Commits · flax-community/roberta-base-thai

Adding `safetensors` variant of this model (#1)

ee83601
verified

yhavinga

SFconvertbot commited on 8 days ago

update tfevent

3d400e8

sakares commited on Jul 17, 2021

update event

596725d

sakares commited on Jul 17, 2021

fix run.sh

6665f93

sakares commited on Jul 17, 2021

new event

6e833e6

sakares commited on Jul 17, 2021

use proper tokenizer.json and change to batch_size=256

de940b9

sakares commited on Jul 16, 2021

train with 10K batch_size

89c14b0

sakares commited on Jul 15, 2021

remove old tokenizer.json. it's not trained properly

a615d86

sakares commited on Jul 15, 2021

change config and fix training tokenizer script

13d02d7

sakares commited on Jul 15, 2021

remove tfevents

70b401c

sakares commited on Jul 15, 2021

rename previous tfevents

d12a293

sakares commited on Jul 15, 2021

add 4th tfevents

8dc7108

sakares commited on Jul 15, 2021

increaaase bs=128 and change dtype to fp32

2071335

sakares commited on Jul 15, 2021

remove old model

8887f4f

sakares commited on Jul 15, 2021

change to proper vocab_size

9fa8adc

sakares commited on Jul 15, 2021

update pt model

d6a778b

sakares commited on Jul 15, 2021

add pt model from 15K trained steps

1aa50f9

sakares commited on Jul 15, 2021

update rule

5a8de81

sakares commited on Jul 15, 2021

update evnt

24b543a

sakares commited on Jul 15, 2021

update event

578dba5

sakares commited on Jul 15, 2021

update event

2c966a4

sakares commited on Jul 14, 2021

Saving weights and logs of step 15000

8664a01

sakares commited on Jul 14, 2021

check event

d74794e

sakares commited on Jul 14, 2021

Saving weights and logs of step 10000

908d1a5

sakares commited on Jul 13, 2021

update event

be4ec62

sakares commited on Jul 13, 2021

update tfevent

e60615a

sakares commited on Jul 13, 2021

Saving weights and logs of step 5000

acba88d

sakares commited on Jul 12, 2021

update event

14c6643

sakares commited on Jul 12, 2021

Create Thai Tokenizer script

5ad5b72

sakares commited on Jul 12, 2021

check event

e76044d

sakares commited on Jul 12, 2021

init tfevents for experiment#3

bbeb147

sakares commited on Jul 12, 2021

remove old exp&model and lower steps to save

64934c7

sakares commited on Jul 12, 2021

fix script

9c0b789

sakares commited on Jul 12, 2021

make it resumable training

51e884f

sakares commited on Jul 12, 2021

Saving weights and logs of step 10000

3040f73

sakares commited on Jul 12, 2021

update event

a9d2138

sakares commited on Jul 11, 2021

update tfevent

f8a3350

sakares commited on Jul 11, 2021

update event

bae02ca

sakares commited on Jul 10, 2021

remove old event

23b10e3

sakares commited on Jul 10, 2021

add experiment_2 tfevents

e0c32ef

sakares commited on Jul 10, 2021

update script

8976790

sakares commited on Jul 10, 2021

update script regarding to https://github.com/huggingface/transformers/pull/12608

6aba5c0

sakares commited on Jul 10, 2021

lower lr regarding explode gradient. also new seed number

b0729e9

sakares commited on Jul 10, 2021

Merge branch 'main' of https://huggingface.co./flax-community/roberta-base-thai into main

bfd5d73

sakares commited on Jul 10, 2021

update tfevents

0d2b32a

sakares commited on Jul 10, 2021

Update .gitattributes

1e02891

patrickvonplaten commited on Jul 9, 2021

update event

d9006b0

sakares commited on Jul 9, 2021

update event

33260f5

sakares commited on Jul 9, 2021

update event

0b153f6

sakares commited on Jul 8, 2021

update tfevents

e0cfb09

sakares commited on Jul 8, 2021

Commit History

Adding `safetensors` variant of this model (#1) ee83601 verified

update tfevent 3d400e8

update event 596725d

fix run.sh 6665f93

new event 6e833e6

use proper tokenizer.json and change to batch_size=256 de940b9

train with 10K batch_size 89c14b0

remove old tokenizer.json. it's not trained properly a615d86

change config and fix training tokenizer script 13d02d7

remove tfevents 70b401c

rename previous tfevents d12a293

add 4th tfevents 8dc7108

increaaase bs=128 and change dtype to fp32 2071335

remove old model 8887f4f

change to proper vocab_size 9fa8adc

update pt model d6a778b

add pt model from 15K trained steps 1aa50f9

update rule 5a8de81

update evnt 24b543a

update event 578dba5

update event 2c966a4

Saving weights and logs of step 15000 8664a01

check event d74794e

Saving weights and logs of step 10000 908d1a5

update event be4ec62

update tfevent e60615a

Saving weights and logs of step 5000 acba88d

update event 14c6643

Create Thai Tokenizer script 5ad5b72

check event e76044d

init tfevents for experiment#3 bbeb147

remove old exp&model and lower steps to save 64934c7

fix script 9c0b789

make it resumable training 51e884f

Saving weights and logs of step 10000 3040f73

update event a9d2138

update tfevent f8a3350

update event bae02ca

remove old event 23b10e3

add experiment_2 tfevents e0c32ef

update script 8976790

update script regarding to https://github.com/huggingface/transformers/pull/12608 6aba5c0

lower lr regarding explode gradient. also new seed number b0729e9

Merge branch 'main' of https://huggingface.co./flax-community/roberta-base-thai into main bfd5d73

update tfevents 0d2b32a

Update .gitattributes 1e02891

update event d9006b0

update event 33260f5

update event 0b153f6

update tfevents e0cfb09

Adding `safetensors` variant of this model (#1)

ee83601
verified

update tfevent

3d400e8

update event

596725d

fix run.sh

6665f93

new event

6e833e6

use proper tokenizer.json and change to batch_size=256

de940b9

train with 10K batch_size

89c14b0

remove old tokenizer.json. it's not trained properly

a615d86

change config and fix training tokenizer script

13d02d7

remove tfevents

70b401c

rename previous tfevents

d12a293

add 4th tfevents

8dc7108

increaaase bs=128 and change dtype to fp32

2071335

remove old model

8887f4f

change to proper vocab_size

9fa8adc

update pt model

d6a778b

add pt model from 15K trained steps

1aa50f9

update rule

5a8de81

update evnt

24b543a

update event

578dba5

update event

2c966a4

Saving weights and logs of step 15000

8664a01

check event

d74794e

Saving weights and logs of step 10000

908d1a5

update event

be4ec62

update tfevent

e60615a

Saving weights and logs of step 5000

acba88d

update event

14c6643

Create Thai Tokenizer script

5ad5b72

check event

e76044d

init tfevents for experiment#3

bbeb147

remove old exp&model and lower steps to save

64934c7

fix script

9c0b789

make it resumable training

51e884f

Saving weights and logs of step 10000

3040f73

update event

a9d2138

update tfevent

f8a3350

update event

bae02ca

remove old event

23b10e3

add experiment_2 tfevents

e0c32ef

update script

8976790

update script regarding to https://github.com/huggingface/transformers/pull/12608

6aba5c0

lower lr regarding explode gradient. also new seed number

b0729e9

Merge branch 'main' of https://huggingface.co./flax-community/roberta-base-thai into main

bfd5d73

update tfevents

0d2b32a

Update .gitattributes

1e02891

update event

d9006b0

update event

33260f5

update event

0b153f6

update tfevents

e0cfb09