[ Tech stack ]
PyTorch
The deep-learning framework that dominates neural-network research and training.
PyTorch builds and trains neural networks with dynamic computation graphs (define-by-run) and autograd, running tensors on GPU (CUDA/ROCm). It's the de-facto standard for AI research, and torch.compile paired with TorchScript and ONNX export takes a model from prototype to production without a rewrite.
[ Why PyTorch at Dexon ]
What this technology does well, and why we use it.
Typical usage: Train and deploy custom vision, NLP or LLM models.
- 01
Dynamic graphs and autograd: natural debugging, fast iteration.
- 02
GPU tensors and native distributed multi-GPU training.
- 03
Rich ecosystem: torchvision, torchaudio, Hugging Face, Lightning.
- 04
torch.compile, TorchScript and ONNX export for production.
[ Complementary technologies ]
The building blocks we often mobilise alongside.
A stack rarely exists alone. Here are the technologies Dexon most often pairs with this one, through pipeline habits, usage similarity or internal mastery. Click on a building block to see its scope.
[ Reassurance ]
- 0+
- custom projects delivered
- 30
- engineers, designers, project managers
- 80 %
- from top French schools
- 24 h
- average reply time
[ They trust us ]
More than 100 French and European companies trust us
[ Press ]
They talk about us.
Nationwide coverage by BFM Business, Le Figaro, Challenges, La Tribune and CNews. An outside reading of our work and our innovations.
