metadata
library_name: transformers
license: mit
base_model: gpt2
tags:
- text-generation
- generated_from_trainer
model-index:
- name: gpt2-pt-code-completion-causal-lm
results: []
gpt2-pt-code-completion-causal-lm
This model is a fine-tuned version of gpt2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.8459
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: 0.0005
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 50
- num_epochs: 1
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 5.4833 | 0.1900 | 200 | 4.2306 |
| 3.678 | 0.3801 | 400 | 3.5269 |
| 3.1166 | 0.5701 | 600 | 3.1171 |
| 2.8183 | 0.7602 | 800 | 2.9114 |
| 2.664 | 0.9502 | 1000 | 2.8459 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.4.0+cu121
- Datasets 3.0.0
- Tokenizers 0.19.1