Upload folder using huggingface_hub

README.md

@@ -76,7 +76,7 @@ The model expects the following input format:
 - `<|GENE|>`: Optional gene name (truncated to a maximum of 10 characters in training).
 - `<|FLANK_BEFORE|>` and `<|FLANK_AFTER|>`: Optional upstream/downstream context sequences. Maximum of 25 nucleotides.
 
-The model should predict the next token as the class label: 0 (Exon) or 1 (Intron).
+The model should predict the class label: 0 (Exon) or 1 (Intron).
 
 ---
 
@@ -112,6 +112,8 @@ The model was trained on a processed version of GenBank sequences spanning multi
 | **Intron** | 0.9994    | 0.9994 | 0.9994   |
 | **Exon**   | 0.9997    | 0.9997 | 0.9997   |
 
+---
+
 ### Notes
 
 - Metrics were computed on a full isolated test set.

config.json

@@ -3,7 +3,7 @@
   "attention_probs_dropout_prob": 0.1,
   "classifier_dropout": null,
   "custom_pipelines": {
-    "gpt2-exon-intron-classification": {
+    "bert-exon-intron-classification": {
       "impl": "bert_exon_intron_classification.BERTExonIntronClassificationPipeline",
       "pt": ["BertForSequenceClassification"]
     }