qwen25_math_acc_rm

This model is a fine-tuned version of Jennny/qwen25_7b_sft_math on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 2.2156
  • Accuracy: 0.65

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: 2
  • eval_batch_size: 2
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 8
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 128
  • total_eval_batch_size: 16
  • optimizer: Use OptimizerNames.PAGED_ADAMW 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: 0.03
  • num_epochs: 2

Training results

Training Loss Epoch Step Validation Loss Accuracy
0.3308 0.2424 10 1.3608 0.66
0.0063 0.4848 20 1.4237 0.7167
0.0703 0.7273 30 2.4097 0.6267
0.0751 0.9697 40 0.9876 0.67
0.0047 1.1939 50 1.9047 0.68
0.0009 1.4364 60 2.5391 0.6267
0.1065 1.6788 70 2.4082 0.64
0.1204 1.9212 80 2.2156 0.65

Framework versions

  • Transformers 4.49.0
  • Pytorch 2.5.1+cu124
  • Datasets 3.3.2
  • Tokenizers 0.21.1
Downloads last month
-
Safetensors
Model size
7B params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Jennny/qwen25_math_acc_rm

Base model

Qwen/Qwen2.5-7B
Finetuned
(5)
this model