Instructions to use nsalerni/loudink-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use nsalerni/loudink-v1 with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir loudink-v1 nsalerni/loudink-v1
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
| { | |
| "title": "loudink-v1 v0.4 \u2014 daily-Mac product max", | |
| "date": "2026-07-08", | |
| "version": "0.4.0-daily", | |
| "bars": { | |
| "writing_p50_ms_le_250": true, | |
| "writing_neural_daily_ge_80": true, | |
| "ir_neural_ge_45": true, | |
| "daily_mac_expanded": true, | |
| "writing_daily_shipped": 1.0, | |
| "writing_daily_neural": 0.8636, | |
| "writing_daily_p50_ms": 181, | |
| "computer_daily_shipped": 1.0, | |
| "multi_step_shipped": 1.0, | |
| "ir_neural_n112": 0.4643, | |
| "overall_daily_mac": 1.0, | |
| "n_daily_mac": 56 | |
| }, | |
| "stack": { | |
| "writer": "artifacts/sft_loudink_v1_rft_daily/adapters", | |
| "ir": "artifacts/sft_loudink_v1_ir_honesty/adapters", | |
| "polish": "structural" | |
| }, | |
| "methods": [ | |
| "lean_chat_prompts", | |
| "adaptive_max_tokens", | |
| "honesty_safe_polish_fast_path", | |
| "daily_rft_expert_iteration", | |
| "micro_gold_sft_residual_fails", | |
| "ir_honesty_transcript_first_gold_sft", | |
| "sequence_plan_to_compact_ir", | |
| "finder_gmail_calendar_shortcuts_search_fp" | |
| ], | |
| "measurements": { | |
| "daily_mac": "artifacts/benchmarks/loudink_v1/daily_mac/loudink_v1_daily_mac_measurement.json", | |
| "ir_honesty": "artifacts/benchmarks/loudink_v1/ir_honesty/loudink_v1_ir_honesty_measurement.json", | |
| "scorecard": "artifacts/loudink_v1/product_scorecard.json" | |
| } | |
| } | |