AIAny
Icon for item

Hermes Function-Calling V1

Structured dataset for training and evaluating LLM agentic behavior: function-calling conversations, JSON-mode structured outputs, and extraction samples for teaching models to generate tool calls and strict structured responses. Includes single-turn and multi-turn scenarios across several configs.

Introduction

Oops! Something went wrong

[next-mdx-remote-client] error compiling MDX: Expected a closing tag for `<tool_response>` (6:74-6:89) before the end of `paragraph` 4 | ## What sets it apart 5 | - Multi-config mix: five dataset configs (func_calling_singleturn, func_calling, glaive_func_calling, json_mode_agentic, json_mode_singleturn) that cover single-turn calls, multi-turn agentic dialogues, cleaned glaive samples, and advanced JSON-mode responses. This variety helps models learn both simple and complex function-calling patterns. > 6 | - Explicit tool-call format: examples use XML-like `<tools>`/<tool_call>/<tool_response> conventions and pydantic-style schemas so models learn to emit machine-parseable function calls and to parse tool responses back into natural language or structured outputs. | ^ 7 | - Focus on structured extraction and agent skills: includes JSON-mode agentic samples and schema-driven extraction tasks useful for building reliable agents, RAG pipelines, and function-calling integrations. 8 | - Dataset scale & license: ~11,578 examples (size category 10K–100K), Apache-2.0 license, prepared for Hugging Face Datasets consumption with common libraries like pandas and polars. More information: https://mdxjs.com/docs/troubleshooting-mdx

Information

  • Websitehuggingface.co
  • OrganizationsNousResearch
  • Authorsinterstellarninja, Teknium, THEODOROS
  • Published date2024/08/14

Categories

More Items

Hugging Face

Selected ~245B-token corpus of algorithmically relevant code across 11 languages from The Stack v3, produced by file-level language-adaptive selection (UltraData-Code-L2-Classifier). Concentrates problem-solving implementations for LLM code pretraining while preserving broad language coverage; released under ODC-By.

Hugging Face

Contains 73 images used in Hugging Face library documentation, packaged in Parquet and imagefolder formats for examples, demos, and small-scale image experiments; distributed under CC BY‑NC‑SA 4.0.

Hugging Face

A monthly-updated, hive-partitioned parquet corpus of billions of standard rated chess games from Lichess for data analysis and ML — includes move lists with clock info, player Glicko2 ratings, openings, timestamps, and some Stockfish evals under CC0.