modifying parquets

README.md

@@ -35,77 +35,160 @@ task_ids:
 - semantic-similarity-scoring
 configs:
 - config_name: engu
+  default: true
   data_files:
   - split: train
-    path: train.engu.df.short.tsv
+    path: parquet/engu/train.parquet
   - split: validation
-    path: dev.engu.df.short.tsv
+    path: parquet/engu/dev.parquet
   - split: test
-    path: test.engu.df.short.tsv
+    path: parquet/engu/test.parquet
+  features:
+    index: string
+    original: string
+    translation: string
+    scores:
+      - int32
+    mean: float32
+    z_scores:
+      - float32
+    z_mean: float32
 - config_name: enhi
   data_files:
   - split: train
-    path: train.enhi.df.short.tsv
+    path: parquet/enhi/train.parquet
   - split: validation
-    path: dev.enhi.df.short.tsv
+    path: parquet/enhi/dev.parquet
   - split: test
-    path: test.enhi.df.short.tsv
+    path: parquet/enhi/test.parquet
+  features:
+    index: string
+    original: string
+    translation: string
+    scores:
+      - int32
+    mean: float32
+    z_scores:
+      - float32
+    z_mean: float32
 - config_name: enmr
   data_files:
   - split: train
-    path: train.enmr.df.short.tsv
+    path: parquet/enmr/train.parquet
   - split: validation
-    path: dev.enmr.df.short.tsv
+    path: parquet/enmr/dev.parquet
   - split: test
-    path: test.enmr.df.short.tsv
+    path: parquet/enmr/test.parquet
+  features:
+    index: string
+    original: string
+    translation: string
+    pe: string
+    scores:
+      - int32
+    mean: float32
+    z_scores:
+      - float32
+    z_mean: float32
 - config_name: enta
   data_files:
   - split: train
-    path: train.enta.df.short.tsv
+    path: parquet/enta/train.parquet
   - split: validation
-    path: dev.enta.df.short.tsv
+    path: parquet/enta/dev.parquet
   - split: test
-    path: test.enta.df.short.tsv
+    path: parquet/enta/test.parquet
 - config_name: ente
   data_files:
   - split: train
-    path: train.ente.df.short.tsv
+    path: parquet/ente/train.parquet
   - split: validation
-    path: dev.ente.df.short.tsv
+    path: parquet/ente/dev.parquet
   - split: test
-    path: test.ente.df.short.tsv
+    path: parquet/ente/test.parquet
+  features:
+    index: string
+    original: string
+    translation: string
+    scores:
+      - int32
+    mean: float32
+    z_scores:
+      - float32
+    z_mean: float32
 - config_name: eten
   data_files:
   - split: train
-    path: train.eten.df.short.tsv
+    path: parquet/eten/train.parquet
   - split: validation
-    path: dev.eten.df.short.tsv
+    path: parquet/eten/dev.parquet
   - split: test
-    path: test.eten.df.short.tsv
+    path: parquet/eten/test.parquet
+  features:
+    index: string
+    original: string
+    translation: string
+    scores:
+      - int32
+    mean: float32
+    z_scores:
+      - float32
+    z_mean: float32
+    model_scores: float32
 - config_name: neen
   data_files:
   - split: train
-    path: train.neen.df.short.tsv
+    path: parquet/neen/train.parquet
   - split: validation
-    path: dev.neen.df.short.tsv
+    path: parquet/neen/dev.parquet
   - split: test
-    path: test.neen.df.short.tsv
+    path: parquet/neen/test.parquet
+  features:
+    index: string
+    original: string
+    translation: string
+    scores:
+      - int32
+    mean: float32
+    z_scores:
+      - float32
+    z_mean: float32
+    model_scores: float32
 - config_name: sien
   data_files:
   - split: train
-    path: train.sien.df.short.tsv
+    path: parquet/sien/train.parquet
   - split: validation
-    path: dev.sien.df.short.tsv
+    path: parquet/sien/dev.parquet
   - split: test
-    path: test.sien.df.short.tsv
+    path: parquet/sien/test.parquet
+  features:
+    index: string
+    original: string
+    translation: string
+    scores:
+      - int32
+    mean: float32
+    z_scores:
+      - float32
+    z_mean: float32
+    model_scores: float32
 - config_name: multilingual
   data_files:
   - split: train
-    path: train.*.df.short.tsv
+    path: parquet/multilingual/train.parquet
   - split: validation
-    path: dev.*.df.short.tsv
+    path: parquet/multilingual/dev.parquet
   - split: test
-    path: test.*.df.short.tsv
+    path:
+    - parquet/multilingual/test_engu.parquet
+    - parquet/multilingual/test_enhi.parquet
+    - parquet/multilingual/test_enmr.parquet
+    - parquet/multilingual/test_enta.parquet
+    - parquet/multilingual/test_ente.parquet
+    - parquet/multilingual/test_eten.parquet
+    - parquet/multilingual/test_neen.parquet
+    - parquet/multilingual/test_sien.parquet
 ---
 
 # Low-resource QE-DA Dataset

parquet/enmr/test.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7968aa9fdc37d8aaac4d594d7219982b27706f2c30f0f21940efb65692a6d83b
-size 303418
+oid sha256:1e2d9c9a230b7de020db99aa755f9a2eb0ff8031965293956ee2c8c201e54aa3
+size 195493