Gary

size mismatch for encoder.embedding.weight exception when using miven model and NO Chinse voice but only noise

1

1)Use miven's model: https://pan.baidu.com/s/1PI-hM3sn5wbeChRryX-RCQ code: 2021 2)Got exception size mismatch for encoder.embedding.weight: copying a param with shape torch.Size([70, 512]) from checkpoint, the shape in current model is torch.Size([75, 512]). 3)NOT...

When using multi-thread to save data The cl.save() method generate same "_id" value, and cause DUPLICATE key error this is really a critical bug.

Python driver: Same issue in many methods like query(), query_one() etc The method usage/description says "selected" but in code, the key name is "selector"

Collection save() remove "_id" field from the original given data dict

1

**Problem:** The save() method remove the "_id" field from the original doc, and because in python the given parameter: doc is a kind of byref, so the original doc is...

setup.sh and unset.sh by default clean all the sequoiadb-3.4 files

2

Each time when run setup.sh or unset.sh, all the sequoiadb-3.4 files are cleaned. Shouldn't the files be kept? or give a option to user to decide whether remove those files...

as the title says, the unset.sh doesn't do a clean unset, the sdbadmin and sdbadmin_group are remained.

Step1: install sdb via following http://doc.sequoiadb.com/cn/sequoiadb-cat_id-1519612914-edition_id-304 Step2: run quickDeploy.sh on the single node, try twice and always get this error. sdbadmin@dbserver:/home/guang$ cd /opt/sequoiadb sdbadmin@dbserver:/opt/sequoiadb$ ./tools/deploy/quickDeploy.sh Execute command: /opt/sequoiadb/./tools/deploy/../../bin/sdb -f /opt/sequoiadb/./tools/deploy/quickDeploy.js...

tag 0.0.1 和最新版, 选择不同合成器模型,总会在不同的参数上出现 size mismatch, 导致语音合成失败

2

错误信息: size mismatch for encoder.embedding.weight: copying a param with shape torch.Size([70, 512]) from checkpoint, the shape in current model is torch.Size([75, 512]). 请问 1)这个是什么原因导致的呢? 是编码器和合成器不匹配? 2),这个有办法从代码层面调整么? 3)这个问题怎样解决呢?