Update README.md

README.md

@@ -36,15 +36,18 @@ See [mathqa HF dataset](https://huggingface.co/datasets/math_qa) and [official w
 
 Columns:
 
-- `problem` - description of a mathematical problem in natural language
+- `question` - th description of a mathematical problem in natural language
+- `chain` - Solution in the form of step-by-step calculations encoded in simple html-like language. computed from `annotated_formula` column
+- `result` - the result of the problem as string (can be integer, floating number, fraction, ...)
+- `result_float` - the result converted to a float
 - `options` - dictionary with choices 'a' to 'e' as possible solutions
-- `correct` - correct answer, one of 'a', ..., 'e'
+- `options_num` - same as 'options', but with simple parsing to extract the number from string. This is best-effort only - not all values are (or can be) extracted correctly
+- `correct_option` - correct options, one of 'a', ..., 'e', should match with `result`
 - `rationale` - human-annotated free-text reasoning that leads to the correct answer
 - `annotated_formula` - human-annotated nested expression that (approximately) evaluates to the selected correct answer
 - `linear_formula` - same as `annotated_formula`, but linearized. Provided by original math_qa authors
-- `chain` - linearized `annotated_formula`, provided by us. Converted to HTML-like language with expressions that can be evaluated using our sympy-based calculator
 - `index` - index of the example in the original math_qa dataset
-- `options_float` - same as 'options', but with simple parsing and evaluation applied to convert options to float. This is best-effort only - not all values are (or can be) extracted correctly
+