feat: LyCORIS/kohya OFT network support

1 files changed, 26 insertions, 82 deletions

diff --git a/extensions-builtin/Lora/network_oft.py b/extensions-builtin/Lora/network_oft.py
index c45a8d23..05c37811 100644
--- a/extensions-builtin/Lora/network_oft.py
+++ b/extensions-builtin/Lora/network_oft.py
@@ -11,8 +11,8 @@ class ModuleTypeOFT(network.ModuleType):
 
         return None
 
-# adapted from kohya-ss' implementation https://github.com/kohya-ss/sd-scripts/blob/main/networks/oft.py
-# and KohakuBlueleaf's implementation https://github.com/KohakuBlueleaf/LyCORIS/blob/dev/lycoris/modules/diag_oft.py
+# Supports both kohya-ss' implementation of COFT  https://github.com/kohya-ss/sd-scripts/blob/main/networks/oft.py
+# and KohakuBlueleaf's implementation of OFT/COFT https://github.com/KohakuBlueleaf/LyCORIS/blob/dev/lycoris/modules/diag_oft.py
 class NetworkModuleOFT(network.NetworkModule):
     def __init__(self,  net: network.Network, weights: network.NetworkWeights):
 
@@ -25,117 +25,61 @@ class NetworkModuleOFT(network.NetworkModule):
         if "oft_blocks" in weights.w.keys():
             self.is_kohya = True
             self.oft_blocks = weights.w["oft_blocks"] # (num_blocks, block_size, block_size)
-            self.alpha = weights.w["alpha"]
+            self.alpha = weights.w["alpha"] # alpha is constraint
             self.dim = self.oft_blocks.shape[0] # lora dim
-            #self.oft_blocks = rearrange(self.oft_blocks, 'k m ... -> (k m) ...')
+        # LyCORIS
         elif "oft_diag" in weights.w.keys():
             self.is_kohya = False
-            self.oft_blocks = weights.w["oft_diag"] # (num_blocks, block_size, block_size)
-
-            # alpha is rank if alpha is 0 or None
-            if self.alpha is None:
-                pass
-            self.dim = self.oft_blocks.shape[1] # FIXME: almost certainly incorrect, assumes tensor is shape [*, m, n]
-        else:
-            raise ValueError("oft_blocks or oft_diag must be in weights dict")
+            self.oft_blocks = weights.w["oft_diag"]
+            # self.alpha is unused
+            self.dim = self.oft_blocks.shape[1] # (num_blocks, block_size, block_size)
 
         is_linear = type(self.sd_module) in [torch.nn.Linear, torch.nn.modules.linear.NonDynamicallyQuantizableLinear]
         is_conv = type(self.sd_module) in [torch.nn.Conv2d]
-        is_other_linear = type(self.sd_module) in [torch.nn.MultiheadAttention]
+        is_other_linear = type(self.sd_module) in [torch.nn.MultiheadAttention] # unsupported
 
         if is_linear:
             self.out_dim = self.sd_module.out_features
-        elif is_other_linear:
-            self.out_dim = self.sd_module.embed_dim
         elif is_conv:
             self.out_dim = self.sd_module.out_channels
-        else:
-            raise ValueError("sd_module must be Linear or Conv")
+        elif is_other_linear:
+            self.out_dim = self.sd_module.embed_dim
 
         if self.is_kohya:
             self.constraint = self.alpha * self.out_dim
-            self.num_blocks, self.block_size = factorization(self.out_dim, self.dim)
+            self.num_blocks = self.dim
+            self.block_size = self.out_dim // self.dim
         else:
             self.constraint = None
             self.block_size, self.num_blocks = factorization(self.out_dim, self.dim)
 
-    def merge_weight(self, R_weight, org_weight):
-        R_weight = R_weight.to(org_weight.device, dtype=org_weight.dtype)
-        if org_weight.dim() == 4:
-            weight = torch.einsum("oihw, op -> pihw", org_weight, R_weight)
-        else:
-            weight = torch.einsum("oi, op -> pi", org_weight, R_weight)
-        return weight
-
-    def get_weight(self, oft_blocks, multiplier=None):
-        if self.constraint is not None:
-            constraint = self.constraint.to(oft_blocks.device, dtype=oft_blocks.dtype)
-
-        block_Q = oft_blocks - oft_blocks.transpose(1, 2)
-        norm_Q = torch.norm(block_Q.flatten())
-        if self.constraint is not None:
-            new_norm_Q = torch.clamp(norm_Q, max=constraint)
-        else:
-            new_norm_Q = norm_Q
-        block_Q = block_Q * ((new_norm_Q + 1e-8) / (norm_Q + 1e-8))
-        m_I = torch.eye(self.num_blocks, device=oft_blocks.device).unsqueeze(0).repeat(self.block_size, 1, 1)
-        #m_I = torch.eye(self.block_size, device=oft_blocks.device).unsqueeze(0).repeat(self.num_blocks, 1, 1)
-        block_R = torch.matmul(m_I + block_Q, (m_I - block_Q).inverse())
+    def calc_updown_kb(self, orig_weight, multiplier):
+        oft_blocks = self.oft_blocks.to(orig_weight.device, dtype=orig_weight.dtype)
+        oft_blocks = oft_blocks - oft_blocks.transpose(1, 2) # ensure skew-symmetric orthogonal matrix
 
-        block_R_weighted = multiplier * block_R + (1 - multiplier) * m_I
-        R = torch.block_diag(*block_R_weighted)
-        return R
+        R = oft_blocks.to(orig_weight.device, dtype=orig_weight.dtype)
+        R = R * multiplier + torch.eye(self.block_size, device=orig_weight.device)
 
-    def calc_updown_kohya(self, orig_weight, multiplier):
-        R = self.get_weight(self.oft_blocks, multiplier)
-        merged_weight = self.merge_weight(R, orig_weight)
+        # This errors out for MultiheadAttention, might need to be handled up-stream
+        merged_weight = rearrange(orig_weight, '(k n) ... -> k n ...', k=self.num_blocks, n=self.block_size)
+        merged_weight = torch.einsum(
+            'k n m, k n ... -> k m ...',
+            R,
+            merged_weight
+        )
+        merged_weight = rearrange(merged_weight, 'k m ... -> (k m) ...')
 
         updown = merged_weight.to(orig_weight.device, dtype=orig_weight.dtype) - orig_weight
         output_shape = orig_weight.shape
-        orig_weight = orig_weight
-        return self.finalize_updown(updown, orig_weight, output_shape)
-
-    def calc_updown_kb(self, orig_weight, multiplier):
-        is_other_linear = type(self.sd_module) in [torch.nn.MultiheadAttention]
-
-        if not is_other_linear:
-            oft_blocks = self.oft_blocks.to(orig_weight.device, dtype=orig_weight.dtype)
-
-            # ensure skew-symmetric matrix
-            oft_blocks = oft_blocks - oft_blocks.transpose(1, 2)
-
-            R = oft_blocks.to(orig_weight.device, dtype=orig_weight.dtype)
-            R = R * multiplier + torch.eye(self.block_size, device=orig_weight.device)
-
-            merged_weight = rearrange(orig_weight, '(k n) ... -> k n ...', k=self.num_blocks, n=self.block_size)
-            merged_weight = torch.einsum(
-                'k n m, k n ... -> k m ...',
-                R,
-                merged_weight
-            )
-            merged_weight = rearrange(merged_weight, 'k m ... -> (k m) ...')
-
-            updown = merged_weight.to(orig_weight.device, dtype=orig_weight.dtype) - orig_weight
-            output_shape = orig_weight.shape
-        else:
-            # FIXME: skip MultiheadAttention for now
-            #up = self.lin_module.weight.to(orig_weight.device, dtype=orig_weight.dtype)
-            updown = torch.zeros([orig_weight.shape[1], orig_weight.shape[1]], device=orig_weight.device, dtype=orig_weight.dtype)
-            output_shape = (orig_weight.shape[1], orig_weight.shape[1])
-
         return self.finalize_updown(updown, orig_weight, output_shape)
 
     def calc_updown(self, orig_weight):
-        # if alpha is a very small number as in coft, calc_scale will return a almost zero number so we ignore it
-        #multiplier = self.multiplier() * self.calc_scale()
+        # if alpha is a very small number as in coft, calc_scale() will return a almost zero number so we ignore it
         multiplier = self.multiplier()
-
         return self.calc_updown_kb(orig_weight, multiplier)
 
     # override to remove the multiplier/scale factor; it's already multiplied in get_weight
     def finalize_updown(self, updown, orig_weight, output_shape, ex_bias=None):
-        #return super().finalize_updown(updown, orig_weight, output_shape, ex_bias)
-
         if self.bias is not None:
             updown = updown.reshape(self.bias.shape)
             updown += self.bias.to(orig_weight.device, dtype=orig_weight.dtype)