Open Weights

Model Downloads

Every released model, with its exact parameters and download link. Run them on your own hardware — CPU-friendly, no GPU required.

M-Series ● Live on site

M1-128M

From-scratch hybrid — attention + liquid core in every layer. 50K-step run on WikiText-103. English text continuation.

Best for English text continuation & inference-cost demos.

128.6M params12 layers · 832d512-token window
Download weights ↓ Try it live → 514MB serve checkpoint · gpt2 tokenizer
M-Series ● Live on site

M1 · TinyLlama-1.1B + MT v2s

Frozen TinyLlama-1.1B-Chat base with trained liquid-core residual adapters + LoRA. The conversational model — English and Chinese.

Best for Bilingual conversation with cross-turn memory.

1.1B base + 2.3M adapterchat-capablebilingual EN/ZH
Download adapter ↓ Try it live → 24.8MB checkpoint · replicate the live demo
O-Series ↓ Download

O1-48M

Attention-free O-series edge model. Constant 0.381MB carried state regardless of context — the O(1) memory line. Runs on CPU.

Best for Always-on edge streaming, CPU-only inference.

48M paramsattention-freeCPU-friendly
Download weights ↓ 194MB checkpoint · gpt2 tokenizer
M-Series ◐ Training

M1-2B

1.9B-parameter hybrid (2912d × 35L) in training. Learning-rate scheduling bug fixed — validation PPL falling steadily (107 → 74.75, -30%). Weights when the run converges.

Best for Stronger capability once training converges.

1.9B params35 layers · 2912dconverging · PPL -30%
Weights when it converges Training on rented A100

Family comparison

One glance: what each model is for, how big it is, and where to get it.

Model Best for Params Window Carried state Status
M1-128M English text continuation 128.6M 512 tokens 0.381 MB (O(1)) ● Live on site
M1 · TinyLlama adapter Bilingual conversation, cross-turn memory 1.1B + 2.3M — — ● Live on site
O1-48M Always-on edge streaming, CPU-only 48M — 0.381 MB (O(1)) ↓ Download
M1-2B Stronger capability — training in progress 1.9B — — ◐ Training

Research family (O1-Anti, C1, O1-Sound, AwareLiquid-Physic): clone & run from github.com/AwareLiquid.

Quick start

git clone https://github.com/AwareLiquid/M1 && cd M1
pip install -r requirements.txt
# drop the downloaded .pt into checkpoints/, then:
CKPT_PATH=checkpoints/hybrid_125m_serve.pt TOKENIZER=gpt2 \
python -m uvicorn serve.server:app --port 8000

All models run CPU-only. The 128M hybrid needs ~1GB RAM; the TinyLlama adapter needs ~5GB. Full instructions in the repo README.