Upload HATForCausalLM

config.json

@@ -0,0 +1,86 @@
+{
+  "auto_map": {
+    "AutoConfig": "config.HATArchitectureConfig",
+    "AutoModelForCausalLM": "model.HATForCausalLM"
+  },
+  "backbone_config": {
+    "hidden_size": 4096,
+    "intermediate_size": 14336,
+    "max_position_embeddings": 32900,
+    "mlp_bias": false,
+    "num_attention_heads": 32,
+    "num_hidden_layers": 32,
+    "num_key_value_heads": 8,
+    "rms_norm_eps": 1e-05,
+    "rope_scaling": {
+      "factor": 8.0,
+      "high_freq_factor": 4.0,
+      "low_freq_factor": 1.0,
+      "original_max_position_embeddings": 8192,
+      "rope_type": "llama3"
+    },
+    "rope_theta": 500000,
+    "sliding_window": null,
+    "transformers_version": null,
+    "use_cache": true,
+    "vocab_size": 0
+  },
+  "decoder_config": {
+    "cross_attention_config": {
+      "attention_num_kv_heads": 8,
+      "hidden_size": 1024,
+      "hidden_size_kv": 4096,
+      "hidden_size_q": 1024,
+      "num_attention_heads": 8,
+      "word_window_size": 1
+    },
+    "cross_attn_every_layer": true,
+    "hidden_size": 1024,
+    "intermediate_size": 2816,
+    "max_position_embeddings": 262144,
+    "mlp_bias": false,
+    "num_attention_heads": 8,
+    "num_hidden_layers": 4,
+    "num_key_value_heads": 8,
+    "rms_norm_eps": 1e-05,
+    "rope_scaling": {
+      "rope_type": "default"
+    },
+    "rope_theta": 100000,
+    "sliding_window": 768,
+    "transformers_version": null,
+    "use_cache": true,
+    "vocab_size": 256
+  },
+  "encoder_config": {
+    "cross_attention_config": {
+      "attention_num_kv_heads": 32,
+      "hidden_size": 4096,
+      "hidden_size_kv": 1024,
+      "hidden_size_q": 4096,
+      "num_attention_heads": 32,
+      "word_window_size": 1
+    },
+    "hidden_size": 1024,
+    "intermediate_size": 2816,
+    "max_position_embeddings": 262144,
+    "mlp_bias": false,
+    "num_attention_heads": 8,
+    "num_hidden_layers": 6,
+    "num_key_value_heads": 8,
+    "rms_norm_eps": 1e-05,
+    "rope_scaling": {
+      "rope_type": "default"
+    },
+    "rope_theta": 100000,
+    "sliding_window": 768,
+    "transformers_version": null,
+    "use_cache": true,
+    "vocab_size": 256
+  },
+  "model_type": "hierarchical_autoregressive_transformer",
+  "special_token_dict": {
+    "<|eot_id|>": 192
+  },
+  "transformers_version": "4.46.3"
+}

config.py

@@ -0,0 +1,232 @@
+from dataclasses import dataclass
+
+import torch.nn as nn
+from transformers.configuration_utils import PretrainedConfig
+from transformers.models.llama.configuration_llama import LlamaConfig
+
+
+@dataclass
+class TransformerHATModelConfig(LlamaConfig):
+    def __init__(
+        self,
+        hidden_size: int,
+        num_hidden_layers: int,
+        num_attention_heads: int,
+        num_key_value_heads: int,
+        rms_norm_eps: float,
+        intermediate_size: int,
+        max_position_embeddings: int,
+        rope_scaling: dict,
+        rope_theta: float,
+        mlp_bias: bool,
+        use_cache: bool = True,
+        sliding_window: int | None = None,
+        vocab_size: int = 0,
+        hidden_act: str = "silu",
+        **kwargs,
+    ):
+        super().__init__(
+            vocab_size=vocab_size,
+            hidden_size=hidden_size,
+            num_hidden_layers=num_hidden_layers,
+            num_attention_heads=num_attention_heads,
+            num_key_value_heads=num_key_value_heads,
+            hidden_act=hidden_act,
+            rms_norm_eps=rms_norm_eps,
+            intermediate_size=intermediate_size,
+            max_position_embeddings=max_position_embeddings,
+            rope_scaling=rope_scaling,
+            rope_theta=rope_theta,
+            mlp_bias=mlp_bias,
+            use_cache=use_cache,
+            **kwargs,
+        )
+
+        self.sliding_window = sliding_window
+
+    def to_dict(self):
+        config_dict = {
+            "vocab_size": self.vocab_size,
+            "hidden_size": self.hidden_size,
+            "num_hidden_layers": self.num_hidden_layers,
+            "num_attention_heads": self.num_attention_heads,
+            "num_key_value_heads": self.num_key_value_heads,
+            "rms_norm_eps": self.rms_norm_eps,
+            "intermediate_size": self.intermediate_size,
+            "max_position_embeddings": self.max_position_embeddings,
+            "rope_scaling": self.rope_scaling,
+            "rope_theta": self.rope_theta,
+            "mlp_bias": self.mlp_bias,
+            "use_cache": self.use_cache,
+            "sliding_window": self.sliding_window,
+            "transformers_version": self.transformers_version,
+        }
+        return config_dict
+
+
+@dataclass
+class CrossAttentionConfig:
+    def __init__(
+        self,
+        hidden_size: int,
+        hidden_size_q: int,
+        hidden_size_kv: int,
+        num_attention_heads: int,
+        attention_num_kv_heads: int,
+        word_window_size: int,
+    ):
+        self.hidden_size = hidden_size
+        self.hidden_size_q = hidden_size_q
+        self.hidden_size_kv = hidden_size_kv
+        self.num_attention_heads = num_attention_heads
+        self.attention_num_kv_heads = attention_num_kv_heads
+        self.word_window_size = word_window_size
+
+    def to_dict(self):
+        return {
+            "hidden_size_q": self.hidden_size_q,
+            "hidden_size_kv": self.hidden_size_kv,
+            "hidden_size": self.hidden_size,
+            "num_attention_heads": self.num_attention_heads,
+            "attention_num_kv_heads": self.attention_num_kv_heads,
+            "word_window_size": self.word_window_size,
+        }
+
+
+@dataclass
+class DecoderHATModelConfig(TransformerHATModelConfig):
+    def __init__(
+        self,
+        num_attention_heads: int,
+        num_key_value_heads: int,
+        sliding_window: int,
+        cross_attention_config: CrossAttentionConfig,
+        cross_attn_every_layer: bool,
+        **kwargs,
+    ):
+        super().__init__(
+            num_attention_heads=num_attention_heads,
+            num_key_value_heads=num_key_value_heads,
+            sliding_window=sliding_window,
+            **kwargs,
+        )
+        self.cross_attn_every_layer = cross_attn_every_layer
+        self.cross_attention_config = cross_attention_config
+
+    def to_dict(self):
+        config_dict = super().to_dict()
+        config_dict["cross_attn_every_layer"] = self.cross_attn_every_layer
+        config_dict["cross_attention_config"] = self.cross_attention_config.to_dict()
+        return config_dict
+
+    @classmethod
+    def from_dict(cls, config_dict, **kwargs):
+        config_dict = config_dict.copy()  # Avoid modifying the original dict
+        config_dict.update(kwargs)  # Apply overrides
+        dict_config = config_dict.pop("cross_attention_config", {})
+        cross_attention_config = CrossAttentionConfig(**dict_config)
+        config_dict["cross_attention_config"] = cross_attention_config
+        return cls(**config_dict)
+
+
+@dataclass
+class EncoderHATModelConfig(TransformerHATModelConfig):
+    def __init__(
+        self,
+        cross_attention_config: CrossAttentionConfig,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+        self.cross_attention_config = cross_attention_config
+
+    @classmethod
+    def from_dict(cls, config_dict, **kwargs):
+        config_dict = config_dict.copy()  # Avoid modifying the original dict
+        config_dict.update(kwargs)  # Apply overrides
+        dict_config = config_dict.pop("cross_attention_config", {})
+        cross_attention_config = CrossAttentionConfig(**dict_config)
+        config_dict["cross_attention_config"] = cross_attention_config
+
+        return cls(**config_dict)
+
+    def to_dict(self):
+        config_dict = super().to_dict()
+        if self.cross_attention_config:
+            config_dict["cross_attention_config"] = self.cross_attention_config.to_dict()
+        return config_dict
+
+
+@dataclass
+class HATArchitectureConfig(PretrainedConfig):
+    model_type: str
+
+    def __init__(
+        self,
+        special_token_dict : dict | None = None,
+        encoder_config: EncoderHATModelConfig | None = None,
+        backbone_config: TransformerHATModelConfig | None = None,
+        decoder_config: DecoderHATModelConfig | None = None,
+        model_type: str = "hierarchical_autoregressive_transformer",
+        eos_token_id: int = 192,
+        max_word_size: int = 100,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+        self.encoder_config = encoder_config
+        self.backbone_config = backbone_config
+        self.decoder_config = decoder_config
+        self.model_type = model_type
+        self.eos_token_id = eos_token_id
+        self.max_word_size = max_word_size
+        self.special_token_dict = special_token_dict
+        self.transformers_version = "4.46.3"
+
+    @classmethod
+    def from_dict(cls, config_dict, **kwargs):
+        """
+        Instantiates a HATArchitectureConfig from a Python dictionary of parameters.
+
+        Overrides the base `from_dict` to correctly handle nested config objects.
+        """
+        config_dict = config_dict.copy()  # Avoid modifying the original dict
+        config_dict.update(kwargs)  # Apply overrides
+
+        # Pop and instantiate nested config dictionaries
+        encoder_dict = config_dict.pop("encoder_config", {})
+        backbone_dict = config_dict.pop("backbone_config", {})
+        decoder_dict = config_dict.pop("decoder_config", {})
+
+        # Instantiate nested configs
+        encoder_config = EncoderHATModelConfig.from_dict(encoder_dict) if encoder_dict else None
+        backbone_config = TransformerHATModelConfig.from_dict(backbone_dict) if backbone_dict else None
+        decoder_config = DecoderHATModelConfig.from_dict(decoder_dict) if decoder_dict else None
+        special_token_dict = config_dict.pop("special_token_dict", {"<|eot_id|>": 192})
+        max_word_size = config_dict.pop("max_word_size", 100)
+        return cls(
+            encoder_config=encoder_config,
+            backbone_config=backbone_config,
+            decoder_config=decoder_config,
+            special_token_dict=special_token_dict,
+            max_word_size=max_word_size,
+            **config_dict,
+        ), {}
+
+    def to_dict(self):
+        config_dict = {}
+        if self.encoder_config:
+            config_dict["encoder_config"] = self.encoder_config.to_dict()
+        if self.backbone_config:
+            config_dict["backbone_config"] = self.backbone_config.to_dict()
+        if self.decoder_config:
+            config_dict["decoder_config"] = self.decoder_config.to_dict()
+        config_dict["model_type"] = self.model_type
+        config_dict["transformers_version"] = self.transformers_version
+        config_dict["auto_map"] = {"AutoConfig": "config.HATArchitectureConfig", "AutoModelForCausalLM": "model.HATForCausalLM"}
+        config_dict["special_token_dict"] = self.special_token_dict
+        return config_dict
+
+
+class EncoderHATModel(nn.Module):
+    def __init__(self, config: HATArchitectureConfig, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        self.config = config

generation_config.json

@@ -0,0 +1,4 @@
+{
+  "_from_model_config": true,
+  "transformers_version": "4.46.3"
+}

model-00001-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f521d1f8718e6f1a5c01024634d4d93601510e7a5220cf9bff1bb1d99127da0
+size 4919609440

model-00002-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a34cfe3c6655a656ca8cde7015c38e41cff670c3be80dd387346c310fdda2307
+size 4999819656

model-00003-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e80d1987d9ecbc99651a16bc1b7181d4e975ebd1fa8c4ef821678b8ea7054407
+size 4465610896

model.safetensors.index.json

@@ -0,0 +1,418 @@
+{
+  "metadata": {
+    "total_size": 14384990208
+  },
+  "weight_map": {
+    "backbone.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.10.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.10.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.10.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.10.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.10.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.10.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.10.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.10.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.10.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.11.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.11.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.11.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.11.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.11.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.12.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.12.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.12.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.12.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.12.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.12.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.12.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.13.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.13.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.13.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.13.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.14.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.14.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.14.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.16.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.17.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.18.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.19.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.19.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.19.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.20.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.20.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.20.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.20.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.20.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.21.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.21.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.21.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.22.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.22.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.22.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.22.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.22.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.22.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
+    "backbone.layers.23.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.23.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.23.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.23.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.23.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.23.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.23.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.23.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.23.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.24.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.24.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.24.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.24.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.24.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.24.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.24.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.24.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.24.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.25.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.25.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.25.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.25.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.25.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.25.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.25.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.25.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.25.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.26.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.26.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.26.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.26.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.26.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.26.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.26.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.26.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.26.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.27.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.27.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.27.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.27.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.27.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.27.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.27.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.27.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.27.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.28.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.28.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.28.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.28.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.28.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.28.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.28.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.28.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.28.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.29.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.29.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.29.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.29.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.29.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.29.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.29.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.29.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.29.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.30.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.30.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.30.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.30.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.30.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.30.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.30.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.30.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.30.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.31.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.31.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.31.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.31.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.31.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.31.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.31.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.31.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.31.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "backbone.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.4.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.4.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.5.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.6.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.6.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.6.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.8.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.8.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.8.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.9.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.9.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.9.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "backbone.layers.9.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "decoder.decoder_layers.0.cross_attention.k_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.cross_attention.o_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.cross_attention.q_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.cross_attention.v_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.kv_norm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.llama_layer.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.llama_layer.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.llama_layer.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.llama_layer.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.llama_layer.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.llama_layer.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.llama_layer.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.llama_layer.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.llama_layer.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.0.query_norm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.cross_attention.k_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.cross_attention.o_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.cross_attention.q_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.cross_attention.v_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.kv_norm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.llama_layer.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.llama_layer.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.llama_layer.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.llama_layer.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.llama_layer.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.llama_layer.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.llama_layer.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.llama_layer.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.llama_layer.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.1.query_norm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.cross_attention.k_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.cross_attention.o_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.cross_attention.q_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.cross_attention.v_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.kv_norm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.llama_layer.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.llama_layer.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.llama_layer.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.llama_layer.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.llama_layer.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.llama_layer.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.llama_layer.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.llama_layer.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.llama_layer.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.2.query_norm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.cross_attention.k_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.cross_attention.o_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.cross_attention.q_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.cross_attention.v_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.kv_norm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.llama_layer.input_layernorm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.llama_layer.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.llama_layer.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.llama_layer.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.llama_layer.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.llama_layer.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.llama_layer.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.llama_layer.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.llama_layer.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
+    "decoder.decoder_layers.3.query_norm.weight": "model-00003-of-00003.safetensors",
+    "decoder_connector.first_word_embedding": "model-00003-of-00003.safetensors",
+    "encoder.embedding_layer.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder_connector.cross_attention_encoder_connector.k_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder_connector.cross_attention_encoder_connector.o_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder_connector.cross_attention_encoder_connector.q_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder_connector.cross_attention_encoder_connector.v_proj.weight": "model-00001-of-00003.safetensors",
+    "encoder_connector.latent_query": "model-00001-of-00003.safetensors",
+    "layer_norm.weight": "model-00003-of-00003.safetensors",
+    "lm_head.weight": "model-00003-of-00003.safetensors"
+  }
+}