diff --git a/docs/source/en/training/t2i_adapters.md b/docs/source/en/training/t2i_adapters.md
index 0f65ad8ed31d..03f4537cb28d 100644
--- a/docs/source/en/training/t2i_adapters.md
+++ b/docs/source/en/training/t2i_adapters.md
@@ -12,7 +12,7 @@ specific language governing permissions and limitations under the License.
 
 # T2I-Adapter
 
-[T2I-Adapter]((https://hf.co/papers/2302.08453)) is a lightweight adapter model that provides an additional conditioning input image (line art, canny, sketch, depth, pose) to better control image generation. It is similar to a ControlNet, but it is a lot smaller (~77M parameters and ~300MB file size) because its only inserts weights into the UNet instead of copying and training it.
+[T2I-Adapter](https://hf.co/papers/2302.08453) is a lightweight adapter model that provides an additional conditioning input image (line art, canny, sketch, depth, pose) to better control image generation. It is similar to a ControlNet, but it is a lot smaller (~77M parameters and ~300MB file size) because its only inserts weights into the UNet instead of copying and training it.
 
 The T2I-Adapter is only available for training with the Stable Diffusion XL (SDXL) model.