import math
from collections.abc import Callable

import torch
import torch.nn.functional as F
from torch import nn

from ... import initialization as init
from ...cache_utils import Cache
from ...modeling_flash_attention_utils import FlashAttentionKwargs
from ...modeling_layers import GenericForSequenceClassification, GenericForTokenClassification
from ...modeling_utils import ALL_ATTENTION_FUNCTIONS, PreTrainedModel
from ...processing_utils import Unpack
from ...utils import logging
from ...utils.generic import is_flash_attention_requested
from ..llama.modeling_llama import (
    LlamaDecoderLayer,
    LlamaForCausalLM,
    LlamaModel,
    LlamaPreTrainedModel,
    LlamaRMSNorm,
    LlamaRotaryEmbedding,
    apply_rotary_pos_emb,
    eager_attention_forward,
    rotate_half,
)
from ..mixtral.modeling_mixtral import MixtralExperts
from ..qwen2_moe.modeling_qwen2_moe import Qwen2MoeMLP
from .configuration_deepseek_v3 import DeepseekV3Config


logger = logging.get_logger(__name__)


class DeepseekV3RMSNorm(LlamaRMSNorm):
    pass


class DeepseekV3RotaryEmbedding(LlamaRotaryEmbedding):
    pass


class DeepseekV3MLP(Qwen2MoeMLP):
    pass


def apply_rotary_pos_emb_interleave(q, k, cos, sin, position_ids=None, unsqueeze_dim=1):
    r"""
    TODO let's just use the original freqcis computation to not have the view
    transpose + reshape! This is not optimized!
    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.
        position_ids (`torch.Tensor`):
            The position indices of the tokens corresponding to the query and key tensors. For example, this can be
            used to pass offsetted position ids when working with a KV-cache.
        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 of the query and key tensors rotated using the Rotary Position Embedding.
    """
    cos = cos.unsqueeze(unsqueeze_dim)
    sin = sin.unsqueeze(unsqueeze_dim)

    b, h, s, d = q.shape
    q = q.view(b, h, s, d // 2, 2).transpose(4, 3).reshape(b, h, s, d)

    b, h, s, d = k.shape
    k = k.view(b, h, s, d // 2, 2).transpose(4, 3).reshape(b, h, s, d)

    q_embed = (q * cos) + (rotate_half(q) * sin)
    k_embed = (k * cos) + (rotate_half(k) * sin)
    return q_embed, k_embed


def yarn_get_mscale(scale=1, mscale=1):
    if scale <= 1:
        return 1.0
    return 0.1 * mscale * math.log(scale) + 1.0


class DeepseekV3TopkRouter(nn.Module):
    def __init__(self, config):
        super().__init__()
        self.config = config
        self.n_routed_experts = config.n_routed_experts

        self.weight = nn.Parameter(torch.empty((self.n_routed_experts, config.hidden_size)))
        self.register_buffer("e_score_correction_bias", torch.zeros(self.n_routed_experts))

    def forward(self, hidden_states):
        hidden_states = hidden_states.view(-1, self.config.hidden_size)
        router_logits = F.linear(hidden_states.type(torch.float32), self.weight.type(torch.float32))
        return router_logits


class DeepseekV3NaiveMoe(MixtralExperts):
    def __init__(self, config):
        super().__init__(config)
        self.num_experts = config.num_local_experts
        self.intermediate_dim = config.moe_intermediate_size


class DeepseekV3MoE(nn.Module):
    """
    A mixed expert module containing shared experts.
    """

    def __init__(self, config):
        super().__init__()
        self.config = config
        self.experts = DeepseekV3NaiveMoe(config)
        self.gate = DeepseekV3TopkRouter(config)
        self.shared_experts = DeepseekV3MLP(
            config=config, intermediate_size=config.moe_intermediate_size * config.n_shared_experts
        )
        self.n_routed_experts = config.n_routed_experts
        self.n_group = config.n_group
        self.topk_group = config.topk_group
        self.norm_topk_prob = config.norm_topk_prob
        self.routed_scaling_factor = config.routed_scaling_factor
        self.top_k = config.num_experts_per_tok

    def route_tokens_to_experts(self, router_logits):
        router_logits = router_logits.sigmoid()
        router_logits_for_choice = router_logits + self.gate.e_score_correction_bias
        group_scores = (
            router_logits_for_choice.view(-1, self.n_group, self.n_routed_experts // self.n_group)
            .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)
        score_mask = (
            group_mask.unsqueeze(-1)
            .expand(-1, self.n_group, self.n_routed_experts // self.n_group)
            .reshape(-1, self.n_routed_experts)
        )
        scores_for_choice = router_logits_for_choice.masked_fill(~score_mask.bool(), 0.0)
        topk_indices = torch.topk(scores_for_choice, k=self.top_k, dim=-1, sorted=False)[1]
        topk_weights = router_logits.gather(1, topk_indices)
        if self.norm_topk_prob:
            denominator = topk_weights.sum(dim=-1, keepdim=True) + 1e-20
            topk_weights /= denominator
        topk_weights = topk_weights * self.routed_scaling_factor
        return topk_indices, topk_weights

    def forward(self, hidden_states):
        residuals = hidden_states
        orig_shape = hidden_states.shape
        router_logits = self.gate(hidden_states)
        topk_indices, topk_weights = self.route_tokens_to_experts(router_logits)
        hidden_states = hidden_states.view(-1, hidden_states.shape[-1])
        hidden_states = self.experts(hidden_states, topk_indices, topk_weights).view(*orig_shape)
        hidden_states = hidden_states + self.shared_experts(residuals)
        return hidden_states


class DeepseekV3Attention(nn.Module):
    """Multi-headed attention from 'Attention Is All You Need' paper"""

    def __init__(self, config: DeepseekV3Config, layer_idx: int):
        super().__init__()
        self.config = config
        self.layer_idx = layer_idx
        self.num_key_value_groups = config.num_attention_heads // config.num_key_value_heads
        self.attention_dropout = config.attention_dropout
        self.num_heads = config.num_attention_heads

        self.q_lora_rank = config.q_lora_rank
        self.qk_rope_head_dim = config.qk_rope_head_dim
        self.kv_lora_rank = config.kv_lora_rank
        self.v_head_dim = config.v_head_dim
        self.qk_nope_head_dim = config.qk_nope_head_dim
        self.qk_head_dim = config.qk_head_dim

        self.is_causal = True
        if self.q_lora_rank is None:
            self.q_proj = nn.Linear(config.hidden_size, self.num_heads * self.qk_head_dim, bias=False)
        else:
            self.q_a_proj = nn.Linear(config.hidden_size, config.q_lora_rank, bias=config.attention_bias)
            self.q_a_layernorm = DeepseekV3RMSNorm(config.q_lora_rank)
            self.q_b_proj = nn.Linear(config.q_lora_rank, self.num_heads * self.qk_head_dim, bias=False)

        self.kv_a_proj_with_mqa = nn.Linear(
            config.hidden_size,
            self.kv_lora_rank + self.qk_rope_head_dim,
            bias=config.attention_bias,
        )
        self.kv_a_layernorm = DeepseekV3RMSNorm(self.kv_lora_rank)
        self.kv_b_proj = nn.Linear(
            self.kv_lora_rank,
            self.num_heads * (self.qk_nope_head_dim + self.v_head_dim),
            bias=False,
        )

        self.o_proj = nn.Linear(
            self.num_heads * self.v_head_dim,
            config.hidden_size,
            bias=config.attention_bias,
        )

        self.scaling = self.qk_head_dim ** (-0.5)
        if self.config.rope_parameters.get("rope_type", "default") != "default":
            mscale_all_dim = self.config.rope_parameters.get("mscale_all_dim", 0)
            scaling_factor = self.config.rope_parameters["factor"]
            if mscale_all_dim:
                mscale = yarn_get_mscale(scaling_factor, mscale_all_dim)
                self.scaling = self.scaling * mscale * mscale

    def forward(
        self,
        hidden_states: torch.Tensor,
        position_embeddings: tuple[torch.Tensor, torch.Tensor],
        attention_mask: torch.Tensor | None,
        past_key_values: Cache | None = None,
        cache_position: torch.LongTensor | None = None,
        **kwargs: Unpack[FlashAttentionKwargs],
    ) -> tuple[torch.Tensor, torch.Tensor | None, tuple[torch.Tensor] | None]:
        batch_size, seq_length = hidden_states.shape[:-1]
        query_shape = (batch_size, seq_length, -1, self.qk_head_dim)
        key_shape = (batch_size, seq_length, -1, self.qk_nope_head_dim + self.v_head_dim)

        if self.q_lora_rank is None:
            q_states = self.q_proj(hidden_states)
        else:
            q_states = self.q_b_proj(self.q_a_layernorm(self.q_a_proj(hidden_states)))
        q_states = q_states.view(query_shape).transpose(1, 2)
        q_pass, q_rot = torch.split(q_states, [self.qk_nope_head_dim, self.qk_rope_head_dim], dim=-1)

        compressed_kv = self.kv_a_proj_with_mqa(hidden_states)
        k_pass, k_rot = torch.split(compressed_kv, [self.kv_lora_rank, self.qk_rope_head_dim], dim=-1)

        k_pass = self.kv_b_proj(self.kv_a_layernorm(k_pass)).view(key_shape).transpose(1, 2)
        k_pass, value_states = torch.split(k_pass, [self.qk_nope_head_dim, self.v_head_dim], dim=-1)

        k_rot = k_rot.view(batch_size, 1, seq_length, self.qk_rope_head_dim)

        cos, sin = position_embeddings
        if self.config.rope_interleave:  # support using interleaved weights for efficiency
            q_rot, k_rot = apply_rotary_pos_emb_interleave(q_rot, k_rot, cos, sin)
        else:
            q_rot, k_rot = apply_rotary_pos_emb(q_rot, k_rot, cos, sin)
        k_rot = k_rot.expand(*k_pass.shape[:-1], -1)

        query_states = torch.cat((q_pass, q_rot), dim=-1)
        key_states = torch.cat((k_pass, k_rot), dim=-1)

        if past_key_values is not None:
            # sin and cos are specific to RoPE models; cache_position needed for the static cache
            cache_kwargs = {"sin": sin, "cos": cos, "cache_position": cache_position}
            key_states, value_states = past_key_values.update(key_states, value_states, self.layer_idx, cache_kwargs)

        if is_flash_attention_requested(self.config) and self.qk_head_dim != self.v_head_dim:
            value_states = F.pad(value_states, [0, self.qk_head_dim - self.v_head_dim])

        attention_interface: Callable = ALL_ATTENTION_FUNCTIONS.get_interface(
            self.config._attn_implementation, eager_attention_forward
        )

        attn_output, attn_weights = attention_interface(
            self,
            query_states,
            key_states,
            value_states,
            attention_mask,
            dropout=0.0 if not self.training else self.attention_dropout,
            scaling=self.scaling,
            **kwargs,
        )

        if is_flash_attention_requested(self.config) and self.qk_head_dim != self.v_head_dim:
            attn_output = attn_output[:, :, :, : self.v_head_dim]

        attn_output = attn_output.reshape(batch_size, seq_length, -1).contiguous()
        attn_output = self.o_proj(attn_output)
        return attn_output, attn_weights


class DeepseekV3DecoderLayer(LlamaDecoderLayer):
    def __init__(self, config: DeepseekV3Config, layer_idx: int):
        nn.Module.__init__(self)
        self.hidden_size = config.hidden_size

        self.self_attn = DeepseekV3Attention(config=config, layer_idx=layer_idx)

        if layer_idx >= config.first_k_dense_replace:
            self.mlp = DeepseekV3MoE(config)
        else:
            self.mlp = DeepseekV3MLP(config)

        self.input_layernorm = DeepseekV3RMSNorm(config.hidden_size, eps=config.rms_norm_eps)
        self.post_attention_layernorm = DeepseekV3RMSNorm(config.hidden_size, eps=config.rms_norm_eps)


class DeepseekV3PreTrainedModel(LlamaPreTrainedModel):
    _keep_in_fp32_modules_strict = ["e_score_correction_bias"]
    _keys_to_ignore_on_load_unexpected = [r"model\.layers\.61.*"]

    @torch.no_grad()
    def _init_weights(self, module):
        PreTrainedModel._init_weights(self, module)
        if isinstance(module, DeepseekV3TopkRouter):
            init.normal_(module.weight, mean=0.0, std=self.config.initializer_range)
            init.zeros_(module.e_score_correction_bias)
        elif isinstance(module, DeepseekV3NaiveMoe):
            init.normal_(module.gate_up_proj, mean=0.0, std=self.config.initializer_range)
            init.normal_(module.down_proj, mean=0.0, std=self.config.initializer_range)


class DeepseekV3Model(LlamaModel):
    pass


class DeepseekV3ForCausalLM(LlamaForCausalLM):
    pass


class DeepseekV3ForSequenceClassification(GenericForSequenceClassification, DeepseekV3PreTrainedModel):
    pass


class DeepseekV3ForTokenClassification(GenericForTokenClassification, DeepseekV3PreTrainedModel):
    pass


__all__ = [
    "DeepseekV3PreTrainedModel",
    "DeepseekV3Model",
    "DeepseekV3ForCausalLM",
    "DeepseekV3ForSequenceClassification",
    "DeepseekV3ForTokenClassification",
]