Skip to content

Issues: huggingface/alignment-handbook

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

86 Open 36 Closed
86 Open 36 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

https://github.com/huggingface/alignment-handbook
#176 opened Jun 26, 2024 by ratterdull78
1
TRL/Alignment-Handbook torch.dtype Issues
#175 opened Jun 26, 2024 by neelsjain
Question about torch_dtype when runnging run_orpo.py
#174 opened Jun 23, 2024 by sylee96
Wrong exception handling when loading dataset from local disk
#173 opened Jun 8, 2024 by ganler
Question on "mlm" in continued pre-training
#172 opened Jun 3, 2024 by tanliboy
2
Question about sft with deepspeed
#170 opened May 27, 2024 by XXares
1
Cannot flatten integer dtype tensors
#169 opened May 17, 2024 by jaywongs
1
Released model weights for ablations of KTO/IPO/DPO cannot be found
#168 opened May 17, 2024 by ChenDRAG
CI failing due to mistralai/Mistral-7B-Instruct-v0.2 being gated now
#166 opened May 8, 2024 by alvarobartt
FSDP + QDoRA Support
#159 opened Apr 24, 2024 by iseesaw
6
How to work with local data
#158 opened Apr 18, 2024 by pretidav
1
Clarification on dataset mixer
#157 opened Apr 18, 2024 by deep-diver
5
Dependency updates for QLoRA+FSDP
#155 opened Apr 16, 2024 by deep-diver
CPT training is giving pretty unstalbe results with the learning rate 2e-5.
#149 opened Apr 4, 2024 by shamanez
1
Multi-GPU Training with DPO Full Parameter Stucks
#147 opened Mar 31, 2024 by Taishi-N324
Constitutional AI models do not achieve MT-Bench scores as reported
#145 opened Mar 27, 2024 by JingtongSu
Can we please add the option to work with a tokenized dataset, escpailly for the CPT task.
#144 opened Mar 27, 2024 by shamanez
Efficient dialog data format for KTO training
#142 opened Mar 26, 2024 by DavidFarago
Can any one share the script what params should be passed to run_dpo.py
#141 opened Mar 26, 2024 by Oscarjia
1
How to select parts to bp in sft
#138 opened Mar 19, 2024 by Fu-Dayuan
Is there a way to freeze some layers of a model ?
#134 opened Mar 12, 2024 by shamanez
Early Stopping Issue when used with ConstantLengthDataset
#133 opened Mar 8, 2024 by sankydesai
Not able to run Zephyr 7B Gemma with 4 80GB A100s
#132 opened Mar 7, 2024 by TJ-Solergibert
1
Downloading latest CUDA version (11.6 or above) for MacOS to use FlashAttention
#130 opened Mar 5, 2024 by shubhamcs162
Minor question about PAD token and EOS token.
#127 opened Feb 26, 2024 by HaniItani
2
ProTip! What’s not been updated in a month: updated:<2024-05-30.