ExtractableMistralAttention.py

import torch
from torch import nn

from typing import Optional, Tuple
import warnings
import math

from transformers.cache_utils import Cache
from transformers.models.mistral.modeling_mistral import apply_rotary_pos_emb, repeat_kv

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,
    **kwargs,
) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
    if "padding_mask" in kwargs:
        warnings.warn(
            "Passing `padding_mask` is deprecated and will be removed in v4.37. Please make sure use `attention_mask` instead.`"
        )
    bsz, q_len, _ = hidden_states.size()

    query_states = self.q_proj(hidden_states)
    key_states = self.k_proj(hidden_states)
    value_states = self.v_proj(hidden_states)

    query_states = query_states.view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)
    key_states = key_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
    value_states = value_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)

    kv_seq_len = key_states.shape[-2]
    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."
            )
        kv_seq_len += past_key_value.get_usable_length(kv_seq_len, self.layer_idx)
    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)

    if past_key_value is not None:
        cache_kwargs = {"sin": sin, "cos": cos}  # Specific to RoPE models
        key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs)

    # repeat k/v heads if n_kv_heads < n_heads
    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)

    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)

    d = value_states.size(-1)                                                                   # Custom

    averaging_matrix = torch.full(value_states.shape, 1/d)                                      # Custom
    custom_attn_output = torch.matmul(attn_weights.to('cpu'), averaging_matrix)                 # Custom

    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, self.hidden_size)

    custom_attn_output = custom_attn_output.transpose(1, 2).contiguous()                        # Custom
    custom_attn_output = custom_attn_output.reshape(bsz, q_len, self.hidden_size)               # Custom
    self.attn_vec = custom_attn_output                                                          # Custom

    attn_output = self.o_proj(attn_output)

    if not output_attentions:
        attn_weights = None

    return attn_output, attn_weights, past_key_value