Quantized Llama 2 (7B/13B/70B) at 2-bit via HARP, a learnable orthogonal preprocessor for extreme LLM quantization. EMNLP 2026.
AI & ML interests
None defined yet.
Recent Activity
View all activity
models 7
brain-lab/Llama-2-70b-QuIP-HARP-2Bit
Text Generation • 3B • Updated • 114
brain-lab/Llama-2-13b-QuIP-HARP-2Bit
Text Generation • 0.9B • Updated • 112
brain-lab/Llama-2-7b-QuIP-HARP-2Bit
Text Generation • 0.6B • Updated • 117
brain-lab/qjl-qwen3-gradient-faithful
4B • Updated • 9
brain-lab/kivi-qwen3-gradient-faithful
4B • Updated • 11
brain-lab/qjl-llama-gradient-faithful
8B • Updated • 10
brain-lab/kivi-llama-gradient-faithful
8B • Updated • 7
datasets 0
None public yet