AOLM3
This model is a fine-tuned version of allenai/OLMo-1B on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.1426
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: 80
- num_epochs: 3
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
2.9989 | 0.09 | 10 | 1.0026 |
0.3997 | 0.18 | 20 | 0.1602 |
0.1542 | 0.27 | 30 | 0.1613 |
0.1535 | 0.36 | 40 | 0.1525 |
0.1526 | 0.45 | 50 | 0.1486 |
0.1508 | 0.54 | 60 | 0.1496 |
0.1494 | 0.63 | 70 | 0.1490 |
0.1493 | 0.73 | 80 | 0.1514 |
0.1471 | 0.82 | 90 | 0.1513 |
0.1477 | 0.91 | 100 | 0.1507 |
0.1493 | 1.0 | 110 | 0.1496 |
0.1453 | 1.09 | 120 | 0.1498 |
0.1446 | 1.18 | 130 | 0.1523 |
0.147 | 1.27 | 140 | 0.1481 |
0.1478 | 1.36 | 150 | 0.1476 |
0.146 | 1.45 | 160 | 0.1496 |
0.1462 | 1.54 | 170 | 0.1466 |
0.1462 | 1.63 | 180 | 0.1452 |
0.1471 | 1.72 | 190 | 0.1515 |
0.1453 | 1.81 | 200 | 0.1469 |
0.1487 | 1.9 | 210 | 0.1475 |
0.1461 | 1.99 | 220 | 0.1483 |
0.1444 | 2.08 | 230 | 0.1468 |
0.139 | 2.18 | 240 | 0.1450 |
0.1411 | 2.27 | 250 | 0.1461 |
0.1405 | 2.36 | 260 | 0.1464 |
0.1392 | 2.45 | 270 | 0.1446 |
0.1379 | 2.54 | 280 | 0.1441 |
0.1368 | 2.63 | 290 | 0.1444 |
0.1389 | 2.72 | 300 | 0.1427 |
0.1387 | 2.81 | 310 | 0.1421 |
0.1367 | 2.9 | 320 | 0.1425 |
0.1396 | 2.99 | 330 | 0.1426 |
Framework versions
- Transformers 4.36.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.14.1
Model tree for Litzy619/AOLM3
Base model
allenai/OLMo-1B