Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / tatsu-lab/alpaca_farm issues and pull requests
#94 - question for the dataset and training with rank-list data
Issue -
State: open - Opened by Hyfred 3 months ago
#94 - question for the dataset and training with rank-list data
Issue -
State: open - Opened by Hyfred 3 months ago
#93 - shape error: 32000 vs 32001
Issue -
State: closed - Opened by yiwan-rl 6 months ago
#91 - Inquiry Regarding Supervised Fine-Tuning with AlpacaFarm Framework for Pythia Models
Issue -
State: closed - Opened by hank0316 10 months ago
- 1 comment
#90 - Using pretrained models
Issue -
State: closed - Opened by syleedandekar 11 months ago
- 3 comments
#90 - Using pretrained models
Issue -
State: closed - Opened by syleedandekar 11 months ago
- 3 comments
#89 - Error downloading pre-trained weights
Issue -
State: closed - Opened by syleedandekar 11 months ago
#88 - [BREAKING] 003 depreciation: changing autoannotators to AlpacaEval 1.
Pull Request -
State: closed - Opened by YannDubs 12 months ago
#88 - [BREAKING] 003 depreciation: changing autoannotators to AlpacaEval 1.
Pull Request -
State: closed - Opened by YannDubs 12 months ago
#87 - Repeated Deprecation Error
Issue -
State: closed - Opened by syleedandekar 12 months ago
- 3 comments
#87 - Repeated Deprecation Error
Issue -
State: closed - Opened by syleedandekar 12 months ago
- 3 comments
#86 - PairwiseAutoAnnotator always "Annotating 0 examples with gpt4_3"
Issue -
State: closed - Opened by CharlesQ9 about 1 year ago
- 1 comment
#86 - PairwiseAutoAnnotator always "Annotating 0 examples with gpt4_3"
Issue -
State: closed - Opened by CharlesQ9 about 1 year ago
- 1 comment
#85 - BaseAnnotator.__init__() got an unexpected keyword argument 'other_keys_to_keep'
Issue -
State: closed - Opened by xukefaker about 1 year ago
- 4 comments
#85 - BaseAnnotator.__init__() got an unexpected keyword argument 'other_keys_to_keep'
Issue -
State: closed - Opened by xukefaker about 1 year ago
- 4 comments
#84 - chore: add commitlint based on husky
Pull Request -
State: closed - Opened by lxuechen about 1 year ago
#84 - chore: add commitlint based on husky
Pull Request -
State: closed - Opened by lxuechen about 1 year ago
#83 - Fix bug of DPO training script in README.
Pull Request -
State: closed - Opened by stceum about 1 year ago
#83 - Fix bug of DPO training script in README.
Pull Request -
State: closed - Opened by stceum about 1 year ago
#82 - fix: silly mistake
Pull Request -
State: closed - Opened by lxuechen about 1 year ago
#82 - fix: silly mistake
Pull Request -
State: closed - Opened by lxuechen about 1 year ago
#81 - feat: DPO
Pull Request -
State: closed - Opened by lxuechen about 1 year ago
- 1 comment
#81 - feat: DPO
Pull Request -
State: closed - Opened by lxuechen about 1 year ago
- 1 comment
#80 - RecursionError: maximum recursion depth exceeded
Issue -
State: open - Opened by qlduNLP over 1 year ago
#80 - RecursionError: maximum recursion depth exceeded
Issue -
State: open - Opened by qlduNLP over 1 year ago
#79 - score with reward model
Issue -
State: closed - Opened by qlduNLP over 1 year ago
- 2 comments
#79 - score with reward model
Issue -
State: closed - Opened by qlduNLP over 1 year ago
- 2 comments
#78 - integrity_check error with sft10k
Issue -
State: open - Opened by bpucla over 1 year ago
#78 - integrity_check error with sft10k
Issue -
State: open - Opened by bpucla over 1 year ago
#77 - lower cuda version?
Issue -
State: closed - Opened by dorothee-sigg over 1 year ago
- 1 comment
#77 - lower cuda version?
Issue -
State: closed - Opened by dorothee-sigg over 1 year ago
- 1 comment
#76 - Why use FSDP instead of Deepspeed?
Issue -
State: closed - Opened by nrailg over 1 year ago
- 2 comments
#76 - Why use FSDP instead of Deepspeed?
Issue -
State: closed - Opened by nrailg over 1 year ago
- 2 comments
#75 - RewardModel.from_pretrained() loads redundant weights (incurs extra ~30GB of RAM)
Issue -
State: open - Opened by angie-chen55 over 1 year ago
#74 - minor typo
Pull Request -
State: closed - Opened by YannDubs over 1 year ago
#74 - minor typo
Pull Request -
State: closed - Opened by YannDubs over 1 year ago
#73 - Huge memory demand of recover_model_weights.py?
Issue -
State: open - Opened by mensch72 over 1 year ago
#73 - Huge memory demand of recover_model_weights.py?
Issue -
State: open - Opened by mensch72 over 1 year ago
#72 - recover_model_weights.py gives WARNING:root:Your base LLaMA checkpoint is converted with transformers==4.27.0.dev0
Issue -
State: open - Opened by mensch72 over 1 year ago
#71 - Confusing detail preference mapping
Issue -
State: closed - Opened by dorothee-sigg over 1 year ago
- 1 comment
#71 - Confusing detail preference mapping
Issue -
State: closed - Opened by dorothee-sigg over 1 year ago
- 1 comment
#70 - KeyError: 'llama' in /recover_model_weights.py
Issue -
State: closed - Opened by mensch72 over 1 year ago
- 2 comments
#69 - Use with Llama-2-70b-hf?
Issue -
State: closed - Opened by mensch72 over 1 year ago
- 4 comments
#69 - Use with Llama-2-70b-hf?
Issue -
State: closed - Opened by mensch72 over 1 year ago
- 4 comments
#68 - Reproducibility of pretuned reward model
Issue -
State: closed - Opened by jp7c5 over 1 year ago
#67 - Question about KL term
Issue -
State: closed - Opened by SimengSun over 1 year ago
- 1 comment
#67 - Question about KL term
Issue -
State: closed - Opened by SimengSun over 1 year ago
- 1 comment
#66 - [Fix] Import Error
Pull Request -
State: closed - Opened by SingL3 over 1 year ago
- 3 comments
#66 - [Fix] Import Error
Pull Request -
State: closed - Opened by SingL3 over 1 year ago
- 3 comments
#65 - [Bug] Error importing is_deepspeed_zero3_enabled
Issue -
State: closed - Opened by SingL3 over 1 year ago
- 1 comment
#65 - [Bug] Error importing is_deepspeed_zero3_enabled
Issue -
State: closed - Opened by SingL3 over 1 year ago
- 1 comment
#64 - Problem with PairwiseAutoAnnotator
Issue -
State: closed - Opened by langhaobeijing over 1 year ago
- 4 comments
#64 - Problem with PairwiseAutoAnnotator
Issue -
State: closed - Opened by langhaobeijing over 1 year ago
- 4 comments
#63 - tried to use bnb, QLora on SFT but have errors
Issue -
State: closed - Opened by qiuruiyu over 1 year ago
- 3 comments
#62 - [BUG] padding with input / noinput
Pull Request -
State: closed - Opened by YannDubs over 1 year ago
#62 - [BUG] padding with input / noinput
Pull Request -
State: closed - Opened by YannDubs over 1 year ago
#61 - Problem with Simulation case
Issue -
State: closed - Opened by qiuruiyu over 1 year ago
- 2 comments
#60 - model selection of PPO in Table 2
Issue -
State: closed - Opened by langhaobeijing over 1 year ago
- 1 comment
#59 - log bug fix in recover_model_weights.py
Pull Request -
State: closed - Opened by rtaori over 1 year ago
- 1 comment
#59 - log bug fix in recover_model_weights.py
Pull Request -
State: closed - Opened by rtaori over 1 year ago
- 1 comment
#58 - fix zero grad bug. resolves #57
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#58 - fix zero grad bug. resolves #57
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#57 - Possible issue with gradient accumulation
Issue -
State: closed - Opened by rosinality over 1 year ago
- 3 comments
#56 - [Discussion] about compute_logprobs
Issue -
State: closed - Opened by snowkcon over 1 year ago
#56 - [Discussion] about compute_logprobs
Issue -
State: closed - Opened by snowkcon over 1 year ago
#55 - [Discussion] Adding more diverse annotators representing subpopulations?
Issue -
State: closed - Opened by mensch72 over 1 year ago
- 1 comment
#55 - [Discussion] Adding more diverse annotators representing subpopulations?
Issue -
State: closed - Opened by mensch72 over 1 year ago
- 1 comment
#54 - Alternatively use Claude for annotation?
Issue -
State: closed - Opened by mensch72 over 1 year ago
- 4 comments
#53 - Update chatml_b5_diana_with_inputs.txt
Pull Request -
State: closed - Opened by mensch72 over 1 year ago
- 1 comment
#52 - code question for compute_loss in ppo_trainer
Issue -
State: closed - Opened by Yuhuajoe over 1 year ago
- 4 comments
#52 - code question for compute_loss in ppo_trainer
Issue -
State: closed - Opened by Yuhuajoe over 1 year ago
- 4 comments
#51 - [tokenization] preprocessing inputs and labels
Issue -
State: closed - Opened by ArashAhmadian over 1 year ago
- 2 comments
#51 - [tokenization] preprocessing inputs and labels
Issue -
State: closed - Opened by ArashAhmadian over 1 year ago
- 2 comments
#50 - Running PPO with fewer GPUs
Issue -
State: closed - Opened by shunzh over 1 year ago
- 3 comments
#49 - [ENH] migrate annotator to alpaca eval
Pull Request -
State: closed - Opened by YannDubs over 1 year ago
- 1 comment
#49 - [ENH] migrate annotator to alpaca eval
Pull Request -
State: closed - Opened by YannDubs over 1 year ago
- 1 comment
#48 - Generation Issue (probability tensor contains either `inf`, `nan` or element < 0) of Flash-LLaMA with Model Parallelism
Issue -
State: closed - Opened by Zhiyuan-Zeng over 1 year ago
- 6 comments
#48 - Generation Issue (probability tensor contains either `inf`, `nan` or element < 0) of Flash-LLaMA with Model Parallelism
Issue -
State: closed - Opened by Zhiyuan-Zeng over 1 year ago
- 6 comments
#47 - fix import; resolves #46
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#46 - cannot import name 'PairwiseAutoAnnotator' from partially initialized module 'alpaca_farm.auto_annotations' (most likely due to a circular import)
Issue -
State: closed - Opened by simplelifetime over 1 year ago
- 1 comment
#46 - cannot import name 'PairwiseAutoAnnotator' from partially initialized module 'alpaca_farm.auto_annotations' (most likely due to a circular import)
Issue -
State: closed - Opened by simplelifetime over 1 year ago
- 1 comment
#45 - Differences in results between the code and the leaderboard
Issue -
State: closed - Opened by yifan123 over 1 year ago
- 1 comment
#45 - Differences in results between the code and the leaderboard
Issue -
State: closed - Opened by yifan123 over 1 year ago
- 1 comment
#44 - remove side effect by copying dictionary
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#44 - remove side effect by copying dictionary
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#43 - clean up Quark code
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#43 - clean up Quark code
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#42 - improve doc regarding based checkpoint transformers version for checkpoint recovery; resolves #41
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#41 - Use of decapoda-research/llama-7b-hf checkpoint for the LLaMa-7B
Issue -
State: closed - Opened by twidddj over 1 year ago
- 10 comments
#40 - fix bad style
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#39 - expert iteration documentation
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#38 - remove recursion logic; resolves #34
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#38 - remove recursion logic; resolves #34
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#37 - fix missing configs; resolves #30
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#37 - fix missing configs; resolves #30
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#36 - clarify model names in auto-eval leaderboard; resolves #32
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#35 - fix bug in regex pattern that causes failure of parsing.
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#35 - fix bug in regex pattern that causes failure of parsing.
Pull Request -
State: closed - Opened by lxuechen over 1 year ago
#34 - `_name_or_path` is not stored in llama config.json any more
Issue -
State: closed - Opened by nrailg over 1 year ago
- 2 comments
#34 - `_name_or_path` is not stored in llama config.json any more
Issue -
State: closed - Opened by nrailg over 1 year ago
- 2 comments