README update

README.md

@@ -44,8 +44,8 @@ model-index:
 - Less stable on very long or morphologically complex words
 
 > Development information
-> - 🚧 **Current version:** v2 (stage 3)
-> - ⏳ **Upcoming release:** v3 (stage 4)
+> - 🚧 **Current version:** Baseline (stage 1)
+> - ⏳ **Upcoming release:** v1 (stage 2)
 >
 > **Note:** As of May 19, 2026, AramT5's training process, which was at stage 4, was reset a baseline level due to inconsistencies found in previous versions of the Serto-Madnḥaya mapping code and lack of data for individual words, which mostly invalidated prior learning efforts
 
@@ -149,4 +149,4 @@ uv run python src/train_t5.py --stage 2 --hf-model your-username/model-name
 
 ## 📋 Version Changelog
 
-* **AramT5 Baseline (May 18, 2026):** T5 fine-tuned on 20k records, across 30 epochs, leveraging the stage 1 configuration. Baseline version with a rather poor understanding of how to transliterate properly, yet shown to capture some roots and Syriac morphology in a limited manner
+* **AramT5 Baseline (May 20, 2026):** T5 fine-tuned on 20k records, across 30 epochs, leveraging the stage 1 configuration. Baseline version with a surprisingly good initial understanding of how to transliterate properly, shown to capture some roots and Syriac morphology in a limited manner

src/test_t5.py

@@ -1,7 +1,7 @@
 from transformers import AutoTokenizer, T5ForConditionalGeneration, pipeline
 
 # HF Hub path config
-model_path = "checkpoints/stage4-final"
+model_path = "crossroderick/aramt5"
 
 # Unicode directional formatting for RTL text (Syriac)
 RLI = "\u2067"  # Right-to-Left Isolate