howanching-clara's picture
Upload tokenizer
c1d1397 verified
|
raw
history blame
3.03 kB
metadata
tags:
  - generated_from_trainer
base_model: sentence-transformers/multi-qa-MiniLM-L6-cos-v1
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: all_keywords_multi-qa-MiniLM-L6-cos-v1_another
    results: []

all_keywords_multi-qa-MiniLM-L6-cos-v1_another

This model is a fine-tuned version of sentence-transformers/multi-qa-MiniLM-L6-cos-v1 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 3.2408
  • Accuracy: 0.5259
  • Precision: 0.5259
  • Recall: 0.5259
  • F1: 0.5259

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: 4
  • eval_batch_size: 4
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 15

Training results

Training Loss Epoch Step Validation Loss Accuracy Precision Recall F1
2.3237 1.0 712 1.9365 0.4123 0.4123 0.4123 0.4123
2.0274 2.0 1424 1.8781 0.4502 0.4502 0.4502 0.4502
1.7498 3.0 2136 1.7664 0.4656 0.4656 0.4656 0.4656
1.6009 4.0 2848 1.7801 0.4656 0.4656 0.4656 0.4656
1.2981 5.0 3560 1.9004 0.4600 0.4600 0.4600 0.4600
1.0895 6.0 4272 2.0621 0.4642 0.4642 0.4642 0.4642
0.9878 7.0 4984 2.4640 0.4572 0.4572 0.4572 0.4572
0.784 8.0 5696 2.5635 0.5091 0.5091 0.5091 0.5091
0.6569 9.0 6408 2.5690 0.5273 0.5273 0.5273 0.5273
0.5847 10.0 7120 2.9342 0.5063 0.5063 0.5063 0.5063
0.5107 11.0 7832 2.9652 0.5091 0.5091 0.5091 0.5091
0.4954 12.0 8544 3.1480 0.5161 0.5161 0.5161 0.5161
0.4274 13.0 9256 3.2199 0.4993 0.4993 0.4993 0.4993
0.433 14.0 9968 3.2185 0.5217 0.5217 0.5217 0.5217
0.3615 15.0 10680 3.2408 0.5259 0.5259 0.5259 0.5259

Framework versions

  • Transformers 4.39.3
  • Pytorch 2.2.1+cu118
  • Datasets 2.14.7
  • Tokenizers 0.15.2