|
|
--- |
|
|
license: apache-2.0 |
|
|
base_model: allenai/OLMo-1B |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
model-index: |
|
|
- name: AOLM1 |
|
|
results: [] |
|
|
--- |
|
|
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
|
|
# AOLM1 |
|
|
|
|
|
This model is a fine-tuned version of [allenai/OLMo-1B](https://huggingface.co/allenai/OLMo-1B) on an unknown dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 0.1410 |
|
|
|
|
|
## 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 | |
|
|
|:-------------:|:-----:|:----:|:---------------:| |
|
|
| 2.9303 | 0.09 | 10 | 0.9380 | |
|
|
| 0.3692 | 0.18 | 20 | 0.1548 | |
|
|
| 0.1539 | 0.27 | 30 | 0.1633 | |
|
|
| 0.1561 | 0.36 | 40 | 0.1560 | |
|
|
| 0.1518 | 0.45 | 50 | 0.1541 | |
|
|
| 0.1518 | 0.54 | 60 | 0.1478 | |
|
|
| 0.1488 | 0.63 | 70 | 0.1481 | |
|
|
| 0.1486 | 0.73 | 80 | 0.1550 | |
|
|
| 0.1467 | 0.82 | 90 | 0.1513 | |
|
|
| 0.1476 | 0.91 | 100 | 0.1491 | |
|
|
| 0.149 | 1.0 | 110 | 0.1483 | |
|
|
| 0.1456 | 1.09 | 120 | 0.1493 | |
|
|
| 0.1442 | 1.18 | 130 | 0.1517 | |
|
|
| 0.1474 | 1.27 | 140 | 0.1478 | |
|
|
| 0.1482 | 1.36 | 150 | 0.1495 | |
|
|
| 0.1455 | 1.45 | 160 | 0.1479 | |
|
|
| 0.1455 | 1.54 | 170 | 0.1474 | |
|
|
| 0.1467 | 1.63 | 180 | 0.1452 | |
|
|
| 0.1464 | 1.72 | 190 | 0.1486 | |
|
|
| 0.145 | 1.81 | 200 | 0.1469 | |
|
|
| 0.1485 | 1.9 | 210 | 0.1460 | |
|
|
| 0.1453 | 1.99 | 220 | 0.1480 | |
|
|
| 0.1432 | 2.08 | 230 | 0.1456 | |
|
|
| 0.1376 | 2.18 | 240 | 0.1444 | |
|
|
| 0.1392 | 2.27 | 250 | 0.1451 | |
|
|
| 0.1385 | 2.36 | 260 | 0.1441 | |
|
|
| 0.137 | 2.45 | 270 | 0.1441 | |
|
|
| 0.1352 | 2.54 | 280 | 0.1420 | |
|
|
| 0.1338 | 2.63 | 290 | 0.1423 | |
|
|
| 0.1352 | 2.72 | 300 | 0.1410 | |
|
|
| 0.1351 | 2.81 | 310 | 0.1407 | |
|
|
| 0.1317 | 2.9 | 320 | 0.1409 | |
|
|
| 0.1361 | 2.99 | 330 | 0.1410 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.36.0.dev0 |
|
|
- Pytorch 2.1.2+cu121 |
|
|
- Datasets 2.14.6 |
|
|
- Tokenizers 0.14.1 |
|
|
|
