-
python predict_downstream_condition.py --ckpt_path model_name_roberta-base_taskname_qqp_lr_3e-05_seed_42_numsteps_2000_sample_Categorical_schedule_mutual_hybridlambda_0.0003_wordfreqlambda_0.0_fromscr…
-
## Description
I was experimenting with the `sentence-transformers/msmarco-roberta-base-ance-firstp` model and observed some discrepancies between the outputs of the tokenizer depending on how the …
-
我按照README文件的知识, 然后将chinese-roberta-wwm-ext-large 预训练模型文件下载好了放在./chinese-roberta-wwm-ext-large 目录下,然后我运行了python3 main.py --no_pair --seed 1 --use_apex_amp --apex_amp_opt_level O1 --batch_size 32 --max_…
-
### 🐛 Describe the bug
When running Roberta Question Answering (and also other Huggingface models) in CPU inference mode, I get an extra output returned by dynamo, that did not happen in a previous…
-
作者您好,在您的论文中写道对于 Reuters-21578,我们使用 RoBERTa
作为编码器并截断超过 RoBERTa 最大长度的句子。但是在实验结果那里,您又写道在Reuters-21578数据集中,S-GCN与其编码器DOC-BERT相比,结果提高了0.55,表明S-GCN中的全局信息融合为MLTC任务带来了显著的提升。所以请问关于这个数据集您用的是哪一个编码器,
-
Hello,
Have been looking into this model from few days. Thought of training this model on a new data set so I took a sample dataset with 15 pairs of sentences including human gold score. Trained mo…
-
请问这个代码所需的文件是不全的吗?如./state_dict/文件下的roberta_wwm_vocab.txt等?
-
@JRosenkranz This looks amazing, learned about this lib at vllm yesterday. I am trying to run `bge-m3` using this custom modeling code for https://github.com/michaelfeil/infinity . I am aware that thi…
-
### Feature request
It would be amazing if 0-shot text classifiers that are designed to work with the [HF zeroshot pipeline](https://huggingface.co/docs/transformers/en/main_classes/pipelines#transfo…
-
### Describe the issue
Following the [issue 155](https://github.com/microsoft/LLMLingua/issues/155), I'm trying to reproduce the results of the official [llmlingua-2-xlm-roberta-large-meetingbank](…