forked from CompVis/stable-diffusion
-
Notifications
You must be signed in to change notification settings - Fork 5
/
quantize.py
445 lines (370 loc) · 17.8 KB
/
quantize.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
import torch
import torch.nn as nn
import torch.nn.functional as F
import numpy as np
from torch import einsum
from einops import rearrange
class VectorQuantizer(nn.Module):
"""
see https://github.com/MishaLaskin/vqvae/blob/d761a999e2267766400dc646d82d3ac3657771d4/models/quantizer.py
____________________________________________
Discretization bottleneck part of the VQ-VAE.
Inputs:
- n_e : number of embeddings
- e_dim : dimension of embedding
- beta : commitment cost used in loss term, beta * ||z_e(x)-sg[e]||^2
_____________________________________________
"""
# NOTE: this class contains a bug regarding beta; see VectorQuantizer2 for
# a fix and use legacy=False to apply that fix. VectorQuantizer2 can be
# used wherever VectorQuantizer has been used before and is additionally
# more efficient.
def __init__(self, n_e, e_dim, beta):
super(VectorQuantizer, self).__init__()
self.n_e = n_e
self.e_dim = e_dim
self.beta = beta
self.embedding = nn.Embedding(self.n_e, self.e_dim)
self.embedding.weight.data.uniform_(-1.0 / self.n_e, 1.0 / self.n_e)
def forward(self, z):
"""
Inputs the output of the encoder network z and maps it to a discrete
one-hot vector that is the index of the closest embedding vector e_j
z (continuous) -> z_q (discrete)
z.shape = (batch, channel, height, width)
quantization pipeline:
1. get encoder input (B,C,H,W)
2. flatten input to (B*H*W,C)
"""
# reshape z -> (batch, height, width, channel) and flatten
z = z.permute(0, 2, 3, 1).contiguous()
z_flattened = z.view(-1, self.e_dim)
# distances from z to embeddings e_j (z - e)^2 = z^2 + e^2 - 2 e * z
d = torch.sum(z_flattened ** 2, dim=1, keepdim=True) + \
torch.sum(self.embedding.weight**2, dim=1) - 2 * \
torch.matmul(z_flattened, self.embedding.weight.t())
## could possible replace this here
# #\start...
# find closest encodings
min_encoding_indices = torch.argmin(d, dim=1).unsqueeze(1)
min_encodings = torch.zeros(
min_encoding_indices.shape[0], self.n_e).to(z)
min_encodings.scatter_(1, min_encoding_indices, 1)
# dtype min encodings: torch.float32
# min_encodings shape: torch.Size([2048, 512])
# min_encoding_indices.shape: torch.Size([2048, 1])
# get quantized latent vectors
z_q = torch.matmul(min_encodings, self.embedding.weight).view(z.shape)
#.........\end
# with:
# .........\start
#min_encoding_indices = torch.argmin(d, dim=1)
#z_q = self.embedding(min_encoding_indices)
# ......\end......... (TODO)
# compute loss for embedding
loss = torch.mean((z_q.detach()-z)**2) + self.beta * \
torch.mean((z_q - z.detach()) ** 2)
# preserve gradients
z_q = z + (z_q - z).detach()
# perplexity
e_mean = torch.mean(min_encodings, dim=0)
perplexity = torch.exp(-torch.sum(e_mean * torch.log(e_mean + 1e-10)))
# reshape back to match original input shape
z_q = z_q.permute(0, 3, 1, 2).contiguous()
return z_q, loss, (perplexity, min_encodings, min_encoding_indices)
def get_codebook_entry(self, indices, shape):
# shape specifying (batch, height, width, channel)
# TODO: check for more easy handling with nn.Embedding
min_encodings = torch.zeros(indices.shape[0], self.n_e).to(indices)
min_encodings.scatter_(1, indices[:,None], 1)
# get quantized latent vectors
z_q = torch.matmul(min_encodings.float(), self.embedding.weight)
if shape is not None:
z_q = z_q.view(shape)
# reshape back to match original input shape
z_q = z_q.permute(0, 3, 1, 2).contiguous()
return z_q
class GumbelQuantize(nn.Module):
"""
credit to @karpathy: https://github.com/karpathy/deep-vector-quantization/blob/main/model.py (thanks!)
Gumbel Softmax trick quantizer
Categorical Reparameterization with Gumbel-Softmax, Jang et al. 2016
https://arxiv.org/abs/1611.01144
"""
def __init__(self, num_hiddens, embedding_dim, n_embed, straight_through=True,
kl_weight=5e-4, temp_init=1.0, use_vqinterface=True,
remap=None, unknown_index="random"):
super().__init__()
self.embedding_dim = embedding_dim
self.n_embed = n_embed
self.straight_through = straight_through
self.temperature = temp_init
self.kl_weight = kl_weight
self.proj = nn.Conv2d(num_hiddens, n_embed, 1)
self.embed = nn.Embedding(n_embed, embedding_dim)
self.use_vqinterface = use_vqinterface
self.remap = remap
if self.remap is not None:
self.register_buffer("used", torch.tensor(np.load(self.remap)))
self.re_embed = self.used.shape[0]
self.unknown_index = unknown_index # "random" or "extra" or integer
if self.unknown_index == "extra":
self.unknown_index = self.re_embed
self.re_embed = self.re_embed+1
print(f"Remapping {self.n_embed} indices to {self.re_embed} indices. "
f"Using {self.unknown_index} for unknown indices.")
else:
self.re_embed = n_embed
def remap_to_used(self, inds):
ishape = inds.shape
assert len(ishape)>1
inds = inds.reshape(ishape[0],-1)
used = self.used.to(inds)
match = (inds[:,:,None]==used[None,None,...]).long()
new = match.argmax(-1)
unknown = match.sum(2)<1
if self.unknown_index == "random":
new[unknown]=torch.randint(0,self.re_embed,size=new[unknown].shape).to(device=new.device)
else:
new[unknown] = self.unknown_index
return new.reshape(ishape)
def unmap_to_all(self, inds):
ishape = inds.shape
assert len(ishape)>1
inds = inds.reshape(ishape[0],-1)
used = self.used.to(inds)
if self.re_embed > self.used.shape[0]: # extra token
inds[inds>=self.used.shape[0]] = 0 # simply set to zero
back=torch.gather(used[None,:][inds.shape[0]*[0],:], 1, inds)
return back.reshape(ishape)
def forward(self, z, temp=None, return_logits=False):
# force hard = True when we are in eval mode, as we must quantize. actually, always true seems to work
hard = self.straight_through if self.training else True
temp = self.temperature if temp is None else temp
logits = self.proj(z)
if self.remap is not None:
# continue only with used logits
full_zeros = torch.zeros_like(logits)
logits = logits[:,self.used,...]
soft_one_hot = F.gumbel_softmax(logits, tau=temp, dim=1, hard=hard)
if self.remap is not None:
# go back to all entries but unused set to zero
full_zeros[:,self.used,...] = soft_one_hot
soft_one_hot = full_zeros
z_q = einsum('b n h w, n d -> b d h w', soft_one_hot, self.embed.weight)
# + kl divergence to the prior loss
qy = F.softmax(logits, dim=1)
diff = self.kl_weight * torch.sum(qy * torch.log(qy * self.n_embed + 1e-10), dim=1).mean()
ind = soft_one_hot.argmax(dim=1)
if self.remap is not None:
ind = self.remap_to_used(ind)
if self.use_vqinterface:
if return_logits:
return z_q, diff, (None, None, ind), logits
return z_q, diff, (None, None, ind)
return z_q, diff, ind
def get_codebook_entry(self, indices, shape):
b, h, w, c = shape
assert b*h*w == indices.shape[0]
indices = rearrange(indices, '(b h w) -> b h w', b=b, h=h, w=w)
if self.remap is not None:
indices = self.unmap_to_all(indices)
one_hot = F.one_hot(indices, num_classes=self.n_embed).permute(0, 3, 1, 2).float()
z_q = einsum('b n h w, n d -> b d h w', one_hot, self.embed.weight)
return z_q
class VectorQuantizer2(nn.Module):
"""
Improved version over VectorQuantizer, can be used as a drop-in replacement. Mostly
avoids costly matrix multiplications and allows for post-hoc remapping of indices.
"""
# NOTE: due to a bug the beta term was applied to the wrong term. for
# backwards compatibility we use the buggy version by default, but you can
# specify legacy=False to fix it.
def __init__(self, n_e, e_dim, beta, remap=None, unknown_index="random",
sane_index_shape=False, legacy=True):
super().__init__()
self.n_e = n_e
self.e_dim = e_dim
self.beta = beta
self.legacy = legacy
self.embedding = nn.Embedding(self.n_e, self.e_dim)
self.embedding.weight.data.uniform_(-1.0 / self.n_e, 1.0 / self.n_e)
self.remap = remap
if self.remap is not None:
self.register_buffer("used", torch.tensor(np.load(self.remap)))
self.re_embed = self.used.shape[0]
self.unknown_index = unknown_index # "random" or "extra" or integer
if self.unknown_index == "extra":
self.unknown_index = self.re_embed
self.re_embed = self.re_embed+1
print(f"Remapping {self.n_e} indices to {self.re_embed} indices. "
f"Using {self.unknown_index} for unknown indices.")
else:
self.re_embed = n_e
self.sane_index_shape = sane_index_shape
def remap_to_used(self, inds):
ishape = inds.shape
assert len(ishape)>1
inds = inds.reshape(ishape[0],-1)
used = self.used.to(inds)
match = (inds[:,:,None]==used[None,None,...]).long()
new = match.argmax(-1)
unknown = match.sum(2)<1
if self.unknown_index == "random":
new[unknown]=torch.randint(0,self.re_embed,size=new[unknown].shape).to(device=new.device)
else:
new[unknown] = self.unknown_index
return new.reshape(ishape)
def unmap_to_all(self, inds):
ishape = inds.shape
assert len(ishape)>1
inds = inds.reshape(ishape[0],-1)
used = self.used.to(inds)
if self.re_embed > self.used.shape[0]: # extra token
inds[inds>=self.used.shape[0]] = 0 # simply set to zero
back=torch.gather(used[None,:][inds.shape[0]*[0],:], 1, inds)
return back.reshape(ishape)
def forward(self, z, temp=None, rescale_logits=False, return_logits=False):
assert temp is None or temp==1.0, "Only for interface compatible with Gumbel"
assert rescale_logits==False, "Only for interface compatible with Gumbel"
assert return_logits==False, "Only for interface compatible with Gumbel"
# reshape z -> (batch, height, width, channel) and flatten
z = rearrange(z, 'b c h w -> b h w c').contiguous()
z_flattened = z.view(-1, self.e_dim)
# distances from z to embeddings e_j (z - e)^2 = z^2 + e^2 - 2 e * z
d = torch.sum(z_flattened ** 2, dim=1, keepdim=True) + \
torch.sum(self.embedding.weight**2, dim=1) - 2 * \
torch.einsum('bd,dn->bn', z_flattened, rearrange(self.embedding.weight, 'n d -> d n'))
min_encoding_indices = torch.argmin(d, dim=1)
z_q = self.embedding(min_encoding_indices).view(z.shape)
perplexity = None
min_encodings = None
# compute loss for embedding
if not self.legacy:
loss = self.beta * torch.mean((z_q.detach()-z)**2) + \
torch.mean((z_q - z.detach()) ** 2)
else:
loss = torch.mean((z_q.detach()-z)**2) + self.beta * \
torch.mean((z_q - z.detach()) ** 2)
# preserve gradients
z_q = z + (z_q - z).detach()
# reshape back to match original input shape
z_q = rearrange(z_q, 'b h w c -> b c h w').contiguous()
if self.remap is not None:
min_encoding_indices = min_encoding_indices.reshape(z.shape[0],-1) # add batch axis
min_encoding_indices = self.remap_to_used(min_encoding_indices)
min_encoding_indices = min_encoding_indices.reshape(-1,1) # flatten
if self.sane_index_shape:
min_encoding_indices = min_encoding_indices.reshape(
z_q.shape[0], z_q.shape[2], z_q.shape[3])
return z_q, loss, (perplexity, min_encodings, min_encoding_indices)
def get_codebook_entry(self, indices, shape):
# shape specifying (batch, height, width, channel)
if self.remap is not None:
indices = indices.reshape(shape[0],-1) # add batch axis
indices = self.unmap_to_all(indices)
indices = indices.reshape(-1) # flatten again
# get quantized latent vectors
z_q = self.embedding(indices)
if shape is not None:
z_q = z_q.view(shape)
# reshape back to match original input shape
z_q = z_q.permute(0, 3, 1, 2).contiguous()
return z_q
class EmbeddingEMA(nn.Module):
def __init__(self, num_tokens, codebook_dim, decay=0.99, eps=1e-5):
super().__init__()
self.decay = decay
self.eps = eps
weight = torch.randn(num_tokens, codebook_dim)
self.weight = nn.Parameter(weight, requires_grad = False)
self.cluster_size = nn.Parameter(torch.zeros(num_tokens), requires_grad = False)
self.embed_avg = nn.Parameter(weight.clone(), requires_grad = False)
self.update = True
def forward(self, embed_id):
return F.embedding(embed_id, self.weight)
def cluster_size_ema_update(self, new_cluster_size):
self.cluster_size.data.mul_(self.decay).add_(new_cluster_size, alpha=1 - self.decay)
def embed_avg_ema_update(self, new_embed_avg):
self.embed_avg.data.mul_(self.decay).add_(new_embed_avg, alpha=1 - self.decay)
def weight_update(self, num_tokens):
n = self.cluster_size.sum()
smoothed_cluster_size = (
(self.cluster_size + self.eps) / (n + num_tokens * self.eps) * n
)
#normalize embedding average with smoothed cluster size
embed_normalized = self.embed_avg / smoothed_cluster_size.unsqueeze(1)
self.weight.data.copy_(embed_normalized)
class EMAVectorQuantizer(nn.Module):
def __init__(self, n_embed, embedding_dim, beta, decay=0.99, eps=1e-5,
remap=None, unknown_index="random"):
super().__init__()
self.codebook_dim = codebook_dim
self.num_tokens = num_tokens
self.beta = beta
self.embedding = EmbeddingEMA(self.num_tokens, self.codebook_dim, decay, eps)
self.remap = remap
if self.remap is not None:
self.register_buffer("used", torch.tensor(np.load(self.remap)))
self.re_embed = self.used.shape[0]
self.unknown_index = unknown_index # "random" or "extra" or integer
if self.unknown_index == "extra":
self.unknown_index = self.re_embed
self.re_embed = self.re_embed+1
print(f"Remapping {self.n_embed} indices to {self.re_embed} indices. "
f"Using {self.unknown_index} for unknown indices.")
else:
self.re_embed = n_embed
def remap_to_used(self, inds):
ishape = inds.shape
assert len(ishape)>1
inds = inds.reshape(ishape[0],-1)
used = self.used.to(inds)
match = (inds[:,:,None]==used[None,None,...]).long()
new = match.argmax(-1)
unknown = match.sum(2)<1
if self.unknown_index == "random":
new[unknown]=torch.randint(0,self.re_embed,size=new[unknown].shape).to(device=new.device)
else:
new[unknown] = self.unknown_index
return new.reshape(ishape)
def unmap_to_all(self, inds):
ishape = inds.shape
assert len(ishape)>1
inds = inds.reshape(ishape[0],-1)
used = self.used.to(inds)
if self.re_embed > self.used.shape[0]: # extra token
inds[inds>=self.used.shape[0]] = 0 # simply set to zero
back=torch.gather(used[None,:][inds.shape[0]*[0],:], 1, inds)
return back.reshape(ishape)
def forward(self, z):
# reshape z -> (batch, height, width, channel) and flatten
#z, 'b c h w -> b h w c'
z = rearrange(z, 'b c h w -> b h w c')
z_flattened = z.reshape(-1, self.codebook_dim)
# distances from z to embeddings e_j (z - e)^2 = z^2 + e^2 - 2 e * z
d = z_flattened.pow(2).sum(dim=1, keepdim=True) + \
self.embedding.weight.pow(2).sum(dim=1) - 2 * \
torch.einsum('bd,nd->bn', z_flattened, self.embedding.weight) # 'n d -> d n'
encoding_indices = torch.argmin(d, dim=1)
z_q = self.embedding(encoding_indices).view(z.shape)
encodings = F.one_hot(encoding_indices, self.num_tokens).type(z.dtype)
avg_probs = torch.mean(encodings, dim=0)
perplexity = torch.exp(-torch.sum(avg_probs * torch.log(avg_probs + 1e-10)))
if self.training and self.embedding.update:
#EMA cluster size
encodings_sum = encodings.sum(0)
self.embedding.cluster_size_ema_update(encodings_sum)
#EMA embedding average
embed_sum = encodings.transpose(0,1) @ z_flattened
self.embedding.embed_avg_ema_update(embed_sum)
#normalize embed_avg and update weight
self.embedding.weight_update(self.num_tokens)
# compute loss for embedding
loss = self.beta * F.mse_loss(z_q.detach(), z)
# preserve gradients
z_q = z + (z_q - z).detach()
# reshape back to match original input shape
#z_q, 'b h w c -> b c h w'
z_q = rearrange(z_q, 'b h w c -> b c h w')
return z_q, loss, (perplexity, encodings, encoding_indices)