File size: 1,161 Bytes
87d0422
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
46
47
48
adam_beta1: '0.95'
adam_beta2: '0.999'
assistant_tag: gpt
bf16: 'True'
content_tag: value
cutoff_len: '2048'
dataset: mlfoundations-dev/oh-dcft-v1.2_no-curation_gpt-4o-mini
dataset_dir: ONLINE
ddp_timeout: '180000000'
deepspeed: /opt/ml/code/zero3.json
do_train: 'True'
enable_liger_kernel: 'False'
eval_strategy: epoch
finetuning_type: full
formatting: sharegpt
global_batch_size: '512'
gradient_accumulation_steps: '8'
gradient_checkpointing: 'True'
hub_model_id: mlfoundations-dev/hp_ablations_qwen_adambeta1_0.95_dcftv1.2
learning_rate: 5e-06
logging_steps: '10'
lr_scheduler_type: constant
max_grad_norm: '1'
messages: conversations
model_name_or_path: Qwen/Qwen2.5-7B
neat_packing: 'True'
num_train_epochs: '3.0'
output_dir: /opt/ml/model
overwrite_cache: 'True'
overwrite_output_dir: 'True'
packing: 'True'
per_device_train_batch_size: '8'
plot_loss: 'True'
preprocessing_num_workers: '16'
push_to_db: 'True'
push_to_hub: 'True'
report_to: wandb
role_tag: from
run_name: hp_ablations_qwen_adambeta1_0.95_dcftv1.2
save_strategy: epoch
stage: sft
template: qwen
user_tag: human
val_size: '0.05'
warmup_ratio: '0.1'
warmup_steps: '1738'
weight_decay: '0.1'