kdf/javascript-docstring-generation

1---2license: apache-2.03widget:4- text: "<|endoftext|>\nfunction getDateAfterNDay(n){\n    return moment().add(n, 'day')\n}\n// docstring\n/**"5---6 7## Basic info8 9model based [Salesforce/codegen-350M-mono](https://huggingface.co/Salesforce/codegen-350M-mono)10 11fine-tuned with data [codeparrot/github-code-clean](https://huggingface.co/datasets/codeparrot/github-code-clean)12 13data filter by JavaScript and TypeScript14 15## Usage16 17```python18from transformers import AutoTokenizer, AutoModelForCausalLM19 20model_type = 'kdf/javascript-docstring-generation'21tokenizer = AutoTokenizer.from_pretrained(model_type)22model = AutoModelForCausalLM.from_pretrained(model_type)23 24inputs = tokenizer('''<|endoftext|>25function getDateAfterNDay(n){26    return moment().add(n, 'day')27}28 29// docstring30/**''', return_tensors='pt')31 32doc_max_length = 12833 34generated_ids = model.generate(35    **inputs,36    max_length=inputs.input_ids.shape[1] + doc_max_length,37    do_sample=False,38    return_dict_in_generate=True,39    num_return_sequences=1,40    output_scores=True,41    pad_token_id=50256,42    eos_token_id=50256  # <|endoftext|>43)44 45ret = tokenizer.decode(generated_ids.sequences[0], skip_special_tokens=False)46print(ret)47 48```49 50## Prompt51 52You could give model a style or a specific language, for example:53 54```python55inputs = tokenizer('''<|endoftext|>56function add(a, b){57    return a + b;58}59// docstring60/**61  * Calculate number add.62  * @param a {number} the first number to add63  * @param b {number} the second number to add64  * @return the result of a + b65  */66<|endoftext|>67function getDateAfterNDay(n){68    return moment().add(n, 'day')69}70// docstring71/**''', return_tensors='pt')72 73doc_max_length = 12874 75generated_ids = model.generate(76    **inputs,77    max_length=inputs.input_ids.shape[1] + doc_max_length,78    do_sample=False,79    return_dict_in_generate=True,80    num_return_sequences=1,81    output_scores=True,82    pad_token_id=50256,83    eos_token_id=50256  # <|endoftext|>84)85 86ret = tokenizer.decode(generated_ids.sequences[0], skip_special_tokens=False)87print(ret)88 89inputs = tokenizer('''<|endoftext|>90function add(a, b){91    return a + b;92}93// docstring94/**95  * 计算数字相加96  * @param a {number} 第一个加数97  * @param b {number} 第二个加数98  * @return 返回 a + b 的结果99  */100<|endoftext|>101function getDateAfterNDay(n){102    return moment().add(n, 'day')103}104// docstring105/**''', return_tensors='pt')106 107doc_max_length = 128108 109generated_ids = model.generate(110    **inputs,111    max_length=inputs.input_ids.shape[1] + doc_max_length,112    do_sample=False,113    return_dict_in_generate=True,114    num_return_sequences=1,115    output_scores=True,116    pad_token_id=50256,117    eos_token_id=50256  # <|endoftext|>118)119 120ret = tokenizer.decode(generated_ids.sequences[0], skip_special_tokens=False)121print(ret)122 123```