423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2026-07-31 13:16:12 +08:00
Code Issues Packages Projects Releases Wiki Activity

[v1] refactor registry plugin structure and params (#10641)

Browse Source
This commit is contained in:
Jiaqi
2026-07-24 15:23:21 +08:00
committed by GitHub
parent 19e9fe3ced
commit 3f77101580
50 changed files with 843 additions and 1007 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
examples/v1/train_freeze/train_freeze_sft.yaml
View File

@@ -1,7 +1,6 @@
model: Qwen/Qwen3-4B
model_class: llm
# Freeze Configuration
peft_config:
name: freeze
@@ -12,7 +11,6 @@ peft_config:
# Kernel Config
kernel_config:
name: auto
include_kernels: auto
# FSDP Config
dist_config:
@@ -25,7 +23,6 @@ train_dataset: data/v1_sft_demo.yaml
### training
output_dir: ./outputs/test_freeze
micro_batch_size: 1
global_batch_size: 4
cutoff_len: 2048
learning_rate: 2.0e-5
max_steps: 10