GDN_pico_ByT5_tokenizer is an open-weight model from Kurogane Himaki, released under Apache License 2.0. It has 10M parameters and a 2,048-token context. At 16-bit it needs about 0 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.
use attahed notebook.
Runs On
What it takes to serve GDN_pico_ByT5_tokenizer (10M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 0.0 GB | 0.0 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 0.0 GB | 0.0 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.0 GB | 0.0 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Oct 18, 2026.
GDN_pico_ByT5_tokenizer on every accelerator the SAVRN Index prices, at every precision
Model Card
By Kurogane Himaki, published under apache-2.0, revision 9e5249cbaa35.
use attahed notebook.
Read Kurogane Himaki's full model card
use attahed notebook.
Configuration
- Architecture
- GatedDeltaNetForCausalLM
- Context length (tokens)
- 2,048
- Layers
- 11
- Hidden size
- 256
- Feed-forward size
- 704
- Head dimension
- 32
- Vocabulary size
- 264
- Model type
- gated_deltanet
Identity and Version
- Repository
- kurogane/GDN_pico_ByT5_tokenizer
- Publisher
- Kurogane Himaki
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- 10M parameters
- Languages
- ja
- Revision
- 9e5249cbaa35fa739e6905833a51e2ea160622c3
- First published
- 2026-09-22
- Last updated
- 2026-09-22
Files and Weights
12 files, 41.7 MB in total. The weights are 1 file totalling 41.7 MB in safetensors.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 41.7 MB | 1e2c5b18295c |
| added_tokens.json | Configuration | 93 B | — |
| byte_chat.py | Configuration | 7.9 KB | — |
| config.json | Configuration | 877 B | — |
| gdn_compat.py | Configuration | 1.8 KB | — |
| gdn_greedy.py | Configuration | 5.3 KB | — |
| generation_config.json | Configuration | 228 B | — |
| README.md | Documentation | 181 B | — |
| chat_template.jinja | Other | 571 B | — |
| generate_gdn.ipynb | Other | 3.2 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
| tokenizer_config.json | Tokenizer | 1.8 KB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 41.7 MB
Released by Kurogane Himaki through its official repository on Hugging Face. Read the license.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 41.7 MB |
| 16-bit | 0.0 GB |
| 8-bit | 0.0 GB |
| 4-bit | 0.0 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About GDN_pico_ByT5_tokenizer
How much GPU memory does GDN_pico_ByT5_tokenizer need?
About 0 GB at 16-bit and 0 GB at 4-bit: the weights (10M parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run GDN_pico_ByT5_tokenizer on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use GDN_pico_ByT5_tokenizer commercially?
Yes. GDN_pico_ByT5_tokenizer is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.
What is GDN_pico_ByT5_tokenizer's context length?
2,048 tokens, from the maximum position embeddings in its published configuration.