49 lines
978 B
YAML
49 lines
978 B
YAML
### model
|
|
model_name_or_path: HoangHa/Pensez-v0.1-init
|
|
trust_remote_code: true
|
|
enable_liger_kernel: true
|
|
use_unsloth_gc: true
|
|
|
|
### method
|
|
stage: sft
|
|
do_train: true
|
|
finetuning_type: full
|
|
deepspeed: examples/deepspeed/ds_z3_offload_config.json # choices: [ds_z0_config.json, ds_z2_config.json, ds_z3_config.json]
|
|
|
|
### dataset
|
|
dataset: pensez
|
|
template: qwen
|
|
cutoff_len: 16384
|
|
# max_samples: 1000
|
|
overwrite_cache: true
|
|
preprocessing_num_workers: 16
|
|
neat_packing: true
|
|
|
|
### output
|
|
output_dir: saves/fr-8b/full/sft
|
|
logging_steps: 1
|
|
# save_steps: 500
|
|
save_strategy: "epoch"
|
|
plot_loss: true
|
|
overwrite_output_dir: true
|
|
|
|
### train
|
|
per_device_train_batch_size: 2
|
|
gradient_accumulation_steps: 1
|
|
learning_rate: 1.0e-5
|
|
num_train_epochs: 5.0
|
|
lr_scheduler_type: cosine
|
|
warmup_ratio: 0.05
|
|
bf16: true
|
|
ddp_timeout: 180000000
|
|
neftune_noise_alpha: 5
|
|
|
|
### eval
|
|
# val_size: 0.1
|
|
# per_device_eval_batch_size: 1
|
|
# eval_strategy: steps
|
|
# eval_steps: 500
|
|
|
|
report_to: wandb
|
|
run_name: fr-pensez # optional
|