Custom sentence2vec projection model

so far

model-step-1
query+doc(pos/neg) -> vec -> cos_sim

train a model for loss of cos_sim btw q-n, q-p with margin
device: cpu
epoch size: 3
input_dim or emb dim: 25
projection_dim: 128
margin=0.2
batch size: 32
neg doc size: 10 rows
overall training size: 82_325
time spent to run: 29m49s
Epoch 1, Loss: 0.0960
Epoch 2, Loss: 0.0559
Epoch 3, Loss: 0.0526

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support