This model is a fine-tuned version of meta-llama/Llama-2-7b-hf on the HuggingFaceH4/ultrafeedback_binarized dataset. It achieves the following results on the evaluation set:

Loss: 0.6073
Rewards/chosen: 0.0705
Rewards/rejected: -0.1795
Rewards/accuracies: 0.8214
Rewards/margins: 0.2499
Logps/rejected: -821.0361
Logps/chosen: -523.9339
Logits/rejected: -0.3815
Logits/chosen: -0.1586

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-07
train_batch_size: 8
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
num_devices: 4
gradient_accumulation_steps: 2
total_train_batch_size: 64
total_eval_batch_size: 32
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
0.6065	0.6	100	0.6297	0.0796	-0.0955	0.7679	0.1751	-812.6376	-523.0222	-0.3805	-0.1364

Framework versions

Transformers 4.39.0.dev0
Pytorch 2.1.2
Datasets 2.14.6
Tokenizers 0.15.2

RichardErkhov/Minbyul_-_llama2-7b-dpo-full-wo-live_qa-ep3-4bits

Original model description:

llama2-7b-dpo-full-wo-live_qa-ep3

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

RichardErkhov/
Minbyul_-_llama2-7b-dpo-full-wo-live_qa-ep3-4bits