flowersteam/lamorel
Lamorel is a Python library designed for RL practitioners eager to use Large Language Models (LLMs).
PythonMIT
Issues
- 2
- 3
Runtime Error with PPO Finetuning on Single GPU: ChildFailedError and TCP Connection Closed
#42 opened by oceank - 0
Stability issues in PPO examples
#40 opened by ClementRomac - 1
- 0
Missing log_softmax in score
#37 opened by ClementRomac - 0
- 3
Finetuned Weights Loading Error
#32 opened by AiBo123456 - 2
Expand to multi-agent scenarios.
#28 opened by ewanlee - 2
Using API in lamorel
#30 opened by nuomizai - 2
Connection closed by peer [127.0.1.1]: 14734
#27 opened by ewanlee - 6
Device 0 is not recognized
#24 opened by giobin - 13
Connection error
#23 opened by yone456 - 2
- 2
why should we have two configs?
#18 opened by HCHCXY - 0
Remove the need of custom Accelerate version for single machine with single GPU
#20 opened by ClementRomac - 6
- 3
AssertionError: torch distributed must be used!
#22 opened by Jugg1er - 3
A syntax error in __call_model
#19 opened by Clayfigure - 1
- 4
- 0
Pre-encoding inputs crashes
#10 opened by ClementRomac - 0
Using an encoder-decoder LLM with `pre_encode_inputs: true` doesn't work when `model_parallelism_size` > 1
#8 opened by ClementRomac - 1
Inefficient PPO example
#5 opened by ClementRomac - 0