Model Intelligence Sheet
cdanis/Ornith-1.5-397B-GGUF-imatrix overview
Ornith 1.5 397B MTP GGUF vision: yes MTP: yes best quality / byte: yes. pretty sure. bartowski imatrix data Unsloth style mixed quants PPL vs size quality vs s…
Runs locally from ~879.0 MB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
12 GGUF files detected
Direct downloads for local inference
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| Ornith-1.5-397B-IQ1_M-UD.gguf | GGUF | IQ1_M | 94.01 GB | Download |
| Ornith-1.5-397B-IQ1_S-UD.gguf | GGUF | IQ1_S | 85.90 GB | Download |
| Ornith-1.5-397B-IQ2_M-UD.gguf | GGUF | IQ2_M | 130.57 GB | Download |
| Ornith-1.5-397B-IQ2_S-UD.gguf | GGUF | IQ2_S | 119.76 GB | Download |
| Ornith-1.5-397B-IQ2_XXS-UD.gguf | GGUF | IQ2_XXS | 107.52 GB | Download |
| Ornith-1.5-397B-IQ3_S-UD.gguf | GGUF | IQ3_S | 168.54 GB | Download |
| Ornith-1.5-397B-IQ3_XXS-UD.gguf | GGUF | IQ3_XXS | 151.54 GB | Download |
| Ornith-1.5-397B-IQ4_XS-UD.gguf | GGUF | IQ4_XS | 209.33 GB | Download |
| Ornith-1.5-397B-Q3_K_XL-UD.gguf | GGUF | Q3_K_XL | 174.30 GB | Download |
| Ornith-1.5-397B-Q4_K_XL-UD.gguf | GGUF | Q4_K_XL | 233.80 GB | Download |
| imatrix.gguf | GGUF | GGUF | 1.06 GB | Download |
| mmproj-Ornith-1.5-397B-BF16.gguf | GGUF | BF16 | 879.0 MB | Download |
Model Details
Model README
---
base_model:
- ornith-ai/Ornith-1.5-397B
base_model_relation: quantized
library_name: llama.cpp
tags:
- gguf
- imatrix
- quantization
- mtp
---
Ornith-1.5-397B MTP GGUF
- vision: yes
- MTP: yes!
- best quality / byte: yes. pretty sure.
- bartowski imatrix data
- Unsloth-style mixed quants
Sources
- Base model: ornith-ai/Ornith-1.5-397B-GGUF —
Ornith-1.5-397B-Q8_0.gguf - Calibration corpus: bartowski/Ornith-1.5-397B-GGUF —
Ornith-1.5-397B-calibration-v6.txt(mirrored here) - Tensor-protection recipe lineage: unsloth/Ornith-1.0-397B-GGUF UD tables (same
qwen35moearch), read from shard headers
Run cdanis/Ornith-1.5-397B-GGUF-imatrix with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models