# Model Card - Source: [https://arxiv.org/abs/2509.02046](https://arxiv.org/abs/2509.02046) - Optimizer: `mini` - Model size: `520m` - Data size: `85B` ## Best configuration | Hyperparameter | Value | |---|---| | beta1 | `0.9` | | beta2 | `0.98` | | epsilon | `1e-10` | | learning_rate | `0.004` | | max_grad_norm | `1` | | min_lr_ratio | `0` | | nesterov | `False` | | train_batch_size | `256` | | warmup | `4000` | | weight_decay | `0.1` |