Eikos - OSS Jev-like model
Eikos - OSS Jev-like model caiovicentino / eikos Public Notifications You must be signed in to change notification settings Fork 1 Star 31 Branches Tags Open more actions menu Latest commit History 7 Commits 7 Commits Folders and files Name Name Last commit message Last commit date assets assets data_pipeline data_pipeline eikos eikos evaluation evaluation examples examples quantization quantization release_tools release_tools scripts scripts training training .env.example .env.example .gitignore .gitignore CHANGELOG.md CHANGELOG.md LICENSE LICENSE NOTICE NOTICE README.md README.md pyproject.toml pyproject.toml Repository files navigation Eikos Eikos (εἰκός, "the probable" ) is a family of open typed-decision models, released under MIT. Each model: answers a structured question about a given state in one forward pass ; returns a calibrated probability for every option , so a caller can act on confident decisions and escalate the rest.
This SecurityIncident is relevant to the technology intelligence record because it involves Apple, Hugging Face, GitHub, OpenAI. The source article should remain the factual reference for follow-up coverage.
- caiovicentino / eikos Public Notifications You must be signed in to change notification settings Fork 1 Star 31 Branches Tags Open more actions menu Latest commit History 7 Commits 7 Commits Folders and files Name Name Last commit message Last commit date assets assets data_pipeline data_pipeline eikos eikos evaluation evaluation examples examples quantization quantization release_tools release_tools scripts scripts training training .env.example .env.example .gitignore .gitignore CHANGELOG.md CHANGELOG.md LICENSE LICENSE NOTICE NOTICE README.md README.md pyproject.toml pyproject.toml Repository files navigation Eikos Eikos (εἰκός, "the probable" ) is a family of open typed-decision models, released under MIT.
- Each model: answers a structured question about a given state in one forward pass ; returns a calibrated probability for every option , so a caller can act on confident decisions and escalate the rest.
- It comes in two sizes, Eikos-4B and Eikos-27B, each with bf16, FP8 and INT4 GPU builds and MLX builds for Apple Silicon.
- The focus is global finance, trading and trade finance: applying stated rules, policies and rulebooks to a case.
- This repository contains everything used to build the models: the serving and inference code; the data pipeline; training; quantization; the evaluation harness.
- Models: caiovicentino1/Eikos-4B , caiovicentino1/Eikos-27B (with -FP8 , -INT4 and -MLX builds).