|
|
import re |
|
|
|
|
|
import pytest |
|
|
import torch |
|
|
from flash_attn.models.vit import vit_base_patch16_224 as flash_vit_base_patch16_224 |
|
|
from timm.models.vision_transformer import vit_base_patch16_224 |
|
|
|
|
|
|
|
|
@pytest.mark.parametrize("fused_mlp", [False, True]) |
|
|
|
|
|
@pytest.mark.parametrize("optimized", [False, True]) |
|
|
|
|
|
def test_vit(optimized, fused_mlp): |
|
|
"""Check that our implementation of ViT matches the timm's implementation: |
|
|
the output of our forward pass in fp16 should be around the same as |
|
|
timm' forward pass in fp16, when compared to timm's forward pass in fp32. |
|
|
""" |
|
|
dtype = torch.float16 |
|
|
device = "cuda" |
|
|
|
|
|
kwargs = {} |
|
|
if optimized: |
|
|
kwargs = dict(use_flash_attn=True, fused_bias_fc=True, fused_dropout_add_ln=True) |
|
|
kwargs["fused_mlp"] = fused_mlp |
|
|
model = flash_vit_base_patch16_224(**kwargs).to(device=device, dtype=dtype) |
|
|
|
|
|
model_ref = vit_base_patch16_224(pretrained=True).to(device=device) |
|
|
model_timm = vit_base_patch16_224(pretrained=True).to(device=device, dtype=dtype) |
|
|
|
|
|
model.load_state_dict(model_ref.state_dict()) |
|
|
|
|
|
model.eval() |
|
|
model_ref.eval() |
|
|
model_timm.eval() |
|
|
|
|
|
torch.manual_seed(0) |
|
|
batch_size = 2 |
|
|
x = torch.randn(batch_size, 3, 224, 224, device=device, dtype=dtype) |
|
|
out = model(x) |
|
|
out_timm = model_timm(x) |
|
|
out_ref = model_ref(x.float()) |
|
|
|
|
|
print(f"Output max diff: {(out - out_ref).abs().max().item()}") |
|
|
print(f"Output mean diff: {(out - out_ref).abs().mean().item()}") |
|
|
print(f"timm fp16 max diff: {(out_timm - out_ref).abs().max().item()}") |
|
|
print(f"timm fp16 mean diff: {(out_timm - out_ref).abs().mean().item()}") |
|
|
rtol = 2 if not fused_mlp else 8 |
|
|
assert (out - out_ref).abs().max().item() < rtol * (out_timm - out_ref).abs().max().item() |
|
|
|
