README.md for sample BLEU challenge

README.md

@@ -150,8 +150,6 @@ Then let Gonito pull them and evaluate your results.
 
 ## `geval` options
 
-geval - stand-alone evaluation tool for tests in Gonito platform
-
     geval [--init] [--out-directory OUT-DIRECTORY]
           [--expected-directory EXPECTED-DIRECTORY] [--test-name NAME]
           [--out-file OUT] [--expected-file EXPECTED] [--metric METRIC]
@@ -172,3 +170,5 @@ geval - stand-alone evaluation tool for tests in Gonito platform
     --expected-file EXPECTED The name of the file with expected
                              results (default: "expected.tsv")
     --metric METRIC          Metric to be used - RMSE, MSE or BLEU (default: RMSE)
+
+If you need another metric, let me know, or do it yourself!

src/GEval/CreateChallenge.hs

@@ -41,6 +41,28 @@ createFile filePath contents = do
   writeFile filePath contents
 
 readmeMDContents :: Metric -> String -> String
+readmeMDContents BLEU testName = [i|
+GEval — sample machine translation challenge
+============================================
+
+This is a sample challenge (translating from Finnish to Māori) for Gonito framework. Replace it with
+the description of your challenge.
+
+Directory structure
+-------------------
+
+* `README.md` — this file
+* `config.txt` — configuration file
+* `train/` — directory with training data
+* `train/train.tsv` — sample parallel corpus (Finnish text in the first column, Māori text in the second one)
+* `dev-0/` — directory with dev (test) data
+* `dev-0/in.tsv` — Finnish input text for the dev set
+* `dev-0/expected.tsv` — Māori reference translation for the dev set
+* `${testName}` — directory with test data
+* `${testName}/in.tsv` — Finnish input data for the test set
+* `${testName}/expected.tsv` — Māori reference translation for the test set
+|]
+
 readmeMDContents _ testName = [i|
 GEval — sample challenge
 ========================