UraionLabs/Uraion-Agent-Steer-GGUF overview
Legacy / unsupported — not a faithful H Res model Status as of 2026 07 27: archived in place for provenance and download continuity. The pre audit card explici…
Runs locally from ~2.81 GB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| Uraion-Agent-Steer-Q2_K.gguf | GGUF | Q2_K | 2.81 GB | Download |
| Uraion-Agent-Steer-Q3_K_M.gguf | GGUF | Q3_K_M | 3.55 GB | Download |
| Uraion-Agent-Steer-Q4_K_M.gguf | GGUF | Q4_K_M | 4.36 GB | Download |
| Uraion-Agent-Steer-Q5_K_M.gguf | GGUF | Q5_K_M | 5.07 GB | Download |
| Uraion-Agent-Steer-Q6_K.gguf | GGUF | Q6_K | 5.82 GB | Download |
| Uraion-Agent-Steer-Q8_0.gguf | GGUF | Q8_0 | 7.54 GB | Download |
Model Details
| Model ID | UraionLabs/Uraion-Agent-Steer-GGUF |
|---|---|
| Author | UraionLabs |
| Pipeline | text-generation |
| License | apache-2.0 |
| Base model | UraionLabs/Uraion-Agent-Steer |
| Last modified | 2026-07-27T08:35:20.000Z |
Model README
---
license: apache-2.0
base_model: UraionLabs/Uraion-Agent-Steer
base_model_relation: quantized
library_name: llama.cpp
language:
- en
pipeline_tag: text-generation
tags:
- legacy
- unsupported
- not-for-production
- gguf
---
Legacy / unsupported — not a faithful H-Res model
Status as of 2026-07-27: archived in place for provenance and download continuity.
The pre-audit card explicitly states that these GGUF conversions omit every nonlinear H-Res tensor
because llama.cpp does not implement that adapter path. The files therefore cannot reproduce the
claimed Uraion-Agent-Steer adaptation. No base-model comparison, tool-use benchmark, raw
predictions, or conversion validation is published.
Do not treat these files as a quantized H-Res release or as evidence of model quality. The full
pre-audit card is preserved in LEGACY_CARD.md.
Current Uraion Labs work is FinStruct, with
versioned SEC-filing benchmarks and evidence-linked outputs. See
Run UraionLabs/Uraion-Agent-Steer-GGUF with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models