F1-24B-Base / README.md
lianghsun's picture
Update README.md
2f4d22a verified
metadata
base_model:
  - mistralai/Mistral-Small-24B-Base-2501

sft-2025-03-20-1

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: 8e-06
  • train_batch_size: 2
  • eval_batch_size: 8
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 192 :p
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 3072
  • total_eval_batch_size: 1536
  • optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=min_lr_ratio=0.1
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_ratio: 0.2
  • num_epochs: 3.0

Training results

Framework versions

  • Transformers 4.49.0
  • Pytorch 2.5.1+cu124
  • Datasets 3.2.0
  • Tokenizers 0.21.0