---
library_name: transformers
tags:
- generated_from_trainer
datasets:
- kanishka/babylm2-rewritten-clean-spacy_random-removal-num-adj
metrics:
- accuracy
model-index:
- name: opt-babylm2-rewritten-clean-spacy_random-removal-num-adj-earlystop-bpe_seed-42_1e-3
  results:
  - task:
      name: Causal Language Modeling
      type: text-generation
    dataset:
      name: kanishka/babylm2-rewritten-clean-spacy_random-removal-num-adj
      type: kanishka/babylm2-rewritten-clean-spacy_random-removal-num-adj
    metrics:
    - name: Accuracy
      type: accuracy
      value: 0.47825513461811886
---

<!-- 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. -->

# opt-babylm2-rewritten-clean-spacy_random-removal-num-adj-earlystop-bpe_seed-42_1e-3

This model was trained from scratch on the kanishka/babylm2-rewritten-clean-spacy_random-removal-num-adj dataset.
It achieves the following results on the evaluation set:
- Loss: 2.6863
- Accuracy: 0.4783

## 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: 0.001
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 256
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 32000
- num_epochs: 20.0
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch   | Step  | Validation Loss | Accuracy |
|:-------------:|:-------:|:-----:|:---------------:|:--------:|
| 4.1066        | 0.9997  | 2240  | 3.8160          | 0.3611   |
| 3.446         | 1.9997  | 4480  | 3.3024          | 0.4093   |
| 3.1301        | 2.9997  | 6720  | 3.0883          | 0.4305   |
| 2.9677        | 3.9997  | 8960  | 2.9857          | 0.4410   |
| 2.8452        | 4.9997  | 11200 | 2.9248          | 0.4470   |
| 2.7915        | 5.9997  | 13440 | 2.8839          | 0.4515   |
| 2.746         | 6.9997  | 15680 | 2.8591          | 0.4544   |
| 2.7113        | 7.9997  | 17920 | 2.8384          | 0.4566   |
| 2.6925        | 8.9997  | 20160 | 2.8258          | 0.4583   |
| 2.6723        | 9.9997  | 22400 | 2.8152          | 0.4594   |
| 2.657         | 10.9997 | 24640 | 2.8094          | 0.4603   |
| 2.6394        | 11.9997 | 26880 | 2.7989          | 0.4615   |
| 2.6428        | 12.9997 | 29120 | 2.7990          | 0.4618   |
| 2.636         | 13.9997 | 31360 | 2.7911          | 0.4622   |
| 2.6195        | 14.9997 | 33600 | 2.7706          | 0.4649   |
| 2.5716        | 15.9997 | 35840 | 2.7441          | 0.4685   |
| 2.5238        | 16.9997 | 38080 | 2.7221          | 0.4713   |
| 2.4675        | 17.9997 | 40320 | 2.7048          | 0.4741   |
| 2.4042        | 18.9997 | 42560 | 2.6900          | 0.4769   |
| 2.3318        | 19.9997 | 44800 | 2.6863          | 0.4783   |


### Framework versions

- Transformers 4.48.0
- Pytorch 2.5.1
- Datasets 3.2.0
- Tokenizers 0.21.0