[Bugfix] Fix fully sharded LoRA bug (#10352)

Signed-off-by: Jee Jee Li <[email protected]>

vllm/lora/fully_sharded_layers.py

@@ -165,15 +165,14 @@ class MergedColumnParallelLinearWithShardedLoRA(
     def slice_lora_a(
         self, lora_a: List[Union[torch.Tensor, None]]
     ) -> List[Union[torch.Tensor, None]]:
-        if lora_a[0] is None or lora_a[1] is None:
-            return lora_a
+        #NOTE: lora_a contains 2 subloras, and each sublora could be None.
         output_shard_size = self.lora_a_stacked[0].shape[2]
         output_start_idx = self.tp_rank * output_shard_size
         lora_a = [
-            lora_a[0][:,
-                      output_start_idx:output_start_idx + output_shard_size],
-            lora_a[1][:,
-                      output_start_idx:output_start_idx + output_shard_size],
+            lora_a[0][:, output_start_idx:output_start_idx +
+                      output_shard_size] if lora_a[0] is not None else None,
+            lora_a[1][:, output_start_idx:output_start_idx +
+                      output_shard_size] if lora_a[1] is not None else None,
         ]
         return lora_a
 
@@ -261,14 +260,16 @@ class MergedQKVParallelLinearWithShardedLora(MergedQKVParallelLinearWithLora):
     def slice_lora_a(
         self, lora_a: List[Union[torch.Tensor, None]]
     ) -> List[Union[torch.Tensor, None]]:
-        if lora_a[0] is None or lora_a[1] is None or lora_a[2] is None:
-            return lora_a
+        # NOTE: lora_a contains 3 subloras, and each sublora could be None.
         shard_size = [self.lora_a_stacked[i].shape[2] for i in range(3)]
         start_idx = [self.tp_rank * shard_size[i] for i in range(3)]
         lora_a = [
-            lora_a[0][:, start_idx[0]:start_idx[0] + shard_size[0]],
-            lora_a[1][:, start_idx[1]:start_idx[1] + shard_size[1]],
-            lora_a[2][:, start_idx[2]:start_idx[2] + shard_size[2]],
+            lora_a[0][:, start_idx[0]:start_idx[0] +
+                      shard_size[0]] if lora_a[0] is not None else None,
+            lora_a[1][:, start_idx[1]:start_idx[1] +
+                      shard_size[1]] if lora_a[1] is not None else None,
+            lora_a[2][:, start_idx[2]:start_idx[2] +
+                      shard_size[2]] if lora_a[2] is not None else None,
         ]
         return lora_a
 

vllm/lora/layers.py

@@ -685,26 +685,27 @@ def slice_lora_a(
     def slice_lora_b(
         self, lora_b: List[Union[torch.Tensor, None]]
     ) -> List[Union[torch.Tensor, None]]:
-        if lora_b[0] is None or lora_b[1] is None:
-            return lora_b
+        #NOTE: lora_b contains 2 subloras, and each sublora could be None.
         shard_size = self.output_dim
         start_idx = self.tp_rank * shard_size
         end_idx = (self.tp_rank + 1) * shard_size
         lora_b = [
-            lora_b[0][:, start_idx:end_idx],
-            lora_b[1][:, start_idx:end_idx],
+            lora_b[0][:, start_idx:end_idx] if lora_b[0] is not None else None,
+            lora_b[1][:, start_idx:end_idx] if lora_b[1] is not None else None,
         ]
         return lora_b
 
     def slice_bias(
         self, bias: List[Union[torch.Tensor,
                                None]]) -> List[Union[torch.Tensor, None]]:
-        if bias[0] is None or bias[1] is None:
-            return bias
+        # NOTE : each bias could be None.
         shard_size = self.output_dim
         start_idx = self.tp_rank * shard_size
         end_idx = (self.tp_rank + 1) * shard_size
-        bias = [bias[0][start_idx:end_idx], bias[1][start_idx:end_idx]]
+        bias = [
+            bias[0][start_idx:end_idx] if bias[0] is not None else None,
+            bias[1][start_idx:end_idx] if bias[1] is not None else None
+        ]
         return bias
 
     def set_lora(

vllm/worker/worker.py

@@ -232,7 +232,7 @@ def determine_num_available_blocks(self) -> Tuple[int, int]:
         logger.info(
             "Memory profiling results: total_gpu_memory=%.2fGiB"
             " initial_memory_usage=%.2fGiB peak_torch_memory=%.2fGiB"
-            " memory_usage_post_profile=%.2fGib"
+            " memory_usage_post_profile=%.2fGiB"
             " non_torch_memory=%.2fGiB kv_cache_size=%.2fGiB"
             " gpu_memory_utilization=%.2f", total_gpu_memory / (1024**3),
             (total_gpu_memory - free_memory_pre_profile) / (1024**3),