-
Update evaluation/run_eval_kaggle.py with user enhancements
-
Add official 9-Pillar Base Model Forensic Audit Report
-
Fix vocab_size mismatch (64000 base vs 64003 tokenizer)
-
Fix checkpoint weight subkey extraction and Kaggle runner auto-refresh
-
Fix checkpoint weight subkey extraction and Kaggle runner auto-refresh
-
Fix checkpoint weight subkey extraction and Kaggle runner auto-refresh
-
Feature: Add 9-Pillar Base Model Forensic Diagnostic Suite & Kaggle Runner
