From 014e0c08138b84534f86e2f434c7590ec86f76f8 Mon Sep 17 00:00:00 2001 From: abhishekreddyallu Date: Tue, 1 Sep 2026 16:48:18 +0200 Subject: [PATCH 1/5] :bug: Alexbek taxonomy discovery learning-rate scheduler bug fix --- ontolearner/learner/taxonomy_discovery/alexbek.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ontolearner/learner/taxonomy_discovery/alexbek.py b/ontolearner/learner/taxonomy_discovery/alexbek.py index ea892fe4..58ef0431 100644 --- a/ontolearner/learner/taxonomy_discovery/alexbek.py +++ b/ontolearner/learner/taxonomy_discovery/alexbek.py @@ -560,7 +560,7 @@ def lr_lambda(step): if step < warmup_steps: return (step + 1) / (warmup_steps + 1) else: - progress = (step - warmup_steps) / (total_steps - warmup_steps) + progress = (step - warmup_steps) / max(1,(total_steps - warmup_steps)) return max(0.1, 0.5 * (1 + math.cos(math.pi * progress))) scheduler = torch.optim.lr_scheduler.LambdaLR(optimizer, lr_lambda=lr_lambda) From c4d3a07556fc53d8793083e09ef8ebd877c1c3d9 Mon Sep 17 00:00:00 2001 From: abhishekreddyallu Date: Tue, 1 Sep 2026 17:04:42 +0200 Subject: [PATCH 2/5] :bug: Alexbek term typing token handling --- ontolearner/learner/term_typing/alexbek.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/ontolearner/learner/term_typing/alexbek.py b/ontolearner/learner/term_typing/alexbek.py index 0db694b9..cfb8218c 100644 --- a/ontolearner/learner/term_typing/alexbek.py +++ b/ontolearner/learner/term_typing/alexbek.py @@ -823,7 +823,7 @@ def load( # Tokenizer self.tokenizer = AutoTokenizer.from_pretrained( - self.cfg["llm_model_id"], padding_side="left", token=self.cfg["token"] + self.cfg["llm_model_id"], padding_side="left", token=self.cfg["token"] or None ) if self.tokenizer.pad_token is None: self.tokenizer.pad_token = self.tokenizer.eos_token @@ -833,7 +833,7 @@ def load( self.cfg["llm_model_id"], device_map=device_map, torch_dtype=torch_dtype, - token=self.cfg["token"], + token=self.cfg["token"] or None, ) # Deterministic decoding defaults From a0cef9c08c566aa073bd2759bbe03f2871175819 Mon Sep 17 00:00:00 2001 From: abhishekreddyallu Date: Tue, 1 Sep 2026 17:14:37 +0200 Subject: [PATCH 3/5] :bug: rwthdbis taxonomy discovery --- ontolearner/learner/taxonomy_discovery/rwthdbis.py | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/ontolearner/learner/taxonomy_discovery/rwthdbis.py b/ontolearner/learner/taxonomy_discovery/rwthdbis.py index 14e47a8d..6832ba2a 100644 --- a/ontolearner/learner/taxonomy_discovery/rwthdbis.py +++ b/ontolearner/learner/taxonomy_discovery/rwthdbis.py @@ -1078,7 +1078,7 @@ def tokenize_batch(batch: Dict[str, List[str]]): ) tokenized_dataset = dataset_dict.map( - tokenize_batch, batched=True, remove_columns=dataset_dict["train"].column_names + tokenize_batch, batched=True, remove_columns=["text"] ) data_collator = DataCollatorWithPadding(self.tokenizer) @@ -1109,14 +1109,13 @@ def tokenize_batch(batch: Dict[str, List[str]]): fp16=self.fp16, bf16=self.bf16, report_to="none", - save_safetensors=True, ) trainer = Trainer( model=self.model, args=training_args, train_dataset=tokenized_dataset["train"], - tokenizer=self.tokenizer, + processing_class=self.tokenizer, data_collator=data_collator, ) trainer.train() From d19419bf1d258da62e261428d01acfa8c1f4aaa1 Mon Sep 17 00:00:00 2001 From: abhishekreddyallu Date: Tue, 1 Sep 2026 17:17:35 +0200 Subject: [PATCH 4/5] :bug: rwthdbis term typing trainer --- ontolearner/learner/term_typing/rwthdbis.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ontolearner/learner/term_typing/rwthdbis.py b/ontolearner/learner/term_typing/rwthdbis.py index 09ff40fd..77ce3164 100644 --- a/ontolearner/learner/term_typing/rwthdbis.py +++ b/ontolearner/learner/term_typing/rwthdbis.py @@ -1027,7 +1027,7 @@ def tokenize_batch(batch: Dict[str, List[str]]): model=self.model, args=training_args, train_dataset=tokenized["train"], - tokenizer=self.tokenizer, + processing_class=self.tokenizer, data_collator=data_collator, ) From 356ca26afd210b856ffd94562ffb73315560620a Mon Sep 17 00:00:00 2001 From: abhishekreddyallu Date: Tue, 1 Sep 2026 17:20:38 +0200 Subject: [PATCH 5/5] :bug: Skhnlp taxonomy discovery invalid argument --- ontolearner/learner/taxonomy_discovery/skhnlp.py | 1 - 1 file changed, 1 deletion(-) diff --git a/ontolearner/learner/taxonomy_discovery/skhnlp.py b/ontolearner/learner/taxonomy_discovery/skhnlp.py index cade6bba..aa739adb 100644 --- a/ontolearner/learner/taxonomy_discovery/skhnlp.py +++ b/ontolearner/learner/taxonomy_discovery/skhnlp.py @@ -181,7 +181,6 @@ def __init__( per_device_eval_batch_size=per_device_eval_batch_size, warmup_steps=warmup_steps, weight_decay=weight_decay, - logging_dir=logging_dir, logging_steps=logging_steps, eval_strategy=eval_strategy, save_strategy=save_strategy,