Welcome to Efficient-Transformers Documentation!

Getting Started

• Introduction Qualcomm efficient-transformers library
• Validated Models
• Models Coming Soon

Installation

• Pre-requisites
• Installation
  • 1. Download Apps SDK
  • 2. Install Efficient-Transformers
• Sanity Check

Upgrade Efficient-Transformers

• Using GitHub Repository

Quick start

• Transformed models and QPC storage
• Command Line Interface
  • QEfficient.cloud.infer
  • QEfficient.cloud.execute
  • QEfficient.cloud.finetune
  • Multi-Qranium Inference
  • Continuous Batching
  • QNN Compilation
• Python API
  • 1. Model download and Optimize for Cloud AI 100
  • 2. Export and Compile with one API
  • 3. Execute
  • Draft-Based Speculative Decoding

Command Line Interface Use (CLI)

• QEfficient.cloud.infer
• QEfficient.cloud.execute
• QEfficient.cloud.compile
• QEfficient.cloud.export
• QEfficient.cloud.finetune

Python API

• High Level API
  • QEFFAutoModelForCausalLM
  • QEFFAutoModel
  • QEffAutoPeftModelForCausalLM
  • QEffAutoLoraModelForCausalLM
  • export
  • compile
  • Execute
• Low Level API
  • convert_to_cloud_kvstyle
  • convert_to_cloud_bertstyle
  • utils

QAIC Finetune

• Finetune Infra
  • Installation
  • Finetuning
  • Dataset Details
  • Usage
  • Visualization

Blogs

• Train anywhere, Infer on Qualcomm Cloud AI 100
• How to Quadruple LLM Decoding Performance with Speculative Decoding (SpD) and Microscaling (MX) Formats on Qualcomm® Cloud AI 100
• Power-efficient acceleration for large language models – Qualcomm Cloud AI SDK
• Qualcomm Cloud AI 100 Accelerates Large Language Model Inference by ~2x Using Microscaling (Mx) Formats
• Qualcomm Cloud AI Introduces Efficient Transformers: One API, Infinite Possibilities

Reference

• Qualcomm Cloud AI home
• Qualcomm Cloud AI SDK download
• Qualcomm Cloud AI API reference
• User Guide
• OCP Microscaling Formats (MX) Specification