jumelet commited on
Commit
0f28642
·
verified ·
1 Parent(s): 1031d53

BabyLM-community/eus-baseline-small

Browse files
Files changed (6) hide show
  1. README.md +15 -1
  2. merges.txt +9 -9
  3. model.safetensors +1 -1
  4. tokenizer.json +0 -0
  5. training_args.bin +1 -1
  6. vocab.json +0 -0
README.md CHANGED
@@ -13,6 +13,8 @@ should probably proofread and complete it, then remove this comment. -->
13
  # eus-baseline-small
14
 
15
  This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
 
 
16
 
17
  ## Model description
18
 
@@ -32,7 +34,7 @@ More information needed
32
 
33
  The following hyperparameters were used during training:
34
  - learning_rate: 0.0001
35
- - train_batch_size: 64
36
  - eval_batch_size: 8
37
  - seed: 42
38
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
@@ -41,6 +43,18 @@ The following hyperparameters were used during training:
41
 
42
  ### Training results
43
 
 
 
 
 
 
 
 
 
 
 
 
 
44
 
45
 
46
  ### Framework versions
 
13
  # eus-baseline-small
14
 
15
  This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
16
+ It achieves the following results on the evaluation set:
17
+ - Loss: 3.4351
18
 
19
  ## Model description
20
 
 
34
 
35
  The following hyperparameters were used during training:
36
  - learning_rate: 0.0001
37
+ - train_batch_size: 16
38
  - eval_batch_size: 8
39
  - seed: 42
40
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 
43
 
44
  ### Training results
45
 
46
+ | Training Loss | Epoch | Step | Validation Loss |
47
+ |:-------------:|:-----:|:-----:|:---------------:|
48
+ | 5.1695 | 1.0 | 2279 | 4.4543 |
49
+ | 4.2664 | 2.0 | 4558 | 4.0723 |
50
+ | 3.9875 | 3.0 | 6837 | 3.8665 |
51
+ | 3.8134 | 4.0 | 9116 | 3.7303 |
52
+ | 3.6885 | 5.0 | 11395 | 3.6272 |
53
+ | 3.5941 | 6.0 | 13674 | 3.5577 |
54
+ | 3.5221 | 7.0 | 15953 | 3.5073 |
55
+ | 3.4682 | 8.0 | 18232 | 3.4691 |
56
+ | 3.4281 | 9.0 | 20511 | 3.4452 |
57
+ | 3.3999 | 10.0 | 22790 | 3.4351 |
58
 
59
 
60
  ### Framework versions
merges.txt CHANGED
@@ -177,7 +177,6 @@ u a
177
  ar ra
178
  b er
179
  Ġ on
180
- Ġ Ċ
181
  al a
182
  Ġ U
183
  at ik
@@ -200,6 +199,7 @@ i ak
200
  Ġbeh ar
201
  Ġ os
202
  i m
 
203
  b il
204
  Ġ or
205
  Ġ O
@@ -285,6 +285,7 @@ i b
285
  l ea
286
  aaaaaaaa aaaaaaaa
287
  Ġz it
 
288
  eta ko
289
  T H
290
  Ġes p
@@ -641,6 +642,7 @@ uz ten
641
  un tza
642
  ak it
643
  Ġor ain
 
644
  a st
645
  ak um
646
  ar ik
@@ -1396,6 +1398,7 @@ B eg
1396
  Ġb orro
1397
  K aixo
1398
  U r
 
1399
  Ġ ul
1400
  Ġj ap
1401
  Ġe ki
@@ -1409,7 +1412,6 @@ g arria
1409
  Ġzeh ar
1410
  ĠKan po
1411
  Ġondor ioz
1412
- Ġ Ġ
1413
  Ġadi bidez
1414
  tu ko
1415
  Ġzait ut
@@ -2879,6 +2881,7 @@ n i
2879
  Ġtxirrindu lari
2880
  Ġliter atura
2881
  G o
 
2882
  Ġ197 9
2883
  Ġetx ean
2884
  u f
@@ -3291,6 +3294,7 @@ st an
3291
  Ġm ultz
3292
  Ġeh un
3293
  Ġfor ma
 
3294
  Ġ ;
3295
  itz ar
3296
  Ġpro g
@@ -3721,6 +3725,7 @@ itz aile
3721
  ĠSan t
3722
  Ġhel buru
3723
  ĠRo d
 
3724
  Ġs obietar
3725
  iko ki
3726
  Ġal tx
@@ -3737,7 +3742,6 @@ us s
3737
  Ġ196 3
3738
  Ġ190 7
3739
  Ġnaz ional
3740
- ĠĠ ĠĠ
3741
  Ġer ori
3742
  ĠD om
3743
  Ġne ur
@@ -5607,6 +5611,7 @@ ion alak
5607
  ĠV en
5608
  Ġburu j
5609
  Beh in
 
5610
  Ġzehark atzen
5611
  Ġd in
5612
  un dea
@@ -5926,7 +5931,6 @@ p ia
5926
  ĠA ra
5927
  ab es
5928
  Ġhar k
5929
- Âł Ġ
5930
  Ġip uin
5931
  Ġkom ik
5932
  Ja una
@@ -5985,6 +5989,7 @@ ron te
5985
  Ġpilot aria
5986
  D em
5987
  N ek
 
5988
  Ġg an
5989
  H o
5990
  Ġer or
@@ -7925,8 +7930,3 @@ gar et
7925
  gos lav
7926
  Ġpedago goa
7927
  P as
7928
- Ġe b
7929
- al etan
7930
- al ardo
7931
- is io
7932
- Ġi x
 
177
  ar ra
178
  b er
179
  Ġ on
 
180
  al a
181
  Ġ U
182
  at ik
 
199
  Ġbeh ar
200
  Ġ os
201
  i m
202
+ Ċ Ċ
203
  b il
204
  Ġ or
205
  Ġ O
 
285
  l ea
286
  aaaaaaaa aaaaaaaa
287
  Ġz it
288
+ Ġ Ċ
289
  eta ko
290
  T H
291
  Ġes p
 
642
  un tza
643
  ak it
644
  Ġor ain
645
+ Ġ ĊĊ
646
  a st
647
  ak um
648
  ar ik
 
1398
  Ġb orro
1399
  K aixo
1400
  U r
1401
+ Ġ Ġ
1402
  Ġ ul
1403
  Ġj ap
1404
  Ġe ki
 
1412
  Ġzeh ar
1413
  ĠKan po
1414
  Ġondor ioz
 
1415
  Ġadi bidez
1416
  tu ko
1417
  Ġzait ut
 
2881
  Ġtxirrindu lari
2882
  Ġliter atura
2883
  G o
2884
+ ĠĊ ĠĊ
2885
  Ġ197 9
2886
  Ġetx ean
2887
  u f
 
3294
  Ġm ultz
3295
  Ġeh un
3296
  Ġfor ma
3297
+ Ċ Ġ
3298
  Ġ ;
3299
  itz ar
3300
  Ġpro g
 
3725
  ĠSan t
3726
  Ġhel buru
3727
  ĠRo d
3728
+ ĠĠ ĠĠ
3729
  Ġs obietar
3730
  iko ki
3731
  Ġal tx
 
3742
  Ġ196 3
3743
  Ġ190 7
3744
  Ġnaz ional
 
3745
  Ġer ori
3746
  ĠD om
3747
  Ġne ur
 
5611
  ĠV en
5612
  Ġburu j
5613
  Beh in
5614
+ ĠĊĠĊ ĠĊĠĊ
5615
  Ġzehark atzen
5616
  Ġd in
5617
  un dea
 
5931
  ĠA ra
5932
  ab es
5933
  Ġhar k
 
5934
  Ġip uin
5935
  Ġkom ik
5936
  Ja una
 
5989
  Ġpilot aria
5990
  D em
5991
  N ek
5992
+ Ġ Âł
5993
  Ġg an
5994
  H o
5995
  Ġer or
 
7930
  gos lav
7931
  Ġpedago goa
7932
  P as
 
 
 
 
 
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8a62f8aa6fe1b9bc9f12fbd14181e8c2b4d31426710fb6fe9cb53817b1f7d45b
3
  size 68273200
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:344eaa52a9bc971ff469a3bcdbbc36de1c03089f8b1a14b49a803dde31cd4e62
3
  size 68273200
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff
 
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4b524d67415728a55d3b8c3e818f5d1175b709c382bb754f51b5678406c09fab
3
  size 5777
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:493d7d460f519c95de5f8a28f9ff1a73abbeaef2be1ffda276f61ac56c186e62
3
  size 5777
vocab.json CHANGED
The diff for this file is too large to render. See raw diff