920 B
Compatibility
TorchMLX targets the common transformer operations used by GPT-2, Llama 3, Qwen 3, and GPT-OSS style implementations.
Supported MLX operations include embeddings, linear layers, normalization building blocks, dropout, activations, causal attention, tensor shape operations, masks, top-k routing, and AdamW training through Trainer.
MLX arrays remain native arrays. Torch-style tensor methods are installed on the native array type for the supported subset.
Boolean expert routing and unique execute eagerly because their output shapes control Python flow.
Set TORCHMLX_BACKEND=torch before import to use native PyTorch for unsupported programs. TorchMLX never changes backend during an operation.
The referenced OpenArch model files contain source errors independent of TorchMLX, including invalid constructor calls and undefined attributes. Correct those errors before using either backend.