refactor: rename tokenizer to token_estimator for clarity

- Rename tokenizer.js to token_estimator.js to reflect its purpose
- Update import in translation.js
- Add development journal documenting NVIDIA 4090 performance metrics

journal.rst

@@ -0,0 +1,20 @@
+====================
+Development Journal
+====================
+
+July 2, 2025
+============
+
+Token Estimator Implementation
+-------------------------------
+
+Built `token_estimator.js` - a simplified BPE-based token counter for translation ETA estimation. 
+
+**Not compatible with tiktoken** - this is estimation-only, ~85% accuracy. Good enough for progress bars, not for exact tokenization.
+
+**NVIDIA 4090 Performance Metrics:**
+- Prompt processing: 3,144 tokens/sec
+- Generation: 121 tokens/sec  
+- Translation estimates: 1K tokens ≈ 11 seconds
+
+The estimator uses real production metrics from our Hermes machine to provide realistic translation timing.

src/translation.js

@@ -2,7 +2,7 @@
 import { sendMessage } from "./chat.js";
 import { API_KEY } from "./config.js";
 import { getSelectedModel, getSelectedModelEndpoint } from "./models.js";
-import { countTokens } from "./tokenizer.js";
+import { countTokens } from "./token_estimator.js";
 
 // Languages supported by Hermes 3 model (from Russell's implementation)
 export const SUPPORTED_LANGUAGES = {