pipeline_tag: text-generation
library_name: transformers
tags:
- code
- industrial-ai
- code-generation
InCoder-32B: Industrial Code Foundation Model
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, they often struggle with hardware semantics, specialized language constructs, and strict resource constraints.
InCoder-32B is designed to unify 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.
Resources
- Paper: InCoder-32B: Code Foundation Model for Industrial Scenarios
- Repository: GitHub - Industrial-Coder
- Project Page: IndustrialCoder Project Page
Quickstart
To use InCoder-32B with the transformers library, you can follow the snippet below. Note that trust_remote_code=True is required to load the custom model architecture.
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "Multilingual-Multimodal-NLP/IndustrialCoder"
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto",
trust_remote_code=True,
)
messages = [{"role": "user", "content": "Optimize this CUDA kernel for better memory coalescing."}]
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer([text], return_tensors="pt").to(model.device)
with torch.no_grad():
out = model.generate(**inputs, max_new_tokens=2048, temperature=0.6, top_p=0.85, top_k=20)
print(tokenizer.decode(out[0][inputs["input_ids"].shape[-1]:], skip_special_tokens=True))
Performance
Industrial Code Benchmarks
InCoder-32B establishes strong open-source baselines across specialized industrial domains, often surpassing proprietary models like Claude-Sonnet-4.6 in specific tasks.
| Domain | Benchmark | InCoder-32B | Claude-Sonnet-4.6 |
|---|---|---|---|
| Chip Design | VeriScope Score | 80.7 | 87.7 |
| GPU Optim. | KernelBench L1/L2/L3 | 22.2/36.0/14.0 | 11.1/28.0/2.0 |
| 3D Modeling | CAD-Coder Compile (%) | 82.0 | 77.0 |
| 3D Modeling | CAD-Coder IoU | 53.5 | 32.4 |
Citation
If you find InCoder-32B useful in your research, please cite:
@article{yang2025incoder,
title={InCoder-32B: Code Foundation Model for Industrial Scenarios},
author={Yang, Jian and Zhang, Wei and Wu, Jiajun and Cheng, Junhang and Guo, Shawn and Wang, Haowen and others},
journal={arXiv preprint arXiv:2603.16790},
year={2025}
}
Disclaimer
The model may generate incorrect or unsafe code. Always review and test outputs in a sandboxed environment before production use. Industrial code (RTL, embedded firmware, GPU kernels) requires expert human review before deployment in physical systems or hardware synthesis.