Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / huggingface/peft issues and pull requests

#1681 - FIX Use trl version of tiny random llama

Pull Request - State: closed - Opened by BenjaminBossan 5 months ago - 2 comments

#1680 - FIX / Workflow: Fix Mac-OS CI issues

Pull Request - State: closed - Opened by younesbelkada 5 months ago - 1 comment

#1679 - Update deepspeed.md

Pull Request - State: closed - Opened by sanghyuk-choi 5 months ago - 2 comments

#1677 - DoRA support for Embedding

Issue - State: closed - Opened by ShayekhBinIslam 5 months ago - 5 comments
Labels: contributions-welcome

#1676 - Clarification needed on Adapter Heads in PEFT

Issue - State: closed - Opened by lenglaender 5 months ago - 2 comments

#1675 - FEAT: Add EETQ support in PEFT

Pull Request - State: closed - Opened by younesbelkada 5 months ago - 1 comment

#1674 - [FSDP+QLoRA] ValueError: Expected a cuda device, but got: cpu

Issue - State: closed - Opened by iseesaw 5 months ago - 7 comments

#1672 - Using PEFT causes model to not predict EOS

Issue - State: closed - Opened by Km3888 5 months ago - 15 comments

#1670 - Updating only one adapter using multi adapter

Issue - State: closed - Opened by LameloBally 5 months ago - 2 comments

#1669 - Inquiry on Inserting P-Tuning Soft Prompts at Any Position in Input

Issue - State: closed - Opened by Betty1202 5 months ago - 2 comments

#1667 - ENH: Add multi-backend tests for bnb

Pull Request - State: closed - Opened by younesbelkada 5 months ago - 1 comment

#1666 - supports learning the combination weights of pre-trained LoRA modules

Pull Request - State: closed - Opened by mahdibeit 5 months ago - 7 comments

#1665 - `bitsandbytes` is imported eagerly in `peft/utils/loftq_utils.py`

Issue - State: closed - Opened by function2-llx 5 months ago - 5 comments

#1664 - Apply deprecated `evaluation_strategy`

Pull Request - State: closed - Opened by muellerzr 6 months ago - 10 comments

#1663 - FEAT Show adapter layer and model status

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 2 comments

#1662 - add safetensor load in multitask_prompt_tuning

Pull Request - State: closed - Opened by sywangyi 6 months ago - 2 comments

#1660 - Remove `dreambooth` Git link

Pull Request - State: closed - Opened by charliermarsh 6 months ago - 1 comment

#1659 - FIX: Bug with handling of active adapters

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 1 comment

#1658 - save and load base model with revision

Pull Request - State: closed - Opened by mnoukhov 6 months ago - 13 comments

#1657 - Integrating ReFT

Pull Request - State: closed - Opened by raven38 6 months ago - 17 comments

#1656 - RuntimeError: CUDA error: invalid device ordinal

Issue - State: closed - Opened by yumath 6 months ago - 2 comments

#1655 - Learning the combination weights of pre-trained LoRA Modules

Issue - State: closed - Opened by mahdibeit 6 months ago - 5 comments

#1654 - Integrating ReFT

Issue - State: closed - Opened by raven38 6 months ago - 9 comments

#1653 - FIX: Initialize DoRA weights in float32 if float16 is being used

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 1 comment

#1652 - FIX Issues with AdaLora initialization

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 1 comment

#1651 - Training with out Trainer class. (VRAM usage issue)

Issue - State: closed - Opened by venzino-han 6 months ago - 4 comments

#1650 - Backward compatibility on saved config.

Issue - State: closed - Opened by xkszltl 6 months ago - 5 comments

#1649 - model prediction differs before and after saving the model

Issue - State: closed - Opened by IamExperimenting 6 months ago - 2 comments

#1648 - print_trainable_parameters - format `%` to be sensible

Pull Request - State: closed - Opened by stas00 6 months ago - 1 comment

#1647 - Problem with merging AdaLoRA adapters

Issue - State: closed - Opened by h4iku 6 months ago - 1 comment

#1646 - Adapters cannot be correctly loaded using PeftModel.from_pretrained()

Issue - State: closed - Opened by HenryCai11 6 months ago - 3 comments

#1645 - [docs] usability

Issue - State: closed - Opened by stas00 6 months ago - 3 comments

#1645 - [docs] usability

Issue - State: open - Opened by stas00 6 months ago

#1644 - [doc] issues with rendering docs in FF

Issue - State: open - Opened by stas00 6 months ago

#1644 - [doc] issues with rendering docs in FF

Issue - State: closed - Opened by stas00 6 months ago - 6 comments

#1643 - Support EETQ QLoRA

Issue - State: closed - Opened by thincal 6 months ago - 11 comments

#1642 - update figure assets of BOFT

Pull Request - State: closed - Opened by YuliangXiu 6 months ago - 1 comment

#1641 - Don't use deprecated `Repository` anymore

Pull Request - State: closed - Opened by Wauplin 6 months ago - 3 comments

#1636 - Error running the deepspeed qlora example

Issue - State: closed - Opened by AnirudhVIyer 6 months ago - 2 comments

#1636 - Error running the deepspeed qlora example

Issue - State: closed - Opened by AnirudhVIyer 6 months ago - 2 comments

#1635 - FIX Correctly call element_size

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 3 comments

#1635 - FIX Correctly call element_size

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 3 comments

#1634 - fix bf16 model type issue for ia3

Pull Request - State: closed - Opened by sywangyi 6 months ago - 4 comments

#1633 - Prompt Tuning with Zero3 didn't work

Issue - State: open - Opened by tricktreat 6 months ago - 2 comments

#1633 - Prompt Tuning with Zero3 didn't work

Issue - State: closed - Opened by tricktreat 6 months ago - 4 comments

#1631 - fix: allow load_adapter to use different device

Pull Request - State: closed - Opened by yhZhai 6 months ago - 3 comments

#1631 - fix: allow load_adapter to use different device

Pull Request - State: open - Opened by yhZhai 6 months ago - 3 comments

#1630 - itemsize is torch>=2.1, use element_size()

Pull Request - State: closed - Opened by winglian 6 months ago - 7 comments

#1630 - itemsize is torch>=2.1, use element_size()

Pull Request - State: closed - Opened by winglian 6 months ago - 6 comments

#1629 - FIX Errors in the transformers integration docs

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 1 comment

#1628 - FIX: Send results to correct channel

Pull Request - State: closed - Opened by younesbelkada 6 months ago - 1 comment

#1628 - FIX: Send results to correct channel

Pull Request - State: closed - Opened by younesbelkada 6 months ago - 1 comment

#1627 - Seek help for a more efficient way to use caching to train my model

Issue - State: closed - Opened by whyiug 6 months ago - 4 comments

#1626 - Adding PiSSA as an optional initialization method of LoRA

Pull Request - State: closed - Opened by fxmeng 6 months ago - 41 comments

#1625 - Adalora deepspeed

Pull Request - State: closed - Opened by sywangyi 6 months ago - 1 comment

#1624 - FIX In the image classification example, Change the model to the LoRA…

Pull Request - State: closed - Opened by changhwa 6 months ago - 2 comments

#1622 - Remove duplicated import

Pull Request - State: closed - Opened by nzw0301 6 months ago - 1 comment

#1620 - FEAT: Allow ignoring mismatched sizes when loading

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 1 comment

#1618 - Add LoRA support to HQQ Quantization

Pull Request - State: closed - Opened by fahadh4ilyas 6 months ago - 20 comments

#1617 - replace_lora_weights_loftq seems not present in transformers

Issue - State: closed - Opened by RonanKMcGovern 6 months ago - 2 comments

#1616 - Support HQQ method.

Issue - State: closed - Opened by Minami-su 6 months ago - 11 comments

#1615 - FIX Multiple adapters and modules_to_save

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 2 comments

#1614 - fix the torch_dtype and quant_storage_dtype

Pull Request - State: closed - Opened by pacman100 6 months ago - 2 comments

#1613 - peft v0.10 takes up too much GPU memory than v0.3.0

Issue - State: closed - Opened by azuryl 6 months ago - 14 comments

#1612 - LoRA layer param.grad=None after loss.backward() using lora for mamba 2.8B

Issue - State: closed - Opened by jinlHe 6 months ago - 4 comments

#1610 - Prefix tuning configuration issue

Issue - State: closed - Opened by AswiniNLP 6 months ago - 1 comment

#1609 - AttributeError: ModulesToSaveWrapper has no attribute layers

Issue - State: closed - Opened by pxy963 6 months ago - 4 comments

#1608 - Configuration issue

Issue - State: closed - Opened by AswiniNLP 6 months ago - 11 comments

#1607 - p_tuning layers disappear after loading peft model

Issue - State: closed - Opened by victoriousMan 6 months ago - 4 comments

#1606 - Using LoRA on custom models

Issue - State: closed - Opened by wsuSaiman 6 months ago - 2 comments

#1604 - [Feature Request] Support `torch.nn.Conv1d` layers for LoRA

Issue - State: closed - Opened by dg845 6 months ago - 2 comments

#1603 - FIX: bnb config wrong argument names

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 2 comments

#1602 - TST: Add simple BNB regression tests

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 5 comments

#1601 - LISA

Issue - State: closed - Opened by timothelaborie 6 months ago - 7 comments

#1600 - MNT: Update GH bug report template

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 1 comment

#1598 - examples/sft/run_peft.sh model load dtype error

Issue - State: closed - Opened by galtay 6 months ago - 7 comments

#1597 - RuntimeError: "addmm_impl_cpu_" not implemented for 'Half

Issue - State: closed - Opened by azuryl 6 months ago - 11 comments

#1595 - Merging models and feature extraction

Issue - State: closed - Opened by tyemelya 6 months ago - 4 comments

#1594 - model.print_trainable_parameters() is incomplete.

Issue - State: closed - Opened by Minami-su 6 months ago - 5 comments

#1593 - Getting Dora Model Is Very Slow

Issue - State: closed - Opened by mallorbc 6 months ago - 10 comments

#1592 - ...

Issue - State: closed - Opened by Minami-su 6 months ago

#1591 - fix deepspeed zero3+prompt tuning bug. word_embeddings.weight shape i…

Pull Request - State: closed - Opened by sywangyi 6 months ago - 4 comments

#1589 - TaskType attribute in Lora config for question answering is wrong

Issue - State: closed - Opened by aryangoyal7 6 months ago - 4 comments

#1588 - FIX Make DoRA work with Conv1D layers

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 1 comment

#1587 - DOC: Short section on using transformers pipeline

Pull Request - State: closed - Opened by BenjaminBossan 6 months ago - 1 comment