A tiny pirate-themed GPT trained from scratch on a piratized version of TinyStories, then SFT-tuned. Built as a learning project — closer to nanoGPT than to a production LM. - model.safetensors — model weights. - config.json — architecture config (load into training.config.Config). - piratebpe.json — tokenizer (load with tokenizers.Tokenizer.fromfile). - trainingmetadata.json — full training config + metrics snapshot. - banner.png — the banner above. This model is not a transformers model — it uses the custom GPT class from this repo. - Trained on a small synthetic corpus (TinyStories, piratized). Vocabulary, grammar, and world knowledge are extremely narrow. - Short context window (256…
hypa-tiny-keys is an open-weight model for text generation from Hypa-Intelligence. It has 15M parameters and a 512-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. It draws 4.4k downloads a month.
This is the model card of a transformers model that has been pushed on the Hub. Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use the code below to get started with the model.
Parameters15M
Context512
Weights58.2 MB
License—
AccessOpen weights
Monthly Downloads4.4k
Runs On
What it takes to serve hypa-tiny-keys (15M 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 7, 2026.
hypa-tiny-keys on every accelerator the SAVRN Index prices, at every precision
Model Card
This is the model card of a transformers model that has been pushed on the Hub. Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use the code below to get started with the model. Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).
Excerpt from the card by Hypa-Intelligence.
Configuration
- Architecture
- LlamaForCausalLM
- Context length (tokens)
- 512
- Layers
- 8
- Hidden size
- 320
- Feed-forward size
- 960
- Attention heads
- 8
- Key/value heads
- 2
- Head dimension
- 40
- Vocabulary size
- 16,000
- Model type
- llama
Identity and Version
- Repository
- hypaai/hypa-tiny-keys
- Publisher
- Hypa-Intelligence
- Task
- Text generation
- Modality
- Text
- Library
- transformers
- Parameters
- 15M parameters
- Languages
- Not stated by the source
- Revision
- 7f3f290f2b4f5972a284face2399f9850db9a95a
- First published
- 2026-09-24
- Last updated
- 2026-09-27
Files and Weights
11 files, 59.3 MB in total. The weights are 2 files totalling 58.2 MB in bin, safetensors.
Weights2 files · 58.2 MB
Configuration2 files · 943 B
Tokenizer2 files · 1.1 MB
Documentation1 file · 5.2 KB
Other3 files · 21.8 KB
Repository1 file · 1.5 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 58.2 MB | c3f21bb34c30 |
| training_args.bin | Weights | 5.3 KB | 4acee6ce7241 |
| config.json | Configuration | 717 B | — |
| generation_config.json | Configuration | 226 B | — |
| README.md | Documentation | 5.2 KB | — |
| chat_template.jinja | Other | 859 B | — |
| runs/Sep24_23-16-43_9b921b0a70e7/events.out.tfevents.1790291803.9b921b0a70e7.3485.0 | Other | 4.3 KB | 8c9f04ded14d |
| runs/Sep24_23-32-21_9b921b0a70e7/events.out.tfevents.1790292741.9b921b0a70e7.4048.0 | Other | 16.6 KB | 8c340d5d91e6 |
| .gitattributes | Repository | 1.5 KB | — |
| tokenizer.json | Tokenizer | 1.1 MB | — |
| tokenizer_config.json | Tokenizer | 811 B | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 58.2 MB
Download from Hypa-Intelligence
Released by Hypa-Intelligence through its official repository on Hugging Face.
Built From
- Described by arXiv:1910.09700
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 58.2 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 hypa-tiny-keys
How much GPU memory does hypa-tiny-keys need?
About 0 GB at 16-bit and 0 GB at 4-bit: the weights (15M parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run hypa-tiny-keys 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.
What is hypa-tiny-keys's context length?
512 tokens, from the maximum position embeddings in its published configuration.
Similar Models
This is the model card of a transformers model that has been pushed on the Hub. Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use the code below to get started with the model. Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).
A 10,284,480-parameter LLaMA-style text model, trained from scratch. This is a verified first checkpoint for the LDT-10M request (model-requests #12, DedeProGames) — real weights, real training, but undertrained (see the honest status below). It is not a quality release yet; the card states that plainly. Standard LLaMA block, no sliding window, no GQA: The parameter count is the learnable total: the raw safetensors sum is 14,216,640, which double-counts the tied embedding (tok.weight 12288×320 = 3,932,160) that head.weight aliases. Tied, the true count is 10,284,480. - Trained from scratch (no base model). The model learned real context — val loss 4.602 is well below the 7.38 unigram floor…
Hush-Nano-Chat is an English, single-turn instruction-tuned version of Soulitude/Hush-Nano. It starts from the 22M-parameter pretrained model and uses supervised fine-tuning (SFT) on instruction–response pairs. Due to the model's limited parameters, its response can be inaccurate, incomplete, or inconsistent. Hush-Nano-Chat has the following features: The base model was pretrained on 8.5B tokens (8,554,042,292) drawn from the following subsets: Then it was fine-tuned on a mixture of the following datasets: unsloth/alpaca-cleaned, databricks/databricks-dolly-15k, and HuggingFaceH4/norobots. Only the assistant response and ending EOS token contribute to the training loss. Zero-shot normalized…
Aloha! Today, we are releasing Ornith-1.0, a self-improving family of open-source models for agentic coding. This model card documents Ornith-1.0-9B, the most lightweight member of the Ornith family, designed for efficient single-GPU deployment. Ornith-1.0-9B is a dense ~9B model (≈19 GB in bf16), so it serves comfortably on a single 80GB GPU. The recipes below stand up an OpenAI-compatible server; add --tensor-parallel-size / --tp if you want to shard across more GPUs. For a quick local test (or to script offline generation), load the model directly with Transformers. Make sure you have a recent release installed — see the Transformers installation guide; Ornith-1.0-9B requires…
This package uses the standard Transformers Llama causal-language-model architecture with OpenWALDO's schema-1 byte tokenizer. Load the tokenizer with trustremotecode=True. BOM.json inventories every release file and EU-BOM.json contains the EU GPAI training-content disclosure mapping.