| --- |
| license: mit |
| tags: |
| - generated_from_trainer |
| model-index: |
| - name: Fuxi |
| results: [] |
| --- |
| |
| # Fuxi |
|
|
| 负屃是一个Python代码生成模型，训练过程主要参考了[Training a causal language model from scratch](https://huggingface.co/learn/nlp-course/en/chapter7/6) |
|
|
| 在中国神话中，负屃是龙的第八个儿子。似龙形，平生好文。负屃十分爱好闪耀着艺术光彩的碑文，它甘愿化做图案文龙去衬托这些传世的文学珍品。 |
|
|
| Fuxi is a Python code generation model, and the training process mainly refers to [Training a causal language model from scratch](https://huggingface.co/learn/nlp-course/en/chapter7/6). |
|
|
| In Chinese mythology, Fuxi is the eighth son of the Dragon. He resembles a dragon, and has a strong affinity for literature. Fuxi is particularly fond of inscriptions shining with artistic brilliance, and is willing to transform into an ornamental dragon to set off these enduring literary treasures. |
|
|
| ## 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: 64 |
| - eval_batch_size: 64 |
| - seed: 42 |
| - gradient_accumulation_steps: 8 |
| - total_train_batch_size: 512 |
| - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
| - lr_scheduler_type: cosine |
| - lr_scheduler_warmup_steps: 1000 |
| - num_epochs: 1 |
|
|
| ### Training results |
|
|
| | Training Loss | Epoch | Step | Validation Loss | |
| |:-------------:|:-----:|:-----:|:---------------:| |
| | 2.3403 | 0.15 | 5000 | 1.5401 | |
| | 1.5052 | 0.31 | 10000 | 1.3514 | |
| | 1.3657 | 0.46 | 15000 | 1.2464 | |
| | 1.2715 | 0.61 | 20000 | 1.1665 | |
| | 1.1977 | 0.77 | 25000 | 1.1059 | |
| | 1.1498 | 0.92 | 30000 | 1.0777 | |
|
|
|
|
| ### Framework versions |
|
|
| - Transformers 4.30.2 |
| - Pytorch 2.0.1+cu117 |
| - Datasets 2.13.1 |
| - Tokenizers 0.13.3 |
|
|
