-
按照示例中的代码运行cross_encoder的训练,报 中断执行 的错误,显示 corrupted double linked list
-
Hi,Great job, I find that you release Chinese retrieval model trained on Dureader, Could you please also release your preprocess code or processed datasets.
-
您好!
我读了RocketQA的原文,发现RocketQA介绍的是关于dual-encoder的训练(即[RocketQA/research/DuReader-Retrieval-Baseline](https://github.com/PaddlePaddle/RocketQA/tree/main/research/DuReader-Retrieval-Baseline)中的step 1)。请…
-
Hi, I wonder what are the hyperparams of k1 and b set in your BM25 baseline?
-
您好,感谢提供实验代码!
我发现在单卡训练cross encoder的时候,结果与baseline的4*V100各项指标上都差不少,请问有原因吗?
运行环境 Tesla A100 80G,CUDA11.1 Cudnn8.0.5,paddle 2.2.0
提供模型infer的结果:
{"MRR@10": 0.7284, "QueriesRanked": 2000, "recall@1…
-
```
$ k get pods -w
...
datasets-server-prod-datasets-worker-776b774978-g7mpk 1/1 Evicted 0 73m │DEBUG: 2022-06-16 18:42:46,966 - dat…
-
版本
paddlehub 2.0.4
paddlenlp 2.0.1
paddlepaddle-gpu 2.1.0.post101
FGM代码:
import paddle
class FGM():
"""针对embedding层梯度上升干扰的对抗训练方法,Fast G…
-
export CUDA_VISIBLE_DEVICES=0
TRAIN_SET="dureader-retrieval-baseline-dataset/train/dual.train.tsv"
MODEL_PATH="pretrained-models/ernie_base_1.0_twin_CN/params"
sh script/run_dual_encoder_train.sh $…
-
## Introduciton
when loading ErnieTinyTokenizer instance `from_pretrained('ernie-tiny')`, and there is already resouce file in cache dir, eg: `special_tokens_map.json`, it will raise the following…
-
参考[如何在CPU上运行paddlenlp_pipelines_gpu?](https://github.com/PaddlePaddle/PaddleNLP/issues/3290),在CPU上跑起来了模型。但是系统运行时出现了问题,浏览器上显示“Connection Error. Is pipelines running? An error occurred during the reques…