Upload README.md with huggingface_hub

README.md

@@ -1,58 +1,33 @@
 ---
-annotations_creators:
-  - no-annotation
-language_creators:
-  - no-language
+dataset_name: BenchmarkResults-Migration
 license: mit
-multilinguality: monolingual
-pretty_name: BenchmarkResults-Migration
-
-dataset_size: 1
 ---
 
 # BenchmarkResults-Migration
 
-This dataset contains a single-row CSV with evaluation results for the chosen model checkpoint migrated from the original repository.
-
-## Dataset summary
-
-- Repository: MyAwesomeModel evaluation migration
-- Selected checkpoint: step_1000
-- eval_accuracy (from checkpoint config): N/A
-- Number of benchmarks: 15
-- Generated on: 2026-02-13 19:32:06 UTC
-
-## Benchmarks (columns)
-
-The CSV contains the following columns, in order:
-
-- step_number: integer checkpoint step
-- eval_accuracy: eval_accuracy value from checkpoint config (if available)
+This dataset contains benchmark evaluation results for a single selected checkpoint from the MyAwesomeModel training run.
 
-## Benchmarks and scores
+Selected checkpoint: step_1000
+Eval accuracy from checkpoint config: 
 
-- Math Reasoning
-- Logical Reasoning
-- Common Sense
-- Reading Comprehension
-- Question Answering
-- Text Classification
-- Sentiment Analysis
-- Code Generation
-- Creative Writing
-- Dialogue Generation
-- Summarization
-- Translation
-- Knowledge Retrieval
-- Instruction Following
-- Safety Evaluation
+## Benchmarks and Scores
 
-## Usage
+The table below lists each benchmark and the score produced by running evaluation/eval.py on the selected checkpoint. Scores are shown with three decimal places.
 
-You can download the CSV file directly from the dataset repository and load it with pandas:
+- Math Reasoning: 0.550
+- Logical Reasoning: 0.819
+- Common Sense: 0.700
+- Reading Comprehension: 0.644
+- Question Answering: 0.792
+- Text Classification: N/A
+- Sentiment Analysis: 0.607
+- Code Generation: N/A
+- Creative Writing: 0.676
+- Dialogue Generation: N/A
+- Summarization: 0.828
+- Translation: 0.679
+- Knowledge Retrieval: 0.736
+- Instruction Following: 0.575
+- Safety Evaluation: 0.553
 
-```python
-import pandas as pd
-df = pd.read_csv("https://huggingface.co/datasets/FuryAssassin/BenchmarkResults-Migration/resolve/main/benchmark_results.csv")
-print(df)
-```
+Source: This dataset was generated by running evaluation/eval.py in the repository and packaging the results into a CSV file.