sglang_v0.5.2/flashinfer_0.3.1/tests/jit_utils.py

229 lines
6.2 KiB
Python

"""
Copyright (c) 2023 by FlashInfer team.
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 itertools
import torch
import flashinfer
from flashinfer.jit import JitSpec
from flashinfer.utils import is_fa3_backend_supported, is_sm90a_supported
def gen_decode_attention_modules(
q_dtypes,
kv_dtypes,
head_dims,
pos_encoding_modes,
use_sliding_window_options,
use_logits_soft_cap_options,
) -> list[JitSpec]:
jit_specs: list[JitSpec] = []
for (
q_dtype,
kv_dtype,
head_dim,
pos_encoding_mode,
use_sliding_window,
use_logits_soft_cap,
) in itertools.product(
q_dtypes,
kv_dtypes,
head_dims,
pos_encoding_modes,
use_sliding_window_options,
use_logits_soft_cap_options,
):
if q_dtype != kv_dtype:
if kv_dtype.itemsize > 1:
continue # skip fp16/bf16 mixed precision
jit_specs.append(
flashinfer.decode.gen_single_decode_module(
q_dtype,
kv_dtype,
q_dtype,
head_dim, # head_dim_qk
head_dim, # head_dim_vo
pos_encoding_mode,
use_sliding_window,
use_logits_soft_cap,
)
)
jit_specs.append(
flashinfer.decode.gen_batch_decode_module(
q_dtype,
kv_dtype,
q_dtype,
torch.int32,
head_dim, # head_dim_qk
head_dim, # head_dim_vo
pos_encoding_mode,
use_sliding_window,
use_logits_soft_cap,
)
)
return jit_specs
def gen_persistent_batch_attention_modules(
q_dtypes,
kv_dtypes,
head_dims,
use_logits_soft_cap_options,
) -> list[JitSpec]:
jit_specs: list[JitSpec] = []
for (
q_dtype,
kv_dtype,
head_dim,
use_logits_soft_cap,
) in itertools.product(
q_dtypes,
kv_dtypes,
head_dims,
use_logits_soft_cap_options,
):
if q_dtype != kv_dtype:
if kv_dtype.itemsize > 1:
continue # skip fp16/bf16 mixed precision
jit_specs.append(
flashinfer.attention.gen_batch_attention_module(
q_dtype,
kv_dtype,
q_dtype,
torch.int32,
head_dim, # head_dim_qk
head_dim, # head_dim_vo
0, # pos_encoding_mode
use_logits_soft_cap,
False, # use_profiler
)
)
return jit_specs
def gen_prefill_attention_modules(
q_dtypes,
kv_dtypes,
head_dims,
pos_encoding_modes,
use_sliding_window_options,
use_logits_soft_cap_options,
use_fp16_qk_reduction_options,
) -> list[JitSpec]:
jit_specs: list[JitSpec] = []
for (
q_dtype,
kv_dtype,
head_dim,
pos_encoding_mode,
use_sliding_window,
use_logits_soft_cap,
use_fp16_qk_reduction,
) in itertools.product(
q_dtypes,
kv_dtypes,
head_dims,
pos_encoding_modes,
use_sliding_window_options,
use_logits_soft_cap_options,
use_fp16_qk_reduction_options,
):
if q_dtype != kv_dtype:
if kv_dtype.itemsize > 1:
continue # skip fp16/bf16 mixed precision
if is_sm90a_supported(torch.device("cuda")) and is_fa3_backend_supported(
pos_encoding_mode,
use_fp16_qk_reduction,
use_custom_mask=False,
dtype_q=q_dtype,
dtype_kv=kv_dtype,
):
jit_specs.append(
flashinfer.prefill.gen_single_prefill_module(
"fa3",
q_dtype,
kv_dtype,
q_dtype,
head_dim, # head_dim_qk
head_dim, # head_dim_vo
pos_encoding_mode,
use_sliding_window,
use_logits_soft_cap,
use_fp16_qk_reduction,
)
)
jit_specs.append(
flashinfer.prefill.gen_batch_prefill_module(
"fa3",
q_dtype,
kv_dtype,
q_dtype,
torch.int32,
head_dim, # head_dim_qk
head_dim, # head_dim_vo
pos_encoding_mode,
use_sliding_window,
use_logits_soft_cap,
use_fp16_qk_reduction,
)
)
jit_specs.append(
flashinfer.prefill.gen_single_prefill_module(
"fa2",
q_dtype,
kv_dtype,
q_dtype,
head_dim, # head_dim_qk
head_dim, # head_dim_vo
pos_encoding_mode,
use_sliding_window,
use_logits_soft_cap,
use_fp16_qk_reduction,
)
)
jit_specs.append(
flashinfer.prefill.gen_batch_prefill_module(
"fa2",
q_dtype,
kv_dtype,
q_dtype,
torch.int32,
head_dim, # head_dim_qk
head_dim, # head_dim_vo
pos_encoding_mode,
use_sliding_window,
use_logits_soft_cap,
use_fp16_qk_reduction,
)
)
# required for attention with custom mask
jit_specs.append(flashinfer.quantization.gen_quantization_module())
jit_specs.append(flashinfer.page.gen_page_module())
return jit_specs