README.md

# checkpoint-4600

## Checkpoint Information

**Checkpoint Name**: `checkpoint-4600`

**Repository Name**: `aylinakkus/qwen_2_5_math_epoch_4600`

**Checkpoint Path**: `/home/mert/aylin/capability-erosion-sft/LLaMA-Factory/saves/qwen2.5-1.5b/full/sft/checkpoint-4600`

## Model Configuration

This checkpoint was extracted from a Qwen 2.5 1.5B model training run.

- **Base Model**: Qwen 2.5 1.5B
- **Training Framework**: LLaMA-Factory
- **Task**: Math fine-tuning

## Description

This repository contains the model state dict extracted from the training checkpoint.

### Files

- `model_state_dict.pt`: PyTorch state dictionary containing the model weights
- `README.md`: This file

## Usage

```python
import torch

# Load the model state dict
state_dict = torch.load("model_state_dict.pt", map_location='cpu')

# Use with your model architecture
# model.load_state_dict(state_dict)
```

## Notes

- This checkpoint was automatically uploaded using the `upload_checkpoints.py` script
- Checkpoint extracted from: `checkpoint-4600`
- Original path: `/home/mert/aylin/capability-erosion-sft/LLaMA-Factory/saves/qwen2.5-1.5b/full/sft/checkpoint-4600`