https://cybertoolchain.io/tools/llm-guard New releases and features in LLM Guard, tracked by The Cyber Toolchain. en Mon, 17 Jun 2024 00:00:00 GMT https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Mon, 17 Jun 2024 00:00:00 GMT LLM Guard 0.3.14 adds per-scan scanner suppression and custom uvicorn config; drops gunicorn and caching • Adds the ability to suppress specific scanners during analysis in the API. • Allows custom `uvicorn` configuration in the API deployment. Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Tue, 23 Apr 2024 00:00:00 GMT LLM Guard 0.3.12 adds `BanCode` scanner, `NoRefusalLight`, lazy loading, and `POST /scan/prompt` API endpoint • Adds `BanCode` scanner to detect and block code snippets in the prompt • Adds `NoRefusalLight` scanner to detect refusal using common phrases • Adds lazy loading of models in the API deployment • Adds support for `gunicorn` in the API deployment • Adds `POST /scan/prompt` endpoint to scan prompts without sanitizing them • Adds local folder model loading support to the `API` Breaking changes: • The `model_kwargs` and `pipeline_kwargs` parameters are removed as they are now part of the new `Model` object Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Thu, 14 Mar 2024 00:00:00 GMT LLM Guard 0.3.10 adds Gibberish scanner, AI4Privacy NER models, and local model loading. • Adds `Gibberish` scanner to detect gibberish in input text. • Adds support for Isotonic/distilbert_finetuned_ai4privacy_v2 and Isotonic/deberta-v3-base_finetuned_ai4privacy_v2 NER models for the Anonymize capability. • Enables loading models from local folders instead of requiring HuggingFace access. • Allows passing `pipeline` and `model` kwargs to scanners for better control. Breaking changes: • `Anonymize` scanner patterns are now configured in Python instead of a JSON file. Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Thu, 08 Feb 2024 00:00:00 GMT LLM Guard 0.3.9 adds API auth, rate limiting, ANONYMIZE zh support, and NoRefusal match_type control. • API: Adds HTTP basic authentication and API key authentication endpoints • API: Adds rate limiting configuration via `auth` section in config • API: Adds OpenTelemetry support for tracing and metrics • Anonymize: Adds Chinese (`zh`) language support to `Anonymize` function • Anonymize: Adds new regex patterns `PO_BOX_RE`, `PRICE_RE`, `HEX_COLOR`, `TIME_RE`, `DATE_RE`, `URL_RE`, `PHONE_NUMBER_WITH_EXT`, `BTC_ADDRESS` for data masking • NoRefusal: Adds `match_type` parameter (values: `sentence`, `all`) to control matching scope Breaking changes: • `Code`: updated model requires updating the `languages` parameter to match [philomath-1209/programming-language-identification](https://huggingface.org/philomath-1209/programming-language-identification) vocab • `API` configuration changes with separate sections for `auth`, `rate_limit` and `cache` Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Thu, 21 Dec 2023 00:00:00 GMT LLM Guard 0.3.4 adds matching strategies to Deanonymize scanner and GPU ONNX support • `Deanonymize` scanner now supports configurable matching strategies. • Adds `ONNX` runtime support on GPU for the `Deanonymize` scanner, enabling faster inference. Breaking changes: • The `Anonymize` scanner no longer uses the `dbmdz/bert-large-cased-finetuned-conll03-english` model. Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Wed, 15 Nov 2023 00:00:00 GMT LLM Guard 0.3.2 adds ONNX inference support for multiple scanners, improves MaliciousURLs model, and updates Langchain integration. • Adds `ONNX` inference support to `BanTopics`, `NoRefusal`, `FactualConsistency`, and `Relevance` scanners • Switches the `MaliciousURLs` scanner to use the `DunnBC22/codebert-base-Malicious_URLs` model • Updates the `Language` and `LanguageSame` scanners to use the `papluca/xlm-roberta-base-language-detection` model • Provides an up-to-date Langchain integration using LCEL Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Thu, 09 Nov 2023 00:00:00 GMT Adds `use_onnx` parameter to enable ONNX runtime for faster CPU inference in scanners. • Enables ONNX runtime support in scanners by adding the `use_onnx` parameter. • Switches `PromptInjection` scanner to use `deepset/deberta-v3-base-injection` model. Breaking changes: • Removes the `PromptInjectionV2` scanner; users must now use the single `PromptInjection` scanner. Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Sat, 14 Oct 2023 00:00:00 GMT LLM Guard 0.3.0 adds Regex, Language, JSON, and LanguageSame scanners plus configurable embedding models. • Adds `Regex` scanner to the prompt • Adds `Language` scanners for both prompt and output • Adds `JSON` output scanner capable of repairing broken JSON • Adds `LanguageSame` output scanner to verify prompt and output languages match • Enables configurable embedding models in the `Relevance` scanner Breaking changes: • The `Refutation` scanner is renamed to `FactualConsistency` • Removed chunking behavior in `Anonymize` and `Sensitive` scanners which previously affected redaction output Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Sat, 07 Oct 2023 00:00:00 GMT LLM Guard v0.2.4 adds prompt injection scanner v2, sub-string redaction via BanSubstrings, and new secrets detectors. • Adds `BanSubstrings` capability to redact found substrings in evaluated content • Introduces new secrets detectors for identifying sensitive information in prompts or outputs • Adds Langchain LCEL integration example demonstrating expression language usage Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Sat, 23 Sep 2023 00:00:00 GMT LLM Guard 0.2.3 adds `fail_fast` flag to stop execution on first failure and updates API/Playground support. • Adds `fail_fast` flag to stop the execution after the first failure Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Thu, 21 Sep 2023 00:00:00 GMT LLM Guard 0.2.1 adds `Regex` output scanner with redaction and `threshold` param to `Code` scanner. • `Regex` output scanner can now redact text using a user-provided regular expression pattern. • Adds `threshold` parameter to `Code` scanners to control the similarity threshold for detection. Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Fri, 15 Sep 2023 00:00:00 GMT LLM Guard 0.2.0 adds GPU inference support and Streamlit demo app • Adds `GPU` inference support for scanners to accelerate processing • Includes a Streamlit demo application deployed to HuggingFace Spaces Breaking changes: • The `URL` entity type is removed from the `Anonymize` scanner, which may cause false-negative results for URLs that were previously detected. Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Sat, 12 Aug 2023 00:00:00 GMT LLM Guard v0.1.0 introduces Refutation, MaliciousURLs, Secrets scanners with risk scoring and faker support • Adds `Refutation output scanner` for factual consistency checking • Adds `MaliciousURLs output scanner` to detect malicious URLs in model outputs • Adds `Secrets prompt scanner` to detect secrets in prompts • Introduces `risk score` (0-1) for all prompt and output scanners • Adds `use_faker` parameter to Anonymize prompt scanner to use faker instead of placeholders • Updates Anonymize prompt scanner to use transformer based Spacy model `en_core_web_trf` Breaking changes: • `Anonymize prompt scanner` no longer uses `FILE_EXTENSION` entity type Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Mon, 07 Aug 2023 00:00:00 GMT LLM Guard 0.0.1 introduces the first input and output scanner classes for detecting prompt injection, toxicity, and sensitive data in LLM interactions. • `InputScanners` with tests: `Anonymize`, `BanSubstrings`, `BanTopics`, `Code`, `PromptInjection`, `Sentiment`, `TokenLimit`, and `Toxicity` • `OutputScanners` with tests: `BanSubstrings`, `BanTopics`, `Code`, `Deanonymize`, `NoRefusal`, `Regex`, `Relevance`, `Sensitive`, and `Toxicity` Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md https://github.com/protectai/llm-guard/blob/HEAD/docs/changelog.md Sun, 15 Jan 2023 00:00:00 GMT LLM Guard 0.3.7 adds URLReachability, BanCompetitors, InvisibleText, and ReadingTime scanners. • `URLReachability` scanner to check if the URL is reachable • `BanCompetitors` scanner to check if the prompt or output contains competitors' names • `InvisibleText` scanner to check if the prompt contains invisible unicode characters (steganography attack) • `ReadingTime` scanner to check if the output can be read in less than a certain amount of time Breaking changes: • `Regex` scanners have a new signature. It accepts `patterns`, `is_blocked` and `match_type`. • `BanSubstrings`: `match_type` parameter became `Enum` instead of `str`. • `Code` scanners have a new signature. It accepts `languages` and `is_blocked` instead of 2 separate lists. • Using keyword arguments for better readability of the code e.g. scanner = BanSubstrings(["a", "b", "c"], "str", False, True, False) would raise an error. Full history on Cyber Toolchain: https://cybertoolchain.io/tools/llm-guard