hailin
/
jd_train
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

.

This commit is contained in:
hailin 2025-09-09 22:57:43 +08:00
parent 591b6ceb00
commit 515565b3cd
1 changed files with 4 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
ds_config_zero3_lora.json
View File

@ -9,17 +9,13 @@
"stage": 3, "stage": 3,
"overlap_comm": true, "overlap_comm": true,
"contiguous_gradients": true, "contiguous_gradients": true,
"allgather_partitions": true, "reduce_bucket_size": 500000000,
"reduce_scatter": true, "stage3_prefetch_bucket_size": 200000000,
"round_robin_gradients": true,
"reduce_bucket_size": 150000000,
"stage3_prefetch_bucket_size": 100000000,
"stage3_param_persistence_threshold": 1000000, "stage3_param_persistence_threshold": 1000000,
"offload_optimizer": { "device": "none" }, "offload_optimizer": { "device": "none" },
"offload_param": { "device": "none" } "offload_param": { "device": "none" }
}, },
"stage3_gather_16bit_weights_on_model_save": false,
"gradient_clipping": 1.0, "gradient_clipping": 1.0,
"wall_clock_breakdown": false "wall_clock_breakdown": false
} }