support alphafold v2.3 param (#128)

fastfold/config.py

@@ -575,7 +575,7 @@ multimer_model_config_update = {
         "tm": {
             "c_z": c_z,
             "no_bins": aux_distogram_bins,
-            "enabled": tm_enabled,
+            "enabled": True,
         },
         "masked_msa": {
             "c_m": c_m,

fastfold/model/nn/triangular_multiplicative_update.py

@@ -22,6 +22,16 @@ import torch.nn as nn
 from fastfold.model.nn.primitives import Linear, LayerNorm
 from fastfold.utils.tensor_utils import permute_final_dims
 
+_FUSED_TRIANGLE_MULTIPLICATION = False
+
+
+def set_fused_triangle_multiplication():
+    global _FUSED_TRIANGLE_MULTIPLICATION
+    _FUSED_TRIANGLE_MULTIPLICATION = True
+
+def is_fused_triangle_multiplication():
+    global _FUSED_TRIANGLE_MULTIPLICATION
+    return _FUSED_TRIANGLE_MULTIPLICATION
 
 class TriangleMultiplicativeUpdate(nn.Module):
     """
@@ -40,6 +50,11 @@ class TriangleMultiplicativeUpdate(nn.Module):
         self.c_hidden = c_hidden
         self._outgoing = _outgoing
 
+        if _FUSED_TRIANGLE_MULTIPLICATION:
+            self.linear_p = Linear(self.c_z, 2 * self.c_hidden)
+            self.linear_g = Linear(self.c_z, 2 * self.c_hidden, init="gating")
+            self.linear_gate = Linear(self.c_z, self.c_z, init="gating")
+        else:
             self.linear_a_p = Linear(self.c_z, self.c_hidden)
             self.linear_a_g = Linear(self.c_z, self.c_hidden, init="gating")
             self.linear_b_p = Linear(self.c_z, self.c_hidden)
@@ -77,13 +92,24 @@ class TriangleMultiplicativeUpdate(nn.Module):
         mask = mask.unsqueeze(-1)
 
         z = self.layer_norm_in(z)
+        if _FUSED_TRIANGLE_MULTIPLICATION:
+            a = self.linear_p(z) * mask
+            a = self.sigmoid(self.linear_g(z))
+
+            a, b = a.chunk(2, dim=-1)
+        else:
             a = self.linear_a_p(z) * self.sigmoid(self.linear_a_g(z))
             a = a * mask
             b = self.linear_b_p(z) * self.sigmoid(self.linear_b_g(z))
             b = b * mask
+
         x = self._combine_projections(a, b)
         x = self.layer_norm_out(x)
         x = self.linear_z(x)
+        
+        if _FUSED_TRIANGLE_MULTIPLICATION:
+            g = self.sigmoid(self.linear_gate(z))
+        else:
             g = self.sigmoid(self.linear_g(z))
         z = x * g
 

fastfold/utils/import_weights.py

@@ -20,6 +20,7 @@ import numpy as np
 import torch
 from typing import Union, List
 
+from fastfold.model.nn.triangular_multiplicative_update import is_fused_triangle_multiplication
 
 _NPZ_KEY_PREFIX = "alphafold/alphafold_iteration/"
 
@@ -188,6 +189,29 @@ def get_translation_dict(model, version):
         "attention": AttentionGatedParams(tri_att.mha),
     }
     
+    if is_fused_triangle_multiplication():
+        TriMulOutParams = lambda tri_mul: {
+            "left_norm_input": LayerNormParams(tri_mul.layer_norm_in),
+            "projection": LinearParams(tri_mul.linear_p),
+            "gate": LinearParams(tri_mul.linear_g),
+            "center_norm": LayerNormParams(tri_mul.layer_norm_out),
+            "output_projection": LinearParams(tri_mul.linear_z),
+            "gating_linear": LinearParams(tri_mul.linear_gate),
+        }
+
+        # see commit b88f8da on the Alphafold repo
+        # Alphafold swaps the pseudocode's a and b between the incoming/outcoming
+        # iterations of triangle multiplication, which is confusing and not
+        # reproduced in our implementation.
+        TriMulInParams = lambda tri_mul: {
+            "left_norm_input": LayerNormParams(tri_mul.layer_norm_in),
+            "projection": LinearParams(tri_mul.linear_p),
+            "gate": LinearParams(tri_mul.linear_g),
+            "center_norm": LayerNormParams(tri_mul.layer_norm_out),
+            "output_projection": LinearParams(tri_mul.linear_z),
+            "gating_linear": LinearParams(tri_mul.linear_gate),
+        }
+    else:
         TriMulOutParams = lambda tri_mul: {
             "layer_norm_input": LayerNormParams(tri_mul.layer_norm_in),
             "left_projection": LinearParams(tri_mul.linear_a_p),
@@ -553,7 +577,7 @@ def get_translation_dict(model, version):
             if "template_" in k:
                 evo_dict.pop(k)
 
-    if "_ptm" in version:
+    if "_ptm" in version or is_multimer:
         translations["predicted_aligned_error_head"] = {
             "logits": LinearParams(model.aux_heads.tm.linear)
         }

fastfold/utils/inject_fastnn.py

@@ -18,6 +18,7 @@ from fastfold.model.fastnn import EvoformerStack, ExtraMSAStack
 from fastfold.model.fastnn.embedders import TemplateEmbedder
 from fastfold.model.fastnn.embedders_multimer import TemplateEmbedderMultimer
 from fastfold.model.fastnn.ops import RecyclingEmbedder, InputEmbedder
+from fastfold.model.nn.triangular_multiplicative_update import is_fused_triangle_multiplication
 
 
 def copy_layernorm(model_fast, model_ori):
@@ -72,12 +73,17 @@ def copy_transition(model_fast, model_ori):
 def copy_triangle(model_fast, model_ori):
     copy_layernorm(model_fast.layernorm1, model_ori.layer_norm_in)
     copy_layernorm(model_fast.layernorm2, model_ori.layer_norm_out)
-    copy_linear(model_fast.output_gate, model_ori.linear_g)
+    
     copy_linear(model_fast.output_projection, model_ori.linear_z)
     model_fast.output_bias.copy_(model_ori.linear_z.bias)
 
+    if is_fused_triangle_multiplication():
+        copy_linear(model_fast.output_gate, model_ori.linear_gate)
+        copy_linear(model_fast.left_right_projection, model_ori.linear_p)
+        copy_linear(model_fast.left_right_gate, model_ori.linear_g)
+    else:
+        copy_linear(model_fast.output_gate, model_ori.linear_g)
         copy_left_right(model_fast.left_right_projection, model_ori.linear_a_p, model_ori.linear_b_p)
-
         copy_left_right(model_fast.left_right_gate, model_ori.linear_a_g, model_ori.linear_b_g)
 
 

inference.py

@@ -34,6 +34,7 @@ import fastfold.relax.relax as relax
 from fastfold.common import protein, residue_constants
 from fastfold.config import model_config
 from fastfold.model.fastnn import set_chunk_size
+from fastfold.model.nn.triangular_multiplicative_update import set_fused_triangle_multiplication
 from fastfold.data import data_pipeline, feature_pipeline, templates
 from fastfold.data.tools import hhsearch, hmmsearch
 from fastfold.workflow.template import FastFoldDataWorkFlow, FastFoldMultimerDataWorkFlow
@@ -117,6 +118,10 @@ def inference_model(rank, world_size, result_q, batch, args):
     config = model_config(args.model_name)
     if args.chunk_size:
         config.globals.chunk_size = args.chunk_size
+
+    if "v3" in args.param_path:
+        set_fused_triangle_multiplication()
+
     config.globals.inplace = args.inplace
     config.globals.is_multimer = args.model_preset == 'multimer'
     model = AlphaFold(config)