This model is a fine-tuned version of alignment-handbook/zephyr-7b-sft-qlora on the updated and the original datasets. It achieves the following results on the evaluation set:

Loss: 0.5386
Rewards/chosen: -0.9159
Rewards/rejected: -1.5101
Rewards/accuracies: 0.6960
Rewards/margins: 0.5943
Logps/rejected: -392.2019
Logps/chosen: -355.0374
Logits/rejected: 0.1234
Logits/chosen: -0.3064

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e-06
train_batch_size: 2
eval_batch_size: 2
seed: 42
distributed_type: multi-GPU
num_devices: 4
gradient_accumulation_steps: 32
total_train_batch_size: 256
total_eval_batch_size: 8
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.1
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.5464	0.42	100	0.5636	-0.6892	-1.1960	0.7060	0.5068	-360.7926	-332.3759	-0.6169	-0.9747
0.5369	0.84	200	0.5386	-0.9159	-1.5101	0.6960	0.5943	-392.2019	-355.0374	0.1234	-0.3064

Framework versions

PEFT 0.7.1
Transformers 4.36.2
Pytorch 2.2.2+cu121
Datasets 2.14.6
Tokenizers 0.15.2

YYYYYYibo/zephyr-7b-dpo-qlora-training-on-full-binary

zephyr-7b-dpo-qlora-training-on-full-binary

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

YYYYYYibo/
zephyr-7b-dpo-qlora-training-on-full-binary