NEW BITNET MODELS!
Summary
New BitCPM4-CANN models (1B, 3B, 8B) from OpenBMB released on Hugging Face; awaiting llamacpp support for testing.
Similar Articles
OpenBMB presents the model BitCPM-CANN 1.58 bit
OpenBMB introduced BitCPM-CANN, a 1.58-bit model being tested on Huawei Ascend 910B hardware.
@AdinaYakup: BitCPM4-CANN Native 1.58-bit LLM training system on Ascend NPUs https://huggingface.co/collections/openbmb/bitcpm4-cann…
OpenBMB releases BitCPM4-CANN, a collection of natively trained 1.58-bit ternary quantized LLMs (0.5B to 8B) optimized for Ascend NPUs via CANN, achieving 6× memory reduction at inference and minimal training overhead.
openbmb released MiniCPM5-2B, not yet available at huggingface
OpenBMB released MiniCPM5-2B, a 2 billion parameter language model, currently not yet available on Hugging Face.
1-bit / 2-bit / Ternary / Bitnet Models - Updates & Tracking
This article tracks updates on various low-bit AI models, including Bonsai, BitCPM, and others, with performance improvements and compatibility updates for backends like llama.cpp.
MiniCPM5 1B - what is it?
MiniCPM5-1B is a new small language model from OpenBMB, apparently built from scratch with its own tokenizer and distinct behavior, generating excitement as a capable 1B model.