fe1ixxu/ALMA

Issues

• Unable to Reproduce ALMA-7b-LoRA Performance, Seeking Assistance

  #58 opened 2 months ago by liangyingshao
  6

• Fail to load ALMA-13B

  #50 opened 2 months ago by wygao8
  1

• error on pretraining

  #57 opened 2 months ago by tsbiosky
  1

• Using custom monolingual data instead of OSCAR dataset

  #28 opened 8 months ago by learnercat
  3

• 单语数据集的构建

  #41 opened 7 months ago by ywlq
  2

• Issues with Translation Quality Using ALMA/ALMA-R Models on Multi-Domain Dataset

  #56 opened 3 months ago by cocaer
  2

• citation for CPO paper

  #48 opened 3 months ago by kashif
  1

• [Bug/Feature] The dataset isn't reading the same cache_dir

  #35 opened 3 months ago by alvations
  2

• GPUs used during parallel data fine-tuning

  #55 opened 3 months ago by liangyingshao
  1

• i change the sh for evaluate my data ,but met error

  #54 opened 4 months ago by DengNingyuan
  0

• torch版本问题

  #53 opened 4 months ago by DengNingyuan
  0

• Unknown source language

  #52 opened 4 months ago by noahdasanaike
  0

• predict problem

  #51 opened 4 months ago by leee-SeungHyeon
  0

• Questions about Inference

  #47 opened 5 months ago by kira-lin
  1

• preprocess_cpo_data

  #46 opened 5 months ago by martimfasantos
  1

• Question about ALMA(R)

  #45 opened 5 months ago by mru4913
  2

• OOM 问题, 显卡是A00 40G

  #42 opened 6 months ago by gongye19
  4

• OSError: Error no file named pytorch_model.bin, tf_model.h5, model.ckpt.index or flax_model.msgpack found in directory

  #43 opened 6 months ago by gongye19
  2

• CPO question

  #44 opened 6 months ago by gongye19
  2

• Error on running the evaluation command

  #39 opened 7 months ago by Amrit-Bhaskar-abhask10
  2

• [Question] Exception during parallel data finetuning

  #40 opened 7 months ago by aiyubx
  1

• Question on cpo loss

  #36 opened 7 months ago by vince62s
  1

• No such file or directory

  #38 opened 7 months ago by hxue3
  0

• Training metrics currently not logged?

  #37 opened 7 months ago by SirRob1997
  2

• Loading ALMA-7B-R (LORA merged) through huggingface downloads Pretrained + LORA

  #30 opened 7 months ago by tranvaj
  1

• How much parallel data?

  #27 opened 7 months ago by zidsi
  1

• [Question] Replicating ALMA by training from scratch

  #34 opened 7 months ago by alvations
  2

• Got Errors when pretraining LLaMA-2 on Monolingual Dataset

  #33 opened 8 months ago by vhientran
  4

• [Question] Suggested machine and GPUs to run the training

  #6 opened a year ago by alvations
  2

• Runing parallel_ft_lora.sh

  #31 opened 8 months ago by zhengkid
  2

• Pretraining inquiry.

  #32 opened 8 months ago by gyupro
  3

• A couple of questions for your theory

  #29 opened 8 months ago by gyupro
  2

• Incomplete Translation from English to Chinese although `max_tokens` is enough

  #24 opened 8 months ago by DeyangKong
  3

• running `runs/parallel_ft_lora.sh`gives overflow

  #25 opened 9 months ago by hndrstwn
  2

• How much CPO data set is expected to be needed when creating a one-to-one machine translator?

  #26 opened 9 months ago by qwopqwop200
  3

• Fine-tuning on longer contexts for better performance?

  #23 opened 9 months ago by NilanEkanayake
  1

• </s> or eos needed for other base models?

  #22 opened 9 months ago by zidsi
  2

• DPODataCollatorWithPadding( TypeError: __init__() got an unexpected keyword argument 'max_length'

  #21 opened 9 months ago by sahsaeedi
  2

• The English-Chinese translation is incomplete.

  #20 opened 9 months ago by detectRecog
  2

• How to fix error to access huggingface?

  #18 opened 9 months ago by vhientran
  1

• Polite form selection

  #13 opened 10 months ago by cmp-nct
  5

• Regarding the memory usage of full-weight fine-tuning

  #17 opened a year ago by Franciscus-Carolus
  2

• Suggestion on foundation model

  #16 opened a year ago by cmp-nct
  4

• Release `Random` and `Filtered` parallel corpora

  #14 opened a year ago by zwhe99
  2

• NotImplementedError: all_exhausted stopping strategy in `interleave_datasets` is not implemented yet with a list of <class 'datasets.iterable_dataset.IterableDataset'>.

  #15 opened a year ago by zwhe99
  2

• About how to specify pairs in `Parallel_ft.sh`.

  #12 opened a year ago by kyoto7250
  2

• questions about reproduce the results of paper

  #9 opened a year ago by ZeroneBo
  4

• What do i need to add a new language ?

  #8 opened a year ago by MohamedAliRashad
  1

• 7B or 13B ?

  #7 opened a year ago by geronimi73
  1

• [Question] Is "this" needed in the general prompt?

  #5 opened a year ago by alvations
  1