foundation-model-stack / fastsafetensorsLinks
High-performance safetensors model loader
☆99Updated 3 weeks ago
Alternatives and similar repositories for fastsafetensors
Users that are interested in fastsafetensors are comparing it to the libraries listed below
Sorting:
- A safetensors extension to efficiently store sparse quantized tensors on disk☆238Updated this week
- Genai-bench is a powerful benchmark tool designed for comprehensive token-level performance evaluation of large language model (LLM) serv…☆263Updated this week
- CUDA checkpoint and restore utility☆410Updated 4 months ago
- Module, Model, and Tensor Serialization/Deserialization☆286Updated 5 months ago
- ☆31Updated 9 months ago
- The driver for LMCache core to run in vLLM☆60Updated last year
- Fast and memory-efficient exact attention☆111Updated last week
- torchcomms: a modern PyTorch communications API☆327Updated this week
- ☆280Updated this week
- extensible collectives library in triton☆95Updated 10 months ago
- Boosting 4-bit inference kernels with 2:4 Sparsity☆93Updated last year
- Accelerating MoE with IO and Tile-aware Optimizations☆569Updated 3 weeks ago
- ArcticInference: vLLM plugin for high-throughput, low-latency inference☆384Updated last week
- DeeperGEMM: crazy optimized version☆73Updated 9 months ago
- ☆206Updated 9 months ago
- NVIDIA NVSHMEM is a parallel programming interface for NVIDIA GPUs based on OpenSHMEM. NVSHMEM can significantly reduce multi-process com…☆462Updated last month
- ☆48Updated last year
- A unified library for building, evaluating, and storing speculative decoding algorithms for LLM inference in vLLM☆220Updated last week
- Perplexity GPU Kernels☆554Updated 3 months ago
- An early research stage expert-parallel load balancer for MoE models based on linear programming.☆495Updated 2 months ago
- KV cache compression for high-throughput LLM inference☆153Updated last year
- Tilus is a tile-level kernel programming language with explicit control over shared memory and registers.☆440Updated last month
- Offline optimization of your disaggregated Dynamo graph☆177Updated last week
- NVIDIA Inference Xfer Library (NIXL)☆876Updated this week
- [ICLR2025] Breaking Throughput-Latency Trade-off for Long Sequences with Speculative Decoding☆141Updated last year
- ☆342Updated last week
- Accepted to MLSys 2026☆70Updated last week
- NVIDIA Resiliency Extension is a python package for framework developers and users to implement fault-tolerant features. It improves the …☆262Updated this week
- ☆118Updated 8 months ago
- kernels, of the mega variety☆665Updated last week