From c0484f1b986ce7acb0e3596f6089a191279f5442 Mon Sep 17 00:00:00 2001 From: brkirch Date: Mon, 10 Oct 2022 22:48:54 -0400 Subject: Add cross-attention optimization from InvokeAI * Add cross-attention optimization from InvokeAI (~30% speed improvement on MPS) * Add command line option for it * Make it default when CUDA is unavailable --- modules/sd_hijack.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'modules/sd_hijack.py') diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index f07ec041..5a1b167f 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -30,8 +30,11 @@ def apply_optimizations(): elif cmd_opts.opt_split_attention_v1: print("Applying v1 cross attention optimization.") 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_invokeai or not torch.cuda.is_available()): + print("Applying cross attention optimization (InvokeAI).") + ldm.modules.attention.CrossAttention.forward = sd_hijack_optimizations.split_cross_attention_forward_invokeAI elif not cmd_opts.disable_opt_split_attention and (cmd_opts.opt_split_attention or torch.cuda.is_available()): - print("Applying cross attention optimization.") + print("Applying cross attention optimization (Doggettx).") ldm.modules.attention.CrossAttention.forward = sd_hijack_optimizations.split_cross_attention_forward ldm.modules.diffusionmodules.model.AttnBlock.forward = sd_hijack_optimizations.cross_attention_attnblock_forward -- cgit v1.2.3 From 98fd5cde72d5bda1620ab78416c7828fdc3dc10b Mon Sep 17 00:00:00 2001 From: brkirch Date: Mon, 10 Oct 2022 23:55:48 -0400 Subject: Add check for psutil --- modules/sd_hijack.py | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'modules/sd_hijack.py') diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index 5a1b167f..ac70f876 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -10,6 +10,7 @@ from torch.nn.functional import silu 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 modules.sd_hijack_optimizations import invokeAI_mps_available import ldm.modules.attention import ldm.modules.diffusionmodules.model @@ -31,8 +32,13 @@ def apply_optimizations(): print("Applying v1 cross attention optimization.") 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_invokeai or not torch.cuda.is_available()): - print("Applying cross attention optimization (InvokeAI).") - ldm.modules.attention.CrossAttention.forward = sd_hijack_optimizations.split_cross_attention_forward_invokeAI + if not invokeAI_mps_available and shared.device.type == 'mps': + print("The InvokeAI cross attention optimization for MPS requires the psutil package which is not installed.") + print("Applying v1 cross attention optimization.") + ldm.modules.attention.CrossAttention.forward = sd_hijack_optimizations.split_cross_attention_forward_v1 + else: + print("Applying cross attention optimization (InvokeAI).") + ldm.modules.attention.CrossAttention.forward = sd_hijack_optimizations.split_cross_attention_forward_invokeAI elif not cmd_opts.disable_opt_split_attention and (cmd_opts.opt_split_attention or torch.cuda.is_available()): print("Applying cross attention optimization (Doggettx).") ldm.modules.attention.CrossAttention.forward = sd_hijack_optimizations.split_cross_attention_forward -- cgit v1.2.3 From 80f3cf2bb2ce3f00d801cae2c3a8c20a8d4167d8 Mon Sep 17 00:00:00 2001 From: hentailord85ez <112723046+hentailord85ez@users.noreply.github.com> Date: Tue, 11 Oct 2022 19:48:53 +0100 Subject: Account when lines are mismatched --- modules/sd_hijack.py | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) (limited to 'modules/sd_hijack.py') diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index ac70f876..2753d4fa 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -321,7 +321,17 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): fixes.append(fix[1]) self.hijack.fixes.append(fixes) - z1 = self.process_tokens([x[:75] for x in remade_batch_tokens], [x[:75] for x in batch_multipliers]) + tokens = [] + multipliers = [] + for i in range(len(remade_batch_tokens)): + if len(remade_batch_tokens[i]) > 0: + tokens.append(remade_batch_tokens[i][:75]) + multipliers.append(batch_multipliers[i][:75]) + else: + tokens.append([self.wrapped.tokenizer.eos_token_id] * 75) + multipliers.append([1.0] * 75) + + z1 = self.process_tokens(tokens, multipliers) z = z1 if z is None else torch.cat((z, z1), axis=-2) remade_batch_tokens = rem_tokens -- cgit v1.2.3 From 429442f4a6aab7301efb89d27bef524fe827e81a Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Wed, 12 Oct 2022 13:38:03 +0300 Subject: fix iterator bug for #2295 --- modules/sd_hijack.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'modules/sd_hijack.py') diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index 2753d4fa..c81722a0 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -323,10 +323,10 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): tokens = [] multipliers = [] - for i in range(len(remade_batch_tokens)): - if len(remade_batch_tokens[i]) > 0: - tokens.append(remade_batch_tokens[i][:75]) - multipliers.append(batch_multipliers[i][:75]) + for j in range(len(remade_batch_tokens)): + if len(remade_batch_tokens[j]) > 0: + tokens.append(remade_batch_tokens[j][:75]) + multipliers.append(batch_multipliers[j][:75]) else: tokens.append([self.wrapped.tokenizer.eos_token_id] * 75) multipliers.append([1.0] * 75) -- cgit v1.2.3