peasantsmith/Hy4-Flash-Maya-GGUF
image-text-to-text model by peasantsmith
From the publisher
Model card & documentation
Source previewRead the publisher’s intended use, setup instructions, evaluations and limitations. The original model card is the source of truth.
GGUF quants of tencent/Hy4-flash (220B parameters: 155.5B language backbone with 512 experts per layer, 63.5B n-gram "Engram" tables, a vision encoder and the DFly2 drafter; 1M context) made for Project Maya. The files also load in llama.cpp with the Hy4-flash pull request (see "Running it"). Every size uses the same n-gram table file. It is Tencent's FP8 table, value for value; it is read from the SSD page by page while the model runs and does not need to fit in memory. Table: Hardware requirements (units = total memory: RAM + VRAM, or unified memory). The n-gram table is not counted: it stays on the SSD. slower, because parts of the model are then read from disk. about 110 GB. Tencent's defaults, which Project Maya applies by itself when the model is selected: Measured against the original full-precision model on held-out conversations that were never used for calibration (110,230 answer…
Read the full model card ↗ · Preview checked 2026-11-21T09:53:36.404Z
Inside the original model card — Document outline
- Hy4-flash · Maya GGUF
- Sizes
- :gear: Usage Guide
- Recommended Settings
- Quality
- How they were made
- Files
- Running it
- License
Headings are captured from the source. Links open the publisher’s document, not a locally hosted copy.
Links from the model card
References supplied by the publisher, not independently verified endorsements. Check the destination before downloading files or entering credentials.
Documentation belongs to its respective authors. Reported project/model license: other. A listing is not a grant of reuse or training rights. Confirm the document’s own terms at the source.
Model overview
peasantsmith/Hy4-Flash-Maya-GGUF is a image-text-to-text model repository published by peasantsmith on Hugging Face. The source reports the gguf library.
This page summarizes Hub metadata. For intended use, training data, evaluation results and limitations, consult the original model card.
Model facts
- Task
- image-text-to-text
- Library
- gguf
- Recent downloads (30 days)
- 19,139
- Cumulative likes
- 32
- Hugging Face trending score
- 31
- Reported safetensors parameters
- Not reported
- Architecture
- Not reported
- License
- other
- Access
- Not gated by Hugging Face
- Created
- 2026-11-18T23:53:57.000Z
- Last modified
- 2026-11-19T20:34:17.000Z
Compatibility and lineage
Language coverage was not reported.
Reported base models: tencent/Hy4-flash
Parameter count is not a RAM/VRAM requirement. Check precision, quantization, context length and runtime compatibility in the model card; no hardware or API-cost claim is inferred here.
Use and evaluate
Check license terms and access requirements first. Review the model card’s documented loading instructions, evaluations and safety limitations. Benchmark scores are not imported by this directory, and popularity is not an accuracy ranking.
No model weights are downloaded or executed by AltAPIs. Never enable remote model code without reviewing it.
Source and freshness
Source: Hugging Face Hub. Metadata observed 2026-11-21T07:00:33.109Z. Daily imports are snapshots, not real-time monitoring.
Popularity and source listings do not establish security, suitability, licensing rights or benchmark performance.