[ViT] Support timm checkpoint, add tests

flash_attn/models/vit.py

@@ -1,9 +1,12 @@
 # Copyright (c) 2022, Tri Dao.
 # Inspired by / adapted from https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
 import math
+import re
 from functools import partial
 from copy import deepcopy
 
+from collections import OrderedDict
+
 import torch
 import torch.nn as nn
 import torch.nn.functional as F
@@ -218,6 +221,7 @@ class VisionTransformer(nn.Module):
         hidden_states = self._pos_embed(x)
         residual = None
         if self.global_pool != 'token' or all_tokens:
+        # if True:
             for block in self.blocks:
                 hidden_states, residual = block(hidden_states, residual)
         else:
@@ -225,10 +229,8 @@ class VisionTransformer(nn.Module):
                 hidden_states, residual = block(hidden_states, residual)
             # For the last layer, we only want the 1st token of the output. So we do cross-attention
             # where the query is the 1st token and the key/value is the whole sequence.
-            hidden_states_1st = rearrange(hidden_states[:, 0], 'b d -> b 1 d')
+            hidden_states, residual = self.blocks[-1](hidden_states, residual,
-            residual_1st = rearrange(residual[:, 0], 'b d -> b 1 d')
+                                                      mixer_subset=slice(0, 1))
-            hidden_states, residual = self.blocks[-1](hidden_states_1st, residual_1st,
-                                                      mixer_kwargs={'x_kv': hidden_states})
         if not self.fused_dropout_add_ln:
             residual = self.drop_path(self.dropout(hidden_states)) + residual
             hidden_states = self.norm(residual.to(dtype=self.norm.weight.dtype))
@@ -258,6 +260,29 @@ class VisionTransformer(nn.Module):
         x = self.forward_head(x)
         return x
 
+    def load_state_dict(self, state_dict, strict=True):
+        patch_embed_weight = state_dict['patch_embed.proj.weight']
+        if patch_embed_weight.dim() == 4:
+            # convert from Conv2d to Linear
+            state_dict['patch_embed.proj.weight'] = rearrange(patch_embed_weight,
+                                                              'o c h w -> o (c h w)')
+        def key_mapping_attn(key):
+            key = re.sub(r'^blocks.(\d+).attn.qkv.', r'blocks.\1.mixer.Wqkv.', key)
+            key = re.sub(r'^blocks.(\d+).attn.proj.', r'blocks.\1.mixer.out_proj.', key)
+            return key
+        state_dict = OrderedDict((key_mapping_attn(k), v) for k, v in state_dict.items())
+        n_layer = len(self.blocks)
+        # Convert from Wqkv to Wq and Wkv for cross attention (last layer)
+        if (self.blocks[-1].mixer.cross_attn
+            and f'blocks.{n_layer - 1}.mixer.Wqkv.weight' in state_dict):
+            Wqkv = state_dict.pop(f'blocks.{n_layer - 1}.mixer.Wqkv.weight')
+            bqkv = state_dict.pop(f'blocks.{n_layer - 1}.mixer.Wqkv.bias')
+            state_dict[f'blocks.{n_layer - 1}.mixer.Wq.weight'] = Wqkv[:self.embed_dim]
+            state_dict[f'blocks.{n_layer - 1}.mixer.Wkv.weight'] = Wqkv[self.embed_dim:]
+            state_dict[f'blocks.{n_layer - 1}.mixer.Wq.bias'] = bqkv[:self.embed_dim]
+            state_dict[f'blocks.{n_layer - 1}.mixer.Wkv.bias'] = bqkv[self.embed_dim:]
+        return super().load_state_dict(state_dict, strict=strict)
+
 
 def init_weights_vit_timm(module: nn.Module, name: str = ''):
     """ ViT weight initialization, original timm impl (for reproducibility) """

flash_attn/modules/block.py

@@ -89,12 +89,15 @@ class Block(nn.Module):
                     p._shared_params = True
 
     def forward(self, hidden_states: Tensor, residual: Optional[Tensor] = None,
-                mixer_kwargs=None):
+                mixer_subset=None, mixer_kwargs=None):
         r"""Pass the input through the encoder layer.
 
         Args:
             hidden_states: the sequence to the encoder layer (required).
             residual: if postnorm, residual=None, If prenorm, hidden_states = Attn/MLP(LN(residual))
+            mixer_subset: for cross-attention only. If not None, will take a subset of x
+                before applying the query projection. Useful for e.g., ViT where we only care
+                about the CLS token in the last layer.
         """
         if self.prenorm:
             if not self.fused_dropout_add_ln:
@@ -116,8 +119,12 @@ class Block(nn.Module):
                     self.dropout1.p if self.training else 0.0, self.norm1.eps,
                     rowscale=rowscale1, prenorm=True, residual_in_fp32=self.residual_in_fp32
                 )
-            hidden_states = self.mixer(hidden_states,
+            if mixer_kwargs is None:
-                                       **(mixer_kwargs if mixer_kwargs is not None else {}))
+                mixer_kwargs = {}
+            mixer_kwargs['mixer_subset'] = mixer_subset
+            hidden_states = self.mixer(hidden_states, **mixer_kwargs)
+            if mixer_subset is not None:
+                residual = residual[:, mixer_subset]
             if not isinstance(self.mlp, nn.Identity):
                 if not self.fused_dropout_add_ln:
                     dropped = self.drop_path2(self.dropout2(hidden_states))

flash_attn/modules/mha.py

@@ -420,7 +420,7 @@ class MHA(nn.Module):
         return _update_kv_cache(kv, inference_params, self.layer_idx)
 
     def forward(self, x, x_kv=None, key_padding_mask=None, cu_seqlens=None, max_seqlen=None,
-                inference_params=None, **kwargs):
+                mixer_subset=None, inference_params=None, **kwargs):
         """
         Arguments:
             x: (batch, seqlen, hidden_dim) (where hidden_dim = num heads * head dim) if
@@ -433,6 +433,9 @@ class MHA(nn.Module):
             max_seqlen: int. Maximum sequence length in the batch.
             key_padding_mask: boolean mask, True means to keep, False means to mask out.
                 (batch, seqlen). Only applicable when not using FlashAttention.
+            mixer_subset: for cross-attention only. If not None, will take a subset of x
+                before applying the query projection. Useful for e.g., ViT where we only care
+                about the CLS token in the last layer.
             inference_params: for generation. Adapted from Megatron-LM (and Apex)
             https://github.com/NVIDIA/apex/blob/3ff1a10f72ec07067c4e44759442329804ac5162/apex/transformer/testing/standalone_transformer_lm.py#L470
         """
@@ -454,6 +457,7 @@ class MHA(nn.Module):
         kwargs = ({'cu_seqlens': cu_seqlens, 'max_seqlen': max_seqlen, **kwargs}
                   if self.use_flash_attn else {'key_padding_mask': key_padding_mask, **kwargs})
         if not self.cross_attn:
+            assert x_kv is None and mixer_subset is None
             if not self.return_residual:
                 qkv = self.Wqkv(x)
             else:
@@ -491,14 +495,14 @@ class MHA(nn.Module):
                     context = rearrange(context, 'b h d -> b 1 h d')
         else:
             if not self.return_residual:
-                q = self.Wq(x)
+                q = self.Wq(x if mixer_subset is None else x[:, mixer_subset])
                 kv = self.Wkv(x_kv if x_kv is not None else x)
             else:
                 if x_kv is not None:
                     kv, x_kv = self.Wkv(x_kv)
                 else:
                     kv, x = self.Wkv(x)
-                q = self.Wq(x)
+                q = self.Wq(x if mixer_subset is None else x[:, mixer_subset])
             q = rearrange(q, '... (h d) -> ... h d', d=self.head_dim)
             kv = rearrange(kv, '... (two h d) -> ... two h d', two=2, d=self.head_dim)
             if self.dwconv:

tests/models/test_opt.py

@@ -26,7 +26,7 @@ def test_opt_state_dict(model_name):
 @pytest.mark.parametrize('model_name', ["facebook/opt-125m", "facebook/opt-350m", "facebook/opt-1.3b"])
 # @pytest.mark.parametrize('model_name', ["facebook/opt-350m"])
 def test_opt_optimized(model_name):
-    """Check that our implementation of OPT (without any optimizations enabled) matches the
+    """Check that our implementation of OPT (without all optimizations enabled) matches the
     HF implementation: the output of our forward pass in fp16 should be around the same as the HF
     forward pass in fp16, when compared to the HF forward pass in fp32.
     """

tests/models/test_vit.py

@@ -0,0 +1,49 @@
+import re
+
+import torch
+import pytest
+
+from timm.models.vision_transformer import vit_base_patch16_224
+
+from flash_attn.models.vit import vit_base_patch16_224 as flash_vit_base_patch16_224
+
+
+@pytest.mark.parametrize('fused_dense_gelu_dense', [False, True])
+# @pytest.mark.parametrize('fused_dense_gelu_dense', [False])
+@pytest.mark.parametrize('optimized', [False, True])
+# @pytest.mark.parametrize('optimized', [True])
+def test_vit(optimized, fused_dense_gelu_dense):
+    """Check that our implementation of ViT matches the timm's implementation:
+    the output of our forward pass in fp16 should be around the same as
+    timm' forward pass in fp16, when compared to timm's forward pass in fp32.
+    """
+    dtype = torch.float16
+    device = 'cuda'
+
+    kwargs = {}
+    if optimized:
+        kwargs = dict(use_flash_attn=True, fused_bias_fc=True, fused_dropout_add_ln=True)
+    kwargs['fused_dense_gelu_dense'] = fused_dense_gelu_dense
+    model = flash_vit_base_patch16_224(**kwargs).to(device=device, dtype=dtype)
+
+    model_ref = vit_base_patch16_224(pretrained=True).to(device=device)
+    model_timm = vit_base_patch16_224(pretrained=True).to(device=device, dtype=dtype)
+
+    model.load_state_dict(model_ref.state_dict())
+
+    model.eval()
+    model_ref.eval()
+    model_timm.eval()
+
+    torch.manual_seed(0)
+    batch_size = 2
+    x = torch.randn(batch_size, 3, 224, 224, device=device, dtype=dtype)
+    out = model(x)
+    out_timm = model_timm(x)
+    out_ref = model_ref(x.float())
+
+    print(f'Output max diff: {(out - out_ref).abs().max().item()}')
+    print(f'Output mean diff: {(out - out_ref).abs().mean().item()}')
+    print(f'timm fp16 max diff: {(out_timm - out_ref).abs().max().item()}')
+    print(f'timm fp16 mean diff: {(out_timm - out_ref).abs().mean().item()}')
+    assert (out - out_ref).abs().max().item() < 3 * (out_timm - out_ref).abs().max().item()