Discussions

• 1 You must be logged in to vote

  🙏

  tie_word_embeddings with LoRA + new special tokens: keep or untie?

  Xxxxsir asked Apr 10, 2026 in Q&A · Unanswered

  

  

  1
• 1 You must be logged in to vote

  🙌

  Hugging Face PEFT Integration of KappaTune

  oswaldoludwig started Mar 13, 2026 in Show and tell

  

  

  3
• 1 You must be logged in to vote

  🙏

  Does T5 actually utilize Encoder prefixes in PrefixTuning? past_key_values seems ignored by the Encoder.

  david900125 asked Jan 5, 2026 in Q&A · Unanswered

  

  

  4
• 1 You must be logged in to vote

  💬

  Reproducible gradient norm spike in QLoRA at step 44 on Mistral-7B — spectral norm constraint as a fix

  fourwheels2512 started Feb 22, 2026 in General

  

  

  1
• 1 You must be logged in to vote

  🙏

  Comparison between different PEFT and Adapter techniques

  ImamaDev asked Jan 7, 2025 in Q&A · Unanswered

  

  

  

  

  2
• 1 You must be logged in to vote

  🙏

  How to fine tune a PEFT model in stages with LoRA/QLoRA

  supreme-gg-gg asked Sep 7, 2025 in Q&A · Answered

  

  

  

  

  4
• 1 You must be logged in to vote

  🙏

  Help setting up SDXL LoRA training on Google Cloud (Vertex AI Workbench)

  aici23 asked Aug 21, 2025 in Q&A · Unanswered

  

  

  1
• 1 You must be logged in to vote

  🙌

  Fused Qwen3 MoE layer with LoRA

  woct0rdho started Jun 29, 2025 in Show and tell

  

  

  

  

  6
• 2 You must be logged in to vote

  🙏

  QLoRA + FSDP with compressed-tensors format quantized models

  kmehant asked Jun 26, 2025 in Q&A · Closed · Unanswered

  

  

  2
• 1 You must be logged in to vote

  🙏

  How to load a PEFT adapter with per-layer rank variation via from_pretrained?

  yuxuan-z19 asked Jun 9, 2025 in Q&A · Answered

  

  

  3
• 1 You must be logged in to vote

  🙏

  Where can I find the existing community experiment results for method_comparision?

  2catycm asked Jun 14, 2025 in Q&A · Answered

  

  

  7
• 1 You must be logged in to vote

  🙏

  Fail to merge for a custom model.

  Boltzmachine asked Jun 11, 2025 in Q&A · Unanswered

  

  

  3
• 3 You must be logged in to vote

  🙏

  peft provides the position of the model input?

  ciaoyizhen asked Mar 19, 2024 in Q&A · Answered

  

  

  1
• 1 You must be logged in to vote

  🙌

  PEFT code evolution

  andrehora started Apr 23, 2025 in Show and tell

  

  

  0
• 1 You must be logged in to vote

  🙏

  Fuse Similar Modules for Inference?

  Beinsezii asked Mar 27, 2025 in Q&A · Unanswered

  

  

  

  

  9
• 2 You must be logged in to vote

  💡

  LLama Adapters for GPT

  efraimdahl started Mar 10, 2025 in Ideas

  

  

  3
• 1 You must be logged in to vote

  💡

  Multi-modal projections for Llama-Adapter

  efraimdahl started Mar 11, 2025 in Ideas

  

  

  1
• 1 You must be logged in to vote

  💬

  ONNX support

  Valoisa started Feb 26, 2025 in General

  

  

  1
• 1 You must be logged in to vote

  🙏

  Is it possible to obtain separate features for multiple adapters in one inference?

  foolishcqx asked Feb 21, 2025 in Q&A · Answered

  

  

  4
• 1 You must be logged in to vote

  🙏

  Comparison between different instantiation of Peftmodel

  shedding-ash asked Feb 19, 2025 in Q&A · Answered

  

  

  1
• 1 You must be logged in to vote

  🙌

  PermuLoRA

  JBedichek started Feb 8, 2025 in Show and tell

  

  

  2
• 1 You must be logged in to vote

  💬

  Not able to set early_stopping_patience because error with load_best_model_at_end=True. Avoiding over-fitting.

  sneha117 started Jan 29, 2025 in General

  

  

  3
• 2 You must be logged in to vote

  🙏

  QLoRa with Vision Transformers

  anwai98 asked Apr 23, 2024 in Q&A · Closed · Unanswered

  

  

  

  

  6
• 1 You must be logged in to vote

  🙏

  How to train multiple LoRA at the same time, then esemble outputs to make prediction?

  quanphamnus asked Jan 1, 2025 in Q&A · Unanswered

  

  

  1
• 1 You must be logged in to vote

  🙏

  Help Needed: Implementing QLoRA on AutoGPTQ Model - Unsupported Module Error

  SkanderGhariani asked Jun 14, 2024 in Q&A · Answered

  

  

  

  

  8