stanfordnlp/CoreNLP · error · RuntimeIOException

Couldn't load word vectors

Error message

Couldn't load word vectors

What it means

Thrown in ColumnDataClassifier.loadWordVectors when the word-vector file given via -wordVector cannot be opened or parsed (unreadable path, wrong format, or missing file). It signals that pretrained embedding initialization cannot proceed.

Solutions

1. Verify the word vector file path is correct and the file exists and is readable
2. Use a supported vector format (e.g., GloVe text format with word followed by floats)
3. Check the classpath/working directory if using a relative path
4. Download the intended embedding file if it is missing

When it happens

Trigger: Thrown at src/edu/stanford/nlp/classify/ColumnDataClassifier.java:1561 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of stanfordnlp/CoreNLP@1b7edd19c4 (2026-09-10). Data as JSON: /api/errors/a1ae4193fc9d9721. Report an issue: GitHub.

Appendix: source

Thrown at src/edu/stanford/nlp/classify/ColumnDataClassifier.java:1561

      for (String line; (line = br.readLine()) != null; ) {
        String[] fields = line.split("\\s+");
        if (numDimensions < 0) {
          numDimensions = fields.length - 1;
        } else {
          if (numDimensions != fields.length -1 && ! warned) {
            logger.info("loadWordVectors: Inconsistent vector size: " + numDimensions +
                    " vs. " + (fields.length - 1));
            warned = true;
          }
        }
        float[] vector = new float[fields.length - 1];
        for (int i = 1; i < fields.length; i++) {
          vector[i-1] = Float.parseFloat(fields[i]);
        }
        map.put(fields[0], vector);
      }
    } catch (IOException ioe) {
      throw new RuntimeIOException("Couldn't load word vectors", ioe);
    }
    timing.done("Loading word vectors from " + filename + " ... ");
    return map;
  }


  /**
   * Initialize using values in Properties file.
   *
   * @param props Properties, with the special format of column.flag used in ColumnDataClassifier
   * @return An array of flags for each data column, with additional global flags in element [0]
   */
  private static Pair<Flags[], Classifier<String,String>> setProperties(Properties props) {
    Flags[] myFlags;
    Classifier<String,String> classifier = null;
    boolean myUsesRealValues = false;

    Pattern prefix;

View on GitHub (pinned to 1b7edd19c4)