Search results

[Gemma3N] Not able to add new special tokens to model/tokenizer due to projection error

bug

Good Second IssueIssues that are more difficult to do than "Good First" issues - give it a try if you want!Issues that are more difficult to do than "Good First" issues - give it a try if you want!

UsageGeneral questions about the libraryGeneral questions about the library

Status: Open.

#39921 In huggingface/transformers;

· debasisdwivedy opened on Aug 5, 2025

../aten/src/ATen/native/cuda/Indexing.cu:1289: indexSelectLargeIndex: block: [267,0,0], thread: [25,0,0] Assertion <code>srcIndex &lt; srcSelectDimSize</code> failed.

bug

Good Second IssueIssues that are more difficult to do than "Good First" issues - give it a try if you want!Issues that are more difficult to do than "Good First" issues - give it a try if you want!

UsageGeneral questions about the libraryGeneral questions about the library

Status: Open.

#33985 In huggingface/transformers;

· JHW5981 opened on Oct 6, 2024

Automatic dynamic batch size selection for DataCollatorWithFlattening

Feature requestRequest for a new featureRequest for a new feature

UsageGeneral questions about the libraryGeneral questions about the library

Status: Open.

#33945 In huggingface/transformers;

· alex-hh opened on Oct 4, 2024

How to implement weight decay towards the pre-trained model?

Feature requestRequest for a new featureRequest for a new feature

UsageGeneral questions about the libraryGeneral questions about the library

Status: Open.

#33909 In huggingface/transformers;

· sedol1339 opened on Oct 3, 2024

Enable changing the loss function by making the hard-coded <code>loss_fct</code> an attribute of <code>BertForTokenClassification</code>.

Core: ModelingInternals of the library; Models.Internals of the library; Models.

Feature requestRequest for a new featureRequest for a new feature

UsageGeneral questions about the libraryGeneral questions about the library

Status: Open.

#33642 In huggingface/transformers;

· tom-010 opened on Sep 21, 2024

oom when using adafactor optimizer in deepspeed

bug

Good First Issue

UsageGeneral questions about the libraryGeneral questions about the library

Status: Open.

#33290 In huggingface/transformers;

· zhangvia opened on Sep 4, 2024

How to use hugginface for training: google-t5/t5-base

Feature requestRequest for a new featureRequest for a new feature

UsageGeneral questions about the libraryGeneral questions about the library

Status: Open.

#33232 In huggingface/transformers;

· gg22mm opened on Aug 31, 2024