Add perplexity scores

calculate_perplexity.py

@@ -0,0 +1,24 @@
+import datasets
+import kenlm
+
+kenlm_model = kenlm.Model("/researchdisk/kenlm_clean_data_models/kenlm_clean_data_model_correct.bin")
+
+def get_perplexity(example):
+    """Get perplexity of text field."""
+    example["perplexity_kenlm_full"] = kenlm_model.perplexity(example["text"].lower())
+    return example
+
+dataset = datasets.load_dataset("csv", data_files={"train": "train.csv", "validation": "valid.csv"})
+
+# TRAIN SPLIT PERPLEXITY
+
+dataset["train"] = dataset["train"].map(get_perplexity, num_proc=64, writer_batch_size=100000)
+
+# VALIDATION SPLIT PERPLEXITY
+
+dataset["validation"] = dataset["validation"].map(get_perplexity, num_proc=64, writer_batch_size=100000)
+
+# SAVE DATASET
+
+dataset["train"].to_csv("train.csv", num_proc=64, index=False)
+dataset["validation"].to_csv("valid.csv", num_proc=64, index=False)

train.csv

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:857db09dc9d10a5606768ff8885d6fe1162f5dbbbb2aa82977278499530a3fd9
-size 64553318366
+oid sha256:d4719836b089156cf5f7cb332af6b639295e7f82eb37cd4a32f20ed4dc667b1e
+size 64886601971

valid.csv

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:816cb816f2f002e158bba01e19eee6cec58d18a37919f1284de45decbe331289
-size 63928522
+oid sha256:60e519a0379423f3ee4380edec58e5f30ad574e5c19b797b4ed5a6dedeff0fbe
+size 64259420