Instructions to use noahmclaughlin/Mixed-Messenger-4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use noahmclaughlin/Mixed-Messenger-4B with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] hf download noahmclaughlin/Mixed-Messenger-4B --local-dir Mixed-Messenger-4B
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Download evaluation/pause-for-performance.json from noahmclaughlin/Mixed-Messenger-4B: direct link, hf CLI and curl.
- Browser
- Download file 469 Bytes
-
https://huggingface.co/noahmclaughlin/Mixed-Messenger-4B/resolve/main/evaluation/pause-for-performance.json
- Command line
-
hf download hf://noahmclaughlin/Mixed-Messenger-4B/evaluation/pause-for-performance.json
-
curl -L -o pause-for-performance.json https://huggingface.co/noahmclaughlin/Mixed-Messenger-4B/resolve/main/evaluation/pause-for-performance.json
469 Bytes
| { | |
| "paused_at": "2026-11-15T06:19:57.455731+00:00", | |
| "reason": "User requires meaningful improvement; initial best development gain was only 0.05445 points. Publication paused and development reopened.", | |
| "archived_baseline_run": "runs/paused-initial-qwen3-4b-base-final-de152c4ceb49002d", | |
| "saved_baseline_episode_files": { | |
| "none": 45, | |
| "agent": 25 | |
| }, | |
| "adapted_model_evaluation_started": false, | |
| "final_scores_used_for_new_training_decisions": false | |
| } | |