PollardWeights/K2-Horizon-7B-Pollard

K2-Horizon-7B — Pollard

### Pollard shrank this model: 18.00 GB (f16) → 4.14 GB — 77% smaller, 4.4× down. The smallest rung here; larger, higher-fidelity rungs are listed below. | format | this model's size | |---|---:| | f16 | 18.00 GB | | Q80 | ~9.54 GB | | Q6K | ~7.38 GB | | Q4KM | ~5.22 GB | | PollardMix (this repo's IQ2_S) | 4.14 GB |

Pollard builds of IFM/K2-Horizon-7B made with Pollard Weights — a ladder of measured-allocation quants (bits placed by per-layer sensitivity, not a uniform crush).

These files need [the MBZUAI-IFM llama.cpp fork](https://github.com/MBZUAI-IFM/llama.cpp). This model's architecture (k2-horizon) is not one upstream llama.cpp knows, so stock llama.cpp -- and therefore Ollama and LM Studio -- cannot load them whatever the quant types are. The quants themselves are ordinary K-quants.

Model details

Which file should I choose?

Every rung is the same weights, sized to a different RAM budget by the measured allocation. Pick the largest one that fits your machine with room for context:

• —~9 GB RAM / VRAM → `Q6_K` (7.39 GB). (needs the MBZUAI-IFM llama.cpp fork) near-lossless
• —~7 GB RAM / VRAM → `IQ3_S` (5.03 GB). (needs the MBZUAI-IFM llama.cpp fork) recommended default
• —~6 GB RAM / VRAM → `IQ2_S` (4.14 GB). (needs the MBZUAI-IFM llama.cpp fork) smallest - 4.4x down from f16

Available files (wikitext-2 test, ctx 512)

f16 reference PPL 9.7673...

Measured notes

f16 reference PPL 9.7673.

Measured notes

f16 reference PPL 9.7673..

Download a specific file

pip install -U "huggingface_hub[cli]"
hf download PollardWeights/K2-Horizon-7B-Pollard \
  --include "K2-Horizon-7B-Pollard-IQ3_S.gguf" --local-dir ./

How to run

This model's architecture (k2-horizon) needs the MBZUAI-IFM llama.cpp fork, so every file here runs there:

llama-cli    -m K2-Horizon-7B-Pollard-IQ3_S.gguf -ngl 99 -p "Explain why the sky is blue."
llama-server -m K2-Horizon-7B-Pollard-IQ3_S.gguf -ngl 99

No rung in this repo loads in stock llama.cpp, so Ollama and LM Studio cannot run these files.

ARM / AVX

llama.cpp repacks weights into an interleaved layout at load time for faster inference on ARM and AVX machines — no special file needed, online repacking covers these quants. The old Q4_0_4_4/4_8/8_8 variants are not required.

Errata

• —general.architecture is k2-horizon, which upstream llama.cpp does not implement, so these files load only in the MBZUAI-IFM llama.cpp fork — the quant types are ordinary and irrelevant to that. Checked with pollard-ggufcheck, which reads the architecture and the tensor types out of the header.
• —Measured allocation places bits by per-layer sensitivity under a size budget.
• —Single machine; replication invited.

Credits & license

• —Base model: `IFM/K2-Horizon-7B`
• —Quantization tooling: llama.cpp (ggml-org)
• —Method + tooling: Pollard Weights — measure first, no claim before a number.
• —License: apache-2.0, inherited from the base model.

Built with [Pollard Weights](https://github.com/WestWaters/pollard-weights) — frontier models, small hardware, no compromise.