Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
M
macaon
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Build
Pipelines
Jobs
Pipeline schedules
Artifacts
Deploy
Releases
Container Registry
Model registry
Operate
Environments
Monitor
Incidents
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
Franck Dary
macaon
Graph
12f80204b1964d919b3f987638664a25e06a0097
Select Git revision
Branches
3
loss
master
default
protected
producer
3 results
You can move around the graph by using the arrow keys.
Begin with the selected commit
Created with Raphaël 2.2.0
30
Jul
28
15
14
9
7
2
1
30
Jun
29
26
23
22
21
19
18
17
16
15
14
13
12
10
9
8
7
5
4
3
2
28
May
25
24
23
21
18
17
16
15
14
13
12
11
10
8
6
5
4
3
30
Apr
29
27
26
24
23
22
21
20
17
16
15
14
13
12
11
10
9
8
7
6
5
3
2
1
31
Mar
30
29
26
21
20
19
17
16
15
14
12
11
10
9
8
7
5
4
3
2
28
Feb
27
26
25
24
23
22
21
20
19
17
16
15
14
13
12
11
8
7
6
5
4
31
Jan
30
29
28
27
26
22
21
20
14
13
12
9
8
7
6
2
21
Dec
20
19
18
17
16
15
14
13
12
11
allow multiple pretrained embeddings file for ContextModule
removed word2vec from documentation
<unk> in pretrained embeddgins is used for Dict::unknownValueStr
Action EOS adds text metadata
Added Transformer MyModule
Moving nn to correct device
Making sure the nn are loaded to the correct device
Corrected bug
Explore different transitions only for parser
Sending torch object to CPU before saving them to disk
ADDCHARTOWORD can add multiple letters at once
Corrected bug in ContextualModule
Fixed dynamic oracle for tokenization transitions
Senting neural network to correct device
Corrected bug with multiple classifiers
Support for multiple classifiers
Corrected a bug where a tensor was not send to the correct device
Shuffling batches Tensor in ConfigDataset
different null value for each column in ContextModule
Added cnn
Added seed program argument
Batched indexing for ContextualModule
Adding None to dict in cotnextual module
Added Contextual module
Corrected bug on CustomHingeLoss where it would not run on gpu
In ContextModule : left-most and right-most child features are only taken into account if they are respectively to the left or to the right of their parents
Removed debug print
Added custom hinge loss and added explorationThreshold program argument
Added bce and mse losses
Added gold parsing transitions
Make sure nothing is added to the dict mid training
Removed cycle checking in Action::attach and fixed index problem in some transition cost
Only use dynamic oracle and may have corrected a problem
Corrected order of trainStrategy applications : removed possibility of parameter reset before dynamic example extracting
Defaulting staticCost to dynamicCost
w2v in ContextModule is now relative path
Added special case for EOS and ID in ContextModule
ContextModule : now possible to refer to childrens
Added prefix in distanceModule values to avoid them being treated as generic numbers bu the Dict
Added Adagrad optimizer
Loading