From 2f3d1bd0ec3f92f97b5fb562e9ff7dabe3a5d0dd Mon Sep 17 00:00:00 2001 From: binbin Deng <108676127+plusbang@users.noreply.github.com> Date: Mon, 2 Sep 2024 18:11:08 +0800 Subject: [PATCH] hotfix qwen2-7b weight setting (#11991) --- python/llm/src/ipex_llm/transformers/npu_models/qwen2_mp.py | 1 + 1 file changed, 1 insertion(+) diff --git a/python/llm/src/ipex_llm/transformers/npu_models/qwen2_mp.py b/python/llm/src/ipex_llm/transformers/npu_models/qwen2_mp.py index c6cb74c8..e1c6f9b8 100644 --- a/python/llm/src/ipex_llm/transformers/npu_models/qwen2_mp.py +++ b/python/llm/src/ipex_llm/transformers/npu_models/qwen2_mp.py @@ -587,6 +587,7 @@ def run_decode( (attn_layer.v_proj.weight, attn_layer.v_proj.scale), (attn_layer.o_proj.weight, attn_layer.o_proj.scale), (mlp_layer.gate_proj.weight, mlp_layer.gate_proj.scale), + (mlp_layer.up_proj.weight, mlp_layer.up_proj.scale), (mlp_layer.down_proj_0.weight, mlp_layer.down_proj_0.scale), (mlp_layer.down_proj_1.weight, mlp_layer.down_proj_1.scale) ]