-
Notifications
You must be signed in to change notification settings - Fork 1.3k
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
1 parent
2b8ad87
commit b9c6699
Showing
4 changed files
with
234 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
157 changes: 157 additions & 0 deletions
157
python/llm/src/ipex_llm/transformers/npu_models/phi3.py
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,157 @@ | ||
| # | ||
| # 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/v4.40.0/src/transformers/models/llama/modeling_llama.py | ||
| # which is licensed under Apache License 2.0: | ||
| # | ||
| # Copyright 2021 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. | ||
|
|
||
|
|
||
| from typing import Optional, Tuple, List | ||
| import torch | ||
| from torch import nn | ||
| import math | ||
| import importlib | ||
| from transformers.cache_utils import Cache | ||
| from ipex_llm.utils.common.log4Error import invalidInputError | ||
|
|
||
|
|
||
| def phi3_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, | ||
| ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: | ||
| modeling_module_name = self.__class__.__module__ | ||
| module = importlib.import_module(modeling_module_name) | ||
| apply_rotary_pos_emb, repeat_kv = module.apply_rotary_pos_emb, module.repeat_kv | ||
| bsz, q_len, _ = hidden_states.size() | ||
|
|
||
| qkv = self.qkv_proj(hidden_states) | ||
| query_pos = self.num_heads * self.head_dim | ||
| query_states = qkv[..., :query_pos] | ||
| key_states = qkv[..., query_pos:query_pos + self.num_key_value_heads * self.head_dim] | ||
| value_states = qkv[..., query_pos + self.num_key_value_heads * self.head_dim:] | ||
|
|
||
| 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: | ||
| invalidInputError( | ||
| False, | ||
| f"The cache structure has changed since version v4.36." | ||
| f"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, position_ids, 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) | ||
|
|
||
| if attention_mask is not None: | ||
| causal_mask = attention_mask[:, :, :, : key_states.shape[-2]] | ||
| else: | ||
| causal_mask = None | ||
|
|
||
| if query_states.size(2) == key_states.size(2): | ||
| # first token | ||
| from intel_npu_acceleration_library.functional import scaled_dot_product_attention | ||
| attn_output = scaled_dot_product_attention( | ||
| query_states, | ||
| key_states, | ||
| value_states, | ||
| attn_mask=attention_mask, | ||
| is_causal=self.is_causal and causal_mask is None and q_len > 1, | ||
| ) | ||
| attn_weights = None | ||
| else: | ||
|
|
||
| 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): | ||
| invalidInputError( | ||
| False, | ||
| f"Attention weights should be of" | ||
| f"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): | ||
| invalidInputError( | ||
| False, | ||
| 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(value_states.dtype) | ||
| attn_weights = nn.functional.dropout(attn_weights, | ||
| p=self.attention_dropout, training=self.training) | ||
|
|
||
| attn_output = torch.matmul(attn_weights, value_states) | ||
|
|
||
| if attn_output.size() != (bsz, self.num_heads, q_len, self.head_dim): | ||
| invalidInputError( | ||
| False, | ||
| 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) | ||
|
|
||
| attn_output = self.o_proj(attn_output) | ||
|
|
||
| if not output_attentions: | ||
| attn_weights = None | ||
|
|
||
| return attn_output, attn_weights, past_key_value |