chore: clean-up a conflict from the rebase

assets/compile_stats.py

@@ -2,133 +2,6 @@
 import json
 import os
 
-text_types = {
-    "ocr": [
-        "AmericanStories",
-        "Eurovoc",
-        "GallicaPress",
-        "GallicaMonographies",
-        "HAL",
-        # 'OtherFr',
-        "Persee",
-        "Theses",
-    ],
-    "mixed": ["PeS2o"],
-}
-
-datasets_categories = {
-    "technical": [
-        "HAL", 
-        "NIH_ExPorter",
-        "OpenEdition", 
-        "Persee", 
-        "PeS2o", 
-        "PhilPapers", 
-        "Theses", 
-        "USPTO_Backgrounds",
-        ],
-    "legi_written": [
-        "amendements_parlement", 
-        "Eurovoc.de",
-        "Eurovoc.en",
-        "Eurovoc.es",
-        "Eurovoc.it",
-        "FreeLaw",
-        "LEGI", 
-        "OpenData",
-        "questions_ecrites_parlement", 
-        ],
-    "legi_spoken": [
-        "DiscoursPublics",
-        "interventions_parlement",   
-        ],
-    "legi_dialogue": [
-        "AssembleeNationale",
-        "Europarl.en",
-        "Europarl.es",
-        "Europarl.de", 
-        "Europarl.fr",
-        "FREDSum",
-        "Senat", 
-        ],
-    "dialogue": [
-        "Claire.en", 
-        "Claire.fr", 
-        "Stac",
-        "ValidatedYouTube.fr",
-        ],
-    "book": [
-        "GallicaMonographies", 
-        "Gutenberg.en", 
-        "Gutenberg.de", 
-        "Gutenberg.it", 
-        "Gutenberg.es", 
-        "Gutenberg.fr",
-        ],
-    "newspaper": [
-        "AmericanStories", 
-        "GallicaPress",
-        ],
-    "forum": [
-        "StackExchange",
-        "Ubuntu_IRC", 
-        ],
-    "wiki": [
-        "Wikiother.fr", 
-        "Wikipedia.en", 
-        "Wikipedia.es", 
-        "Wikipedia.de", 
-        "Wikipedia.it", 
-        "Wikipedia.fr",
-        ],
-    "programming": [
-        "TheStack",
-        ],
-    "math": [
-        "DM_Mathematics",
-        "MathPile", 
-        ],
-    "aligned": [
-        "CroissantAligned",
-        "EuroparlAligned.fr-en",
-        "EuroparlAligned.es-en",
-        "EuroparlAligned.it-en",
-        "EuroparlAligned.de-fr",
-    ]
-}
-
-
-def _norm_string(s):
-    return s.lower().replace("_", " ")
-
-
-text_types = {k: [_norm_string(x) for x in v] for k, v in text_types.items()}
-datasets_categories = {k: [_norm_string(x) for x in v] for k, v in datasets_categories.items()}
-
-
-def is_ocr_dataset(name, subset):
-    if name in ["---", "", None]:
-        return ""
-    name = _norm_string(name)
-    res = "false"
-    if any(d in name for d in text_types["ocr"]):
-        res = "true"
-    if any(d in name for d in text_types["mixed"]):
-        res = "mixed"
-    return res
-
-
-def get_dataset_category(name, subset):
-    if name in ["---", "", None]:
-        return ""
-    name = _norm_string(name)
-    if name in ["pile", "otherfr"] and subset:
-        name = _norm_string(subset)
-    for cat, datasets in datasets_categories.items():
-        if name in datasets:
-            return cat
-    return None
-
 
 # Ignore datasets
 def ignore_datasets(name):
@@ -706,4 +579,4 @@ def sort_function(row):
             for row in rows_detailed:
                 row = compute_extra_stats(row, tokencount_folder)
                 row = format_stats_display(row, ONLY_DETAILED)
-                writer.writerow(row)
+                writer.writerow(row)