File size: 859 Bytes
f651cd7 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 |
### model
model_name_or_path: google/gemma-7b-it
adapter_name_or_path: chchen/Gemma-7B-It-ORPO
### method
stage: sft
do_train: true
finetuning_type: lora
lora_target: all
### dataset
dataset: bct_non_cot_sft_1000
dataset_dir: data_private
template: gemma
cutoff_len: 1024
# max_samples: 1000
overwrite_cache: true
preprocessing_num_workers: 16
### output
output_dir: saves/Gemma-7B-It/lora/orpo-sft
logging_steps: 10
save_steps: 500
plot_loss: true
overwrite_output_dir: true
save_total_limit: 3
load_best_model_at_end: true
push_to_hub: true
hub_model_id: chchen/Gemma-7B-It-ORPO-SFT
### train
per_device_train_batch_size: 2
gradient_accumulation_steps: 8
learning_rate: 0.000005
num_train_epochs: 3.0
lr_scheduler_type: cosine
warmup_steps: 0.1
fp16: true
### eval
val_size: 0.1
per_device_eval_batch_size: 2
evaluation_strategy: steps
eval_steps: 500
|