lmind_hotpot_train8000_eval7405_v1_qa_Qwen_Qwen1.5-4B_5e-5_lora2

This model is a fine-tuned version of Qwen/Qwen1.5-4B on the tyzhu/lmind_hotpot_train8000_eval7405_v1_qa dataset. It achieves the following results on the evaluation set:

  • Loss: 3.9023
  • Accuracy: 0.4878

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: 5e-05
  • train_batch_size: 1
  • eval_batch_size: 2
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 4
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 32
  • total_eval_batch_size: 8
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: constant
  • lr_scheduler_warmup_ratio: 0.05
  • num_epochs: 20.0

Training results

Training Loss Epoch Step Validation Loss Accuracy
2.2742 1.0 250 2.3355 0.5148
2.1673 2.0 500 2.3223 0.5169
2.0189 3.0 750 2.3526 0.5157
1.8684 4.0 1000 2.4020 0.5140
1.7028 5.0 1250 2.4910 0.5100
1.5685 6.0 1500 2.5835 0.5072
1.4161 7.0 1750 2.6927 0.5036
1.2904 8.0 2000 2.8047 0.5009
1.1397 9.0 2250 2.9097 0.4979
1.0366 10.0 2500 3.0555 0.4951
0.9114 11.0 2750 3.1526 0.4928
0.8301 12.0 3000 3.2417 0.4928
0.7283 13.0 3250 3.3687 0.4894
0.6717 14.0 3500 3.5007 0.4899
0.6055 15.0 3750 3.5470 0.4894
0.5667 16.0 4000 3.6736 0.4876
0.5271 17.0 4250 3.7100 0.4870
0.5081 18.0 4500 3.7920 0.4877
0.4758 19.0 4750 3.8548 0.4868
0.4682 20.0 5000 3.9023 0.4878

Framework versions

  • PEFT 0.5.0
  • Transformers 4.40.2
  • Pytorch 2.3.0
  • Datasets 2.19.1
  • Tokenizers 0.19.1
Downloads last month
0
Inference API
Unable to determine this model’s pipeline type. Check the docs .

Model tree for tyzhu/lmind_hotpot_train8000_eval7405_v1_qa_Qwen_Qwen1.5-4B_5e-5_lora2

Base model

Qwen/Qwen1.5-4B
Adapter
(272)
this model

Dataset used to train tyzhu/lmind_hotpot_train8000_eval7405_v1_qa_Qwen_Qwen1.5-4B_5e-5_lora2

Evaluation results

  • Accuracy on tyzhu/lmind_hotpot_train8000_eval7405_v1_qa
    self-reported
    0.488