G0513HMA19H / README.md
Litzy619's picture
End of training
ace0f9d verified
metadata
license: gemma
base_model: google/gemma-2b
tags:
  - generated_from_trainer
model-index:
  - name: G0513HMA19H
    results: []

G0513HMA19H

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: 0.1260

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: 0.0003
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 16
  • total_train_batch_size: 128
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine_with_restarts
  • lr_scheduler_warmup_steps: 60
  • num_epochs: 3
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
3.1408 0.09 10 2.7234
2.2777 0.18 20 1.6102
1.1251 0.27 30 0.5583
0.3105 0.36 40 0.1700
0.1607 0.45 50 0.1518
0.1511 0.54 60 0.1510
0.1503 0.63 70 0.1517
0.1515 0.73 80 0.1494
0.1429 0.82 90 0.1487
0.1455 0.91 100 0.1479
0.149 1.0 110 0.1491
0.1437 1.09 120 0.1489
0.1447 1.18 130 0.1486
0.1445 1.27 140 0.1475
0.1484 1.36 150 0.1462
0.1417 1.45 160 0.1461
0.1436 1.54 170 0.1454
0.1444 1.63 180 0.1448
0.1456 1.72 190 0.1476
0.1439 1.81 200 0.1442
0.1421 1.9 210 0.1423
0.1406 1.99 220 0.1393
0.135 2.08 230 0.1362
0.131 2.18 240 0.1343
0.128 2.27 250 0.1325
0.1349 2.36 260 0.1333
0.1302 2.45 270 0.1316
0.1237 2.54 280 0.1280
0.12 2.63 290 0.1267
0.1219 2.72 300 0.1260
0.1235 2.81 310 0.1259
0.1244 2.9 320 0.1260
0.1228 2.99 330 0.1260

Framework versions

  • Transformers 4.36.0.dev0
  • Pytorch 2.1.2+cu121
  • Datasets 2.14.6
  • Tokenizers 0.14.0