update lora logic

TODO.md

@@ -25,8 +25,6 @@ Stuff to be added...
 
 Stuff to be investigated...
 
-- Best practices for **Apple M1**
-- Best practices for **AMD GPUs**
 - Torch Compile
 - `Torch-DirectML`
 - `TensorRT`
@@ -64,6 +62,7 @@ Tech that can be integrated as part of the core workflow...
   should fix most issues with NaN or black images  
 - add built-in Gradio themes  
 - fix setup race conditions
-- fix requirements  
+- reduce requirements  
 - more AMD specific work
+- initial work on Apple platform support
 - additional PR merges

extensions-builtin/Lora/lora.py

@@ -1,8 +1,8 @@
 import glob
 import os
 import re
-import torch
 from typing import Union
+import torch
 
 from modules import shared, devices, sd_models, errors
 
@@ -139,9 +139,7 @@ def load_lora(name, filename):
         lora_key_parts = key_diffusers.split(".", 1)
         key_diffusers_without_lora_parts = lora_key_parts[0]
         lora_key = lora_key_parts[1] if len(lora_key_parts) > 1 else ""
-        
         key = convert_diffusers_name_to_compvis(key_diffusers_without_lora_parts, is_sd2)
-
         sd_module = shared.sd_model.lora_layer_mapping.get(key, None)
 
         if sd_module is None:
@@ -173,7 +171,6 @@ def load_lora(name, filename):
         else:
             print(f'Lora layer {key_diffusers} matched a layer with unsupported type: {type(sd_module).__name__}')
             continue
-            assert False, f'Lora layer {key_diffusers} matched a layer with unsupported type: {type(sd_module).__name__}'
 
         with torch.no_grad():
             module.weight.copy_(weight)
@@ -321,10 +318,10 @@ def lora_reset_cached_weight(self: Union[torch.nn.Conv2d, torch.nn.Linear]):
     setattr(self, "lora_weights_backup", None)
 
 
-def lora_Linear_forward(self, input):
+def lora_Linear_forward(self, lora_input):
     lora_apply_weights(self)
 
-    return torch.nn.Linear_forward_before_lora(self, input)
+    return torch.nn.Linear_forward_before_lora(self, lora_input)
 
 
 def lora_Linear_load_state_dict(self, *args, **kwargs):
@@ -333,10 +330,10 @@ def lora_Linear_load_state_dict(self, *args, **kwargs):
     return torch.nn.Linear_load_state_dict_before_lora(self, *args, **kwargs)
 
 
-def lora_Conv2d_forward(self, input):
+def lora_Conv2d_forward(self, lora_input):
     lora_apply_weights(self)
 
-    return torch.nn.Conv2d_forward_before_lora(self, input)
+    return torch.nn.Conv2d_forward_before_lora(self, lora_input)
 
 
 def lora_Conv2d_load_state_dict(self, *args, **kwargs):