-
Colbert的思路是将每个token存成embedding,做late interaction,问题是消耗过多存储
Dense embedding则是过早的interaction,导致查询精度下降。
有没有可能做Sentence level的colbert,单个chunk可能会有10-100个sentence embedding,然后query用token level embeddi…
-
### Issue
Not able to change distance model when creating a collection with FastEmbed.
### Minimal steps to reproduce
```
from qdrant_client import QdrantClient
from qdrant_client.models import…
-
After reading the [BERT, Pre-training of Deep Bidirectional Transformers fo r Language Understanding](https://arxiv.org/pdf/1810.04805.pdf) paper, I had a fundamental question want to figure out.
B…
-
HI
Can i pass token position from where i want to extract embedding ?
instead of [CLS] if i want to extract from other special token , how can i do it in model.encode
-
Using version `2.2.2`. Several places in the docs show computing similarity like this (example here https://www.sbert.net/index.html)
```python
# 3. Calculate the embedding similarities
similarit…
-
### Initial Checks
- [ ] I have searched GitHub for a duplicate issue and I'm sure this is something new
- [ ] I have read and followed [the docs & demos](https://github.com/modelscope/modelscope-age…
-
Hi. In the tfhub of elmo (https://tfhub.dev/google/elmo/2) there is an output like you provide:
_elmo: the weighted sum of the 3 layers, where the weights are trainable. This tensor has shape [batch_…
-
Hi @nreimers,
I have one doubt about the behaviour of performing a paragraph embedding composed as the sum/mean of the embeddings created by `distiluse-multilingual` of the sentences in the paragr…
-
A Structured Self-attentive Sentence Embedding
Zhouhan Lin, Minwei Feng, Cicero Nogueira dos Santos, Mo Yu, Bing Xiang, Bowen Zhou, Yoshua Bengio, ICLR 2017
https://arxiv.org/abs/1703.03130
![se](h…
-
If I use `SoftmaxLoss` in NLI example as shown in https://github.com/UKPLab/sentence-transformers/blob/master/examples/training/nli/training_nli.py , during training the base model `Transformer+poolin…