bf16: true
cutoff_len: 20000
dataset: ALS
dataset_dir: data
ddp_timeout: 180000000
deepspeed: cache/ds_z2_offload_config.json
do_train: true
finetuning_type: lora
flash_attn: fa2
gradient_accumulation_steps: 8
include_num_input_tokens_seen: true
learning_rate: 5.0e-05
logging_steps: 10
lora_alpha: 16
lora_dropout: 0
lora_rank: 8
lora_target: all
lr_scheduler_type: cosine
max_grad_norm: 1.0
max_samples: 100000
model_name_or_path: microsoft/Phi-3-medium-128k-instruct
num_train_epochs: 3.0
optim: adamw_torch
output_dir: saves/Custom/lora/train_2024-07-24-15-00-21
packing: false
per_device_train_batch_size: 2
plot_loss: true
preprocessing_num_workers: 16
quantization_bit: 4
quantization_method: bitsandbytes
report_to: none
save_steps: 500
stage: sft
template: default
warmup_steps: 50