ericlbuehler/xlora issues and pull requests

#34 - Is there a training script?

Issue - State: open - Opened by phbst 2 months ago

#33 - Generation error

Issue - State: open - Opened by MaxwellJryao 3 months ago

#32 - seq_len is always 1

Issue - State: closed - Opened by Zehrooo 5 months ago

#31 - How to get Token-by-token scalings pic like Readme

Issue - State: open - Opened by Zehrooo 5 months ago

#30 - request for the training code

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

#29 - Support for seq2seq models

Issue - State: open - Opened by maximus-21 6 months ago - 2 comments

#28 - does xlora train not support llama2?

Issue - State: open - Opened by crossxxd 7 months ago - 4 comments

#27 - Support loading from HF hub

Pull Request - State: closed - Opened by EricLBuehler 8 months ago

#26 - cross entropy loss during training with xlora

Issue - State: open - Opened by crossxxd 8 months ago - 4 comments

#25 - Support quantized models

Pull Request - State: open - Opened by EricLBuehler 8 months ago
Labels: enhancement

#24 - Would you kindly update Xlora to support Quantized Models?

Issue - State: open - Opened by Abdullah-kwl 8 months ago - 9 comments
Labels: enhancement

#23 - Can XLora work fine or same if LoRA -> xLoRA complete: Swapped 0 LoRA layers:

Issue - State: closed - Opened by Abdullah-kwl 9 months ago - 1 comment

#22 - xLoRA complete: Swapped 0 LoRA layers (out of 1897 modules)

Issue - State: closed - Opened by Abdullah-kwl 9 months ago - 1 comment

#21 - Fix topk

Pull Request - State: closed - Opened by EricLBuehler 9 months ago
Labels: bug

#20 - Removal of commented sections

Pull Request - State: closed - Opened by EricLBuehler 9 months ago
Labels: enhancement

#19 - Specifying adapters path on HF hub does not work

Issue - State: closed - Opened by EricLBuehler 10 months ago - 1 comment
Labels: bug

#18 - Any plans to share code for training?

Issue - State: closed - Opened by JY-CCK 10 months ago - 8 comments
Labels: triaged

#17 - Add set topk

Pull Request - State: closed - Opened by EricLBuehler 10 months ago

#16 - Add manual (global) LoRA weight

Pull Request - State: closed - Opened by EricLBuehler 10 months ago
Labels: enhancement

#15 - Update requirements.txt

Pull Request - State: closed - Opened by lamm-mit 10 months ago

#14 - Update scalings logging implementation

Pull Request - State: closed - Opened by EricLBuehler 10 months ago
Labels: enhancement

#13 - Update README.md, reference/citation

Pull Request - State: closed - Opened by lamm-mit 10 months ago
Labels: enhancement

#12 - Save all adapters

Pull Request - State: closed - Opened by EricLBuehler 10 months ago - 1 comment
Labels: enhancement, major-change

#11 - Load from hf hub feature

Issue - State: closed - Opened by EricLBuehler 10 months ago
Labels: tracking

#10 - Implement seqlen-wise scalings

Pull Request - State: closed - Opened by EricLBuehler 10 months ago
Labels: enhancement, major-change

#9 - `use_mean_pool` is broken

Issue - State: closed - Opened by EricLBuehler 10 months ago
Labels: bug

#8 - Multi-device training tracking issue

Issue - State: closed - Opened by EricLBuehler 10 months ago - 1 comment
Labels: optimization, tracking

#7 - Remove all global state