Upload folder using huggingface_hub

.gitattributes

@@ -34,3 +34,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 ring-mini-2.0-9b/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

chat_template.jinja

@@ -0,0 +1,2 @@
+{% for message in messages %}{% set role = message['role'] | lower %}{% if role == 'user' %}{% set role = 'HUMAN' %}{% endif %}{% set role = role | upper %}{{ '<role>' + role + '</role>' + message['content'] }}{% endfor %}{% if add_generation_prompt %}{{ '<role>ASSISTANT</role><think>
+' }}{% endif %}

config.json

@@ -1,50 +1,54 @@
 {
-    "architectures": [
-        "BailingMoeV2ForCausalLM"
-    ],
-    "attention_dropout": 0.0,
-    "auto_map": {
-        "AutoConfig": "configuration_bailing_moe_v2.BailingMoeV2Config",
-        "AutoModel": "modeling_bailing_moe_v2.BailingMoeV2Model",
-        "AutoModelForCausalLM": "modeling_bailing_moe_v2.BailingMoeV2ForCausalLM"
-    },
-    "num_hidden_layers": 20,
-    "hidden_size": 2048,
-    "intermediate_size": 5120,
-    "eos_token_id": 156892,
-    "pad_token_id": 156892,
-    "first_k_dense_replace": 1,
-    "hidden_act": "silu",
-    "max_position_embeddings": 32768,
-    "model_type": "bailing_moe",
-    "moe_intermediate_size": 512,
-    "norm_topk_prob": true,
-    "num_experts_per_tok": 8,
-    "num_attention_heads": 16,
-    "num_experts": 256,
-    "num_key_value_heads": 4,
-    "rope_theta": 600000,
-    "rope_scaling": null,
-    "tie_word_embeddings": false,
-    "torch_dtype": "bfloat16",
-    "transformers_version": "4.52.3",
-    "use_bias": false,
-    "use_rmsnorm": true,
-    "rms_norm_eps": 1e-06,
-    "head_dim": 128,
-    "num_shared_experts": 1,
-    "use_cache": true,
-    "use_qkv_bias": false,
-    "embedding_dropout": 0.0,
-    "output_dropout": 0.0,
-    "vocab_size": 157184,
-    "partial_rotary_factor": 0.5,
-    "router_dtype": "fp32",
-    "moe_router_enable_expert_bias": true,
-    "routed_scaling_factor": 2.5,
-    "n_group": 8,
-    "topk_group": 4,
-    "use_qk_norm": true,
-    "score_function": "sigmoid",
-    "moe_shared_expert_intermediate_size": 512
-}
+  "architectures": [
+    "BailingMoeV2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "auto_map": {
+    "AutoConfig": "configuration_bailing_moe_v2.BailingMoeV2Config",
+    "AutoModel": "modeling_bailing_moe_v2.BailingMoeV2Model",
+    "AutoModelForCausalLM": "modeling_bailing_moe_v2.BailingMoeV2ForCausalLM"
+  },
+  "dtype": "bfloat16",
+  "embedding_dropout": 0.0,
+  "eos_token_id": 156892,
+  "first_k_dense_replace": 1,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 5120,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 20,
+  "moe_intermediate_size": 512,
+  "moe_router_enable_expert_bias": true,
+  "moe_shared_expert_intermediate_size": 512,
+  "mtp_loss_scaling_factor": 0,
+  "n_group": 8,
+  "norm_topk_prob": true,
+  "num_attention_heads": 16,
+  "num_experts": 128,
+  "num_experts_per_tok": 8,
+  "num_hidden_layers": 20,
+  "num_key_value_heads": 4,
+  "num_nextn_predict_layers": 0,
+  "num_shared_experts": 1,
+  "output_dropout": 0.0,
+  "output_router_logits": false,
+  "pad_token_id": 156892,
+  "partial_rotary_factor": 0.5,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 600000,
+  "routed_scaling_factor": 2.5,
+  "router_dtype": "fp32",
+  "score_function": "sigmoid",
+  "tie_word_embeddings": false,
+  "topk_group": 4,
+  "transformers_version": "4.56.2",
+  "use_bias": false,
+  "use_cache": true,
+  "use_qk_norm": true,
+  "use_qkv_bias": false,
+  "use_rmsnorm": true,
+  "vocab_size": 157184
+}

expert_selection.json

@@ -0,0 +1,2472 @@
+{
+  "layer_1": [
+    22,
+    13,
+    29,
+    1,
+    9,
+    20,
+    27,
+    16,
+    19,
+    21,
+    17,
+    28,
+    11,
+    14,
+    31,
+    15,
+    57,
+    42,
+    45,
+    55,
+    48,
+    38,
+    34,
+    61,
+    62,
+    63,
+    51,
+    47,
+    49,
+    58,
+    52,
+    46,
+    71,
+    73,
+    81,
+    87,
+    86,
+    65,
+    67,
+    64,
+    92,
+    85,
+    95,
+    74,
+    94,
+    91,
+    76,
+    77,
+    104,
+    107,
+    126,
+    102,
+    105,
+    121,
+    116,
+    117,
+    110,
+    127,
+    106,
+    122,
+    96,
+    103,
+    109,
+    115,
+    147,
+    134,
+    155,
+    148,
+    152,
+    157,
+    156,
+    146,
+    142,
+    149,
+    145,
+    153,
+    158,
+    135,
+    143,
+    133,
+    166,
+    190,
+    163,
+    187,
+    167,
+    174,
+    170,
+    183,
+    169,
+    182,
+    188,
+    168,
+    171,
+    186,
+    162,
+    161,
+    197,
+    209,
+    211,
+    201,
+    221,
+    210,
+    206,
+    192,
+    213,
+    217,
+    207,
+    212,
+    193,
+    216,
+    196,
+    218,
+    236,
+    237,
+    242,
+    245,
+    229,
+    227,
+    248,
+    250,
+    247,
+    243,
+    255,
+    244,
+    232,
+    249,
+    239,
+    234
+  ],
+  "layer_2": [
+    4,
+    24,
+    18,
+    29,
+    26,
+    8,
+    16,
+    12,
+    7,
+    31,
+    28,
+    5,
+    9,
+    3,
+    17,
+    13,
+    58,
+    35,
+    57,
+    63,
+    48,
+    40,
+    45,
+    50,
+    38,
+    47,
+    41,
+    53,
+    39,
+    46,
+    62,
+    43,
+    83,
+    86,
+    91,
+    85,
+    93,
+    73,
+    94,
+    64,
+    71,
+    69,
+    81,
+    66,
+    95,
+    68,
+    89,
+    92,
+    124,
+    126,
+    111,
+    122,
+    97,
+    110,
+    112,
+    119,
+    101,
+    116,
+    117,
+    127,
+    100,
+    123,
+    98,
+    121,
+    140,
+    128,
+    138,
+    134,
+    158,
+    154,
+    144,
+    145,
+    137,
+    157,
+    153,
+    159,
+    150,
+    146,
+    156,
+    133,
+    181,
+    161,
+    189,
+    190,
+    177,
+    167,
+    173,
+    191,
+    176,
+    165,
+    180,
+    187,
+    168,
+    178,
+    186,
+    170,
+    216,
+    194,
+    195,
+    211,
+    219,
+    212,
+    223,
+    197,
+    215,
+    221,
+    203,
+    199,
+    220,
+    207,
+    210,
+    218,
+    252,
+    225,
+    240,
+    250,
+    254,
+    227,
+    247,
+    244,
+    249,
+    228,
+    242,
+    226,
+    233,
+    232,
+    248,
+    251
+  ],
+  "layer_3": [
+    8,
+    2,
+    11,
+    16,
+    10,
+    12,
+    21,
+    1,
+    3,
+    24,
+    29,
+    23,
+    17,
+    13,
+    5,
+    14,
+    50,
+    45,
+    49,
+    52,
+    61,
+    53,
+    32,
+    42,
+    56,
+    63,
+    58,
+    41,
+    47,
+    40,
+    36,
+    59,
+    89,
+    66,
+    69,
+    83,
+    82,
+    72,
+    94,
+    75,
+    87,
+    92,
+    73,
+    85,
+    67,
+    88,
+    71,
+    91,
+    123,
+    96,
+    121,
+    118,
+    112,
+    127,
+    114,
+    122,
+    100,
+    103,
+    109,
+    107,
+    111,
+    113,
+    117,
+    115,
+    131,
+    158,
+    157,
+    153,
+    138,
+    129,
+    149,
+    156,
+    137,
+    154,
+    152,
+    145,
+    147,
+    142,
+    150,
+    134,
+    185,
+    177,
+    178,
+    181,
+    173,
+    164,
+    186,
+    161,
+    165,
+    172,
+    190,
+    184,
+    188,
+    179,
+    175,
+    182,
+    213,
+    222,
+    203,
+    216,
+    221,
+    198,
+    211,
+    199,
+    220,
+    207,
+    205,
+    210,
+    214,
+    208,
+    217,
+    204,
+    250,
+    229,
+    252,
+    243,
+    233,
+    227,
+    245,
+    242,
+    230,
+    231,
+    244,
+    255,
+    224,
+    246,
+    247,
+    234
+  ],
+  "layer_4": [
+    21,
+    31,
+    23,
+    5,
+    17,
+    22,
+    28,
+    24,
+    8,
+    1,
+    14,
+    18,
+    29,
+    2,
+    10,
+    19,
+    61,
+    58,
+    43,
+    44,
+    54,
+    34,
+    53,
+    55,
+    41,
+    42,
+    45,
+    35,
+    63,
+    32,
+    37,
+    62,
+    93,
+    90,
+    86,
+    83,
+    76,
+    89,
+    70,
+    66,
+    71,
+    92,
+    88,
+    64,
+    72,
+    79,
+    95,
+    75,
+    127,
+    109,
+    110,
+    101,
+    107,
+    100,
+    123,
+    99,
+    108,
+    115,
+    120,
+    119,
+    113,
+    96,
+    98,
+    116,
+    144,
+    134,
+    152,
+    141,
+    151,
+    159,
+    156,
+    153,
+    136,
+    149,
+    148,
+    157,
+    138,
+    128,
+    139,
+    155,
+    184,
+    174,
+    178,
+    175,
+    170,
+    186,
+    187,
+    189,
+    177,
+    190,
+    176,
+    188,
+    172,
+    183,
+    160,
+    182,
+    200,
+    217,
+    207,
+    218,
+    214,
+    212,
+    208,
+    197,
+    204,
+    206,
+    196,
+    220,
+    192,
+    213,
+    210,
+    194,
+    226,
+    242,
+    248,
+    246,
+    234,
+    253,
+    250,
+    255,
+    249,
+    245,
+    233,
+    241,
+    231,
+    239,
+    237,
+    235
+  ],
+  "layer_5": [
+    26,
+    4,
+    27,
+    2,
+    22,
+    11,
+    29,
+    14,
+    6,
+    8,
+    25,
+    5,
+    18,
+    21,
+    20,
+    24,
+    62,
+    40,
+    46,
+    51,
+    63,
+    55,
+    50,
+    32,
+    37,
+    49,
+    33,
+    52,
+    59,
+    43,
+    36,
+    41,
+    81,
+    95,
+    74,
+    68,
+    72,
+    76,
+    70,
+    77,
+    93,
+    65,
+    94,
+    87,
+    82,
+    88,
+    84,
+    71,
+    118,
+    108,
+    125,
+    109,
+    117,
+    121,
+    96,
+    97,
+    113,
+    99,
+    122,
+    101,
+    111,
+    123,
+    114,
+    106,
+    152,
+    145,
+    158,
+    139,
+    130,
+    138,
+    142,
+    136,
+    148,
+    133,
+    144,
+    146,
+    143,
+    128,
+    135,
+    134,
+    168,
+    172,
+    186,
+    162,
+    187,
+    173,
+    175,
+    183,
+    160,
+    166,
+    169,
+    179,
+    180,
+    164,
+    188,
+    184,
+    201,
+    213,
+    202,
+    217,
+    206,
+    199,
+    223,
+    207,
+    212,
+    197,
+    221,
+    219,
+    200,
+    193,
+    198,
+    196,
+    230,
+    227,
+    251,
+    238,
+    245,
+    228,
+    255,
+    226,
+    232,
+    246,
+    224,
+    225,
+    243,
+    254,
+    236,
+    234
+  ],
+  "layer_6": [
+    21,
+    11,
+    22,
+    14,
+    19,
+    3,
+    5,
+    27,
+    29,
+    6,
+    15,
+    0,
+    12,
+    4,
+    2,
+    31,
+    61,
+    44,
+    59,
+    41,
+    33,
+    42,
+    53,
+    47,
+    54,
+    55,
+    57,
+    38,
+    48,
+    63,
+    56,
+    34,
+    89,
+    95,
+    80,
+    71,
+    77,
+    64,
+    74,
+    90,
+    66,
+    67,
+    75,
+    93,
+    91,
+    69,
+    92,
+    79,
+    102,
+    109,
+    110,
+    126,
+    120,
+    108,
+    98,
+    125,
+    123,
+    112,
+    118,
+    113,
+    97,
+    104,
+    96,
+    127,
+    155,
+    138,
+    147,
+    151,
+    132,
+    136,
+    128,
+    140,
+    154,
+    156,
+    150,
+    139,
+    130,
+    149,
+    157,
+    129,
+    176,
+    178,
+    175,
+    168,
+    187,
+    169,
+    190,
+    170,
+    180,
+    179,
+    166,
+    165,
+    172,
+    167,
+    160,
+    185,
+    196,
+    200,
+    212,
+    205,
+    219,
+    193,
+    214,
+    215,
+    210,
+    222,
+    217,
+    206,
+    208,
+    203,
+    221,
+    201,
+    232,
+    231,
+    248,
+    227,
+    249,
+    250,
+    235,
+    239,
+    229,
+    240,
+    225,
+    228,
+    242,
+    254,
+    244,
+    247
+  ],
+  "layer_7": [
+    22,
+    21,
+    14,
+    28,
+    3,
+    31,
+    29,
+    26,
+    27,
+    15,
+    20,
+    0,
+    30,
+    10,
+    5,
+    18,
+    50,
+    37,
+    39,
+    60,
+    33,
+    53,
+    62,
+    63,
+    43,
+    42,
+    57,
+    34,
+    61,
+    52,
+    47,
+    35,
+    68,
+    73,
+    87,
+    78,
+    74,
+    94,
+    85,
+    69,
+    91,
+    84,
+    95,
+    77,
+    93,
+    86,
+    83,
+    71,
+    97,
+    111,
+    123,
+    118,
+    110,
+    125,
+    119,
+    122,
+    96,
+    112,
+    109,
+    114,
+    120,
+    108,
+    98,
+    99,
+    143,
+    153,
+    135,
+    129,
+    148,
+    156,
+    133,
+    150,
+    137,
+    157,
+    138,
+    146,
+    132,
+    130,
+    128,
+    159,
+    180,
+    181,
+    186,
+    168,
+    184,
+    166,
+    189,
+    160,
+    173,
+    185,
+    178,
+    163,
+    183,
+    177,
+    172,
+    162,
+    199,
+    203,
+    202,
+    209,
+    204,
+    207,
+    220,
+    197,
+    200,
+    212,
+    215,
+    218,
+    214,
+    219,
+    211,
+    193,
+    238,
+    240,
+    231,
+    224,
+    227,
+    226,
+    251,
+    253,
+    254,
+    252,
+    232,
+    249,
+    244,
+    247,
+    248,
+    225
+  ],
+  "layer_8": [
+    21,
+    12,
+    24,
+    13,
+    0,
+    22,
+    18,
+    8,
+    15,
+    7,
+    17,
+    5,
+    2,
+    16,
+    6,
+    9,
+    42,
+    50,
+    33,
+    58,
+    56,
+    47,
+    34,
+    60,
+    54,
+    32,
+    62,
+    55,
+    41,
+    43,
+    48,
+    45,
+    67,
+    71,
+    80,
+    93,
+    87,
+    76,
+    69,
+    72,
+    79,
+    78,
+    90,
+    77,
+    82,
+    92,
+    81,
+    83,
+    126,
+    120,
+    101,
+    110,
+    97,
+    124,
+    112,
+    122,
+    117,
+    102,
+    115,
+    121,
+    104,
+    99,
+    96,
+    111,
+    143,
+    158,
+    136,
+    138,
+    146,
+    153,
+    137,
+    150,
+    130,
+    135,
+    151,
+    132,
+    142,
+    152,
+    154,
+    131,
+    186,
+    177,
+    191,
+    172,
+    181,
+    175,
+    167,
+    179,
+    190,
+    188,
+    171,
+    170,
+    189,
+    161,
+    187,
+    173,
+    209,
+    207,
+    223,
+    214,
+    215,
+    196,
+    205,
+    194,
+    192,
+    193,
+    202,
+    213,
+    217,
+    222,
+    219,
+    221,
+    254,
+    226,
+    249,
+    229,
+    244,
+    253,
+    250,
+    224,
+    248,
+    251,
+    235,
+    232,
+    243,
+    227,
+    238,
+    230
+  ],
+  "layer_9": [
+    27,
+    29,
+    26,
+    16,
+    4,
+    0,
+    13,
+    28,
+    5,
+    11,
+    24,
+    19,
+    21,
+    7,
+    30,
+    15,
+    33,
+    51,
+    40,
+    37,
+    38,
+    43,
+    48,
+    49,
+    61,
+    56,
+    54,
+    59,
+    39,
+    52,
+    63,
+    50,
+    93,
+    80,
+    83,
+    79,
+    78,
+    77,
+    74,
+    87,
+    95,
+    72,
+    94,
+    92,
+    90,
+    73,
+    86,
+    69,
+    120,
+    122,
+    99,
+    123,
+    96,
+    124,
+    112,
+    103,
+    108,
+    114,
+    98,
+    110,
+    101,
+    113,
+    117,
+    105,
+    137,
+    156,
+    131,
+    150,
+    129,
+    151,
+    159,
+    146,
+    135,
+    141,
+    132,
+    139,
+    158,
+    140,
+    152,
+    153,
+    181,
+    169,
+    191,
+    162,
+    184,
+    161,
+    170,
+    185,
+    164,
+    177,
+    183,
+    186,
+    160,
+    166,
+    175,
+    176,
+    203,
+    222,
+    216,
+    209,
+    218,
+    195,
+    204,
+    220,
+    214,
+    193,
+    205,
+    192,
+    215,
+    197,
+    221,
+    211,
+    237,
+    230,
+    228,
+    229,
+    225,
+    248,
+    234,
+    233,
+    246,
+    245,
+    226,
+    227,
+    236,
+    252,
+    240,
+    241
+  ],
+  "layer_10": [
+    18,
+    12,
+    15,
+    6,
+    16,
+    20,
+    25,
+    24,
+    7,
+    9,
+    21,
+    0,
+    23,
+    13,
+    26,
+    2,
+    44,
+    34,
+    62,
+    47,
+    63,
+    38,
+    40,
+    55,
+    35,
+    41,
+    53,
+    58,
+    39,
+    37,
+    50,
+    51,
+    89,
+    78,
+    93,
+    65,
+    64,
+    90,
+    80,
+    83,
+    95,
+    84,
+    79,
+    86,
+    91,
+    74,
+    81,
+    67,
+    104,
+    116,
+    126,
+    102,
+    111,
+    115,
+    119,
+    121,
+    109,
+    103,
+    122,
+    110,
+    100,
+    107,
+    108,
+    101,
+    143,
+    154,
+    131,
+    138,
+    130,
+    142,
+    155,
+    151,
+    159,
+    133,
+    134,
+    129,
+    150,
+    135,
+    149,
+    153,
+    183,
+    167,
+    174,
+    166,
+    175,
+    179,
+    164,
+    188,
+    176,
+    181,
+    160,
+    184,
+    185,
+    165,
+    173,
+    189,
+    218,
+    196,
+    201,
+    193,
+    211,
+    223,
+    212,
+    216,
+    219,
+    208,
+    213,
+    205,
+    207,
+    220,
+    199,
+    214,
+    253,
+    231,
+    232,
+    251,
+    230,
+    250,
+    247,
+    244,
+    226,
+    246,
+    241,
+    227,
+    225,
+    249,
+    229,
+    254
+  ],
+  "layer_11": [
+    23,
+    14,
+    3,
+    16,
+    17,
+    2,
+    6,
+    18,
+    31,
+    29,
+    4,
+    0,
+    8,
+    30,
+    5,
+    21,
+    62,
+    49,
+    55,
+    54,
+    60,
+    48,
+    45,
+    51,
+    53,
+    58,
+    46,
+    37,
+    35,
+    47,
+    32,
+    36,
+    94,
+    79,
+    90,
+    70,
+    95,
+    88,
+    91,
+    84,
+    87,
+    93,
+    81,
+    64,
+    82,
+    89,
+    66,
+    83,
+    109,
+    115,
+    124,
+    99,
+    105,
+    127,
+    106,
+    121,
+    97,
+    100,
+    108,
+    98,
+    125,
+    103,
+    120,
+    102,
+    128,
+    155,
+    129,
+    136,
+    138,
+    146,
+    148,
+    133,
+    132,
+    158,
+    137,
+    156,
+    142,
+    141,
+    143,
+    134,
+    170,
+    180,
+    179,
+    183,
+    169,
+    184,
+    190,
+    175,
+    185,
+    181,
+    160,
+    165,
+    172,
+    171,
+    168,
+    167,
+    192,
+    199,
+    214,
+    194,
+    202,
+    204,
+    223,
+    215,
+    205,
+    221,
+    208,
+    195,
+    200,
+    212,
+    196,
+    210,
+    236,
+    235,
+    252,
+    225,
+    227,
+    234,
+    241,
+    253,
+    228,
+    242,
+    249,
+    248,
+    244,
+    254,
+    231,
+    233
+  ],
+  "layer_12": [
+    24,
+    10,
+    25,
+    14,
+    8,
+    5,
+    1,
+    20,
+    30,
+    16,
+    7,
+    22,
+    23,
+    15,
+    13,
+    21,
+    61,
+    50,
+    36,
+    42,
+    56,
+    60,
+    34,
+    35,
+    53,
+    39,
+    37,
+    45,
+    48,
+    59,
+    40,
+    51,
+    87,
+    74,
+    90,
+    89,
+    67,
+    72,
+    79,
+    95,
+    71,
+    94,
+    78,
+    86,
+    65,
+    64,
+    91,
+    69,
+    96,
+    102,
+    118,
+    109,
+    123,
+    105,
+    108,
+    106,
+    120,
+    117,
+    115,
+    103,
+    114,
+    97,
+    119,
+    107,
+    128,
+    148,
+    152,
+    158,
+    145,
+    151,
+    138,
+    149,
+    154,
+    157,
+    142,
+    133,
+    143,
+    156,
+    131,
+    147,
+    169,
+    176,
+    173,
+    180,
+    163,
+    164,
+    165,
+    160,
+    190,
+    177,
+    187,
+    168,
+    182,
+    171,
+    172,
+    191,
+    199,
+    218,
+    221,
+    210,
+    211,
+    203,
+    212,
+    202,
+    219,
+    208,
+    197,
+    200,
+    214,
+    205,
+    217,
+    213,
+    251,
+    249,
+    230,
+    228,
+    252,
+    255,
+    231,
+    226,
+    225,
+    246,
+    247,
+    233,
+    250,
+    242,
+    245,
+    229
+  ],
+  "layer_13": [
+    7,
+    23,
+    11,
+    26,
+    3,
+    15,
+    12,
+    17,
+    24,
+    21,
+    22,
+    27,
+    1,
+    2,
+    9,
+    8,
+    43,
+    58,
+    38,
+    53,
+    32,
+    51,
+    36,
+    60,
+    40,
+    61,
+    41,
+    48,
+    54,
+    49,
+    50,
+    55,
+    67,
+    87,
+    81,
+    83,
+    68,
+    65,
+    80,
+    86,
+    64,
+    92,
+    72,
+    94,
+    85,
+    89,
+    77,
+    66,
+    105,
+    106,
+    96,
+    100,
+    109,
+    99,
+    127,
+    124,
+    101,
+    113,
+    102,
+    110,
+    126,
+    120,
+    108,
+    115,
+    142,
+    139,
+    130,
+    133,
+    148,
+    150,
+    135,
+    159,
+    149,
+    153,
+    158,
+    141,
+    147,
+    155,
+    131,
+    156,
+    164,
+    186,
+    173,
+    170,
+    174,
+    172,
+    183,
+    187,
+    177,
+    167,
+    178,
+    166,
+    176,
+    165,
+    191,
+    161,
+    200,
+    223,
+    215,
+    201,
+    196,
+    212,
+    207,
+    205,
+    211,
+    218,
+    219,
+    221,
+    206,
+    222,
+    208,
+    213,
+    246,
+    226,
+    238,
+    235,
+    242,
+    239,
+    232,
+    255,
+    227,
+    231,
+    230,
+    252,
+    224,
+    243,
+    237,
+    225
+  ],
+  "layer_14": [
+    19,
+    2,
+    12,
+    16,
+    3,
+    31,
+    6,
+    8,
+    7,
+    24,
+    20,
+    29,
+    17,
+    21,
+    9,
+    26,
+    44,
+    61,
+    59,
+    33,
+    45,
+    58,
+    48,
+    54,
+    49,
+    35,
+    55,
+    43,
+    53,
+    63,
+    39,
+    41,
+    70,
+    69,
+    85,
+    76,
+    73,
+    81,
+    94,
+    88,
+    82,
+    91,
+    64,
+    78,
+    75,
+    80,
+    95,
+    83,
+    113,
+    110,
+    126,
+    106,
+    119,
+    104,
+    125,
+    103,
+    97,
+    107,
+    98,
+    118,
+    96,
+    108,
+    102,
+    109,
+    136,
+    151,
+    156,
+    158,
+    147,
+    132,
+    130,
+    153,
+    141,
+    140,
+    138,
+    155,
+    145,
+    131,
+    143,
+    146,
+    181,
+    183,
+    180,
+    169,
+    175,
+    174,
+    171,
+    168,
+    166,
+    190,
+    188,
+    173,
+    189,
+    187,
+    178,
+    185,
+    214,
+    205,
+    207,
+    209,
+    221,
+    201,
+    217,
+    206,
+    192,
+    211,
+    216,
+    203,
+    199,
+    208,
+    218,
+    194,
+    255,
+    246,
+    224,
+    230,
+    250,
+    240,
+    225,
+    253,
+    247,
+    254,
+    251,
+    252,
+    238,
+    236,
+    228,
+    239
+  ],
+  "layer_15": [
+    5,
+    9,
+    31,
+    14,
+    29,
+    11,
+    13,
+    10,
+    8,
+    12,
+    27,
+    22,
+    19,
+    7,
+    2,
+    3,
+    56,
+    47,
+    43,
+    61,
+    45,
+    58,
+    41,
+    36,
+    51,
+    63,
+    39,
+    33,
+    35,
+    34,
+    40,
+    57,
+    79,
+    77,
+    85,
+    86,
+    73,
+    75,
+    89,
+    92,
+    66,
+    95,
+    94,
+    90,
+    87,
+    80,
+    67,
+    71,
+    110,
+    96,
+    104,
+    102,
+    124,
+    126,
+    115,
+    113,
+    120,
+    122,
+    98,
+    99,
+    107,
+    105,
+    123,
+    127,
+    150,
+    155,
+    139,
+    134,
+    141,
+    144,
+    146,
+    148,
+    159,
+    145,
+    142,
+    129,
+    152,
+    130,
+    154,
+    157,
+    185,
+    182,
+    167,
+    179,
+    181,
+    171,
+    186,
+    177,
+    178,
+    165,
+    173,
+    162,
+    180,
+    175,
+    191,
+    164,
+    197,
+    205,
+    219,
+    194,
+    214,
+    198,
+    217,
+    196,
+    204,
+    222,
+    213,
+    210,
+    202,
+    207,
+    199,
+    223,
+    228,
+    248,
+    234,
+    251,
+    250,
+    227,
+    237,
+    243,
+    246,
+    254,
+    231,
+    230,
+    239,
+    240,
+    225,
+    224
+  ],
+  "layer_16": [
+    12,
+    1,
+    6,
+    29,
+    8,
+    14,
+    27,
+    18,
+    3,
+    15,
+    20,
+    31,
+    5,
+    4,
+    26,
+    10,
+    36,
+    59,
+    48,
+    54,
+    41,
+    49,
+    46,
+    53,
+    37,
+    50,
+    33,
+    44,
+    42,
+    38,
+    52,
+    35,
+    86,
+    68,
+    93,
+    71,
+    85,
+    79,
+    70,
+    78,
+    72,
+    83,
+    77,
+    74,
+    91,
+    89,
+    95,
+    65,
+    98,
+    103,
+    96,
+    108,
+    102,
+    124,
+    99,
+    127,
+    100,
+    97,
+    101,
+    125,
+    116,
+    105,
+    111,
+    112,
+    138,
+    147,
+    137,
+    149,
+    136,
+    143,
+    140,
+    151,
+    135,
+    128,
+    141,
+    134,
+    145,
+    139,
+    142,
+    129,
+    170,
+    160,
+    182,
+    185,
+    163,
+    191,
+    169,
+    171,
+    186,
+    189,
+    187,
+    178,
+    190,
+    179,
+    168,
+    174,
+    213,
+    211,
+    194,
+    221,
+    204,
+    192,
+    212,
+    218,
+    205,
+    217,
+    198,
+    216,
+    193,
+    203,
+    220,
+    195,
+    231,
+    247,
+    250,
+    235,
+    224,
+    241,
+    251,
+    245,
+    248,
+    227,
+    252,
+    244,
+    234,
+    253,
+    243,
+    230
+  ],
+  "layer_17": [
+    24,
+    17,
+    22,
+    0,
+    6,
+    8,
+    2,
+    4,
+    14,
+    15,
+    9,
+    7,
+    29,
+    12,
+    11,
+    28,
+    37,
+    33,
+    59,
+    49,
+    45,
+    51,
+    53,
+    50,
+    43,
+    57,
+    34,
+    38,
+    44,
+    52,
+    46,
+    36,
+    80,
+    85,
+    86,
+    81,
+    68,
+    71,
+    67,
+    75,
+    88,
+    93,
+    74,
+    64,
+    77,
+    70,
+    83,
+    90,
+    109,
+    111,
+    103,
+    107,
+    123,
+    125,
+    116,
+    126,
+    114,
+    117,
+    104,
+    108,
+    118,
+    101,
+    99,
+    113,
+    147,
+    136,
+    146,
+    143,
+    154,
+    130,
+    135,
+    133,
+    128,
+    142,
+    153,
+    134,
+    129,
+    131,
+    159,
+    139,
+    165,
+    162,
+    177,
+    184,
+    171,
+    185,
+    161,
+    176,
+    183,
+    170,
+    169,
+    188,
+    172,
+    187,
+    167,
+    179,
+    204,
+    193,
+    220,
+    222,
+    213,
+    217,
+    214,
+    206,
+    197,
+    194,
+    198,
+    215,
+    195,
+    209,
+    199,
+    211,
+    239,
+    227,
+    224,
+    250,
+    233,
+    235,
+    231,
+    251,
+    234,
+    229,
+    238,
+    247,
+    225,
+    245,
+    236,
+    241
+  ],
+  "layer_18": [
+    25,
+    22,
+    30,
+    29,
+    1,
+    19,
+    9,
+    10,
+    23,
+    28,
+    12,
+    16,
+    4,
+    24,
+    2,
+    17,
+    39,
+    51,
+    42,
+    46,
+    61,
+    43,
+    60,
+    50,
+    37,
+    55,
+    58,
+    40,
+    54,
+    34,
+    44,
+    62,
+    64,
+    78,
+    77,
+    86,
+    89,
+    65,
+    70,
+    69,
+    93,
+    83,
+    95,
+    81,
+    94,
+    71,
+    84,
+    75,
+    114,
+    123,
+    124,
+    127,
+    120,
+    103,
+    106,
+    119,
+    97,
+    107,
+    121,
+    122,
+    115,
+    99,
+    110,
+    96,
+    145,
+    158,
+    148,
+    142,
+    132,
+    129,
+    154,
+    146,
+    150,
+    152,
+    136,
+    131,
+    147,
+    128,
+    130,
+    156,
+    170,
+    171,
+    163,
+    161,
+    185,
+    172,
+    174,
+    180,
+    179,
+    166,
+    164,
+    169,
+    173,
+    191,
+    183,
+    187,
+    192,
+    203,
+    216,
+    220,
+    215,
+    202,
+    213,
+    218,
+    209,
+    206,
+    208,
+    200,
+    219,
+    195,
+    193,
+    201,
+    227,
+    228,
+    234,
+    251,
+    243,
+    230,
+    240,
+    254,
+    252,
+    232,
+    250,
+    246,
+    236,
+    233,
+    226,
+    229
+  ],
+  "layer_19": [
+    2,
+    14,
+    20,
+    28,
+    12,
+    17,
+    24,
+    19,
+    27,
+    29,
+    21,
+    5,
+    30,
+    31,
+    16,
+    4,
+    49,
+    42,
+    52,
+    50,
+    54,
+    38,
+    36,
+    53,
+    43,
+    47,
+    58,
+    56,
+    62,
+    46,
+    61,
+    35,
+    75,
+    71,
+    78,
+    72,
+    68,
+    86,
+    66,
+    64,
+    95,
+    89,
+    82,
+    85,
+    83,
+    90,
+    69,
+    80,
+    121,
+    118,
+    104,
+    107,
+    127,
+    114,
+    96,
+    102,
+    115,
+    124,
+    105,
+    97,
+    108,
+    109,
+    117,
+    125,
+    137,
+    133,
+    140,
+    149,
+    134,
+    146,
+    151,
+    131,
+    129,
+    153,
+    147,
+    128,
+    132,
+    136,
+    156,
+    130,
+    172,
+    177,
+    164,
+    190,
+    188,
+    181,
+    182,
+    183,
+    189,
+    191,
+    174,
+    173,
+    167,
+    175,
+    176,
+    166,
+    216,
+    194,
+    222,
+    219,
+    207,
+    205,
+    215,
+    206,
+    211,
+    213,
+    210,
+    197,
+    196,
+    218,
+    195,
+    202,
+    239,
+    231,
+    253,
+    237,
+    229,
+    242,
+    249,
+    232,
+    243,
+    255,
+    245,
+    228,
+    235,
+    238,
+    252,
+    225
+  ]
+}

generation_config.json

@@ -1,9 +1,9 @@
 {
-    "bos_token_id": 156891,
-    "eos_token_id": [
-        156892,
-        156895
-    ],
-    "pad_token_id": 156892,
-    "transformers_version": "4.52.3"
+  "bos_token_id": 156891,
+  "eos_token_id": [
+    156892,
+    156895
+  ],
+  "pad_token_id": 156892,
+  "transformers_version": "4.56.2"
 }

model-00001-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cae8ccb790ea396e88288a81e62356445631c34020744690f133d67bf2e9c494
+size 4998340808

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a461d1c12b98b5bd67e7b00f6f36bef0d7c2cd46a98adff63e73a7e60aee387
+size 4998907736

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c92b3d0a351c1ab831365415fd7278a2f1b81f777ce62421971108243ef2a78
+size 4998909232

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd15d9dc0dc949a6b508fc2dba53ce732a385695d8dcc37b756e810064bcc1d8
+size 2205269520

modeling_bailing_moe_v2.py

@@ -48,7 +48,7 @@ from transformers.utils import (
     replace_return_docstrings,
 )
 from transformers.utils.import_utils import is_torch_fx_available
-from configuration_bailing_moe_v2 import BailingMoeV2Config
+from .configuration_bailing_moe_v2 import BailingMoeV2Config
 from transformers.generation.utils import GenerationMixin
 from dataclasses import dataclass
 from transformers.utils import ModelOutput
@@ -1530,3 +1530,4 @@ class BailingMoeV2ForCausalLM(BailingMoeV2PreTrainedModel, GenerationMixin):
             attentions=outputs.attentions,
             router_logits=outputs.router_logits,
         )
+

pruning-code/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 inclusionAI
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

pruning-code/config.json

@@ -0,0 +1,50 @@
+{
+    "architectures": [
+        "BailingMoeV2ForCausalLM"
+    ],
+    "attention_dropout": 0.0,
+    "auto_map": {
+        "AutoConfig": "configuration_bailing_moe_v2.BailingMoeV2Config",
+        "AutoModel": "modeling_bailing_moe_v2.BailingMoeV2Model",
+        "AutoModelForCausalLM": "modeling_bailing_moe_v2.BailingMoeV2ForCausalLM"
+    },
+    "num_hidden_layers": 20,
+    "hidden_size": 2048,
+    "intermediate_size": 5120,
+    "eos_token_id": 156892,
+    "pad_token_id": 156892,
+    "first_k_dense_replace": 1,
+    "hidden_act": "silu",
+    "max_position_embeddings": 32768,
+    "model_type": "bailing_moe",
+    "moe_intermediate_size": 512,
+    "norm_topk_prob": true,
+    "num_experts_per_tok": 8,
+    "num_attention_heads": 16,
+    "num_experts": 256,
+    "num_key_value_heads": 4,
+    "rope_theta": 600000,
+    "rope_scaling": null,
+    "tie_word_embeddings": false,
+    "torch_dtype": "bfloat16",
+    "transformers_version": "4.52.3",
+    "use_bias": false,
+    "use_rmsnorm": true,
+    "rms_norm_eps": 1e-06,
+    "head_dim": 128,
+    "num_shared_experts": 1,
+    "use_cache": true,
+    "use_qkv_bias": false,
+    "embedding_dropout": 0.0,
+    "output_dropout": 0.0,
+    "vocab_size": 157184,
+    "partial_rotary_factor": 0.5,
+    "router_dtype": "fp32",
+    "moe_router_enable_expert_bias": true,
+    "routed_scaling_factor": 2.5,
+    "n_group": 8,
+    "topk_group": 4,
+    "use_qk_norm": true,
+    "score_function": "sigmoid",
+    "moe_shared_expert_intermediate_size": 512
+}

pruning-code/configuration_bailing_moe_v2.py

@@ -0,0 +1,84 @@
+"""Bailing MoE V2 model configuration"""
+
+from transformers.configuration_utils import PretrainedConfig
+
+
+class BailingMoeV2Config(PretrainedConfig):
+
+    def __init__(
+        self,
+        vocab_size=157184,
+        hidden_size=2048,
+        intermediate_size=5120,
+        num_hidden_layers=20,
+        num_attention_heads=16,
+        num_key_value_heads=4,
+        hidden_act="silu",
+        use_qkv_bias=False,  # bailing only
+        use_bias=False,  # bailing only
+        rms_norm_eps=1e-06,
+        tie_word_embeddings=False,  # PretrainedConfig key, here change default value.
+        embedding_dropout=0.0,
+        attention_dropout=0.0,
+        output_dropout=0.0,
+        initializer_range=0.02,
+        max_position_embeddings=32768,
+        rope_theta=600000.0,
+        use_cache=True,
+        max_window_layers=20,
+        rope_scaling=None,
+        pad_token_id=156892,
+        eos_token_id=156892,
+        num_experts=256,
+        num_shared_experts=1,
+        num_experts_per_tok=8,
+        n_group=8,
+        topk_group=4,
+        moe_intermediate_size=512,
+        first_k_dense_replace=1,
+        head_dim=128,
+        output_router_logits=False,
+        use_qk_norm=True,
+        num_nextn_predict_layers=0,
+        mtp_loss_scaling_factor=0,
+        moe_router_enable_expert_bias=True,
+        routed_scaling_factor=1.0,
+        **kwargs,
+    ):
+        self.num_hidden_layers = num_hidden_layers
+        self.vocab_size = vocab_size
+        self.hidden_size = hidden_size
+        self.intermediate_size = intermediate_size
+        self.num_attention_heads = num_attention_heads
+        self.num_key_value_heads = num_key_value_heads
+        self.hidden_act = hidden_act
+        self.use_qkv_bias = use_qkv_bias
+        self.use_bias = use_bias
+        self.rms_norm_eps = rms_norm_eps
+        self.embedding_dropout = embedding_dropout
+        self.attention_dropout = attention_dropout
+        self.output_dropout = output_dropout
+        self.num_nextn_predict_layers = num_nextn_predict_layers
+        self.mtp_loss_scaling_factor = mtp_loss_scaling_factor
+        self.initializer_range = initializer_range
+        self.max_position_embeddings = max_position_embeddings
+        self.rope_theta = rope_theta
+        self.use_cache = use_cache
+        self.max_window_layers = max_window_layers
+        self.head_dim = head_dim or self.hidden_size // self.num_attention_heads
+        self.rope_scaling = rope_scaling
+        self.use_qk_norm = use_qk_norm
+        self.moe_router_enable_expert_bias = moe_router_enable_expert_bias
+        self.routed_scaling_factor = routed_scaling_factor
+
+        # MoE configs
+        self.num_experts = num_experts
+        self.num_shared_experts = num_shared_experts
+        self.num_experts_per_tok = num_experts_per_tok
+        self.n_group = n_group
+        self.topk_group = topk_group
+        self.moe_intermediate_size = moe_intermediate_size
+        self.first_k_dense_replace = first_k_dense_replace
+        self.output_router_logits = output_router_logits
+
+        super().__init__(pad_token_id=pad_token_id, eos_token_id=eos_token_id, tie_word_embeddings=tie_word_embeddings, **kwargs)

pruning-code/generation_config.json

@@ -0,0 +1,9 @@
+{
+    "bos_token_id": 156891,
+    "eos_token_id": [
+        156892,
+        156895
+    ],
+    "pad_token_id": 156892,
+    "transformers_version": "4.52.3"
+}

pruning-code/modeling_bailing_moe_v2.py

@@ -0,0 +1,1532 @@
+# coding=utf-8
+# Copyright 2025 Antgroup and The HuggingFace Inc. team. All rights reserved.
+#
+# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
+# and OPT implementations in this library. It has been modified from its
+# original forms to accommodate minor architectural differences compared
+# to GPT-NeoX and OPT used by the Meta AI team that trained the model.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""PyTorch BailingMoE model."""
+
+import math
+import warnings
+from typing import List, Optional, Tuple, Union
+
+import torch
+import torch.nn.functional as F
+from torch import nn
+
+from transformers.activations import ACT2FN
+from transformers.cache_utils import Cache, DynamicCache
+from transformers.modeling_attn_mask_utils import (
+    AttentionMaskConverter,
+    _prepare_4d_attention_mask,
+    _prepare_4d_causal_attention_mask,
+    _prepare_4d_causal_attention_mask_for_sdpa,
+)
+from transformers.modeling_outputs import MoeModelOutputWithPast
+from transformers.modeling_rope_utils import ROPE_INIT_FUNCTIONS, dynamic_rope_update
+from transformers.modeling_utils import PreTrainedModel
+from transformers.pytorch_utils import ALL_LAYERNORM_LAYERS, is_torch_greater_or_equal_than_1_13
+from transformers.utils import (
+    add_start_docstrings,
+    add_start_docstrings_to_model_forward,
+    is_flash_attn_2_available,
+    is_flash_attn_greater_or_equal_2_10,
+    logging,
+    replace_return_docstrings,
+)
+from transformers.utils.import_utils import is_torch_fx_available
+from configuration_bailing_moe_v2 import BailingMoeV2Config
+from transformers.generation.utils import GenerationMixin
+from dataclasses import dataclass
+from transformers.utils import ModelOutput
+
+
+if is_flash_attn_2_available():
+    from flash_attn import flash_attn_func, flash_attn_varlen_func
+    from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input  # noqa
+
+
+# This makes `_prepare_4d_causal_attention_mask` a leaf function in the FX graph.
+# It means that the function will not be traced through and simply appear as a node in the graph.
+if is_torch_fx_available():
+    if not is_torch_greater_or_equal_than_1_13:
+        import torch.fx
+
+    _prepare_4d_causal_attention_mask = torch.fx.wrap(_prepare_4d_causal_attention_mask)
+
+
+logger = logging.get_logger(__name__)
+
+_CONFIG_FOR_DOC = "BailingMoeV2Config"
+
+
+def roll_tensor(tensor, shifts=-1, dims=-1, fill_value=0):
+    """Roll the tensor input along the given dimension(s).
+    Inserted elements are set to be 0.0.
+    """
+    rolled_tensor = torch.roll(tensor, shifts=shifts, dims=dims)
+    rolled_tensor.select(dims, shifts).fill_(fill_value)
+    return rolled_tensor, rolled_tensor.sum()
+
+
+@dataclass
+class MoEV2CausalLMOutputWithPast(ModelOutput):
+    """
+    Base class for causal language model (or autoregressive) outputs as well as Mixture of Expert's router hidden
+    states terms, to train a MoE model.
+
+    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 (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+            It is a [`~cache_utils.Cache`] instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache).
+
+            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.
+        z_loss (`torch.FloatTensor`, *optional*, returned when `labels` is provided):
+            z_loss for the sparse modules.
+        aux_loss (`torch.FloatTensor`, *optional*, returned when `labels` is provided):
+            aux_loss for the sparse modules.
+        router_logits (`tuple(torch.FloatTensor)`, *optional*, returned when `output_router_logits=True` is passed or when `config.add_router_probs=True`):
+            Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, sequence_length, num_experts)`.
+
+            Router logits of the encoder model, useful to compute the auxiliary loss and the z_loss for the sparse
+            modules.
+    """
+
+    loss: Optional[torch.FloatTensor] = None
+    logits: Optional[torch.FloatTensor] = None
+    past_key_values: Optional[Cache] = None
+    hidden_states: Optional[tuple[torch.FloatTensor, ...]] = None
+    attentions: Optional[tuple[torch.FloatTensor, ...]] = None
+    z_loss: Optional[torch.FloatTensor] = None
+    aux_loss: Optional[torch.FloatTensor] = None
+    router_logits: Optional[tuple[torch.FloatTensor]] = None
+    mtp_loss: Optional[torch.FloatTensor] = None
+    mtp_logits: Optional[tuple[torch.FloatTensor, ...]] = None
+
+
+class MoeV2ModelOutputWithPast(MoeModelOutputWithPast):
+
+    def __init__(self, mtp_hidden_states=None, **kwargs):
+        super().__init__(**kwargs)
+        self.mtp_hidden_states = mtp_hidden_states
+
+
+def _get_unpad_data(attention_mask):
+    seqlens_in_batch = attention_mask.sum(dim=-1, dtype=torch.int32)
+    indices = torch.nonzero(attention_mask.flatten(), as_tuple=False).flatten()
+    max_seqlen_in_batch = seqlens_in_batch.max().item()
+    cu_seqlens = F.pad(torch.cumsum(seqlens_in_batch, dim=0, dtype=torch.torch.int32), (1, 0))
+    return (
+        indices,
+        cu_seqlens,
+        max_seqlen_in_batch,
+    )
+
+
+def _expand_mask(mask: torch.Tensor, dtype: torch.dtype, tgt_len: Optional[int] = None):
+    warnings.warn(
+        "Calling `transformers.models.BailingMoeV2.modeling_BailingMoeV2._prepare_4d_attention_mask` is deprecated and will be removed in v4.37. Use `transformers.modeling_attn_mask_utils._prepare_4d_attention_mask"
+    )
+    return _prepare_4d_attention_mask(mask=mask, dtype=dtype, tgt_len=tgt_len)
+
+
+def _make_causal_mask(
+    input_ids_shape: torch.Size, dtype: torch.dtype, device: torch.device, past_key_values_length: int = 0
+):
+    warnings.warn(
+        "Calling `transformers.models.BailingMoeV2.modeling_BailingMoeV2._make_causal_mask` is deprecated and will be removed in v4.37. Use `transformers.models.BailingMoeV2.modeling_BailingMoeV2.AttentionMaskConverter._make_causal_mask"
+    )
+    return AttentionMaskConverter._make_causal_mask(
+        input_ids_shape=input_ids_shape, dtype=dtype, device=device, past_key_values_length=past_key_values_length
+    )
+
+
+class BailingMoeV2RMSNorm(nn.Module):
+    def __init__(self, hidden_size, eps=1e-6):
+        """
+        BailingMoeV2RMSNorm is equivalent to T5LayerNorm
+        """
+        super().__init__()
+        self.weight = nn.Parameter(torch.ones(hidden_size))
+        self.variance_epsilon = eps
+
+    def forward(self, hidden_states):
+        input_dtype = hidden_states.dtype
+        hidden_states = hidden_states.to(torch.float32)
+        variance = hidden_states.pow(2).mean(-1, keepdim=True)
+        hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)
+        return self.weight * hidden_states.to(input_dtype)
+
+
+ALL_LAYERNORM_LAYERS.append(BailingMoeV2RMSNorm)
+
+
+class BailingMoeV2RotaryEmbedding(nn.Module):
+    def __init__(self, config: BailingMoeV2Config, device=None):
+        super().__init__()
+        # BC: "rope_type" was originally "type"
+        if hasattr(config, "rope_scaling") and config.rope_scaling is not None:
+            self.rope_type = config.rope_scaling.get("rope_type", config.rope_scaling.get("type"))
+        else:
+            self.rope_type = "default"
+        self.max_seq_len_cached = config.max_position_embeddings
+        self.original_max_seq_len = config.max_position_embeddings
+
+        self.config = config
+        self.rope_init_fn = ROPE_INIT_FUNCTIONS[self.rope_type]
+
+        inv_freq, self.attention_scaling = self.rope_init_fn(self.config, device)
+        self.register_buffer("inv_freq", inv_freq, persistent=False)
+        self.original_inv_freq = self.inv_freq
+
+    @torch.no_grad()
+    @dynamic_rope_update  # power user: used with advanced RoPE types (e.g. dynamic rope)
+    def forward(self, x, position_ids):
+        inv_freq_expanded = self.inv_freq[None, :, None].float().expand(position_ids.shape[0], -1, 1).to(x.device)
+        position_ids_expanded = position_ids[:, None, :].float()
+
+        device_type = x.device.type if isinstance(x.device.type, str) and x.device.type != "mps" else "cpu"
+        with torch.autocast(device_type=device_type, enabled=False):  # Force float32
+            freqs = (inv_freq_expanded.float() @ position_ids_expanded.float()).transpose(1, 2)
+            emb = torch.cat((freqs, freqs), dim=-1)
+            cos = emb.cos() * self.attention_scaling
+            sin = emb.sin() * self.attention_scaling
+
+        return cos.to(dtype=x.dtype), sin.to(dtype=x.dtype)
+
+
+# Copied from transformers.models.llama.modeling_llama.rotate_half
+def rotate_half(x):
+    """Rotates half the hidden dims of the input."""
+    x1 = x[..., : x.shape[-1] // 2]
+    x2 = x[..., x.shape[-1] // 2 :]
+    return torch.cat((-x2, x1), dim=-1)
+
+
+# Copied from transformers.models.llama.modeling_llama.apply_rotary_pos_emb
+def apply_rotary_pos_emb(q, k, cos, sin, unsqueeze_dim=1):
+    """Applies Rotary Position Embedding to the query and key tensors.
+
+    Args:
+        q (`torch.Tensor`): The query tensor.
+        k (`torch.Tensor`): The key tensor.
+        cos (`torch.Tensor`): The cosine part of the rotary embedding.
+        sin (`torch.Tensor`): The sine part of the rotary embedding.
+        unsqueeze_dim (`int`, *optional*, defaults to 1):
+            The 'unsqueeze_dim' argument specifies the dimension along which to unsqueeze cos[position_ids] and
+            sin[position_ids] so that they can be properly broadcasted to the dimensions of q and k. For example, note
+            that cos[position_ids] and sin[position_ids] have the shape [batch_size, seq_len, head_dim]. Then, if q and
+            k have the shape [batch_size, heads, seq_len, head_dim], then setting unsqueeze_dim=1 makes
+            cos[position_ids] and sin[position_ids] broadcastable to the shapes of q and k. Similarly, if q and k have
+            the shape [batch_size, seq_len, heads, head_dim], then set unsqueeze_dim=2.
+    Returns:
+        `tuple(torch.Tensor)` comprising the query and key tensors rotated using the Rotary Position Embedding.
+    """
+    cos = cos.unsqueeze(unsqueeze_dim)
+    sin = sin.unsqueeze(unsqueeze_dim)
+
+    # Keep half or full tensor for later concatenation
+    rotary_dim = cos.shape[-1]
+    q_rot, q_pass = q[..., :rotary_dim], q[..., rotary_dim:]
+    k_rot, k_pass = k[..., :rotary_dim], k[..., rotary_dim:]
+
+    # Apply rotary embeddings on the first half or full tensor
+    q_embed = (q_rot * cos) + (rotate_half(q_rot) * sin)
+    k_embed = (k_rot * cos) + (rotate_half(k_rot) * sin)
+
+    # Concatenate back to full shape
+    q_embed = torch.cat([q_embed, q_pass], dim=-1)
+    k_embed = torch.cat([k_embed, k_pass], dim=-1)
+    return q_embed, k_embed
+
+
+class BailingMoeV2MLP(nn.Module):
+    def __init__(self, config: BailingMoeV2Config, intermediate_size: int):
+        super().__init__()
+        self.config = config
+        self.hidden_size = config.hidden_size
+        self.intermediate_size = intermediate_size
+
+        self.gate_proj = nn.Linear(self.hidden_size, self.intermediate_size, bias=False)
+        self.up_proj = nn.Linear(self.hidden_size, self.intermediate_size, bias=False)
+        self.down_proj = nn.Linear(self.intermediate_size, self.hidden_size, bias=False)
+        self.act_fn = ACT2FN[config.hidden_act]
+
+    def forward(self, x):
+        return self.down_proj(self.act_fn(self.gate_proj(x)) * self.up_proj(x))
+
+
+class BailingMoeV2Gate(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.config = config
+        self.top_k = config.num_experts_per_tok
+        self.num_experts = config.num_experts
+
+        self.n_group = config.n_group
+        self.topk_group = config.topk_group
+
+        # topk selection algorithm
+        self.gating_dim = config.hidden_size
+        self.weight = nn.Parameter(torch.empty((self.num_experts, self.gating_dim)))
+        self.routed_scaling_factor = config.routed_scaling_factor
+
+        self.register_buffer("expert_bias", torch.zeros((self.num_experts)))
+        self.reset_parameters()
+
+    def reset_parameters(self) -> None:
+        import torch.nn.init as init
+
+        init.kaiming_uniform_(self.weight, a=math.sqrt(5))
+
+    def group_limited_topk(
+        self,
+        scores: torch.Tensor,
+    ):
+        num_tokens, _ = scores.size()
+        # Organize the experts into groups
+        group_scores = scores.view(num_tokens, self.n_group, -1).topk(2, dim=-1)[0].sum(dim=-1)
+        group_idx = torch.topk(group_scores, k=self.topk_group, dim=-1, sorted=False)[1]
+        group_mask = torch.zeros_like(group_scores)
+        group_mask.scatter_(1, group_idx, 1)
+
+        # Mask the experts based on selection groups
+        score_mask = (
+            group_mask.unsqueeze(-1)
+            .expand(num_tokens, self.n_group, self.num_experts // self.n_group)
+            .reshape(num_tokens, -1)
+        )
+
+        masked_scores = scores.masked_fill(~score_mask.bool(), float('-inf'))
+        probs, top_indices = torch.topk(masked_scores, k=self.top_k, dim=-1)
+
+        return probs, top_indices
+
+    def forward(self, hidden_states):
+        # compute gating score
+        hidden_states = hidden_states.view(-1, hidden_states.shape[-1])
+        logits = F.linear(hidden_states.type(torch.float32), self.weight.type(torch.float32))
+
+        scores = torch.sigmoid(logits.float()).type_as(logits)
+
+        scores_for_routing = scores + self.expert_bias
+        _, topk_idx = self.group_limited_topk(scores_for_routing)
+
+        scores = torch.gather(scores, dim=1, index=topk_idx).type_as(logits)
+
+        topk_weight = scores / (scores.sum(dim=-1, keepdim=True) + 1e-20) if self.top_k > 1 else scores
+        topk_weight = topk_weight * self.routed_scaling_factor
+
+        return topk_idx, topk_weight, logits
+
+
+class BailingMoeV2SparseMoeBlock(nn.Module):
+    """
+    A mixed expert module containing shared experts.
+    """
+
+    def __init__(self, config: BailingMoeV2Config):
+        super().__init__()
+        self.config = config
+        self.num_experts_per_tok = config.num_experts_per_tok
+        self._setup_experts()
+        self.gate = BailingMoeV2Gate(config)
+        if config.num_shared_experts is not None:
+            self.shared_experts = BailingMoeV2MLP(
+                config=config, intermediate_size=config.moe_intermediate_size * config.num_shared_experts
+            )
+
+    def _setup_experts(self):
+        self.experts = nn.ModuleList(
+            [
+                BailingMoeV2MLP(config=self.config, intermediate_size=self.config.moe_intermediate_size)
+                for _ in range(self.config.num_experts)
+            ]
+        )
+
+    def forward(self, hidden_states):
+        identity = hidden_states
+        bsz, seq_len, h = hidden_states.shape
+        topk_idx, topk_weight, router_logits = self.gate(hidden_states)
+        hidden_states = hidden_states.view(-1, hidden_states.shape[-1])
+        flat_topk_idx = topk_idx.view(-1)
+        if self.training:
+            hidden_states = hidden_states.repeat_interleave(self.num_experts_per_tok, dim=0)
+            y = torch.empty_like(hidden_states)
+            for i, expert in enumerate(self.experts):
+                y[flat_topk_idx == i] = expert(hidden_states[flat_topk_idx == i])
+            y = (y.view(*topk_weight.shape, -1) * topk_weight.unsqueeze(-1)).sum(dim=1)
+            y = y.to(hidden_states.dtype).view(bsz, seq_len, h)
+        else:
+            y = self.moe_infer(hidden_states, topk_idx, topk_weight).view(bsz, seq_len, h)
+        if self.config.num_shared_experts is not None:
+            y = y + self.shared_experts(identity)
+        return y, (router_logits.view(bsz, seq_len, -1), topk_idx.view(bsz, seq_len, -1))
+
+    @torch.no_grad()
+    def moe_infer(self, x, topk_ids, topk_weight):
+        cnts = topk_ids.new_zeros((topk_ids.shape[0], len(self.experts)))
+        cnts.scatter_(1, topk_ids, 1)
+        tokens_per_expert = cnts.sum(dim=0)
+        idxs = topk_ids.view(-1).argsort()
+        sorted_tokens = x[idxs // topk_ids.shape[1]]
+        tokens_per_expert = tokens_per_expert.cpu().numpy()
+        outputs = []
+        start_idx = 0
+        for i, num_tokens in enumerate(tokens_per_expert):
+            end_idx = start_idx + num_tokens
+            if num_tokens == 0:
+                continue
+            expert = self.experts[i]
+            tokens_for_this_expert = sorted_tokens[start_idx:end_idx]
+            expert_out = expert(tokens_for_this_expert)
+            outputs.append(expert_out.to(x.device))
+            start_idx = end_idx
+
+        outs = torch.cat(outputs, dim=0) if len(outputs) else sorted_tokens.new_empty(0)
+        new_x = torch.empty_like(outs)
+        new_x[idxs] = outs
+        final_out = (
+            new_x.view(*topk_ids.shape, -1)
+            .type(topk_weight.dtype)
+            .mul_(topk_weight.unsqueeze(dim=-1))
+            .sum(dim=1)
+            .type(new_x.dtype)
+        )
+        return final_out
+
+
+# Copied from transformers.models.llama.modeling_llama.repeat_kv
+def repeat_kv(hidden_states: torch.Tensor, n_rep: int) -> torch.Tensor:
+    """
+    This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch,
+    num_key_value_heads, seqlen, head_dim) to (batch, num_attention_heads, seqlen, head_dim)
+    """
+    batch, num_key_value_heads, slen, head_dim = hidden_states.shape
+    if n_rep == 1:
+        return hidden_states
+    hidden_states = hidden_states[:, :, None, :, :].expand(batch, num_key_value_heads, n_rep, slen, head_dim)
+    return hidden_states.reshape(batch, num_key_value_heads * n_rep, slen, head_dim)
+
+
+# Copied from transformers.models.llama.modeling_llama.LlamaAttention with Llama->BailingMoeV2
+class BailingMoeV2Attention(nn.Module):
+    """Multi-headed attention from 'Attention Is All You Need' paper"""
+
+    def __init__(self, config: BailingMoeV2Config, layer_idx: Optional[int] = None):
+        super().__init__()
+        self.config = config
+        self.layer_idx = layer_idx
+        if layer_idx is None:
+            logger.warning_once(
+                f"Instantiating {self.__class__.__name__} without passing `layer_idx` is not recommended and will "
+                "to errors during the forward call, if caching is used. Please make sure to provide a `layer_idx` "
+                "when creating this class."
+            )
+
+        self.attention_dropout = config.attention_dropout
+        self.hidden_size = config.hidden_size
+        self.num_heads = config.num_attention_heads
+        self.head_dim = config.head_dim or self.hidden_size // self.num_heads
+        partial_rotary_factor = config.partial_rotary_factor if hasattr(config, "partial_rotary_factor") else 1.0
+        self.rope_dim = int(self.head_dim * partial_rotary_factor)
+        self.num_key_value_heads = config.num_key_value_heads
+        self.num_key_value_groups = self.num_heads // self.num_key_value_heads
+        self.max_position_embeddings = config.max_position_embeddings
+        self.rope_theta = config.rope_theta
+        self.is_causal = True
+
+        self.query_key_value = nn.Linear(
+            self.hidden_size,
+            (self.num_heads + 2 * self.num_key_value_heads) * self.head_dim,
+            bias=config.use_qkv_bias,
+        )
+
+        if self.config.use_qk_norm:
+            self.query_layernorm = BailingMoeV2RMSNorm(self.head_dim, eps=config.rms_norm_eps)
+            self.key_layernorm = BailingMoeV2RMSNorm(self.head_dim, eps=config.rms_norm_eps)
+        self.dense = nn.Linear(self.num_heads * self.head_dim, self.hidden_size, bias=config.use_bias)
+
+    def _shape(self, tensor: torch.Tensor, seq_len: int, bsz: int):
+        return tensor.view(bsz, seq_len, self.num_heads, self.head_dim).transpose(1, 2).contiguous()
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+        position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None,  # necessary, but kept here for BC
+        **kwargs,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+
+        bsz, q_len, _ = hidden_states.size()
+
+        qkv = self.query_key_value(hidden_states)
+        qkv = qkv.view(bsz, q_len, self.num_heads + 2 * self.num_key_value_heads, self.head_dim)
+
+        query_states, key_states, value_states = qkv.split(
+            [self.num_heads, self.num_key_value_heads, self.num_key_value_heads], dim=-2
+        )
+        query_states = query_states.transpose(1, 2)
+        key_states = key_states.transpose(1, 2)
+        value_states = value_states.transpose(1, 2)
+
+        if self.config.use_qk_norm:
+            query_states = self.query_layernorm(query_states)
+            key_states = self.key_layernorm(key_states)
+
+        cos, sin = position_embeddings
+        query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin)
+
+        if past_key_value is not None:
+            if self.layer_idx is None:
+                raise ValueError(
+                    f"The cache structure has changed since version v4.36. If you are using {self.__class__.__name__} "
+                    "for auto-regressive decoding with k/v caching, please make sure to initialize the attention class "
+                    "with a layer index."
+                )
+            cache_kwargs = {"sin": sin, "cos": cos}
+            key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs)
+
+        key_states = repeat_kv(key_states, self.num_key_value_groups)
+        value_states = repeat_kv(value_states, self.num_key_value_groups)
+
+        attn_weights = torch.matmul(query_states, key_states.transpose(2, 3)) / math.sqrt(self.head_dim)
+
+        kv_seq_len = key_states.shape[-2]
+        if attn_weights.size() != (bsz, self.num_heads, q_len, kv_seq_len):
+            raise ValueError(
+                f"Attention weights should be of size {(bsz, self.num_heads, q_len, kv_seq_len)}, but is"
+                f" {attn_weights.size()}"
+            )
+
+        if attention_mask is not None:
+            if attention_mask.size() != (bsz, 1, q_len, kv_seq_len):
+                raise ValueError(
+                    f"Attention mask should be of size {(bsz, 1, q_len, kv_seq_len)}, but is {attention_mask.size()}"
+                )
+            attn_weights = attn_weights + attention_mask
+
+        # upcast attention to fp32
+        attn_weights = nn.functional.softmax(attn_weights, dim=-1, dtype=torch.float32).to(query_states.dtype)
+        attn_weights = nn.functional.dropout(attn_weights, p=self.attention_dropout, training=self.training)
+        attn_output = torch.matmul(attn_weights, value_states)
+
+        if attn_output.size() != (bsz, self.num_heads, q_len, self.head_dim):
+            raise ValueError(
+                f"`attn_output` should be of size {(bsz, self.num_heads, q_len, self.head_dim)}, but is"
+                f" {attn_output.size()}"
+            )
+
+        attn_output = attn_output.transpose(1, 2).contiguous()
+
+        attn_output = attn_output.reshape(bsz, q_len, -1)
+
+        attn_output = self.dense(attn_output)
+
+        if not output_attentions:
+            attn_weights = None
+
+        return attn_output, attn_weights, past_key_value
+
+
+# Copied from transformers.models.llama.modeling_llama.LlamaFlashAttention2 with Llama->BailingMoeV2
+class BailingMoeV2FlashAttention2(BailingMoeV2Attention):
+    """
+    BailingMoeV2 flash attention module. This module inherits from `BailingMoeV2Attention` as the weights of the module stays
+    untouched. The only required change would be on the forward pass where it needs to correctly call the public API of
+    flash attention and deal with padding tokens in case the input contains any of them.
+    """
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+
+        # TODO: Should be removed once Flash Attention for RoCm is bumped to 2.1.
+        # flash_attn<2.1 generates top-left aligned causal mask, while what is needed here is bottom-right alignement, that was made default for flash_attn>=2.1. This attribute is used to handle this difference. Reference: https://github.com/Dao-AILab/flash-attention/releases/tag/v2.1.0.
+        # Beware that with flash_attn<2.1, using q_seqlen != k_seqlen (except for the case q_seqlen == 1) produces a wrong mask (top-left).
+        self._flash_attn_uses_top_left_mask = not is_flash_attn_greater_or_equal_2_10()
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.LongTensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+        position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None,  # necessary, but kept here for BC
+        **kwargs,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+        # BailingMoeV2FlashAttention2 attention does not support output_attentions
+        output_attentions = False
+
+        bsz, q_len, _ = hidden_states.size()
+
+        # Flash attention requires the input to have the shape
+        # batch_size x seq_length x head_dim x hidden_dim
+        # therefore we just need to keep the original shape
+
+        qkv = self.query_key_value(hidden_states)
+        qkv = qkv.view(bsz, q_len, self.num_heads + 2 * self.num_key_value_heads, self.head_dim)
+
+        query_states, key_states, value_states = qkv.split(
+            [self.num_heads, self.num_key_value_heads, self.num_key_value_heads], dim=-2
+        )
+        query_states = query_states.transpose(1, 2)
+        key_states = key_states.transpose(1, 2)
+        value_states = value_states.transpose(1, 2)
+
+        if self.config.use_qk_norm:
+            query_states = self.query_layernorm(query_states)
+            key_states = self.key_layernorm(key_states)
+
+        cos, sin = position_embeddings
+        query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin)
+
+        if past_key_value is not None:
+            cache_kwargs = {"sin": sin, "cos": cos}
+            key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs)
+
+        # TODO: These transpose are quite inefficient but Flash Attention requires the layout [batch_size, sequence_length, num_heads, head_dim]. We would need to refactor the KV cache
+        # to be able to avoid many of these transpose/reshape/view.
+        query_states = query_states.transpose(1, 2)
+        key_states = key_states.transpose(1, 2)
+        value_states = value_states.transpose(1, 2)
+
+        dropout_rate = self.attention_dropout if self.training else 0.0
+
+        # In PEFT, usually we cast the layer norms in float32 for training stability reasons
+        # therefore the input hidden states gets silently cast in float32. Hence, we need
+        # cast them back in the correct dtype just to be sure everything works as expected.
+        # This might slow down training & inference so it is recommended to not cast the LayerNorms
+        # in fp32. (BailingMoeV2RMSNorm handles it correctly)
+
+        input_dtype = query_states.dtype
+        if input_dtype == torch.float32:
+            # Handle the case where the model is quantized
+            if hasattr(self.config, "_pre_quantization_dtype"):
+                target_dtype = self.config._pre_quantization_dtype
+            elif torch.is_autocast_enabled():
+                target_dtype = torch.get_autocast_gpu_dtype()
+            else:
+                target_dtype = self.query_key_value.weight.dtype
+
+            logger.warning_once(
+                f"The input hidden states seems to be silently casted in float32, this might be related to"
+                f" the fact you have upcasted embedding or layer norm layers in float32. We will cast back the input in"
+                f" {target_dtype}."
+            )
+
+            query_states = query_states.to(target_dtype)
+            key_states = key_states.to(target_dtype)
+            value_states = value_states.to(target_dtype)
+
+        attn_output = self._flash_attention_forward(
+            query_states, key_states, value_states, attention_mask, q_len, dropout=dropout_rate
+        )
+
+        attn_output = attn_output.reshape(bsz, q_len, -1).contiguous()
+        attn_output = self.dense(attn_output)
+
+        if not output_attentions:
+            attn_weights = None
+
+        return attn_output, attn_weights, past_key_value
+
+    def _flash_attention_forward(
+        self, query_states, key_states, value_states, attention_mask, query_length, dropout=0.0, softmax_scale=None
+    ):
+        """
+        Calls the forward method of Flash Attention - if the input hidden states contain at least one padding token
+        first unpad the input, then computes the attention scores and pad the final attention scores.
+
+        Args:
+            query_states (`torch.Tensor`):
+                Input query states to be passed to Flash Attention API
+            key_states (`torch.Tensor`):
+                Input key states to be passed to Flash Attention API
+            value_states (`torch.Tensor`):
+                Input value states to be passed to Flash Attention API
+            attention_mask (`torch.Tensor`):
+                The padding mask - corresponds to a tensor of size `(batch_size, seq_len)` where 0 stands for the
+                position of padding tokens and 1 for the position of non-padding tokens.
+            dropout (`int`, *optional*):
+                Attention dropout
+            softmax_scale (`float`, *optional*):
+                The scaling of QK^T before applying softmax. Default to 1 / sqrt(head_dim)
+            query_length (`int`):
+                The length of the query sequence in terms of tokens. This represents the number of tokens in the
+                `query_states` tensor along the sequence dimension. It is used to determine the effective sequence
+                length for attention computations.
+        """
+        if not self._flash_attn_uses_top_left_mask:
+            causal = self.is_causal
+        else:
+            # TODO: Remove the `query_length != 1` check once Flash Attention for RoCm is bumped to 2.1. For details, please see the comment in BailingMoeV2FlashAttention2 __init__.
+            causal = self.is_causal and query_length != 1
+
+        # Contains at least one padding token in the sequence
+        if attention_mask is not None:
+            batch_size = query_states.shape[0]
+            query_states, key_states, value_states, indices_q, cu_seq_lens, max_seq_lens = self._upad_input(
+                query_states, key_states, value_states, attention_mask, query_length
+            )
+
+            cu_seqlens_q, cu_seqlens_k = cu_seq_lens
+            max_seqlen_in_batch_q, max_seqlen_in_batch_k = max_seq_lens
+
+            attn_output_unpad = flash_attn_varlen_func(
+                query_states,
+                key_states,
+                value_states,
+                cu_seqlens_q=cu_seqlens_q,
+                cu_seqlens_k=cu_seqlens_k,
+                max_seqlen_q=max_seqlen_in_batch_q,
+                max_seqlen_k=max_seqlen_in_batch_k,
+                dropout_p=dropout,
+                softmax_scale=softmax_scale,
+                causal=causal,
+            )
+
+            attn_output = pad_input(attn_output_unpad, indices_q, batch_size, query_length)
+        else:
+            attn_output = flash_attn_func(
+                query_states, key_states, value_states, dropout, softmax_scale=softmax_scale, causal=causal
+            )
+
+        return attn_output
+
+    def _upad_input(self, query_layer, key_layer, value_layer, attention_mask, query_length):
+        indices_k, cu_seqlens_k, max_seqlen_in_batch_k = _get_unpad_data(attention_mask)
+        batch_size, kv_seq_len, num_key_value_heads, head_dim = key_layer.shape
+
+        key_layer = index_first_axis(
+            key_layer.reshape(batch_size * kv_seq_len, num_key_value_heads, head_dim), indices_k
+        )
+        value_layer = index_first_axis(
+            value_layer.reshape(batch_size * kv_seq_len, num_key_value_heads, head_dim), indices_k
+        )
+        if query_length == kv_seq_len:
+            query_layer = index_first_axis(
+                query_layer.reshape(batch_size * kv_seq_len, self.num_heads, head_dim), indices_k
+            )
+            cu_seqlens_q = cu_seqlens_k
+            max_seqlen_in_batch_q = max_seqlen_in_batch_k
+            indices_q = indices_k
+        elif query_length == 1:
+            max_seqlen_in_batch_q = 1
+            cu_seqlens_q = torch.arange(
+                batch_size + 1, dtype=torch.int32, device=query_layer.device
+            )  # There is a memcpy here, that is very bad.
+            indices_q = cu_seqlens_q[:-1]
+            query_layer = query_layer.squeeze(1)
+        else:
+            # The -q_len: slice assumes left padding.
+            attention_mask = attention_mask[:, -query_length:]
+            query_layer, indices_q, cu_seqlens_q, max_seqlen_in_batch_q = unpad_input(query_layer, attention_mask)
+
+        return (
+            query_layer,
+            key_layer,
+            value_layer,
+            indices_q,
+            (cu_seqlens_q, cu_seqlens_k),
+            (max_seqlen_in_batch_q, max_seqlen_in_batch_k),
+        )
+
+
+# Copied from transformers.models.llama.modeling_llama.LlamaSdpaAttention with Llama->BailingMoeV2
+class BailingMoeV2SdpaAttention(BailingMoeV2Attention):
+    """
+    BailingMoeV2 attention module using torch.nn.functional.scaled_dot_product_attention. This module inherits from
+    `BailingMoeV2Attention` as the weights of the module stays untouched. The only changes are on the forward pass to adapt to
+    SDPA API.
+    """
+
+    # Adapted from BailingMoeV2Attention.forward
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+        position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None,  # necessary, but kept here for BC
+        **kwargs,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+        if output_attentions:
+            # TODO: Improve this warning with e.g. `model.config.attn_implementation = "manual"` once this is implemented.
+            logger.warning_once(
+                "BailingMoeV2Model is using BailingMoeV2SdpaAttention, but `torch.nn.functional.scaled_dot_product_attention` does not support `output_attentions=True`. Falling back to the manual attention implementation, "
+                'but specifying the manual implementation will be required from Transformers version v5.0.0 onwards. This warning can be removed using the argument `attn_implementation="eager"` when loading the model.'
+            )
+            return super().forward(
+                hidden_states=hidden_states,
+                attention_mask=attention_mask,
+                position_ids=position_ids,
+                past_key_value=past_key_value,
+                output_attentions=output_attentions,
+                use_cache=use_cache,
+            )
+
+        bsz, q_len, _ = hidden_states.size()
+
+        qkv = self.query_key_value(hidden_states)
+        qkv = qkv.view(bsz, q_len, self.num_heads + 2 * self.num_key_value_heads, self.head_dim)
+
+        query_states, key_states, value_states = qkv.split(
+            [self.num_heads, self.num_key_value_heads, self.num_key_value_heads], dim=-2
+        )
+        query_states = query_states.transpose(1, 2)
+        key_states = key_states.transpose(1, 2)
+        value_states = value_states.transpose(1, 2)
+
+        if self.config.use_qk_norm:
+            query_states = self.query_layernorm(query_states)
+            key_states = self.key_layernorm(key_states)
+
+        cos, sin = position_embeddings
+        query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin)
+
+        if past_key_value is not None:
+            cache_kwargs = {"sin": sin, "cos": cos}
+            key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs)
+
+        key_states = repeat_kv(key_states, self.num_key_value_groups)
+        value_states = repeat_kv(value_states, self.num_key_value_groups)
+
+        if attention_mask is not None:
+            kv_seq_len = key_states.shape[-2]
+            if attention_mask.size() != (bsz, 1, q_len, kv_seq_len):
+                raise ValueError(
+                    f"Attention mask should be of size {(bsz, 1, q_len, kv_seq_len)}, but is {attention_mask.size()}"
+                )
+
+        # SDPA with memory-efficient backend is currently (torch==2.1.2) bugged with non-contiguous inputs with custom attn_mask,
+        # Reference: https://github.com/pytorch/pytorch/issues/112577.
+        if query_states.device.type == "cuda" and attention_mask is not None:
+            query_states = query_states.contiguous()
+            key_states = key_states.contiguous()
+            value_states = value_states.contiguous()
+
+        attn_output = torch.nn.functional.scaled_dot_product_attention(
+            query_states,
+            key_states,
+            value_states,
+            attn_mask=attention_mask,
+            dropout_p=self.attention_dropout if self.training else 0.0,
+            # The q_len > 1 is necessary to match with AttentionMaskConverter.to_causal_4d that does not create a causal mask in case q_len == 1.
+            is_causal=self.is_causal and attention_mask is None and q_len > 1,
+        )
+
+        attn_output = attn_output.transpose(1, 2).contiguous()
+        attn_output = attn_output.reshape(bsz, q_len, -1)
+
+        attn_output = self.dense(attn_output)
+
+        return attn_output, None, past_key_value
+
+
+ATTENTION_CLASSES = {
+    "eager": BailingMoeV2Attention,
+    "flash_attention_2": BailingMoeV2FlashAttention2,
+    "sdpa": BailingMoeV2SdpaAttention,
+}
+
+
+class BailingMoeV2MTPLayer(nn.Module):
+    def __init__(self, config: BailingMoeV2Config, layer_idx: int):
+        super().__init__()
+        self.layer_idx = layer_idx
+        self.input_layernorm = BailingMoeV2RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
+        self.enorm = BailingMoeV2RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
+
+        self.eh_proj = nn.Linear(config.hidden_size * 2, config.hidden_size, bias=False)
+        self.post_attention_layernorm = BailingMoeV2RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
+        self.attention = ATTENTION_CLASSES[config._attn_implementation](config=config, layer_idx=layer_idx)
+        self.mlp = BailingMoeV2SparseMoeBlock(config)
+
+        self.hnorm = BailingMoeV2RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
+        self.final_layernorm = BailingMoeV2RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
+
+    def forward(
+        self,
+        input_embeds,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Tuple[torch.Tensor]] = None,
+        output_attentions: Optional[bool] = False,
+        output_router_logits: Optional[bool] = False,
+        use_cache: Optional[bool] = False,
+        position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None,  # necessary, but kept here for BC
+        **kwargs,
+    ) -> Tuple[torch.FloatTensor, Optional[Tuple[torch.FloatTensor, torch.FloatTensor]]]:
+        input_embeds = self.enorm(input_embeds)
+        hidden_states = self.hnorm(hidden_states)
+        hidden_states = self.eh_proj(torch.cat([input_embeds, hidden_states], dim=-1))
+        residual = hidden_states
+
+        hidden_states = self.input_layernorm(hidden_states)
+
+        # Self Attention
+        hidden_states, self_attn_weights, present_key_value = self.attention(
+            hidden_states=hidden_states,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_value=past_key_value,
+            output_attentions=output_attentions,
+            position_embeddings=position_embeddings,
+            use_cache=use_cache,
+        )
+        hidden_states = residual + hidden_states
+
+        # Fully Connected
+        residual = hidden_states
+        hidden_states = self.post_attention_layernorm(hidden_states)
+        hidden_states = self.mlp(hidden_states)
+        if isinstance(hidden_states, tuple):
+            hidden_states, router_logits = hidden_states
+        else:
+            router_logits = None
+        hidden_states = residual + hidden_states.to(residual.device)
+        hidden_states = self.final_layernorm(hidden_states)
+
+        outputs = (hidden_states,)
+
+        if output_attentions:
+            outputs += (self_attn_weights,)
+
+        if use_cache:
+            outputs += (present_key_value,)
+
+        if output_router_logits:
+            outputs += (router_logits,)
+
+        return outputs
+
+
+class BailingMoeV2DecoderLayer(nn.Module):
+    def __init__(self, config: BailingMoeV2Config, layer_idx: int):
+        super().__init__()
+        self.hidden_size = config.hidden_size
+
+        self.attention = ATTENTION_CLASSES[config._attn_implementation](config=config, layer_idx=layer_idx)
+
+        self.mlp = (
+            BailingMoeV2SparseMoeBlock(config)
+            if (config.num_experts is not None and layer_idx >= config.first_k_dense_replace)
+            else BailingMoeV2MLP(config=config, intermediate_size=config.intermediate_size)
+        )
+        self.input_layernorm = BailingMoeV2RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
+        self.post_attention_layernorm = BailingMoeV2RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Tuple[torch.Tensor]] = None,
+        output_attentions: Optional[bool] = False,
+        output_router_logits: Optional[bool] = False,
+        use_cache: Optional[bool] = False,
+        position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None,  # necessary, but kept here for BC
+        **kwargs,
+    ) -> Tuple[torch.FloatTensor, Optional[Tuple[torch.FloatTensor, torch.FloatTensor]]]:
+        """
+        Args:
+            hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)`
+            attention_mask (`torch.FloatTensor`, *optional*):
+                attention mask of size `(batch_size, sequence_length)` if flash attention is used or `(batch_size, 1,
+                query_sequence_length, key_sequence_length)` if default attention is used.
+            position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+                Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
+                config.n_positions - 1]`.
+            past_key_value (`Tuple(torch.FloatTensor)`, *optional*):
+                cached past key and value projection states
+            output_attentions (`bool`, *optional*):
+                Whether to return the attentions tensors of all attention layers. See `attentions` under
+                returned tensors for more detail.
+            output_router_logits (`bool`, *optional*):
+                Whether or not to return the logits of all the routers. They are useful for computing the router loss,
+                and should not be returned during inference.
+            use_cache (`bool`, *optional*):
+                If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding
+                (see `past_key_values`).
+        """
+        residual = hidden_states
+
+        hidden_states = self.input_layernorm(hidden_states)
+
+        # Self Attention
+        hidden_states, self_attn_weights, present_key_value = self.attention(
+            hidden_states=hidden_states,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_value=past_key_value,
+            output_attentions=output_attentions,
+            position_embeddings=position_embeddings,
+            use_cache=use_cache,
+        )
+        hidden_states = residual + hidden_states
+
+        # Fully Connected
+        residual = hidden_states
+        hidden_states = self.post_attention_layernorm(hidden_states)
+        hidden_states = self.mlp(hidden_states)
+        if isinstance(hidden_states, tuple):
+            hidden_states, router_logits = hidden_states
+        else:
+            router_logits = None
+        hidden_states = residual + hidden_states.to(residual.device)
+
+        outputs = (hidden_states,)
+
+        if output_attentions:
+            outputs += (self_attn_weights,)
+
+        if use_cache:
+            outputs += (present_key_value,)
+
+        if output_router_logits:
+            outputs += (router_logits,)
+
+        return outputs
+
+
+BAILINGMOEV2_START_DOCSTRING = r"""
+    This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the
+    library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
+    etc.)
+
+    This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
+    Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
+    and behavior.
+
+    Parameters:
+        config ([`BailingMoeV2Config`]):
+            Model configuration class with all the parameters of the model. Initializing with a config file does not
+            load the weights associated with the model, only the configuration. Check out the
+            [`~PreTrainedModel.from_pretrained`] method to load the model weights.
+"""
+
+
+@add_start_docstrings(
+    "The bare BailingMoeV2 Model outputting raw hidden-states without any specific head on top.",
+    BAILINGMOEV2_START_DOCSTRING,
+)
+class BailingMoeV2PreTrainedModel(PreTrainedModel):
+    config_class = BailingMoeV2Config
+    base_model_prefix = "model"
+    supports_gradient_checkpointing = True
+    _no_split_modules = ["BailingMoeV2DecoderLayer"]
+    _skip_keys_device_placement = "past_key_values"
+    _supports_flash_attn_2 = True
+    _supports_sdpa = True
+    _supports_cache_class = True
+
+    def _init_weights(self, module):
+        std = self.config.initializer_range
+        if isinstance(module, nn.Linear):
+            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_()
+
+
+BAILINGMOEV2_INPUTS_DOCSTRING = r"""
+    Args:
+        input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`):
+            Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you provide
+            it.
+
+            Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
+            [`PreTrainedTokenizer.__call__`] for details.
+
+            [What are input IDs?](../glossary#input-ids)
+        attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:
+
+            - 1 for tokens that are **not masked**,
+            - 0 for tokens that are **masked**.
+
+            [What are attention masks?](../glossary#attention-mask)
+
+            Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
+            [`PreTrainedTokenizer.__call__`] for details.
+
+            If `past_key_values` is used, optionally only the last `input_ids` have to be input (see
+            `past_key_values`).
+
+            If you want to change padding behavior, you should read [`modeling_opt._prepare_decoder_attention_mask`]
+            and modify to your needs. See diagram 1 in [the paper](https://arxiv.org/abs/1910.13461) for more
+            information on the default strategy.
+
+            - 1 indicates the head is **not masked**,
+            - 0 indicates the head is **masked**.
+        position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
+            config.n_positions - 1]`.
+
+            [What are position IDs?](../glossary#position-ids)
+        past_key_values (`Cache` or `tuple(tuple(torch.FloatTensor))`, *optional*):
+            Pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention
+            blocks) that can be used to speed up sequential decoding. This typically consists in the `past_key_values`
+            returned by the model at a previous stage of decoding, when `use_cache=True` or `config.use_cache=True`.
+
+            Two formats are allowed:
+            - a [`~cache_utils.Cache`] instance;
+            - 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)`). This is also known as the legacy
+            cache format.
+
+            The model will output the same cache format that is fed as input. If no `past_key_values` are passed, the
+            legacy cache format will be returned.
+
+            If `past_key_values` are used, the user can optionally input only the last `input_ids` (those that don't
+            have their past key value states given to this model) of shape `(batch_size, 1)` instead of all `input_ids`
+            of shape `(batch_size, sequence_length)`.
+        inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
+            Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This
+            is useful if you want more control over how to convert `input_ids` indices into associated vectors than the
+            model's internal embedding lookup matrix.
+        use_cache (`bool`, *optional*):
+            If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
+            `past_key_values`).
+        output_attentions (`bool`, *optional*):
+            Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned
+            tensors for more detail.
+        output_hidden_states (`bool`, *optional*):
+            Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
+            more detail.
+        return_dict (`bool`, *optional*):
+            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
+"""
+
+
+@add_start_docstrings(
+    "The bare BailingMoeV2 Model outputting raw hidden-states without any specific head on top.",
+    BAILINGMOEV2_START_DOCSTRING,
+)
+class BailingMoeV2Model(BailingMoeV2PreTrainedModel):
+    """
+    Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`BailingMoeV2DecoderLayer`]
+
+    Args:
+        config: BailingMoeV2Config
+    """
+
+    def __init__(self, config: BailingMoeV2Config):
+        super().__init__(config)
+        self.padding_idx = config.pad_token_id
+        self.vocab_size = config.vocab_size
+        self.num_nextn_predict_layers = config.num_nextn_predict_layers
+
+        self.word_embeddings = nn.Embedding(config.vocab_size, config.hidden_size, self.padding_idx)
+        self.layers = []
+        for layer_idx in range(config.num_hidden_layers + config.num_nextn_predict_layers):
+            layer_cls = BailingMoeV2DecoderLayer if layer_idx < config.num_hidden_layers else BailingMoeV2MTPLayer
+            self.layers.append(layer_cls(config, layer_idx))
+
+        self.layers = nn.ModuleList(self.layers)
+
+        self._use_sdpa = config._attn_implementation == "sdpa"
+        self._use_flash_attention_2 = config._attn_implementation == "flash_attention_2"
+        self.norm = BailingMoeV2RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
+        self.rotary_emb = BailingMoeV2RotaryEmbedding(config=config)
+        self.gradient_checkpointing = False
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def get_input_embeddings(self):
+        return self.word_embeddings
+
+    def set_input_embeddings(self, value):
+        self.word_embeddings = value
+
+    @add_start_docstrings_to_model_forward(BAILINGMOEV2_INPUTS_DOCSTRING)
+    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,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        output_router_logits: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        **kwargs,
+    ) -> Union[Tuple, MoeV2ModelOutputWithPast]:
+        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
+        )
+        output_router_logits = (
+            output_router_logits if output_router_logits is not None else self.config.output_router_logits
+        )
+        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
+
+        # retrieve input_ids and inputs_embeds
+        if input_ids is not None and inputs_embeds is not None:
+            raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
+        elif input_ids is not None:
+            batch_size, seq_length = input_ids.shape[:2]
+        elif inputs_embeds is not None:
+            batch_size, seq_length = inputs_embeds.shape[:2]
+        else:
+            raise ValueError("You have to specify either input_ids or inputs_embeds")
+
+        if self.gradient_checkpointing and self.training:
+            if use_cache:
+                logger.warning_once(
+                    "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`transformers."
+                )
+                use_cache = False
+
+        if use_cache and past_key_values is None:
+            past_key_values = DynamicCache()
+
+        if inputs_embeds is None:
+            inputs_embeds = self.word_embeddings(input_ids)
+
+        past_seen_tokens = past_key_values.get_seq_length() if past_key_values is not None else 0
+
+        if position_ids is None:
+            position_ids = torch.arange(
+                past_seen_tokens, past_seen_tokens + inputs_embeds.shape[1], device=inputs_embeds.device
+            )
+            position_ids = position_ids.unsqueeze(0)
+
+        if self._use_flash_attention_2:
+            # 2d mask is passed through the layers
+            attention_mask = attention_mask if (attention_mask is not None and 0 in attention_mask) else None
+        elif self._use_sdpa and not output_attentions:
+            # output_attentions=True can not be supported when using SDPA, and we fall back on
+            # the manual implementation that requires a 4D causal mask in all cases.
+            attention_mask = _prepare_4d_causal_attention_mask_for_sdpa(
+                attention_mask,
+                (batch_size, seq_length),
+                inputs_embeds,
+                past_seen_tokens,
+            )
+        else:
+            # 4d mask is passed through the layers
+            attention_mask = _prepare_4d_causal_attention_mask(
+                attention_mask, (batch_size, seq_length), inputs_embeds, past_seen_tokens
+            )
+
+        # embed positions
+        hidden_states = inputs_embeds
+
+        # create position embeddings to be shared across the decoder layers
+        position_embeddings = self.rotary_emb(hidden_states, position_ids)
+
+        # decoder layers
+        all_hidden_states = () if output_hidden_states else None
+        all_self_attns = () if output_attentions else None
+        all_router_logits = () if output_router_logits else None
+        next_decoder_cache = None
+        layers = self.layers[: -self.num_nextn_predict_layers] if self.num_nextn_predict_layers > 0 else self.layers
+        mtp_layers = self.layers[-self.num_nextn_predict_layers :] if self.num_nextn_predict_layers > 0 else None
+
+        for decoder_layer in layers:
+            if output_hidden_states:
+                all_hidden_states += (hidden_states,)
+
+            if self.gradient_checkpointing and self.training:
+                layer_outputs = self._gradient_checkpointing_func(
+                    decoder_layer.__call__,
+                    hidden_states,
+                    attention_mask,
+                    position_ids,
+                    past_key_values,
+                    output_attentions,
+                    output_router_logits,
+                    use_cache,
+                    position_embeddings,
+                )
+            else:
+                layer_outputs = decoder_layer(
+                    hidden_states,
+                    attention_mask=attention_mask,
+                    position_ids=position_ids,
+                    past_key_value=past_key_values,
+                    output_attentions=output_attentions,
+                    output_router_logits=output_router_logits,
+                    use_cache=use_cache,
+                    position_embeddings=position_embeddings,
+                )
+            hidden_states = layer_outputs[0]
+
+            if use_cache:
+                next_decoder_cache = layer_outputs[2 if output_attentions else 1]
+
+            if output_attentions:
+                all_self_attns += (layer_outputs[1],)
+
+            if output_router_logits and layer_outputs[-1] is not None:
+                all_router_logits += (layer_outputs[-1],)
+
+        hidden_states = self.norm(hidden_states)
+        main_hidden_states = hidden_states
+
+        # add hidden states from the last decoder layer
+        if output_hidden_states:
+            all_hidden_states += (main_hidden_states,)
+
+        mtp_hidden_states = None
+
+        if mtp_layers:
+            for decoder_layer in mtp_layers:
+                input_ids, _ = roll_tensor(input_ids, shifts=-1, dims=-1)
+                inputs_embeds = self.word_embeddings(input_ids)
+
+                if self.gradient_checkpointing and self.training:
+                    layer_outputs = self._gradient_checkpointing_func(
+                        decoder_layer.__call__,
+                        inputs_embeds,
+                        hidden_states,
+                        attention_mask,
+                        position_ids,
+                        past_key_values,
+                        output_attentions,
+                        output_router_logits,
+                        use_cache,
+                        position_embeddings,
+                    )
+                else:
+                    layer_outputs = decoder_layer(
+                        inputs_embeds,
+                        hidden_states,
+                        attention_mask=attention_mask,
+                        position_ids=position_ids,
+                        past_key_value=past_key_values,
+                        output_attentions=output_attentions,
+                        output_router_logits=output_router_logits,
+                        use_cache=use_cache,
+                        position_embeddings=position_embeddings,
+                    )
+                if mtp_hidden_states is None:
+                    mtp_hidden_states = []
+                hidden_states = layer_outputs[0]
+                mtp_hidden_states.append(hidden_states)
+
+                if output_hidden_states:
+                    all_hidden_states += (hidden_states,)
+
+                if use_cache:
+                    next_decoder_cache = layer_outputs[2 if output_attentions else 1]
+
+                if output_attentions:
+                    all_self_attns += (layer_outputs[1],)
+
+                if output_router_logits and layer_outputs[-1] is not None:
+                    all_router_logits += (layer_outputs[-1],)
+
+        next_cache = None
+        if use_cache:
+            next_cache = next_decoder_cache
+        if not return_dict:
+            return tuple(
+                v
+                for v in [main_hidden_states, next_cache, all_hidden_states, all_self_attns, all_router_logits]
+                if v is not None
+            )
+        return MoeV2ModelOutputWithPast(
+            last_hidden_state=main_hidden_states,
+            past_key_values=next_cache,
+            hidden_states=all_hidden_states,
+            mtp_hidden_states=mtp_hidden_states,
+            attentions=all_self_attns,
+            router_logits=all_router_logits,
+        )
+
+
+class BailingMoeV2ForCausalLM(BailingMoeV2PreTrainedModel, GenerationMixin):
+    _tied_weights_keys = ["lm_head.weight"]
+
+    def __init__(self, config: BailingMoeV2Config):
+        super().__init__(config)
+        self.model = BailingMoeV2Model(config)
+        self.vocab_size = config.vocab_size
+        self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
+        self.num_nextn_predict_layers = config.num_nextn_predict_layers
+        self.mtp_loss_scaling_factor = config.mtp_loss_scaling_factor
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def get_input_embeddings(self):
+        return self.model.word_embeddings
+
+    def set_input_embeddings(self, value):
+        self.model.word_embeddings = value
+
+    def get_output_embeddings(self):
+        return self.lm_head
+
+    def set_output_embeddings(self, new_embeddings):
+        self.lm_head = new_embeddings
+
+    def set_decoder(self, decoder):
+        self.model = decoder
+
+    def get_decoder(self):
+        return self.model
+
+    @add_start_docstrings_to_model_forward(BAILINGMOEV2_INPUTS_DOCSTRING)
+    @replace_return_docstrings(output_type=MoEV2CausalLMOutputWithPast, config_class=_CONFIG_FOR_DOC)
+    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,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        output_router_logits: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        **kwargs,
+    ) -> Union[Tuple, MoEV2CausalLMOutputWithPast]:
+        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 -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
+                (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.
+
+        Returns:
+
+        Example:
+
+        ```python
+        >>> from transformers import AutoTokenizer
+
+        >>> model = BailingMoeV2ForCausalLM.from_pretrained(PATH_TO_CONVERTED_WEIGHTS)
+        >>> tokenizer = AutoTokenizer.from_pretrained(PATH_TO_CONVERTED_TOKENIZER)
+
+        >>> prompt = "Hey, are you conscious? Can you talk to me?"
+        >>> inputs = tokenizer(prompt, return_tensors="pt")
+
+        >>> # Generate
+        >>> generate_ids = model.generate(inputs.input_ids, max_length=30)
+        >>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
+        "Hey, are you conscious? Can you talk to me?\nI'm not conscious, but I can talk to you."
+        ```"""
+        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
+        )
+        output_router_logits = (
+            output_router_logits if output_router_logits is not None else self.config.output_router_logits
+        )
+        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,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            output_router_logits=output_router_logits,
+            return_dict=return_dict,
+            **kwargs,
+        )
+
+        loss = None
+        all_mtp_loss = None
+        aux_loss = None
+        hidden_states = outputs[0]
+        logits = self.lm_head(hidden_states)
+        logits = logits.float()
+
+        if labels is not None:
+            loss = self.loss_function(logits, labels, self.config.vocab_size, **kwargs)
+
+        all_mtp_logits = None
+        if self.num_nextn_predict_layers > 0:
+            mtp_hidden_states = outputs.mtp_hidden_states
+            shift_labels_mtp = None
+            for i in range(self.num_nextn_predict_layers):
+                mtp_hidden_states = mtp_hidden_states[i]
+                mtp_logits = self.lm_head(mtp_hidden_states).float()
+                if all_mtp_logits is None:
+                    all_mtp_logits = []
+                all_mtp_logits.append(mtp_logits)
+                if labels is not None:
+                    if shift_labels_mtp is None:
+                        shift_labels_mtp = labels.clone()
+                    shift_labels_mtp, _ = roll_tensor(shift_labels_mtp, shifts=-1, dims=-1, fill_value=-100)
+                    mtp_logits_ = mtp_logits.view(-1, self.config.vocab_size)
+                    mtp_loss = self.loss_function(mtp_logits_, shift_labels_mtp.to(mtp_logits_.device).view(-1), self.config.vocab_size, **kwargs)
+                    if loss is not None:
+                        loss += self.mtp_loss_scaling_factor * mtp_loss
+                    else:
+                        loss = self.mtp_loss_scaling_factor * mtp_loss
+
+                    if all_mtp_loss is None:
+                        all_mtp_loss = []
+                    all_mtp_loss.append(mtp_loss)
+
+        if not return_dict:
+            output = (logits,) + outputs[1:]
+            if output_router_logits:
+                output = (aux_loss,) + output
+            return (loss,) + output if loss is not None else output
+
+        return MoEV2CausalLMOutputWithPast(
+            loss=loss,
+            mtp_loss=all_mtp_loss,
+            aux_loss=aux_loss,
+            logits=logits,
+            mtp_logits=all_mtp_logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            router_logits=outputs.router_logits,
+        )

pruning-code/prune_ring_moe.py

@@ -0,0 +1,549 @@
+#!/usr/bin/env python3
+"""
+Calibration-based structured pruning for Ring-mini-2.0 (BailingMoeV2 MoE).
+
+This script prunes Mixture-of-Experts (MoE) experts per layer based on routing
+statistics gathered from a small calibration dataset, and writes out a new
+checkpoint with fewer experts (e.g., 256 -> 144 per layer) to hit a target
+total parameter budget (e.g., ~10B total params instead of ~16.8B).
+
+Key features:
+- Collects per-layer expert importance via router outputs (prob mass or usage).
+- Keeps the same number of experts per group to preserve gating layout.
+- Rebuilds each MoE block with the selected experts and remaps gate weights.
+- Saves a standalone pruned model with updated config.
+
+Usage example (experts -> 144 per layer):
+
+  python prune_ring_moe.py \
+    --model . \
+    --output_dir ./ring-mini-2.0-pruned-144e \
+    --calib_path ./calib_data \
+    --samples 200 --seq_len 512 \
+    --target_experts 144 \
+    --score route_count
+
+Alternatively, specify a target total parameter budget (approx.):
+
+  python prune_ring_moe.py \
+    --model . \
+    --output_dir ./ring-mini-2.0-pruned-10B \
+    --calib_path ./calib_data \
+    --samples 200 --seq_len 512 \
+    --target_total_params 10000000000
+
+Notes:
+- The script expects Ring-mini-2.0 code in the working directory (trust_remote_code=True).
+- Calibration path can be a folder of .txt files or a .jsonl file with {"text": ...} lines.
+- Ensure enough VRAM/system RAM to load the model and run short eval passes.
+"""
+
+from __future__ import annotations
+
+import argparse
+import copy
+import json
+import math
+import os
+import random
+import sys
+from tqdm import tqdm
+from dataclasses import dataclass
+from typing import Dict, Iterable, List, Optional, Sequence, Tuple
+
+import torch
+from torch import nn
+
+try:
+    from transformers import AutoModelForCausalLM, AutoTokenizer
+except Exception as e:  # pragma: no cover - CLI convenience
+    print("Please install transformers: pip install transformers", file=sys.stderr)
+    raise
+
+
+# ----------------------------
+# Calibration dataset
+# ----------------------------
+
+def iter_texts(calib_path: str) -> Iterable[str]:
+    """Yield texts from a directory of .txt files or a .jsonl file with {"text": ...} lines."""
+    if os.path.isdir(calib_path):
+        for root, _, files in os.walk(calib_path):
+            for fn in files:
+                if fn.lower().endswith(".txt"):
+                    p = os.path.join(root, fn)
+                    try:
+                        with open(p, "r", encoding="utf-8", errors="ignore") as f:
+                            txt = f.read().strip()
+                            if txt:
+                                yield txt
+                    except Exception:
+                        continue
+    elif os.path.isfile(calib_path) and calib_path.lower().endswith(".jsonl"):
+        with open(calib_path, "r", encoding="utf-8", errors="ignore") as f:
+            for line in f:
+                line = line.strip()
+                if not line:
+                    continue
+                try:
+                    obj = json.loads(line)
+                    txt = obj.get("text", "").strip()
+                    if txt:
+                        yield txt
+                except Exception:
+                    continue
+    else:
+        raise ValueError(f"Unsupported calib_path: {calib_path}")
+
+
+@dataclass
+class PruneArgs:
+    model: str
+    output_dir: str
+    calib_path: Optional[str]
+    tokenizer: Optional[str]
+    device: str
+    dtype: str
+    device_map: Optional[str]
+    seed: int
+    samples: int
+    seq_len: int
+    batch_size: int
+    score: str  # 'route_count' | 'prob_mass' | 'combined'
+    alpha: float  # weight for route_count in combined
+    target_experts: Optional[int]
+    target_total_params: Optional[int]
+    dry_run: bool
+
+
+def set_seed(seed: int) -> None:
+    random.seed(seed)
+    torch.manual_seed(seed)
+    torch.cuda.manual_seed_all(seed)
+
+
+def infer_dtype(dtype_str: str) -> torch.dtype:
+    if dtype_str.lower() in ("auto",):
+        return torch.float16 if torch.cuda.is_available() else torch.float32
+    m = dtype_str.lower()
+    if m in ("fp16", "float16", "half"): return torch.float16
+    if m in ("bf16", "bfloat16"): return torch.bfloat16
+    if m in ("fp32", "float32", "full"): return torch.float32
+    raise ValueError(f"Unsupported dtype: {dtype_str}")
+
+
+# ----------------------------
+# Param count estimation
+# ----------------------------
+
+def estimate_total_params_for_config(
+    *,
+    vocab_size: int,
+    hidden_size: int,
+    intermediate_size: int,
+    num_hidden_layers: int,
+    num_attention_heads: int,
+    num_key_value_heads: int,
+    head_dim: int,
+    use_qkv_bias: bool,
+    use_bias: bool,
+    n_group: int,
+    num_experts: int,
+    moe_intermediate_size: int,
+    num_shared_experts: int,
+    first_k_dense_replace: int,
+) -> int:
+    """Compute an approximate parameter count from config, without instantiating the model.
+
+    The calculation mirrors the shapes in modeling_bailing_moe_v2.py and is sufficient
+    to solve for a target experts-per-layer given a total param budget.
+    """
+    L = num_hidden_layers
+    HS = hidden_size
+    NH = num_attention_heads
+    NKV = num_key_value_heads
+    Dh = head_dim
+    V = vocab_size
+
+    # Embedding + LM head + final RMS norm
+    emb = V * HS
+    lm_head = HS * V
+    final_norm = HS
+
+    # Per-layer attention params
+    qkv_out = (NH + 2 * NKV) * Dh
+    qkv = HS * qkv_out + (qkv_out if use_qkv_bias else 0)
+    o_proj = (NH * Dh) * HS + (HS if use_bias else 0)
+    # QK norms (weights only)
+    qk_norm = 2 * Dh
+    # Pre/post RMS norms (weights only)
+    layer_norms = 2 * HS
+    attn_block = qkv + o_proj + qk_norm + layer_norms
+
+    # Dense MLP (first_k_dense_replace layers)
+    dense_mlp_per_layer = 3 * HS * intermediate_size
+
+    # MoE per layer: gate + experts + shared experts
+    gate = num_experts * HS  # expert_bias buffer omitted
+    expert_mlp_per_expert = 3 * HS * moe_intermediate_size
+    shared_experts = 3 * HS * moe_intermediate_size * max(1, num_shared_experts)
+    moe_block = gate + num_experts * expert_mlp_per_expert + shared_experts
+
+    total = 0
+    total += emb + lm_head + final_norm
+    total += L * attn_block
+    total += first_k_dense_replace * dense_mlp_per_layer
+    total += (L - first_k_dense_replace) * moe_block
+    return int(total)
+
+
+def solve_target_experts_for_budget(cfg, target_total_params: int) -> int:
+    """Given a config and a target total params, solve for experts per layer.
+
+    Returns a value divisible by cfg.n_group and >= cfg.num_experts_per_tok.
+    """
+    def total_for_experts(E: int) -> int:
+        return estimate_total_params_for_config(
+            vocab_size=cfg.vocab_size,
+            hidden_size=cfg.hidden_size,
+            intermediate_size=cfg.intermediate_size,
+            num_hidden_layers=cfg.num_hidden_layers,
+            num_attention_heads=cfg.num_attention_heads,
+            num_key_value_heads=cfg.num_key_value_heads,
+            head_dim=cfg.head_dim,
+            use_qkv_bias=cfg.use_qkv_bias,
+            use_bias=cfg.use_bias,
+            n_group=cfg.n_group,
+            num_experts=E,
+            moe_intermediate_size=cfg.moe_intermediate_size,
+            num_shared_experts=cfg.num_shared_experts or 0,
+            first_k_dense_replace=cfg.first_k_dense_replace,
+        )
+
+    # Binary search between [num_experts_per_tok, current_num_experts]
+    lo = max(cfg.num_experts_per_tok, cfg.n_group)  # at least one per group
+    hi = cfg.num_experts
+    best = hi
+    while lo <= hi:
+        mid = (lo + hi) // 2
+        # enforce divisibility by n_group
+        mid = (mid // cfg.n_group) * cfg.n_group
+        mid = max(mid, cfg.n_group)
+        mid = max(mid, cfg.num_experts_per_tok)
+        if mid == 0:
+            mid = cfg.n_group
+        t = total_for_experts(mid)
+        if t > target_total_params:
+            hi = mid - cfg.n_group
+        else:
+            best = mid
+            lo = mid + cfg.n_group
+    return max(cfg.num_experts_per_tok, best)
+
+
+# ----------------------------
+# Scoring and selection
+# ----------------------------
+
+@torch.no_grad()
+def collect_moe_stats(
+    model, tokenizer, texts: Iterable[str], *,
+    max_samples: int, seq_len: int, batch_size: int, device: str
+) -> Tuple[List[torch.Tensor], List[torch.Tensor]]:
+    """Accumulate per-layer stats for MoE experts.
+
+    Returns (prob_mass_list, route_count_list), each a list of tensors of
+    shape [num_experts] for each MoE layer.
+    """
+    model.eval()
+    prob_mass: List[torch.Tensor] = []
+    route_cnt: List[torch.Tensor] = []
+
+    # Determine MoE layers using config
+    cfg = model.config
+    moe_layer_indices = list(range(cfg.first_k_dense_replace, cfg.num_hidden_layers))
+    if not moe_layer_indices:
+        return prob_mass, route_cnt
+
+    # Initialize accumulators lazily after first forward (to know num_experts)
+    initialized = False
+
+    batch_texts: List[str] = []
+    n_seen = 0
+    for txt in tqdm(texts, desc="Calculating MoE Expert Weights"):
+        if n_seen >= max_samples:
+            break
+        if not txt:
+            continue
+        batch_texts.append(txt)
+        if len(batch_texts) < batch_size and (n_seen + len(batch_texts)) < max_samples:
+            continue
+
+        enc = tokenizer(
+            batch_texts,
+            padding=True,
+            truncation=True,
+            max_length=seq_len,
+            return_tensors="pt",
+            return_token_type_ids=False,
+        )
+        enc = {k: v.to(device) for k, v in enc.items()}
+        out = model(**enc, use_cache=False, output_router_logits=True, return_dict=True)
+
+        # out.router_logits is a tuple per layer: (router_logits, topk_idx)
+        if not initialized:
+            # Create accumulators per MoE layer
+            for (r_logits, topk_idx) in out.router_logits:
+                num_experts_layer = r_logits.shape[-1]
+                prob_mass.append(torch.zeros(num_experts_layer, dtype=torch.float64))
+                route_cnt.append(torch.zeros(num_experts_layer, dtype=torch.float64))
+            initialized = True
+
+        for layer_i, (r_logits, topk_idx) in enumerate(out.router_logits):
+            probs = torch.sigmoid(r_logits).sum(dim=(0, 1)).double().cpu()  # sum over batch, seq
+            prob_mass[layer_i] += probs
+            # Count selections via bincount on flattened topk indices
+            flat = topk_idx.reshape(-1).to(torch.int64).cpu()
+            bc = torch.bincount(flat, minlength=prob_mass[layer_i].numel()).double()
+            route_cnt[layer_i] += bc
+
+        n_seen += len(batch_texts)
+        batch_texts.clear()
+
+    if not initialized:
+        raise RuntimeError("No calibration samples were processed; check calib_path.")
+
+    return prob_mass, route_cnt
+
+
+def per_group_topk_indices(
+    scores: torch.Tensor,
+    *,
+    n_group: int,
+    keep_per_group: int,
+) -> List[int]:
+    """Select top experts per group based on scores.
+
+    - scores: tensor [E]
+    - n_group: number of groups
+    - keep_per_group: how many experts to keep in each group
+
+    Returns list of kept expert indices in new order grouped contiguously by group.
+    """
+    E = scores.numel()
+    if E % n_group != 0:
+        raise ValueError(f"num_experts={E} must be divisible by n_group={n_group}")
+    per_group = E // n_group
+    keep: List[int] = []
+    for g in range(n_group):
+        start = g * per_group
+        end = start + per_group
+        sl = scores[start:end]
+        vals, idx = torch.topk(sl, k=keep_per_group, dim=0, largest=True, sorted=True)
+        keep.extend((start + idx).tolist())
+    return keep
+
+
+def build_pruned_mlp_block(
+    old_block: nn.Module,
+    *,
+    new_cfg,
+    keep_indices: Sequence[int],
+) -> nn.Module:
+    """Create a new BailingMoeV2SparseMoeBlock with selected experts and remapped gate.
+
+    Assumes old_block is an instance of BailingMoeV2SparseMoeBlock.
+    """
+    # Lazy import to avoid import-time side effects when running help
+    from modeling_bailing_moe_v2 import BailingMoeV2SparseMoeBlock
+
+    new_block = BailingMoeV2SparseMoeBlock(new_cfg)
+
+    # Copy shared experts (if present)
+    if hasattr(old_block, "shared_experts") and hasattr(new_block, "shared_experts"):
+        for name in ("gate_proj", "up_proj", "down_proj"):
+            getattr(new_block.shared_experts, name).weight.data.copy_(
+                getattr(old_block.shared_experts, name).weight.data
+            )
+
+    # Copy selected experts in the new order
+    for new_idx, old_idx in enumerate(keep_indices):
+        old_exp = old_block.experts[old_idx]
+        new_exp = new_block.experts[new_idx]
+        for name in ("gate_proj", "up_proj", "down_proj"):
+            getattr(new_exp, name).weight.data.copy_(getattr(old_exp, name).weight.data)
+
+    # Remap gating weights and biases (if any)
+    new_block.gate.weight.data.copy_(old_block.gate.weight.data[keep_indices, :])
+    if hasattr(old_block.gate, "expert_bias") and hasattr(new_block.gate, "expert_bias"):
+        new_block.gate.expert_bias.data.copy_(old_block.gate.expert_bias.data[keep_indices])
+
+    return new_block
+
+
+def prune_model_experts(
+    model, *,
+    keep_per_group: int,
+    scores_prob_mass: List[torch.Tensor],
+    scores_route_cnt: List[torch.Tensor],
+    score_mode: str,
+    alpha: float,
+) -> Tuple[object, Dict[str, List[int]]]:
+    """Prune experts per MoE layer in-place and return selection mapping.
+
+    - keep_per_group: number of experts kept per group in every MoE layer.
+    - score_mode: 'route_count' | 'prob_mass' | 'combined'
+    - alpha: weight for route_count in 'combined'
+    """
+    cfg = model.config
+    if keep_per_group * cfg.n_group < cfg.num_experts_per_tok:
+        raise ValueError("Keeping fewer experts than `num_experts_per_tok` is invalid.")
+
+    # Prepare a config clone with new num_experts
+    new_cfg = copy.deepcopy(cfg)
+    new_cfg.num_experts = keep_per_group * cfg.n_group
+
+    selection: Dict[str, List[int]] = {}
+
+    # Iterate MoE layers
+    moe_layers = list(range(cfg.first_k_dense_replace, cfg.num_hidden_layers))
+    for local_i, layer_idx in tqdm(enumerate(moe_layers), desc="Rebuilding MoE Layers"):
+        layer = model.model.layers[layer_idx]
+        old_block = layer.mlp
+        # Compute score
+        if score_mode == "route_count":
+            score_vec = scores_route_cnt[local_i]
+        elif score_mode == "prob_mass":
+            score_vec = scores_prob_mass[local_i]
+        elif score_mode == "combined":
+            # normalize then combine
+            rc = scores_route_cnt[local_i]
+            pm = scores_prob_mass[local_i]
+            rc = rc / (rc.sum() + 1e-9)
+            pm = pm / (pm.sum() + 1e-9)
+            score_vec = alpha * rc + (1 - alpha) * pm
+        else:
+            raise ValueError(f"Unknown score mode: {score_mode}")
+
+        keep_idx = per_group_topk_indices(
+            score_vec,
+            n_group=cfg.n_group,
+            keep_per_group=keep_per_group,
+        )
+        selection[f"layer_{layer_idx}"] = keep_idx
+
+        # Rebuild this MoE block and swap
+        new_block = build_pruned_mlp_block(old_block, new_cfg=new_cfg, keep_indices=keep_idx)
+        layer.mlp = new_block
+
+    # Update top-level config
+    model.config.num_experts = new_cfg.num_experts
+    return model, selection
+
+
+def main(argv: Optional[Sequence[str]] = None) -> None:
+    p = argparse.ArgumentParser(description="Prune Ring-mini-2.0 MoE experts using a calibration set")
+    p.add_argument("--model", type=str, required=True, help="Model path or HF repo id (trust_remote_code=True)")
+    p.add_argument("--output_dir", type=str, required=True, help="Where to save the pruned model")
+    p.add_argument("--calib_path", type=str, default=None, help="Path to calibration data (.txt dir or .jsonl)")
+    p.add_argument("--tokenizer", type=str, default=None, help="Optional tokenizer path; defaults to --model")
+    p.add_argument("--device", type=str, default="cuda" if torch.cuda.is_available() else "cpu")
+    p.add_argument("--dtype", type=str, default="auto", help="auto|fp16|bf16|fp32")
+    p.add_argument("--device_map", type=str, default=None, help="Optional accelerate device_map like 'auto'")
+    p.add_argument("--seed", type=int, default=42)
+    p.add_argument("--samples", type=int, default=200, help="Number of texts to use for calibration")
+    p.add_argument("--seq_len", type=int, default=512)
+    p.add_argument("--batch_size", type=int, default=4)
+    p.add_argument("--score", type=str, default="route_count", choices=["route_count", "prob_mass", "combined"])
+    p.add_argument("--alpha", type=float, default=0.7, help="Weight for route_count in combined scoring")
+    p.add_argument("--target_experts", type=int, default=None, help="Experts per layer to keep (divisible by n_group)")
+    p.add_argument("--target_total_params", type=int, default=None, help="Approximate total params budget, e.g. 10000000000 for ~10B")
+    p.add_argument("--dry_run", action="store_true", help="Collect stats and print plan, but do not prune or save")
+    args = p.parse_args(argv)
+
+    set_seed(args.seed)
+    dtype = infer_dtype(args.dtype)
+
+    # Load tokenizer
+    tok_name = args.tokenizer or args.model
+    tokenizer = AutoTokenizer.from_pretrained(tok_name, trust_remote_code=True)
+
+    # Load model (weights required to get routing behavior)
+    model_kwargs = dict(torch_dtype=dtype, trust_remote_code=True)
+    if args.device_map:
+        model_kwargs["device_map"] = args.device_map
+    model = AutoModelForCausalLM.from_pretrained(args.model, **model_kwargs)
+    if args.device_map is None:
+        model.to(args.device)
+
+    cfg = model.config
+
+    # Determine target experts per layer
+    if args.target_experts is not None and args.target_total_params is not None:
+        print("Both --target_experts and --target_total_params provided; --target_experts takes precedence.")
+
+    if args.target_experts is not None:
+        target_experts = int(args.target_experts)
+    elif args.target_total_params is not None:
+        target_experts = solve_target_experts_for_budget(cfg, int(args.target_total_params))
+        print(f"Resolved experts/layer for target params ≈ {args.target_total_params}: {target_experts}")
+    else:
+        # Default to a 144 experts/layer target (≈10B for Ring-mini-2.0 defaults)
+        target_experts = 144
+        print("No target specified; defaulting to --target_experts 144 (~10B for default config)")
+
+    if target_experts % cfg.n_group != 0:
+        raise ValueError(f"target_experts={target_experts} must be divisible by n_group={cfg.n_group}")
+    if target_experts < cfg.num_experts_per_tok:
+        raise ValueError(f"target_experts={target_experts} must be >= num_experts_per_tok={cfg.num_experts_per_tok}")
+
+    keep_per_group = target_experts // cfg.n_group
+    print(f"Target experts per layer: {target_experts} ({keep_per_group} per group × {cfg.n_group} groups)")
+
+    # Collect calibration stats
+    if args.calib_path is None:
+        raise ValueError("--calib_path is required to compute routing-based scores")
+    texts = iter_texts(args.calib_path)
+    prob_mass, route_cnt = collect_moe_stats(
+        model, tokenizer, texts,
+        max_samples=args.samples, seq_len=args.seq_len, batch_size=args.batch_size, device=args.device,
+    )
+    print("Collected MoE routing statistics.")
+
+    if args.dry_run:
+        print("--dry_run set; skipping pruning and save.")
+        return
+
+    # Prune experts per MoE layer
+    model, selection = prune_model_experts(
+        model,
+        keep_per_group=keep_per_group,
+        scores_prob_mass=prob_mass,
+        scores_route_cnt=route_cnt,
+        score_mode=args.score,
+        alpha=args.alpha,
+    )
+
+    os.makedirs(args.output_dir, exist_ok=True)
+    # Save selection mapping for transparency
+    with open(os.path.join(args.output_dir, "expert_selection.json"), "w", encoding="utf-8") as f:
+        json.dump(selection, f, indent=2)
+
+    # Save model + tokenizer
+    # Default to saving weights in BF16 regardless of compute dtype.
+    # This keeps runtime/calibration flexible while ensuring compact BF16 checkpoints.
+    try:
+        model.to(torch.bfloat16)
+    except Exception as e:
+        print(f"Warning: failed to cast model to BF16 before saving; saving current dtype. Error: {e}")
+    model.save_pretrained(args.output_dir, safe_serialization=True)
+    tokenizer.save_pretrained(args.output_dir)
+    print(f"Saved pruned model to {args.output_dir}")
+
+    # Print number of parameters of the model
+    total_params = sum(p.numel() for p in model.parameters())
+    print(f"Total number of parameters in the model: {total_params}")
+
+
+if __name__ == "__main__":
+    main()

pruning-code/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "bos_token": "<|startoftext|>",
+  "cls_token": "[CLS]",
+  "eos_token": "<|endoftext|>",
+  "gmask_token": "[gMASK]",
+  "pad_token": "<|endoftext|>"
+}

pruning-code/tokenizer_config.json

@@ -0,0 +1,17 @@
+{
+  "add_bos_token": false,
+  "add_eos_token": false,
+  "bos_token": "<|startoftext|>",
+  "chat_template": "{% for message in messages %}{% set role = message['role'] | lower %}{% if role == 'user' %}{% set role = 'HUMAN' %}{% endif %}{% set role = role | upper %}{{ '<role>' + role + '</role>' + message['content'] }}{% endfor %}{% if add_generation_prompt %}{{ '<role>ASSISTANT</role><think>\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "[CLS]",
+  "eos_token": "<|endoftext|>",
+  "fast_tokenizer": true,
+  "gmask_token": "[gMASK]",
+  "merges_file": null,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|endoftext|>",
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "trust_remote_code": true,
+  "vocab_file": null
+}

special_tokens_map.json

@@ -1,7 +1,30 @@
 {
-  "bos_token": "<|startoftext|>",
-  "cls_token": "[CLS]",
-  "eos_token": "<|endoftext|>",
-  "gmask_token": "[gMASK]",
-  "pad_token": "<|endoftext|>"
+  "bos_token": {
+    "content": "<|startoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "cls_token": {
+    "content": "[CLS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
 }

tokenizer_config.json

@@ -1,17 +1,2114 @@
 {
   "add_bos_token": false,
   "add_eos_token": false,
+  "added_tokens_decoder": {
+    "156891": {
+      "content": "<|startoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156892": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156893": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156894": {
+      "content": "[gMASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156895": {
+      "content": "<|reserved_token_0|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156896": {
+      "content": "<|reserved_token_1|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156897": {
+      "content": "<|reserved_token_2|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156898": {
+      "content": "<|reserved_token_3|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156899": {
+      "content": "<|reserved_token_4|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156900": {
+      "content": "<|reserved_token_5|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156901": {
+      "content": "<|reserved_token_6|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156902": {
+      "content": "<|reserved_token_7|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156903": {
+      "content": "<|reserved_token_8|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156904": {
+      "content": "<|reserved_token_9|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156905": {
+      "content": "<|reserved_token_10|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156906": {
+      "content": "<|reserved_token_11|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156907": {
+      "content": "<|reserved_token_12|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156908": {
+      "content": "<|reserved_token_13|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156909": {
+      "content": "<|reserved_token_14|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156910": {
+      "content": "<|reserved_token_15|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156911": {
+      "content": "<|reserved_token_16|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156912": {
+      "content": "<|reserved_token_17|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156913": {
+      "content": "<|reserved_token_18|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156914": {
+      "content": "<|reserved_token_19|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156915": {
+      "content": "<|reserved_token_20|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156916": {
+      "content": "<|reserved_token_21|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156917": {
+      "content": "<|reserved_token_22|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156918": {
+      "content": "<|reserved_token_23|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156919": {
+      "content": "<|reserved_token_24|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156920": {
+      "content": "<|reserved_token_25|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156921": {
+      "content": "<|reserved_token_26|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156922": {
+      "content": "<|reserved_token_27|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156923": {
+      "content": "<|reserved_token_28|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156924": {
+      "content": "<|reserved_token_29|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156925": {
+      "content": "<|reserved_token_30|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156926": {
+      "content": "<|reserved_token_31|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156927": {
+      "content": "<|reserved_token_32|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156928": {
+      "content": "<|reserved_token_33|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156929": {
+      "content": "<|reserved_token_34|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156930": {
+      "content": "<|reserved_token_35|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156931": {
+      "content": "<|reserved_token_36|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156932": {
+      "content": "<|reserved_token_37|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156933": {
+      "content": "<|reserved_token_38|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156934": {
+      "content": "<|reserved_token_39|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156935": {
+      "content": "<|reserved_token_40|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156936": {
+      "content": "<|reserved_token_41|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156937": {
+      "content": "<|reserved_token_42|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156938": {
+      "content": "<|reserved_token_43|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156939": {
+      "content": "<|reserved_token_44|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156940": {
+      "content": "<|reserved_token_45|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156941": {
+      "content": "<|reserved_token_46|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156942": {
+      "content": "<|reserved_token_47|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156943": {
+      "content": "<|reserved_token_48|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156944": {
+      "content": "<|reserved_token_49|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156945": {
+      "content": "<|reserved_token_50|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156946": {
+      "content": "<|reserved_token_51|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156947": {
+      "content": "<|reserved_token_52|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156948": {
+      "content": "<|reserved_token_53|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156949": {
+      "content": "<|reserved_token_54|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156950": {
+      "content": "<|reserved_token_55|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156951": {
+      "content": "<|reserved_token_56|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156952": {
+      "content": "<|reserved_token_57|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156953": {
+      "content": "<|reserved_token_58|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156954": {
+      "content": "<|reserved_token_59|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156955": {
+      "content": "<|reserved_token_60|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156956": {
+      "content": "<|reserved_token_61|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156957": {
+      "content": "<|reserved_token_62|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156958": {
+      "content": "<|reserved_token_63|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156959": {
+      "content": "<|reserved_token_64|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156960": {
+      "content": "<|reserved_token_65|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156961": {
+      "content": "<|reserved_token_66|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156962": {
+      "content": "<|reserved_token_67|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156963": {
+      "content": "<|reserved_token_68|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156964": {
+      "content": "<|reserved_token_69|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156965": {
+      "content": "<|reserved_token_70|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156966": {
+      "content": "<|reserved_token_71|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156967": {
+      "content": "<|reserved_token_72|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156968": {
+      "content": "<|reserved_token_73|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156969": {
+      "content": "<|reserved_token_74|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156970": {
+      "content": "<|reserved_token_75|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156971": {
+      "content": "<|reserved_token_76|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156972": {
+      "content": "<|reserved_token_77|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156973": {
+      "content": "<|reserved_token_78|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156974": {
+      "content": "<|reserved_token_79|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156975": {
+      "content": "<|reserved_token_80|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156976": {
+      "content": "<|reserved_token_81|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156977": {
+      "content": "<|reserved_token_82|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156978": {
+      "content": "<|reserved_token_83|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156979": {
+      "content": "<|reserved_token_84|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156980": {
+      "content": "<|reserved_token_85|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156981": {
+      "content": "<|reserved_token_86|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156982": {
+      "content": "<|reserved_token_87|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156983": {
+      "content": "<|reserved_token_88|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156984": {
+      "content": "<|reserved_token_89|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156985": {
+      "content": "<|reserved_token_90|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156986": {
+      "content": "<|reserved_token_91|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156987": {
+      "content": "<|reserved_token_92|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156988": {
+      "content": "<|reserved_token_93|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156989": {
+      "content": "<|reserved_token_94|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156990": {
+      "content": "<|reserved_token_95|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156991": {
+      "content": "<|reserved_token_96|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156992": {
+      "content": "<|reserved_token_97|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156993": {
+      "content": "<|reserved_token_98|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156994": {
+      "content": "<|reserved_token_99|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156995": {
+      "content": "<|reserved_token_100|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156996": {
+      "content": "<|reserved_token_101|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156997": {
+      "content": "<|reserved_token_102|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156998": {
+      "content": "<|reserved_token_103|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "156999": {
+      "content": "<|reserved_token_104|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157000": {
+      "content": "<|reserved_token_105|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157001": {
+      "content": "<|reserved_token_106|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157002": {
+      "content": "<|reserved_token_107|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157003": {
+      "content": "<|reserved_token_108|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157004": {
+      "content": "<|reserved_token_109|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157005": {
+      "content": "<|reserved_token_110|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157006": {
+      "content": "<|reserved_token_111|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157007": {
+      "content": "<|reserved_token_112|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157008": {
+      "content": "<|reserved_token_113|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157009": {
+      "content": "<|reserved_token_114|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157010": {
+      "content": "<|reserved_token_115|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157011": {
+      "content": "<|reserved_token_116|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157012": {
+      "content": "<|reserved_token_117|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157013": {
+      "content": "<|reserved_token_118|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157014": {
+      "content": "<|reserved_token_119|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157015": {
+      "content": "<|reserved_token_120|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157016": {
+      "content": "<|reserved_token_121|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157017": {
+      "content": "<|reserved_token_122|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157018": {
+      "content": "<|reserved_token_123|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157019": {
+      "content": "<|reserved_token_124|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157020": {
+      "content": "<|reserved_token_125|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157021": {
+      "content": "<|reserved_token_126|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157022": {
+      "content": "<|reserved_token_127|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157023": {
+      "content": "<|reserved_token_128|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157024": {
+      "content": "<|reserved_token_129|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157025": {
+      "content": "<|reserved_token_130|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157026": {
+      "content": "<|reserved_token_131|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157027": {
+      "content": "<|reserved_token_132|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157028": {
+      "content": "<|reserved_token_133|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157029": {
+      "content": "<|reserved_token_134|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157030": {
+      "content": "<|reserved_token_135|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157031": {
+      "content": "<|reserved_token_136|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157032": {
+      "content": "<|reserved_token_137|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157033": {
+      "content": "<|reserved_token_138|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157034": {
+      "content": "<|reserved_token_139|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157035": {
+      "content": "<|reserved_token_140|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157036": {
+      "content": "<|reserved_token_141|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157037": {
+      "content": "<|reserved_token_142|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157038": {
+      "content": "<|reserved_token_143|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157039": {
+      "content": "<|reserved_token_144|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157040": {
+      "content": "<|reserved_token_145|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157041": {
+      "content": "<|reserved_token_146|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157042": {
+      "content": "<|reserved_token_147|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157043": {
+      "content": "<|reserved_token_148|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157044": {
+      "content": "<|reserved_token_149|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157045": {
+      "content": "<|reserved_token_150|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157046": {
+      "content": "<|reserved_token_151|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157047": {
+      "content": "<|reserved_token_152|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157048": {
+      "content": "<|reserved_token_153|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157049": {
+      "content": "<|reserved_token_154|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157050": {
+      "content": "<|reserved_token_155|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157051": {
+      "content": "<|reserved_token_156|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157052": {
+      "content": "<|reserved_token_157|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157053": {
+      "content": "<|reserved_token_158|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157054": {
+      "content": "<|reserved_token_159|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157055": {
+      "content": "<|reserved_token_160|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157056": {
+      "content": "<|reserved_token_161|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157057": {
+      "content": "<|reserved_token_162|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157058": {
+      "content": "<|reserved_token_163|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157059": {
+      "content": "<|reserved_token_164|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157060": {
+      "content": "<|reserved_token_165|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157061": {
+      "content": "<|reserved_token_166|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157062": {
+      "content": "<|reserved_token_167|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157063": {
+      "content": "<|reserved_token_168|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157064": {
+      "content": "<|reserved_token_169|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157065": {
+      "content": "<|reserved_token_170|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157066": {
+      "content": "<|reserved_token_171|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157067": {
+      "content": "<|reserved_token_172|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157068": {
+      "content": "<|reserved_token_173|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157069": {
+      "content": "<|reserved_token_174|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157070": {
+      "content": "<|reserved_token_175|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157071": {
+      "content": "<|reserved_token_176|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157072": {
+      "content": "<|reserved_token_177|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157073": {
+      "content": "<|reserved_token_178|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157074": {
+      "content": "<|reserved_token_179|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157075": {
+      "content": "<|reserved_token_180|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157076": {
+      "content": "<|reserved_token_181|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157077": {
+      "content": "<|reserved_token_182|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157078": {
+      "content": "<|reserved_token_183|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157079": {
+      "content": "<|reserved_token_184|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157080": {
+      "content": "<|reserved_token_185|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157081": {
+      "content": "<|reserved_token_186|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157082": {
+      "content": "<|reserved_token_187|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157083": {
+      "content": "<|reserved_token_188|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157084": {
+      "content": "<|reserved_token_189|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157085": {
+      "content": "<|reserved_token_190|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157086": {
+      "content": "<|reserved_token_191|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157087": {
+      "content": "<|reserved_token_192|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157088": {
+      "content": "<|reserved_token_193|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157089": {
+      "content": "<|reserved_token_194|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157090": {
+      "content": "<|reserved_token_195|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157091": {
+      "content": "<|reserved_token_196|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157092": {
+      "content": "<|reserved_token_197|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157093": {
+      "content": "<|reserved_token_198|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157094": {
+      "content": "<|reserved_token_199|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157095": {
+      "content": "<|reserved_token_200|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157096": {
+      "content": "<|reserved_token_201|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157097": {
+      "content": "<|reserved_token_202|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157098": {
+      "content": "<|reserved_token_203|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157099": {
+      "content": "<|reserved_token_204|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157100": {
+      "content": "<|reserved_token_205|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157101": {
+      "content": "<|reserved_token_206|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157102": {
+      "content": "<|reserved_token_207|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157103": {
+      "content": "<|reserved_token_208|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157104": {
+      "content": "<|reserved_token_209|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157105": {
+      "content": "<|reserved_token_210|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157106": {
+      "content": "<|reserved_token_211|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157107": {
+      "content": "<|reserved_token_212|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157108": {
+      "content": "<|reserved_token_213|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157109": {
+      "content": "<|reserved_token_214|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157110": {
+      "content": "<|reserved_token_215|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157111": {
+      "content": "<|reserved_token_216|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157112": {
+      "content": "<|reserved_token_217|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157113": {
+      "content": "<|reserved_token_218|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157114": {
+      "content": "<|reserved_token_219|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157115": {
+      "content": "<|reserved_token_220|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157116": {
+      "content": "<|reserved_token_221|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157117": {
+      "content": "<|reserved_token_222|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157118": {
+      "content": "<|reserved_token_223|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157119": {
+      "content": "<|reserved_token_224|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157120": {
+      "content": "<|reserved_token_225|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157121": {
+      "content": "<|reserved_token_226|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157122": {
+      "content": "<|reserved_token_227|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157123": {
+      "content": "<|reserved_token_228|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157124": {
+      "content": "<|reserved_token_229|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157125": {
+      "content": "<|reserved_token_230|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157126": {
+      "content": "<|reserved_token_231|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157127": {
+      "content": "<|reserved_token_232|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157128": {
+      "content": "<|reserved_token_233|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157129": {
+      "content": "<|reserved_token_234|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157130": {
+      "content": "<|reserved_token_235|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157131": {
+      "content": "<|reserved_token_236|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157132": {
+      "content": "<|reserved_token_237|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157133": {
+      "content": "<|reserved_token_238|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157134": {
+      "content": "<|reserved_token_239|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157135": {
+      "content": "<|reserved_token_240|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157136": {
+      "content": "<|reserved_token_241|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157137": {
+      "content": "<|reserved_token_242|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157138": {
+      "content": "<|reserved_token_243|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157139": {
+      "content": "<|reserved_token_244|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157140": {
+      "content": "<|reserved_token_245|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157141": {
+      "content": "<|reserved_token_246|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157142": {
+      "content": "<|reserved_token_247|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157143": {
+      "content": "<|reserved_token_248|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157144": {
+      "content": "<|reserved_token_249|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157145": {
+      "content": "<|reserved_token_250|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157146": {
+      "content": "<|reserved_token_251|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157147": {
+      "content": "<|reserved_token_252|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157148": {
+      "content": "<|reserved_token_253|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157149": {
+      "content": "<|reserved_token_254|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157150": {
+      "content": "<|reserved_token_255|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157151": {
+      "content": "<role>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "157152": {
+      "content": "</role>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
   "bos_token": "<|startoftext|>",
-  "chat_template": "{% for message in messages %}{% set role = message['role'] | lower %}{% if role == 'user' %}{% set role = 'HUMAN' %}{% endif %}{% set role = role | upper %}{{ '<role>' + role + '</role>' + message['content'] }}{% endfor %}{% if add_generation_prompt %}{{ '<role>ASSISTANT</role><think>\n' }}{% endif %}",
   "clean_up_tokenization_spaces": false,
   "cls_token": "[CLS]",
   "eos_token": "<|endoftext|>",
+  "extra_special_tokens": {},
   "fast_tokenizer": true,
   "gmask_token": "[gMASK]",
   "merges_file": null,
   "model_max_length": 1000000000000000019884624838656,
   "pad_token": "<|endoftext|>",
   "tokenizer_class": "PreTrainedTokenizerFast",
-  "trust_remote_code": true,
-  "vocab_file": null
+  "trust_remote_code": true
 }