from typing import List, Optional, Tuple

import torch
from torch import nn

import transformers
from transformers.models.llama.modeling_llama import apply_rotary_pos_emb
import torch
from torch import nn
import torch.nn.functional as F
import math
import transformers
from transformers.models.llama.modeling_llama import apply_rotary_pos_emb
    

def forward_2(
        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: bool = False,
        use_cache: bool = False,
) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
    bsz, q_len, _ = hidden_states.size()

    query_states = self.q_proj(hidden_states).view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)
    key_states = self.k_proj(hidden_states).view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)
    value_states = self.v_proj(hidden_states).view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)

    kv_seq_len = key_states.shape[-2]
    if past_key_value is not None:
        kv_seq_len += past_key_value[0].shape[-2]
    cos, sin = self.rotary_emb(value_states, seq_len=kv_seq_len)
    query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin, position_ids)

    assert not output_attentions, "output_attentions is not supported"
    assert not use_cache, "use_cache is not supported"
    assert past_key_value is None, "past_key_value is not supported"


    if past_key_value is not None:
        # reuse k, v, self_attention
        key_states = torch.cat([past_key_value[0], key_states], dim=2)
        value_states = torch.cat([past_key_value[1], value_states], dim=2)

    past_key_value = (key_states, value_states) if use_cache else None
    attn_output= F.scaled_dot_product_attention(query_states,key_states,value_states,dropout_p=0.0, is_causal=True)
    attn_weights = None
    
    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)
    attn_output = attn_output.reshape(bsz, q_len, self.hidden_size)

    attn_output = self.o_proj(attn_output)

    if not output_attentions:
        attn_weights = None

    return attn_output, attn_weights, past_key_value


def _prepare_decoder_attention_mask(self, attention_mask, input_shape,
                                    inputs_embeds, past_key_values_length):
    # [bsz, seq_len]
    return attention_mask

def replace_llama_attn_with_flash_attn():
    transformers.models.llama.modeling_llama.LlamaModel._prepare_decoder_attention_mask = _prepare_decoder_attention_mask
    transformers.models.llama.modeling_llama.LlamaAttention.forward = forward_2