ros/LLaMA-Factory
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

add more llava-next series template

Browse Source 
Former-commit-id: 93f64f2aebf41582d39aa8a2c6059e562ca694b0
This commit is contained in:
BUAADreamer
2024-09-29 21:29:29 +08:00
parent 4d43317aa1
commit ab5d042cd3
6 changed files with 141 additions and 48 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/llamafactory/model/model_utils/misc.py
View File

@@ -34,7 +34,7 @@ def find_all_linear_modules(model: "PreTrainedModel", freeze_vision_tower: bool)
forbidden_modules.add("output_layer")
elif model_type == "internlm2":
forbidden_modules.add("output")
elif model_type in ["llava", "llava_next", "llava_next_video", "paligemma", "video_llava"]:
elif model_type in ["llava", "paligemma", "video_llava"] or "llava_next" in model_type:
forbidden_modules.add("multi_modal_projector")
elif model_type == "qwen2_vl":
forbidden_modules.add("merger")