stablelm-base-alpha-3b-v2 / modeling_stablelm_alpha.py
Jonathan Tow
refactor: clean nn package access
997d959
# coding=utf-8
# Copyright 2023 Stability AI, EleutherAI, and The HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
# This code is based off the following work:
# https://github.com/huggingface/transformers/blob/main/src/transformers/models/llama/modeling_llama.py
# https://github.com/huggingface/transformers/blob/main/src/transformers/models/gpt_neox/modeling_gpt_neox.py
""" PyTorch StableLM-Alpha model. """
from typing import Optional, Tuple, Union
import math
import torch
import torch.utils.checkpoint
from torch import nn
from torch.nn import CrossEntropyLoss
from transformers.modeling_outputs import (
BaseModelOutputWithPast,
CausalLMOutputWithPast,
)
from transformers.modeling_utils import PreTrainedModel
from transformers.utils import logging
from .configuration_stablelm_alpha import StableLMAlphaConfig
logger = logging.get_logger(__name__)
def _expand_mask(mask: torch.Tensor, dtype: torch.dtype, tgt_len: Optional[int] = None):
"""Expands attention_mask from `[batch_size, seq_len]` to `[batch_size, 1, tgt_seq_len, src_seq_len]`."""
batch_size, src_len = mask.size()
tgt_len = tgt_len if tgt_len is not None else src_len
expanded_mask = mask[:, None, None, :].expand(batch_size, 1, tgt_len, src_len).to(dtype)
inverted_mask = 1.0 - expanded_mask
return inverted_mask.masked_fill(inverted_mask.to(torch.bool), torch.finfo(dtype).min)
class LayerNorm(nn.LayerNorm):
def __init__(self, normalized_shape: torch.Size, bias: bool = True, **kwargs):
r"""
bias (`bool`, default = True): whether to use the bias term.
"""
super().__init__(normalized_shape, **kwargs)
if not bias:
self.bias = None
class DecoderLayer(nn.Module):
def __init__(self, config: StableLMAlphaConfig):
super().__init__()
self.norm = LayerNorm(config.hidden_size, eps=config.norm_eps)
self.attention = Attention(config)
self.mlp = MLP(config)
def forward(
self,
hidden_states: Optional[torch.FloatTensor],
attention_mask: Optional[torch.FloatTensor] = None,
position_ids: Optional[torch.LongTensor] = None,
past_key_value: Optional[Tuple[torch.Tensor]] = None,
output_attentions: Optional[bool] = False,
use_cache: Optional[bool] = False,
) -> Union[Tuple[torch.Tensor], Optional[Tuple[torch.Tensor, Tuple[torch.FloatTensor, ...]]]]:
residual = hidden_states
# Pre-Norm
hidden_states = self.norm(hidden_states)
# Self-Attention
attn_output, attn_weights, present_key_value = self.attention(
hidden_states=hidden_states,
attention_mask=attention_mask,
position_ids=position_ids,
past_key_value=past_key_value,
use_cache=use_cache,
output_attentions=output_attentions,
)
# Feed-forward
mlp_output = self.mlp(hidden_states)
hidden_states = residual + attn_output + mlp_output
outputs = (hidden_states,)
if output_attentions:
outputs += (attn_weights,)
if use_cache:
outputs += (present_key_value,)
return outputs # hidden_states, (optional: attn_weights), (optional: present_key_value)
class MLP(nn.Module):
def __init__(self, config: StableLMAlphaConfig):
super().__init__()
hidden_size = config.hidden_size
multiple_of = 256
ff_dim = int(8 * hidden_size / 3)
intermediate_size = multiple_of * ((ff_dim + multiple_of - 1) // multiple_of)
self.gate_proj = nn.Linear(hidden_size, 2 * intermediate_size, bias=False)
self.out_proj = nn.Linear(intermediate_size, hidden_size, bias=False)
self.act = nn.SiLU()
def forward(self, x: torch.Tensor) -> torch.Tensor:
ff, ff_gate = self.gate_proj(x).chunk(2, dim=-1)
return self.out_proj(ff * self.act(ff_gate))
class RotaryEmbedding(nn.Module):
def __init__(
self,
dim: int,
max_position_embeddings: int,
base: int = 10_000,
device: Optional[torch.device] = None,
):
super().__init__()
self.dim = dim
self.max_position_embeddings = max_position_embeddings
self.base = base
inv_freq = 1.0 / (self.base ** (torch.arange(0, self.dim, 2, device=device, dtype=torch.float32) / self.dim))
self.register_buffer("inv_freq", inv_freq, persistent=False)
# Build here to make `torch.jit.trace` work.
self._set_cos_sin_cache(
seq_len=max_position_embeddings, device=self.inv_freq.device, dtype=torch.get_default_dtype()
)
def _set_cos_sin_cache(self, seq_len: int, device: torch.device, dtype: torch.dtype):
self.max_seq_len_cached = seq_len
t = torch.arange(self.max_seq_len_cached, device=device, dtype=self.inv_freq.dtype)
freqs = torch.outer(t, self.inv_freq)
# Different from paper, but it uses a different permutation in order to obtain the same calculation
emb = torch.cat((freqs, freqs), dim=-1)
self.register_buffer("cos_cached", emb.cos()[None, None, :, :].to(dtype), persistent=False)
self.register_buffer("sin_cached", emb.sin()[None, None, :, :].to(dtype), persistent=False)
def forward(self, x: torch.Tensor, seq_len: Optional[int] = None):
# x: [batch_size, num_heads, seq_len, head_size]
if seq_len > self.max_seq_len_cached:
self._set_cos_sin_cache(seq_len=seq_len, device=x.device, dtype=torch.get_default_dtype())
return (
self.cos_cached[:, :, :seq_len, ...].to(dtype=x.dtype),
self.sin_cached[:, :, :seq_len, ...].to(dtype=x.dtype),
)
def rotate_half(x: torch.Tensor):
"""Rotates half the hidden dims of the input."""
x1, x2 = torch.chunk(x, 2, dim=-1)
return torch.cat((-x2, x1), dim=-1)
def apply_rotary_pos_emb(q, k, cos, sin, position_ids):
# The first two dimensions of cos and sin are always 1, so we can `squeeze` them.
cos = cos.squeeze(1).squeeze(0) # [seq_len, dim]
sin = sin.squeeze(1).squeeze(0) # [seq_len, dim]
cos = cos[position_ids].unsqueeze(1) # [batch_size, 1, seq_len, dim]
sin = sin[position_ids].unsqueeze(1) # [batch_size, 1, seq_len, dim]
q_embed = (q * cos) + (rotate_half(q) * sin)
k_embed = (k * cos) + (rotate_half(k) * sin)
return q_embed, k_embed
class Attention(nn.Module):
def __init__(self, config: StableLMAlphaConfig):
super().__init__()
self.config = config
self.hidden_size = config.hidden_size
self.num_heads = config.num_heads
self.head_dim = self.hidden_size // self.num_heads
self.max_position_embeddings = config.max_position_embeddings
if self.hidden_size % self.num_heads != 0:
raise ValueError(
"`hidden_size` is not divisble by the number of attention heads! Make sure to update them"
)
self.qkv_proj = nn.Linear(self.hidden_size, 3 * self.hidden_size, bias=False)
self.out_proj = nn.Linear(self.hidden_size, self.hidden_size, bias=False)
self._init_rope()
def _init_rope(self):
self.rotary_ndims = int(self.head_dim * self.config.rotary_pct)
self.rotary_emb = RotaryEmbedding(
self.rotary_ndims,
max_position_embeddings=self.config.max_position_embeddings,
base=self.config.rotary_emb_base,
)
def forward(
self,
hidden_states: torch.FloatTensor,
attention_mask: torch.FloatTensor,
position_ids: torch.LongTensor,
past_key_value: Optional[Tuple[torch.Tensor]] = None,
output_attentions: Optional[bool] = False,
use_cache: Optional[bool] = False,
) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
has_past_key_value = past_key_value is not None
# Compute QKV
# [batch_size, seq_len, (num_heads * 3 * head_dim)]
qkv = self.qkv_proj(hidden_states)
# [batch_size, seq_len, num_heads, 3 * head_dim]
new_qkv_shape = qkv.size()[:-1] + (self.num_heads, 3 * self.head_dim)
qkv = qkv.view(*new_qkv_shape)
# 3 * [batch_size, num_heads, seq_len, head_dim]
query = qkv[..., : self.head_dim].permute(0, 2, 1, 3)
key = qkv[..., self.head_dim:(2 * self.head_dim)].permute(0, 2, 1, 3)
value = qkv[..., (2 * self.head_dim):].permute(0, 2, 1, 3)
# Compute rotary embeddings on rotary_ndims
# [batch_size, num_heads, seq_len, rotary_ndims]
query_rot = query[..., :self.rotary_ndims]
query_pass = query[..., self.rotary_ndims:]
key_rot = key[..., :self.rotary_ndims]
key_pass = key[..., self.rotary_ndims:]
# Compute token offset for rotary embeddings (when decoding)
kv_seq_len = key.shape[-2]
if has_past_key_value:
kv_seq_len += past_key_value[0].shape[-2]
# Add rotary embeddings to query and key
cos, sin = self.rotary_emb(value, seq_len=kv_seq_len)
query, key = apply_rotary_pos_emb(query_rot, key_rot, cos, sin, position_ids)
# Concatenate rotary embeddings with pass-through query and key
# [batch_size, num_heads, seq_len, head_dim]
query = torch.cat((query, query_pass), dim=-1)
key = torch.cat((key, key_pass), dim=-1)
# Reuse past key-value states
if has_past_key_value:
key = torch.cat((past_key_value[0], key), dim=2)
value = torch.cat((past_key_value[1], value), dim=2)
present_key_value = (key, value) if use_cache else None
# [batch_size, num_heads, seq_len, head_dim]
query = query.transpose(1, 2).contiguous()
key = key.transpose(1, 2).contiguous()
value = value.transpose(1, 2).contiguous()
# Compute attention
softmax_scale = 1 / math.sqrt(self.head_dim)
attn_scores = torch.einsum('bthd,bshd->bhts', query, key * softmax_scale)
# Apply the attention mask
if attention_mask is not None:
attn_scores = attn_scores + attention_mask
attn_weights = nn.functional.softmax(attn_scores, dim=-1, dtype=torch.float32).to(query.dtype)
attn_output = torch.einsum('bhts,bshd->bthd', attn_weights, value)
# Merge heads
attn_output = attn_output.reshape(attn_output.shape[0], attn_output.shape[1], -1)
# Final linear projection
attn_output = self.out_proj(attn_output)
if not output_attentions:
attn_weights = None
return attn_output, attn_weights, present_key_value
def attention_mask_func(attention_scores: torch.Tensor, ltor_mask: torch.Tensor):
attention_scores.masked_fill_(~ltor_mask, torch.finfo(attention_scores.dtype).min)
return attention_scores
class StableLMAlphaPreTrainedModel(PreTrainedModel):
"""An abstract class to handle weights initialization and a simple interface
for downloading and loading pretrained models.
"""
config_class = StableLMAlphaConfig
base_model_prefix = "transformer"
supports_gradient_checkpointing = True
_no_split_modules = ["DecoderLayer"]
_skip_keys_device_placement = "past_key_values"
def _init_weights(self, module: nn.Module):
"""Initialize the weights"""
if isinstance(module, nn.Linear):
module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
if module.bias is not None:
module.bias.data.zero_()
elif isinstance(module, nn.Embedding):
module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
if module.padding_idx is not None:
module.weight.data[module.padding_idx].zero_()
elif isinstance(module, nn.LayerNorm):
module.bias.data.zero_()
module.weight.data.fill_(1.0)
def _set_gradient_checkpointing(self, module: nn.Module, value=False):
if isinstance(module, StableLMAlphaModel):
module.gradient_checkpointing = value
def _make_causal_mask(
input_ids_shape: torch.Size,
dtype: torch.dtype,
device: torch.device,
past_key_values_length: int = 0
):
"""Make causal mask used for bi-directional self-attention."""
batch_size, tgt_len = input_ids_shape
mask = torch.full((tgt_len, tgt_len), torch.finfo(torch.float16).min, device=device)
mask_cond = torch.arange(mask.size(-1), device=device)
mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
mask = mask.to(dtype)
if past_key_values_length > 0:
mask = torch.cat([torch.zeros(tgt_len, past_key_values_length, dtype=dtype, device=device), mask], dim=-1)
return mask[None, None, :, :].expand(batch_size, 1, tgt_len, tgt_len + past_key_values_length)
class StableLMAlphaModel(StableLMAlphaPreTrainedModel):
def __init__(self, config: StableLMAlphaConfig):
super().__init__(config)
self.config = config
self.embed = nn.Embedding(config.vocab_size, config.hidden_size)
self.layers = nn.ModuleList([DecoderLayer(config) for _ in range(config.num_hidden_layers)])
self.final_norm = LayerNorm(config.hidden_size, eps=config.norm_eps)
self.gradient_checkpointing = False
self.post_init()
def get_input_embeddings(self):
return self.embed
def set_input_embeddings(self, value: nn.Module):
self.embed = value
# Copied from transformers.models.bart.modeling_bart.BartDecoder._prepare_decoder_attention_mask
def _prepare_decoder_attention_mask(
self,
attention_mask: torch.Tensor,
input_shape: torch.Size,
inputs_embeds: torch.Tensor,
past_key_values_length: int,
):
# Create causal mask
# [batch_size, seq_len] -> [batch_size, 1, tgt_seq_len, src_seq_len]
combined_attention_mask = None
if input_shape[-1] > 1:
combined_attention_mask = _make_causal_mask(
input_shape,
inputs_embeds.dtype,
device=inputs_embeds.device,
past_key_values_length=past_key_values_length,
)
if attention_mask is not None:
# [batch_size, seq_len] -> [batch_size, 1, tgt_seq_len, src_seq_len]
expanded_attn_mask = _expand_mask(attention_mask, inputs_embeds.dtype, tgt_len=input_shape[-1]).to(
inputs_embeds.device
)
combined_attention_mask = (
expanded_attn_mask if combined_attention_mask is None else expanded_attn_mask + combined_attention_mask
)
return combined_attention_mask
def forward(
self,
input_ids: Optional[torch.LongTensor] = None,
attention_mask: Optional[torch.FloatTensor] = None,
position_ids: Optional[torch.LongTensor] = None,
inputs_embeds: Optional[torch.FloatTensor] = None,
past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None,
use_cache: Optional[bool] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
) -> Union[Tuple, BaseModelOutputWithPast]:
r"""
past_key_values (`tuple(tuple(torch.FloatTensor))` of length `config.n_layers`
with each tuple having 4 tensors of shape `(batch_size, num_heads, sequence_length - 1, embed_size_per_head)`):
Contains precomputed key and value hidden states of the attention blocks.
Can be used to speed up decoding. If `past_key_values` are used, the user
can optionally input only the last `decoder_input_ids` (those that don't
have their past key value states given to this model) of shape `(batch_size, 1)`
instead of all `decoder_input_ids` of shape `(batch_size, sequence_length)`.
use_cache (`bool`, *optional*):
If set to `True`, `past_key_values` key value states are returned and
can be used to speed up decoding (see `past_key_values`).
"""
output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
output_hidden_states = (
output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
)
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
use_cache = use_cache if use_cache is not None else self.config.use_cache
if input_ids is not None and inputs_embeds is not None:
raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
elif input_ids is not None:
input_shape = input_ids.size()
elif inputs_embeds is not None:
input_shape = inputs_embeds.size()[:-1]
else:
raise ValueError("You have to specify either input_ids or inputs_embeds")
batch_size, seq_length = input_shape
if past_key_values is None:
past_key_values_length = 0
past_key_values = tuple([None] * self.config.num_hidden_layers)
seq_length_with_past = seq_length
else:
past_key_values_length = past_key_values[0][0].shape[2]
seq_length_with_past = seq_length + past_key_values_length
if position_ids is None:
device = input_ids.device if input_ids is not None else inputs_embeds.device
position_ids = torch.arange(past_key_values_length, seq_length + past_key_values_length, dtype=torch.long, device=device)
position_ids = position_ids.unsqueeze(0).view(-1, seq_length)
else:
position_ids = position_ids.view(-1, seq_length).long()
if inputs_embeds is None:
inputs_embeds = self.embed(input_ids)
# Attention mask.
if attention_mask is None:
attention_mask = torch.ones(
(batch_size, seq_length_with_past), dtype=torch.bool, device=inputs_embeds.device
)
attention_mask = self._prepare_decoder_attention_mask(
attention_mask, (batch_size, seq_length), inputs_embeds, past_key_values_length
)
hidden_states = inputs_embeds
if self.gradient_checkpointing and self.training:
if use_cache:
logger.warning(
"`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..."
)
use_cache = False
all_hidden_states = () if output_hidden_states else None
all_attentions = () if output_attentions else None
present_key_values = () if use_cache else None
for _, (decoder_layer, past_key_value) in enumerate(zip(self.layers, past_key_values)):
if output_hidden_states:
all_hidden_states = all_hidden_states + (hidden_states,)
if self.gradient_checkpointing and self.training:
def create_custom_forward(module):
def custom_forward(*inputs):
# `None` for `use_cache`
return module(*inputs, output_attentions, None)
return custom_forward
outputs = torch.utils.checkpoint.checkpoint(
create_custom_forward(decoder_layer),
hidden_states,
attention_mask,
position_ids,
# `None` for `past_key_value`
None,
)
else:
outputs = decoder_layer(
hidden_states,
attention_mask=attention_mask,
position_ids=position_ids,
past_key_value=past_key_value,
output_attentions=output_attentions,
use_cache=use_cache,
)
hidden_states = outputs[0]
if output_attentions:
all_attentions = all_attentions + (outputs[1],)
if use_cache:
present_key_values += (outputs[2 if output_attentions else 1],)
hidden_states = self.final_norm(hidden_states)
# Add last hidden state
if output_hidden_states:
all_hidden_states += (hidden_states,)
present_key_values = present_key_values if use_cache else None
if not return_dict:
return tuple(v for v in [hidden_states, present_key_values, all_hidden_states, all_attentions] if v is not None)
return BaseModelOutputWithPast(
last_hidden_state=hidden_states,
past_key_values=present_key_values,
hidden_states=all_hidden_states,
attentions=all_attentions,
)
class StableLMAlphaForCausalLM(StableLMAlphaPreTrainedModel):
_tied_weights_keys = ["lm_head.weight"]
def __init__(self, config: StableLMAlphaConfig):
super().__init__(config)
self.transformer = StableLMAlphaModel(config)
self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
self.post_init()
def get_output_embeddings(self):
return self.lm_head
def set_output_embeddings(self, new_embeddings: nn.Module):
self.lm_head = new_embeddings
def forward(
self,
input_ids: Optional[torch.LongTensor] = None,
attention_mask: Optional[torch.FloatTensor] = None,
position_ids: Optional[torch.LongTensor] = None,
inputs_embeds: Optional[torch.FloatTensor] = None,
past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None,
labels: Optional[torch.LongTensor] = None,
use_cache: Optional[bool] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
) -> Union[Tuple, CausalLMOutputWithPast]:
r"""
Example:
```python
>>> from transformers import AutoTokenizer, StableLMAlphaForCausalLM, StableLMAlphaConfig
>>> import torch
>>> tokenizer = AutoTokenizer.from_pretrained("stabilityai/stablelm-base-alpha-3b-v2", trust_remote_code=True)
>>> config = StableLMAlphaConfig.from_pretrained("stabilityai/stablelm-base-alpha-3b-v2")
>>> config.is_decoder = True
>>> model = StableLMAlphaForCausalLM.from_pretrained("stabilityai/stablelm-base-alpha-3b-v2", config=config)
>>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
>>> outputs = model(**inputs)
>>> logits = outputs.logits
```
"""
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
outputs = self.transformer(
input_ids,
attention_mask=attention_mask,
position_ids=position_ids,
inputs_embeds=inputs_embeds,
past_key_values=past_key_values,
use_cache=use_cache,
output_attentions=output_attentions,
output_hidden_states=output_hidden_states,
return_dict=return_dict,
)
hidden_states = outputs[0]
logits = self.lm_head(hidden_states)
lm_loss = None
if labels is not None:
# move labels to correct device to enable model parallelism
labels = labels.to(logits.device)
# we are doing next-token prediction; shift prediction scores and input ids by one
shift_logits = logits[:, :-1, :].contiguous()
labels = labels[:, 1:].contiguous()
loss_fct = CrossEntropyLoss()
lm_loss = loss_fct(shift_logits.view(-1, shift_logits.size(-1)), labels.view(-1))
if not return_dict:
output = (logits,) + outputs[1:]
return ((lm_loss,) + output) if lm_loss is not None else output
return CausalLMOutputWithPast(
loss=lm_loss,
logits=logits,
past_key_values=outputs.past_key_values,
hidden_states=outputs.hidden_states,
attentions=outputs.attentions,
)
def prepare_inputs_for_generation(
self,
input_ids,
past_key_values: Optional[torch.Tensor] = None,
attention_mask: Optional[torch.Tensor] = None,
inputs_embeds: Optional[torch.Tensor] = None,
**kwargs
):
# Cut decoder_input_ids if past is used
if past_key_values and past_key_values[0] is not None:
input_ids = input_ids[:, -1:]
position_ids = kwargs.get("position_ids", None)
if attention_mask is not None and position_ids is None:
# Create position_ids on the fly for batch generation
position_ids = attention_mask.long().cumsum(-1) - 1
position_ids.masked_fill_(attention_mask == 0, 1)
if past_key_values:
position_ids = position_ids[:, -1].unsqueeze(-1)
# If `inputs_embeds` are passed, we only want to use them in the 1st generation step
if inputs_embeds is not None and past_key_values is None:
model_inputs = {"inputs_embeds": inputs_embeds}
else:
model_inputs = {"input_ids": input_ids}
model_inputs.update(
{
"attention_mask": attention_mask,
"past_key_values": past_key_values,
"position_ids": position_ids,
}
)
return model_inputs
def _reorder_cache(self, past_key_values: torch.Tensor, beam_idx: int):
reordered_past = ()
for past_key_value in past_key_values:
reordered_past += (
tuple(past_state.index_select(0, beam_idx) for past_state in past_key_value[:2]) + past_key_value[2:],
)
return reordered_past
StableLMAlphaConfig.register_for_auto_class()
StableLMAlphaForCausalLM.register_for_auto_class("AutoModelForCausalLM")