Fix _prepare_4d_causal_attention_mask_for_sdpa
Browse files- modeling_plamo.py +98 -2
modeling_plamo.py
CHANGED
@@ -6,13 +6,109 @@ from torch import nn
|
|
6 |
from torch.nn import functional as F
|
7 |
from transformers import AutoTokenizer, PretrainedConfig, PreTrainedModel
|
8 |
from transformers.modeling_attn_mask_utils import (
|
|
|
9 |
_prepare_4d_causal_attention_mask,
|
10 |
-
_prepare_4d_causal_attention_mask_for_sdpa,
|
11 |
)
|
12 |
from transformers.modeling_outputs import BaseModelOutputWithPast
|
13 |
from transformers.tokenization_utils_base import BatchEncoding
|
14 |
|
15 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
def _swiglu(h: torch.Tensor) -> torch.Tensor:
|
17 |
h0, h1 = h.chunk(2, dim=-1)
|
18 |
return torch.nn.functional.silu(h0) * h1
|
@@ -817,7 +913,7 @@ class ModifiedAttention(Attention):
|
|
817 |
|
818 |
|
819 |
PLAMO_ATTENTION_CLASSES = {
|
820 |
-
"sdpa":
|
821 |
}
|
822 |
|
823 |
|
|
|
6 |
from torch.nn import functional as F
|
7 |
from transformers import AutoTokenizer, PretrainedConfig, PreTrainedModel
|
8 |
from transformers.modeling_attn_mask_utils import (
|
9 |
+
AttentionMaskConverter,
|
10 |
_prepare_4d_causal_attention_mask,
|
|
|
11 |
)
|
12 |
from transformers.modeling_outputs import BaseModelOutputWithPast
|
13 |
from transformers.tokenization_utils_base import BatchEncoding
|
14 |
|
15 |
|
16 |
+
# From: https://github.com/McGill-NLP/llm2vec/blob/main/llm2vec/models/attn_mask_utils.py
|
17 |
+
def _prepare_4d_causal_attention_mask_for_sdpa(
|
18 |
+
attention_mask: Optional[torch.Tensor],
|
19 |
+
input_shape: Union[torch.Size, Tuple, List],
|
20 |
+
inputs_embeds: torch.Tensor,
|
21 |
+
past_key_values_length: int,
|
22 |
+
sliding_window: Optional[int] = None,
|
23 |
+
):
|
24 |
+
"""
|
25 |
+
Prepares the correct `attn_mask` argument to be used by `torch.nn.functional.scaled_dot_product_attention`.
|
26 |
+
|
27 |
+
In case no token is masked in the `attention_mask` argument, we simply set it to `None` for the cases `query_length == 1` and
|
28 |
+
`key_value_length == query_length`, and rely instead on SDPA `is_causal` argument to use causal/non-causal masks,
|
29 |
+
allowing to dispatch to the flash attention kernel (that can otherwise not be used if a custom `attn_mask` is passed).
|
30 |
+
"""
|
31 |
+
attn_mask_converter = AttentionMaskConverter(
|
32 |
+
is_causal=False, sliding_window=sliding_window
|
33 |
+
) # is_causal=True in original implementation
|
34 |
+
|
35 |
+
key_value_length = input_shape[-1] + past_key_values_length
|
36 |
+
batch_size, query_length = input_shape
|
37 |
+
|
38 |
+
# torch.jit.trace, symbolic_trace and torchdynamo with fullgraph=True are unable to capture the controlflow `is_causal=attention_mask is None and q_len > 1`
|
39 |
+
# used as an SDPA argument. We keep compatibility with these tracing tools by always using SDPA's `attn_mask` argument in case we are tracing.
|
40 |
+
# TODO: For dynamo, rather use a check on fullgraph=True once this is possible (https://github.com/pytorch/pytorch/pull/120400).
|
41 |
+
is_tracing = (
|
42 |
+
torch.jit.is_tracing()
|
43 |
+
or isinstance(inputs_embeds, torch.fx.Proxy)
|
44 |
+
or (hasattr(torch, "_dynamo") and torch._dynamo.is_compiling())
|
45 |
+
)
|
46 |
+
|
47 |
+
if attention_mask is not None:
|
48 |
+
# 4d mask is passed through
|
49 |
+
if len(attention_mask.shape) == 4:
|
50 |
+
expected_shape = (input_shape[0], 1, input_shape[1], key_value_length)
|
51 |
+
if tuple(attention_mask.shape) != expected_shape:
|
52 |
+
raise ValueError(
|
53 |
+
f"Incorrect 4D attention_mask shape: {tuple(attention_mask.shape)}; expected: {expected_shape}."
|
54 |
+
)
|
55 |
+
else:
|
56 |
+
# if the 4D mask has correct shape - invert it and fill with negative infinity
|
57 |
+
inverted_mask = 1.0 - attention_mask.to(inputs_embeds.dtype)
|
58 |
+
attention_mask = inverted_mask.masked_fill(
|
59 |
+
inverted_mask.to(torch.bool), torch.finfo(inputs_embeds.dtype).min
|
60 |
+
)
|
61 |
+
return attention_mask
|
62 |
+
|
63 |
+
elif not is_tracing and torch.all(attention_mask == 1):
|
64 |
+
if query_length == 1:
|
65 |
+
# For query_length == 1, causal attention and bi-directional attention are the same.
|
66 |
+
attention_mask = None
|
67 |
+
elif key_value_length == query_length:
|
68 |
+
attention_mask = None
|
69 |
+
else:
|
70 |
+
# Unfortunately, for query_length > 1 and key_value_length != query_length, we cannot generally ignore the attention mask, as SDPA causal mask generation
|
71 |
+
# may be wrong. We will set `is_causal=False` in SDPA and rely on Transformers attention_mask instead, hence not setting it to None here.
|
72 |
+
# Reference: https://github.com/pytorch/pytorch/issues/108108
|
73 |
+
pass
|
74 |
+
elif query_length > 1 and key_value_length != query_length:
|
75 |
+
# See the comment above (https://github.com/pytorch/pytorch/issues/108108).
|
76 |
+
# Ugly: we set it to True here to dispatch in the following controlflow to `to_causal_4d`.
|
77 |
+
attention_mask = True
|
78 |
+
elif is_tracing:
|
79 |
+
raise ValueError(
|
80 |
+
'Attention using SDPA can not be traced with torch.jit.trace when no attention_mask is provided. To solve this issue, please either load your model with the argument `attn_implementation="eager"` or pass an attention_mask input when tracing the model.'
|
81 |
+
)
|
82 |
+
|
83 |
+
if attention_mask is None:
|
84 |
+
expanded_4d_mask = None
|
85 |
+
elif attention_mask is True:
|
86 |
+
expanded_4d_mask = attn_mask_converter.to_causal_4d(
|
87 |
+
input_shape[0],
|
88 |
+
input_shape[-1],
|
89 |
+
key_value_length,
|
90 |
+
dtype=inputs_embeds.dtype,
|
91 |
+
device=inputs_embeds.device,
|
92 |
+
)
|
93 |
+
else:
|
94 |
+
expanded_4d_mask = attn_mask_converter.to_4d(
|
95 |
+
attention_mask,
|
96 |
+
input_shape[-1],
|
97 |
+
dtype=inputs_embeds.dtype,
|
98 |
+
key_value_length=key_value_length,
|
99 |
+
)
|
100 |
+
|
101 |
+
# Attend to all tokens in masked rows from the causal_mask, for example the relevant first rows when
|
102 |
+
# using left padding. This is required by F.scaled_dot_product_attention memory-efficient attention path.
|
103 |
+
# Details: https://github.com/pytorch/pytorch/issues/110213
|
104 |
+
if not is_tracing and expanded_4d_mask.device.type == "cuda":
|
105 |
+
expanded_4d_mask = AttentionMaskConverter._unmask_unattended(
|
106 |
+
expanded_4d_mask, min_dtype=torch.finfo(inputs_embeds.dtype).min
|
107 |
+
)
|
108 |
+
|
109 |
+
return expanded_4d_mask
|
110 |
+
|
111 |
+
|
112 |
def _swiglu(h: torch.Tensor) -> torch.Tensor:
|
113 |
h0, h1 = h.chunk(2, dim=-1)
|
114 |
return torch.nn.functional.silu(h0) * h1
|
|
|
913 |
|
914 |
|
915 |
PLAMO_ATTENTION_CLASSES = {
|
916 |
+
"sdpa": ModifiedAttention,
|
917 |
}
|
918 |
|
919 |
|