--- title: README emoji: ๐ŸงŠ colorFrom: indigo colorTo: blue sdk: static pinned: false --- # Sipsa Labs Sipsa Labs builds SIP, the operating system for the physical world. UltraCompress is retired. Its public materials preserve historical research; new API access and paid pilots are no longer offered. The research results below are historical and apply to their recorded versions and evaluation conditions. --- ## Historical UltraCompress research **Historical research tiers:** - A **genuinely lossless archival tier** โ€” reconstructs the original weights bit-for-bit (~1.3โ€“1.6ร— smaller) for exact storage and hand-off. - A **near-lossless 5-bit tier** โ€” ~1% perplexity vs bf16 (lossy), ~3ร— smaller, for running frontier-scale models on small GPUs. **23 architectures verified**: 22 PPL-verified end-to-end (21 transformer โ€” 17 dense + 4 MoE โ€” plus 1 SSM with a comparator-note caveat) and 1 Vision Transformer cosine-verified (DINOv2-Large). Dense + MoE + SSM + ViT, 0.6B to 405B. The public v0.6.27 CLI checks basic pack structure and computes SHA-256 fingerprints unless hashing is skipped. It does not reconstruct weights, reproduce the evaluations below, or automatically compare its fingerprint with a trusted reference. ### Verified 5-bit PPL ratios (sample) *(perplexity ratio = compressed PPL / bf16 baseline PPL; FineWeb-edu held-out tail; seq_len = 1024; n = 30โ€“50; seed = 42)* | Model | Params | Type | PPL ratio | Notes | |---|---:|:---|---:|:---| | **Mixtral-8x7B** | 47B | MoE | **1.00368ร—** | tightest MoE result | | Qwen3-1.7B-Base | 1.7B | dense | **1.00401ร—** | small-decoder record | | Qwen3-14B | 14B | dense | **1.00403ร—** | 14B-class record | | Yi-1.5-9B | 8.8B | dense | **1.00414ร—** | >8B record | | Qwen3-8B | 8B | dense | **1.00440ร—** | 8B-class record | | Mistral-7B-v0.3 | 7B | dense | **1.00548ร—** | strong dense 7B-class result | | Phi-3-mini-4k | 3.8B | dense | **1.00262ร—** | tightest dense ratio (seq_len=128 caveat) | | **Hermes-3-Llama-3.1-405B** | 405B | dense | **1.0066ร—** | 405B-class near-lossless on a single 32 GB GPU (streamed from disk) | | Mamba-2.8B | 2.8B | SSM | **1.0059ร—** | state-space model โ€” SSM-compatible comparator (the canonical transformer pipeline is architecture-incompatible with SSMs) | | Llama-3.1-8B | 8B | dense | **1.0125ร—** | honest worst case (pipeline-specific floor, not architectural) | Full matrix at [github.com/sipsalabs/ultracompress](https://github.com/sipsalabs/ultracompress). ### Inspect a historical pack ```bash pip install ultracompress hf download SipsaLabs/qwen3-1.7b-base-uc-v3-bpw5 --local-dir ./pack uc verify ./pack ``` Expected result category: basic pack-structure checks and a computed fingerprint, not a reconstruction proof or an automatic comparison with a trusted reference. Without an API key, `uc try` prints a recorded reference response, not a live inference result. --- ## Historical API and services New API access and paid pilots are no longer offered. The public repository and model cards preserve historical materials. --- ## License + IP - **PyPI v0.6+** under BUSL-1.1 with an **Additional Use Grant**: free for sub-$1M-ARR companies, research, and individuals. Auto-converts to Apache 2.0 four years after each release. - **v0.5.x** stays Apache-2.0 on the `legacy/0.5.x` branch. - Codec internals are patent-protected (continuations pending). --- ## Contact - **Project inquiries** โ†’ founder@sipsalabs.com - **Patents / licensing** โ†’ legal@sipsalabs.com - **Press / media** โ†’ press@sipsalabs.com - **Security disclosure** โ†’ security@sipsalabs.com - **General** โ†’ hello@sipsalabs.com [sipsalabs.com](https://sipsalabs.com) ยท [GitHub](https://github.com/sipsalabs) ยท [PyPI](https://pypi.org/project/ultracompress/) ยท [Historical repository](https://github.com/sipsalabs/ultracompress)