void-main / fastertransformer_backend
☆21Updated last year
Related projects: ⓘ
- Transformer related optimization, including BERT, GPT☆58Updated last year
- ☆123Updated 3 months ago
- Transformer related optimization, including BERT, GPT☆39Updated last year
- ☆70Updated 9 months ago
- Transformer related optimization, including BERT, GPT☆17Updated last year
- ☆140Updated 4 months ago
- A LLaMA1/LLaMA12 Megatron implement.☆26Updated 9 months ago
- Train llm (bloom, llama, baichuan2-7b, chatglm3-6b) with deepspeed pipeline mode. Faster than zero/zero++/fsdp.☆90Updated 7 months ago
- ☆105Updated last week
- FlagScale is a large model toolkit based on open-sourced projects.☆129Updated last week
- ☆60Updated last month
- Compare different hardware platforms via the Roofline Model for LLM inference tasks.☆71Updated 6 months ago
- ☆108Updated 6 months ago
- ☆82Updated last year
- An easy-to-use package for implementing SmoothQuant for LLMs☆78Updated 4 months ago
- Ongoing research training transformer language models at scale, including: BERT & GPT-2☆67Updated last year
- Disaggregated serving system for Large Language Models (LLMs).☆278Updated last month
- Performance of the C++ interface of flash attention and flash attention v2 in large language model (LLM) inference scenarios.☆20Updated 2 weeks ago
- [USENIX ATC '24] Accelerating the Training of Large Language Models using Efficient Activation Rematerialization and Optimal Hybrid Paral…☆41Updated last month
- PyTorch bindings for CUTLASS grouped GEMM.☆57Updated 2 months ago
- A general 2-8 bits quantization toolbox with GPTQ/AWQ/HQQ, and export to onnx/onnx-runtime easily.☆141Updated 3 weeks ago
- ☆67Updated last week
- ☆251Updated last week
- Latency and Memory Analysis of Transformer Models for Training and Inference☆338Updated 3 months ago
- A MoE impl for PyTorch, [ATC'23] SmartMoE☆56Updated last year
- Summary of system papers/frameworks/codes/tools on training or serving large model☆56Updated 9 months ago
- LLM Inference benchmark☆331Updated last month
- ☆145Updated last month
- ☆170Updated this week
- USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inference☆309Updated this week