MemTensor/MemT-1.0-Nano
main boardmodel
MemT-1.0 is a new generation of memory foundation models that makes persistent, dynamically evolving memory an intrinsic part of model computation. 35B params. License: apache-2.0. Vendor-reported: "Across 11 memory benchmarks, MemT-1.0-Flash and MemT-1.0-Nano rank first and second among the evaluated parametric-memory methods, respectively."
- Vendor-reported: Across 11 memory benchmarks, MemT-1.0-Flash and MemT-1.0-Nano rank first and second among the evaluated parametric-memory methods, respectively.
- Org
- MemTensor
- Released
- 2026-10-15
- Status
- main board
- License
- apache-2.0 (permissive)
- Parameters
- 35.1B
- Architecture
- MemTQwen3_5MoeForCausalLM
- Task
- text-generation · text
- Signal
- 110 HF likes · 18765 downloads
Runs on
mixed: real files + estimate; need = weights + FP16 KV cache for 8K tokens + 1GB. A sizing calculation, not tested.
- Q4 ~25.1 GB (estimated size): 32GB GPU (5090), Mac 36GB, Mac 48GB, CPU only, 32GB RAM (slow)
- Q8 ~41.3 GB (estimated size): 48GB GPU (RTX 6000 Ada / A6000), Mac 64GB (M1 Max), Mac 96GB, CPU only, 64GB RAM (slow)
- FP16 ~74.2 GB: 96GB GPU (RTX Pro 6000 Blackwell), 1× DGX Spark (128GB), 2× DGX Spark (256GB), Strix Halo 128GB unified (~96GB usable), Mac 128GB, Mac 192GB, Mac 256GB, Mac 512GB, CPU only, 128GB RAM (slow), CPU only, 256GB RAM (slow)
Links
First published here 2026-10-15 · updated 2026-10-19 · markdown · JSON