eaddario/imatrix-calibration
Importance Matrix Calibration Datasets This repository provides calibration datasets used to generate importance matrices (imatrix), which are required to minimize errors when quantizing models with LLaMA C++. The llama-imatrix program cannot handle parquet files directly and thus requires them to be converted into text format first. There are many ways to do this but a simple approach is to use DuckDB with the following command: duckdb -noheader -ascii -c "SELECT content FROM… See the full description on the dataset page: https://huggingface.co/datasets/eaddario/imatrix-calibration.
| Author / organization | eaddario |
|---|---|
| License (as declared) | mit |
| Downloads | 7.4K (as reported by Hugging Face) |
| Likes | 73 |
| Created | 29 January 2025 |
| Last updated at source | 5 May 2026 |
Popularity, last 2 days
7,374 now (+37 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Connected in the knowledge graph
- eaddario/Hammer2.1-7b-GGUFModel · used to train
- eaddario/OLMo-2-1124-7B-Instruct-GGUFModel · used to train
- eaddario/gemma-3-12b-it-GGUFModel · used to train
- eaddario/Mistral-Small-3.2-24B-Instruct-2506-pruned-GGUFModel · used to train
- eaddario/Spark-X2.5-4B-GGUFModel · used to train
- theLittleStone/Qwen3.6-27B-AEON-Ultimate-Uncensored-MTP-i1-GGUFModel · used to train
- eaddario/Dolphin3.0-R1-Mistral-24B-GGUFModel · used to train
- KryptykAngel/Qwen3.6-27B-AEON-Ultimate-Uncensored-MTP-i1-GGUFModel · used to train
- ENOSYS/Qwen3.6-35B-A3B-heretic-750-v1-GGUFModel · used to train
- eaddario/Watt-Tool-8B-GGUFModel · used to train
- ENOSYS/GigaChat3.1-10B-A1.8B-750-v1-GGUFModel · used to train
- eaddario/DeepSeek-R1-Distill-Qwen-7B-GGUFModel · used to train
- gooseyai/MiniCPM5-2B-GGUFModel · used to train
- eaddario/granite-4.1-8b-GGUFModel · used to train
- eaddario/Qwen3.6-35B-A3B-GGUFModel · used to train
- eaddario/gemma-4-26B-A4B-it-GGUFModel · used to train
- agentlans/Apertus-v1.1-4B-Instruct-GGUFModel · used to train
- eaddario/GLM-4.6V-Flash-GGUFModel · used to train
- eaddario/MiniCPM5-2B-GGUFModel · used to train
- eaddario/Devstral-Small-2505-GGUFModel · used to train
- Koshkasa/MuXodious_GLM-4.7-Flash-absolute-heresy-mixed-trellis-GGUFModel · used to train
- eaddario/Olmo-3-7B-Instruct-GGUFModel · used to train
- ENOSYS/Kimi-Linear-48B-A3B-Instruct-abliterated-1000-v1-GGUFModel · used to train
- ENOSYS/Qwen3-Embedding-0.6B-250-v1-GGUFModel · used to train
- eaddario/Dolphin3.0-Mistral-24B-GGUFModel · used to train
- eaddario/gemma-3-12b-it-pruned-GGUFModel · used to train
- eaddario/MiniCPM5-1B-GGUFModel · used to train
- Alexs28/Qwen3.5-0.8B-heretic-250-v1-GGUFModel · used to train
- eaddario/Llama-xLAM-2-8b-fc-r-GGUFModel · used to train
- eaddario/Olmo-3-7B-Think-GGUFModel · used to train
- eaddario/Dolphin-Mistral-24B-Venice-Edition-GGUFModel · used to train
- ENOSYS/Octen-Embedding-0.6B-750-v1-GGUFModel · used to train
- ENOSYS/Qwen3.5-0.8B-heretic-250-v1-GGUFModel · used to train
- eaddario/Spark-X2.5-1.7B-GGUFModel · used to train
- eaddario/Qwen3.8-27B-GGUFModel · used to train
- eaddario/gemma-4-E2B-it-GGUFModel · used to train
- ENOSYS/Octen-Embedding-4B-750-v1-GGUFModel · used to train
- eaddario/Ministral-3-14B-Reasoning-2512-GGUFModel · used to train
- eaddario/granite-4.1-30b-GGUFModel · used to train
- eaddario/Qwen3.6-27B-GGUFModel · used to train
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.