regex-llm


Nameregex-llm JSON
Version 0.1.0 PyPI version JSON
download
home_pagehttps://github.com/yanyongyu/regex-llm
SummaryControlled language model generation using regex
upload_time2023-05-30 07:10:20
maintainer
docs_urlNone
authoryanyongyu
requires_python>=3.8,<4.0
licenseMIT
keywords llm regex language model generation
VCS
bugtrack_url
requirements No requirements were recorded.
Travis-CI No Travis.
coveralls test coverage No coveralls.
            # regex llm

Constrained language model generation with regex.

## Installation

```bash
pip install regex-llm
```

## Usage

```python
import regex
from transformers import (
    AutoTokenizer,
    GenerationMixin,
    PreTrainedTokenizer,
    AutoModelForCausalLM,
)

from regex_llm import RegexConstraint

model: GenerationMixin = AutoModelForCausalLM.from_pretrained("your model")
tokenizer: PreTrainedTokenizer = AutoTokenizer.from_pretrained("your model")

controller = RegexConstraint(tokenizer=tokenizer, model=model)

prompt = '帮我生成3种不同的水果\n输出json格式如下:\n[{"name":"xxx","description":"xxx"}]'
pattern = r"\[(,?\s*\{\"name\":\s*\"[^\"]+\",\s*\"description\":\s*\"[^\"]+\"\\s*})+\s*\]"
output = controller.generate(
    prompt=prompt,
    pattern=pattern,
    do_sample=True,
    num_beams=1,
    top_p=0.7,
    temperature=0.95,
)
print(output)
```

            

Raw data

            {
    "_id": null,
    "home_page": "https://github.com/yanyongyu/regex-llm",
    "name": "regex-llm",
    "maintainer": "",
    "docs_url": null,
    "requires_python": ">=3.8,<4.0",
    "maintainer_email": "",
    "keywords": "llm,regex,language model,generation",
    "author": "yanyongyu",
    "author_email": "yyy@yyydl.top",
    "download_url": "https://files.pythonhosted.org/packages/13/7b/5a6dd6bd808dc55ef97bef558198282502631ccceb43b5f6a76c81c3d6e7/regex_llm-0.1.0.tar.gz",
    "platform": null,
    "description": "# regex llm\n\nConstrained language model generation with regex.\n\n## Installation\n\n```bash\npip install regex-llm\n```\n\n## Usage\n\n```python\nimport regex\nfrom transformers import (\n    AutoTokenizer,\n    GenerationMixin,\n    PreTrainedTokenizer,\n    AutoModelForCausalLM,\n)\n\nfrom regex_llm import RegexConstraint\n\nmodel: GenerationMixin = AutoModelForCausalLM.from_pretrained(\"your model\")\ntokenizer: PreTrainedTokenizer = AutoTokenizer.from_pretrained(\"your model\")\n\ncontroller = RegexConstraint(tokenizer=tokenizer, model=model)\n\nprompt = '\u5e2e\u6211\u751f\u62103\u79cd\u4e0d\u540c\u7684\u6c34\u679c\\n\u8f93\u51fajson\u683c\u5f0f\u5982\u4e0b\uff1a\\n[{\"name\":\"xxx\",\"description\":\"xxx\"}]'\npattern = r\"\\[(,?\\s*\\{\\\"name\\\":\\s*\\\"[^\\\"]+\\\",\\s*\\\"description\\\":\\s*\\\"[^\\\"]+\\\"\\\\s*})+\\s*\\]\"\noutput = controller.generate(\n    prompt=prompt,\n    pattern=pattern,\n    do_sample=True,\n    num_beams=1,\n    top_p=0.7,\n    temperature=0.95,\n)\nprint(output)\n```\n",
    "bugtrack_url": null,
    "license": "MIT",
    "summary": "Controlled language model generation using regex",
    "version": "0.1.0",
    "project_urls": {
        "Documentation": "https://github.com/yanyongyu/regex-llm",
        "Homepage": "https://github.com/yanyongyu/regex-llm",
        "Repository": "https://github.com/yanyongyu/regex-llm"
    },
    "split_keywords": [
        "llm",
        "regex",
        "language model",
        "generation"
    ],
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "blake2b_256": "fcb1a21f033c6c471befccb2e55b53320d86095d12f98c49b05fc1672409c1c2",
                "md5": "691484e081bec8dd0f4e799c66e2830b",
                "sha256": "f5d0b1ebbc57fd14ef7ddb43d8a0c1baf791aab4d65e21aa097651ebc082b52d"
            },
            "downloads": -1,
            "filename": "regex_llm-0.1.0-py3-none-any.whl",
            "has_sig": false,
            "md5_digest": "691484e081bec8dd0f4e799c66e2830b",
            "packagetype": "bdist_wheel",
            "python_version": "py3",
            "requires_python": ">=3.8,<4.0",
            "size": 2946,
            "upload_time": "2023-05-30T07:10:18",
            "upload_time_iso_8601": "2023-05-30T07:10:18.918427Z",
            "url": "https://files.pythonhosted.org/packages/fc/b1/a21f033c6c471befccb2e55b53320d86095d12f98c49b05fc1672409c1c2/regex_llm-0.1.0-py3-none-any.whl",
            "yanked": false,
            "yanked_reason": null
        },
        {
            "comment_text": "",
            "digests": {
                "blake2b_256": "137b5a6dd6bd808dc55ef97bef558198282502631ccceb43b5f6a76c81c3d6e7",
                "md5": "1a47ff47c3ecb5f06ac20a5e4fede517",
                "sha256": "d12b2d69d956fa7d3924cd551f3712a2629a0064fc0c0ef69cd53fc05a65afca"
            },
            "downloads": -1,
            "filename": "regex_llm-0.1.0.tar.gz",
            "has_sig": false,
            "md5_digest": "1a47ff47c3ecb5f06ac20a5e4fede517",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": ">=3.8,<4.0",
            "size": 2784,
            "upload_time": "2023-05-30T07:10:20",
            "upload_time_iso_8601": "2023-05-30T07:10:20.966018Z",
            "url": "https://files.pythonhosted.org/packages/13/7b/5a6dd6bd808dc55ef97bef558198282502631ccceb43b5f6a76c81c3d6e7/regex_llm-0.1.0.tar.gz",
            "yanked": false,
            "yanked_reason": null
        }
    ],
    "upload_time": "2023-05-30 07:10:20",
    "github": true,
    "gitlab": false,
    "bitbucket": false,
    "codeberg": false,
    "github_user": "yanyongyu",
    "github_project": "regex-llm",
    "travis_ci": false,
    "coveralls": false,
    "github_actions": true,
    "lcname": "regex-llm"
}
        
Elapsed time: 0.11387s