kingoflolz/mesh-transformer-jax

Issues

• Web demo is not launching any results. Might be disconnected from the model.

  #241 opened 2 years ago by Gertie01
  43

• lm_eval missing

  #264 opened 3 months ago by falv706
  0

• GPT-J used in "Domain-Specific Text Generation for Machine Translation"

  #250 opened 8 months ago by ymoslem
  0

• AttributeError: module 'jax.random' has no attribute 'KeyArray' while fine tuning.

  #221 opened 3 years ago by samyakai
  15

• 6b.eleuther.ai mystic model is down for GPT-J-6B.

  #263 opened a year ago by Gertie01
  0

• Web demo must be fixed.

  #261 opened a year ago by Gertie01
  0

• TPU-V4

  #255 opened 2 years ago by wimjan123
  11

• 6b.eleuther.ai mystic model is down for GPT-J-6B

  #226 opened 3 years ago by orionnelson
  4

• What about a Hugging Face Spaces demo so we can test this?

  #262 opened a year ago by Gertie01
  0

• About rope embedding

  #260 opened a year ago by eyuansu62
  0

• Framework

  #259 opened a year ago by T3fo0ls7766
  0

• Finetuning Hardware Recomendations

  #258 opened a year ago by greyweb
  0

• How to stop model generating

  #228 opened 3 years ago by jingrongchen
  1

• Download Link for Model Weights in howto_finetune.md is broken

  #251 opened 2 years ago by torakoneko
  2

• The PILE dataset is full of racist content and thus GPT-J produces racist thinking.

  #240 opened 2 years ago by azeemh
  2

• How to infer with GPT-J on TPU_driver0.2 or nightly?

  #256 opened 2 years ago by mosmos6
  1

• tpu_driver0.1 is not initialized on colab (cannot infer with GPT-J on Colab) [Again]

  #252 opened 2 years ago by mosmos6
  7

• Discrepancy between results reported in this repo and in the NeoX paper

  #257 opened 2 years ago by william-cerebras
  2

• Resolving dependency issues

  #246 opened 2 years ago by rinapch
  6

• Can we please get a quickstart guide?

  #243 opened 2 years ago by tswallen
  2

• Which version of Python does this work with?

  #253 opened 2 years ago by chrisbward
  2

• Quantization for training / finetuning

  #254 opened 2 years ago by torphix
  0

• TypeError: __init__() takes 2 positional arguments but 4 were given

  #225 opened 3 years ago
  1

• Could not find a version that satisfies the requirement ray[default]==1.4.1

  #245 opened 2 years ago by Maxim-Mazurok
  5

• Fine-tuning on conversations (format of conversations)

  #248 opened 2 years ago by Eichhof
  2

• Do you have any plans to create the open source version of chatGPT ?

  #244 opened 2 years ago by stc2001
  2

• training stuck at validation step 1

  #218 opened 3 years ago by Selimonder
  4

• TPU not found on VM (jax version 0.2.16)

  #242 opened 2 years ago by Eichhof
  0

• Project dependencies may have API risk issues

  #239 opened 2 years ago by PyDeps
  0

• AttributeError: module 'jaxlib.pocketfft' has no attribute 'pocketfft'

  #233 opened 2 years ago by umm-maybe
  4

• Dead link to weights?

  #238 opened 2 years ago by samacqua
  1

• TPU Instance Creation

  #237 opened 2 years ago by zzj0402
  2

• Update the readme with required and recommended hardware list

  #236 opened 2 years ago by sxiii
  3

• on number of training tokens of gpt-j-6b and gpt-neox-20b

  #235 opened 2 years ago by xiaoda99
  0

• Is the treatment of embedding bias in to_hf_weights.py correct?

  #234 opened 2 years ago by xiaoda99
  2

• TypeError: Cannot subclass <class 'typing._SpecialForm'> while fine tuning

  #222 opened 3 years ago by samyakai
  9

• GPT-J-6B Inference Demo notebook giving errors when cores_per_replica=1

  #232 opened 2 years ago by batrasakshi
  1

• Google Colab Error: optax is throwing an attribute error.

  #230 opened 2 years ago by prajjwalgeek
  2

• Typo in 'to_hf_weights.py '

  #231 opened 2 years ago by AmoArt
  1

• [Feature Request] Multilingual assistance.

  #229 opened 3 years ago by phly95
  0

• Training data format for generating Scenario based MCQ's

  #224 opened 3 years ago by shrey10926
  2

• Finetuning GPT Neo 20B Using TPU V3-8s

  #227 opened 3 years ago by nikhilanayak
  0

• `TypeError: Cannot subclass <class 'typing._SpecialForm'>` in `slim_model.py `

  #212 opened 3 years ago by danyaljj
  3

• GPT-J inference on TPU

  #219 opened 3 years ago by airesearch38
  3

• CausalTransformerV2 or CausalTransformer?

  #220 opened 3 years ago by leejason
  0

• Can "slim_model.py" work with "d_model" as 768?

  #217 opened 3 years ago by leejason
  0

• Running on Colab TPU only gives random words and nonsensical outputs

  #216 opened 3 years ago by JohnnyRacer
  2

• Error while `to_hf_weights.py`: `ValueError: cannot reshape array of size 25804800 into shape (1,4096,50400)`

  #214 opened 3 years ago by danyaljj
  1

• `OSError: libmkl_intel_lp64.so.1: cannot open shared object file` when using `to_hf_weights.py`

  #215 opened 3 years ago by danyaljj
  1

• `Incompatible checkpoints` error when running `slim_model.py`

  #213 opened 3 years ago by danyaljj
  1