Jeon Sangmin comments

Results 9 comments of


                                            Jeon Sangmin

Mergekit-Evolve with vLLM enabled causes error if merge_method is linear

They can raise errors whether base model is defined or not. I modified this code evo/actors.py line 202 ```python tok = transformers.AutoTokenizer.from_pretrained( self.genome.definition.base_model.model.path, use_fast=True ) ``` to ```python if self.genome.definition.base_model...

how to merge for different rope scaling?

I just changed the rope scailing of basemodel and model1 before merging but not working well it seems especially, I used merge-evolve. Is there a no consideration about [rope](https://github.com/vllm-project/vllm/pull/555) when...

TypeError: unsupported operand type(s) for /: 'Tensor' and 'NoneType' when full finetuning gemma3

@cslin0915 sorry but we are talking when `full_finetuning` is `True`

TypeError: unsupported operand type(s) for /: 'Tensor' and 'NoneType' when full finetuning gemma3

@BenBatsir yes, It worked out. I think the version matters. It was not working. `pip install unsloth` But It was working. `pip install unsloth==2025.3.18`

TypeError: unsupported operand type(s) for /: 'Tensor' and 'NoneType' when full finetuning gemma3

thank you

[Usage]: Llama4 tool parser

222

[BFCL] Introducing custom handlers

> Thanks for the PR @sangmandu ! Could you provide an example `BFCL_HANDLER_CONFIG`? @HuanzhiMao Wow, rapid reply! Thanks. It looks like our timing is a little off, and I've provided...

Volta Error: volta-shim Invoked Directly by Reflex on reflex run

the code @iCarossio said is here /venv/lib/python3.10/site-packages/reflex/utils/prerequisites.py:231:def get_package_manager(on_failure_return_none: bool = False) -> str | None: and it worked out for me. thx

multi-GPU training

I am sooooooo looking forward to 2.0. They said it comes out on end of April. - https://github.com/unslothai/unsloth/issues/2300