Fix bloom ffn fusion (#620)

Author: LJ-underdog

intel_extension_for_transformers/llm/runtime/graph/models/bloom/bloom.cpp

@@ -213,13 +213,15 @@ static bool bloom_model_eval_internal(model_context& lctx, const model_token* to
  model.layers[il].ffn[5], cur);
  } else {
  cur = ne_mul_mat(ctx0, model.layers[il].ffn[2], cur);
+
  cur = ne_add(ctx0, ne_repeat(ctx0, model.layers[il].ffn[3], cur), cur);
 
  cur = ne_gelu(ctx0, cur);
 
  cur = ne_mul_mat(ctx0, model.layers[il].ffn[4], cur);
+
+ cur = ne_add(ctx0, ne_repeat(ctx0, model.layers[il].ffn[5], cur), cur);
  }
- cur = ne_add(ctx0, ne_repeat(ctx0, model.layers[il].ffn[5], cur), cur);
  }
 
  cur = ne_add(ctx0, cur, inpFF);