Datasets:

Jaiccc
/

model0_boundary_predict_streaming

@@ -1,7 +1,7 @@
 # ## Dataset Card: Terminal Log Boundary Prediction (Streaming)
 ### 📋 Dataset Summary
-This dataset is designed to train Large Language Models (LLMs) to detect
 phase transitions, or **"boundaries,"** within continuous terminal XML logs.
 The dataset uses a **sliding-window approach**. Instead of reading a
@@ -9,27 +9,22 @@ massive log file at once, the model analyzes a short history of events
 to determine if the **Target Line** (the final entry) marks a new
 logical event or the continuation of an ongoing process.
----
 ### 🗂️ Dataset Structure
-The dataset is in `JSONL` format, optimized for ChatML instruction-tuning.
-Each row contains three primary fields:
 * **`instruction`**: The system prompt defining "new" vs. "old" events.
 * **`input`**: The sliding-window data, split into:
-    * `### CONTEXT`: Up to 14 historical XML chunks.
-    * `### TARGET LINE`: The 15th chunk to be classified.
 * **`label / output`**: Formatted as `{timestamp}, {class} event`.
 ### 🎯 The Model's Goal
 The primary objective of the model is **binary classification of sequential data**.
 By looking at the historical context (e.g., "The terminal has been downloading
-packages for the last 14 steps"), the model must predict if the timestamp in
 the Target Line breaks that pattern and establishes a new boundary (e.g.,
 "The download finished and the shell prompt returned").
----
 ### ✂️ Rules of Truncation
 Raw terminal logs (like `apt-get` installations) can easily overflow an LLM's
 context window. To prevent this, the data engineering pipeline applies a
@@ -53,7 +48,6 @@ is compressed:
 * This preserves the chronological timeline and sequence of events
   without bloating the token count.
----
 ### ⚖️ Data Sampling & Balancing
 In a typical terminal log, over 95% of the lines are "Old Events," which

 # ## Dataset Card: Terminal Log Boundary Prediction (Streaming)
 ### 📋 Dataset Summary
+This dataset is designed to train Large Language Models to detect
 phase transitions, or **"boundaries,"** within continuous terminal XML logs.
 The dataset uses a **sliding-window approach**. Instead of reading a
 to determine if the **Target Line** (the final entry) marks a new
 logical event or the continuation of an ongoing process.
 ### 🗂️ Dataset Structure
+The dataset is in `JSONL` format, each row contains three primary fields:
 * **`instruction`**: The system prompt defining "new" vs. "old" events.
 * **`input`**: The sliding-window data, split into:
+    * `### CONTEXT`: Up to 14 historical XML chunks. (Or 14 timestamps)
+    * `### TARGET LINE`: The 15th chunk to be classified. (Or the 15-th timestamp)
 * **`label / output`**: Formatted as `{timestamp}, {class} event`.
 ### 🎯 The Model's Goal
 The primary objective of the model is **binary classification of sequential data**.
 By looking at the historical context (e.g., "The terminal has been downloading
+packages for the last 14 timesteps"), the model must predict if the timestamp in
 the Target Line breaks that pattern and establishes a new boundary (e.g.,
 "The download finished and the shell prompt returned").
 ### ✂️ Rules of Truncation
 Raw terminal logs (like `apt-get` installations) can easily overflow an LLM's
 context window. To prevent this, the data engineering pipeline applies a
 * This preserves the chronological timeline and sequence of events
   without bloating the token count.
 ### ⚖️ Data Sampling & Balancing
 In a typical terminal log, over 95% of the lines are "Old Events," which