mirror of
https://github.com/wassname/torch-neuralpointprocess.git
synced 2026-06-27 16:47:35 +08:00
doc
This commit is contained in:
@@ -1,12 +1,14 @@
|
|||||||
Modified by wassname from the below:
|
Modified by wassname from the below:
|
||||||
|
|
||||||
Changes:
|
Changes:
|
||||||
- [ ] clamp weight with epsilon for stablity
|
- [x] clamp weight with epsilon for stablity
|
||||||
- from `p.data *= (p.data>=0)`
|
- from `p.data *= (p.data>=0)`
|
||||||
- to `p.data = torch.clamp(p.data, min=eps)`
|
- to `p.data = torch.clamp(p.data, min=eps)`
|
||||||
- [ ] try log t
|
- [ ] try log t
|
||||||
- [ ] try not mean as much in intensity layer
|
- [x] try not mean as much in intensity layer
|
||||||
- [ ] use pytorch lightning
|
- [x] use pytorch lightning
|
||||||
|
- [x] fix potential data leak where label is fed to lsm
|
||||||
|
- [x] commit notebook with plots of val
|
||||||
|
|
||||||
|
|
||||||
# Fully Neural Network based Model for General Temporal Point Process(Neurips 2019,Takahiro Omi)
|
# Fully Neural Network based Model for General Temporal Point Process(Neurips 2019,Takahiro Omi)
|
||||||
|
|||||||
Reference in New Issue
Block a user