world2vec

Hi, how to extend this model to edit image by text like 'add a horse in the image' and reuse the pretrained model weights?

implementation questions

2

Hi, Thanks for your good idea and paper, have some questions about TPLinkerPlus 1. I see you use logsum to calculate the loss, is it better than BCE? 2. Is...

Hi, This maybe the reason why you get worse result. You can refer the source code of transformers [BertPooler](https://github.com/huggingface/transformers/blob/1c191efc3abc391072ff0094a8108459bc08e3fa/src/transformers/models/bert/modeling_bert.py#L624)

trained weights for base

3

Good work. I only see weights for large, could you also provider weights for base? That will be much easy to handle for debugging. Thanks.

Env: torch2.4 cuda 12.4 unsloth main below is the code errored ``` from unsloth import FastLanguageModel import torch model_id="unsloth/gemma-2-2b-it-bnb-4bit" model, tokenizer = FastLanguageModel.from_pretrained(model_id, dtype=torch.float16, use_cache=False, max_seq_length=1024, load_in_4bit=True) FastLanguageModel.for_inference(model) input_text =...

What is llmxmapreduce? Any reference?

1

Congrats to your great work!

grpo training without network

1

I am working in a server only have local networking. I used fast_inference=True, load_in_4bit=True, max_lora_rank=32, and I have already download the model, but looks unsloth faill with below errors even...