Skills tagged "pytorch"

forward-forward-learning - Forward-Forward Algorithm for Local Layer-Wise Learning
forward-forward-algorithm local-learning neural-networks pytorch
★ 66 · Updated 2026-09-23
Implements Hinton's Forward-Forward algorithm for local layer-wise neural network training without backpropagation.
⚙ Train layers without backpropagation ⚙ Compute layer goodness functions ⚙ Generate self-contrastive negative samples
cnn-inference-on-spectral-data - Generate molecular embeddings from MS/MS spectra with CNN
mass-spectrometry metabolomics cnn inference
★ 15 · Updated 2026-09-19
Passes preprocessed MS/MS spectra through a trained CNN model to extract multidimensional molecular embedding vectors for downstream metabolite matching.
⚙ Load preprocessed MS/MS spectral data ⚙ Initialize trained CNN model checkpoint ⚙ Extract molecular embedding vectors
positional-encoding-for-spectral-sequences - Positional encoding for mass spectrum tensors in IDSL_MINT
metabolomics mass-spectrometry positional-encoding transformer
★ 15 · Updated 2026-09-17
Embeds ordinal and sequential information into mass spectrum m/z-intensity pairs for transformer encoder layers in IDSL_MINT.
⚙ Compute sinusoidal positional encodings ⚙ Add positional encodings to embeddings ⚙ Prepare spectrum input tensors
pytorch-lightning - PyTorch Lightning Deep Learning Framework
pytorch-lightning pytorch deep-learning distributed-training
★ 0 · Updated 2026-09-17
Organize PyTorch code into LightningModules, configure Trainers, implement data pipelines, callbacks, logging, and distributed training strategies.
⚙ Organize PyTorch models into LightningModules ⚙ Configure Trainers for multi-device training ⚙ Implement data pipelines using LightningDataModules
mhc-algorithm - Manifold-Constrained Hyper-Connections for Deep Networks
neural-networks pytorch deep-learning hyper-connections
★ 0 · Updated 2026-09-14
Implement mHC to stabilize deep network training using doubly stochastic matrices via the Sinkhorn-Knopp algorithm.
⚙ Implement Sinkhorn-Knopp algorithm ⚙ Project matrices to doubly stochastic ⚙ Build PyTorch HyperConnections module
mhc-algorithm - Implement Manifold-Constrained Hyper-Connections
mhc hyper-connections sinkhorn-knopp pytorch
★ 0 · Updated 2026-09-14
Stabilize deep network training by constraining residual mixing matrices to be doubly stochastic via log-space Sinkhorn-Knopp algorithm.
⚙ Project matrices using Sinkhorn-Knopp ⚙ Implement PyTorch HyperConnections module ⚙ Wrap attention and MLP layers
mhc-algorithm - Implement mHC for stabilizing deep network training
mhc hyper-connections sinkhorn-knopp pytorch
★ 48 · Updated 2026-09-14
Implement Manifold-Constrained Hyper-Connections using doubly stochastic matrices via Sinkhorn-Knopp algorithm to stabilize deep network training.
⚙ Implement HyperConnections module ⚙ Apply Sinkhorn-Knopp algorithm ⚙ Wrap attention and MLP layers
vmas-simulator-guide - VMAS Vectorized Multi-Agent Simulator Guide
vmas multi-agent-rl simulator pytorch
★ 1 · Updated 2026-09-13
Provides instructions for using VMAS, a PyTorch-based GPU-accelerated vectorized simulator for multi-agent reinforcement learning.
⚙ Create parallel vectorized environments ⚙ Step multi-agent environments ⚙ Integrate with TorchRL and RLlib
arxiv-2608-19889-write-once-run-everywhere-the-axon-dsl-for-shape-s - Axon DSL for Shape-Safe Framework-Agnostic LLMs
dsl llm pytorch jax
★ 3 · Updated 2026-09-11
Presents the Axon DSL for compiling shape-safe, framework-agnostic LLM architectures to PyTorch, JAX, MLX, and vLLM.
⚙ Compile LLMs to PyTorch implementations ⚙ Compile LLMs to JAX implementations ⚙ Compile LLMs to MLX implementations
metal-kernel - Implement Metal Kernels for PyTorch Operators
metal-kernels mps pytorch apple-silicon
★ 1 · Updated 2026-06-30
Write native Metal kernels and implement MPS device support for PyTorch operators on Apple Silicon
⚙ update native_functions.yaml dispatch ⚙ write Metal kernel files ⚙ implement host-side stub