From 820f1dc96b1979d7e92170c161db281ee8bd988b Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 15:03:39 +0300 Subject: initial support for training textual inversion --- modules/sd_hijack.py | 324 ++++++++------------------------------------------- 1 file changed, 51 insertions(+), 273 deletions(-) (limited to 'modules/sd_hijack.py') diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index fa7eaeb8..fd57e5c5 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -6,244 +6,41 @@ import torch import numpy as np from torch import einsum -from modules import prompt_parser +import modules.textual_inversion.textual_inversion +from modules import prompt_parser, devices, sd_hijack_optimizations, shared from modules.shared import opts, device, cmd_opts -from ldm.util import default -from einops import rearrange import ldm.modules.attention import ldm.modules.diffusionmodules.model +attention_CrossAttention_forward = ldm.modules.attention.CrossAttention.forward +diffusionmodules_model_nonlinearity = ldm.modules.diffusionmodules.model.nonlinearity +diffusionmodules_model_AttnBlock_forward = ldm.modules.diffusionmodules.model.AttnBlock.forward -# see https://github.com/basujindal/stable-diffusion/pull/117 for discussion -def split_cross_attention_forward_v1(self, x, context=None, mask=None): - h = self.heads - q = self.to_q(x) - context = default(context, x) - k = self.to_k(context) - v = self.to_v(context) - del context, x +def apply_optimizations(): + if cmd_opts.opt_split_attention_v1: + ldm.modules.attention.CrossAttention.forward = sd_hijack_optimizations.split_cross_attention_forward_v1 + elif not cmd_opts.disable_opt_split_attention and (cmd_opts.opt_split_attention or torch.cuda.is_available()): + ldm.modules.attention.CrossAttention.forward = sd_hijack_optimizations.split_cross_attention_forward + ldm.modules.diffusionmodules.model.nonlinearity = sd_hijack_optimizations.nonlinearity_hijack + ldm.modules.diffusionmodules.model.AttnBlock.forward = sd_hijack_optimizations.cross_attention_attnblock_forward - q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v)) - r1 = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device) - for i in range(0, q.shape[0], 2): - end = i + 2 - s1 = einsum('b i d, b j d -> b i j', q[i:end], k[i:end]) - s1 *= self.scale +def undo_optimizations(): + ldm.modules.attention.CrossAttention.forward = attention_CrossAttention_forward + ldm.modules.diffusionmodules.model.nonlinearity = diffusionmodules_model_nonlinearity + ldm.modules.diffusionmodules.model.AttnBlock.forward = diffusionmodules_model_AttnBlock_forward - s2 = s1.softmax(dim=-1) - del s1 - - r1[i:end] = einsum('b i j, b j d -> b i d', s2, v[i:end]) - del s2 - - r2 = rearrange(r1, '(b h) n d -> b n (h d)', h=h) - del r1 - - return self.to_out(r2) - - -# taken from https://github.com/Doggettx/stable-diffusion -def split_cross_attention_forward(self, x, context=None, mask=None): - h = self.heads - - q_in = self.to_q(x) - context = default(context, x) - k_in = self.to_k(context) * self.scale - v_in = self.to_v(context) - del context, x - - q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q_in, k_in, v_in)) - del q_in, k_in, v_in - - r1 = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device, dtype=q.dtype) - - stats = torch.cuda.memory_stats(q.device) - mem_active = stats['active_bytes.all.current'] - mem_reserved = stats['reserved_bytes.all.current'] - mem_free_cuda, _ = torch.cuda.mem_get_info(torch.cuda.current_device()) - mem_free_torch = mem_reserved - mem_active - mem_free_total = mem_free_cuda + mem_free_torch - - gb = 1024 ** 3 - tensor_size = q.shape[0] * q.shape[1] * k.shape[1] * q.element_size() - modifier = 3 if q.element_size() == 2 else 2.5 - mem_required = tensor_size * modifier - steps = 1 - - if mem_required > mem_free_total: - steps = 2 ** (math.ceil(math.log(mem_required / mem_free_total, 2))) - # print(f"Expected tensor size:{tensor_size/gb:0.1f}GB, cuda free:{mem_free_cuda/gb:0.1f}GB " - # f"torch free:{mem_free_torch/gb:0.1f} total:{mem_free_total/gb:0.1f} steps:{steps}") - - if steps > 64: - max_res = math.floor(math.sqrt(math.sqrt(mem_free_total / 2.5)) / 8) * 64 - raise RuntimeError(f'Not enough memory, use lower resolution (max approx. {max_res}x{max_res}). ' - f'Need: {mem_required / 64 / gb:0.1f}GB free, Have:{mem_free_total / gb:0.1f}GB free') - - slice_size = q.shape[1] // steps if (q.shape[1] % steps) == 0 else q.shape[1] - for i in range(0, q.shape[1], slice_size): - end = i + slice_size - s1 = einsum('b i d, b j d -> b i j', q[:, i:end], k) - - s2 = s1.softmax(dim=-1, dtype=q.dtype) - del s1 - - r1[:, i:end] = einsum('b i j, b j d -> b i d', s2, v) - del s2 - - del q, k, v - - r2 = rearrange(r1, '(b h) n d -> b n (h d)', h=h) - del r1 - - return self.to_out(r2) - -def nonlinearity_hijack(x): - # swish - t = torch.sigmoid(x) - x *= t - del t - - return x - -def cross_attention_attnblock_forward(self, x): - h_ = x - h_ = self.norm(h_) - q1 = self.q(h_) - k1 = self.k(h_) - v = self.v(h_) - - # compute attention - b, c, h, w = q1.shape - - q2 = q1.reshape(b, c, h*w) - del q1 - - q = q2.permute(0, 2, 1) # b,hw,c - del q2 - - k = k1.reshape(b, c, h*w) # b,c,hw - del k1 - - h_ = torch.zeros_like(k, device=q.device) - - stats = torch.cuda.memory_stats(q.device) - mem_active = stats['active_bytes.all.current'] - mem_reserved = stats['reserved_bytes.all.current'] - mem_free_cuda, _ = torch.cuda.mem_get_info(torch.cuda.current_device()) - mem_free_torch = mem_reserved - mem_active - mem_free_total = mem_free_cuda + mem_free_torch - - tensor_size = q.shape[0] * q.shape[1] * k.shape[2] * q.element_size() - mem_required = tensor_size * 2.5 - steps = 1 - - if mem_required > mem_free_total: - steps = 2**(math.ceil(math.log(mem_required / mem_free_total, 2))) - - slice_size = q.shape[1] // steps if (q.shape[1] % steps) == 0 else q.shape[1] - for i in range(0, q.shape[1], slice_size): - end = i + slice_size - - w1 = torch.bmm(q[:, i:end], k) # b,hw,hw w[b,i,j]=sum_c q[b,i,c]k[b,c,j] - w2 = w1 * (int(c)**(-0.5)) - del w1 - w3 = torch.nn.functional.softmax(w2, dim=2, dtype=q.dtype) - del w2 - - # attend to values - v1 = v.reshape(b, c, h*w) - w4 = w3.permute(0, 2, 1) # b,hw,hw (first hw of k, second of q) - del w3 - - h_[:, :, i:end] = torch.bmm(v1, w4) # b, c,hw (hw of q) h_[b,c,j] = sum_i v[b,c,i] w_[b,i,j] - del v1, w4 - - h2 = h_.reshape(b, c, h, w) - del h_ - - h3 = self.proj_out(h2) - del h2 - - h3 += x - - return h3 class StableDiffusionModelHijack: - ids_lookup = {} - word_embeddings = {} - word_embeddings_checksums = {} fixes = None comments = [] - dir_mtime = None layers = None circular_enabled = False clip = None - def load_textual_inversion_embeddings(self, dirname, model): - mt = os.path.getmtime(dirname) - if self.dir_mtime is not None and mt <= self.dir_mtime: - return - - self.dir_mtime = mt - self.ids_lookup.clear() - self.word_embeddings.clear() - - tokenizer = model.cond_stage_model.tokenizer - - def const_hash(a): - r = 0 - for v in a: - r = (r * 281 ^ int(v) * 997) & 0xFFFFFFFF - return r - - def process_file(path, filename): - name = os.path.splitext(filename)[0] - - data = torch.load(path, map_location="cpu") - - # textual inversion embeddings - if 'string_to_param' in data: - param_dict = data['string_to_param'] - if hasattr(param_dict, '_parameters'): - param_dict = getattr(param_dict, '_parameters') # fix for torch 1.12.1 loading saved file from torch 1.11 - assert len(param_dict) == 1, 'embedding file has multiple terms in it' - emb = next(iter(param_dict.items()))[1] - # diffuser concepts - elif type(data) == dict and type(next(iter(data.values()))) == torch.Tensor: - assert len(data.keys()) == 1, 'embedding file has multiple terms in it' - - emb = next(iter(data.values())) - if len(emb.shape) == 1: - emb = emb.unsqueeze(0) - - self.word_embeddings[name] = emb.detach().to(device) - self.word_embeddings_checksums[name] = f'{const_hash(emb.reshape(-1)*100)&0xffff:04x}' - - ids = tokenizer([name], add_special_tokens=False)['input_ids'][0] - - first_id = ids[0] - if first_id not in self.ids_lookup: - self.ids_lookup[first_id] = [] - self.ids_lookup[first_id].append((ids, name)) - - for fn in os.listdir(dirname): - try: - fullfn = os.path.join(dirname, fn) - - if os.stat(fullfn).st_size == 0: - continue - - process_file(fullfn, fn) - except Exception: - print(f"Error loading emedding {fn}:", file=sys.stderr) - print(traceback.format_exc(), file=sys.stderr) - continue - - print(f"Loaded a total of {len(self.word_embeddings)} textual inversion embeddings.") + embedding_db = modules.textual_inversion.textual_inversion.EmbeddingDatabase(cmd_opts.embeddings_dir) def hijack(self, m): model_embeddings = m.cond_stage_model.transformer.text_model.embeddings @@ -253,12 +50,7 @@ class StableDiffusionModelHijack: self.clip = m.cond_stage_model - if cmd_opts.opt_split_attention_v1: - ldm.modules.attention.CrossAttention.forward = split_cross_attention_forward_v1 - elif not cmd_opts.disable_opt_split_attention and (cmd_opts.opt_split_attention or torch.cuda.is_available()): - ldm.modules.attention.CrossAttention.forward = split_cross_attention_forward - ldm.modules.diffusionmodules.model.nonlinearity = nonlinearity_hijack - ldm.modules.diffusionmodules.model.AttnBlock.forward = cross_attention_attnblock_forward + apply_optimizations() def flatten(el): flattened = [flatten(children) for children in el.children()] @@ -296,7 +88,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): def __init__(self, wrapped, hijack): super().__init__() self.wrapped = wrapped - self.hijack = hijack + self.hijack: StableDiffusionModelHijack = hijack self.tokenizer = wrapped.tokenizer self.max_length = wrapped.max_length self.token_mults = {} @@ -317,7 +109,6 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): if mult != 1.0: self.token_mults[ident] = mult - def tokenize_line(self, line, used_custom_terms, hijack_comments): id_start = self.wrapped.tokenizer.bos_token_id id_end = self.wrapped.tokenizer.eos_token_id @@ -339,28 +130,19 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): while i < len(tokens): token = tokens[i] - possible_matches = self.hijack.ids_lookup.get(token, None) + embedding = self.hijack.embedding_db.find_embedding_at_position(tokens, i) - if possible_matches is None: + if embedding is None: remade_tokens.append(token) multipliers.append(weight) + i += 1 else: - found = False - for ids, word in possible_matches: - if tokens[i:i + len(ids)] == ids: - emb_len = int(self.hijack.word_embeddings[word].shape[0]) - fixes.append((len(remade_tokens), word)) - remade_tokens += [0] * emb_len - multipliers += [weight] * emb_len - i += len(ids) - 1 - found = True - used_custom_terms.append((word, self.hijack.word_embeddings_checksums[word])) - break - - if not found: - remade_tokens.append(token) - multipliers.append(weight) - i += 1 + emb_len = int(embedding.vec.shape[0]) + fixes.append((len(remade_tokens), embedding)) + remade_tokens += [0] * emb_len + multipliers += [weight] * emb_len + used_custom_terms.append((embedding.name, embedding.checksum())) + i += emb_len if len(remade_tokens) > maxlen - 2: vocab = {v: k for k, v in self.wrapped.tokenizer.get_vocab().items()} @@ -431,32 +213,23 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): while i < len(tokens): token = tokens[i] - possible_matches = self.hijack.ids_lookup.get(token, None) + embedding = self.hijack.embedding_db.find_embedding_at_position(tokens, i) mult_change = self.token_mults.get(token) if opts.enable_emphasis else None if mult_change is not None: mult *= mult_change - elif possible_matches is None: + i += 1 + elif embedding is None: remade_tokens.append(token) multipliers.append(mult) + i += 1 else: - found = False - for ids, word in possible_matches: - if tokens[i:i+len(ids)] == ids: - emb_len = int(self.hijack.word_embeddings[word].shape[0]) - fixes.append((len(remade_tokens), word)) - remade_tokens += [0] * emb_len - multipliers += [mult] * emb_len - i += len(ids) - 1 - found = True - used_custom_terms.append((word, self.hijack.word_embeddings_checksums[word])) - break - - if not found: - remade_tokens.append(token) - multipliers.append(mult) - - i += 1 + emb_len = int(embedding.vec.shape[0]) + fixes.append((len(remade_tokens), embedding)) + remade_tokens += [0] * emb_len + multipliers += [mult] * emb_len + used_custom_terms.append((embedding.name, embedding.checksum())) + i += emb_len if len(remade_tokens) > maxlen - 2: vocab = {v: k for k, v in self.wrapped.tokenizer.get_vocab().items()} @@ -464,6 +237,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): overflowing_words = [vocab.get(int(x), "") for x in ovf] overflowing_text = self.wrapped.tokenizer.convert_tokens_to_string(''.join(overflowing_words)) hijack_comments.append(f"Warning: too many input tokens; some ({len(overflowing_words)}) have been truncated:\n{overflowing_text}\n") + token_count = len(remade_tokens) remade_tokens = remade_tokens + [id_end] * (maxlen - 2 - len(remade_tokens)) remade_tokens = [id_start] + remade_tokens[0:maxlen-2] + [id_end] @@ -484,7 +258,6 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): else: batch_multipliers, remade_batch_tokens, used_custom_terms, hijack_comments, hijack_fixes, token_count = self.process_text(text) - self.hijack.fixes = hijack_fixes self.hijack.comments = hijack_comments @@ -517,14 +290,19 @@ class EmbeddingsWithFixes(torch.nn.Module): inputs_embeds = self.wrapped(input_ids) - if batch_fixes is not None: - for fixes, tensor in zip(batch_fixes, inputs_embeds): - for offset, word in fixes: - emb = self.embeddings.word_embeddings[word] - emb_len = min(tensor.shape[0]-offset-1, emb.shape[0]) - tensor[offset+1:offset+1+emb_len] = self.embeddings.word_embeddings[word][0:emb_len] + if batch_fixes is None or len(batch_fixes) == 0 or max([len(x) for x in batch_fixes]) == 0: + return inputs_embeds + + vecs = [] + for fixes, tensor in zip(batch_fixes, inputs_embeds): + for offset, embedding in fixes: + emb = embedding.vec + emb_len = min(tensor.shape[0]-offset-1, emb.shape[0]) + tensor = torch.cat([tensor[0:offset+1], emb[0:emb_len], tensor[offset+1+emb_len:]]) + + vecs.append(tensor) - return inputs_embeds + return torch.stack(vecs) def add_circular_option_to_conv_2d(): -- cgit v1.2.1