---
title: LongWriter-V-7B-DPO
canonical_url: "https://www.modelscope.ai/models/THU-KEG/LongWriter-V-7B-DPO"
md_url: "https://www.modelscope.ai/models/THU-KEG/LongWriter-V-7B-DPO.md"
repository: THU-KEG/LongWriter-V-7B-DPO
last_updated: 2025-12-26
license: other
model_type:
  - qwen2_5_vl
architectures:
  - Qwen2_5_VLForConditionalGeneration
base_model:
  - THU-KEG/LongWriter-V-7B
base_model_relation: finetune
parameters: 8.3B
tensor_type:
  - BF16
library_name:
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 130
stars: 0
tags:
  - llama-factory
  - full
  - generated_from_trainer
---

# LongWriter-V-7B-DPO

> LongWriter-V-7B-DPO - An open-source model by THU-KEG on ModelScope. This model is a fine-tuned version of THU-KEG/LongWriter-V-7B on the LongWriter-V-DPO dataset.

THU-KEG/LongWriter-V-7B-DPO is a 8.3B-parameter machine learning model on ModelScope. licensed under other. derived from THU-KEG/LongWriter-V-7B.

- **Repository**: THU-KEG/LongWriter-V-7B-DPO
- **License**: other
- **Parameters**: 8.3B
- **Base model**: THU-KEG/LongWriter-V-7B
- **Tags**: llama-factory, full, generated_from_trainer
- **Downloads**: 130
- **Stars**: 0
- **Last updated**: 2025-12-26

Source: https://www.modelscope.ai/models/THU-KEG/LongWriter-V-7B-DPO

---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# LongWriter-V-7B-DPO

This model is a fine-tuned version of [THU-KEG/LongWriter-V-7B](https://huggingface.co/THU-KEG/LongWriter-V-7B) on the LongWriter-V-DPO dataset.

## 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: 3e-06
- train_batch_size: 1
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- total_eval_batch_size: 64
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3.0

### Training results



### Framework versions

- Transformers 4.49.0.dev0
- Pytorch 2.5.1+cu124
- Datasets 3.2.0
- Tokenizers 0.21.0