wassname/openai-transformer-lm-gutenberg-erotic
1
0
Fork 0
mirror of https://github.com/wassname/openai-transformer-lm-gutenberg-erotic.git synced 2026-06-27 16:10:19 +08:00
Code Issues Packages Projects Releases Wiki Activity
83 Commits 1 Branch 0 Tags
master
Commit Graph

83 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
nottombrown 2439b3c971 Remove globals 2018-06-26 19:49:29 -07:00
nottombrown 9e5c14dd48 Refactor 2018-06-26 19:44:23 -07:00
nottombrown 1199c206b0 Cleanup 2018-06-26 19:31:46 -07:00
thomwolf b0cdf6bf60 Merge pull request #8 from rodgzilla/master 
Using negative index to reshape the input tensor.
 2018-06-26 09:11:57 +02:00
Grégory Châtel 71bcad84f9 Using negative index to reshape the input tensor. 2018-06-22 15:33:29 +02:00
thomwolf d74d00c8b5 Updated test accuracy to 85.84 with batch size 20 2018-06-19 22:20:09 +02:00
thomwolf 7b1d4775d9 small tweaks 2018-06-19 22:02:52 +02:00
thomwolf f2a1dcf22f updated test accuracy 2018-06-19 21:31:27 +02:00
thomwolf f413a81700 clean up tied weights 2018-06-19 17:55:20 +02:00
thomwolf cca20335bf update optimizer scope in training loop 2018-06-19 17:48:44 +02:00
thomwolf 7702cc141b Merge pull request #5 from cclauss/patch-1 
Create .travis.yml
 2018-06-19 17:45:01 +02:00
thomwolf a0c2b61fd2 Merge pull request #6 from soskek/patch-2 
fix the scope of optimizer
 2018-06-19 17:44:41 +02:00
Sosuke Kobayashi 39891aae3c fix the scope of optimizer 
The target of an optimizer should contain `clf_head` (new task-specific output matrix) in addition to `model` (Transformer encoder).
The code might fail to do that, right?
 2018-06-19 21:29:16 +09:00
cclauss a9055bd370 Create .travis.yml 
Output: https://travis-ci.org/huggingface/neuralcoref/jobs/393000348
 2018-06-16 09:32:00 +02:00
thomwolf e339361770 Merge pull request #4 from soskek/patch-1 
fix arguments for initializing a model
 2018-06-16 08:53:30 +02:00
Sosuke Kobayashi 3c485eb1d1 fix arguments for initializing a model 
In the sample ROCStory script, the line for model construction seems wrong. This will fix it.
 2018-06-16 08:26:08 +09:00
thomwolf 2ce292bfcb typos 2018-06-15 21:13:29 +02:00
thomwolf 3388d7e1d2 Update readme 2018-06-15 21:13:29 +02:00
Mark Neumann 5b9cbd7468 fixes to load the model easier 2018-06-15 21:13:29 +02:00
thomwolf aa800ef759 more details in readme 2018-06-14 17:29:36 +02:00
thomwolf d05519131b instruction on how to download openai weights 2018-06-14 17:27:19 +02:00
thomwolf 19cbffe6f8 updated readme 2018-06-14 17:24:17 +02:00
thomwolf 3d8d70937c fix model and training 2018-06-14 16:40:00 +02:00
thomwolf 6a20d66253 updated readme 2018-06-14 15:09:50 +02:00
thomwolf 8d44206590 updated readme 2018-06-14 15:03:06 +02:00
thomwolf 49d628f68a fixed modified Adam + added evaluation code 2018-06-14 14:29:53 +02:00
thomwolf 0704c84ab9 added openAIAdam optimizer 2018-06-14 13:19:07 +02:00
thomwolf 0b730579c5 model train run - no eval 2018-06-14 12:10:05 +02:00
thomwolf 89ab479aad refactoring 2018-06-14 03:50:41 +02:00
thomwolf 1f209c42b1 first version of model + weights transfer 2018-06-14 00:18:48 +02:00
thomwolf 5eb7937f07 loss computation 2018-06-13 16:48:44 +02:00
thomwolf 8858f99438 code 2018-06-13 16:07:58 +02:00
thomwolf 83202ae27e gitignore 2018-06-13 16:07:26 +02:00