Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,8 +1,3 @@
----
-license: apache-2.0
-language:
-- en
-base_model:
-- Qwen/Qwen2.5-VL-7B-Instruct
-- Qwen/Qwen2-Audio-7B-Instruct
----
+---
+license: apache-2.0
+---

config.json

@@ -0,0 +1,156 @@
+{
+  "architectures": [
+    "HithinkOmniForConditionalGeneration"
+  ],
+  "attention_dropout": 0.0,
+  "audio_config": {
+    "activation_dropout": 0.0,
+    "activation_function": "gelu",
+    "attention_dropout": 0.0,
+    "d_model": 1280,
+    "dropout": 0.0,
+    "encoder_attention_heads": 20,
+    "encoder_ffn_dim": 5120,
+    "encoder_layerdrop": 0.0,
+    "encoder_layers": 32,
+    "init_std": 0.02,
+    "max_source_positions": 1500,
+    "model_type": "hithink_audio_encoder",
+    "num_hidden_layers": 32,
+    "num_mel_bins": 128,
+    "scale_embedding": false
+  },
+  "audio_decoder_config": {
+    "_attn_implementation_autoset": false,
+    "_name_or_path": "",
+    "add_cross_attention": false,
+    "architectures": null,
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": null,
+    "chunk_size_feed_forward": 0,
+    "codebook_size": 1024,
+    "cross_attention_hidden_size": null,
+    "decoder_start_token_id": null,
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "early_stopping": false,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": null,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "min_length": 0,
+    "model_type": "hithink_omni_audio_decoder",
+    "no_repeat_ngram_size": 0,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_codebooks": 8,
+    "num_hidden_layers": 6,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": null,
+    "prefix": null,
+    "problem_type": null,
+    "pruned_heads": {},
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "sep_token_id": null,
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "tf_legacy_loss": false,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": true,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torch_dtype": null,
+    "torchscript": false,
+    "typical_p": 1.0,
+    "use_bfloat16": false
+  },
+  "audio_token_index": 151665,
+  "auto_map": {
+    "AutoConfig": "configuration_hithinkomni.HithinkOmniConfig",
+    "AutoModel": "modeling_hithinkomni.HithinkOmniForConditionalGeneration"
+  },
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_token_id": 151655,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 128000,
+  "max_window_layers": 28,
+  "model_type": "hithink_omni",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "pad_token_id": 151643,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": {
+    "mrope_section": [
+      16,
+      24,
+      24
+    ],
+    "rope_type": "default",
+    "type": "default"
+  },
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.50.3",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "video_token_id": 151656,
+  "vision_config": {
+    "depth": 32,
+    "fullatt_block_indexes": [
+      7,
+      15,
+      23,
+      31
+    ],
+    "hidden_act": "silu",
+    "hidden_size": 1280,
+    "in_channels": 3,
+    "in_chans": 3,
+    "intermediate_size": 3420,
+    "model_type": "hithink_omni",
+    "num_heads": 16,
+    "out_hidden_size": 3584,
+    "patch_size": 14,
+    "spatial_merge_size": 2,
+    "spatial_patch_size": 14,
+    "temporal_patch_size": 2,
+    "tokens_per_second": 2,
+    "window_size": 112
+  },
+  "vision_end_token_id": 151653,
+  "vision_start_token_id": 151652,
+  "vision_token_id": 151654,
+  "vocab_size": 151665,
+  "vocab_size_ext": 3
+}

configuration_hithinkomni.py

@@ -0,0 +1,370 @@
+import os
+from typing import Union
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.models.auto import CONFIG_MAPPING
+from transformers.modeling_rope_utils import rope_config_validation
+from transformers.utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+
+class HithinkOmniVisionConfig(PretrainedConfig):
+    model_type = "hithink_omni"
+    base_config_key = "vision_config"
+
+    def __init__(
+        self,
+        depth=32,
+        hidden_size=3584,
+        hidden_act="silu",
+        intermediate_size=3420,
+        num_heads=16,
+        in_channels=3,
+        patch_size=14,
+        spatial_merge_size=2,
+        temporal_patch_size=2,
+        tokens_per_second=4,
+        window_size=112,
+        out_hidden_size=3584,
+        fullatt_block_indexes=[7, 15, 23, 31],
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+
+        self.depth = depth
+        self.hidden_size = hidden_size
+        self.hidden_act = hidden_act
+        self.intermediate_size = intermediate_size
+        self.num_heads = num_heads
+        self.in_channels = in_channels
+        self.patch_size = patch_size
+        self.spatial_merge_size = spatial_merge_size
+        self.temporal_patch_size = temporal_patch_size
+        self.tokens_per_second = tokens_per_second
+        self.window_size = window_size
+        self.fullatt_block_indexes = fullatt_block_indexes
+        self.out_hidden_size = out_hidden_size
+
+
+class HithinkAudioEncoderConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`HithinkAudioEncoder`]. It is used to instantiate a
+    HithinkAudio audio encoder according to the specified arguments, defining the model architecture. Instantiating a
+    configuration with the defaults will yield a similar configuration to that of the audio encoder of the HithinkAudio
+    architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+    Args:
+        num_mel_bins (`int`, *optional*, defaults to 128):
+            Number of mel features used per input features. Should correspond to the value used in the
+            `HithinkOmniProcessor` class.
+        encoder_layers (`int`, *optional*, defaults to 32):
+            Number of encoder layers.
+        encoder_attention_heads (`int`, *optional*, defaults to 20):
+            Number of attention heads for each attention layer in the Transformer encoder.
+        encoder_ffn_dim (`int`, *optional*, defaults to 5120):
+            Dimensionality of the "intermediate" (often named feed-forward) layer in encoder.
+        encoder_layerdrop (`float`, *optional*, defaults to 0.0):
+            The LayerDrop probability for the encoder. See the [LayerDrop paper](see https://arxiv.org/abs/1909.11556)
+            for more details.
+        d_model (`int`, *optional*, defaults to 1280):
+            Dimensionality of the layers.
+        dropout (`float`, *optional*, defaults to 0.0):
+            The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
+        attention_dropout (`float`, *optional*, defaults to 0.0):
+            The dropout ratio for the attention probabilities.
+        activation_function (`str`, *optional*, defaults to `"gelu"`):
+            The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
+            `"relu"`, `"silu"` and `"gelu_new"` are supported.
+        activation_dropout (`float`, *optional*, defaults to 0.0):
+            The dropout ratio for activations inside the fully connected layer.
+        scale_embedding (`bool`, *optional*, defaults to `False`):
+            Scale embeddings by diving by sqrt(d_model).
+        init_std (`float`, *optional*, defaults to 0.02):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+        max_source_positions (`int`, *optional*, defaults to 1500):
+            The maximum sequence length of log-mel filter-bank features that this model might ever be used with.
+
+    Example:
+
+    ```python
+    >>> from transformers import HithinkAudioEncoderConfig, HithinkAudioEncoder
+
+    >>> # Initializing a HithinkAudioEncoderConfig
+    >>> configuration = HithinkAudioEncoderConfig()
+
+    >>> # Initializing a HithinkAudioEncoder (with random weights)
+    >>> model = HithinkAudioEncoder(configuration)
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "hithink_audio_encoder"
+
+    def __init__(
+        self,
+        num_mel_bins=128,
+        encoder_layers=32,
+        encoder_attention_heads=20,
+        encoder_ffn_dim=5120,
+        encoder_layerdrop=0.0,
+        d_model=1280,
+        dropout=0.0,
+        attention_dropout=0.0,
+        activation_function="gelu",
+        activation_dropout=0.0,
+        scale_embedding=False,
+        init_std=0.02,
+        max_source_positions=1500,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+
+        self.num_mel_bins = num_mel_bins
+        self.d_model = d_model
+        self.encoder_layers = encoder_layers
+        self.encoder_attention_heads = encoder_attention_heads
+        self.encoder_ffn_dim = encoder_ffn_dim
+        self.dropout = dropout
+        self.attention_dropout = attention_dropout
+        self.activation_function = activation_function
+        self.activation_dropout = activation_dropout
+        self.encoder_layerdrop = encoder_layerdrop
+        self.num_hidden_layers = encoder_layers
+        self.init_std = init_std
+        self.scale_embedding = scale_embedding  # scale factor will be sqrt(d_model) if True
+        self.max_source_positions = max_source_positions
+
+
+class HithinkAudioDecoderConfig(PretrainedConfig):
+
+    model_type = "hithink_omni_audio_decoder"
+
+    def __init__(
+        self,
+        num_hidden_layers=6,
+        codebook_size=1024,
+        num_codebooks=8,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+        self.num_hidden_layers = num_hidden_layers
+        self.codebook_size = codebook_size
+        self.num_codebooks = num_codebooks
+
+
+class HithinkOmniConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`HithinkOmniModel`]. It is used to instantiate a
+    HithinkOmni model according to the specified arguments, defining the model architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+
+    Args:
+        vocab_size (`int`, *optional*, defaults to 152064):
+            Vocabulary size of the HithinkOmni model. Defines the number of different tokens that can be represented by the
+            `inputs_ids` passed when calling [`HithinkOmniModel`]
+        hidden_size (`int`, *optional*, defaults to 8192):
+            Dimension of the hidden representations.
+        intermediate_size (`int`, *optional*, defaults to 29568):
+            Dimension of the MLP representations.
+        num_hidden_layers (`int`, *optional*, defaults to 80):
+            Number of hidden layers in the Transformer encoder.
+        num_attention_heads (`int`, *optional*, defaults to 64):
+            Number of attention heads for each attention layer in the Transformer encoder.
+        num_key_value_heads (`int`, *optional*, defaults to 8):
+            This is the number of key_value heads that should be used to implement Grouped Query Attention. If
+            `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if
+            `num_key_value_heads=1` the model will use Multi Query Attention (MQA) otherwise GQA is used. When
+            converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed
+            by meanpooling all the original heads within that group. For more details checkout [this
+            paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to `32`.
+        hidden_act (`str` or `function`, *optional*, defaults to `"silu"`):
+            The non-linear activation function (function or string) in the decoder.
+        max_position_embeddings (`int`, *optional*, defaults to 32768):
+            The maximum sequence length that this model might ever be used with.
+        initializer_range (`float`, *optional*, defaults to 0.02):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+        rms_norm_eps (`float`, *optional*, defaults to 1e-05):
+            The epsilon used by the rms normalization layers.
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should return the last key/values attentions (not used by all models). Only
+            relevant if `config.is_decoder=True`.
+        tie_word_embeddings (`bool`, *optional*, defaults to `False`):
+            Whether the model's input and output word embeddings should be tied.
+        rope_theta (`float`, *optional*, defaults to 1000000.0):
+            The base period of the RoPE embeddings.
+        use_sliding_window (`bool`, *optional*, defaults to `False`):
+            Whether to use sliding window attention.
+        sliding_window (`int`, *optional*, defaults to 4096):
+            Sliding window attention (SWA) window size. If not specified, will default to `4096`.
+        max_window_layers (`int`, *optional*, defaults to 80):
+            The number of layers that use SWA (Sliding Window Attention). The bottom layers use SWA while the top use full attention.
+        attention_dropout (`float`, *optional*, defaults to 0.0):
+            The dropout ratio for the attention probabilities.
+        vision_config (`Dict`, *optional*):
+            The config for the visual encoder initialization.
+        rope_scaling (`Dict`, *optional*):
+            Dictionary containing the scaling configuration for the RoPE embeddings. NOTE: if you apply new rope type
+            and you expect the model to work on longer `max_position_embeddings`, we recommend you to update this value
+            accordingly.
+            Expected contents:
+                `rope_type` (`str`):
+                    The sub-variant of RoPE to use. Can be one of ['default', 'linear', 'dynamic', 'yarn', 'longrope',
+                    'llama3'], with 'default' being the original RoPE implementation.
+                `factor` (`float`, *optional*):
+                    Used with all rope types except 'default'. The scaling factor to apply to the RoPE embeddings. In
+                    most scaling types, a `factor` of x will enable the model to handle sequences of length x *
+                    original maximum pre-trained length.
+                `original_max_position_embeddings` (`int`, *optional*):
+                    Used with 'dynamic', 'longrope' and 'llama3'. The original max position embeddings used during
+                    pretraining.
+                `attention_factor` (`float`, *optional*):
+                    Used with 'yarn' and 'longrope'. The scaling factor to be applied on the attention
+                    computation. If unspecified, it defaults to value recommended by the implementation, using the
+                    `factor` field to infer the suggested value.
+                `beta_fast` (`float`, *optional*):
+                    Only used with 'yarn'. Parameter to set the boundary for extrapolation (only) in the linear
+                    ramp function. If unspecified, it defaults to 32.
+                `beta_slow` (`float`, *optional*):
+                    Only used with 'yarn'. Parameter to set the boundary for interpolation (only) in the linear
+                    ramp function. If unspecified, it defaults to 1.
+                `short_factor` (`List[float]`, *optional*):
+                    Only used with 'longrope'. The scaling factor to be applied to short contexts (<
+                    `original_max_position_embeddings`). Must be a list of numbers with the same length as the hidden
+                    size divided by the number of attention heads divided by 2
+                `long_factor` (`List[float]`, *optional*):
+                    Only used with 'longrope'. The scaling factor to be applied to long contexts (<
+                    `original_max_position_embeddings`). Must be a list of numbers with the same length as the hidden
+                    size divided by the number of attention heads divided by 2
+                `low_freq_factor` (`float`, *optional*):
+                    Only used with 'llama3'. Scaling factor applied to low frequency components of the RoPE
+                `high_freq_factor` (`float`, *optional*):
+                    Only used with 'llama3'. Scaling factor applied to high frequency components of the RoPE
+
+    ```python
+    >>> from transformers import HithinkOmniForConditionalGeneration, HithinkOmniConfig
+
+    >>> # Initializing a HithinkOmni style configuration
+    >>> configuration = HithinkOmniConfig()
+
+    >>> # Initializing a model from the HithinkOmni-7B style configuration
+    >>> model = HithinkOmniForConditionalGeneration(configuration)
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "hithink_omni"
+    sub_configs = {"vision_config": HithinkOmniVisionConfig}
+    keys_to_ignore_at_inference = ["past_key_values"]
+    # Default tensor parallel plan for base model `HithinkOmni`
+    base_model_tp_plan = {
+        "layers.*.self_attn.q_proj": "colwise",
+        "layers.*.self_attn.k_proj": "colwise",
+        "layers.*.self_attn.v_proj": "colwise",
+        "layers.*.self_attn.o_proj": "rowwise",
+        "layers.*.mlp.gate_proj": "colwise",
+        "layers.*.mlp.up_proj": "colwise",
+        "layers.*.mlp.down_proj": "rowwise",
+    }
+
+    def __init__(
+        self,
+        vocab_size=152064,
+        vocab_size_ext=None,
+        hidden_size=8192,
+        intermediate_size=29568,
+        num_hidden_layers=80,
+        num_attention_heads=64,
+        num_key_value_heads=8,
+        hidden_act="silu",
+        max_position_embeddings=32768,
+        initializer_range=0.02,
+        rms_norm_eps=1e-05,
+        use_cache=True,
+        tie_word_embeddings=False,
+        rope_theta=1000000.0,
+        use_sliding_window=False,
+        sliding_window=4096,
+        max_window_layers=80,
+        attention_dropout=0.0,
+        vision_config=None,
+        rope_scaling=None,
+        audio_config=None,
+        audio_token_index=151665,
+        audio_decoder_config=None,
+        **kwargs,
+    ):
+        if isinstance(vision_config, dict):
+            self.vision_config = self.sub_configs["vision_config"](**vision_config)
+        elif vision_config is None:
+            self.vision_config = self.sub_configs["vision_config"]()
+
+        self.vocab_size = vocab_size
+        self.vocab_size_ext = vocab_size_ext
+        self.max_position_embeddings = max_position_embeddings
+        self.hidden_size = hidden_size
+        self.intermediate_size = intermediate_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+        self.use_sliding_window = use_sliding_window
+        self.sliding_window = sliding_window
+        self.max_window_layers = max_window_layers
+
+        # for backward compatibility
+        if num_key_value_heads is None:
+            num_key_value_heads = num_attention_heads
+
+        self.num_key_value_heads = num_key_value_heads
+        self.hidden_act = hidden_act
+        self.initializer_range = initializer_range
+        self.rms_norm_eps = rms_norm_eps
+        self.use_cache = use_cache
+        self.rope_theta = rope_theta
+        self.attention_dropout = attention_dropout
+        self.rope_scaling = rope_scaling
+
+        # define audio config
+        self.audio_token_index = audio_token_index
+        self.ignore_index = -100
+        if isinstance(audio_config, dict):
+            audio_config = HithinkAudioEncoderConfig(**audio_config)
+        elif audio_config is None:
+            audio_config = HithinkAudioEncoderConfig(
+                d_model=1280,
+                encoder_attention_heads=20,
+                encoder_ffn_dim=5120,
+                encoder_layerdrop=0.0,
+                encoder_layers=32,
+                num_mel_bins=128,
+                max_source_positions=1500,
+                scale_embedding=False,
+                activation_function="gelu",
+            )
+        self.audio_config = audio_config
+
+        if isinstance(audio_decoder_config, dict):
+            self.audio_decoder_config = HithinkAudioDecoderConfig(**audio_decoder_config)
+        else:
+            self.audio_decoder_config = None
+
+        # Validate the correctness of rotary position embeddings parameters
+        # BC: if there is a 'type' field, move it to 'rope_type'.
+        # and change type from 'mrope' to 'default' because `mrope` does defeault RoPE calculations
+        # one can set it to "linear"/"dynamic" etc. to have scaled RoPE
+        # TODO: @raushan update config in the hub
+        if self.rope_scaling is not None and "type" in self.rope_scaling:
+            if self.rope_scaling["type"] == "mrope":
+                self.rope_scaling["type"] = "default"
+            self.rope_scaling["rope_type"] = self.rope_scaling["type"]
+        rope_config_validation(self, ignore_keys={"mrope_section"})
+
+        super().__init__(tie_word_embeddings=tie_word_embeddings, **kwargs)

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "bos_token_id": 151643,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "transformers_version": "4.50.3"
+}

model-00001-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a962f71e35583dcdec95dc44f51b033c7a35a65e4ae238e4796c988e793f8be
+size 9977516952

model-00002-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f817ff1bea9de1b7faa07b69dc21ae055479fb4d768b3dc7d4e08ec64cff6919
+size 9943393624

model-00003-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3736c5b8f17c19392d11b217f212660a135c844aa44a26d9bf1697c45a99592
+size 855522352

preprocessor_config.json

@@ -0,0 +1,34 @@
+{
+  "auto_map": {
+    "AutoProcessor": "processing_hithinkomni.HithinkOmniProcessor"
+  },
+  "chunk_length": 30,
+  "dither": 0.0,
+  "feature_extractor_type": "WhisperFeatureExtractor",
+  "feature_size": 128,
+  "hop_length": 160,
+  "image_mean": [
+    0.48145466,
+    0.4578275,
+    0.40821073
+  ],
+  "image_processor_type": "Qwen2VLImageProcessor",
+  "image_std": [
+    0.26862954,
+    0.26130258,
+    0.27577711
+  ],
+  "max_pixels": 12845056,
+  "merge_size": 2,
+  "min_pixels": 3136,
+  "n_fft": 400,
+  "n_samples": 480000,
+  "nb_max_frames": 3000,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "patch_size": 14,
+  "processor_class": "HithinkOmniProcessor",
+  "return_attention_mask": false,
+  "sampling_rate": 16000,
+  "temporal_patch_size": 2
+}

processing_hithinkomni.py

@@ -0,0 +1,262 @@
+from typing import List, Optional, Union
+
+import numpy as np
+
+from transformers import BatchFeature
+from transformers.tokenization_utils_base import PaddingStrategy, PreTokenizedInput, TextInput
+from transformers.image_utils import ImageInput, VideoInput
+from transformers.processing_utils import ProcessingKwargs, ProcessorMixin, Unpack, VideosKwargs
+
+
+class HithinkOmniVideosProcessorKwargs(VideosKwargs, total=False):
+    fps: Union[List[float], float]
+
+
+class HithinkOmniProcessorKwargs(ProcessingKwargs, total=False):
+    videos_kwargs: HithinkOmniVideosProcessorKwargs
+    _defaults = {
+        "text_kwargs": {
+            "padding": False,
+        },
+        "videos_kwargs": {"fps": 2.0},
+    }
+
+
+class HithinkOmniProcessor(ProcessorMixin):
+    r"""
+    Constructs a HithinkOmni processor which wraps a Qwen2.5-VL image processor and a HithinkOmni tokenizer into a single processor.
+    [`HithinkOmniProcessor`] offers all the functionalities of [`Qwen2VLImageProcessor`] and [`PreTrainedTokenizerFast`]. See the
+    [`~HithinkOmniProcessor.__call__`] and [`~HithinkOmniProcessor.decode`] for more information.
+    Args:
+        image_processor ([`Qwen2VLImageProcessor`], *optional*):
+            The image processor is a required input.
+        feature_extractor ([`WhisperFeatureExtractor`], *optional*):
+            The feature extractor is a required input.
+        tokenizer ([`PreTrainedTokenizerFast`], *optional*):
+            The tokenizer is a required input.
+        chat_template (`str`, *optional*): A Jinja template which will be used to convert lists of messages
+            in a chat into a tokenizable string.
+    """
+
+    attributes = ["image_processor", "feature_extractor", "tokenizer"]
+    valid_kwargs = ["chat_template"]
+
+    image_processor_class = "Qwen2VLImageProcessor"
+    feature_extractor_class = "WhisperFeatureExtractor"
+    tokenizer_class = "PreTrainedTokenizerFast"
+
+    def __init__(self, image_processor=None, feature_extractor=None, tokenizer=None, chat_template=None, **kwargs):
+        tokenizer.model_input_names = ["input_ids", "attention_mask"]  # do not include token_type_ids
+        super().__init__(image_processor, feature_extractor, tokenizer, chat_template=chat_template)
+        self.image_token = getattr(tokenizer, 'image_token', '<|image_pad|>')
+        self.video_token = getattr(tokenizer, 'video_token', '<|video_pad|>')
+        self.chat_template = tokenizer.chat_template if chat_template is None else chat_template
+
+    def __call__(
+        self,
+        images: ImageInput = None,
+        text: Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]] = None,
+        videos: VideoInput = None,
+        audios: Union[np.ndarray, List[np.ndarray]] = None,
+        sampling_rate: Optional[int] = None,
+        **kwargs: Unpack[HithinkOmniProcessorKwargs],
+    ) -> BatchFeature:
+        """
+        Main method to prepare for the model one or several sequences(s) and image(s). This method forwards the `text`
+        and `kwargs` arguments to PreTrainedTokenizerFast's [`~PreTrainedTokenizerFast.__call__`] if `text` is not `None` to encode
+        the text. To prepare the vision inputs, this method forwards the `vision_infos` and `kwrags` arguments to
+        Qwen2VLImageProcessor's [`~Qwen2VLImageProcessor.__call__`] if `vision_infos` is not `None`.
+
+        Args:
+            images (`PIL.Image.Image`, `np.ndarray`, `torch.Tensor`, `List[PIL.Image.Image]`, `List[np.ndarray]`, `List[torch.Tensor]`):
+                The image or batch of images to be prepared. Each image can be a PIL image, NumPy array or PyTorch
+                tensor. Both channels-first and channels-last formats are supported.
+            text (`str`, `List[str]`, `List[List[str]]`):
+                The sequence or batch of sequences to be encoded. Each sequence can be a string or a list of strings
+                (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set
+                `is_split_into_words=True` (to lift the ambiguity with a batch of sequences).
+            videos (`np.ndarray`, `torch.Tensor`, `List[np.ndarray]`, `List[torch.Tensor]`):
+                The image or batch of videos to be prepared. Each video can be a 4D NumPy array or PyTorch
+                tensor, or a nested list of 3D frames. Both channels-first and channels-last formats are supported.
+            audios (`np.ndarray`, `List[np.ndarray]`):
+                The audio or batch of audios to be prepared. Each audio can be a NumPy array.
+            sampling_rate (`int`, defaults to 16000):
+                The sampling rate at which the audio files should be digitalized expressed in hertz (Hz).
+            return_tensors (`str` or [`~utils.TensorType`], *optional*):
+                If set, will return tensors of a particular framework. Acceptable values are:
+                - `'tf'`: Return TensorFlow `tf.constant` objects.
+                - `'pt'`: Return PyTorch `torch.Tensor` objects.
+                - `'np'`: Return NumPy `np.ndarray` objects.
+                - `'jax'`: Return JAX `jnp.ndarray` objects.
+
+        Returns:
+            [`BatchFeature`]: A [`BatchFeature`] with the following fields:
+
+            - **input_ids** -- List of token ids to be fed to a model. Returned when `text` is not `None`.
+            - **attention_mask** -- List of indices specifying which tokens should be attended to by the model (when
+              `return_attention_mask=True` or if *"attention_mask"* is in `self.model_input_names` and if `text` is not
+              `None`).
+            - **pixel_values** -- Pixel values to be fed to a model. Returned when `images` is not `None`.
+            - **pixel_values_videos** -- Pixel values of videos to be fed to a model. Returned when `videos` is not `None`.
+            - **image_grid_thw** -- List of image 3D grid in LLM. Returned when `images` is not `None`.
+            - **video_grid_thw** -- List of video 3D grid in LLM. Returned when `videos` is not `None`.
+            - **second_per_grid_ts** -- List of video seconds per time grid. Returned when `videos` is not `None`.
+        """
+        output_kwargs = self._merge_kwargs(
+            HithinkOmniProcessorKwargs,
+            tokenizer_init_kwargs=self.tokenizer.init_kwargs,
+            **kwargs,
+        )
+        if images is not None:
+            image_inputs = self.image_processor(images=images, videos=None, **output_kwargs["images_kwargs"])
+            image_grid_thw = image_inputs["image_grid_thw"]
+        else:
+            image_inputs = {}
+            image_grid_thw = None
+
+        if videos is not None:
+            videos_inputs = self.image_processor(images=None, videos=videos, **output_kwargs["images_kwargs"])
+            video_grid_thw = videos_inputs["video_grid_thw"]
+
+            fps = output_kwargs["videos_kwargs"].pop("fps", 2.0)
+            if isinstance(fps, (int, float)):
+                second_per_grid_ts = [self.image_processor.temporal_patch_size / fps] * len(video_grid_thw)
+            elif hasattr(fps, "__len__") and len(fps) == len(video_grid_thw):
+                second_per_grid_ts = [self.image_processor.temporal_patch_size / tmp for tmp in fps]
+            else:
+                raise ValueError(
+                    f"The length of fps ({len(fps) if hasattr(fps, '__len__') else fps}) must be equal to the length of video_grid_thw ({len(video_grid_thw)}) or fps should be a single number."
+                )
+            videos_inputs.update({"second_per_grid_ts": second_per_grid_ts})
+
+        else:
+            videos_inputs = {}
+            video_grid_thw = None
+
+        if not isinstance(text, list):
+            text = [text]
+
+        if image_grid_thw is not None:
+            merge_length = self.image_processor.merge_size**2
+            index = 0
+            for i in range(len(text)):
+                while self.image_token in text[i]:
+                    text[i] = text[i].replace(
+                        self.image_token,
+                        "<|placeholder|>" * (image_grid_thw[index].prod() // merge_length),
+                        1,
+                    )
+                    index += 1
+                text[i] = text[i].replace("<|placeholder|>", self.image_token)
+
+        if video_grid_thw is not None:
+            merge_length = self.image_processor.merge_size**2
+            index = 0
+            for i in range(len(text)):
+                while self.video_token in text[i]:
+                    text[i] = text[i].replace(
+                        self.video_token,
+                        "<|placeholder|>" * (video_grid_thw[index].prod() // merge_length),
+                        1,
+                    )
+                    index += 1
+                text[i] = text[i].replace("<|placeholder|>", self.video_token)
+
+        if audios is not None:
+            audio_inputs = self.feature_extractor(
+                audios, sampling_rate=sampling_rate, return_attention_mask=True, padding="max_length", **kwargs
+            )
+            audio_inputs["feature_attention_mask"] = audio_inputs.pop(
+                "attention_mask"
+            )  # rename attention_mask to prevent conflicts later on
+            audio_output_lengths = self.get_feat_extract_output_lengths(
+                audio_inputs['feature_attention_mask'].sum(-1)
+            )
+            index = 0
+            for i in range(len(text)):
+                while "<|AUDIO|>" in text[i]:
+                    text[i] = text[i].replace(
+                        "<|AUDIO|>", "<|placeholder|>" * audio_output_lengths[index], 1
+                    )
+                    index += 1
+                text[i] = text[i].replace("<|placeholder|>", "<|AUDIO|>")
+        else:
+            audio_inputs = {}
+
+        text_inputs =self.tokenizer(text, **output_kwargs["text_kwargs"])
+
+        return BatchFeature(data={**text_inputs, **image_inputs, **videos_inputs, **audio_inputs})
+
+    @staticmethod
+    def get_feat_extract_input_length(audio_length):
+        """
+        Computes the input length of the audio encoder (i.e. output of the feature extractor)
+        e.g. 30-second audio has 480,000 samples (sampling_rate = 16,000), the feature length will be 3,000
+        """
+        return int(np.ceil((audio_length - 40) / 160))  # 第一帧需要200样本，后续每帧需要160样本
+
+    @staticmethod
+    def get_feat_extract_output_lengths(input_lengths):
+        """
+        Computes the output length of the convolutional layers and the output length of the audio encoder
+        """
+        input_lengths = (input_lengths - 1) // 2 + 1
+        output_lengths = (input_lengths - 2) // 2 + 1
+        return output_lengths
+
+    def featurize_audio_chunk(self, audio: np.ndarray, is_last: bool, n_extracted_frames: int = 0, **kwargs):
+        """
+        Extract the features from the audio chunk during streaming inference
+        """
+        n_frames = (len(audio) - 40) / 160  # 第一帧需要200样本，后续每帧需要160样本
+        n_frames = int(np.ceil(n_frames) if is_last else np.floor(n_frames))
+        n_new_frames = n_frames - n_extracted_frames
+        i_end = n_frames * 160 + 40
+        i_start = max(0, (n_extracted_frames + 1 - 3) * 160)  # 滑窗需要400样本，即最少3帧
+        if n_new_frames <= 0 or n_frames < 2:
+            return
+        a = audio[i_start: i_end]  # 截取计算new frames需要的chunk
+        if is_last and (n_pad := int(np.ceil(len(a) / 160)) * 160 - len(a)):  # pad to multiple of 160
+            a = np.pad(a, [0, n_pad])
+        features = self.feature_extractor(
+            a, sampling_rate=self.feature_extractor.sampling_rate, padding='do_not_pad', **kwargs
+        )['input_features']
+        return features[:, :, -n_new_frames:]
+
+    def batch_decode(self, *args, **kwargs):
+        """
+        This method forwards all its arguments to PreTrainedTokenizerFast's [`~PreTrainedTokenizer.batch_decode`]. Please
+        refer to the docstring of this method for more information.
+        """
+        return self.tokenizer.batch_decode(*args, **kwargs)
+
+    def decode(self, *args, **kwargs):
+        """
+        This method forwards all its arguments to PreTrainedTokenizerFast's [`~PreTrainedTokenizer.decode`]. Please refer to
+        the docstring of this method for more information.
+        """
+        return self.tokenizer.decode(*args, **kwargs)
+
+    def post_process_image_text_to_text(self, generated_outputs):
+        """
+        Post-process the output of the model to decode the text.
+
+        Args:
+            generated_outputs (`torch.Tensor` or `np.ndarray`):
+                The output of the model `generate` function. The output is expected to be a tensor of shape `(batch_size, sequence_length)`
+                or `(sequence_length,)`.
+
+        Returns:
+            `List[str]`: The decoded text.
+        """
+        return self.tokenizer.batch_decode(
+            generated_outputs, skip_special_tokens=True, clean_up_tokenization_spaces=False
+        )
+
+    @property
+    def model_input_names(self):
+        tokenizer_input_names = self.tokenizer.model_input_names
+        image_processor_input_names = self.image_processor.model_input_names
+        feature_extractor_input_names = self.feature_extractor.model_input_names
+        return list(dict.fromkeys(tokenizer_input_names + image_processor_input_names
+                                  + feature_extractor_input_names + ["feature_attention_mask"]))  # audio

processor_config.json

@@ -0,0 +1,6 @@
+{
+  "auto_map": {
+    "AutoProcessor": "processing_hithinkomni.HithinkOmniProcessor"
+  },
+  "processor_class": "HithinkOmniProcessor"
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8382c71e5571bd71a614f60271a48b21e4c21b02f0140ee7ab2f708ce510949f
+size 11422462

tokenizer_config.json

@@ -0,0 +1,239 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151665": {
+      "content": "<|AUDIO|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151666": {
+      "content": "<|audio_bos|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151667": {
+      "content": "<|audio_eos|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>",
+    "<|AUDIO|>",
+    "<|audio_bos|>",
+    "<|audio_eos|>"
+  ],
+  "auto_map": {
+    "AutoProcessor": "processing_hithinkomni.HithinkOmniProcessor"
+  },
+  "bos_token": null,
+  "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% set audio_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'audio' in content or 'audio_url' in content %}{% set audio_count.value = audio_count.value + 1 %}Audio {{ audio_count.value }}: <|audio_bos|><|AUDIO|><|audio_eos|>\n{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "processor_class": "HithinkOmniProcessor",
+  "split_special_tokens": false,
+  "tokenizer_class": "PreTrainedTokenizer",
+  "unk_token": null
+}