423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2026-11-06 15:10:01 +08:00
Code Issues Packages Projects Releases Wiki Activity

docs: add Qwen3.8 LlamaFactory SFT skill (#10901)

Browse Source 
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
This commit is contained in:
Cui-yshoho
2026-11-04 20:39:18 +08:00
committed by GitHub
parent a3e7f60ae6
commit 84024cefa6
12 changed files with 702 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
examples/accelerate/fsdp2_config_qwen38.yaml Normal file
View File

@@ -0,0 +1,20 @@
compute_environment: LOCAL_MACHINE
debug: false
distributed_type: FSDP
downcast_bf16: 'no'
fsdp_config:
fsdp_version: 2
fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
fsdp_transformer_layer_cls_to_wrap: Qwen3_5DecoderLayer,Qwen3_5VisionBlock
fsdp_cpu_ram_efficient_loading: true
fsdp_offload_params: false
fsdp_reshard_after_forward: true
fsdp_state_dict_type: FULL_STATE_DICT
machine_rank: 0
main_training_function: main
mixed_precision: bf16
num_machines: 1
num_processes: 16 # Change together with the visible-device list for your hardware.
rdzv_backend: static
same_network: true
use_cpu: false