---
title: DeepPrune-Judge-4B
canonical_url: "https://www.modelscope.ai/models/THU-KEG/DeepPrune-Judge-4B"
md_url: "https://www.modelscope.ai/models/THU-KEG/DeepPrune-Judge-4B.md"
repository: THU-KEG/DeepPrune-Judge-4B
last_updated: 2025-12-26
license: other
pipeline_tag: text-generation
tasks:
  - text-generation
model_type:
  - qwen3
architectures:
  - Qwen3ForCausalLM
base_model:
  - Qwen/Qwen3-4B-Instruct-2507
base_model_relation: finetune
parameters: 4.0B
tensor_type:
  - BF16
library_name:
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 129
stars: 0
tags:
  - llama-factory
  - full
  - generated_from_trainer
---

# DeepPrune-Judge-4B

> DeepPrune-Judge-4B - An open-source model by THU-KEG on ModelScope. Qwen3-4B-Instruct-2507-fullsft25oversamplingfocalloss

THU-KEG/DeepPrune-Judge-4B is a 4.0B-parameter text-generation model on ModelScope. licensed under other. derived from Qwen/Qwen3-4B-Instruct-2507.

- **Repository**: THU-KEG/DeepPrune-Judge-4B
- **License**: other
- **Tasks**: text-generation
- **Parameters**: 4.0B
- **Base model**: Qwen/Qwen3-4B-Instruct-2507
- **Tags**: llama-factory, full, generated_from_trainer
- **Downloads**: 129
- **Stars**: 0
- **Last updated**: 2025-12-26

Source: https://www.modelscope.ai/models/THU-KEG/DeepPrune-Judge-4B

---

<!-- 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. -->

# Qwen3-4B-Instruct-2507-full_sft_25_oversampling_focal_loss

This model is a fine-tuned version of [Qwen3-4B-Instruct-2507](https://huggingface.co/Qwen/Qwen3-4B-Instruct-2507) on the my_custom_dataset dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0438

## 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: 1e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- total_eval_batch_size: 4
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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 4.55.0
- Pytorch 2.8.0+cu128
- Datasets 3.6.0
- Tokenizers 0.21.1