Files
FastDeploy/tests/operators/test_w4afp8_gemm.py
2025-12-22 13:39:41 +08:00

140 lines
5.6 KiB
Python

# Copyright (c) 2025 PaddlePaddle Authors. 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.
import unittest
import numpy as np
import paddle
from fastdeploy.model_executor.ops.gpu import (
w4afp8_gemm,
w4afp8_gemm_scale_permute,
w4afp8_gemm_weight_convert,
)
class TestW4AFP8GEMM(unittest.TestCase):
def setUp(self):
paddle.seed(0)
self.test_cases = [
[1, 2, 256, 256],
[4096, 128, 256, 5120],
]
self.TokenPadding = 0
def set_data(self, tokens_per_group, Experts, N, K):
self.tokens_per_group = tokens_per_group
self.N = N
self.K = K
self.BATCH = Experts
tokens = [self.tokens_per_group] * self.BATCH
self.tokens_prefix_sum = np.cumsum(tokens)
self.tokens = paddle.to_tensor(tokens, dtype="int64")
self.tokens_prefix_sum = paddle.to_tensor(self.tokens_prefix_sum, dtype="int64")
self.all_tokens = int(self.tokens.sum())
self.input_fp8 = paddle.randn([self.all_tokens, self.K], dtype="bfloat16").astype(paddle.float8_e4m3fn)
self.input_bf16 = self.input_fp8.astype("bfloat16")
self.weight = paddle.randn([self.BATCH, self.N, self.K], dtype="bfloat16")
self.weight_scale = 7 / self.weight.abs().max(axis=-1).reshape([self.BATCH, self.N, 1])
self.weight_quant = (self.weight * self.weight_scale).astype("int")
self.weight_quant = paddle.clip(self.weight_quant, -7, 7)
self.weight_quant_naive = self.weight_quant.astype("float32")
self.weight_quant = self.weight_quant.astype("bfloat16")
self.weight_quant = paddle.where(self.weight_quant > 0, self.weight_quant, 8 - self.weight_quant)
self.weight_dequant_scale = 1 / self.weight_scale.astype("float32")
self.max_tokens = int(self.tokens.max())
def w4afp8_gemm_naive(self, input_bf16, weight_quant, tokens, weight_dequant_scale):
all_tokens = int(tokens.sum())
out = paddle.zeros([all_tokens, self.N], dtype="bfloat16")
pre_fix_token = 0
for i in range(self.BATCH):
input = input_bf16[pre_fix_token : pre_fix_token + tokens[i], :]
weight = weight_quant[i] * weight_dequant_scale[i]
out_i = paddle.matmul(input, weight.astype("bfloat16"), transpose_y=True)
out[pre_fix_token : pre_fix_token + tokens[i], :] = out_i
pre_fix_token += tokens[i]
return out
def permute_scale(self, weight_scale):
weight_scale = weight_scale.reshape([self.BATCH, self.N])
temp = paddle.zeros([16])
for b in range(self.BATCH):
for n in range(0, self.N, 16):
temp[:] = weight_scale[b, n : n + 16]
for j in range(0, 16, 2):
weight_scale[b, n + j] = temp[j // 2]
weight_scale[b, n + j + 1] = temp[j // 2 + 8]
return weight_scale
def get_per_group_scale(self, processed_weight_scale):
processed_weight_scale = processed_weight_scale.repeat_interleave(self.K // 128, axis=-1)
origin_shape = processed_weight_scale.shape
processed_weight_scale = processed_weight_scale.transpose([0, 2, 1])
processed_weight_scale = processed_weight_scale.reshape([-1, processed_weight_scale.shape[-1]])
processed_weight_scale = w4afp8_gemm_scale_permute(processed_weight_scale)
processed_weight_scale = processed_weight_scale.reshape(
[origin_shape[0], origin_shape[2], origin_shape[1] // 128, 128]
)
processed_weight_scale = processed_weight_scale.transpose([0, 2, 1, 3])
return processed_weight_scale
def test_w4afp8_gemm(self):
for test_case in self.test_cases:
tokens_per_group, Experts, N, K = test_case
self.set_data(tokens_per_group, Experts, N, K)
out_naive = self.w4afp8_gemm_naive(
self.input_bf16, self.weight_quant_naive, self.tokens, self.weight_dequant_scale
)
# weight_dequant_scale = paddle.to_tensor(self.permute_scale(self.weight_dequant_scale) * 512)
weight_dequant_scale = self.get_per_group_scale(self.weight_dequant_scale * 512)
weight_int4 = w4afp8_gemm_weight_convert(self.weight_quant.astype("uint8").cpu()).cuda()
if self.TokenPadding == 0:
out_cuda = w4afp8_gemm(
self.input_fp8,
weight_int4,
self.tokens_prefix_sum,
weight_dequant_scale.astype("float32"),
None,
int(self.TokenPadding),
self.all_tokens,
True,
)
else:
out_cuda = w4afp8_gemm(
self.input_fp8,
weight_int4,
self.tokens,
weight_dequant_scale.astype("float32"),
None,
int(self.TokenPadding),
self.max_tokens,
True,
)
gap = (out_cuda - out_naive).abs()
self.assertLess(float(gap.mean()), 0.11)
if __name__ == "__main__":
unittest.main()