423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2026-06-17 12:48:55 +08:00
Code Issues Packages Projects Releases Wiki Activity

[model] Patch GDN for NPU (#10504)

Browse Source 
Co-authored-by: jiaqiw09 <jiaqiw960714@gmail.com>
This commit is contained in:
A1waysBeenHere
2026-06-04 16:39:02 +08:00
committed by GitHub
parent 053d43c0ac
commit 409e8a477f
14 changed files with 3219 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
examples/accelerate/fsdp2_config_qwen35_moe.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_5MoeDecoderLayer,Qwen3_5MoeVisionBlock
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: 8 # Change to match your NPU count (e.g., 8 for A2, 16 for A3)
rdzv_backend: static
same_network: true
use_cpu: false