fix alignment up to input ln

flexflow · Aug 20, 2024 · d0e98ec · d0e98ec
1 parent 9ca3687
commit d0e98ec
Showing 1 changed file with 8 additions and 8 deletions.
diff --git a/tests/peft/peft_alignment_test.py b/tests/peft/peft_alignment_test.py
@@ -368,7 +368,7 @@ def get_hf_tensor(hf_tensor_name, tensor_comparison_idx):
             hf_tensor = torch.load(hf_tensor_path, map_location='cpu')
             return hf_tensor
 
-        def get_ff_tensor(ff_tensor_name, tensor_comparison_idx, hf_shape, tp_type=TPType.REPLICATE, pre=False):
+        def get_ff_tensor(ff_tensor_name, tensor_comparison_idx, hf_shape, tp_type=TPType.REPLICATE, pre=False, shard_axis=0):
             ff_tensor_suffix = f".{tensor_comparison_idx.ff_tensor_type}" if len(tensor_comparison_idx.ff_tensor_type) > 0 else ""
             ff_tensor_idx_suffix = f"_{tensor_comparison_idx.ff_tensor_idx}" if tensor_comparison_idx.ff_tensor_idx is not None else ""
             ff_tensor_filename = f"{ff_tensor_name}{ff_tensor_suffix}{ff_tensor_idx_suffix}"
@@ -381,7 +381,7 @@ def get_ff_tensor(ff_tensor_name, tensor_comparison_idx, hf_shape, tp_type=TPTyp
 
             ff_shape = list(hf_shape)[::-1]
             if tp_type == TPType.PARTITION:
-                ff_shape[0] //= self.tp_degree
+                ff_shape[shard_axis] //= self.tp_degree
 
             # exception: intermediate attention tensors
             intermediate_attention_tensor = (
@@ -405,10 +405,10 @@ def get_ff_tensor(ff_tensor_name, tensor_comparison_idx, hf_shape, tp_type=TPTyp
                     ff_tensor = ff_tensors[0]
                 # if partition, concatenate along the partition dimension
                 elif tp_type == TPType.PARTITION:
-                    ff_tensor = np.concatenate(ff_tensors, axis=0)
+                    ff_tensor = np.concatenate(ff_tensors, axis=shard_axis)
                 # if to_reduce, sum along the partition dimension
                 elif tp_type == TPType.TO_REDUCE:
-                    ff_tensor = np.sum(ff_tensors, axis=0)
+                    ff_tensor = np.sum(ff_tensors, axis=shard_axis)
             else:
                 ff_tensor = ff_tensors[0]
             ff_tensor = torch.from_numpy(ff_tensor)
@@ -551,7 +551,7 @@ def compare(hf_tensor, ff_tensor, label="", additional_ff_tensor=None, tolerance
             mixed_comparison = TensorComparisonIdxs(hf_tensor_type="output_gradient", ff_tensor_type="input_gradient", hf_tensor_idx=0, ff_tensor_idx=0)
             hf_tensor = get_hf_tensor(hf_tensor_name, mixed_comparison)
             hf_tensor = hf_tensor.squeeze().T
-            ff_tensor = get_ff_tensor(ff_tensor_name, mixed_comparison, hf_tensor.shape, tp_type=TPType.PARTITION)
+            ff_tensor = get_ff_tensor(ff_tensor_name, mixed_comparison, hf_tensor.shape, tp_type=TPType.PARTITION, shard_axis=1)
             compare(hf_tensor, ff_tensor, label=f"V-proj {i} gradient input")
 
             # K-proj grads
@@ -562,7 +562,7 @@ def compare(hf_tensor, ff_tensor, label="", additional_ff_tensor=None, tolerance
             hf_tensor = get_hf_tensor(hf_tensor_name, k_proj_comparison)
             hf_tensor = hf_tensor.squeeze().view(self.num_tokens, self.num_attention_heads, self.projsize).transpose(1, 2).contiguous()
             hf_tensor = hf_tensor.T
-            ff_tensor = get_ff_tensor(ff_tensor_name, k_proj_comparison, hf_tensor.shape, tp_type=TPType.PARTITION)
+            ff_tensor = get_ff_tensor(ff_tensor_name, k_proj_comparison, hf_tensor.shape, tp_type=TPType.PARTITION, shard_axis=2)
             compare(hf_tensor, ff_tensor, label=f"K-proj {i} gradient input")
 
             # Q-proj grads
@@ -574,15 +574,15 @@ def compare(hf_tensor, ff_tensor, label="", additional_ff_tensor=None, tolerance
             hf_tensor = get_hf_tensor(hf_tensor_name, q_proj_comparison)
             hf_tensor = hf_tensor.view(self.num_tokens, self.num_attention_heads, self.projsize).transpose(1, 2).contiguous().T
             augmented_hf_tensor_shape = torch.Size([3]+list(hf_tensor.size()))
-            ff_tensor = get_ff_tensor(ff_tensor_name, q_proj_comparison, augmented_hf_tensor_shape, tp_type=TPType.PARTITION)[:,:,:,0]
+            ff_tensor = get_ff_tensor(ff_tensor_name, q_proj_comparison, augmented_hf_tensor_shape, tp_type=TPType.PARTITION, shard_axis=2)[:,:,:,0]
             compare(hf_tensor, ff_tensor, label=f"Q-proj {i} gradient input")
 
             # FF Attn input with HF layernorm out
             hf_tensor_name = f"layers.{i}.input_layernorm"
             ff_tensor_name = f"layers.{i}.layers.{i}.self_attn"
             input_comparison = TensorComparisonIdxs(hf_tensor_type="output_gradient", ff_tensor_type="input_gradient", hf_tensor_idx=0, ff_tensor_idx=0)
             hf_tensor = get_hf_tensor(hf_tensor_name, input_comparison)
-            ff_tensor = get_ff_tensor(ff_tensor_name, input_comparison, hf_tensor.shape, tp_type=TPType.PARTITION)
+            ff_tensor = get_ff_tensor(ff_tensor_name, input_comparison, hf_tensor.shape, tp_type=TPType.TO_REDUCE)
             compare(hf_tensor, ff_tensor, label=f"Attn input {i} gradient input")
 
             if i > 0: