File size: 2,648 Bytes
86a5699 924e4c5 86a5699 006bf74 7dcaf95 86a5699 230d3e3 7dcaf95 230d3e3 86a5699 006bf74 86a5699 006bf74 7dcaf95 86a5699 ce0b0fd a2171f4 ce0b0fd 4cb6da6 ce0b0fd 86a5699 924e4c5 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 |
---
tags:
- generated_from_trainer
datasets:
- jed351/shikoto_zh_hk
metrics:
- accuracy
model-index:
- name: gpt2-shikoto
results:
- task:
name: Causal Language Modeling
type: text-generation
dataset:
name: jed351/shikoto_zh_hk
type: jed351/shikoto_zh_hk
metrics:
- name: Accuracy
type: accuracy
value: 0.37381769930940056
license: openrail
---
<!-- 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. -->
# gpt2-shikoto
This model was trained on a dataset I obtained from an online novel site.
**Please be aware that the stories (training data) might contain inappropriate content. This model is intended for research purposes only.**
The base model can be found [here](https://huggingface.co/jed351/gpt2-tiny-zh-hk), which was obtained by
patching a [GPT2 Chinese model](https://huggingface.co/ckiplab/gpt2-tiny-chinese) and its tokenizer with Cantonese characters.
Refer to the base model for info on the patching process.
## Training procedure
Please refer to the [script](https://github.com/huggingface/transformers/tree/main/examples/pytorch/language-modeling)
provided by Huggingface.
The model was trained for 400,000 steps on 2 NVIDIA Quadro RTX6000 for around 15 hours at the Research Computing Services of Imperial College London.
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 20
- eval_batch_size: 20
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- total_train_batch_size: 40
- total_eval_batch_size: 40
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 400000
- mixed_precision_training: Native AMP
### Training results
### How to use it?
```
from transformers import AutoTokenizer
from transformers import TextGenerationPipeline, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("jed351/gpt2-tiny-zh-hk")
model = AutoModelForCausalLM.from_pretrained("jed351/gpt2_tiny_zh-hk-shikoto")
# try messing around with the parameters
generator = TextGenerationPipeline(model, tokenizer,
max_new_tokens=200,
no_repeat_ngram_size=3) #, device=0) #if you have a GPU
input_string = "your input"
output = generator(input_string)
string = output[0]['generated_text'].replace(' ', '')
print(string)
```
### Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.1
- Datasets 2.8.0
- Tokenizers 0.13.2 |