Update README.md

README.md

@@ -29,7 +29,7 @@ Sources:
 * English: MIMIC CXR
 * English: MIMIC4
 
-All translated (if not with DeepL) with a combination of GeminiFlash 1.5/GPT4o mini, MariaNMT, NLLB200.
+All translated (if not with DeepL) with a combination of GeminiFlash 1.5/2.0/GPT4o mini, MariaNMT, NLLB200.
   
 * Number of tokens: 20B
 * Number of documents: 32M
@@ -42,8 +42,8 @@ All translated (if not with DeepL) with a combination of GeminiFlash 1.5/GPT4o m
 * Learning schedule: linear, with 5_000 warmup steps
 * Num epochs: ~3 (off-premise) followed by 3 (on-premise)
 
-Train perplexity: 2.5
-Validation perplexity: 3.4
+Train perplexity: 2.4
+Validation perplexity: 3.3
 
 # Acknowledgement