Update README.md

README.md

@@ -9,7 +9,7 @@ This dataset contains preprocessed and token-packed `.bin` files intended for us
 
 - Each `.bin` file contains a fixed number of samples, where each sample is exactly 8192 tokens long.
 - Samples are grouped into batches of 125 samples, totaling **1.024 million tokens per batch**.
-- Each file (called a "block") contains 12500 samples (approximately 102.4 million tokens).
+- Each file (called a "block") contains 62500 samples (approximately 512 million tokens).
 - All samples are tokenized using the `GPT2TokenizerFast` from Hugging Face Transformers.
 
 ## Structure
@@ -35,7 +35,7 @@ Tokens were drawn from a diverse mix of high-quality open datasets:
 - `PAQ`
 - `Natural Questions`
 
-Each dataset was assigned a token quota (e.g., 10B tokens) to ensure a balanced mix.
+Each dataset was assigned a token quota to ensure a balanced mix.
 
 ## Preprocessing & Packing Strategy