-
Upload news summarization eval results for 69d59c104adb81dd2de75563
-
Upload news summarization eval results for 69d5986c4adb81dd2de75558
-
Fix HF Jobs result upload path expansion
-
Install runtime deps inside HF Jobs launcher
-
Document Hugging Face Jobs path for news summarization eval
-
Add Hugging Face Jobs launcher for news summarization eval
-
Narrow Colab snapshot download to news summarization files
-
Lazy-load shared providers for lighter notebook imports
-
Use lighter Colab defaults for news summarization notebook
-
Improve Colab runner progress visibility
-
Use valid default model in news summarization notebook
-
Include shared eval code in Colab snapshot download
-
Use HF snapshot download in news summarization notebook
-
Fix notebook result inspection cell
-
Link Colab notebook in news summarization README
-
Add Colab notebook for news summarization eval
-
Add news summarization eval harness
-
Upload local-intelligence eval results for 69d289624adb81dd2de74caf
-
Upload local-intelligence eval results for 69d27d17074a09ff9be109c6
-
Upload local-intelligence eval results for 69d27a3e074a09ff9be109a9
-
Upload local-intelligence eval results for 69d1ea7b4adb81dd2de74b7a
-
Sync local-intelligence eval harness
-
Sync local-intelligence eval harness
-
Upload local-intelligence eval results for 69d1de21074a09ff9be10463
-
Sync local-intelligence eval harness
-
Upload local-intelligence eval results for 69d1c0104adb81dd2de74ade
-
Sync local-intelligence eval harness
-
Sync local-intelligence eval harness
-
Sync local-intelligence eval harness
-
🧪 add segment classifier, protocol processor test harness, and full chain tests
- Part 4 blog: segment classifier synopsis with processor + chain test results
- Segment classifier: training script, eval harness, 50 labeled mixed dictations
- Protocol processor: updated with needs-llm detector, test harness (92 unit + 100 eval)
- Full chain test: 12 cases through classifier → model → reassembly
- Model weights: segment-classifier-model.json, needs-llm-model.json
- Eval data: 100 dictation-to-command pairs across 31 categories
-
🧪 initial commit — voice-to-syntax training lab
Processor, pipeline, eval data, training infrastructure, and blog drafts
from the voice dictation → programming syntax experiments.
Architecture: LLM normalizer → procedural processor (split approach).
Zero-training pipeline achieves 47% on mixed-difficulty eval (2x baseline).