mirror of
https://github.com/open-mmlab/mmsegmentation.git
synced 2025-06-03 22:03:48 +08:00
4d34581897
* [Feature]Segformer re-implementation * Using act_cfg and norm_cfg to control activation and normalization * Split this PR into several little PRs * Fix lint error * Remove SegFormerHead * parameters init refactor * 1. Refactor segformer backbone parameters init; 2. Remove rebundant functions and unit tests; * Remove rebundant codes * 1. Remove rebundant codes; 2. Modify module name; * Refactor the backbone of segformer using mmcv.cnn.bricks.transformer.py * Fix some code logic bugs. * Add mit_convert.py to match pretrain keys of segformer. * Resolve some comments. * 1. Add some assert to ensure right params; 2. Support flexible peconv position; * Add pe_index assert and fix unit test. * 1. Add doc string for MixVisionTransformer; 2. Add some unit tests for MixVisionTransformer; * Use hw_shape to pass shape of feature map. * 1. Fix doc string of MixVisionTransformer; 2. Simplify MixFFN; 3. Modify H, W to hw_shape; * Add more unit tests. * Add doc string for shape convertion functions. * Add some unit tests to improve code coverage. * Fix Segformer backbone pretrain weights match bug. * resolve the shape convertion functions doc string. * Add pad_to_patch_size arg. * Modify default value of pad_to_patch_size arg.
140 lines
4.8 KiB
Python
140 lines
4.8 KiB
Python
from collections import OrderedDict
|
|
|
|
import torch
|
|
|
|
|
|
def swin_convert(ckpt):
|
|
new_ckpt = OrderedDict()
|
|
|
|
def correct_unfold_reduction_order(x):
|
|
out_channel, in_channel = x.shape
|
|
x = x.reshape(out_channel, 4, in_channel // 4)
|
|
x = x[:, [0, 2, 1, 3], :].transpose(1,
|
|
2).reshape(out_channel, in_channel)
|
|
return x
|
|
|
|
def correct_unfold_norm_order(x):
|
|
in_channel = x.shape[0]
|
|
x = x.reshape(4, in_channel // 4)
|
|
x = x[[0, 2, 1, 3], :].transpose(0, 1).reshape(in_channel)
|
|
return x
|
|
|
|
for k, v in ckpt.items():
|
|
if k.startswith('head'):
|
|
continue
|
|
elif k.startswith('layers'):
|
|
new_v = v
|
|
if 'attn.' in k:
|
|
new_k = k.replace('attn.', 'attn.w_msa.')
|
|
elif 'mlp.' in k:
|
|
if 'mlp.fc1.' in k:
|
|
new_k = k.replace('mlp.fc1.', 'ffn.layers.0.0.')
|
|
elif 'mlp.fc2.' in k:
|
|
new_k = k.replace('mlp.fc2.', 'ffn.layers.1.')
|
|
else:
|
|
new_k = k.replace('mlp.', 'ffn.')
|
|
elif 'downsample' in k:
|
|
new_k = k
|
|
if 'reduction.' in k:
|
|
new_v = correct_unfold_reduction_order(v)
|
|
elif 'norm.' in k:
|
|
new_v = correct_unfold_norm_order(v)
|
|
else:
|
|
new_k = k
|
|
new_k = new_k.replace('layers', 'stages', 1)
|
|
elif k.startswith('patch_embed'):
|
|
new_v = v
|
|
if 'proj' in k:
|
|
new_k = k.replace('proj', 'projection')
|
|
else:
|
|
new_k = k
|
|
else:
|
|
new_v = v
|
|
new_k = k
|
|
|
|
new_ckpt[new_k] = new_v
|
|
|
|
return new_ckpt
|
|
|
|
|
|
def vit_convert(ckpt):
|
|
|
|
new_ckpt = OrderedDict()
|
|
|
|
for k, v in ckpt.items():
|
|
if k.startswith('head'):
|
|
continue
|
|
if k.startswith('norm'):
|
|
new_k = k.replace('norm.', 'ln1.')
|
|
elif k.startswith('patch_embed'):
|
|
if 'proj' in k:
|
|
new_k = k.replace('proj', 'projection')
|
|
else:
|
|
new_k = k
|
|
elif k.startswith('blocks'):
|
|
if 'norm' in k:
|
|
new_k = k.replace('norm', 'ln')
|
|
elif 'mlp.fc1' in k:
|
|
new_k = k.replace('mlp.fc1', 'ffn.layers.0.0')
|
|
elif 'mlp.fc2' in k:
|
|
new_k = k.replace('mlp.fc2', 'ffn.layers.1')
|
|
elif 'attn.qkv' in k:
|
|
new_k = k.replace('attn.qkv.', 'attn.attn.in_proj_')
|
|
elif 'attn.proj' in k:
|
|
new_k = k.replace('attn.proj', 'attn.attn.out_proj')
|
|
else:
|
|
new_k = k
|
|
new_k = new_k.replace('blocks.', 'layers.')
|
|
else:
|
|
new_k = k
|
|
new_ckpt[new_k] = v
|
|
|
|
return new_ckpt
|
|
|
|
|
|
def mit_convert(ckpt):
|
|
new_ckpt = OrderedDict()
|
|
# Process the concat between q linear weights and kv linear weights
|
|
for k, v in ckpt.items():
|
|
if k.startswith('head'):
|
|
continue
|
|
elif k.startswith('patch_embed'):
|
|
stage_i = int(k.split('.')[0].replace('patch_embed', ''))
|
|
new_k = k.replace(f'patch_embed{stage_i}', f'layers.{stage_i-1}.0')
|
|
new_v = v
|
|
if 'proj.' in new_k:
|
|
new_k = new_k.replace('proj.', 'projection.')
|
|
elif k.startswith('block'):
|
|
stage_i = int(k.split('.')[0].replace('block', ''))
|
|
new_k = k.replace(f'block{stage_i}', f'layers.{stage_i-1}.1')
|
|
new_v = v
|
|
if 'attn.q.' in new_k:
|
|
sub_item_k = k.replace('q.', 'kv.')
|
|
new_k = new_k.replace('q.', 'attn.in_proj_')
|
|
new_v = torch.cat([v, ckpt[sub_item_k]], dim=0)
|
|
elif 'attn.kv.' in new_k:
|
|
continue
|
|
elif 'attn.proj.' in new_k:
|
|
new_k = new_k.replace('proj.', 'attn.out_proj.')
|
|
elif 'attn.sr.' in new_k:
|
|
new_k = new_k.replace('sr.', 'sr.')
|
|
elif 'mlp.' in new_k:
|
|
string = f'{new_k}-'
|
|
new_k = new_k.replace('mlp.', 'ffn.layers.')
|
|
if 'fc1.weight' in new_k or 'fc2.weight' in new_k:
|
|
new_v = v.reshape((*v.shape, 1, 1))
|
|
new_k = new_k.replace('fc1.', '0.')
|
|
new_k = new_k.replace('dwconv.dwconv.', '1.')
|
|
new_k = new_k.replace('fc2.', '4.')
|
|
string += f'{new_k} {v.shape}-{new_v.shape}'
|
|
# print(string)
|
|
elif k.startswith('norm'):
|
|
stage_i = int(k.split('.')[0].replace('norm', ''))
|
|
new_k = k.replace(f'norm{stage_i}', f'layers.{stage_i-1}.2')
|
|
new_v = v
|
|
else:
|
|
new_k = k
|
|
new_v = v
|
|
new_ckpt[new_k] = new_v
|
|
return new_ckpt
|