Gunsoo Han
Gunsoo Han
안녕하세요 좋은 자료 공유해주셔서 감사의 말씀을 우선 정합니다. Inference.py 에서 ``` convert_keys = {} for k, v in checkpoint['model_state_dict'].items(): new_key_name = k.replace("module.", '') if new_key_name not in model_dict: print("{} is...
안녕하세요 Inference 단계에서 self.crf.decode( )를 통해서 나오는 tag_seq 에 대한 확률값/vieterbi_score 를 얻을 수는 없나요 ? 감사합니다
When I evaluate the blenderbot2_3B model with following command, everything works fine ```bash parlai eval_model -mf zoo:blenderbot2/blenderbot2_3B/model -t msc -v --knowledge-access-method none --rag_retriever_type dpr --log_every_n_secs 60 --batchsize 32 ``` However,...
I was trying to reproduce ppl metric given [here](https://github.com/facebookresearch/ParlAI/blob/main/parlai/zoo/blenderbot2/model_card.md#metrics-used-and-evaluation-results) for blenderbot2.0 but could not find a relevant script. I run the following command ```bash parlai eval_model -mf zoo:blenderbot2/blenderbot2_400M/model -t msc...
Hi Thanks for your work I was wondering if we can compute the probability for the optimal sequence Thanks
First of all, thank you for sharing great work ! I was wondering how would you recommend choosing optimal hyperparams for large batch size ? For example, if i train...
I would like to pre-train r2c2-base-400M/3B from scratch with reddit+robert+cc100en datasets on ParlAI. Is there any way / script to do that ?
On your paper, you mentioned "The encoder is pretrained BertSum" and im not entirely sure on the meaning of "pretrained BertSum". For an input document, you modify BERT as follows:...
I would be great if FlexGen supports Bloom models from huggingface !
Once you collect 52k synthetic dataset, how did you plot the pie chart [here](https://github.com/tatsu-lab/stanford_alpaca/blob/main/assets/parse_analysis.png) ? Thanks !