ipex-llm/python/llm/src/ipex_llm/transformers/models/qwen2.py
2024-06-26 16:52:29 +08:00

521 lines
22 KiB
Python

#
# Copyright 2016 The BigDL Authors.
#
# 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.
#
# Some parts of this file is adapted from
# https://github.com/huggingface/transformers/blob/v4.37.0/src/transformers/models/qwen2/modeling_qwen2.py
# which is licensed under Apache License 2.0:
#
# Copyright 2024 The Qwen team, Alibaba Group and the HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original forms to accommodate minor architectural differences compared
# to GPT-NeoX and OPT used by the Meta AI team that trained the model.
#
# 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.
#
import math
from typing import Optional, Tuple, Union, List
import torch
from torch.nn import CrossEntropyLoss
from torch.nn.functional import scaled_dot_product_attention as sdpa
from ipex_llm.transformers.models.utils import should_use_fuse_rope
from ipex_llm.transformers.models.utils import use_quantize_kv_cache, restore_fp8_kv_cache
from ipex_llm.transformers.models.utils import use_flash_attention, use_sdp, use_sdp_causal
from ipex_llm.transformers.kv import DynamicFp8Cache, DynamicNormalCache
from ipex_llm.utils.common import invalidInputError
from transformers.models.qwen2.modeling_qwen2 import Qwen2Attention, Qwen2MLP
from transformers.models.qwen2.modeling_qwen2 import apply_rotary_pos_emb, repeat_kv
from transformers.models.qwen2.modeling_qwen2 import _prepare_4d_causal_attention_mask_for_sdpa
from transformers.models.qwen2.modeling_qwen2 import _prepare_4d_causal_attention_mask
from transformers.modeling_outputs import BaseModelOutputWithPast, CausalLMOutputWithPast
from transformers.cache_utils import Cache, DynamicCache
from transformers import logging
logger = logging.get_logger(__name__)
def qwen2_model_forward(
self,
input_ids: torch.LongTensor = None,
attention_mask: Optional[torch.Tensor] = None,
position_ids: Optional[torch.LongTensor] = None,
past_key_values: Optional[List[torch.FloatTensor]] = None,
inputs_embeds: Optional[torch.FloatTensor] = None,
use_cache: Optional[bool] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
):
use_cache = use_cache if use_cache is not None else self.config.use_cache
input = input_ids if input_ids is not None else inputs_embeds
use_quantize_kv = (
self.config.hidden_size != 3584 # disable quantize kv in specific model
and use_quantize_kv_cache(self.layers[0].mlp.up_proj, input)
)
if use_cache:
if use_quantize_kv and not isinstance(past_key_values, DynamicFp8Cache):
past_key_values = DynamicFp8Cache.from_legacy_cache(past_key_values)
if not use_quantize_kv and not isinstance(past_key_values, DynamicNormalCache):
past_key_values = DynamicNormalCache.from_legacy_cache(past_key_values)
return qwen2_model_forward_internal(
self=self,
input_ids=input_ids,
attention_mask=attention_mask,
position_ids=position_ids,
past_key_values=past_key_values,
inputs_embeds=inputs_embeds,
use_cache=use_cache,
output_attentions=output_attentions,
output_hidden_states=output_hidden_states,
return_dict=return_dict,
)
def qwen2_model_forward_internal(
self,
input_ids: torch.LongTensor = None,
attention_mask: Optional[torch.Tensor] = None,
position_ids: Optional[torch.LongTensor] = None,
past_key_values: Optional[List[torch.FloatTensor]] = None,
inputs_embeds: Optional[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]:
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
)
use_cache = use_cache if use_cache is not None else self.config.use_cache
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
# retrieve input_ids and inputs_embeds
if input_ids is not None and inputs_embeds is not None:
invalidInputError(False,
"You cannot specify both decoder_input_ids and "
"decoder_inputs_embeds at the same time")
elif input_ids is not None:
batch_size, seq_length = input_ids.shape
elif inputs_embeds is not None:
batch_size, seq_length, _ = inputs_embeds.shape
else:
invalidInputError(False,
"You have to specify either decoder_input_ids or decoder_inputs_embeds")
if self.gradient_checkpointing and self.training:
if use_cache:
logger.warning_once(
"`use_cache=True` is incompatible with gradient checkpointing. "
"Setting `use_cache=False`..."
)
use_cache = False
past_key_values_length = 0
if use_cache:
use_legacy_cache = not isinstance(past_key_values, Cache)
if use_legacy_cache:
past_key_values = DynamicCache.from_legacy_cache(past_key_values)
past_key_values_length = past_key_values.get_usable_length(seq_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_tokens(input_ids)
flash_attn_2 = self._attn_implementation == "flash_attention_2"
if attention_mask is not None and flash_attn_2 and use_cache:
is_padding_right = attention_mask[:, -1].sum().item() != batch_size
if is_padding_right:
invalidInputError(
False,
"You are attempting to perform batched generation with padding_side='right'"
" this may lead to unexpected behaviour for Flash Attention version of Qwen2."
" Make sure to call `tokenizer.padding_side = 'left'` before tokenizing "
"the input. "
)
if self._attn_implementation == "flash_attention_2":
# 2d mask is passed through the layers
attention_mask = attention_mask if (attention_mask is not None and
0 in attention_mask) else None
elif self._attn_implementation == "sdpa" and not output_attentions:
# output_attentions=True can not be supported when using SDPA, and we fall back on
# the manual implementation that requires a 4D causal mask in all cases.
attention_mask = _prepare_4d_causal_attention_mask_for_sdpa(
attention_mask,
(batch_size, seq_length),
inputs_embeds,
past_key_values_length,
)
else:
# 4d mask is passed through the layers
attention_mask = _prepare_4d_causal_attention_mask(
attention_mask,
(batch_size, seq_length),
inputs_embeds,
past_key_values_length,
sliding_window=self.config.sliding_window,
)
hidden_states = inputs_embeds
# decoder layers
all_hidden_states = () if output_hidden_states else None
all_self_attns = () if output_attentions else None
next_decoder_cache = None
for decoder_layer in self.layers:
if output_hidden_states:
all_hidden_states += (hidden_states,)
if self.gradient_checkpointing and self.training:
layer_outputs = self._gradient_checkpointing_func(
decoder_layer.__call__,
hidden_states,
attention_mask,
position_ids,
past_key_values,
output_attentions,
use_cache,
)
else:
# ipex-llm changes
curr_device = decoder_layer.input_layernorm.weight.device
if attention_mask is not None:
attention_mask = attention_mask.to(curr_device)
if position_ids is not None:
position_ids = position_ids.to(curr_device)
# ipex-llm changes end
layer_outputs = decoder_layer(
hidden_states,
attention_mask=attention_mask,
position_ids=position_ids,
past_key_value=past_key_values,
output_attentions=output_attentions,
use_cache=use_cache,
)
hidden_states = layer_outputs[0]
if use_cache:
next_decoder_cache = layer_outputs[2 if output_attentions else 1]
if output_attentions:
all_self_attns += (layer_outputs[1],)
hidden_states = self.norm(hidden_states)
# add hidden states from the last decoder layer
if output_hidden_states:
all_hidden_states += (hidden_states,)
next_cache = None
if use_cache:
next_cache = next_decoder_cache.to_legacy_cache() if use_legacy_cache else \
next_decoder_cache
if not return_dict:
return tuple(v for v in [hidden_states, next_cache,
all_hidden_states, all_self_attns] if v is not None)
return BaseModelOutputWithPast(
last_hidden_state=hidden_states,
past_key_values=next_cache,
hidden_states=all_hidden_states,
attentions=all_self_attns,
)
def qwen2_causal_lm_forward(
self,
input_ids: torch.LongTensor = None,
attention_mask: Optional[torch.Tensor] = None,
position_ids: Optional[torch.LongTensor] = None,
past_key_values: Optional[List[torch.FloatTensor]] = None,
inputs_embeds: Optional[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]:
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
# decoder outputs consists of (dec_features, layer_state, dec_hidden, dec_attn)
outputs = self.model(
input_ids=input_ids,
attention_mask=attention_mask,
position_ids=position_ids,
past_key_values=past_key_values,
inputs_embeds=inputs_embeds,
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)
# ipex-llm changes start: remove `logits.float()` to reduce memory usage with long input
# logits = logits.float()
# ipex-llm changes end
loss = None
if labels is not None:
# Shift so that tokens < n predict n
shift_logits = logits[..., :-1, :].contiguous()
shift_labels = labels[..., 1:].contiguous()
# Flatten the tokens
loss_fct = CrossEntropyLoss()
shift_logits = shift_logits.view(-1, self.config.vocab_size)
shift_labels = shift_labels.view(-1)
# Enable model parallelism
shift_labels = shift_labels.to(shift_logits.device)
loss = loss_fct(shift_logits, shift_labels)
if not return_dict:
output = (logits,) + outputs[1:]
return (loss,) + output if loss is not None else output
return CausalLMOutputWithPast(
loss=loss,
logits=logits,
past_key_values=outputs.past_key_values,
hidden_states=outputs.hidden_states,
attentions=outputs.attentions,
)
def merge_qkv(module: torch.nn.Module):
if isinstance(module, Qwen2Attention):
new_weight = torch.cat([
module.q_proj.weight.data,
module.k_proj.weight.data,
module.v_proj.weight.data,
], dim=0)
new_bias = torch.cat([
module.q_proj.bias.data,
module.k_proj.bias.data,
module.v_proj.bias.data,
], dim=-1)
qkv_proj = torch.nn.Linear(0, 0, bias=True)
qkv_proj.weight = torch.nn.Parameter(new_weight, requires_grad=False)
qkv_proj.bias = torch.nn.Parameter(new_bias, requires_grad=False)
qkv_proj.in_features = new_weight.size(1)
qkv_proj.out_features = new_weight.size(0)
module.qkv_proj = qkv_proj
del module.q_proj, module.k_proj, module.v_proj
# Qwen2 uses pre-computed rope table to accelerate rope,
# original `cos_cached` and `sin_cached` are added by `register_buffer`,
# so they will move to xpu during `model.to('xpu')`.
# But gpu fuse kernel doesn't need this rope table, only cpu needs them,
# so delete them then add them with `=`, so that they will be pinned on CPU,
# this can save about 0.5GB gpu memory usage when running Qwen2
if hasattr(module.rotary_emb, "cos_cached"):
cos_cached = module.rotary_emb.cos_cached
del module.rotary_emb.cos_cached
module.rotary_emb.cos_cached = cos_cached
if hasattr(module.rotary_emb, "sin_cached"):
sin_cached = module.rotary_emb.sin_cached
del module.rotary_emb.sin_cached
module.rotary_emb.sin_cached = sin_cached
def padding_mlp(module: torch.nn.Module):
# for qwen 1.5 14B
if isinstance(module, Qwen2MLP):
hidden_size = module.gate_proj.weight.shape[1]
intermediate_size = module.gate_proj.weight.shape[0]
padding_intermediate_size = (intermediate_size + 256 - 1) // 256 * 256
if intermediate_size % 256 == 0:
return
gate_weight = module.gate_proj.weight.data
new_gate_weight = torch.zeros([padding_intermediate_size, hidden_size],
dtype=gate_weight.dtype, device=gate_weight.device)
new_gate_weight[:intermediate_size, :] = gate_weight
if hasattr(module.gate_proj, 'out_features'):
module.gate_proj.out_features = padding_intermediate_size
module.gate_proj.weight = torch.nn.Parameter(new_gate_weight, requires_grad=False)
up_weight = module.up_proj.weight.data
new_up_weight = torch.zeros([padding_intermediate_size, hidden_size],
dtype=up_weight.dtype, device=up_weight.device)
new_up_weight[:intermediate_size, :] = up_weight
if hasattr(module.gate_proj, 'out_features'):
module.up_proj.out_features = padding_intermediate_size
module.up_proj.weight = torch.nn.Parameter(new_up_weight, requires_grad=False)
down_weight = module.down_proj.weight.data
new_down_weight = torch.zeros([hidden_size, padding_intermediate_size],
dtype=down_weight.dtype, device=down_weight.device)
new_down_weight[:, :intermediate_size] = down_weight
if hasattr(module.gate_proj, 'out_features'):
module.down_proj.in_features = padding_intermediate_size
module.down_proj.weight = torch.nn.Parameter(new_down_weight, requires_grad=False)
def qwen2_attention_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]]]:
bsz, q_len, _ = hidden_states.size()
device = hidden_states.device
if hasattr(self, 'qkv_proj') and self.qkv_proj is not None:
qkv = self.qkv_proj(hidden_states)
qkv = qkv.view(bsz, q_len, self.num_heads + 2 * self.num_key_value_heads, self.head_dim)
qkv = qkv.transpose(1, 2)
query_states, key_states, value_states = qkv.split([self.num_heads,
self.num_key_value_heads,
self.num_key_value_heads], dim=1)
else:
# when quant_method is 'gptq'
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:
kv_seq_len += past_key_value.get_usable_length(kv_seq_len, self.layer_idx)
if should_use_fuse_rope(hidden_states, position_ids, self.training):
import xe_addons
xe_addons.rotary_half_inplaced(self.rotary_emb.inv_freq, position_ids,
query_states, key_states)
else:
cos, sin = self.rotary_emb(value_states, seq_len=kv_seq_len)
cos, sin = cos.to(device), sin.to(device)
query_states, key_states = apply_rotary_pos_emb(query_states, key_states,
cos, sin, position_ids)
if past_key_value is not None:
key_states, value_states = past_key_value.update(key_states, value_states,
self.layer_idx, None)
attn_weights = None
if query_states.device.type == "cpu":
# 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_output = sdpa(query_states,
key_states,
value_states,
attn_mask=attention_mask,
dropout_p=self.attention_dropout if self.training else 0.0,
is_causal=self.is_causal and attention_mask is None and q_len > 1)
elif not self.training and not hidden_states.requires_grad and \
use_flash_attention(query_states, key_states, attention_mask):
# 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_output = sdpa(query_states.to(device, dtype=torch.float16),
key_states.to(device, dtype=torch.float16),
value_states.to(device, dtype=torch.float16),
is_causal=True).to(hidden_states.dtype)
elif use_sdp(q_len, kv_seq_len, self.head_dim, query_states):
import xe_addons
if isinstance(past_key_value, DynamicFp8Cache):
attn_output = xe_addons.sdp_fp8(query_states, key_states, value_states,
attention_mask)
else:
attn_output = xe_addons.sdp(query_states, key_states, value_states,
attention_mask)
elif use_sdp_causal(q_len, kv_seq_len, self.head_dim, query_states, self.training):
import xe_addons
if isinstance(past_key_value, DynamicFp8Cache):
attn_output = xe_addons.sdp_fp8_causal(query_states, key_states,
value_states, attention_mask)
else:
attn_output = xe_addons.sdp_causal(query_states, key_states,
value_states, attention_mask)
else:
if isinstance(past_key_value, DynamicFp8Cache):
key_states, value_states = restore_fp8_kv_cache(key_states, value_states,
query_states.dtype)
# 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 attention_mask is not None:
attn_weights = attn_weights + attention_mask
# upcast attention to fp32
attn_weights = torch.nn.functional.softmax(attn_weights, dim=-1,
dtype=torch.float32).to(query_states.dtype)
attn_weights = torch.nn.functional.dropout(attn_weights, p=self.attention_dropout,
training=self.training)
attn_output = torch.matmul(attn_weights, value_states)
attn_output = attn_output.transpose(1, 2).contiguous()
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