|
from typing import Optional, Tuple, Union, List |
|
|
|
from transformers.models.mpt.modeling_mpt import MptBlock, build_mpt_alibi_tensor |
|
from transformers.modeling_attn_mask_utils import _prepare_4d_causal_attention_mask |
|
import torch |
|
import torch.nn as nn |
|
from torch.nn import CrossEntropyLoss, LayerNorm |
|
from transformers.models.mpt.modeling_mpt import MptBlock, build_mpt_alibi_tensor |
|
from transformers.modeling_outputs import CausalLMOutputWithCrossAttentions, CausalLMOutputWithPast, \ |
|
BaseModelOutputWithPastAndCrossAttentions, BaseModelOutputWithPast |
|
|
|
from transformers import PreTrainedTokenizerFast |
|
import os |
|
import torch.nn.functional as F |
|
|
|
from mpt_7b.modeling_mpt import MPTModel, MPTForCausalLM, gen_attention_mask_in_length |
|
from mpt_7b.configuration_mpt import MPTConfig |
|
from mpt_7b.blocks import MPTBlock |
|
from mpt_7b.norm import NORM_CLASS_REGISTRY |
|
from mpt_7b.custom_embedding import SharedEmbedding |
|
from mpt_7b.attention import ATTN_CLASS_REGISTRY, attn_bias_shape, build_attn_bias, gen_slopes |
|
|
|
import logging |
|
log = logging.getLogger(__name__) |
|
|
|
class Custom_MptModel(MPTModel): |
|
def __init__(self, config: MPTConfig, modality0_dim=128, modality2_dim=1536): |
|
config._validate_config() |
|
super().__init__(config) |
|
self.attn_impl = config.attn_config['attn_impl'] |
|
self.prefix_lm = config.attn_config['prefix_lm'] |
|
self.attn_uses_sequence_id = config.attn_config['attn_uses_sequence_id'] |
|
self.alibi = config.attn_config['alibi'] |
|
self.alibi_bias_max = config.attn_config['alibi_bias_max'] |
|
self.learned_pos_emb = config.learned_pos_emb |
|
if config.init_device == 'mixed': |
|
if dist.get_local_rank() == 0: |
|
config.init_device = 'cpu' |
|
else: |
|
config.init_device = 'meta' |
|
if config.norm_type.lower() not in NORM_CLASS_REGISTRY.keys(): |
|
norm_options = ' | '.join(NORM_CLASS_REGISTRY.keys()) |
|
raise NotImplementedError(f'Requested norm type ({config.norm_type}) is not implemented within this repo (Options: {norm_options}).') |
|
norm_class = NORM_CLASS_REGISTRY[config.norm_type.lower()] |
|
self.embedding_fraction = config.embedding_fraction |
|
self.wte = SharedEmbedding(config.vocab_size, config.d_model, device=config.init_device) |
|
if self.learned_pos_emb: |
|
self.wpe = torch.nn.Embedding(config.max_seq_len, config.d_model, device=config.init_device) |
|
self.emb_drop = nn.Dropout(config.emb_pdrop) |
|
self.blocks = nn.ModuleList([MPTBlock(device=config.init_device, **config.to_dict()) for _ in range(config.n_layers)]) |
|
self.norm_f = norm_class(config.d_model, device=config.init_device) |
|
|
|
|
|
|
|
|
|
for param in self.wte.parameters(): |
|
param.requires_grad = False |
|
|
|
for param in self.blocks.parameters(): |
|
param.requires_grad = False |
|
|
|
|
|
|
|
self.modality0_embedding_projection = nn.ModuleList([nn.Linear(modality0_dim, config.d_model), |
|
|
|
nn.ReLU(), |
|
nn.Linear(config.d_model, config.d_model), |
|
|
|
nn.ReLU(), |
|
nn.Linear(config.d_model, config.d_model)]) |
|
|
|
|
|
self.modality2_embedding_projection = nn.ModuleList([nn.Linear(modality2_dim, config.d_model), |
|
|
|
nn.ReLU(), |
|
nn.Linear(config.d_model, config.d_model), |
|
|
|
nn.ReLU(), |
|
nn.Linear(config.d_model, config.d_model)]) |
|
|
|
|
|
|
|
|
|
self.rope = config.attn_config['rope'] |
|
self.rope_impl = None |
|
if self.rope: |
|
self.rope_impl = config.attn_config['rope_impl'] |
|
self.rotary_embedding = gen_rotary_embedding(rope_head_dim=config.d_model // config.n_heads, rope_impl=self.rope_impl, rope_theta=config.attn_config['rope_theta'], rope_dail_config=config.attn_config['rope_dail_config'], rope_hf_config=config.attn_config['rope_hf_config'], max_seq_len=self.config.max_seq_len) |
|
if config.init_device != 'meta': |
|
log.info(f'We recommend using config.init_device="meta" with Composer + FSDP for faster initialization.') |
|
self.apply(self.param_init_fn) |
|
self.is_causal = not self.prefix_lm |
|
self._attn_bias_initialized = False |
|
self.attn_bias = None |
|
self.attn_bias_shape = attn_bias_shape(self.attn_impl, config.n_heads, config.max_seq_len, self.alibi, prefix_lm=self.prefix_lm, causal=self.is_causal, use_sequence_id=self.attn_uses_sequence_id) |
|
if config.no_bias: |
|
for module in self.modules(): |
|
if hasattr(module, 'bias') and isinstance(module.bias, nn.Parameter): |
|
log.info(f'Removing bias from module={module!r}.') |
|
module.register_parameter('bias', None) |
|
if hasattr(module, 'use_bias'): |
|
log.info(f'Setting use_bias=False for module={module!r}.') |
|
module.use_bias = False |
|
log.debug(self) |
|
log.debug(f"Using {self.config.init_config['name']} initialization.") |
|
|
|
|
|
|
|
|
|
|
|
def get_input_embeddings(self): |
|
return self.wte |
|
|
|
|
|
def set_input_embeddings(self, new_embeddings): |
|
|
|
self.wte.weight = new_embeddings |
|
|
|
|
|
def forward(self, input_ids: Optional[torch.LongTensor]=None, past_key_values: Optional[List[Tuple[torch.FloatTensor]]]=None, |
|
attention_mask: Optional[torch.ByteTensor]=None, prefix_mask: Optional[torch.ByteTensor]=None, |
|
sequence_id: Optional[torch.LongTensor]=None, return_dict: Optional[bool]=None, output_attentions: Optional[bool]=None, |
|
output_hidden_states: Optional[bool]=None, use_cache: Optional[bool]=None, |
|
inputs_embeds: Optional[torch.Tensor]=None, modality0_emb: Optional[bool] = None, |
|
modality0_token_id: Optional[bool] = None, modality1_emb: Optional[bool] = None, modality1_token_id: Optional[bool] = None, |
|
modality2_emb: Optional[bool] = None, modality2_token_id: Optional[bool] = None, modality3_emb: Optional[bool] = None, |
|
modality3_token_id: Optional[bool] = None,) -> BaseModelOutputWithPast: |
|
|
|
return_dict = return_dict if return_dict is not None else self.config.return_dict |
|
use_cache = use_cache if use_cache is not None else self.config.use_cache |
|
if attention_mask is not None: |
|
attention_mask = attention_mask.bool() |
|
if prefix_mask is not None: |
|
prefix_mask = prefix_mask.bool() |
|
if not return_dict: |
|
raise NotImplementedError('return_dict False is not implemented yet for MPT') |
|
if output_attentions: |
|
if self.attn_impl != 'torch': |
|
raise NotImplementedError('output_attentions is not implemented for MPT when using attn_impl `flash` or `triton`.') |
|
if self.training and attention_mask is not None and (attention_mask[:, 0].sum() != attention_mask.shape[0]): |
|
raise NotImplementedError('MPT does not support training with left padding.') |
|
if self.prefix_lm and prefix_mask is None: |
|
raise ValueError('prefix_mask is a required argument when MPT is configured with prefix_lm=True.') |
|
if self.training: |
|
if self.attn_uses_sequence_id and sequence_id is None: |
|
raise ValueError('sequence_id is a required argument when MPT is configured with attn_uses_sequence_id=True ' + 'and the model is in train mode.') |
|
elif self.attn_uses_sequence_id is False and sequence_id is not None: |
|
warnings.warn('MPT received non-None input for `sequence_id` but is configured with attn_uses_sequence_id=False. ' + 'This input will be ignored. If you want the model to use `sequence_id`, set attn_uses_sequence_id to True.') |
|
|
|
|
|
|
|
if modality0_emb is not None: |
|
modality0_emb = torch.tensor(modality0_emb, dtype=torch.bfloat16) |
|
hidden_states = self.wte.weight.detach() |
|
|
|
for layer in self.modality0_embedding_projection: |
|
modality0_emb = layer(modality0_emb) |
|
proj_modality0_emb = modality0_emb |
|
|
|
|
|
hidden_states[modality0_token_id, :] = torch.mean(torch.squeeze(proj_modality0_emb, 1), dim=0) |
|
self.set_input_embeddings(torch.nn.Parameter(hidden_states)) |
|
|
|
if modality1_emb is not None: |
|
modality1_emb = torch.tensor(modality1_emb, dtype=torch.bfloat16) |
|
hidden_states = self.wte.weight.detach() |
|
|
|
for layer in self.modality0_embedding_projection: |
|
modality1_emb = layer(modality1_emb) |
|
proj_modality1_emb = modality1_emb |
|
|
|
|
|
hidden_states[modality1_token_id, :] = torch.mean(torch.squeeze(proj_modality1_emb, 1), dim=0) |
|
self.set_input_embeddings(torch.nn.Parameter(hidden_states)) |
|
|
|
if modality2_emb is not None: |
|
modality2_emb = torch.tensor(modality2_emb, dtype=torch.bfloat16) |
|
hidden_states = self.wte.weight.detach() |
|
|
|
for layer in self.modality2_embedding_projection: |
|
modality2_emb = layer(modality2_emb) |
|
proj_modality2_emb = modality2_emb |
|
|
|
|
|
hidden_states[modality2_token_id, :] = torch.mean(torch.squeeze(proj_modality2_emb, 1), dim=0) |
|
self.set_input_embeddings(torch.nn.Parameter(hidden_states)) |
|
|
|
if modality3_emb is not None: |
|
modality3_emb = torch.tensor(modality3_emb, dtype=torch.bfloat16) |
|
hidden_states = self.wte.weight.detach() |
|
|
|
for layer in self.modality2_embedding_projection: |
|
modality3_emb = layer(modality3_emb) |
|
proj_modality3_emb = modality3_emb |
|
|
|
|
|
hidden_states[modality3_token_id, :] = torch.mean(torch.squeeze(proj_modality3_emb, 1), dim=0) |
|
self.set_input_embeddings(torch.nn.Parameter(hidden_states)) |
|
|
|
|
|
|
|
if input_ids is not None and inputs_embeds is not None: |
|
raise ValueError('You cannot specify both input_ids and inputs_embeds.') |
|
elif input_ids is not None: |
|
bsz = input_ids.size(0) |
|
S = input_ids.size(1) |
|
x = self.wte(input_ids) |
|
input_device = input_ids.device |
|
elif inputs_embeds is not None: |
|
bsz = inputs_embeds.size(0) |
|
S = inputs_embeds.size(1) |
|
x = inputs_embeds |
|
input_device = inputs_embeds.device |
|
else: |
|
raise ValueError('You must specify input_ids or inputs_embeds') |
|
assert S <= self.config.max_seq_len, f'Cannot forward input with seq_len={S}, this model only supports seq_len<={self.config.max_seq_len}' |
|
rotary_emb_w_meta_info = None |
|
past_position = 0 |
|
if past_key_values is not None: |
|
if len(past_key_values) != self.config.n_layers: |
|
raise ValueError(f'past_key_values must provide a past_key_value for each attention ' + f'layer in the network (len(past_key_values)={len(past_key_values)!r}; self.config.n_layers={self.config.n_layers!r}).') |
|
past_position = past_key_values[0][0].size(1) |
|
if self.attn_impl == 'torch': |
|
past_position = past_key_values[0][0].size(3) |
|
if self.learned_pos_emb or self.rope: |
|
if self.learned_pos_emb and S + past_position > self.config.max_seq_len: |
|
raise ValueError(f'Cannot forward input with past sequence length {past_position} and current sequence length ' + f'{S + 1}, this model only supports total sequence length <= {self.config.max_seq_len}.') |
|
if self.learned_pos_emb or (self.rope and self.rope_impl == 'hf'): |
|
pos = torch.arange(past_position, S + past_position, dtype=torch.long, device=input_device).unsqueeze(0) |
|
if attention_mask is not None: |
|
pos = torch.clamp(pos - torch.cumsum((~attention_mask).to(torch.int32), dim=1)[:, past_position:], min=0) |
|
if self.learned_pos_emb: |
|
x = x + self.wpe(pos) |
|
elif self.rope and self.rope_impl == 'hf': |
|
rotary_emb_w_meta_info = {'impl': self.rope_impl, 'rotary_emb': self.rotary_embedding, 'offset_info': pos, 'seq_len': S + past_position} |
|
elif self.rope and self.rope_impl == 'dail': |
|
rotary_emb_w_meta_info = {'impl': self.rope_impl, 'rotary_emb': self.rotary_embedding, 'offset_info': past_position, 'seq_len': S + past_position} |
|
if self.embedding_fraction == 1: |
|
x = self.emb_drop(x) |
|
else: |
|
x_shrunk = x * self.embedding_fraction + x.detach() * (1 - self.embedding_fraction) |
|
assert isinstance(self.emb_drop, nn.Module) |
|
x = self.emb_drop(x_shrunk) |
|
(attn_bias, attention_mask) = self._attn_bias(device=x.device, dtype=torch.float32, attention_mask=attention_mask, prefix_mask=prefix_mask, sequence_id=sequence_id) |
|
attention_mask_in_length = gen_attention_mask_in_length(sequence_id=sequence_id, S=S, attn_uses_sequence_id=self.attn_uses_sequence_id, attn_impl=self.attn_impl, attention_mask=attention_mask) |
|
alibi_slopes = None |
|
if self.alibi and self.attn_impl == 'flash': |
|
alibi_slopes = gen_slopes(n_heads=self.config.n_heads, alibi_bias_max=self.alibi_bias_max, device=x.device, return_1d=True) |
|
|
|
presents = () if use_cache else None |
|
if use_cache and past_key_values is None: |
|
past_key_values = [() for _ in range(self.config.n_layers)] |
|
all_hidden_states = () if output_hidden_states else None |
|
all_self_attns = () if output_attentions else None |
|
flash_attn_padding_info = {} |
|
if self.attn_impl == 'flash': |
|
flash_attn_padding_info = gen_flash_attn_padding_info(bsz, S, past_position, x.device, attention_mask_in_length, attention_mask) |
|
for (b_idx, block) in enumerate(self.blocks): |
|
if output_hidden_states: |
|
assert all_hidden_states is not None |
|
all_hidden_states = all_hidden_states + (x,) |
|
past_key_value = past_key_values[b_idx] if past_key_values is not None else None |
|
(x, attn_weights, present) = block(x, past_key_value=past_key_value, attn_bias=attn_bias, rotary_emb_w_meta_info=rotary_emb_w_meta_info, attention_mask=attention_mask, is_causal=self.is_causal, output_attentions=bool(output_attentions), alibi_slopes=alibi_slopes, flash_attn_padding_info=flash_attn_padding_info) |
|
if presents is not None: |
|
presents += (present,) |
|
if output_attentions: |
|
assert all_self_attns is not None |
|
all_self_attns = all_self_attns + (attn_weights,) |
|
x = self.norm_f(x) |
|
if output_hidden_states: |
|
assert all_hidden_states is not None |
|
all_hidden_states = all_hidden_states + (x,) |
|
return BaseModelOutputWithPast(last_hidden_state=x, past_key_values=presents, hidden_states=all_hidden_states, attentions=all_self_attns) |
|
|
|
|
|
class Custom_MPTForCausalLM(MPTForCausalLM): |
|
|
|
def __init__(self, config: MPTConfig): |
|
super().__init__(config) |
|
|
|
self.transformer: MPTModel = Custom_MptModel(config) |
|
self.lm_head = None |
|
if not config.tie_word_embeddings: |
|
self.lm_head = nn.Linear(config.d_model, config.vocab_size, bias=False, device=config.init_device) |
|
self.lm_head._fsdp_wrap = True |
|
for child in self.transformer.children(): |
|
if isinstance(child, torch.nn.ModuleList): |
|
continue |
|
if isinstance(child, torch.nn.Module): |
|
child._fsdp_wrap = True |
|
self.logit_scale = None |
|
if config.logit_scale is not None: |
|
logit_scale = config.logit_scale |
|
if isinstance(logit_scale, str): |
|
if logit_scale == 'inv_sqrt_d_model': |
|
logit_scale = 1 / math.sqrt(config.d_model) |
|
else: |
|
raise ValueError(f"logit_scale={logit_scale!r} is not recognized as an option; use numeric value or 'inv_sqrt_d_model'.") |
|
self.logit_scale = logit_scale |
|
|
|
def forward(self, input_ids: Optional[torch.LongTensor]=None, past_key_values: Optional[List[Tuple[torch.FloatTensor]]]=None, |
|
attention_mask: Optional[torch.ByteTensor]=None, prefix_mask: Optional[torch.ByteTensor]=None, |
|
sequence_id: Optional[torch.LongTensor]=None, labels: Optional[torch.LongTensor]=None, |
|
return_dict: Optional[bool]=None, output_attentions: Optional[bool]=None, output_hidden_states: Optional[bool]=None, |
|
use_cache: Optional[bool]=None, inputs_embeds: Optional[torch.FloatTensor]=None, |
|
modality0_emb: Optional[bool] = None, modality0_token_id: Optional[bool] = None, |
|
modality1_emb: Optional[bool] = None, modality1_token_id: Optional[bool] = None, |
|
modality2_emb: Optional[bool] = None, modality2_token_id: Optional[bool] = None, |
|
modality3_emb: Optional[bool] = None, modality3_token_id: Optional[bool] = None) -> CausalLMOutputWithPast: |
|
return_dict = return_dict if return_dict is not None else self.config.return_dict |
|
use_cache = use_cache if use_cache is not None else self.config.use_cache |
|
outputs = self.transformer( |
|
input_ids=input_ids, past_key_values=past_key_values, attention_mask=attention_mask, prefix_mask=prefix_mask, |
|
sequence_id=sequence_id, return_dict=return_dict, output_attentions=output_attentions, output_hidden_states=output_hidden_states, |
|
use_cache=use_cache, inputs_embeds=inputs_embeds, |
|
modality0_emb=modality0_emb, |
|
modality0_token_id=modality0_token_id, |
|
modality1_emb=modality1_emb, |
|
modality1_token_id=modality1_token_id, |
|
modality2_emb=modality2_emb, |
|
modality2_token_id=modality2_token_id, |
|
modality3_emb=modality3_emb, |
|
modality3_token_id=modality3_token_id |
|
) |
|
if self.lm_head is not None: |
|
logits = self.lm_head(outputs.last_hidden_state) |
|
else: |
|
out = outputs.last_hidden_state |
|
out = out.to(self.transformer.wte.weight.device) |
|
logits = self.transformer.wte(out, True) |
|
if self.logit_scale is not None: |
|
if self.logit_scale == 0: |
|
warnings.warn(f'Multiplying logits by self.logit_scale={self.logit_scale!r}. This will produce uniform (uninformative) outputs.') |
|
logits *= self.logit_scale |
|
loss = None |
|
if labels is not None: |
|
_labels = torch.roll(labels, shifts=-1) |
|
_labels[:, -1] = -100 |
|
loss = F.cross_entropy(logits.view(-1, logits.size(-1)), _labels.to(logits.device).view(-1)) |
|
return CausalLMOutputWithPast(loss=loss, logits=logits, past_key_values=outputs.past_key_values, hidden_states=outputs.hidden_states, attentions=outputs.attentions) |
|
|