import torch import numpy as np import re import itertools from comfy import model_management from comfy.sdxl_clip import SDXLClipModel, SDXLRefinerClipModel, SDXLClipG try: from comfy.text_encoders.sd3_clip import SD3ClipModel, T5XXLModel except ImportError: from comfy.sd3_clip import SD3ClipModel, T5XXLModel from nodes import NODE_CLASS_MAPPINGS, ConditioningConcat, ConditioningZeroOut, ConditioningSetTimestepRange, ConditioningCombine def _grouper(n, iterable): it = iter(iterable) while True: chunk = list(itertools.islice(it, n)) if not chunk: return yield chunk def _norm_mag(w, n): d = w - 1 return 1 + np.sign(d) * np.sqrt(np.abs(d) ** 2 / n) # return np.sign(w) * np.sqrt(np.abs(w)**2 / n) def divide_length(word_ids, weights): sums = dict(zip(*np.unique(word_ids, return_counts=True))) sums[0] = 1 weights = [[_norm_mag(w, sums[id]) if id != 0 else 1.0 for w, id in zip(x, y)] for x, y in zip(weights, word_ids)] return weights def shift_mean_weight(word_ids, weights): delta = 1 - np.mean([w for x, y in zip(weights, word_ids) for w, id in zip(x, y) if id != 0]) weights = [[w if id == 0 else w + delta for w, id in zip(x, y)] for x, y in zip(weights, word_ids)] return weights def scale_to_norm(weights, word_ids, w_max): top = np.max(weights) w_max = min(top, w_max) weights = [[w_max if id == 0 else (w / top) * w_max for w, id in zip(x, y)] for x, y in zip(weights, word_ids)] return weights def from_zero(weights, base_emb): weight_tensor = torch.tensor(weights, dtype=base_emb.dtype, device=base_emb.device) weight_tensor = weight_tensor.reshape(1, -1, 1).expand(base_emb.shape) return base_emb * weight_tensor def mask_word_id(tokens, word_ids, target_id, mask_token): new_tokens = [[mask_token if wid == target_id else t for t, wid in zip(x, y)] for x, y in zip(tokens, word_ids)] mask = np.array(word_ids) == target_id return (new_tokens, mask) def batched_clip_encode(tokens, length, encode_func, num_chunks): embs = [] for e in _grouper(32, tokens): enc, pooled = encode_func(e) enc = enc.reshape((len(e), length, -1)) embs.append(enc) embs = torch.cat(embs) embs = embs.reshape((len(tokens) // num_chunks, length * num_chunks, -1)) return embs def from_masked(tokens, weights, word_ids, base_emb, length, encode_func, m_token=266): pooled_base = base_emb[0, length - 1:length, :] wids, inds = np.unique(np.array(word_ids).reshape(-1), return_index=True) weight_dict = dict((id, w) for id, w in zip(wids, np.array(weights).reshape(-1)[inds]) if w != 1.0) if len(weight_dict) == 0: return torch.zeros_like(base_emb), base_emb[0, length - 1:length, :] weight_tensor = torch.tensor(weights, dtype=base_emb.dtype, device=base_emb.device) weight_tensor = weight_tensor.reshape(1, -1, 1).expand(base_emb.shape) # m_token = (clip.tokenizer.end_token, 1.0) if clip.tokenizer.pad_with_end else (0,1.0) # TODO: find most suitable masking token here m_token = (m_token, 1.0) ws = [] masked_tokens = [] masks = [] # create prompts for id, w in weight_dict.items(): masked, m = mask_word_id(tokens, word_ids, id, m_token) masked_tokens.extend(masked) m = torch.tensor(m, dtype=base_emb.dtype, device=base_emb.device) m = m.reshape(1, -1, 1).expand(base_emb.shape) masks.append(m) ws.append(w) # batch process prompts embs = batched_clip_encode(masked_tokens, length, encode_func, len(tokens)) masks = torch.cat(masks) embs = (base_emb.expand(embs.shape) - embs) pooled = embs[0, length - 1:length, :] embs *= masks embs = embs.sum(axis=0, keepdim=True) pooled_start = pooled_base.expand(len(ws), -1) ws = torch.tensor(ws).reshape(-1, 1).expand(pooled_start.shape) pooled = (pooled - pooled_start) * (ws - 1) pooled = pooled.mean(axis=0, keepdim=True) return ((weight_tensor - 1) * embs), pooled_base + pooled def mask_inds(tokens, inds, mask_token): clip_len = len(tokens[0]) inds_set = set(inds) new_tokens = [[mask_token if i * clip_len + j in inds_set else t for j, t in enumerate(x)] for i, x in enumerate(tokens)] return new_tokens def down_weight(tokens, weights, word_ids, base_emb, length, encode_func, m_token=266): w, w_inv = np.unique(weights, return_inverse=True) if np.sum(w < 1) == 0: return base_emb, tokens, base_emb[0, length - 1:length, :] # m_token = (clip.tokenizer.end_token, 1.0) if clip.tokenizer.pad_with_end else (0,1.0) # using the comma token as a masking token seems to work better than aos tokens for SD 1.x m_token = (m_token, 1.0) masked_tokens = [] masked_current = tokens for i in range(len(w)): if w[i] >= 1: continue masked_current = mask_inds(masked_current, np.where(w_inv == i)[0], m_token) masked_tokens.extend(masked_current) embs = batched_clip_encode(masked_tokens, length, encode_func, len(tokens)) embs = torch.cat([base_emb, embs]) w = w[w <= 1.0] w_mix = np.diff([0] + w.tolist()) w_mix = torch.tensor(w_mix, dtype=embs.dtype, device=embs.device).reshape((-1, 1, 1)) weighted_emb = (w_mix * embs).sum(axis=0, keepdim=True) return weighted_emb, masked_current, weighted_emb[0, length - 1:length, :] def scale_emb_to_mag(base_emb, weighted_emb): norm_base = torch.linalg.norm(base_emb) norm_weighted = torch.linalg.norm(weighted_emb) embeddings_final = (norm_base / norm_weighted) * weighted_emb return embeddings_final def recover_dist(base_emb, weighted_emb): fixed_std = (base_emb.std() / weighted_emb.std()) * (weighted_emb - weighted_emb.mean()) embeddings_final = fixed_std + (base_emb.mean() - fixed_std.mean()) return embeddings_final def A1111_renorm(base_emb, weighted_emb): embeddings_final = (base_emb.mean() / weighted_emb.mean()) * weighted_emb return embeddings_final def advanced_encode_from_tokens(tokenized, token_normalization, weight_interpretation, encode_func, m_token=266, length=77, w_max=1.0, return_pooled=False, apply_to_pooled=False): tokens = [[t for t, _, _ in x] for x in tokenized] weights = [[w for _, w, _ in x] for x in tokenized] word_ids = [[wid for _, _, wid in x] for x in tokenized] # weight normalization # ==================== # distribute down/up weights over word lengths if token_normalization.startswith("length"): weights = divide_length(word_ids, weights) # make mean of word tokens 1 if token_normalization.endswith("mean"): weights = shift_mean_weight(word_ids, weights) # weight interpretation # ===================== pooled = None if weight_interpretation == "comfy": weighted_tokens = [[(t, w) for t, w in zip(x, y)] for x, y in zip(tokens, weights)] weighted_emb, pooled_base = encode_func(weighted_tokens) pooled = pooled_base else: unweighted_tokens = [[(t, 1.0) for t, _, _ in x] for x in tokenized] base_emb, pooled_base = encode_func(unweighted_tokens) if weight_interpretation == "A1111": weighted_emb = from_zero(weights, base_emb) weighted_emb = A1111_renorm(base_emb, weighted_emb) pooled = pooled_base if weight_interpretation == "compel": pos_tokens = [[(t, w) if w >= 1.0 else (t, 1.0) for t, w in zip(x, y)] for x, y in zip(tokens, weights)] weighted_emb, _ = encode_func(pos_tokens) weighted_emb, _, pooled = down_weight(pos_tokens, weights, word_ids, weighted_emb, length, encode_func) if weight_interpretation == "comfy++": weighted_emb, tokens_down, _ = down_weight(unweighted_tokens, weights, word_ids, base_emb, length, encode_func) weights = [[w if w > 1.0 else 1.0 for w in x] for x in weights] # unweighted_tokens = [[(t,1.0) for t, _,_ in x] for x in tokens_down] embs, pooled = from_masked(unweighted_tokens, weights, word_ids, base_emb, length, encode_func) weighted_emb += embs if weight_interpretation == "down_weight": weights = scale_to_norm(weights, word_ids, w_max) weighted_emb, _, pooled = down_weight(unweighted_tokens, weights, word_ids, base_emb, length, encode_func) if return_pooled: if apply_to_pooled: return weighted_emb, pooled else: return weighted_emb, pooled_base return weighted_emb, None def encode_token_weights_g(model, token_weight_pairs): return model.clip_g.encode_token_weights(token_weight_pairs) def encode_token_weights_l(model, token_weight_pairs): l_out, pooled = model.clip_l.encode_token_weights(token_weight_pairs) return l_out, pooled def encode_token_weights_t5(model, token_weight_pairs): return model.t5xxl.encode_token_weights(token_weight_pairs) def encode_token_weights(model, token_weight_pairs, encode_func): if model.layer_idx is not None: # 2016 [c2cb8e88] 及以上版本去除了sdxl clip的clip_layer方法 # if compare_revision(2016): model.cond_stage_model.set_clip_options({'layer': model.layer_idx}) # else: # model.cond_stage_model.clip_layer(model.layer_idx) model_management.load_model_gpu(model.patcher) return encode_func(model.cond_stage_model, token_weight_pairs) def prepareXL(embs_l, embs_g, pooled, clip_balance): l_w = 1 - max(0, clip_balance - .5) * 2 g_w = 1 - max(0, .5 - clip_balance) * 2 if embs_l is not None: return torch.cat([embs_l * l_w, embs_g * g_w], dim=-1), pooled else: return embs_g, pooled def prepareSD3(out, pooled, clip_balance): lg_w = 1 - max(0, clip_balance - .5) * 2 t5_w = 1 - max(0, .5 - clip_balance) * 2 if out.shape[0] > 1: return torch.cat([out[0] * lg_w, out[1] * t5_w], dim=-1), pooled else: return out, pooled def advanced_encode(clip, text, token_normalization, weight_interpretation, w_max=1.0, clip_balance=.5, apply_to_pooled=True, width=1024, height=1024, crop_w=0, crop_h=0, target_width=1024, target_height=1024, a1111_prompt_style=False, steps=1): # Use clip text encode by smzNodes like same as a1111, when if you need installed the smzNodes if a1111_prompt_style: if "smZ CLIPTextEncode" in NODE_CLASS_MAPPINGS: cls = NODE_CLASS_MAPPINGS['smZ CLIPTextEncode'] embeddings_final, = cls().encode(clip, text, weight_interpretation, True, True, False, False, 6, 1024, 1024, 0, 0, 1024, 1024, '', '', steps) return embeddings_final else: raise Exception(f"[smzNodes Not Found] you need to install 'ComfyUI-smzNodes'") time_start = 0 time_end = 1 match = re.search(r'TIMESTEP.*$', text) if match: timestep = match.group() timestep = timestep.split(' ') timestep = timestep[0] text = text.replace(timestep, '') value = timestep.split(':') if len(value) >= 3: time_start = float(value[1]) time_end = float(value[2]) elif len(value) == 2: time_start = float(value[1]) time_end = 1 elif len(value) == 1: time_start = 0.1 time_end = 1 pass3 = [x.strip() for x in text.split("BREAK")] pass3 = [x for x in pass3 if x != ''] if len(pass3) == 0: pass3 = [''] # pass3_str = [f'[{x}]' for x in pass3] # print(f"CLIP: {str.join(' + ', pass3_str)}") conditioning = None for text in pass3: tokenized = clip.tokenize(text, return_word_ids=True) if SD3ClipModel and isinstance(clip.cond_stage_model, SD3ClipModel): lg_out = None pooled = None out = None if len(tokenized['l']) > 0 or len(tokenized['g']) > 0: if clip.cond_stage_model.clip_l is not None: lg_out, l_pooled = advanced_encode_from_tokens(tokenized['l'], token_normalization, weight_interpretation, lambda x: encode_token_weights(clip, x, encode_token_weights_l), w_max=w_max, return_pooled=True,) else: l_pooled = torch.zeros((1, 768), device=model_management.intermediate_device()) if clip.cond_stage_model.clip_g is not None: g_out, g_pooled = advanced_encode_from_tokens(tokenized['g'], token_normalization, weight_interpretation, lambda x: encode_token_weights(clip, x, encode_token_weights_g), w_max=w_max, return_pooled=True) if lg_out is not None: lg_out = torch.cat([lg_out, g_out], dim=-1) else: lg_out = torch.nn.functional.pad(g_out, (768, 0)) else: g_out = None g_pooled = torch.zeros((1, 1280), device=model_management.intermediate_device()) if lg_out is not None: lg_out = torch.nn.functional.pad(lg_out, (0, 4096 - lg_out.shape[-1])) out = lg_out pooled = torch.cat((l_pooled, g_pooled), dim=-1) # t5xxl if 't5xxl' in tokenized: t5_out, t5_pooled = advanced_encode_from_tokens(tokenized['t5xxl'], token_normalization, weight_interpretation, lambda x: encode_token_weights(clip, x, encode_token_weights_t5), w_max=w_max, return_pooled=True) if lg_out is not None: out = torch.cat([lg_out, t5_out], dim=-2) else: out = t5_out if out is None: out = torch.zeros((1, 77, 4096), device=model_management.intermediate_device()) if pooled is None: pooled = torch.zeros((1, 768 + 1280), device=model_management.intermediate_device()) embeddings_final, pooled = prepareSD3(out, pooled, clip_balance) cond = [[embeddings_final, {"pooled_output": pooled}]] elif isinstance(clip.cond_stage_model, (SDXLClipModel, SDXLRefinerClipModel, SDXLClipG)): embs_l = None embs_g = None pooled = None if 'l' in tokenized and isinstance(clip.cond_stage_model, SDXLClipModel): embs_l, _ = advanced_encode_from_tokens(tokenized['l'], token_normalization, weight_interpretation, lambda x: encode_token_weights(clip, x, encode_token_weights_l), w_max=w_max, return_pooled=False) if 'g' in tokenized: embs_g, pooled = advanced_encode_from_tokens(tokenized['g'], token_normalization, weight_interpretation, lambda x: encode_token_weights(clip, x, encode_token_weights_g), w_max=w_max, return_pooled=True, apply_to_pooled=apply_to_pooled) embeddings_final, pooled = prepareXL(embs_l, embs_g, pooled, clip_balance) cond = [[embeddings_final, {"pooled_output": pooled}]] # cond = [[embeddings_final, # {"pooled_output": pooled, "width": width, "height": height, "crop_w": crop_w, # "crop_h": crop_h, "target_width": target_width, "target_height": target_height}]] else: embeddings_final, pooled = advanced_encode_from_tokens(tokenized['l'], token_normalization, weight_interpretation, lambda x: encode_token_weights(clip, x, encode_token_weights_l), w_max=w_max,return_pooled=True,) cond = [[embeddings_final, {"pooled_output": pooled}]] if conditioning is not None: conditioning = ConditioningConcat().concat(conditioning, cond)[0] else: conditioning = cond # setTimeStepRange if time_start > 0 or time_end < 1: conditioning_2, = ConditioningSetTimestepRange().set_range(conditioning, 0, time_start) conditioning_1, = ConditioningZeroOut().zero_out(conditioning) conditioning_1, = ConditioningSetTimestepRange().set_range(conditioning_1, time_start, time_end) conditioning, = ConditioningCombine().combine(conditioning_1, conditioning_2) return conditioning