states,
+        )
+
+    def prepare_inputs_for_generation(
+        self,
+        input_ids,
+        past_key_values=None,
+        attention_mask=None,
+        inputs_embeds=None,
+        cache_position=None,
+        pixel_values=None,
+        pixel_attention_mask=None,
+        image_hidden_states=None,
+        logits_to_keep=None,
+        **kwargs,
+    ):
+        # Overwritten -- there are mutually exclusive inputs (if the logic to make `image_hidden_states` take
+        # precedence is moved to the model, we can remove this fn)
+
+        model_inputs = super().prepare_inputs_for_generation(
+            input_ids,
+            past_key_values=past_key_values,
+            attention_mask=attention_mask,
+            inputs_embeds=inputs_embeds,
+            cache_position=cache_position,
+            pixel_values=pixel_values,
+            pixel_attention_mask=pixel_attention_mask,
+            image_hidden_states=image_hidden_states,
+            logits_to_keep=logits_to_keep,
+            **kwargs,
+        )
+
+        # if `inputs_embeds` are passed, we only want to use them in the 1st generation step
+        # but IDEFICS requires both ids and embeds to be present
+        if inputs_embeds is not None and cache_position[0] == 0:
+            model_inputs["input_ids"] = input_ids
+
+        if image_hidden_states is not None:
+            model_inputs["pixel_values"] = None
+            model_inputs["pixel_attention_mask"] = None
+
+        return model_inputs
+
+    def _update_model_kwargs_for_generation(self, outputs, model_kwargs, is_encoder_decoder, **kwargs):
+        model_kwargs = super()._update_model_kwargs_for_generation(
+            outputs=outputs,
+            model_kwargs=model_kwargs,
+            is_encoder_decoder=is_encoder_decoder,
+            **kwargs,
+        )
+        # Get the precomputed image_hidden_states
+        model_kwargs["image_hidden_states"] = outputs.image_hidden_states
+        return model_kwargs
+    
+    @staticmethod
+    # Copied from transformers.models.llama.modeling_llama.LlamaForCausalLM._reorder_cache
+    def _reorder_cache(past_key_values, beam_idx):
+        reordered_past = ()
+        for layer_past in past_key_values:
+            reordered_past += (
+                tuple(past_state.index_select(0, beam_idx.to(past_state.device)) for past_state in layer_past),
+            )
+        return reordered_past

vlm-siglip2-sllm_210/opt_step-18000__merged/preprocessor_config.json

@@ -0,0 +1,28 @@
+{
+    "do_convert_rgb": true,
+    "do_image_splitting": true,
+    "do_normalize": true,
+    "do_pad": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "image_mean": [
+        0.5,
+        0.5,
+        0.5
+    ],
+    "image_processor_type": "Idefics3ImageProcessor",
+    "image_std": [
+        0.5,
+        0.5,
+        0.5
+    ],
+    "max_image_size": {
+        "longest_edge": 512
+    },
+    "processor_class": "Idefics3Processor",
+    "resample": 1,
+    "rescale_factor": 0.00392156862745098,
+    "size": {
+        "longest_edge": 2048
+    }
+}

vlm-siglip2-sllm_210/opt_step-18000__merged/processor_config.json

@@ -0,0 +1,4 @@
+{
+    "image_seq_len": 64,
+    "processor_class": "Idefics3Processor"
+}

vlm-siglip2-sllm_210/opt_step-18000__merged/special_tokens_map.json

@@ -0,0 +1,72 @@
+{
+  "additional_special_tokens": [
+    "<global-img>",
+    "<row_1_col_1>",
+    "<row_1_col_2>",
+    "<row_1_col_3>",
+    "<row_1_col_4>",
+    "<row_1_col_5>",
+    "<row_1_col_6>",
+    "<row_2_col_1>",
+    "<row_2_col_2>",
+    "<row_2_col_3>",
+    "<row_2_col_4>",
+    "<row_2_col_5>",
+    "<row_2_col_6>",
+    "<row_3_col_1>",
+    "<row_3_col_2>",
+    "<row_3_col_3>",
+    "<row_3_col_4>",
+    "<row_3_col_5>",
+    "<row_3_col_6>",
+    "<row_4_col_1>",
+    "<row_4_col_2>",
+    "<row_4_col_3>",
+    "<row_4_col_4>",
+    "<row_4_col_5>",
+    "<row_4_col_6>",
+    "<row_5_col_1>",
+    "<row_5_col_2>",
+    "<row_5_col_3>",
+    "<row_5_col_4>",
+    "<row_5_col_5>",
+    "<row_5_col_6>",
+    "<row_6_col_1>",
+    "<row_6_col_2>",
+    "<row_6_col_3>",
+    "<row_6_col_4>",
+    "<row_6_col_5>",
+    "<row_6_col_6>",
+    "<end_of_utterance>",
+    "<fake_token_around_image>",
+    "<image>"
+  ],
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|end_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "<|reserved_special_token_0|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|end_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

vlm-siglip2-sllm_210/opt_step-18000__merged/tokenizer_config.json

@@ -0,0 +1,2429 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|finetune_right_pad_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|eom_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|python_tag|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128256": {
+      "content": "<global-img>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128257": {
+      "content": "<row_1_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128258": {
+      "content": "<row_1_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128259": {
+      "content": "<row_1_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128260": {
+      "content": "<row_1_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128261": {
+      "content": "<row_1_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128262": {
+      "content": "<row_1_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128263": {
+      "content": "<row_2_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128264": {
+      "content": "<row_2_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128265": {
+      "content": "<row_2_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128266": {
+      "content": "<row_2_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128267": {
+      "content": "<row_2_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128268": {
+      "content": "<row_2_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128269": {
+      "content": "<row_3_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128270": {
+      "content": "<row_3_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128271": {
+      "content": "<row_3_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128272": {
+      "content": "<row_3_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128273": {
+      "content": "<row_3_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128274": {
+      "content": "<row_3_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128275": {
+      "content": "<row_4_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128276": {
+      "content": "<row_4_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128277": {
+      "content": "<row_4_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128278": {
+      "content": "<row_4_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128279": {
+      "content": "<row_4_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128280": {
+      "content": "<row_4_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128281": {
+      "content": "<row_5_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128282": {
+      "content": "<row_5_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128283": {
+      "content": "<row_5_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128284": {
+      "content": "<row_5_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128285": {
+      "content": "<row_5_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128286": {
+      "content": "<row_5_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128287": {
+      "content": "<row_6_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128288": {
+      "content": "<row_6_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128289": {
+      "content": "<row_6_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128290": {
+      "content": "<row_6_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128291": {
+      "content": "<row_6_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128292": {
+      "content": "<row_6_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128293": {
+      "content": "<end_of_utterance>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128294": {
+      "content": "<fake_token_around_image>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128295": {
+      "content": "<image>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<global-img>",
+    "<row_1_col_1>",
+    "<row_1_col_2>",
+    "<row_1_col_3>",
+    "<row_1_col_4>",
+    "<row_1_col_5>",
+    "<row_1_col_6>",
+    "<row_2_col_1>",
+    "<row_2_col_2>",
+    "<row_2_col_3>",
+    "<row_2_col_4>",
+    "<row_2_col_5>",
+    "<row_2_col_6>",
+    "<row_3_col_1>",
+    "<row_3_col_2>",
+    "<row_3_col_3>",
+    "<row_3_col_4>",
+    "<row_3_col_5>",
+    "<row_3_col_6>",
+    "<row_4_col_1>",
+    "<row_4_col_2>",
+    "<row_4_col_3>",
+    "<row_4_col_4>",
+    "<row_4_col_5>",
+    "<row_4_col_6>",
+    "<row_5_col_1>",
+    "<row_5_col_2>",
+    "<row_5_col_3>",
+    "<row_5_col_4>",
+    "<row_5_col_5>",
+    "<row_5_col_6>",
+    "<row_6_col_1>",
+    "<row_6_col_2>",
+    "<row_6_col_3>",
+    "<row_6_col_4>",
+    "<row_6_col_5>",
+    "<row_6_col_6>",
+    "<end_of_utterance>",
+    "<fake_token_around_image>",
+    "<image>"
+  ],
+  "bos_token": "<|begin_of_text|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|end_of_text|>",
+  "extra_special_tokens": {},
+  "legacy": false,
+  "mask_token": "<|reserved_special_token_0|>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask",
+    "pixel_values",
+    "pixel_attention_mask"
+  ],
+  "model_max_length": 8192,
+  "pad_token": "<|end_of_text|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}

vlm-siglip2-sllm_210/opt_step-2000/resume_run_infos.json

@@ -0,0 +1,91 @@
+{
+  "train_logs": {
+    "lr": 0.0001,
+    "num_opt_steps": 2000,
+    "num_epochs": 0,
+    "per_token_loss": {
+      "sft": 1.3265778124332428,
+      "all": 1.3265778124332428
+    },
+    "z_loss": {
+      "sft": 0.0,
+      "all": 0.0
+    },
+    "watt/s": {
+      "sft": 311.4078322082747,
+      "all": 311.4078322082747
+    },
+    "tflops": {
+      "sft": 11.076079297876293,
+      "all": 11.076079297876293
+    },
+    "tflop_counter": {
+      "sft": 1830048.8243103027,
+      "all": 1830048.8243103027
+    },
+    "fwd_bwd_time": {
+      "sft": 154524.75284576416,
+      "all": 154524.75284576416
+    },
+    "tflops_acc": {
+      "sft": 11.843078798753556,
+      "all": 11.843078798753556
+    },
+    "num_per_device_batches": {
+      "sft": 128000,
+      "all": 128000
+    },
+    "num_images": {
+      "sft": 7387449,
+      "all": 7387449
+    },
+    "num_image_tokens": {
+      "sft": 472796736,
+      "all": 472796736
+    },
+    "num_tokens": {
+      "sft": 235936850,
+      "all": 235936850
+    },
+    "image_to_text_ratio": {
+      "sft": 0.046299852430820465,
+      "all": 0.046299852430820465
+    },
+    "pixel_values_sum": {
+      "sft": 1924372547680.0,
+      "all": 1924372547680.0
+    },
+    "num_padding": {
+      "sft": 128449227,
+      "all": 128449227
+    },
+    "num_per_device_batches_in_curr_epoch": {
+      "sft": 128000,
+      "all": 128000
+    },
+    "num_batches": {
+      "all": 128000
+    },
+    "num_batches_in_curr_epoch": {
+      "all": 128000
+    },
+    "per_token_loss_acc": {},
+    "z_loss_acc": {},
+    "num_batches_since_training_logged": {},
+    "num_per_device_batches_since_training_logged": {},
+    "tflop_counter_since_training_logged": {},
+    "total_energy_delta_since_training_logged": {},
+    "fwd_bwd_time_since_training_logged": {},
+    "global_batch_size_current": 256
+  },
+  "wandb_run_id": "hp0rcdxj",
+  "seed": 42,
+  "resume_opt_step": 2000,
+  "resume_epoch": 0,
+  "gbs_running": {
+    "global_seen_samples": 512000,
+    "global_batch_size_current": 256,
+    "next_goal_samples": 0,
+    "grad_acc_size_current": 4
+  }
+}

vlm-siglip2-sllm_210/opt_step-20000/resume_run_infos.json

@@ -0,0 +1,91 @@
+{
+  "train_logs": {
+    "lr": 0.0001,
+    "num_opt_steps": 20000,
+    "num_epochs": 0,
+    "per_token_loss": {
+      "sft": 0.8643546402454376,
+      "all": 0.8643546402454376
+    },
+    "z_loss": {
+      "sft": 0.0,
+      "all": 0.0
+    },
+    "watt/s": {
+      "sft": 289.80214395771384,
+      "all": 289.80214395771384
+    },
+    "tflops": {
+      "sft": 10.133296464837661,
+      "all": 10.133296464837661
+    },
+    "tflop_counter": {
+      "sft": 18270358.57293701,
+      "all": 18270358.57293701
+    },
+    "fwd_bwd_time": {
+      "sft": 1541407.9232234955,
+      "all": 1541407.9232234955
+    },
+    "tflops_acc": {
+      "sft": 11.853032735636141,
+      "all": 11.853032735636141
+    },
+    "num_per_device_batches": {
+      "sft": 1280000,
+      "all": 1280000
+    },
+    "num_images": {
+      "sft": 73887963,
+      "all": 73887963
+    },
+    "num_image_tokens": {
+      "sft": 4728829632,
+      "all": 4728829632
+    },
+    "num_tokens": {
+      "sft": 2352720718,
+      "all": 2352720718
+    },
+    "image_to_text_ratio": {
+      "sft": 0.03385593742132187,
+      "all": 0.03385593742132187
+    },
+    "pixel_values_sum": {
+      "sft": 19224060293104.0,
+      "all": 19224060293104.0
+    },
+    "num_padding": {
+      "sft": 1276515058,
+      "all": 1276515058
+    },
+    "num_per_device_batches_in_curr_epoch": {
+      "sft": 1280000,
+      "all": 1280000
+    },
+    "num_batches": {
+      "all": 1280000
+    },
+    "num_batches_in_curr_epoch": {
+      "all": 1280000
+    },
+    "per_token_loss_acc": {},
+    "z_loss_acc": {},
+    "num_batches_since_training_logged": {},
+    "num_per_device_batches_since_training_logged": {},
+    "tflop_counter_since_training_logged": {},
+    "total_energy_delta_since_training_logged": {},
+    "fwd_bwd_time_since_training_logged": {},
+    "global_batch_size_current": 256
+  },
+  "wandb_run_id": "hp0rcdxj",
+  "seed": 42,
+  "resume_opt_step": 20000,
+  "resume_epoch": 0,
+  "gbs_running": {
+    "global_seen_samples": 5120000,
+    "global_batch_size_current": 256,
+    "next_goal_samples": 0,
+    "grad_acc_size_current": 4
+  }
+}

vlm-siglip2-sllm_210/opt_step-20000__merged/chat_template.jinja

@@ -0,0 +1,2 @@
+<|begin_of_text|>{% for message in messages %}{{message['role'] | capitalize}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>
+{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}

vlm-siglip2-sllm_210/opt_step-20000__merged/chat_template.json

@@ -0,0 +1,3 @@
+{
+    "chat_template": "<|begin_of_text|>{% for message in messages %}{{message['role'] | capitalize}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>\n{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}"
+}

vlm-siglip2-sllm_210/opt_step-20000__merged/config.json

@@ -0,0 +1,51 @@
+{
+  "additional_vocab_size": 40,
+  "architectures": [
+    "VLlamaForCausalLM"
+  ],
+  "auto_map": {
+    "AutoConfig": "configuration_vllama.VLlamaConfig",
+    "AutoModel": "modeling_vllama.VLlamaModel",
+    "AutoModelForCausalLM": "modeling_vllama.VLlamaForCausalLM",
+    "AutoModelForVision2Seq": "modeling_vllama.VLlamaForVision2Seq"
+  },
+  "freeze_config": {
+    "freeze_lm_head": true,
+    "freeze_text_layers": true,
+    "freeze_vision_layers": true
+  },
+  "hidden_size": 768,
+  "image_token_id": 128295,
+  "initializer_range": 0.02,
+  "max_position_embeddings": 8192,
+  "model_type": "VLlama",
+  "neftune_noise_alpha": 0.0,
+  "output_attentions": false,
+  "pixel_shuffle_factor": 4,
+  "qk_layer_norms": false,
+  "scale_factor": 4,
+  "text_config": {
+    "hidden_size": 768,
+    "intermediate_size": 3072,
+    "mlp_bias": false,
+    "model_type": "VLlama",
+    "num_hidden_layers": 12,
+    "text_model_name": "SmolVEncoder/decoder-210m",
+    "vocab_size": 128256
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": null,
+  "use_cache": true,
+  "use_resampler": false,
+  "vision_config": {
+    "embed_dim": 768,
+    "image_size": 512,
+    "intermediate_size": 3072,
+    "model_type": "VLlama",
+    "num_hidden_layers": 12,
+    "patch_size": 16,
+    "vision_model_name": "google/siglip2-base-patch16-512"
+  },
+  "vocab_size": 128256
+}

vlm-siglip2-sllm_210/opt_step-20000__merged/configuration_vllama.py

@@ -0,0 +1,233 @@
+import copy
+import os
+
+from typing import Union, Any, Dict
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+from transformers import AutoConfig
+
+logger = logging.get_logger(__name__)
+
+def collect_arg_in_candidates(config, candidates, default = None) -> Any:
+    """ Gets the argument in a config given a list of candidates """
+    for c in candidates:
+        if hasattr(config, c):
+            return getattr(config, c)
+        elif c in config:
+            return config[c]
+    if default is not None:
+        return default
+    raise ValueError("No matching arguments found in candidates. Candidates: {}, Config: {}".format(candidates, config))
+
+class VLlamaTextConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`LlamaModel`]. It is used to instantiate an LLaMA
+    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 LLaMA-7B.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        embed_dim (`int`, *optional*, defaults to 1152):
+            Dimensionality of the encoder layers and the pooler layer. (elsewhere referred to as `embed_dim`)
+        image_size (`int`, *optional*, defaults to 384):
+            The size (resolution) of each image.
+    """
+    model_type = "VLlama"
+
+    def __init__(
+        self,
+        # Case for when vllama3 is from the hub with no vision_model_name
+        text_model_name="HuggingFaceTB/SmolLM2-135M-Instruct",
+        **kwargs,
+    ):
+        self.text_model_name = text_model_name
+        text_config = AutoConfig.from_pretrained(text_model_name, trust_remote_code=True)
+        if hasattr(text_config, "text_config"):
+            text_config = text_config.text_config
+        
+        self.hidden_size = collect_arg_in_candidates(text_config, ["hidden_size", "embed_dim"])
+        self.num_hidden_layers = collect_arg_in_candidates(text_config, ["num_hidden_layers", "num_hidden_blocks"])
+        self.intermediate_size = collect_arg_in_candidates(text_config, ["intermediate_size", "mlp_dim"])
+        self.mlp_bias = collect_arg_in_candidates(text_config, ["mlp_bias", "mlp_hidden_bias"], default = False)
+        self.vocab_size = collect_arg_in_candidates(text_config, ["vocab_size"])
+    
+        super().__init__(text_model_name=text_model_name, **kwargs)
+
+class VLlamaVisionConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`LlamaModel`]. It is used to instantiate an LLaMA
+    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 LLaMA-7B.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        embed_dim (`int`, *optional*, defaults to 1152):
+            Dimensionality of the encoder layers and the pooler layer. (elsewhere referred to as `embed_dim`)
+        image_size (`int`, *optional*, defaults to 384):
+            The size (resolution) of each image.
+    """
+    model_type = "VLlama"
+    attribute_map = {
+        "hidden_size": "embed_dim",
+    }
+
+    def __init__(
+        self,
+        # Case for when vllama3 is from the hub with no vision_model_name
+        vision_model_name="google/siglip2-base-patch16-512",
+        **kwargs,
+    ):
+        self.vision_model_name = vision_model_name
+        vision_config = AutoConfig.from_pretrained(vision_model_name, trust_remote_code=True)
+        if hasattr(vision_config, "vision_config"):
+            vision_config = vision_config.vision_config
+
+        self.embed_dim = collect_arg_in_candidates(vision_config, ["embed_dim", "hidden_size"])
+        self.image_size = collect_arg_in_candidates(vision_config, ["image_size", "img_size"])
+        self.patch_size = collect_arg_in_candidates(vision_config, ["patch_size"])
+        self.num_hidden_layers = collect_arg_in_candidates(vision_config, ["num_hidden_layers", "num_hidden_blocks"])
+        self.intermediate_size = collect_arg_in_candidates(vision_config, ["intermediate_size", "mlp_dim"])
+
+        super().__init__(vision_model_name=vision_model_name, **kwargs)
+
+class VLlamaConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`SmolVLMModel`]. It is used to instantiate a
+    SmolVLM 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 model of the SmolVLM
+    [HuggingFaceTB/SmolVLM2-2.2B-Instruct](https://huggingface.co/HuggingFaceTB/SmolVLM2-2.2B-Instruct) architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should cache the key/value pairs of the attention mechanism. Only
+            relevant if `config.is_decoder=True`.
+        image_token_id (`int`, *optional*, defaults to 128257):
+            The id of the "image" token.
+        tie_word_embeddings (`bool`, *optional*, defaults to `False`):
+            Whether or not to tie the word embeddings with the token embeddings.
+        vision_config (`IdeficsVisionConfig` or `dict`, *optional*, defaults to `IdeficsVisionConfig`):
+            Custom vision config or dict for the vision tower
+        text_config (`PretrainedConfig` or `dict`, *optional*, defaults to `LlamaConfig`):
+            Custom text config or dict for the text model
+        scale_factor (`int`, *optional*, defaults to 2):
+            The scale factor for the image encoder.
+        pad_token_id (`int`, *optional*, defaults to 128002):
+            The id of the padding token.
+
+    Example:
+    ```python
+    >>> from transformers import SmolVLMModel, SmolVLMConfig
+    >>> # Initializing configuration
+    >>> configuration = SmolVLMConfig()
+    >>> # Initializing a model from the configuration
+    >>> model = SmolVLMModel(configuration)
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "VLlama"
+    is_composition = True
+    # sub_configs = {"text_config": VLlamaTextConfig, "vision_config": VLlamaVisionConfig}
+
+    DEFAULT_TEXT_MODEL_NAME = "EuroBERT/EuroBERT-210m"
+    DEFAULT_VISION_MODEL_NAME = "google/siglip2-base-patch16-512"
+
+    def __init__(
+        self,
+        text_config: Union[PretrainedConfig, Dict[str, Any]] = None,
+        vision_config: Union[PretrainedConfig, Dict[str, Any]] = None,
+        image_token_id: int = 128_257,
+        vocab_size=128_256,
+        use_cache = True,
+        tie_word_embeddings = False,
+        freeze_config = None,
+        pad_token_id = None,
+        initializer_range = 0.02,
+        pixel_shuffle_factor = 4,
+        use_resampler = False,
+        additional_vocab_size = 0,
+        neftune_noise_alpha = 0.0,
+        **kwargs,
+    ):        
+        self.image_token_id = image_token_id
+        self.use_cache = use_cache
+        self.tie_word_embeddings = tie_word_embeddings
+        self.scale_factor = pixel_shuffle_factor
+        self.additional_vocab_size = additional_vocab_size
+
+        if text_config is None:
+            text_config = AutoConfig.from_pretrained(self.DEFAULT_TEXT_MODEL_NAME, trust_remote_code=True)
+        elif isinstance(text_config, dict):
+            text_config = VLlamaTextConfig(text_config["text_model_name"])
+        self.text_config = text_config
+        
+        if vision_config is None:
+            vision_config = AutoConfig.from_pretrained(self.DEFAULT_VISION_MODEL_NAME, trust_remote_code=True)
+        elif isinstance(vision_config, dict):
+            vision_config = VLlamaVisionConfig(vision_config["vision_model_name"])
+        self.vision_config = vision_config
+
+        self.freeze_config = freeze_config
+
+        # Pixel shuffle factor
+        self.pixel_shuffle_factor = pixel_shuffle_factor
+        self.use_resampler = use_resampler
+
+        self.neftune_noise_alpha = neftune_noise_alpha
+
+        self.initializer_range = initializer_range
+
+        hidden_size = kwargs.pop("hidden_size", self.text_config.hidden_size)
+
+        super().__init__(
+            **kwargs, 
+            pad_token_id=pad_token_id, 
+            tie_word_embeddings=tie_word_embeddings,
+            vocab_size=vocab_size,
+            hidden_size=hidden_size,
+        )
+
+    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["model_type"] = self.__class__.model_type
+        output["vision_config"] = self.vision_config.to_dict()
+        output["text_config"] = self.text_config.to_dict()
+        # output["freeze_config"] = self.freeze_config.to_dict()
+
+        return output
+    
+    # @classmethod
+    # def from_pretrained(cls, pretrained_model_name_or_path: Union[str, os.PathLike], **kwargs) -> "PretrainedConfig":
+    #     outputs = super(VLlamaConfig, cls).from_pretrained(pretrained_model_name_or_path, **kwargs)
+    #     return outputs
+
+    @classmethod
+    def from_pretrained_models(
+        cls, 
+        text_model_name: Union[str, os.PathLike],
+        vision_model_name: Union[str, os.PathLike],
+        **kwargs
+    ) -> "PretrainedConfig":
+        # text_model_config = AutoConfig.from_pretrained(text_model_name, trust_remote_code=True)
+        # vision_model_config = AutoConfig.from_pretrained(vision_model_name, trust_remote_code=True)
+        text_model_config = VLlamaTextConfig(text_model_name)
+        vision_model_config = VLlamaVisionConfig(vision_model_name)
+        return cls(
+            text_config=text_model_config,
+            vision_config=vision_model_config,
+            **kwargs
+        )

vlm-siglip2-sllm_210/opt_step-20000__merged/modeling_vllama.py

@@ -0,0 +1,895 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from torch.nn import CrossEntropyLoss
+from typing import Optional, Tuple, Union, List
+
+# from transformers.models.smolvlm import SmolVLMModel, SmolVLMPreTrainedModel
+
+from .configuration_vllama import VLlamaConfig
+
+from transformers import AutoModel, AutoConfig, AutoModelForMaskedLM, GenerationMixin
+from transformers.cache_utils import Cache
+from transformers.modeling_utils import PreTrainedModel
+from transformers.modeling_outputs import BaseModelOutput
+from transformers.models.bert.modeling_bert import BaseModelOutputWithPoolingAndCrossAttentions, MaskedLMOutput
+from transformers.models.idefics3.modeling_idefics3 import Idefics3VisionTransformer
+from transformers.processing_utils import Unpack
+from transformers.modeling_flash_attention_utils import FlashAttentionKwargs
+from transformers.utils import LossKwargs
+
+from typing import List, Optional, Tuple, Union
+
+import torch
+import torch.utils.checkpoint
+
+from dataclasses import dataclass
+
+from transformers import logging 
+from transformers.utils import ContextManagers
+
+logger = logging.get_logger(__name__)
+
+class KwargsForCausalLM(FlashAttentionKwargs, LossKwargs): ...
+
+class DecoupledEmbedding(nn.Embedding):
+    # Derived from https://pytorch.org/docs/stable/_modules/torch/nn/modules/sparse.html#Embedding
+    """
+    Implements a decoupling of parameters to allow freezing (or not) a subset of the embeddings.
+    In practise, the regular `weight` can be trained or frozen (i.e. `partially_freeze=True`), and if `num_additional_embeddings` > 0, then it will create `num_additional_embeddings` additional parameters that are always trained.
+    If `num_additional_embeddings=0`, then the module defaults back to the regular behavior of `nn.Embedding`.
+    """
+
+    def __init__(
+        self,
+        num_embeddings,
+        num_additional_embeddings,
+        embedding_dim,
+        partially_freeze=False,
+        device=None,
+        dtype=None,
+        padding_idx=None,
+        **kwargs,
+    ) -> None:
+        """
+        num_additional_embeddings: int. Number of additional embeddings. Only useful when you `partially_freeze=True`.
+        partially_freeze: bool. If True, the regular `weight` will be frozen. `additional_weight` is never frozen.
+        Note: there are a lot of other parameters to initialize a standard `nn.Embedding` such as `padding_idx`, `max_norm` or `norm_type`. We are not supporting these.
+        """
+        if padding_idx is not None and padding_idx > num_embeddings:
+            raise ValueError(f"padding_idx must be within num_embeddings. Got {padding_idx} and {num_embeddings}")
+        super().__init__(
+            num_embeddings=num_embeddings,
+            embedding_dim=embedding_dim,
+            device=device,
+            dtype=dtype,
+            padding_idx=padding_idx,
+            **kwargs,
+        )
+        self.num_embeddings = num_embeddings
+        self.padding_idx = padding_idx
+        self.num_additional_embeddings = num_additional_embeddings
+        self.partially_freeze = partially_freeze
+
+        if partially_freeze:
+            self.weight.requires_grad_(False)
+
+        if self.num_additional_embeddings > 0:
+            self.additional_embedding = nn.Embedding(
+                num_embeddings=self.num_additional_embeddings,
+                embedding_dim=embedding_dim,
+                device=device,
+                dtype=dtype,
+            )
+
+    def forward(self, input_ids):
+        """
+        we have 2 embeddings, with different indices - one pretrained self.weight and another
+        self.additional_embedding.weight that is being trained.
+        in order to make a lookup of the input ids, we:
+        1. find out the indices of the entries belonging to the 2nd embedding
+        2. extract those values while subtracting the size of the first embedding (num_embeddings),
+           since the 2nd embedding starts from 0 and not num_embeddings
+        3. perform the 2nd embedding lookup
+        4. now we handle the 1st embedding, we overwrite indices belonging to the 2nd embedding with a padding index
+        5. perform the 1st embedding lookup
+        6. now we overwrite the values in the 1st embedding lookup with the values of the 2nd embedding lookup
+        note: for the 1st embedding lookup we could have looked up only the low indices and not do
+        the padding, but then we have to create a new tensor and populate it with 2 tensors that are
+        spread out across various indices - i.e. not a simple concat - I haven't benchmarked the
+        complex case if it's any faster, given that seqlens are usually relatively short it's
+        probably not faster or if faster not by much - but might be a good idea to measure.
+        """
+        if self.num_additional_embeddings == 0:
+            return self.additional_embedding(input_ids)
+
+        # Clone so that we don't modify the original input_ids later on
+        input_ids = input_ids.clone()
+        additional_vocab_indices = torch.where(input_ids >= self.num_embeddings)
+        input_ids_additional_vocab = input_ids[additional_vocab_indices]
+        additional_embeddings = self.additional_embedding(input_ids_additional_vocab - self.num_embeddings)
+
+        # for successful lookup replace input_ids with 0, the results of these will be discarded anyway
+        input_ids[additional_vocab_indices] = 0
+        full_vector = F.embedding(input_ids, self.weight)
+
+        # overwrite the records with high indices
+        full_vector[additional_vocab_indices] = additional_embeddings
+
+        return full_vector
+
+    def extra_repr(self) -> str:
+        return "num_embeddings={}, num_additional_embeddings={}, embedding_dim={}, partially_freeze={}".format(
+            self.num_embeddings,
+            self.num_additional_embeddings,
+            self.embedding_dim,
+            self.partially_freeze,
+        )
+
+@dataclass
+class VLlamaBaseModelOutputWithPast(BaseModelOutput):
+    """
+    Base class for VLlama3 model's outputs that may also contain a past key/values (to speed up sequential decoding).
+    Args:
+        last_hidden_state (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
+            Sequence of hidden-states at the output of the last layer of the model.
+            If `past_key_values` is used only the last hidden-state of the sequences of shape `(batch_size, 1,
+            hidden_size)` is output.
+        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+            Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
+            `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and optionally if
+            `config.is_encoder_decoder=True` 2 additional tensors of shape `(batch_size, num_heads,
+            encoder_sequence_length, embed_size_per_head)`.
+            Contains pre-computed hidden-states (key and values in the self-attention blocks and optionally if
+            `config.is_encoder_decoder=True` in the cross-attention blocks) that can be used (see `past_key_values`
+            input) to speed up sequential decoding.
+        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
+            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.
+            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
+        attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
+            Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
+            sequence_length)`.
+            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
+            heads.
+        image_hidden_states (`tuple(torch.FloatTensor)`, *optional*):
+            Tuple of `torch.FloatTensor` (one for the output of the image embeddings, `(batch_size, num_images,
+            sequence_length, hidden_size)`.
+            image_hidden_states of the model produced by the vision encoder, and optionally by the perceiver
+    """
+
+    last_hidden_state: torch.FloatTensor = None
+    past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None
+    hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None
+    attentions: Optional[Tuple[torch.FloatTensor, ...]] = None
+    image_hidden_states: Optional[Tuple[torch.FloatTensor]] = None
+
+@dataclass
+class VLlamaCausalLMOutputWithPast(BaseModelOutput):
+    """
+    Base class for VLlama3 causal language model (or autoregressive) outputs.
+    Args:
+        loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
+            Language modeling loss (for next-token prediction).
+        logits (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`):
+            Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
+        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+            Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
+            `(batch_size, num_heads, sequence_length, embed_size_per_head)`)
+            Contains pre-computed hidden-states (key and values in the self-attention blocks) that can be used (see
+            `past_key_values` input) to speed up sequential decoding.
+        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
+            Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
+            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.
+            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
+        attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
+            Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
+            sequence_length)`.
+            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
+            heads.
+        image_hidden_states (`tuple(torch.FloatTensor)`, *optional*):
+            Tuple of `torch.FloatTensor` (one for the output of the image embeddings, `(batch_size, num_images,
+            sequence_length, hidden_size)`.
+            image_hidden_states of the model produced by the vision encoder, and optionally by the perceiver
+    """
+
+    loss: Optional[torch.FloatTensor] = None
+    logits: torch.FloatTensor = None
+    past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None
+    hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None
+    attentions: Optional[Tuple[torch.FloatTensor, ...]] = None
+    image_hidden_states: Optional[Tuple[torch.FloatTensor]] = None
+
+
+class VLlamaSimpleMLP(nn.Module):
+    def __init__(self, input_size, output_size):
+        super().__init__()
+        self.proj = nn.Linear(input_size, output_size, bias=False)
+
+    def forward(self, x):
+        return self.proj(x)
+
+class VLlamaConnector(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.scale_factor = config.pixel_shuffle_factor
+        self.modality_projection = VLlamaSimpleMLP(
+            input_size=config.vision_config.hidden_size * (config.scale_factor**2),
+            output_size=config.text_config.hidden_size
+        )
+
+    def pixel_shuffle(self, x, scale_factor):
+        bsz, seq, embed_dim = x.size()
+        height = width = int(seq**0.5)
+        x = x.view(bsz, height, width, embed_dim)
+        x = x.view(bsz, height, int(width / scale_factor), embed_dim * scale_factor)
+        x = x.permute(0, 2, 1, 3)
+        x = x.reshape(bsz, int(width / scale_factor), int(height / scale_factor), embed_dim * (scale_factor**2))
+        x = x.permute(0, 2, 1, 3)
+        x = x.reshape(bsz, int(seq / (scale_factor**2)), embed_dim * (scale_factor**2))
+        return x
+
+    def forward(self, image_hidden_states):
+        image_hidden_states = self.pixel_shuffle(image_hidden_states, self.scale_factor)
+        image_hidden_states = self.modality_projection(image_hidden_states)
+        return image_hidden_states
+
+class VLlamaPreTrainedModel(PreTrainedModel):
+    config_class = VLlamaConfig
+    base_model_prefix = "model"
+    supports_gradient_checkpointing = True
+    _no_split_modules = ["VLlamaDecoderLayer"]
+    _skip_keys_device_placement = "past_key_values"
+    _supports_flash_attn_2 = True
+    _supports_sdpa = True
+    _supports_cache_class = True
+
+    def _init_weights(self, module):
+        """Initialize the weights."""
+
+        std = (
+            self.config.initializer_range
+            if hasattr(self.config, "initializer_range")
+            else self.config.text_config.initializer_range
+        )
+
+        if hasattr(module, "class_embedding"):
+            module.class_embedding.data.normal_(mean=0.0, std=std)
+
+        if isinstance(module, (nn.Linear, nn.Conv2d)):
+            module.weight.data.normal_(mean=0.0, std=std)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.Embedding):
+            module.weight.data.normal_(mean=0.0, std=std)
+            if module.padding_idx is not None:
+                module.weight.data[module.padding_idx].zero_()
+
+class VLlamaModel(VLlamaPreTrainedModel):
+    """
+    A subclass of Idefics3Model. We do *not* remove or block the call to inputs_merger
+    in forward. Instead, we override inputs_merger here with custom logic.
+    """
+
+    def __init__(self, config: VLlamaConfig, **kwargs):
+        super().__init__(config)
+
+        self.vision_model = VLlamaModel.init_vision_model(config, **kwargs)
+        self.connector = VLlamaConnector(config)
+        self.text_model = VLlamaModel.init_language_model(config, **kwargs)
+
+        self.image_seq_len = int(
+            ((config.vision_config.image_size // config.vision_config.patch_size) ** 2) / (config.scale_factor**2)
+        )
+        self.image_token_id = self.config.image_token_id
+
+        self.post_init()
+
+    @staticmethod
+    def init_vision_model(config: VLlamaConfig, **kwargs):
+        vision_model_config = AutoConfig.from_pretrained(
+            config.vision_config.vision_model_name,
+            trust_remote_code=True,
+            **kwargs,
+        )
+
+        vision_model = AutoModel.from_config(vision_model_config, trust_remote_code=True, **kwargs)
+
+        if hasattr(vision_model, "vision_model"):
+            # If the model has a vision_model attribute, it means it's a wrapper around another model
+            vision_model = vision_model.vision_model
+
+        return vision_model
+
+    @staticmethod
+    def init_language_model(config: VLlamaConfig, **kwargs):
+        text_model_config = AutoConfig.from_pretrained(
+            config.text_config.text_model_name,
+            trust_remote_code=True,
+            **kwargs,
+        )
+
+        text_model = AutoModel.from_config(text_model_config, trust_remote_code=True, **kwargs)
+        # extractor = regex_lookup(language_model_name, language_model_name2model)
+
+        embed_layer = DecoupledEmbedding(
+            num_embeddings=text_model_config.vocab_size,
+            num_additional_embeddings=config.additional_vocab_size,
+            embedding_dim=config.hidden_size,
+            partially_freeze=config.freeze_config["freeze_text_layers"],
+            padding_idx=config.pad_token_id,
+        )
+
+        text_model.set_input_embeddings(embed_layer)
+
+        return text_model
+
+    def enable_input_require_grads(self):
+        """
+        Enables the gradients for the input embeddings.
+        This is useful for lora when using gradient checkpointing.
+        c.f. https://github.com/huggingface/peft/issues/1402#issuecomment-1913675032
+        Override to set output.requires_grad = True for both the decoder's and vision model's embeddings.
+        """
+
+        def get_lowest_module(module):
+            if len(list(module.children())) == 0:
+                # If the module has no children, it is a leaf module (e.g., Linear, Conv2d, etc.)
+                return module
+            else:
+                # Recursively call the function on each child module
+                return get_lowest_module(list(module.children())[0])
+
+        def make_inputs_require_grads(module, input, output):
+            output.requires_grad_(True)
+
+        self._text_require_grads_hook = self.get_input_embeddings().register_forward_hook(make_inputs_require_grads)
+        self._vision_require_grads_hook = get_lowest_module(self.vision_model).register_forward_hook(
+            make_inputs_require_grads
+        )
+
+    def disable_input_require_grads(self):
+        self._text_require_grads_hook.remove()
+        self._vision_require_grads_hook.remove()
+
+    def get_input_embeddings(self):
+        return self.text_model.get_input_embeddings()
+
+    def set_input_embeddings(self, value):
+        self.text_model.set_input_embeddings(value)
+
+    def inputs_merger(
+        self, input_ids: torch.LongTensor, inputs_embeds: torch.Tensor, image_hidden_states: torch.Tensor
+    ):
+        """
+        This method aims at merging the token embeddings with the image hidden states into one single sequence of vectors that are fed to the transformer LM.
+        The merging happens as follows:
+        - The text token sequence is: `tok_1 tok_2 tok_3 <fake_token_around_image> <image> <image> ... <image> <fake_token_around_image> tok_4`.
+        - We get the image hidden states for the image through the vision encoder and that hidden state, after a pixel shuffle operation, is then projected into the text embedding space.
+        We thus have a sequence of image hidden states of size (1, image_seq_len, hidden_dim), where 1 is for batch_size of 1 image and hidden_dim is the hidden_dim of the LM transformer.
+        - The merging happens so that we obtain the following sequence: `vector_tok_1 vector_tok_2 vector_tok_3 vector_fake_tok_around_image {sequence of image_seq_len image hidden states} vector_fake_toke_around_image vector_tok_4`. That sequence is fed to the LM.
+        - To fit the format of that sequence, `input_ids`, `input_embeds`, `attention_mask` are all 3 adapted to insert the image hidden states.
+        """
+        _, patch_size, _ = image_hidden_states.shape
+
+        image_mask = input_ids == self.image_token_id
+        num_image_tokens = image_mask.sum(dim=1)
+        if not torch.all(num_image_tokens % patch_size == 0):
+            raise ValueError("At least one sample has <image> tokens not divisible by patch_size.")
+
+        blocks_per_sample = num_image_tokens // patch_size
+
+        offsets = torch.nn.functional.pad(blocks_per_sample.cumsum(dim=0), (1, 0), value=0)
+        block_offset = offsets[:-1]
+        row_cum = image_mask.cumsum(dim=-1)
+        chunk_idx = (row_cum - 1) // patch_size
+        local_idx = (row_cum - 1) % patch_size
+        block_idx = block_offset.unsqueeze(1) + chunk_idx
+
+        image_embeds = torch.zeros_like(inputs_embeds)
+        image_embeds[image_mask] = image_hidden_states[block_idx[image_mask], local_idx[image_mask], :]
+
+        merged_embeds = torch.where(image_mask.unsqueeze(-1), image_embeds, inputs_embeds)
+        return merged_embeds
+    
+    def embed_tokens(self, input_ids: torch.LongTensor) -> torch.FloatTensor:
+        """
+        Override the embed_tokens method to use the text model's input embeddings.
+        This is necessary to ensure that the image token ID is correctly handled.
+        """
+        if self.text_model.get_input_embeddings() is None:
+            raise ValueError("The text model does not have input embeddings.")
+        
+        return self.text_model.get_input_embeddings()(input_ids).to(input_ids.device)
+
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[List[torch.FloatTensor]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        pixel_values: Optional[torch.FloatTensor] = None,
+        pixel_attention_mask: Optional[torch.BoolTensor] = None,
+        image_hidden_states: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        cache_position: Optional[torch.LongTensor] = None,
+    ) -> Union[Tuple, VLlamaBaseModelOutputWithPast]:    
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        use_cache = use_cache if use_cache is not None else self.config.use_cache
+
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        if (input_ids is None) ^ (inputs_embeds is not None):
+            raise ValueError(
+                "You cannot specify both input_ids and inputs_embeds at the same time, and must specify either one"
+            )
+
+        if self.training and self.text_model.gradient_checkpointing and use_cache:
+            logger.warning_once(
+                "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`."
+            )
+            use_cache = False
+
+        if inputs_embeds is None:
+            inputs_embeds = self.embed_tokens(input_ids)
+
+        if inputs_embeds is not None and input_ids is None:
+            raise ValueError("When first calling the model, if input_embeds are passed, input_ids should not be None.")
+
+        # START VISUAL INPUTS INTEGRATION
+        if pixel_values is not None and image_hidden_states is not None:
+            raise ValueError("You cannot specify both pixel_values and image_hidden_states at the same time")
+        elif pixel_values is not None:
+            batch_size, num_images, num_channels, height, width = pixel_values.shape
+            pixel_values = pixel_values
+            pixel_values = pixel_values.view(batch_size * num_images, *pixel_values.shape[2:])
+
+            # Remove padding images - padding images are full 0.
+            nb_values_per_image = pixel_values.shape[1:].numel()
+            real_images_inds = (pixel_values == 0.0).sum(dim=(-1, -2, -3)) != nb_values_per_image
+
+            if not any(real_images_inds):
+                # no images, leave one empty image.
+                real_images_inds[0] = True
+
+            pixel_values = pixel_values[real_images_inds].contiguous()
+
+            # Handle the vision attention mask
+            if pixel_attention_mask is None:
+                pixel_attention_mask = torch.ones(
+                    size=[pixel_values.shape[i] for i in (0, 2, 3)],
+                    dtype=torch.bool,
+                    device=pixel_values.device,
+                )
+            else:
+                # Remove padding images from the mask
+                pixel_attention_mask = pixel_attention_mask.view(
+                    batch_size * num_images, *pixel_attention_mask.shape[2:]
+                )
+                pixel_attention_mask = pixel_attention_mask[real_images_inds].contiguous()
+
+            # patch_size = self.config.vision_config.patch_size
+            # patches_subgrid = pixel_attention_mask.unfold(dimension=1, size=patch_size, step=patch_size)
+            # patches_subgrid = patches_subgrid.unfold(dimension=2, size=patch_size, step=patch_size)
+            # patch_attention_mask = (patches_subgrid.sum(dim=(-1, -2)) > 0).bool()
+
+            # Get sequence from the vision encoder
+            image_hidden_states = self.vision_model(
+                pixel_values=pixel_values,
+                # patch_attention_mask=patch_attention_mask,
+            ).last_hidden_state
+
+            # Modality projection & resampling
+            image_hidden_states = self.connector(image_hidden_states)
+
+        elif image_hidden_states is not None:
+            image_hidden_states = image_hidden_states.to(dtype=self.dtype, device=input_ids.device)
+
+        if inputs_embeds is not None and image_hidden_states is not None:
+            # When we embed, we don't want to replace the potential image_token_id that we generated by images
+            # that simply don't exist
+            inputs_embeds = self.inputs_merger(
+                input_ids=input_ids,
+                inputs_embeds=inputs_embeds,
+                image_hidden_states=image_hidden_states,
+            )
+
+        outputs = self.text_model(
+            inputs_embeds=inputs_embeds,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+            past_key_values=past_key_values,
+            use_cache=use_cache,
+            cache_position=cache_position,
+        )
+
+        if not return_dict:
+            return tuple(v for v in [*outputs, image_hidden_states] if v is not None)
+
+        return VLlamaBaseModelOutputWithPast(
+            last_hidden_state=outputs.last_hidden_state,
+            past_key_values=past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            image_hidden_states=image_hidden_states,
+        )
+
+class VLlamaForCausalLM(VLlamaPreTrainedModel):
+    # _tied_weights_keys = ["predictions.decoder.bias", "cls.predictions.decoder.weight"]
+
+    def __init__(self, config, **kwargs):
+        super().__init__(config)
+
+        self.image_token_id = config.image_token_id
+        self.in_features = config.hidden_size
+        self.out_additional_features = config.additional_vocab_size
+        self.vocab_size = config.vocab_size
+
+        self.model = VLlamaModel(config, **kwargs)
+        self.lm_head = VLlamaForCausalLM.init_lm_head(config, **kwargs)
+        if self.out_additional_features > 0:
+            self.additional_fc = nn.Linear(
+                in_features=self.in_features,
+                out_features=self.out_additional_features,
+                bias=False,
+            )
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    @staticmethod
+    def init_lm_head(config, **kwargs):
+        # Get the pretrained model config
+        text_model_config = AutoConfig.from_pretrained(
+            config.text_config.text_model_name,
+            trust_remote_code=True,
+            **kwargs,
+        )
+        model = AutoModelForMaskedLM.from_config(text_model_config, trust_remote_code=True, **kwargs)
+        # Get the lm head
+        lm_head = model.lm_head if hasattr(model, "lm_head") else model.decoder if hasattr(model, "decoder") else None
+        if lm_head is None:
+            logger.warning(f"No lm head was found for {config.text_config.text_model_name}, initializing a new one.")
+            lm_head = nn.Linear(config.hidden_size, config.vocab_size, False)
+        return lm_head
+
+    def forward(
+            self,
+            input_ids: torch.LongTensor = None,
+            attention_mask: Optional[torch.Tensor] = None,
+            position_ids: Optional[torch.LongTensor] = None,
+            past_key_values: Optional[List[torch.FloatTensor]] = None,
+            inputs_embeds: Optional[torch.FloatTensor] = None,
+            pixel_values: Optional[torch.FloatTensor] = None,
+            pixel_attention_mask: Optional[torch.BoolTensor] = None,
+            image_hidden_states: Optional[torch.FloatTensor] = None,
+            labels: Optional[torch.LongTensor] = None,
+            use_cache: Optional[bool] = None,
+            output_attentions: Optional[bool] = None,
+            output_hidden_states: Optional[bool] = None,
+            return_dict: Optional[bool] = None,
+            cache_position: Optional[torch.LongTensor] = None,
+    ) -> Union[Tuple, VLlamaCausalLMOutputWithPast]:
+        r"""
+        Args:
+            labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+                Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
+                config.vocab_size]` or `model.image_token_id` (where `model` is your instance of `Idefics3ForConditionalGeneration`).
+                Tokens with indices set to `model.image_token_id` are ignored (masked), the loss is only
+                computed for the tokens with labels in `[0, ..., config.vocab_size]`.
+        ```"""
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+
+        # Pass the inputs to VLlamaModel
+        outputs = self.model(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            pixel_values=pixel_values,
+            pixel_attention_mask=pixel_attention_mask,
+            image_hidden_states=image_hidden_states,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+            cache_position=cache_position,
+        )
+
+        # Pass the outputs to the MLM head
+        hidden_states = outputs[0]
+
+        logits = self.lm_head(hidden_states)
+        if self.out_additional_features > 0:
+            additional_features = self.additional_fc(hidden_states)
+            logits = torch.cat((logits, additional_features), -1)
+        logits = logits.float()
+
+        loss = None
+        if labels is not None:
+            # Shift so that tokens < n predict n
+            if attention_mask is not None:
+                shift_attention_mask = attention_mask[..., 1:]
+                shift_logits = logits[..., :-1, :][shift_attention_mask != 0].contiguous()
+                shift_labels = labels[..., 1:][shift_attention_mask != 0].contiguous()
+            else:
+                shift_logits = logits[..., :-1, :].contiguous()
+                shift_labels = labels[..., 1:].contiguous()
+            # Flatten the tokens
+            loss_fct = CrossEntropyLoss()
+            loss = loss_fct(shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1))
+
+        if not return_dict:
+            output = (logits,) + outputs[1:]
+            return (loss,) + output if loss is not None else output
+
+        return VLlamaCausalLMOutputWithPast(
+            loss=loss,
+            logits=logits,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            image_hidden_states=outputs.image_hidden_states,
+        )
+
+class VLlamaForVision2Seq(VLlamaPreTrainedModel, GenerationMixin):
+    def __init__(self, config, **kwargs):
+        super().__init__(config)
+
+        self.image_token_id = config.image_token_id
+        self.in_features = config.hidden_size
+        self.out_additional_features = config.additional_vocab_size
+        self.vocab_size = config.vocab_size
+
+        self.model = VLlamaModel(config, **kwargs)
+        self.lm_head = VLlamaForVision2Seq.init_lm_head(config, **kwargs)
+        if self.out_additional_features > 0:
+            self.additional_fc = nn.Linear(
+                in_features=self.in_features,
+                out_features=self.out_additional_features,
+                bias=False,
+            )
+
+        self.loss_fct = CrossEntropyLoss()
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    @staticmethod
+    def init_lm_head(config, **kwargs):
+        # Get the pretrained model config
+        text_model_config = AutoConfig.from_pretrained(
+            config.text_config.text_model_name,
+            trust_remote_code=True,
+            **kwargs,
+        )
+        model = AutoModelForMaskedLM.from_config(text_model_config, trust_remote_code=True, **kwargs)
+        # Get the lm head
+        lm_head = model.lm_head if hasattr(model, "lm_head") else model.decoder if hasattr(model, "decoder") else None
+        if lm_head is None:
+            logger.warning(f"No lm head was found for {config.text_config.text_model_name}, initializing a new one.")
+            lm_head = nn.Linear(config.hidden_size, config.vocab_size, False)
+        return lm_head
+
+    def enable_input_require_grads(self):
+        """
+        Enables the gradients for the input embeddings. This is useful for fine-tuning adapter weights while keeping
+        the model weights fixed.
+        """
+
+        def make_inputs_require_grads(module, input, output):
+            output.requires_grad_(True)
+
+        self._text_require_grads_hook = self.get_input_embeddings().register_forward_hook(make_inputs_require_grads)
+        self._vision_require_grads_hook = self.model.vision_model.get_input_embeddings().register_forward_hook(
+            make_inputs_require_grads
+        )
+
+    def disable_input_require_grads(self):
+        self._text_require_grads_hook.remove()
+        self._vision_require_grads_hook.remove()
+
+    def get_input_embeddings(self):
+        return self.model.text_model.get_input_embeddings()
+
+    def set_input_embeddings(self, value):
+        self.model.text_model.set_input_embeddings(value)
+
+    def get_output_embeddings(self):
+        return self.lm_head
+
+    def set_output_embeddings(self, new_embeddings):
+        self.lm_head = new_embeddings
+
+    def forward(
+        self,
+        input_ids: Optional[torch.LongTensor] = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[List[torch.FloatTensor]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        pixel_values: Optional[torch.FloatTensor] = None,
+        pixel_attention_mask: Optional[torch.BoolTensor] = None,
+        image_hidden_states: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        cache_position: Optional[torch.LongTensor] = None,
+        return_dict: Optional[bool] = None,
+        logits_to_keep: Union[int, torch.Tensor] = 0,
+        **kwargs: Unpack[KwargsForCausalLM],
+    ) -> Union[Tuple, VLlamaCausalLMOutputWithPast]:
+        r"""
+        pixel_attention_mask (`torch.Tensor` of shape `(batch_size, image_size, image_size)`, *optional*):
+            Mask to avoid performing attention on padding pixel indices.
+        image_hidden_states (`torch.FloatTensor` of shape `(batch_size, num_channels, image_size, image_size)`):
+            The hidden states of the image encoder after modality projection.
+        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
+            config.vocab_size]` or `model.image_token_id` (where `model` is your instance of `SmolVLMForConditionalGeneration`).
+            Tokens with indices set to `model.image_token_id` are ignored (masked), the loss is only
+            computed for the tokens with labels in `[0, ..., config.vocab_size]`.
+
+        Example:
+
+        ```python
+        >>> import requests
+        >>> import torch
+        >>> from PIL import Image
+        >>> from io import BytesIO
+
+        >>> from transformers import AutoProcessor, AutoModelForImageTextToText
+        >>> from transformers.image_utils import load_image
+
+        >>> # Note that passing the image urls (instead of the actual pil images) to the processor is also possible
+        >>> image1 = load_image("https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg")
+        >>> image2 = load_image("https://cdn.britannica.com/59/94459-050-DBA42467/Skyline-Chicago.jpg")
+        >>> image3 = load_image("https://cdn.britannica.com/68/170868-050-8DDE8263/Golden-Gate-Bridge-San-Francisco.jpg")
+
+        >>> processor = AutoProcessor.from_pretrained("HuggingFaceTB/SmolVLM2-2.2B-Instruct")
+        >>> model = AutoModelForImageTextToText.from_pretrained("HuggingFaceTB/SmolVLM2-2.2B-Instruct", torch_dtype=torch.bfloat16, device_map="auto")
+
+        >>> # Create inputs
+        >>> messages = [
+        ...     {
+        ...         "role": "user",
+        ...         "content": [
+        ...             {"type": "video", "path": path/to/video},
+        ...             {"type": "text", "text": "What is happening in this video?"},
+        ...         ]
+        ...     }
+        ... ]
+
+        >>> inputs = processor.apply_chat_template([messages], add_generation_prompt=True)
+
+        >>> # Generate
+        >>> generated_ids = model.generate(**inputs, max_new_tokens=256)
+        >>> generated_texts = processor.batch_decode(generated_ids, skip_special_tokens=True)
+
+        >>> print(generated_texts)
+        ```"""
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        # decoder outputs consists of (dec_features, layer_state, dec_hidden, dec_attn)
+        outputs = self.model(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            pixel_values=pixel_values,
+            pixel_attention_mask=pixel_attention_mask,
+            image_hidden_states=image_hidden_states,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            cache_position=cache_position,
+            return_dict=True,
+            **kwargs,
+        )
+
+        hidden_states = outputs[0]
+        slice_indices = slice(-logits_to_keep, None) if isinstance(logits_to_keep, int) else logits_to_keep
+        # Only compute necessary logits, and do not upcast them to float if we are not computing the loss
+        hidden_states = hidden_states[:, slice_indices, :]
+        logits = self.lm_head(hidden_states)
+        if self.out_additional_features > 0:
+            additional_features = self.additional_fc(hidden_states)
+            logits = torch.cat((logits, additional_features), -1)
+        logits = logits.float()
+
+        loss = None
+        if labels is not None:
+            loss = self.loss_fct(
+                logits=logits, labels=labels, vocab_size=self.config.text_config.vocab_size, **kwargs
+            )
+
+        return VLlamaCausalLMOutputWithPast(
+            loss=loss,
+            logits=logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            image_hidden_states=outputs.image_hidden_states,
+        )
+
+    def prepare_inputs_for_generation(
+        self,
+        input_ids,
+        past_key_values=None,
+        attention_mask=None,
+        inputs_embeds=None,
+        cache_position=None,
+        pixel_values=None,
+        pixel_attention_mask=None,
+        image_hidden_states=None,
+        logits_to_keep=None,
+        **kwargs,
+    ):
+        # Overwritten -- there are mutually exclusive inputs (if the logic to make `image_hidden_states` take
+        # precedence is moved to the model, we can remove this fn)
+
+        model_inputs = super().prepare_inputs_for_generation(
+            input_ids,
+            past_key_values=past_key_values,
+            attention_mask=attention_mask,
+            inputs_embeds=inputs_embeds,
+            cache_position=cache_position,
+            pixel_values=pixel_values,
+            pixel_attention_mask=pixel_attention_mask,
+            image_hidden_states=image_hidden_states,
+            logits_to_keep=logits_to_keep,
+            **kwargs,
+        )
+
+        # if `inputs_embeds` are passed, we only want to use them in the 1st generation step
+        # but IDEFICS requires both ids and embeds to be present
+        if inputs_embeds is not None and cache_position[0] == 0:
+            model_inputs["input_ids"] = input_ids
+
+        if image_hidden_states is not None:
+            model_inputs["pixel_values"] = None
+            model_inputs["pixel_attention_mask"] = None
+
+        return model_inputs
+
+    def _update_model_kwargs_for_generation(self, outputs, model_kwargs, is_encoder_decoder, **kwargs):
+        model_kwargs = super()._update_model_kwargs_for_generation(
+            outputs=outputs,
+            model_kwargs=model_kwargs,
+            is_encoder_decoder=is_encoder_decoder,
+            **kwargs,
+        )
+        # Get the precomputed image_hidden_states
+        model_kwargs["image_hidden_states"] = outputs.image_hidden_states
+        return model_kwargs
+    
+    @staticmethod
+    # Copied from transformers.models.llama.modeling_llama.LlamaForCausalLM._reorder_cache
+    def _reorder_cache(past_key_values, beam_idx):
+        reordered_past = ()
+        for layer_past in past_key_values:
+            reordered_past += (
+                tuple(past_state.index_select(0, beam_idx.to(past_state.device)) for past_state in layer_past),
+            )
+        return reordered_past

vlm-siglip2-sllm_210/opt_step-20000__merged/preprocessor_config.json

@@ -0,0 +1,28 @@
+{
+    "do_convert_rgb": true,
+    "do_image_splitting": true,
+    "do_normalize": true,
+    "do_pad": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "image_mean": [
+        0.5,
+        0.5,
+        0.5
+    ],
+    "image_processor_type": "Idefics3ImageProcessor",
+    "image_std": [
+        0.5,
+        0.5,
+        0.5
+    ],
+    "max_image_size": {
+        "longest_edge": 512
+    },
+    "processor_class": "Idefics3Processor",
+    "resample": 1,
+    "rescale_factor": 0.00392156862745098,
+    "size": {
+        "longest_edge": 2048
+    }
+}

vlm-siglip2-sllm_210/opt_step-20000__merged/processor_config.json

@@ -0,0 +1,4 @@
+{
+    "image_seq_len": 64,
+    "processor_class": "Idefics3Processor"
+}

vlm-siglip2-sllm_210/opt_step-20000__merged/special_tokens_map.json

@@ -0,0 +1,72 @@
+{
+  "additional_special_tokens": [
+    "<global-img>",
+    "<row_1_col_1>",
+    "<row_1_col_2>",
+    "<row_1_col_3>",
+    "<row_1_col_4>",
+    "<row_1_col_5>",
+    "<row_1_col_6>",
+    "<row_2_col_1>",
+    "<row_2_col_2>",
+    "<row_2_col_3>",
+    "<row_2_col_4>",
+    "<row_2_col_5>",
+    "<row_2_col_6>",
+    "<row_3_col_1>",
+    "<row_3_col_2>",
+    "<row_3_col_3>",
+    "<row_3_col_4>",
+    "<row_3_col_5>",
+    "<row_3_col_6>",
+    "<row_4_col_1>",
+    "<row_4_col_2>",
+    "<row_4_col_3>",
+    "<row_4_col_4>",
+    "<row_4_col_5>",
+    "<row_4_col_6>",
+    "<row_5_col_1>",
+    "<row_5_col_2>",
+    "<row_5_col_3>",
+    "<row_5_col_4>",
+    "<row_5_col_5>",
+    "<row_5_col_6>",
+    "<row_6_col_1>",
+    "<row_6_col_2>",
+    "<row_6_col_3>",
+    "<row_6_col_4>",
+    "<row_6_col_5>",
+    "<row_6_col_6>",
+    "<end_of_utterance>",
+    "<fake_token_around_image>",
+    "<image>"
+  ],
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|end_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "<|reserved_special_token_0|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|end_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

vlm-siglip2-sllm_210/opt_step-20000__merged/tokenizer_config.json

@@ -0,0 +1,2429 @@
+{
+  "added_tokens_decoder": {
+    "128000": {
+      "content": "<|begin_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128001": {
+      "content": "<|end_of_text|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128002": {
+      "content": "<|reserved_special_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128003": {
+      "content": "<|reserved_special_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128004": {
+      "content": "<|finetune_right_pad_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128005": {
+      "content": "<|reserved_special_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128006": {
+      "content": "<|start_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128007": {
+      "content": "<|end_header_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128008": {
+      "content": "<|eom_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128009": {
+      "content": "<|eot_id|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128010": {
+      "content": "<|python_tag|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128011": {
+      "content": "<|reserved_special_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128012": {
+      "content": "<|reserved_special_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128013": {
+      "content": "<|reserved_special_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128014": {
+      "content": "<|reserved_special_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128015": {
+      "content": "<|reserved_special_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128016": {
+      "content": "<|reserved_special_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128017": {
+      "content": "<|reserved_special_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128018": {
+      "content": "<|reserved_special_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128019": {
+      "content": "<|reserved_special_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128020": {
+      "content": "<|reserved_special_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128021": {
+      "content": "<|reserved_special_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128022": {
+      "content": "<|reserved_special_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128023": {
+      "content": "<|reserved_special_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128024": {
+      "content": "<|reserved_special_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128025": {
+      "content": "<|reserved_special_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128026": {
+      "content": "<|reserved_special_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128027": {
+      "content": "<|reserved_special_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128028": {
+      "content": "<|reserved_special_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128029": {
+      "content": "<|reserved_special_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128030": {
+      "content": "<|reserved_special_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128031": {
+      "content": "<|reserved_special_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128032": {
+      "content": "<|reserved_special_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128033": {
+      "content": "<|reserved_special_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128034": {
+      "content": "<|reserved_special_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128035": {
+      "content": "<|reserved_special_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128036": {
+      "content": "<|reserved_special_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128037": {
+      "content": "<|reserved_special_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128038": {
+      "content": "<|reserved_special_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128039": {
+      "content": "<|reserved_special_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128040": {
+      "content": "<|reserved_special_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128041": {
+      "content": "<|reserved_special_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128042": {
+      "content": "<|reserved_special_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128043": {
+      "content": "<|reserved_special_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128044": {
+      "content": "<|reserved_special_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128045": {
+      "content": "<|reserved_special_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128046": {
+      "content": "<|reserved_special_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128047": {
+      "content": "<|reserved_special_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128048": {
+      "content": "<|reserved_special_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128049": {
+      "content": "<|reserved_special_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128050": {
+      "content": "<|reserved_special_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128051": {
+      "content": "<|reserved_special_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128052": {
+      "content": "<|reserved_special_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128053": {
+      "content": "<|reserved_special_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128054": {
+      "content": "<|reserved_special_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128055": {
+      "content": "<|reserved_special_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128056": {
+      "content": "<|reserved_special_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128057": {
+      "content": "<|reserved_special_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128058": {
+      "content": "<|reserved_special_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128059": {
+      "content": "<|reserved_special_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128060": {
+      "content": "<|reserved_special_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128061": {
+      "content": "<|reserved_special_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128062": {
+      "content": "<|reserved_special_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128063": {
+      "content": "<|reserved_special_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128064": {
+      "content": "<|reserved_special_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128065": {
+      "content": "<|reserved_special_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128066": {
+      "content": "<|reserved_special_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128067": {
+      "content": "<|reserved_special_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128068": {
+      "content": "<|reserved_special_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128069": {
+      "content": "<|reserved_special_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128070": {
+      "content": "<|reserved_special_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128071": {
+      "content": "<|reserved_special_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128072": {
+      "content": "<|reserved_special_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128073": {
+      "content": "<|reserved_special_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128074": {
+      "content": "<|reserved_special_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128075": {
+      "content": "<|reserved_special_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128076": {
+      "content": "<|reserved_special_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128077": {
+      "content": "<|reserved_special_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128078": {
+      "content": "<|reserved_special_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128079": {
+      "content": "<|reserved_special_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128080": {
+      "content": "<|reserved_special_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128081": {
+      "content": "<|reserved_special_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128082": {
+      "content": "<|reserved_special_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128083": {
+      "content": "<|reserved_special_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128084": {
+      "content": "<|reserved_special_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128085": {
+      "content": "<|reserved_special_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128086": {
+      "content": "<|reserved_special_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128087": {
+      "content": "<|reserved_special_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128088": {
+      "content": "<|reserved_special_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128089": {
+      "content": "<|reserved_special_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128090": {
+      "content": "<|reserved_special_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128091": {
+      "content": "<|reserved_special_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128092": {
+      "content": "<|reserved_special_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128093": {
+      "content": "<|reserved_special_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128094": {
+      "content": "<|reserved_special_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128095": {
+      "content": "<|reserved_special_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128096": {
+      "content": "<|reserved_special_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128097": {
+      "content": "<|reserved_special_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128098": {
+      "content": "<|reserved_special_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128099": {
+      "content": "<|reserved_special_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128100": {
+      "content": "<|reserved_special_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128101": {
+      "content": "<|reserved_special_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128102": {
+      "content": "<|reserved_special_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128103": {
+      "content": "<|reserved_special_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128104": {
+      "content": "<|reserved_special_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128105": {
+      "content": "<|reserved_special_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128106": {
+      "content": "<|reserved_special_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128107": {
+      "content": "<|reserved_special_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128108": {
+      "content": "<|reserved_special_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128109": {
+      "content": "<|reserved_special_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128110": {
+      "content": "<|reserved_special_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128111": {
+      "content": "<|reserved_special_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128112": {
+      "content": "<|reserved_special_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128113": {
+      "content": "<|reserved_special_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128114": {
+      "content": "<|reserved_special_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128115": {
+      "content": "<|reserved_special_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128116": {
+      "content": "<|reserved_special_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128117": {
+      "content": "<|reserved_special_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128118": {
+      "content": "<|reserved_special_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128119": {
+      "content": "<|reserved_special_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128120": {
+      "content": "<|reserved_special_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128121": {
+      "content": "<|reserved_special_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128122": {
+      "content": "<|reserved_special_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128123": {
+      "content": "<|reserved_special_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128124": {
+      "content": "<|reserved_special_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128125": {
+      "content": "<|reserved_special_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128126": {
+      "content": "<|reserved_special_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128127": {
+      "content": "<|reserved_special_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128128": {
+      "content": "<|reserved_special_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128129": {
+      "content": "<|reserved_special_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128130": {
+      "content": "<|reserved_special_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128131": {
+      "content": "<|reserved_special_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128132": {
+      "content": "<|reserved_special_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128133": {
+      "content": "<|reserved_special_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128134": {
+      "content": "<|reserved_special_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128135": {
+      "content": "<|reserved_special_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128136": {
+      "content": "<|reserved_special_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128137": {
+      "content": "<|reserved_special_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128138": {
+      "content": "<|reserved_special_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128139": {
+      "content": "<|reserved_special_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128140": {
+      "content": "<|reserved_special_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128141": {
+      "content": "<|reserved_special_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128142": {
+      "content": "<|reserved_special_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128143": {
+      "content": "<|reserved_special_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128144": {
+      "content": "<|reserved_special_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128145": {
+      "content": "<|reserved_special_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128146": {
+      "content": "<|reserved_special_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128147": {
+      "content": "<|reserved_special_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128148": {
+      "content": "<|reserved_special_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128149": {
+      "content": "<|reserved_special_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128150": {
+      "content": "<|reserved_special_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128151": {
+      "content": "<|reserved_special_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128152": {
+      "content": "<|reserved_special_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128153": {
+      "content": "<|reserved_special_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128154": {
+      "content": "<|reserved_special_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128155": {
+      "content": "<|reserved_special_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128156": {
+      "content": "<|reserved_special_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128157": {
+      "content": "<|reserved_special_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128158": {
+      "content": "<|reserved_special_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128159": {
+      "content": "<|reserved_special_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128160": {
+      "content": "<|reserved_special_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128161": {
+      "content": "<|reserved_special_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128162": {
+      "content": "<|reserved_special_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128163": {
+      "content": "<|reserved_special_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128164": {
+      "content": "<|reserved_special_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128165": {
+      "content": "<|reserved_special_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128166": {
+      "content": "<|reserved_special_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128167": {
+      "content": "<|reserved_special_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128168": {
+      "content": "<|reserved_special_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128169": {
+      "content": "<|reserved_special_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128170": {
+      "content": "<|reserved_special_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128171": {
+      "content": "<|reserved_special_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128172": {
+      "content": "<|reserved_special_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128173": {
+      "content": "<|reserved_special_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128174": {
+      "content": "<|reserved_special_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128175": {
+      "content": "<|reserved_special_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128176": {
+      "content": "<|reserved_special_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128177": {
+      "content": "<|reserved_special_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128178": {
+      "content": "<|reserved_special_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128179": {
+      "content": "<|reserved_special_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128180": {
+      "content": "<|reserved_special_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128181": {
+      "content": "<|reserved_special_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128182": {
+      "content": "<|reserved_special_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128183": {
+      "content": "<|reserved_special_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128184": {
+      "content": "<|reserved_special_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128185": {
+      "content": "<|reserved_special_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128186": {
+      "content": "<|reserved_special_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128187": {
+      "content": "<|reserved_special_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128188": {
+      "content": "<|reserved_special_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128189": {
+      "content": "<|reserved_special_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128190": {
+      "content": "<|reserved_special_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128191": {
+      "content": "<|reserved_special_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128192": {
+      "content": "<|reserved_special_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128193": {
+      "content": "<|reserved_special_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128194": {
+      "content": "<|reserved_special_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128195": {
+      "content": "<|reserved_special_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128196": {
+      "content": "<|reserved_special_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128197": {
+      "content": "<|reserved_special_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128198": {
+      "content": "<|reserved_special_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128199": {
+      "content": "<|reserved_special_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128200": {
+      "content": "<|reserved_special_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128201": {
+      "content": "<|reserved_special_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128202": {
+      "content": "<|reserved_special_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128203": {
+      "content": "<|reserved_special_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128204": {
+      "content": "<|reserved_special_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128205": {
+      "content": "<|reserved_special_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128206": {
+      "content": "<|reserved_special_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128207": {
+      "content": "<|reserved_special_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128208": {
+      "content": "<|reserved_special_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128209": {
+      "content": "<|reserved_special_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128210": {
+      "content": "<|reserved_special_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128211": {
+      "content": "<|reserved_special_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128212": {
+      "content": "<|reserved_special_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128213": {
+      "content": "<|reserved_special_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128214": {
+      "content": "<|reserved_special_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128215": {
+      "content": "<|reserved_special_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128216": {
+      "content": "<|reserved_special_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128217": {
+      "content": "<|reserved_special_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128218": {
+      "content": "<|reserved_special_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128219": {
+      "content": "<|reserved_special_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128220": {
+      "content": "<|reserved_special_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128221": {
+      "content": "<|reserved_special_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128222": {
+      "content": "<|reserved_special_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128223": {
+      "content": "<|reserved_special_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128224": {
+      "content": "<|reserved_special_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128225": {
+      "content": "<|reserved_special_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128226": {
+      "content": "<|reserved_special_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128227": {
+      "content": "<|reserved_special_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128228": {
+      "content": "<|reserved_special_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128229": {
+      "content": "<|reserved_special_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128230": {
+      "content": "<|reserved_special_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128231": {
+      "content": "<|reserved_special_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128232": {
+      "content": "<|reserved_special_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128233": {
+      "content": "<|reserved_special_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128234": {
+      "content": "<|reserved_special_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128235": {
+      "content": "<|reserved_special_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128236": {
+      "content": "<|reserved_special_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128237": {
+      "content": "<|reserved_special_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128238": {
+      "content": "<|reserved_special_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128239": {
+      "content": "<|reserved_special_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128240": {
+      "content": "<|reserved_special_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128241": {
+      "content": "<|reserved_special_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128242": {
+      "content": "<|reserved_special_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128243": {
+      "content": "<|reserved_special_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128244": {
+      "content": "<|reserved_special_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128245": {
+      "content": "<|reserved_special_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128246": {
+      "content": "<|reserved_special_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128247": {
+      "content": "<|reserved_special_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128248": {
+      "content": "<|reserved_special_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128249": {
+      "content": "<|reserved_special_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128250": {
+      "content": "<|reserved_special_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128251": {
+      "content": "<|reserved_special_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128252": {
+      "content": "<|reserved_special_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128253": {
+      "content": "<|reserved_special_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128254": {
+      "content": "<|reserved_special_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128255": {
+      "content": "<|reserved_special_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128256": {
+      "content": "<global-img>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128257": {
+      "content": "<row_1_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128258": {
+      "content": "<row_1_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128259": {
+      "content": "<row_1_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128260": {
+      "content": "<row_1_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128261": {
+      "content": "<row_1_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128262": {
+      "content": "<row_1_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128263": {
+      "content": "<row_2_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128264": {
+      "content": "<row_2_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128265": {
+      "content": "<row_2_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128266": {
+      "content": "<row_2_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128267": {
+      "content": "<row_2_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128268": {
+      "content": "<row_2_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128269": {
+      "content": "<row_3_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128270": {
+      "content": "<row_3_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128271": {
+      "content": "<row_3_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128272": {
+      "content": "<row_3_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128273": {
+      "content": "<row_3_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128274": {
+      "content": "<row_3_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128275": {
+      "content": "<row_4_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128276": {
+      "content": "<row_4_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128277": {
+      "content": "<row_4_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128278": {
+      "content": "<row_4_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128279": {
+      "content": "<row_4_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128280": {
+      "content": "<row_4_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128281": {
+      "content": "<row_5_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128282": {
+      "content": "<row_5_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128283": {
+      "content": "<row_5_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128284": {
+      "content": "<row_5_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128285": {
+      "content": "<row_5_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128286": {
+      "content": "<row_5_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128287": {
+      "content": "<row_6_col_1>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128288": {
+      "content": "<row_6_col_2>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128289": {
+      "content": "<row_6_col_3>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128290": {
+      "content": "<row_6_col_4>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128291": {
+      "content": "<row_6_col_5>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128292": {
+      "content": "<row_6_col_6>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128293": {
+      "content": "<end_of_utterance>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128294": {
+      "content": "<fake_token_around_image>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128295": {
+      "content": "<image>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<global-img>",
+    "<row_1_col_1>",
+    "<row_1_col_2>",
+    "<row_1_col_3>",
+    "<row_1_col_4>",
+    "<row_1_col_5>",
+    "<row_1_col_6>",
+    "<row_2_col_1>",
+    "<row_2_col_2>",
+    "<row_2_col_3>",
+    "<row_2_col_4>",
+    "<row_2_col_5>",
+    "<row_2_col_6>",
+    "<row_3_col_1>",
+    "<row_3_col_2>",
+    "<row_3_col_3>",
+    "<row_3_col_4>",
+    "<row_3_col_5>",
+    "<row_3_col_6>",
+    "<row_4_col_1>",
+    "<row_4_col_2>",
+    "<row_4_col_3>",
+    "<row_4_col_4>",
+    "<row_4_col_5>",
+    "<row_4_col_6>",
+    "<row_5_col_1>",
+    "<row_5_col_2>",
+    "<row_5_col_3>",
+    "<row_5_col_4>",
+    "<row_5_col_5>",
+    "<row_5_col_6>",
+    "<row_6_col_1>",
+    "<row_6_col_2>",
+    "<row_6_col_3>",
+    "<row_6_col_4>",
+    "<row_6_col_5>",
+    "<row_6_col_6>",
+    "<end_of_utterance>",
+    "<fake_token_around_image>",
+    "<image>"
+  ],
+  "bos_token": "<|begin_of_text|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|end_of_text|>",
+  "extra_special_tokens": {},
+  "legacy": false,
+  "mask_token": "<|reserved_special_token_0|>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask",
+    "pixel_values",
+    "pixel_attention_mask"
+  ],
+  "model_max_length": 8192,
+  "pad_token": "<|end_of_text|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}

vlm-siglip2-sllm_210/opt_step-2000__merged/chat_template.jinja

@@ -0,0 +1,2 @@
+<|begin_of_text|>{% for message in messages %}{{message['role'] | capitalize}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>
+{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}

vlm-siglip2-sllm_210/opt_step-2000__merged/chat_template.json

@@ -0,0 +1,3 @@
+{
+    "chat_template": "<|begin_of_text|>{% for message in messages %}{{message['role'] | capitalize}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>\n{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}"
+}

vlm-siglip2-sllm_210/opt_step-2000__merged/config.json

@@ -0,0 +1,51 @@
+{
+  "additional_vocab_size": 40,
+  "architectures": [
+    "VLlamaForCausalLM"
+  ],
+  "auto_map": {
+    "AutoConfig": "configuration_vllama.VLlamaConfig",
+    "AutoModel": "modeling_vllama.VLlamaModel",
+    "AutoModelForCausalLM": "modeling_vllama.VLlamaForCausalLM",
+    "AutoModelForVision2Seq": "modeling_vllama.VLlamaForVision2Seq"
+  },
+  "freeze_config": {
+    "freeze_lm_head": true,
+    "freeze_text_layers": true,
+    "freeze_vision_layers": true
+  },
+  "hidden_size": 768,
+  "image_token_id": 128295,
+  "initializer_range": 0.02,
+  "max_position_embeddings": 8192,
+  "model_type": "VLlama",
+  "neftune_noise_alpha": 0.0,
+  "output_attentions": false,
+  "pixel_shuffle_factor": 4,
+  "qk_layer_norms": false,
+  "scale_factor": 4,
+  "text_config": {
+    "hidden_size": 768,
+    "intermediate_size": 3072,
+    "mlp_bias": false,
+    "model_type": "VLlama",
+    "num_hidden_layers": 12,
+    "text_model_name": "SmolVEncoder/decoder-210m",
+    "vocab_size": 128256
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": null,
+  "use_cache": true,
+  "use_resampler": false,
+  "vision_config": {
+    "embed_dim": 768,
+    "image_size": 512,
+    "intermediate_size": 3072,
+    "model_type": "VLlama",
+    "num_hidden_layers": 12,
+    "patch_size": 16,
+    "vision_model_name": "google/siglip2-base-patch16-512"
+  },
+  "vocab_size": 128256
+}