Search results

First layer residual is inconsistently normalized

Status: Open.

#243 In AnswerDotAI/ModernBERT;

· lgienapp opened on Aug 16, 2025

Question about masking - why mask on unpadded sequence?

Status: Open.

#242 In AnswerDotAI/ModernBERT;

· yjoonjang opened on Aug 14, 2025

<code>_pad_token</code> attribute?

Status: Open.

#241 In AnswerDotAI/ModernBERT;

· ceferisbarov opened on Aug 11, 2025

Usage of GTE-ModernColBERT or ModernBERT embedding model for storing file chunks and vector embeddings in a vector store

Status: Open.

#240 In AnswerDotAI/ModernBERT;

· ampulsora opened on Aug 2, 2025

ValueError: Invalid pattern: '**' can only be an entire path component

Status: Open.

#237 In AnswerDotAI/ModernBERT;

· amrzv opened on Jun 8, 2025

question to implementation for ModernBertForMaskedLM vs ModernBertForSequenceClassification

Status: Open.

#235 In AnswerDotAI/ModernBERT;

· geraldstanje opened on May 22, 2025

Pretraining with FlexBert but can't export to HF

Status: Open.

#234 In AnswerDotAI/ModernBERT;

· asafam opened on May 21, 2025

grad_norm Nan question

Status: Open.

#233 In AnswerDotAI/ModernBERT;

· 937739823 opened on May 21, 2025

Pretraining Throughput — Unexpected Drop in Speed

Status: Open.

#228 In AnswerDotAI/ModernBERT;

· ebrarkiziloglu opened on Apr 30, 2025

Training Tokenizer and pretraining

Status: Open.

#226 In AnswerDotAI/ModernBERT;

· Thiago-Reis-Porto opened on Apr 17, 2025

How to continue to train ModernBert base/large in a specific domain?

Status: Open.

#222 In AnswerDotAI/ModernBERT;

· BernieVA opened on Apr 2, 2025

Multi-GPU Pretraining is slower than single GPU training?

Status: Open.

#220 In AnswerDotAI/ModernBERT;

· leannmlindsey opened on Mar 31, 2025