InCoder-32B (Industrial-Coder-32B) is the first 32B-parameter code foundation model purpose-built for industrial code intelligence. While general code LLMs excel at standard programming tasks, their performance often degrades in industrial scenarios that require reasoning about hardware semantics, specialized language constructs, and strict resource constraints. InCoder-32B unifies code intelligence across: — Chip Design (Verilog / RTL) — GPU Kernel Optimization (CUDA / Triton) — Embedded Systems (ARM Cortex-M, STM32) — Compiler Optimization (x86-64 assembly, LLVM) — 3D Modeling (CAD/CAM via CadQuery / OpenCascade) The model supports a native long-context window of up to 128K tokens. — Paper: InCoder-32B: Code Foundation Model for Industrial Scenarios — Repository: GitHub — Industrial-Coder — Project Page: IndustrialCoder InCoder-32B was developed using a three-stage Code-Flow pipeline: 1. Pre-training & Annealing: Curated industrial code mixed with general code pre-training using multi-level deduplication. 2. Mid-training (Context Extension): Progressive context extension from 8K to 128K tokens using synthetic industrial reasoning data and agent trajectories. 3. Post-training:…
Модальности:
Генерация текста
Области применения:
Генерация кода Диалог / чат
Задача: Генерация текста
Автор: Multilingual-Multimodal-NLP
Теги: iquestcoder, code, industrial-code, long-context, conversational, custom_code
Лайков: 4 | Загрузок: 330
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
