metadata

license: apache-2.0
base_model: Qwen/Qwen2-0.5B-Instruct
tags:
  - alignment-handbook
  - trl
  - dpo
  - generated_from_trainer
  - trl
  - dpo
  - generated_from_trainer
datasets:
  - princeton-nlp/llama3-ultrafeedback
model-index:
  - name: qwen2-0.5b-instruct-simpo-lr-5e-07-gamma-1.5
    results: []

Description

This model was trained as part of the Reinforcement Learning - 24 project at Peking University, focusing on [simpo].

Authors

Ejafa Bassam
Yaroslav Ponomarenko

qwen2-0.5b-instruct-simpo-lr-5e-07-gamma-1.5

This model is a fine-tuned version of Qwen/Qwen2-0.5B-Instruct on the princeton-nlp/llama3-ultrafeedback dataset. It achieves the following results on the evaluation set:

Loss: 1.6594
Rewards/chosen: -3.3473
Rewards/rejected: -3.4798
Rewards/accuracies: 0.5282
Rewards/margins: 0.1325
Logps/rejected: -1.3919
Logps/chosen: -1.3389
Logits/rejected: -5.2419
Logits/chosen: -5.3398

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: 2
eval_batch_size: 4
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 8
total_train_batch_size: 128
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
1.6693	0.8549	400	1.6598	-3.3421	-3.4735	0.5282	0.1314	-1.3894	-1.3368	-5.2590	-5.3573

Framework versions

Transformers 4.41.2
Pytorch 2.3.1+cu121
Datasets 2.20.0
Tokenizers 0.19.1