mindi-dev/experience-extractor-1.2b-v1-GGUF

experience-extractor-1.2b-v1 (GGUF)

A small, on-device structured fact extractor for memory engines, fine-tuned from `LiquidAI/LFM2-1.2B-Extract` (LoRA (rank 32) fine-tune (mlx-lm)). It reads a chat transcript and emits every storable fact as JSON in a fixed 8-field schema:

{"facts": [
  {"what": "...", "when": null, "where": null, "why": null,
   "who": ["..."], "fact_type": "world|experience",
   "entities": ["..."], "message_refs": ["id:m07"]}
]}

It powers the `experience` memory engine (EXPERIENCE_EXTRACTOR=lfm25). This repo holds the GGUF quantizations for llama.cpp, Ollama, LM Studio, and the crate's lfm25 path (which adds grammar-constrained decoding).

Evaluation (LongMemEval-cleaned "KU", content-recall)

Run it windowed. Whole-transcript extraction caps a small model near 0.62; sliding a 5-message window and unioning the per-window facts is the recall mechanism and the recommended deploy mode. Pairing the 350M + 1.2B as an ensemble reaches ~0.986 on KU.

Files

Usage

Ollama — ready extractor (this repo ships a Modelfile with the 8-field system prompt baked in):

hf download mindi-dev/experience-extractor-1.2b-v1-GGUF --include "*Q4_0.gguf" Modelfile --local-dir exp-extractor && cd exp-extractor
ollama create experience-extractor-1.2b -f Modelfile
ollama run experience-extractor-1.2b "<paste a rendered transcript>"

Or the raw model (no system prompt): ollama run hf.co/mindi-dev/experience-extractor-1.2b-v1-GGUF:Q4_0

llama.cpp: llama-cli -hf mindi-dev/experience-extractor-1.2b-v1-GGUF:Q4_0

`experience` crate (windowed + constrained 8-field JSON — the validated recall path):

EXPERIENCE_EXTRACTOR=lfm25 EXPERIENCE_EXTRACTOR_MODEL_PATH=./experience-extractor-1.2b-v1-Q4_0.gguf \
EXPERIENCE_EXTRACTION_WINDOW=5 experience serve

Other formats

• —MLX 8-bit (Apple Silicon): `mindi-dev/experience-extractor-1.2b-v1-mlx-8bit`

Training

Full pipeline at mindi-dev/experience (training/). Fine-tuned on real-distribution LongMemEval transcripts (leakage-safe; held-out KU never trained on) with grounded teacher-generated labels.

License

Fine-tune of `LiquidAI/LFM2-1.2B-Extract` under the LFM Open License v1.0. Redistribution permitted with attribution + change notice; commercial use by entities with ≥ US$10M revenue requires a Liquid AI commercial license (Sec. 5). The crate code is MIT and separate. See NOTICE.md and the full LICENSE in this repo.