File size: 647 Bytes
6eae9eb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
cutoff_len: 4096
dataset: LangGPT_community,LangGPT_alpaca
dataset_dir: /datas/wangm/LLM4LangGPT
do_train: true
finetuning_type: lora
flash_attn: auto
fp16: true
gradient_accumulation_steps: 8
learning_rate: 5.0e-05
logging_steps: 5
lora_alpha: 16
lora_dropout: 0
lora_rank: 8
lora_target: q_proj,v_proj
lr_scheduler_type: cosine
max_grad_norm: 1.0
max_samples: 100000
model_name_or_path: Qwen/Qwen1.5-4B-Chat
num_train_epochs: 10.0
optim: adamw_torch
output_dir: /datas/wangm/LLM4LangGPT/output/Qwen1.5-4B-Chat/full_alpaca
packing: false
per_device_train_batch_size: 2
report_to: none
save_steps: 100
stage: sft
template: default
warmup_steps: 0