Open-weight model
DeepSeek-V4.1-Flash-TensorFold-Engram
by Andy Stewart manateelazycat/DeepSeek-V4.1-Flash-TensorFold-Engram
DeepSeek-V4.1-Flash-TensorFold-Engram is an open-weight model from Andy Stewart, released under MIT License. Its published files total 203.1 GB.
The current publication is in source/: original Engram safetensors copied server-side from deepseek-ai/DeepSeek-V4.1-Flash, with the upstream license, provenance and a pinned engram-manifest.json TP2 preparation recipe.
Model Card
By Andy Stewart, published under mit, revision 99d8cfbca033.
The current publication is in source/: original Engram safetensors copied server-side from deepseek-ai/DeepSeek-V4.1-Flash, with the upstream license, provenance and a pinned engram-manifest.json TP2 preparation recipe. The DeepSeek41 runtime prepares only its local rank, copying the original FP8 weight and scale bytes without numerical conversion. Full-file SHA256 checks must match all four TP2 packs from the accepted 192.168.1.160 / 192.168.1.134 deployment. These raw safetensors require the source-aware runtime; they are not directly consumable by the Engram pack reader. Older split-part files, if present at the repository root, are legacy artifacts. Use the exact files and revisions…
Read Andy Stewart's full model card
The current publication is in source/: original Engram safetensors copied
server-side from deepseek-ai/DeepSeek-V4.1-Flash, with the upstream license,
provenance and a pinned engram-manifest.json TP2 preparation recipe.
The DeepSeek41 runtime prepares only its local rank, copying the original FP8 weight and scale bytes without numerical conversion. Full-file SHA256 checks must match all four TP2 packs from the accepted 192.168.1.160 / 192.168.1.134 deployment. These raw safetensors require the source-aware runtime; they are not directly consumable by the Engram pack reader.
Older split-part files, if present at the repository root, are legacy artifacts. Use the exact files and revisions pinned by the selected LPK release. Weight integrity verification does not itself certify runtime or LPK performance. This repository is independent of the DeepSeek vision deployment repositories.
Identity and Version
- Repository
- manateelazycat/DeepSeek-V4.1-Flash-TensorFold-Engram
- Publisher
- Andy Stewart
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- Not stated by the source
- Languages
- Not stated by the source
- Revision
- 99d8cfbca033cffa6318724f83a1614acdb1bc08
- First published
- 2026-10-04
- Last updated
- 2026-10-05
Files and Weights
9 files, 203.1 GB in total. The weights are 2 files totalling 203.1 GB in safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| source/model-00047-of-00048.safetensors | Weights | 101.5 GB | 824db4881320 |
| source/model-00048-of-00048.safetensors | Weights | 101.5 GB | 976330f49543 |
| source/PROVENANCE.json | Configuration | 1.2 KB | — |
| source/config.json | Configuration | 3.3 KB | — |
| source/engram-manifest.json | Configuration | 1.9 KB | — |
| source/model.safetensors.index.json | Configuration | 7.5 MB | — |
| README.md | Documentation | 961 B | — |
| source/LICENSE | Documentation | 1.1 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
License and Download
- License
- mit
- Access
- Open weights, no gate
- Download size
- 203.1 GB
Released by Andy Stewart through its official repository on Hugging Face. Read the license.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 203.1 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About DeepSeek-V4.1-Flash-TensorFold-Engram
Can I use DeepSeek-V4.1-Flash-TensorFold-Engram commercially?
Yes. DeepSeek-V4.1-Flash-TensorFold-Engram is released under MIT License. The MIT License is a short permissive license. It permits commercial use, modification and redistribution, provided the copyright notice and permission notice are included.