BaseRT: Best-in-Class LLM Inference on Apple Silicon via Native Metal
• 3
AI & ML interests
local AI, GPU kernels, performance optimisation, agentic research pipelines, automated research
Recent Activity
View all activity
Articles
basecompute/Qwen3-30B-A3B-Instruct-2507
Updated
basecompute/Qwen3-4B-Thinking-2507
Updated • 15
basecompute/Qwen3-4B-Instruct-2507
Updated • 15
basecompute/Mistral-7B-Instruct-v0.3
Text Generation • Updated • 150
basecompute/Llama-3.1-8B-Instruct
Text Generation • Updated • 143
basecompute/Llama-3.2-3B-Instruct
Text Generation • Updated • 139
basecompute/Qwen3-8B
Text Generation • Updated • 145
basecompute/Qwen3-1.7B
Text Generation • Updated • 141
basecompute/Qwen3-30B-A3B-Thinking-2507
Text Generation • Updated • 144
basecompute/gemma-4-26B-A4B-it
Text Generation • Updated • 139
basecompute/nomic-embed-text-v1.5
Feature Extraction • Updated • 20
basecompute/gemma-4-E4B-it
Text Generation • Updated • 139
basecompute/gemma-4-E2B-it
Text Generation • Updated • 137
basecompute/Qwen3-4B
Text Generation • Updated • 136
basecompute/Llama-3.2-1B-Instruct
Text Generation • Updated • 139
basecompute/gemma-3-1b-it
Text Generation • Updated • 130
basecompute/Qwen3-0.6B
Text Generation • Updated • 137