tatsu-lab/stanford_alpaca

Issues

• ValueError: Trying to set a tensor of shape torch.Size([32769536]) in "weight" (which has shape torch.Size([32001, 4096])), this looks incorrect.

  #319 opened 2 months ago by daidaiershidi
  3

• SFT Mistral；

  #317 opened 6 months ago by feiying12343
  2

• Keyword arguments {'add_special_tokens': False} not recognized.

  #313 opened 9 months ago by cswangxiaowei
  1

• How to finetune with a own private data and then build chatbot on that?

  #296 opened a year ago by rjtshrm
  2

• weight_diff.py state_dict_recovered[key].add_(state_dict_raw[key]) RuntimeError: The size of tensor a (32001) must match the size of tensor b (32000) at non-singleton dimension 0

  #304 opened a year ago by gaodexiaozheng
  3

• RuntimeError: The size of tensor a (65539072) must match the size of tensor b (262156288) at non-singleton dimension 0

  #316 opened 7 months ago by YuyangJ0
  0

• Tensors of the same index must be on the same device and the same dtype except `step` tensors that can be CPU and float32 notwithstanding

  #315 opened 8 months ago by wurevvc
  0

• train.py fails with TypeError: Object of type Tensor is not JSON serializable

  #314 opened 8 months ago by khayamgondal
  0

• openai version

  #312 opened 9 months ago by cswangxiaowei
  1

• Cuda OOM during training

  #308 opened 10 months ago by hychaochao
  0

• Finetune with A100 40G

  #280 opened a year ago by jianchaoji
  4

• The arugment order of Rouge score might be wrong.

  #307 opened a year ago by zhaowei-wang-nlp
  0

• Problems generating my own data offline

  #305 opened a year ago by JieDengsc
  0

• NotImplementedError: offload_to_cpu=True and NO_SHARD is not supported yet

  #306 opened a year ago by mechigonft
  1

• ImportError when using `weight_diff.py` script

  #303 opened a year ago by SebastianRiquelmeM
  2

• How to get the model

  #302 opened a year ago by Guo-Chenxu
  0

• Can you release your evaluation code and data?

  #301 opened a year ago by mshich1
  0

• Loss will suddenly turn 0 during SFT

  #298 opened a year ago by zhangyx0417
  2

• Question about padding the input sequence

  #294 opened a year ago by BaleChen
  3

• AttributeError: 'ModelArguments' object has no attribute 'target_modules'

  #300 opened a year ago by juzidelang
  0

• Confusion about instruction task

  #299 opened a year ago by mitchelldehaven
  1

• TypeError: 'type' object is not subscriptable

  #287 opened a year ago by WYXG233
  2

• Utilize regen.json in finetuning

  #297 opened a year ago by Yijia-Xiao
  0

• Wonder how to inference after finetuning.

  #295 opened a year ago by 5taku
  0

• Python bindings and text classification & summarisation tasks

  #293 opened a year ago by Dee-Ma
  0

• How to provide extra contexrt as a pdf file?

  #292 opened a year ago by gamerjazzar
  0

• [Windows]: RuntimeError: Distributed package doesn't have NCCL built in

  #291 opened a year ago by SkibaSAY
  0

• Training bug for 13b, 30b, and 65b

  #285 opened a year ago by alexgshaw
  5

• BUG: "labels" information leakage into "input_ids" fields - incorrect attention_mask

  #290 opened a year ago by Nsigma-Bill
  1

• incorrect model_max_length

  #289 opened a year ago by joemkwon
  1

• Why do we pass both question and answer as input to the model during training?

  #281 opened a year ago by ruiyigan
  0

• DeepSpeed compilation (cpu_adam issue)

  #288 opened a year ago by JohnTailor
  0

• The OOM problem caused by the Transformers version

  #278 opened a year ago by kiseliu
  2

• Model Training Never Starts - Can't Finetune Anymore

  #268 opened a year ago by cumbersomeamir
  1

• How to fine tuning th model with limited resource?

  #270 opened a year ago by GivanTsai
  1

• where did the code define the wandb?

  #286 opened a year ago by applepieiris
  0

• Location of Log Files for the model?

  #284 opened a year ago by harshaelon
  0

• How to classify all the data?

  #283 opened a year ago by rayrayraykk
  0

• how to determine the basis of fine-tuen is enough?

  #282 opened a year ago by aijianiula0601
  0

• How to finetune using the customizer data?

  #279 opened a year ago by JustinZou1
  0

• Inquiry about license

  #276 opened a year ago by CallMeDek
  1

• why I save two pytorch_model.bin with same size

  #277 opened a year ago by qwjaskzxl
  0

• encounter errors when I try to finetune the model

  #273 opened a year ago by SleepEarlyLiveLong
  2

• Why the model I got after finetune is not good

  #275 opened a year ago by wyzhhhh
  0

• Older Mac

  #272 opened a year ago by betolley
  2

• how to train alpaca with single GPU A100 without FSDP?

  #271 opened a year ago by moseshu
  0

• Can fine-tuning run on multi machine distributedly?

  #269 opened a year ago by ovasty
  0

• Number of trainable parameters is less than 7B

  #266 opened a year ago by yuanzhedong
  1

• I have trained 10000 records from alpaca_data.json; but encountered an unrecognized response

  #267 opened a year ago by gugongerguo
  0

• Why are 28 of the outputs empty?

  #265 opened 2 years ago by tetratorus
  0