Support auto upward transformation, downward transformation for column's data type in datastore

[goldenxinxing]

Describe the bug

ai.starwhale.mlops.exception.SwValidationException: invalid request on subject Starwhale Internal DataStore
conflicting type for column weighted avg/support, expected INT64, actual FLOAT64
request=UpdateTableRequest(tableName=project/starwhale/eval/summary, tableSchemaDesc=TableSchemaDesc(keyColumn=id, columnSchemaList=[ColumnSchemaDesc(name=id, type=STRING, pythonType=null, elementType=null, keyType=null, valueType=null, attributes=null), ColumnSchemaDesc(name=weighted avg/precision, type=FLOAT64, pythonType=null, elementType=null, keyType=null, valueType=null, attributes=null), ColumnSchemaDesc(name=weighted avg/recall, type=FLOAT64, pythonType=null, elementType=null, keyType=null, valueType=null, attributes=null), ColumnSchemaDesc(name=weighted avg/f1-score, type=FLOAT64, pythonType=null, elementType=null, keyType=null, valueType=null, attributes=null), ColumnSchemaDesc(name=weighted avg/support, type=FLOAT64, pythonType=null, elementType=null, keyType=null, valueType=null, attributes=null), ColumnSchemaDesc(name=macro avg/precision, type=FLOAT64, pythonType=null, elementType=null, keyType=null, valueType=null, attributes=null), ColumnSchemaDesc(name=macro avg/recall, type=FLOAT64, pythonType=null, elementType=null, keyType=null, valueType=null, attributes=null), ColumnSchemaDesc(name=macro avg/f1-score, type=FLOAT64, pythonType=null, elementType=null, keyType=null, valueType=null, attributes=null), ColumnSchemaDesc(name=macro avg/support, type=FLOAT64, pythonType=null, elementType=null, keyType=null, valueType=null, attributes=null), ColumnSchemaDesc(name=hamming_loss, type=FLOAT64, pythonType=null, elementType=null, keyType=null, valueType=null, attributes=null), ColumnSchemaDesc(name=cohen_kappa_score, type=FLOAT64, pythonType=null, elementType=null, keyType=null, valueType=null, attributes=null), ColumnSchemaDesc(name=kind, type=STRING, pythonType=null, elementType=null, keyType=null, valueType=null, attributes=null)]), records=[RecordDesc(values=[RecordValueDesc(key=id, value=a1791a3ff1fd4099a317884918d8a8f5), RecordValueDesc(key=weighted avg/precision, value=0000000000000000), RecordValueDesc(key=weighted avg/recall, value=0000000000000000), RecordValueDesc(key=weighted avg/f1-score, value=0000000000000000), RecordValueDesc(key=weighted avg/support, value=0000000000000000), RecordValueDesc(key=macro avg/precision, value=fff8000000000000), RecordValueDesc(key=macro avg/recall, value=fff8000000000000), RecordValueDesc(key=macro avg/f1-score, value=fff8000000000000), RecordValueDesc(key=macro avg/support, value=0000000000000000), RecordValueDesc(key=hamming_loss, value=fff8000000000000), RecordValueDesc(key=cohen_kappa_score, value=fff8000000000000), RecordValueDesc(key=kind, value=multi_classification)])])

To Reproduce

Expected behavior

Environment