diff --git a/convert-mpt-hf-to-gguf.py b/convert-mpt-hf-to-gguf.py
index 60bceb0fa..cbe4a9f11 100755
--- a/convert-mpt-hf-to-gguf.py
+++ b/convert-mpt-hf-to-gguf.py
@@ -121,7 +121,7 @@ gguf_writer.add_embedding_length(hparams["d_model"])
 gguf_writer.add_block_count(block_count)
 gguf_writer.add_feed_forward_length(4 * hparams["d_model"])
 gguf_writer.add_head_count(hparams["n_heads"])
-gguf_writer.add_layer_norm_eps(1e-05) 
+gguf_writer.add_layer_norm_eps(1e-05)
 if hparams["attn_config"]["clip_qkv"] is not None:
     gguf_writer.add_clamp_kqv(hparams["attn_config"]["clip_qkv"])
 gguf_writer.add_max_alibi_bias(hparams["attn_config"]["alibi_bias_max"])
diff --git a/llama.cpp b/llama.cpp
index 81a014d0a..ede95f607 100644
--- a/llama.cpp
+++ b/llama.cpp
@@ -4304,11 +4304,11 @@ static struct ggml_cgraph * llm_build_mpt(
             offload_func_kq(KQ_scaled);
             ggml_set_name(KQ_scaled, "KQ_scaled");
 
-			// TODO: replace with ggml_add()
-			struct ggml_tensor * KQ_scaled_alibi =
-				ggml_alibi(ctx0, KQ_scaled, std::max(kv_head, n_kv - n_tokens), n_head, max_alibi_bias);
+            // TODO: replace with ggml_add()
+            struct ggml_tensor * KQ_scaled_alibi =
+                ggml_alibi(ctx0, KQ_scaled, std::max(kv_head, n_kv - n_tokens), n_head, max_alibi_bias);
             offload_func_kq(KQ_scaled_alibi);
-			ggml_set_name(KQ_scaled_alibi, "KQ_scaled_alibi");
+            ggml_set_name(KQ_scaled_alibi, "KQ_scaled_alibi");
 
             struct ggml_tensor * KQ_masked = ggml_add(ctx0, KQ_scaled_alibi, KQ_mask);
             offload_func_kq(KQ_masked);