Upload folder using huggingface_hub

README.md

@@ -0,0 +1,17 @@
+---
+license: mit
+library_name: mlx
+pipeline_tag: text-generation
+tags:
+- transformers
+- mlx
+base_model: meituan-longcat/LongCat-Flash-Thinking-2601
+---
+
+This model [finding1/LongCat-Flash-Thinking-2601-MLX-5.5bpw](https://huggingface.co/finding1/LongCat-Flash-Thinking-2601-MLX-5.5bpw) was
+converted to MLX format from [meituan-longcat/LongCat-Flash-Thinking-2601](https://huggingface.co/meituan-longcat/LongCat-Flash-Thinking-2601)
+using mlx-lm version **0.30.0** by running
+`mlx_lm.convert --hf-path meituan-longcat/LongCat-Flash-Thinking-2601 --mlx-path LongCat-Flash-Thinking-2601-MLX-5.5bpw --quantize --q-bits 5`
+until it crashed with a `KeyError`;
+[adding `"model_type": "longcat_flash",` to the downloaded `config.json`](https://github.com/ml-explore/mlx-lm/issues/433#issuecomment-3262138687),
+then running the command again.

chat_template.jinja

@@ -0,0 +1,138 @@
+{%- set tool_choice = tool_choice | default('auto') %}
+{%- set enable_thinking = enable_thinking | default(none) %}
+{%- set save_history_reasoning_content = save_history_reasoning_content | default(false) %}
+{%- set ns = namespace(tool_types = [], last_query_index = -1) %}
+{%- for idx in range(messages|length - 1) %}
+    {%- set msg = messages[idx] %}
+    {%- if msg.role == 'assistant' and not msg.tool_calls %}
+        {%- set ns.last_query_index = idx %}
+    {%- endif %}
+{%- endfor %}
+
+{%- if tools and tool_choice != 'none' %}
+    {{- "<longcat_tool_declare>\n"-}}
+    {{- "# Tools\n" }}
+    {{- "You have access to the following tools:\n\n" }}
+    {%- for tool in tools %}
+        {%- if tool.type not in ns.tool_types %}
+            {%- set ns.tool_types = ns.tool_types + [tool.type] %}
+            {{- "## Tool namespace: " ~ tool.type ~ "\n\n" }}
+        {%- endif %}
+        {%- if tool.type == 'code_interpreter' %}
+            {%- set tool = {"type":"code_interpreter","function":{"name":"code_interpreter_preview","description":"The code will be executed in a stateful Jupyter notebook sandbox environment, only supports local computation, data processing, and file operations.\nCode sandbox environment (network isolated) Any external network requests or online API calls are prohibited.\nIf online functionality is needed, please use other permitted tools.\nCode will respond with the output of the execution or time out after 60.0 seconds. ","parameters":{"type":"object","properties":{"language":{"type":"string","description":"The programming language of the code to be executed. Available values: python (Default), java, go, js, ts, c, c++."},"code":{"type":"string","description":"Python code to be executed must not include the following:\n- Importing network libraries such as requests, httplib, etc.\n- Any form of HTTP requests.\n- External API calls.\n- Network port operations. Example: ```python\nimport pandas as pd\npd.DataFrame({'A':[1,2]})\n```"},"timeout":{"type":"number","description":"The maximum execution time of the code, in seconds. Default is 60.0."}}},"required":["code"]}} %}
+        {%- endif %}
+        {{- "### Tool name: " + tool.function.name + "\n" }}
+        {{- "Description: " + tool.function.description + "\n\n" }}
+        {{- "InputSchema: " + tool.function.parameters | tojson(ensure_ascii=False) + "\n\n" }}
+    {%- endfor %}
+    {{- '**Note**: For each function call, output the function name and arguments within the following XML format:\n<longcat_tool_call>{function-name}\n<longcat_arg_key>{arg-key-1}</longcat_arg_key>\n<longcat_arg_value>{arg-value-1}</longcat_arg_value>\n<longcat_arg_key>{arg-key-2}</longcat_arg_key>\n<longcat_arg_value>{arg-value-2}</longcat_arg_value>\n...\n</longcat_tool_call>\n' }}
+    {{- "</longcat_tool_declare>"-}}
+{%- endif %}
+
+{%- for msg in messages %}
+    {%- if msg.role == "system" %}
+        {{- "<longcat_system>" + msg.content }}
+    {%- elif msg.role == "user" %}
+        {{- "<longcat_user>" }}
+        {%- if msg["files"] %}
+            {{- '<longcat_files>\n' ~ msg.files | tojson(indent=2) ~ '\n</longcat_files>' }}
+        {%- endif %}
+        {{- msg.content }}
+        {%- if save_history_reasoning_content and enable_thinking is not none %}
+            {%- if loop.index0 == messages|length - 1 %}
+                {# pass #}
+            {%- else %}
+                {%- if enable_thinking == true %}
+                    {{- " /think_on" }}
+                    {%- if thinking_budget %}
+                        {%- if thinking_budget < 1024 %}
+                            {%- set thinking_budget = 1024 %}
+                        {%- endif %}
+                        {{- "\nthinking_budget: < " ~ thinking_budget ~ "." }}
+                    {%- endif %}
+                    {{- " " }}
+                {%- elif enable_thinking == false %}
+                    {{- " /think_off " }}
+                {%- endif %}
+            {%- endif %}
+        {%- elif loop.index0 > ns.last_query_index and enable_thinking is not none %}
+            {%- if loop.index0 == messages|length - 1 %}
+                {# pass #}
+            {%- else %}
+                {%- if enable_thinking == true %}
+                    {{- " /think_on" }}
+                    {%- if thinking_budget %}
+                        {%- if thinking_budget < 1024 %}
+                            {%- set thinking_budget = 1024 %}
+                        {%- endif %}
+                        {{- "\nthinking_budget: < " ~ thinking_budget ~ "." }}
+                    {%- endif %}
+                    {{- " " }}
+                {%- elif enable_thinking == false %}
+                    {{- " /think_off " }}
+                {%- endif %}
+            {%- endif %}
+        {%- endif %}
+    {%- elif msg.role == "assistant" %}
+        {{- "<longcat_assistant>" }}
+        {%- if save_history_reasoning_content == true or loop.index0 > ns.last_query_index %}
+            {%- if enable_thinking == true and msg.reasoning_content%}
+                {{- "<longcat_think>\n" ~ msg.reasoning_content ~ "\n</longcat_think>\n" }}
+            {%- elif enable_thinking == false %}
+                {{- "<longcat_think>\n\n</longcat_think>\n" }}
+            {%- endif %}
+        {%- endif %}
+        {%- if msg.content %}
+            {{- msg.content }}
+        {%- endif %}
+        {%- if msg.tool_calls %}
+            {%- for tool_call in msg.tool_calls -%}
+                {{- "<longcat_tool_call>" ~ tool_call.function.name ~ "\n" -}}
+                {% set _args = tool_call.function.arguments %}
+                {% for k, v in _args.items() %}
+                    {{- "<longcat_arg_key>" ~ k ~ "</longcat_arg_key>\n" -}}
+                    {{- "<longcat_arg_value>" ~ (v if v is string else v | tojson(ensure_ascii=False)) ~ "</longcat_arg_value>\n" -}}
+                {% endfor %}
+                {{- "</longcat_tool_call>\n" }}
+            {%- endfor %}
+        {%- endif %}
+        {{- "</longcat_s>" -}}
+    {%- elif msg.role == "tool" %}
+        {%- if messages[loop.index0 - 1].role != "tool" %}
+            {{- "<longcat_observation>" -}}
+        {%- endif %}
+        {{- "<longcat_tool_response>" ~ msg.name ~ "\n" ~ msg.content ~ "</longcat_tool_response>"-}}
+        {%- if loop.index0 == messages|length - 1 %}
+            {# pass #}
+        {%- elif save_history_reasoning_content or loop.index0 > ns.last_query_index %}
+            {%- if enable_thinking == true and messages[loop.index0 + 1].role != "tool" %}
+                {{- " /think_on" }}
+                {%- if thinking_budget %}
+                    {%- if thinking_budget < 1024 %}
+                        {%- set thinking_budget = 1024 %}
+                    {%- endif %}
+                    {{- "\nthinking_budget: < " ~ thinking_budget ~ "." }}
+                {%- endif %}
+                {{- " " }}
+            {%- elif enable_thinking == false %}
+                {{- " /think_off " }}
+            {%- endif %}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {%- if enable_thinking == true %}
+        {{- " /think_on" }}
+        {%- if thinking_budget %}
+            {%- if thinking_budget < 1024 %}
+                {%- set thinking_budget = 1024 %}
+            {%- endif %}
+            {{- "\nthinking_budget: < " ~ thinking_budget ~ "." }}
+        {%- endif %}
+        {{- " <longcat_assistant><longcat_think>\n" }}
+    {%- elif enable_thinking == false %}
+        {{- " /think_off <longcat_assistant><longcat_think>\n\n</longcat_think>\n" }}
+    {%- else %}
+        {{- "<longcat_assistant>" }}
+    {%- endif %}
+{%- endif %}

config.json

@@ -0,0 +1,281 @@
+{
+    "architectures": [
+        "LongcatCausalLM"
+    ],
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "attention_method": "MLA",
+    "auto_map": {
+        "AutoConfig": "configuration_longcat.LongcatConfig",
+        "AutoModel": "modeling_longcat.LongcatModel",
+        "AutoModelForCausalLM": "modeling_longcat.LongcatForCausalLM"
+    },
+    "bos_token_id": 1,
+    "eos_token_id": 2,
+    "expert_ffn_hidden_size": 2048,
+    "ffn_hidden_size": 12288,
+    "hidden_size": 6144,
+    "kv_lora_rank": 512,
+    "max_position_embeddings": 983040,
+    "mla_scale_kv_lora": true,
+    "mla_scale_q_lora": true,
+    "model_type": "longcat_flash",
+    "moe_topk": 12,
+    "n_routed_experts": 512,
+    "num_attention_heads": 64,
+    "num_layers": 28,
+    "q_lora_rank": 1536,
+    "qk_nope_head_dim": 128,
+    "qk_rope_head_dim": 64,
+    "quantization": {
+        "group_size": 64,
+        "bits": 5,
+        "mode": "affine",
+        "model.layers.0.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.1.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.2.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.3.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.4.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.5.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.6.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.7.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.8.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.9.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.10.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.11.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.12.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.13.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.14.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.15.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.16.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.17.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.18.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.19.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.20.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.21.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.22.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.23.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.24.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.25.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.26.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.27.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        }
+    },
+    "quantization_config": {
+        "group_size": 64,
+        "bits": 5,
+        "mode": "affine",
+        "model.layers.0.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.1.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.2.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.3.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.4.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.5.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.6.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.7.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.8.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.9.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.10.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.11.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.12.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.13.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.14.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.15.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.16.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.17.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.18.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.19.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.20.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.21.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.22.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.23.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.24.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.25.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.26.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        },
+        "model.layers.27.mlp.router.classifier": {
+            "group_size": 64,
+            "bits": 8
+        }
+    },
+    "rms_norm_eps": 1e-05,
+    "rope_scaling": {
+        "original_max_position_embeddings": 8192,
+        "rope_type": "deepseek_yarn",
+        "factor": 120,
+        "beta_fast": 32,
+        "beta_slow": 1,
+        "mscale": 1,
+        "mscale_all_dim": 1
+    },
+    "rope_theta": 1000000.0,
+    "routed_scaling_factor": 6.0,
+    "use_cache": true,
+    "v_head_dim": 128,
+    "vocab_size": 131072,
+    "zero_expert_num": 256,
+    "zero_expert_type": "identity"
+}

configuration_longcat.py

@@ -0,0 +1,210 @@
+# coding=utf-8
+# Copyright 2025 bzantium and the HuggingFace Inc. team. All rights reserved.
+#
+
+"""Longcat model configuration"""
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.modeling_rope_utils import rope_config_validation
+
+
+LONGCAT_PRETRAINED_CONFIG_ARCHIVE_MAP = {}
+
+
+class LongcatConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`LongcatModel`]. It is used to instantiate an Longcat
+    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 Longcat.
+    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 131072):
+            Vocabulary size of the Deep model. Defines the number of different tokens that can be represented by the
+            `inputs_ids` passed when calling [`LongcatModel`]
+        hidden_size (`int`, *optional*, defaults to 7168):
+            Dimension of the hidden representations.
+        ffn_hidden_size (`int`, *optional*, defaults to 18432):
+            Dimension of the MLP representations.
+        expert_ffn_hidden_size (`int`, *optional*, defaults to 2048):
+            Dimension of the MoE representations.
+        num_layers (`int`, *optional*, defaults to 61):
+            Number of hidden layers in the Transformer decoder.
+        num_attention_heads (`int`, *optional*, defaults to 128):
+            Number of attention heads for each attention layer in the Transformer decoder.
+        num_key_value_heads (`int`, *optional*, defaults to 128):
+            This is the number of key_value heads that should be used to implement Grouped Query Attention. If
+            `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if
+            `num_key_value_heads=1 the model will use Multi Query Attention (MQA) otherwise GQA is used. When
+            converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed
+            by meanpooling all the original heads within that group. For more details checkout [this
+            paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to
+            `num_attention_heads`.
+        n_routed_experts (`int`, *optional*, defaults to 256):
+            Number of routed experts.
+        routed_scaling_factor (`float`, *optional*, defaults to 2.5):
+            Scaling factor or routed experts.
+        kv_lora_rank (`int`, *optional*, defaults to 512):
+            Rank of the LoRA matrices for key and value projections.
+        q_lora_rank (`int`, *optional*, defaults to 1536):
+            Rank of the LoRA matrices for query projections.
+        qk_rope_head_dim (`int`, *optional*, defaults to 64):
+            Dimension of the query/key heads that use rotary position embeddings.
+        v_head_dim (`int`, *optional*, defaults to 128):
+            Dimension of the value heads.
+        qk_nope_head_dim (`int`, *optional*, defaults to 128):
+            Dimension of the query/key heads that don't use rotary position embeddings.
+        norm_topk_prob (`bool`, *optional*, defaults to `True`):
+            Whether to normalize the weights of the routed experts.
+        hidden_act (`str` or `function`, *optional*, defaults to `"silu"`):
+            The non-linear activation function (function or string) in the decoder.
+        max_position_embeddings (`int`, *optional*, defaults to 4096):
+            The maximum sequence length that this model might ever be used with.
+        rms_norm_eps (`float`, *optional*, defaults to 1e-06):
+            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`.
+        pad_token_id (`int`, *optional*):
+            Padding token id.
+        bos_token_id (`int`, *optional*, defaults to 0):
+            Beginning of stream token id.
+        eos_token_id (`int`, *optional*, defaults to 1):
+            End of stream token id.
+        tie_word_embeddings (`bool`, *optional*, defaults to `False`):
+            Whether to tie weight embeddings
+        rope_theta (`float`, *optional*, defaults to 10000.0):
+            The base period of the RoPE embeddings.
+        attention_bias (`bool`, defaults to `False`, *optional*, defaults to `False`):
+            Whether to use a bias in the query, key, value and output projection layers during self-attention.
+        attention_dropout (`float`, *optional*, defaults to 0.0):
+            The dropout ratio for the attention probabilities.
+
+    ```python
+    >>> from transformers import LongcatModel, LongcatConfig
+
+    >>> # Initializing a Longcat style configuration
+    >>> configuration = LongcatConfig()
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "longcat"
+    keys_to_ignore_at_inference = ["past_key_values"]
+    base_model_tp_plan = {  # TODO: only replicate attention layers when > first_k_dense_replace
+        "layers.*.self_attn.k_proj": "colwise",
+        "layers.*.self_attn.v_proj": "colwise",
+        "layers.*.self_attn.o_proj": "rowwise",
+        "layers.*.mlp.experts.*.gate_proj": "local_colwise",
+        "layers.*.mlp.experts.*.up_proj": "local_colwise",
+        "layers.*.mlp.experts.*.down_proj": "local_rowwise",
+        "layers.*.mlps.*.gate_proj": "local_colwise",
+        "layers.*.mlps.*.up_proj": "local_colwise",
+        "layers.*.mlps.*.down_proj": "local_rowwise",
+    }
+    base_model_pp_plan = {
+        "embed_tokens": (["input_ids"], ["inputs_embeds"]),
+        "layers": (["hidden_states", "attention_mask"], ["hidden_states"]),
+        "norm": (["hidden_states"], ["hidden_states"]),
+    }
+
+    def __init__(
+        self,
+        vocab_size=131072,
+        hidden_size=7168,
+        ffn_hidden_size=18432,
+        expert_ffn_hidden_size=2048,
+        num_layers=61,
+        num_attention_heads=128,
+        num_key_value_heads=None,
+        n_routed_experts=256,
+        routed_scaling_factor=1,
+        kv_lora_rank=512,
+        q_lora_rank=1536,
+        qk_rope_head_dim=64,
+        v_head_dim=128,
+        head_dim=128,
+        qk_nope_head_dim=128,
+        mla_scale_q_lora=True,
+        mla_scale_kv_lora=True,
+        moe_topk=8,
+        norm_topk_prob=False,
+        hidden_act="silu",
+        max_position_embeddings=4096,
+        rms_norm_eps=1e-6,
+        use_cache=True,
+        pad_token_id=None,
+        bos_token_id=0,
+        eos_token_id=1,
+        tie_word_embeddings=False,
+        rope_theta=10000.0,
+        attention_bias=False,
+        attention_dropout=0.0,
+        attention_method='GQA',
+        initializer_range=0.006,
+        router_bias=False,
+        zero_expert_num=None,
+        zero_expert_type=None,
+        **kwargs,
+    ):
+        self.vocab_size = vocab_size
+        self.max_position_embeddings = max_position_embeddings
+        self.hidden_size = hidden_size
+        self.ffn_hidden_size = ffn_hidden_size
+        self.expert_ffn_hidden_size = expert_ffn_hidden_size
+        self.num_layers = num_layers
+        self.num_attention_heads = num_attention_heads
+        self.n_routed_experts = n_routed_experts
+        self.routed_scaling_factor = routed_scaling_factor
+        self.kv_lora_rank = kv_lora_rank
+        self.q_lora_rank = q_lora_rank
+        self.qk_rope_head_dim = qk_rope_head_dim
+        self.v_head_dim = v_head_dim
+        self.qk_nope_head_dim = qk_nope_head_dim
+        self.qk_head_dim = qk_nope_head_dim + qk_rope_head_dim
+        self.moe_topk = moe_topk
+        self.norm_topk_prob = norm_topk_prob
+        self.mla_scale_q_lora = mla_scale_q_lora
+        self.mla_scale_kv_lora = mla_scale_kv_lora
+        self.attention_method = attention_method
+        self.initializer_range = initializer_range
+        self.router_bias = router_bias
+        self.zero_expert_num = zero_expert_num
+        self.zero_expert_type = zero_expert_type
+
+        if self.attention_method == "GQA":
+            self.head_dim = head_dim
+        elif self.attention_method == "MLA":
+            self.head_dim = qk_rope_head_dim
+        else:
+            ValueError("attention_method should be one of [\"GQA\", \"MLA\"]")
+
+
+        # 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.rms_norm_eps = rms_norm_eps
+        self.use_cache = use_cache
+        self.rope_theta = rope_theta
+        self.attention_bias = attention_bias
+        self.attention_dropout = attention_dropout
+        # Validate the correctness of rotary position embeddings parameters
+        # BC: if there is a 'type' field, copy it it to 'rope_type'.
+        rope_config_validation(self)
+
+        super().__init__(
+            pad_token_id=pad_token_id,
+            bos_token_id=bos_token_id,
+            eos_token_id=eos_token_id,
+            tie_word_embeddings=tie_word_embeddings,
+            **kwargs,
+        )
+
+
+__all__ = ["LongcatConfig"]

generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 0,
+  "transformers_version": "4.36.0"
+}

model-00001-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6e8624c8e929b284949d7b7a2c68c4f909100948fef217aec34bf7d916d7e7c
+size 4982833873

model-00002-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44cd331a4bb4defc9cb1f26f81efd6a1413eb6c6967da8cc8f59b5ad79878958
+size 4429185427

model-00003-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6fd3aa15d8794ad9baa0b1bd9af68027adf9f1c8e89f0f34040f841d09210e9a
+size 4870523118

model-00004-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96a53cb4d5f7875d64c0f7ff29895b0f9806047e8ae450996d7c77c20b80c0f9
+size 4429185433

model-00005-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1877c984771661eb02d7196d3825ea023a137c0cf9f33dc3613beca2ba62b93
+size 4429185427

model-00006-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05ac51b806faa93334f39e79c096e928afedd3b5a50fa46e9d574769c3542e4e
+size 4870523134

model-00007-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91b7f0e8cbac263ec6bd5b07d1a859ecde8d2d3de8a6686303adad7d71f09b18
+size 4429185433

model-00008-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b405a61878102bff003f80f1fd18f765a585a9bfec6b270a06b2f7687b4b51d
+size 4429185427

model-00009-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f27f3529beaf0fdc12f27ee8502e3c67e5266e4750350afef32cca2fedf6902
+size 4870523102

model-00010-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:285dc29db4c6985d15464346595591b89a1c2168eb20643c507bda88da78c884
+size 4429185433

model-00011-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62ab886c06ba6007bd85095fe767b7245ee578f42a3ef471f006f0b99b57bf9f
+size 4429185427

model-00012-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49658a558d740134bd6dbd9714b8500773579a907cdecad63055ef53631dd1db
+size 4870523134

model-00013-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54d4d29cf62636edcb934a7ffa1a60f5d22c2a6a349e4e82a79896ba2ffd5ea1
+size 4429185433

model-00014-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:323aa9ea053355c896e14dd34240343eb93977f3668c92cb9caa6b8c6e7f61f9
+size 4429185427

model-00015-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8d9ef2ebbe8e374755b019f00d20b961fae4499a2b25ef8a15a1e189c9c23e8
+size 4870523124

model-00016-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d319fc89678d7dcc38b98d8e2d9a030ebd421a6333df20d54ad69a2d73208cd3
+size 4429185433

model-00017-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3104b61cd54a0b78992fa3a6ed1d5178b48c37d93d90727cec1639cb34dfae3
+size 4429185427

model-00018-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3d26a2a67151caea3e680c44728931f140526276b3a58dbffe61df75adedf0f
+size 4870523122

model-00019-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ef1aab652e974860d3f187c5228a59eb886ac682580083efd5ce69c63e5c41c
+size 4429185433

model-00020-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96c6a6618957e8585155ebce25f68f3f3ef2246c2da177854b90aa10ab6de3eb
+size 4429185427

model-00021-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e585332c2b9631d49f7b887fcd15989821419462eddddad7b2b41e1e055c7ab
+size 4870523160

model-00022-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:714ddc016194f2b5b436436a9193847ac5802517670cd86c9dd8ac596e1fd494
+size 4429185433

model-00023-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50d8e256dba88459c543607c6b9a83d490ff1d6ff19f00cdcf4c9579c7c958ae
+size 4429185427

model-00024-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec9f7fac6755c2ef65736b46f4e32eeb2855359ff43fad44d8eaf075aaa7ed72
+size 4870523130

model-00025-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6be05ea82a4b897bc64b0f553b41fd9fa660803f1905b61658cf2b5b1caf5502
+size 4429185433

model-00026-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49a7cb96bc6346284e7e3e0a985dfc4f1ca754fe5365471417824716159e917f
+size 4429185427

model-00027-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:daa3ef46610a039cb0b615e52ddb1ecd031557c0bd4f45ecb3376bd65fdba2d5
+size 4870523106

model-00028-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c11506d62890545a72522c53b1a010728b89d44f8dc496d5d20e4292f7e6ba50
+size 4429185433

model-00029-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e57f5ca87b68cd34d397567943b82d4d3a8785b35c794db9a2ff333d6562c7e2
+size 4429185427

model-00030-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef2dc49086ccb200a7e2f1c54ef0665ae955e0227a6413e81220f2f94bce8028
+size 4870523128

model-00031-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df7e051f9868974cb91a5620f9732b9ec08881d77353b9c63af34a5f146fd5c7
+size 4429185436

model-00032-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ccf75a005d8b781ae975643f4770ee096c296d86da604dc3237d1dbbe3f5c814
+size 4429185430

model-00033-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a265b858f5d0418708a6595d0b2dce3ef4a8ea24882702cd9a2a37787a99214
+size 4870523265

model-00034-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1541a60369bd19ea93d1a3df05d687f37dc67d2cdac1d9e683f320e58a791abe
+size 4429185436

model-00035-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:346302db09fb3c0a9b6c6fb687702f1929950990e3d08d06028b7228845593b2
+size 4429185430

model-00036-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91d3598707ec7ed1ab1d9d04ba3a218a6eeb2d7e9e1d3d1a4df07437c44890ae
+size 4870523213

model-00037-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfa7cfce7de2615f18699e3a7000f5101d3cd046b51bae65f439d24e9110f034
+size 4429185436

model-00038-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cae7162f49d5b68d08ca8bfff6d6e8524769add3534d903766106ac82cc27eef
+size 4429185430

model-00039-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ff6a957a7e384ed0488e3f9e92e64949923f575a0ee803eb266f5daf66c3d96
+size 4870523163

model-00040-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be70f54d7d181dfe4b647686f4c131ef85c12b70becd2f17378844e4299df170
+size 4429185436

model-00041-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdf21b25c2e49847ab93b860c891e0853fc9b41c2ad4ffeb95a2714864885ba2
+size 4429185430

model-00042-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7dd45cb10993aa869b9335681e396ee5306ba5a2808329a5ae02ba7eccf4e3e
+size 4870523155

model-00043-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:798c426c1cf168ab40c24911abdb4c5a7dffbd462d3684e0be411f01d401ed94
+size 4429185436

model-00044-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2664c110716e1a593bfa603abf5d9a7f774739c76216793734e550c2802b2e6a
+size 4429185430

model-00045-of-00085.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8141698545c4eeef5fe90c119ed4338085d26704b48ef44a3fb37277460e521b
+size 4870523163