modeling_backpack_gpt2.py

import math
from dataclasses import dataclass
from typing import Optional, Tuple

import torch
import torch.utils.checkpoint
from torch import nn

from transformers.activations import ACT2FN
from transformers.pytorch_utils import Conv1D
from transformers.utils import (
    ModelOutput,
    logging,
)
from transformers.models.gpt2.modeling_gpt2 import GPT2Model, GPT2PreTrainedModel
from .configuration_backpack_gpt2 import BackpackGPT2Config # From the same directory

logger = logging.get_logger(__name__)


### Backpack-Specific
class BackpackGPT2PreTrainedModel(GPT2PreTrainedModel):
    """
    An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained
    models.
    """
    _keys_to_ignore_on_load_missing = [r"attn.masked_bias", r"attn.bias"]

    config_class = BackpackGPT2Config
    base_model_prefix = "backpack"
    is_parallelizable = True
    supports_gradient_checkpointing = False
    _no_split_modules = ["GPT2Block", "BackpackNoMixBlock"]

    def __init__(self, *inputs, **kwargs):
        super().__init__(*inputs, **kwargs)

class BackpackMLP(nn.Module):

  def __init__(self, embed_dim, intermediate_dim, out_dim, config):
        super().__init__()
        self.c_fc = Conv1D(intermediate_dim, embed_dim)
        self.c_proj = Conv1D(out_dim, intermediate_dim)
        self.act = ACT2FN[config.activation_function]
        self.dropout = nn.Dropout(config.resid_pdrop)

  def forward(self, hidden_states: Optional[Tuple[torch.FloatTensor]]) -> torch.FloatTensor:
      hidden_states = self.c_fc(hidden_states) # run order: 5 
      hidden_states = self.act(hidden_states)
      hidden_states = self.c_proj(hidden_states)
      hidden_states = self.dropout(hidden_states)
      return hidden_states

class BackpackNoMixBlock(nn.Module):

  def __init__(self, config):
    super().__init__()
    self.ln_1 = nn.LayerNorm(config.n_embd, eps=config.layer_norm_epsilon)
    self.ln_2 = nn.LayerNorm(config.n_embd, eps=config.layer_norm_epsilon)
    self.mlp = BackpackMLP(config.n_embd, config.n_embd*4, config.n_embd, config)
    self.resid_dropout1 = nn.Dropout(config.resid_pdrop)
    self.resid_dropout2 = nn.Dropout(config.resid_pdrop)

  def forward(self, hidden_states, residual):
    residual = self.resid_dropout1(hidden_states) + residual # run order: 4
    hidden_states = self.ln_1(residual)
    mlp_out = self.mlp(hidden_states)
    residual = self.resid_dropout2(mlp_out) + residual
    hidden_states = self.ln_2(residual)
    return hidden_states

class BackpackSenseNetwork(nn.Module):
    def __init__(self, config, num_senses, device=None, dtype=None):
        super().__init__()
        self.num_senses = num_senses # paper: k
        #self.embeddings = embeddings
        self.n_embd = config.n_embd # paper: d

        self.dropout = nn.Dropout(config.embd_pdrop)
        self.block = BackpackNoMixBlock(config)
        self.ln = nn.LayerNorm(self.n_embd, eps=config.layer_norm_epsilon)
        self.final_mlp = BackpackMLP(
            embed_dim=config.n_embd,
            intermediate_dim=config.sense_intermediate_scale*config.n_embd,
            out_dim=config.n_embd*config.num_senses,
            config=config,
            )

    def forward(self, input_embeds):
      residual = self.dropout(input_embeds) # run order: 3
      hidden_states = self.ln(residual)
      hidden_states = self.block(hidden_states, residual)
      senses = self.final_mlp(hidden_states)
      bs, s, nvd = senses.shape
      return senses.reshape(bs, s, self.num_senses, self.n_embd).transpose(1,2) # (bs, nv, s, d)

class BackpackWeightNetwork(nn.Module):

  def __init__(self, num_senses, embed_dim):
    super().__init__()
    self.n_embd = embed_dim
    self.num_senses = num_senses
    self.embed_per_sense = embed_dim // num_senses
    self.c_attn = nn.Linear(embed_dim, 2 * num_senses * self.embed_per_sense)
    self.softmax_scale = None

  def forward(self, encoded):
    b, s, d = encoded.shape
    encoded = self.c_attn(encoded) # (b, s, 2*d)
    encoded = encoded.reshape(b, s, 2, self.num_senses, self.embed_per_sense) #(b, s, 2, nv, d//nv)
    batch_size, seqlen = encoded.shape[0], encoded.shape[1]

    # compute scores & mask
    q, k = encoded.unbind(dim=2)
    softmax_scale = self.softmax_scale or 1.0 / math.sqrt(q.shape[-1])
    scores = torch.einsum('bthd,bshd->bhts', q, k * softmax_scale)
    causal_mask = torch.triu(torch.full((seqlen, seqlen), -10000.0, device=scores.device), 1)
    scores = scores + causal_mask.to(dtype=scores.dtype)

    return torch.softmax(scores, dim=-1, dtype=q.dtype)
  

@dataclass
class BackpackGPT2BaseModelOutput(ModelOutput):
    hidden_states: torch.FloatTensor = None
    contextualization: torch.FloatTensor = None

class BackpackGPT2Model(BackpackGPT2PreTrainedModel):
    _keys_to_ignore_on_load_missing = [r".*attn.masked_bias", r".*attn.bias"]

    def __init__(self, config):
        super().__init__(config)

        self.embed_dim = config.n_embd # paper: d
        self.num_senses = config.num_senses # paper: k

        self.gpt2_model = GPT2Model(config) # paper: parameterizing contextualization weights: Transformer Model
        
        self.sense_network = BackpackSenseNetwork(config, self.num_senses, self.gpt2_model.wte) # paper: parametrizing senses
        self.word_embeddings = self.gpt2_model.wte
        self.position_embeddings = self.gpt2_model.wpe
        self.sense_weight_net = BackpackWeightNetwork(self.num_senses, self.embed_dim) # paper: parameterizing contextualization weights
        # Model parallel
        self.model_parallel = False
        self.device_map = None
        self.gradient_checkpointing = False

    def get_num_senses(self):
        return self.num_senses

    def get_word_embeddings(self):
        return self.word_embeddings

    def get_sense_network(self):
        return self.sense_network

    def forward(self, input_ids, position_ids):
        # Compute senses
        sense_input_embeds = self.word_embeddings(input_ids) # GPT2 word emb without pos # run order: 2
        senses = self.sense_network(sense_input_embeds) # (bs, nv, s, d) (batch size, num toks, num senses, sense dim)

        # Compute contextualization weights
        contextl_hidden_states = self.gpt2_model(input_ids, position_ids=position_ids).last_hidden_state # (bs, s, d)
        contextualization = self.sense_weight_net(contextl_hidden_states) # (bs, nv, s, s)

        # Compute resulting outputs
        hidden_states = torch.sum(contextualization @ senses, dim=1) # (bs, nv, s, d) -> (bs, s, d)
        return BackpackGPT2BaseModelOutput(
            hidden_states=hidden_states, # paper: o
            contextualization=contextualization,
        )
    
    def run_with_custom_contextualization(self, input_ids, contextualization):
        # Compute senses
        sense_input_embeds = self.word_embeddings(input_ids)
        senses = self.sense_network(sense_input_embeds) # (bs, nv, s, d)

        # Compute resulting outputs
        hidden_states = torch.sum(contextualization @ senses, dim=1) # (bs, nv, s, d) -> (bs, s, d)
        return BackpackGPT2BaseModelOutput(
            hidden_states=hidden_states,
            contextualization=contextualization,
        )

@dataclass
class BackpackGPT2LMHeadModelOutput(ModelOutput):
    logits: torch.FloatTensor = None
    contextualization: torch.FloatTensor = None

class BackpackGPT2LMHeadModel(BackpackGPT2PreTrainedModel):
  _keys_to_ignore_on_load_missing = [r".*attn.masked_bias", r".*attn.bias"]

  def __init__(self, config):
    super().__init__(config)
    self.backpack = BackpackGPT2Model(config)
    self.lm_head = nn.Linear(config.n_embd, config.vocab_size, bias=False)

    # Model parallel
    self.model_parallel = False
    self.device_map = None

    self.tie_weights()

  def tie_weights(self):
      self.lm_head.weight = self.backpack.word_embeddings.weight # also tied with the underlying underlying transf

  def get_lm_head(self):
      return self.lm_head

  def forward(self, input_ids, position_ids=None): # run order: 1
      outputs = self.backpack(input_ids, position_ids=position_ids)
      hidden_states, contextualization = outputs.hidden_states, outputs.contextualization
      lm_logits = self.lm_head(hidden_states) # (bs, s, V)
      return BackpackGPT2LMHeadModelOutput(
            logits=lm_logits,
            contextualization=contextualization,
        ) # you should return hidden_states as contextualized emb of input words in the specified sequence
            # use these contextualized embeddings as input to classification head, actually you should write a new class similar to this one but
            # use different output size for lm_head param (call it clf_head)

  def run_with_custom_contextualization(self, input_ids, contextualization):
      outputs = self.backpack.run_with_custom_contextualization(input_ids, contextualization)
      hidden_states, contextualization = outputs.hidden_states, outputs.contextualization
      lm_logits = self.lm_head(hidden_states)
      return BackpackGPT2LMHeadModelOutput(
        logits=lm_logits,
        contextualization=contextualization,
    )