Skip to content

Commit

Permalink
refactor baichuan, glm4 and minicpm3 (#12600)
Browse files Browse the repository at this point in the history
  • Loading branch information
@MeouSker77
MeouSker77 authored Dec 24, 2024
1 parent c410d9c commit 7aaf02f
Show file tree
Hide file tree
Showing 4 changed files with 32 additions and 167 deletions.
42 changes: 8 additions & 34 deletions python/llm/src/ipex_llm/transformers/models/baichuan.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -24,16 +24,16 @@
import torch.utils.checkpoint
from torch.nn import functional as F
from transformers.modeling_outputs import BaseModelOutputWithPast
from ipex_llm.transformers.models.common import scaled_dot_product_attention
from ipex_llm.transformers.models.utils import use_quantize_kv_cache, restore_fp8_kv_cache, \
should_use_compresskv, get_compresskv_attn_mask
should_use_compresskv
from ipex_llm.transformers.models.utils import update_past_key_value
from ipex_llm.transformers.models.utils import should_use_fuse_rope
from ipex_llm.transformers.models.utils import use_flash_attention, use_sdp, use_sdp_causal
from ipex_llm.transformers.models.utils import use_flash_attention, use_sdp
from ipex_llm.transformers.models.utils import apply_rotary_pos_emb, SILU
from ipex_llm.transformers.models.utils import mlp_fusion_check
from ipex_llm.transformers.models.utils import is_enough_kv_cache_room_4_36
from ipex_llm.transformers.kv import DynamicCompressFp8Cache, DynamicCompressCache
from ipex_llm.transformers.models.utils import extend_kv_cache, append_kv_cache
import warnings
import os

Expand Down Expand Up @@ -301,42 +301,16 @@ def baichuan_attention_forward_7b(

# IPEX-LLM OPT: sdp
attn_weights = None
if not self.training and not hidden_states.requires_grad and \
use_flash_attention(query_states, key_states, attention_mask):
if use_flash_attention(query_states, key_states, attention_mask):
attn_output = F.scaled_dot_product_attention(query_states.to(dtype=torch.float16),
key_states.to(dtype=torch.float16),
value_states.to(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 use_compresskv:
attention_mask = get_compresskv_attn_mask(key_states, attention_mask)
if use_quantize_kv:
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 use_quantize_kv:
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 use_quantize_kv:
key_states, value_states = restore_fp8_kv_cache(key_states, value_states,
query_states.dtype)
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(value_states.dtype)
attn_output = torch.matmul(attn_weights, value_states)
attn_output = scaled_dot_product_attention(
query_states, key_states, value_states,
attention_mask, q_len == kv_seq_len
)

attn_output = attn_output.transpose(1, 2).contiguous()
attn_output = attn_output.reshape(bsz, q_len, self.hidden_size)
Expand Down
58 changes: 9 additions & 49 deletions python/llm/src/ipex_llm/transformers/models/chatglm4.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -20,15 +20,14 @@
import os
import torch
from typing import Optional, Tuple, Union
from ipex_llm.transformers.models.utils import restore_fp8_kv_cache, update_past_key_value
from ipex_llm.transformers.models.utils import use_quantize_kv_cache, use_sdp, \
use_sdp_causal, should_use_compresskv, is_enough_kv_cache_room_4_36, \
get_compresskv_attn_mask
from ipex_llm.transformers.models.common import scaled_dot_product_attention
from ipex_llm.transformers.models.utils import update_past_key_value
from ipex_llm.transformers.models.utils import use_quantize_kv_cache
from ipex_llm.transformers.models.utils import should_use_compresskv, is_enough_kv_cache_room_4_36
from ipex_llm.transformers.models.utils import should_use_fuse_rope, apply_rotary_pos_emb
from ipex_llm.transformers.models.chatglm2 import repeat_kv
from ipex_llm.transformers.kv import DynamicCompressCache, DynamicCompressFp8Cache
from transformers.modeling_outputs import BaseModelOutputWithPast
import math


KV_CACHE_ALLOC_BLOCK_LENGTH = int(os.environ.get("KV_CACHE_ALLOC_BLOCK_LENGTH", 256))

Expand Down Expand Up @@ -241,49 +240,10 @@ def chatglm4_attention_forward(
past_key_value = None

# IPEX-LLM OPT: sdp
attn_weights = None
if use_sdp(q_len, kv_seq_len, head_dim, query_states):
import xe_addons
if use_compresskv:
attention_mask = get_compresskv_attn_mask(key_states, attention_mask)
if use_quantize_kv:
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, head_dim, query_states, self.training):
import xe_addons
if use_quantize_kv:
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)
elif query_states.device.type == "cpu":
# repeat k/v heads if n_kv_heads < n_heads
key_states = repeat_kv(key_states, n_head // n_kv_head)
value_states = repeat_kv(value_states, n_head // n_kv_head)
if q_len == kv_seq_len:
attn_output = torch.nn.functional.scaled_dot_product_attention(
query_states, key_states, value_states, is_causal=True
)
else:
attn_output = torch.nn.functional.scaled_dot_product_attention(
query_states, key_states, value_states, attention_mask
)
else:
if use_quantize_kv:
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, n_head // n_kv_head)
value_states = repeat_kv(value_states, n_head // n_kv_head)
attn_weights = torch.matmul(query_states / math.sqrt(head_dim),
key_states.transpose(2, 3))
if attention_mask is not None:
attn_weights = attn_weights + attention_mask
attn_weights = torch.nn.functional.softmax(attn_weights, dim=-1,
dtype=torch.float32).to(value_states.dtype)
attn_output = torch.matmul(attn_weights, value_states)
attn_output = scaled_dot_product_attention(
query_states, key_states, value_states,
attention_mask, q_len == kv_seq_len
)

# context_layer's shape: [bsz, n_head, seq_len, head_dim] -> [seq_len, bsz, n_head * head_dim]
attn_output = attn_output.transpose(1, 2).contiguous().view(bsz, q_len, n_head * head_dim)
Expand Down
57 changes: 7 additions & 50 deletions python/llm/src/ipex_llm/transformers/models/chatglm4v.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -20,10 +20,10 @@
import torch
from typing import Optional, Tuple, Union
from ipex_llm.transformers.models.common import merge_qkv_base
from ipex_llm.transformers.models.utils import restore_fp8_kv_cache, update_past_key_value
from ipex_llm.transformers.models.utils import use_quantize_kv_cache, use_sdp, use_sdp_causal
from ipex_llm.transformers.models.common import scaled_dot_product_attention
from ipex_llm.transformers.models.utils import update_past_key_value
from ipex_llm.transformers.models.utils import use_quantize_kv_cache, use_sdp
from ipex_llm.transformers.models.utils import should_use_fuse_rope, apply_rotary_pos_emb
from ipex_llm.transformers.models.chatglm2 import repeat_kv
from ipex_llm.utils.common import invalidInputError
from transformers.modeling_outputs import BaseModelOutputWithPast
import math
Expand Down Expand Up @@ -246,53 +246,10 @@ def chatglm4v_attention_forward(
past_key_value = None

# IPEX-LLM OPT: sdp
attn_weights = None
if use_sdp(q_len, kv_seq_len, head_dim, query_states):
import xe_addons
if use_quantize_kv:
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, head_dim, query_states, self.training):
import xe_addons
if use_quantize_kv:
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)
elif query_states.device.type == "cpu":
# repeat k/v heads if n_kv_heads < n_heads
key_states = repeat_kv(key_states, n_head // n_kv_head)
value_states = repeat_kv(value_states, n_head // n_kv_head)
if q_len == kv_seq_len:
attn_output = torch.nn.functional.scaled_dot_product_attention(
query_states, key_states, value_states, is_causal=True
)
else:
attn_output = torch.nn.functional.scaled_dot_product_attention(
query_states, key_states, value_states, attention_mask
)
else:
if use_quantize_kv:
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, n_head // n_kv_head)
value_states = repeat_kv(value_states, n_head // n_kv_head)
attn_weights = torch.matmul(query_states / math.sqrt(head_dim),
key_states.transpose(2, 3))
if attention_mask is not None:
attn_weights = attn_weights + attention_mask
if kv_seq_len >= 2048 or bsz >= 64:
# for memory considerations, do not upcast attention to fp32
# for long sequences or large batches
attn_weights = torch.nn.functional.softmax(attn_weights, dim=-1)
else:
# upcast attention to fp32
attn_weights = torch.nn.functional.softmax(attn_weights, dim=-1,
dtype=torch.float32).to(value_states.dtype)
attn_output = torch.matmul(attn_weights, value_states)
attn_output = scaled_dot_product_attention(
query_states, key_states, value_states,
attention_mask, q_len == kv_seq_len
)

# context_layer's shape: [bsz, n_head, seq_len, head_dim] -> [seq_len, bsz, n_head * head_dim]
attn_output = attn_output.transpose(1, 2).contiguous().view(bsz, q_len, n_head * head_dim)
Expand Down
42 changes: 8 additions & 34 deletions python/llm/src/ipex_llm/transformers/models/minicpm3.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -6,10 +6,10 @@
from transformers.cache_utils import Cache

from ipex_llm.utils.common.log4Error import invalidInputError
from ipex_llm.transformers.models.common import scaled_dot_product_attention
from ipex_llm.transformers.models.utils import should_use_fuse_rope
from ipex_llm.transformers.models.utils import rotate_half
from ipex_llm.transformers.models.utils import use_sdp, use_sdp_causal
from ipex_llm.transformers.models.utils import use_quantize_kv_cache, restore_fp8_kv_cache
from ipex_llm.transformers.models.utils import use_quantize_kv_cache
from ipex_llm.transformers.kv import DynamicNormalCache, DynamicFp8Cache


Expand All @@ -25,7 +25,7 @@ def pre_compute_inv_freq(module: torch.nn.Module):

def padding_v_head_dim(module: torch.nn.Module):
if module.__class__.__name__ == "MiniCPMAttention":
k_head_dim = module.qk_rope_head_dim + module.qk_nope_head_dim
k_head_dim = module.q_head_dim
v_head_dim = module.v_head_dim
invalidInputError(k_head_dim >= v_head_dim,
f"unsupported k_head_dim and v_head_dim: {k_head_dim} {v_head_dim}")
Expand Down Expand Up @@ -183,37 +183,11 @@ def minicpm3_attention_forward(
self.layer_idx, None)

attn_weights = None
if use_sdp(q_len, kv_seq_len, self.q_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)
attn_output = attn_output[:, :, :, :self.v_head_dim]
elif use_sdp_causal(q_len, kv_seq_len, self.q_head_dim, query_states, False):
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)
attn_output = attn_output[:, :, :, :self.v_head_dim]
else:
if isinstance(past_key_value, DynamicFp8Cache):
key_states, value_states = restore_fp8_kv_cache(key_states, value_states,
query_states.dtype)
attn_weights = torch.matmul(query_states, key_states.transpose(2, 3)) * self.softmax_scale

if attention_mask is not None:
attn_weights = attn_weights + attention_mask

# upcast attention to fp32
attn_weights = nn.functional.softmax(attn_weights,
dim=-1, dtype=torch.float32).to(query_states.dtype)
attn_output = torch.matmul(attn_weights, value_states[:, :, :, :self.v_head_dim])
attn_output = scaled_dot_product_attention(
query_states, key_states, value_states,
attention_mask, q_len == kv_seq_len, self.softmax_scale
)
attn_output = attn_output[:, :, :, :self.v_head_dim]

attn_output = attn_output.transpose(1, 2).contiguous()

Expand Down

0 comments on commit 7aaf02f

Please sign in to comment.