Joosung Yoon

icon indicating a website link eagle705.github.io

icon company NVIDIA icon location Seoul icon location Machine Learning Engineer (NLP, LLMs and Search)

Results 16 comments of Joosung Yoon

Support React Native

@hans00 thank you for your assistance in addressing the issue :) I installed dependencies for RN and then error changed. the error is about "there is no such file" (tokenizer.json,...

Support React Native

@xenova @hans00 is there any plan to merge this branch? I want to use official transformers js in RN. 😀

CUDA out of memory Error when run trainer.hyperparameter_search()

I have the same issue

when "write_predictions_to_file" is true，generate will fail。

I encountered the same issue within 24.05 image

cannot load model

I also failed to load model in simulator

[Feature Request] Qwen3VL GRPO, SFT training

Current issues about Qwen3-VL-30B-A3B (run_vlm_grpo.py) - vllm version: https://github.com/vllm-project/vllm/issues/19793 - log ``` (VllmGenerationWorker pid=3212593, ip=10.52.54.131) File "/opt/ray_venvs/nemo_rl.models.generation.vllm.vllm_worker.VllmGenerationWorker/lib/python3.12/site-packages/transformers/models/auto/configuration_auto.py", line 1273, in from_pretrained (VllmGenerationWorker pid=3212593, ip=10.52.54.131) raise ValueError( (VllmGenerationWorker pid=3212593, ip=10.52.54.131) ValueError:...