ros/LlamaFactory
1
0
Fork 0
mirror of https://github.com/hiyouga/LlamaFactory.git synced 2026-03-24 13:03:08 +00:00
Code Issues Packages Projects Releases Wiki Activity

[fix] fit neat_packing & mrope model packing (#10283)

Browse Source 
Co-authored-by: Yaowei Zheng <hiyouga@buaa.edu.cn>
This commit is contained in:
Kingsley
2026-03-20 16:50:11 +08:00
committed by GitHub
parent d91d8af89e
commit 833f6027b1
15 changed files with 520 additions and 93 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/llamafactory/model/patcher.py
View File

@@ -30,7 +30,6 @@ from .model_utils.embedding import resize_embedding_layer
from .model_utils.kv_cache import configure_kv_cache
from .model_utils.longlora import configure_longlora
from .model_utils.moe import add_z3_leaf_module, configure_moe
from .model_utils.packing import configure_packing
from .model_utils.quantization import configure_quantization
from .model_utils.rope import configure_rope
from .model_utils.valuehead import prepare_valuehead_model
@@ -142,7 +141,6 @@ def patch_config(
configure_quantization(config, tokenizer, model_args, is_trainable, init_kwargs)
configure_moe(config, model_args, is_trainable)
configure_visual_model(config)
configure_packing(model_args, is_trainable)
configure_kv_cache(config, model_args, is_trainable)
if getattr(config, "model_type", None) == "qwen":