zeromem-pointer is an open-weight model from Kumar. Its published files total 137.5 MB.
About 34M parameters, trained from scratch on a laptop. Reads a question plus a chunk whose sentences are tagged [A] [B] [C]... and answers [C] or, so it never retypes (or invents) text.
Model Card
About 34M parameters, trained from scratch on a laptop. Reads a question plus a chunk whose sentences are tagged [A] [B] [C]... and answers [C] or, so it never retypes (or invents) text. best.pt is a PyTorch checkpoint for the ZeroMem code (not a transformers model); tokenizer.json is its byte-level BPE tokenizer.
Excerpt from the card by Kumar.
Identity and Version
- Repository
- balaji9043/zeromem-pointer
- Publisher
- Kumar
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- Not stated by the source
- Languages
- rag
- Revision
- ff684b7626f036c372132c9861c103d333c25114
- First published
- 2026-09-28
- Last updated
- 2026-09-28
Files and Weights
4 files, 137.5 MB in total. The weights are 1 file totalling 136.4 MB in pt.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| best.pt | Weights | 136.4 MB | 3b00fc9605e8 |
| README.md | Documentation | 424 B | — |
| .gitattributes | Repository | 1.5 KB | — |
| tokenizer.json | Tokenizer | 1.1 MB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 136.4 MB
Released by Kumar through its official repository on Hugging Face.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 136.4 MB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.