Dream-Coder 7B is a diffusion LLM for code trained exclusively on open-source data across its development stages—adaptation, supervised fine-tuning, and reinforcement learning. It achieves an impressive 21.4% pass@1 on LiveCodeBench (2410-2505), outperforming other open-source diffusion LLMs by…
The main features of dreamlm/dream-coder are: Code and Formal Reasoning, Language Diffusion Models.
Open-source alternatives to dreamlm/dream-coder include: microsoft/bosquelanguage — Bosque is an experimental programming language and development platform designed for machine-assisted software… bytedance-seed/stable-diffcoder — Stable-DiffCoder: Pushing the Frontier of Code Diffusion Large Language Model. czg1225/dparallel — 🚀 dParallel: Learnable Parallel Decoding for dLLMs. dataarctech/sql-r1 — SQL-R1: Training Natural Language to SQL Reasoning Model By Reinforcement Learning. deepseek-ai/deepseek-prover-v2. ashenweli/discrete-diffusion-models-for-language-genaration — This thesis aimstoinvestigate thepotential of discrete diffusion models in the context ofnaturallanguagegeneration.
Bosque is an experimental programming language and development platform designed for machine-assisted software construction. It combines functional programming semantics with imperative syntax to enforce logic correctness and runtime safety, providing a type-safe environment that utilizes structured data models to maintain information integrity throughout the application lifecycle. The platform distinguishes itself through deep integration with formal verification tools, including automated theorem provers and symbolic execution engines. By transforming source code into a regularized intermed
Stable-DiffCoder: Pushing the Frontier of Code Diffusion Large Language Model
This thesis aimstoinvestigate thepotential of discrete diffusion models in the context ofnaturallanguagegeneration.
