Hugging Face's logo

MAdAiLab
/
SLM_vs_LLM_experiments

Safetensors
Model card Files Community
SLM_vs_LLM_experiments / google /gemma_2b_ledgar /README.md
akkky02's picture
akkky02
Upload folder using huggingface_hub
aa5b5e2 verified
preview code
|
raw
history blame
3.83 kB
metadata
license: other
base_model: google/gemma-2b
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: gemma_2b_ledgar
    results: []

gemma_2b_ledgar

This model is a fine-tuned version of google/gemma-2b on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4836
  • Accuracy: 0.8682
  • F1 Macro: 0.7942
  • F1 Micro: 0.8682

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-06
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 2
  • total_train_batch_size: 64
  • total_eval_batch_size: 64
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 3.0

Training results

Training Loss Epoch Step Validation Loss Accuracy F1 Macro F1 Micro
1.3984 0.11 100 1.1195 0.7373 0.5766 0.7373
0.8017 0.21 200 0.7836 0.7952 0.6837 0.7952
0.7073 0.32 300 0.7108 0.8103 0.7035 0.8103
0.6628 0.43 400 0.6609 0.8197 0.7111 0.8197
0.6884 0.53 500 0.6126 0.8291 0.7293 0.8291
0.6117 0.64 600 0.5862 0.8397 0.7521 0.8397
0.5478 0.75 700 0.5648 0.8437 0.7615 0.8437
0.576 0.85 800 0.5639 0.8409 0.7532 0.8409
0.5161 0.96 900 0.5263 0.8508 0.7735 0.8508
0.2538 1.07 1000 0.5196 0.8594 0.7784 0.8594
0.2496 1.17 1100 0.5230 0.8568 0.7772 0.8568
0.2782 1.28 1200 0.5193 0.8633 0.7831 0.8633
0.2084 1.39 1300 0.5217 0.8593 0.7802 0.8593
0.2054 1.49 1400 0.5111 0.8612 0.7823 0.8612
0.2505 1.6 1500 0.5194 0.8602 0.7760 0.8602
0.2695 1.71 1600 0.5041 0.8616 0.7874 0.8616
0.2427 1.81 1700 0.4895 0.8664 0.7884 0.8664
0.2271 1.92 1800 0.4836 0.8682 0.7942 0.8682
0.0644 2.03 1900 0.4861 0.8686 0.7921 0.8686
0.0451 2.13 2000 0.4934 0.8723 0.7964 0.8723
0.0539 2.24 2100 0.4957 0.8703 0.7968 0.8703
0.0419 2.35 2200 0.4946 0.8717 0.7984 0.8717
0.0639 2.45 2300 0.4900 0.873 0.7999 0.873
0.0422 2.56 2400 0.4956 0.8725 0.7985 0.8725
0.0485 2.67 2500 0.4953 0.8732 0.7994 0.8732
0.0508 2.77 2600 0.4951 0.8738 0.8022 0.8738
0.047 2.88 2700 0.4934 0.8747 0.8027 0.8747
0.043 2.99 2800 0.4926 0.8749 0.8025 0.8749

Framework versions

  • Transformers 4.39.0.dev0
  • Pytorch 2.2.1+cu121
  • Datasets 2.18.0
  • Tokenizers 0.15.2