eduagarcia commited on
Commit
5a9474f
1 Parent(s): acd5eb0

Uploading raw results for mlabonne/Llama-3-8B-Instruct-abliterated-dpomix

Browse files
Files changed (12) hide show
  1. .gitattributes +9 -0
  2. mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_assin2_rte.jsonl +3 -0
  3. mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_assin2_sts.jsonl +3 -0
  4. mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_bluex.jsonl +3 -0
  5. mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_enem_challenge.jsonl +3 -0
  6. mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_faquad_nli.jsonl +3 -0
  7. mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_hatebr_offensive.jsonl +3 -0
  8. mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_oab_exams.jsonl +3 -0
  9. mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_portuguese_hate_speech.jsonl +3 -0
  10. mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_tweetsentbr.jsonl +3 -0
  11. mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/results.json +1244 -0
  12. mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/results_2024-05-27T14-44-42.467656.json +325 -0
.gitattributes CHANGED
@@ -5552,3 +5552,12 @@ mlabonne/Daredevil-8B-abliterated/raw_2024-05-27T13-40-17.084359/pretrained__mla
5552
  mlabonne/Daredevil-8B-abliterated/raw_2024-05-27T13-40-17.084359/pretrained__mlabonne__Daredevil-8B-abliterated,dtype__bfloat16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_oab_exams.jsonl filter=lfs diff=lfs merge=lfs -text
5553
  mlabonne/Daredevil-8B-abliterated/raw_2024-05-27T13-40-17.084359/pretrained__mlabonne__Daredevil-8B-abliterated,dtype__bfloat16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_portuguese_hate_speech.jsonl filter=lfs diff=lfs merge=lfs -text
5554
  mlabonne/Daredevil-8B-abliterated/raw_2024-05-27T13-40-17.084359/pretrained__mlabonne__Daredevil-8B-abliterated,dtype__bfloat16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_tweetsentbr.jsonl filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
5552
  mlabonne/Daredevil-8B-abliterated/raw_2024-05-27T13-40-17.084359/pretrained__mlabonne__Daredevil-8B-abliterated,dtype__bfloat16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_oab_exams.jsonl filter=lfs diff=lfs merge=lfs -text
5553
  mlabonne/Daredevil-8B-abliterated/raw_2024-05-27T13-40-17.084359/pretrained__mlabonne__Daredevil-8B-abliterated,dtype__bfloat16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_portuguese_hate_speech.jsonl filter=lfs diff=lfs merge=lfs -text
5554
  mlabonne/Daredevil-8B-abliterated/raw_2024-05-27T13-40-17.084359/pretrained__mlabonne__Daredevil-8B-abliterated,dtype__bfloat16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_tweetsentbr.jsonl filter=lfs diff=lfs merge=lfs -text
5555
+ mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_assin2_rte.jsonl filter=lfs diff=lfs merge=lfs -text
5556
+ mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_assin2_sts.jsonl filter=lfs diff=lfs merge=lfs -text
5557
+ mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_bluex.jsonl filter=lfs diff=lfs merge=lfs -text
5558
+ mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_enem_challenge.jsonl filter=lfs diff=lfs merge=lfs -text
5559
+ mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_faquad_nli.jsonl filter=lfs diff=lfs merge=lfs -text
5560
+ mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_hatebr_offensive.jsonl filter=lfs diff=lfs merge=lfs -text
5561
+ mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_oab_exams.jsonl filter=lfs diff=lfs merge=lfs -text
5562
+ mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_portuguese_hate_speech.jsonl filter=lfs diff=lfs merge=lfs -text
5563
+ mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_tweetsentbr.jsonl filter=lfs diff=lfs merge=lfs -text
mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_assin2_rte.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be9a2d0378939e8f1a035cca14906f846fd48a5d384c40d73d553e68a13983e9
3
+ size 40560865
mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_assin2_sts.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:77c99b661eb12d4164e270dabfa293d8572f218a997124dc9a17b14f31407f8b
3
+ size 41972541
mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_bluex.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:29c3365dda77107c0156781f334e5ee33ec0db2c06d6f48b20c8964f1bf40f5f
3
+ size 12399069
mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_enem_challenge.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d1413f0f90aad87aa799dbaeefdc77d0f31188f0e277fb0127b97b12b2c031e
3
+ size 24979159
mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_faquad_nli.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d94c08330a6f6514fe6baf7ed03c4894dfc83c89a01cb37d70a513b8fa29b62
3
+ size 12718163
mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_hatebr_offensive.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93f5dcf2e8917f1f6a5f6074946bc177e8f230fc0bcec28fd9f533e50c0f6cbb
3
+ size 23165301
mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_oab_exams.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b8ef5b481e4b45031b252a9fadac6d81a7b98f9e1de63a2f0bd44a295b61f8d6
3
+ size 35105103
mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_portuguese_hate_speech.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5642ca5626059a98311a5dee893466fa14843644bb96226c97d594be6a45f9e2
3
+ size 17256805
mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/pretrained__mlabonne__Llama-3-8B-Instruct-abliterated-dpomix,dtype__float16,device__cuda:0,revision__main,trust_remote_code__True,starting_max_length__2560_tweetsentbr.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2b9c7e1b88a35a3d177775b8fe0ad1ed056e8f300d41940956465fd9919f9a81
3
+ size 38777097
mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/raw_2024-05-27T14-44-42.467656/results.json ADDED
@@ -0,0 +1,1244 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "results": {
3
+ "assin2_rte": {
4
+ "f1_macro,all": 0.9136602188080808,
5
+ "acc,all": 0.9138071895424836,
6
+ "alias": "assin2_rte"
7
+ },
8
+ "assin2_sts": {
9
+ "pearson,all": 0.7676336804192703,
10
+ "mse,all": 0.6167238562091503,
11
+ "alias": "assin2_sts"
12
+ },
13
+ "bluex": {
14
+ "acc,all": 0.5730180806675939,
15
+ "acc,exam_id__UNICAMP_2018": 0.5,
16
+ "acc,exam_id__UNICAMP_2023": 0.627906976744186,
17
+ "acc,exam_id__USP_2023": 0.6818181818181818,
18
+ "acc,exam_id__UNICAMP_2024": 0.6666666666666666,
19
+ "acc,exam_id__USP_2024": 0.6341463414634146,
20
+ "acc,exam_id__UNICAMP_2021_1": 0.5217391304347826,
21
+ "acc,exam_id__USP_2020": 0.5714285714285714,
22
+ "acc,exam_id__UNICAMP_2020": 0.5818181818181818,
23
+ "acc,exam_id__UNICAMP_2022": 0.5641025641025641,
24
+ "acc,exam_id__UNICAMP_2019": 0.54,
25
+ "acc,exam_id__UNICAMP_2021_2": 0.5686274509803921,
26
+ "acc,exam_id__USP_2018": 0.46296296296296297,
27
+ "acc,exam_id__USP_2021": 0.5961538461538461,
28
+ "acc,exam_id__USP_2019": 0.55,
29
+ "acc,exam_id__USP_2022": 0.5714285714285714,
30
+ "alias": "bluex"
31
+ },
32
+ "enem_challenge": {
33
+ "alias": "enem",
34
+ "acc,all": 0.7053883834849545,
35
+ "acc,exam_id__2011": 0.717948717948718,
36
+ "acc,exam_id__2017": 0.6637931034482759,
37
+ "acc,exam_id__2015": 0.7310924369747899,
38
+ "acc,exam_id__2016": 0.7024793388429752,
39
+ "acc,exam_id__2016_2": 0.6585365853658537,
40
+ "acc,exam_id__2009": 0.7130434782608696,
41
+ "acc,exam_id__2012": 0.7241379310344828,
42
+ "acc,exam_id__2010": 0.7435897435897436,
43
+ "acc,exam_id__2013": 0.6944444444444444,
44
+ "acc,exam_id__2014": 0.6972477064220184,
45
+ "acc,exam_id__2022": 0.6616541353383458,
46
+ "acc,exam_id__2023": 0.7555555555555555
47
+ },
48
+ "faquad_nli": {
49
+ "f1_macro,all": 0.7143293906330614,
50
+ "acc,all": 0.7523076923076923,
51
+ "alias": "faquad_nli"
52
+ },
53
+ "hatebr_offensive": {
54
+ "alias": "hatebr_offensive_binary",
55
+ "f1_macro,all": 0.8645687620632841,
56
+ "acc,all": 0.865
57
+ },
58
+ "oab_exams": {
59
+ "acc,all": 0.5097949886104783,
60
+ "acc,exam_id__2011-03": 0.42424242424242425,
61
+ "acc,exam_id__2014-13": 0.45,
62
+ "acc,exam_id__2013-10": 0.45,
63
+ "acc,exam_id__2017-24": 0.4375,
64
+ "acc,exam_id__2017-22": 0.6,
65
+ "acc,exam_id__2012-06a": 0.55,
66
+ "acc,exam_id__2016-20a": 0.425,
67
+ "acc,exam_id__2012-09": 0.45454545454545453,
68
+ "acc,exam_id__2015-16": 0.5,
69
+ "acc,exam_id__2011-04": 0.4625,
70
+ "acc,exam_id__2012-07": 0.5125,
71
+ "acc,exam_id__2014-14": 0.575,
72
+ "acc,exam_id__2014-15": 0.5641025641025641,
73
+ "acc,exam_id__2010-02": 0.57,
74
+ "acc,exam_id__2015-18": 0.525,
75
+ "acc,exam_id__2016-19": 0.5512820512820513,
76
+ "acc,exam_id__2012-06": 0.5875,
77
+ "acc,exam_id__2013-12": 0.5875,
78
+ "acc,exam_id__2011-05": 0.475,
79
+ "acc,exam_id__2017-23": 0.4875,
80
+ "acc,exam_id__2013-11": 0.5375,
81
+ "acc,exam_id__2016-20": 0.5375,
82
+ "acc,exam_id__2016-21": 0.375,
83
+ "acc,exam_id__2018-25": 0.525,
84
+ "acc,exam_id__2010-01": 0.43529411764705883,
85
+ "acc,exam_id__2012-08": 0.525,
86
+ "acc,exam_id__2015-17": 0.6538461538461539,
87
+ "alias": "oab_exams"
88
+ },
89
+ "portuguese_hate_speech": {
90
+ "alias": "portuguese_hate_speech_binary",
91
+ "f1_macro,all": 0.628750468656138,
92
+ "acc,all": 0.6321974148061105
93
+ },
94
+ "tweetsentbr": {
95
+ "f1_macro,all": 0.4905941623046118,
96
+ "acc,all": 0.7174129353233831,
97
+ "alias": "tweetsentbr"
98
+ }
99
+ },
100
+ "configs": {
101
+ "assin2_rte": {
102
+ "task": "assin2_rte",
103
+ "group": [
104
+ "pt_benchmark",
105
+ "assin2"
106
+ ],
107
+ "dataset_path": "assin2",
108
+ "test_split": "test",
109
+ "fewshot_split": "train",
110
+ "doc_to_text": "Premissa: {{premise}}\nHipótese: {{hypothesis}}\nPergunta: A hipótese pode ser inferida pela premissa? Sim ou Não?\nResposta:",
111
+ "doc_to_target": "{{['Não', 'Sim'][entailment_judgment]}}",
112
+ "description": "Abaixo estão pares de premissa e hipótese. Para cada par, indique se a hipótese pode ser inferida a partir da premissa, responda apenas com \"Sim\" ou \"Não\".\n\n",
113
+ "target_delimiter": " ",
114
+ "fewshot_delimiter": "\n\n",
115
+ "fewshot_config": {
116
+ "sampler": "id_sampler",
117
+ "sampler_config": {
118
+ "id_list": [
119
+ 1,
120
+ 3251,
121
+ 2,
122
+ 3252,
123
+ 3,
124
+ 4,
125
+ 5,
126
+ 6,
127
+ 3253,
128
+ 7,
129
+ 3254,
130
+ 3255,
131
+ 3256,
132
+ 8,
133
+ 9,
134
+ 10,
135
+ 3257,
136
+ 11,
137
+ 3258,
138
+ 12,
139
+ 13,
140
+ 14,
141
+ 15,
142
+ 3259,
143
+ 3260,
144
+ 3261,
145
+ 3262,
146
+ 3263,
147
+ 16,
148
+ 17,
149
+ 3264,
150
+ 18,
151
+ 3265,
152
+ 3266,
153
+ 3267,
154
+ 19,
155
+ 20,
156
+ 3268,
157
+ 3269,
158
+ 21,
159
+ 3270,
160
+ 3271,
161
+ 22,
162
+ 3272,
163
+ 3273,
164
+ 23,
165
+ 3274,
166
+ 24,
167
+ 25,
168
+ 3275
169
+ ],
170
+ "id_column": "sentence_pair_id"
171
+ }
172
+ },
173
+ "num_fewshot": 15,
174
+ "metric_list": [
175
+ {
176
+ "metric": "f1_macro",
177
+ "aggregation": "f1_macro",
178
+ "higher_is_better": true
179
+ },
180
+ {
181
+ "metric": "acc",
182
+ "aggregation": "acc",
183
+ "higher_is_better": true
184
+ }
185
+ ],
186
+ "output_type": "generate_until",
187
+ "generation_kwargs": {
188
+ "max_gen_toks": 32,
189
+ "do_sample": false,
190
+ "temperature": 0.0,
191
+ "top_k": null,
192
+ "top_p": null,
193
+ "until": [
194
+ "\n\n"
195
+ ]
196
+ },
197
+ "repeats": 1,
198
+ "filter_list": [
199
+ {
200
+ "name": "all",
201
+ "filter": [
202
+ {
203
+ "function": "find_similar_label",
204
+ "labels": [
205
+ "Sim",
206
+ "Não"
207
+ ]
208
+ },
209
+ {
210
+ "function": "take_first"
211
+ }
212
+ ]
213
+ }
214
+ ],
215
+ "should_decontaminate": false,
216
+ "metadata": {
217
+ "version": 1.1
218
+ }
219
+ },
220
+ "assin2_sts": {
221
+ "task": "assin2_sts",
222
+ "group": [
223
+ "pt_benchmark",
224
+ "assin2"
225
+ ],
226
+ "dataset_path": "assin2",
227
+ "test_split": "test",
228
+ "fewshot_split": "train",
229
+ "doc_to_text": "Frase 1: {{premise}}\nFrase 2: {{hypothesis}}\nPergunta: Quão similares são as duas frases? Dê uma pontuação entre 1,0 a 5,0.\nResposta:",
230
+ "doc_to_target": "<function assin2_float_to_pt_str at 0x7f80d069a200>",
231
+ "description": "Abaixo estão pares de frases que você deve avaliar o grau de similaridade. Dê uma pontuação entre 1,0 e 5,0, sendo 1,0 pouco similar e 5,0 muito similar.\n\n",
232
+ "target_delimiter": " ",
233
+ "fewshot_delimiter": "\n\n",
234
+ "fewshot_config": {
235
+ "sampler": "id_sampler",
236
+ "sampler_config": {
237
+ "id_list": [
238
+ 1,
239
+ 3251,
240
+ 2,
241
+ 3252,
242
+ 3,
243
+ 4,
244
+ 5,
245
+ 6,
246
+ 3253,
247
+ 7,
248
+ 3254,
249
+ 3255,
250
+ 3256,
251
+ 8,
252
+ 9,
253
+ 10,
254
+ 3257,
255
+ 11,
256
+ 3258,
257
+ 12,
258
+ 13,
259
+ 14,
260
+ 15,
261
+ 3259,
262
+ 3260,
263
+ 3261,
264
+ 3262,
265
+ 3263,
266
+ 16,
267
+ 17,
268
+ 3264,
269
+ 18,
270
+ 3265,
271
+ 3266,
272
+ 3267,
273
+ 19,
274
+ 20,
275
+ 3268,
276
+ 3269,
277
+ 21,
278
+ 3270,
279
+ 3271,
280
+ 22,
281
+ 3272,
282
+ 3273,
283
+ 23,
284
+ 3274,
285
+ 24,
286
+ 25,
287
+ 3275
288
+ ],
289
+ "id_column": "sentence_pair_id"
290
+ }
291
+ },
292
+ "num_fewshot": 15,
293
+ "metric_list": [
294
+ {
295
+ "metric": "pearson",
296
+ "aggregation": "pearsonr",
297
+ "higher_is_better": true
298
+ },
299
+ {
300
+ "metric": "mse",
301
+ "aggregation": "mean_squared_error",
302
+ "higher_is_better": false
303
+ }
304
+ ],
305
+ "output_type": "generate_until",
306
+ "generation_kwargs": {
307
+ "max_gen_toks": 32,
308
+ "do_sample": false,
309
+ "temperature": 0.0,
310
+ "top_k": null,
311
+ "top_p": null,
312
+ "until": [
313
+ "\n\n"
314
+ ]
315
+ },
316
+ "repeats": 1,
317
+ "filter_list": [
318
+ {
319
+ "name": "all",
320
+ "filter": [
321
+ {
322
+ "function": "number_filter",
323
+ "type": "float",
324
+ "range_min": 1.0,
325
+ "range_max": 5.0,
326
+ "on_outside_range": "clip",
327
+ "fallback": 5.0
328
+ },
329
+ {
330
+ "function": "take_first"
331
+ }
332
+ ]
333
+ }
334
+ ],
335
+ "should_decontaminate": false,
336
+ "metadata": {
337
+ "version": 1.1
338
+ }
339
+ },
340
+ "bluex": {
341
+ "task": "bluex",
342
+ "group": [
343
+ "pt_benchmark",
344
+ "vestibular"
345
+ ],
346
+ "dataset_path": "eduagarcia-temp/BLUEX_without_images",
347
+ "test_split": "train",
348
+ "fewshot_split": "train",
349
+ "doc_to_text": "<function enem_doc_to_text at 0x7f80d0699bc0>",
350
+ "doc_to_target": "{{answerKey}}",
351
+ "description": "As perguntas a seguir são questões de múltipla escolha de provas de vestibular de universidades brasileiras, selecione a única alternativa correta e responda apenas com as letras \"A\", \"B\", \"C\", \"D\" ou \"E\".\n\n",
352
+ "target_delimiter": " ",
353
+ "fewshot_delimiter": "\n\n",
354
+ "fewshot_config": {
355
+ "sampler": "id_sampler",
356
+ "sampler_config": {
357
+ "id_list": [
358
+ "USP_2018_3",
359
+ "UNICAMP_2018_2",
360
+ "USP_2018_35",
361
+ "UNICAMP_2018_16",
362
+ "USP_2018_89"
363
+ ],
364
+ "id_column": "id",
365
+ "exclude_from_task": true
366
+ }
367
+ },
368
+ "num_fewshot": 3,
369
+ "metric_list": [
370
+ {
371
+ "metric": "acc",
372
+ "aggregation": "acc",
373
+ "higher_is_better": true
374
+ }
375
+ ],
376
+ "output_type": "generate_until",
377
+ "generation_kwargs": {
378
+ "max_gen_toks": 32,
379
+ "do_sample": false,
380
+ "temperature": 0.0,
381
+ "top_k": null,
382
+ "top_p": null,
383
+ "until": [
384
+ "\n\n"
385
+ ]
386
+ },
387
+ "repeats": 1,
388
+ "filter_list": [
389
+ {
390
+ "name": "all",
391
+ "filter": [
392
+ {
393
+ "function": "normalize_spaces"
394
+ },
395
+ {
396
+ "function": "remove_accents"
397
+ },
398
+ {
399
+ "function": "find_choices",
400
+ "choices": [
401
+ "A",
402
+ "B",
403
+ "C",
404
+ "D",
405
+ "E"
406
+ ],
407
+ "regex_patterns": [
408
+ "(?:[Ll]etra|[Aa]lternativa|[Rr]esposta|[Rr]esposta [Cc]orreta|[Rr]esposta [Cc]orreta e|[Oo]pcao):? ([ABCDE])\\b",
409
+ "\\b([ABCDE])\\.",
410
+ "\\b([ABCDE]) ?[.):-]",
411
+ "\\b([ABCDE])$",
412
+ "\\b([ABCDE])\\b"
413
+ ]
414
+ },
415
+ {
416
+ "function": "take_first"
417
+ }
418
+ ],
419
+ "group_by": {
420
+ "column": "exam_id"
421
+ }
422
+ }
423
+ ],
424
+ "should_decontaminate": true,
425
+ "doc_to_decontamination_query": "<function enem_doc_to_text at 0x7f80d0699e40>",
426
+ "metadata": {
427
+ "version": 1.1
428
+ }
429
+ },
430
+ "enem_challenge": {
431
+ "task": "enem_challenge",
432
+ "task_alias": "enem",
433
+ "group": [
434
+ "pt_benchmark",
435
+ "vestibular"
436
+ ],
437
+ "dataset_path": "eduagarcia/enem_challenge",
438
+ "test_split": "train",
439
+ "fewshot_split": "train",
440
+ "doc_to_text": "<function enem_doc_to_text at 0x7f80d069a3e0>",
441
+ "doc_to_target": "{{answerKey}}",
442
+ "description": "As perguntas a seguir são questões de múltipla escolha do Exame Nacional do Ensino Médio (ENEM), selecione a única alternativa correta e responda apenas com as letras \"A\", \"B\", \"C\", \"D\" ou \"E\".\n\n",
443
+ "target_delimiter": " ",
444
+ "fewshot_delimiter": "\n\n",
445
+ "fewshot_config": {
446
+ "sampler": "id_sampler",
447
+ "sampler_config": {
448
+ "id_list": [
449
+ "2022_21",
450
+ "2022_88",
451
+ "2022_143"
452
+ ],
453
+ "id_column": "id",
454
+ "exclude_from_task": true
455
+ }
456
+ },
457
+ "num_fewshot": 3,
458
+ "metric_list": [
459
+ {
460
+ "metric": "acc",
461
+ "aggregation": "acc",
462
+ "higher_is_better": true
463
+ }
464
+ ],
465
+ "output_type": "generate_until",
466
+ "generation_kwargs": {
467
+ "max_gen_toks": 32,
468
+ "do_sample": false,
469
+ "temperature": 0.0,
470
+ "top_k": null,
471
+ "top_p": null,
472
+ "until": [
473
+ "\n\n"
474
+ ]
475
+ },
476
+ "repeats": 1,
477
+ "filter_list": [
478
+ {
479
+ "name": "all",
480
+ "filter": [
481
+ {
482
+ "function": "normalize_spaces"
483
+ },
484
+ {
485
+ "function": "remove_accents"
486
+ },
487
+ {
488
+ "function": "find_choices",
489
+ "choices": [
490
+ "A",
491
+ "B",
492
+ "C",
493
+ "D",
494
+ "E"
495
+ ],
496
+ "regex_patterns": [
497
+ "(?:[Ll]etra|[Aa]lternativa|[Rr]esposta|[Rr]esposta [Cc]orreta|[Rr]esposta [Cc]orreta e|[Oo]pcao):? ([ABCDE])\\b",
498
+ "\\b([ABCDE])\\.",
499
+ "\\b([ABCDE]) ?[.):-]",
500
+ "\\b([ABCDE])$",
501
+ "\\b([ABCDE])\\b"
502
+ ]
503
+ },
504
+ {
505
+ "function": "take_first"
506
+ }
507
+ ],
508
+ "group_by": {
509
+ "column": "exam_id"
510
+ }
511
+ }
512
+ ],
513
+ "should_decontaminate": true,
514
+ "doc_to_decontamination_query": "<function enem_doc_to_text at 0x7f80d069a660>",
515
+ "metadata": {
516
+ "version": 1.1
517
+ }
518
+ },
519
+ "faquad_nli": {
520
+ "task": "faquad_nli",
521
+ "group": [
522
+ "pt_benchmark"
523
+ ],
524
+ "dataset_path": "ruanchaves/faquad-nli",
525
+ "test_split": "test",
526
+ "fewshot_split": "train",
527
+ "doc_to_text": "Pergunta: {{question}}\nResposta: {{answer}}\nA resposta dada satisfaz à pergunta? Sim ou Não?",
528
+ "doc_to_target": "{{['Não', 'Sim'][label]}}",
529
+ "description": "Abaixo estão pares de pergunta e resposta. Para cada par, você deve julgar se a resposta responde à pergunta de maneira satisfatória e aparenta estar correta. Escreva apenas \"Sim\" ou \"Não\".\n\n",
530
+ "target_delimiter": " ",
531
+ "fewshot_delimiter": "\n\n",
532
+ "fewshot_config": {
533
+ "sampler": "first_n",
534
+ "sampler_config": {
535
+ "fewshot_indices": [
536
+ 1893,
537
+ 949,
538
+ 663,
539
+ 105,
540
+ 1169,
541
+ 2910,
542
+ 2227,
543
+ 2813,
544
+ 974,
545
+ 558,
546
+ 1503,
547
+ 1958,
548
+ 2918,
549
+ 601,
550
+ 1560,
551
+ 984,
552
+ 2388,
553
+ 995,
554
+ 2233,
555
+ 1982,
556
+ 165,
557
+ 2788,
558
+ 1312,
559
+ 2285,
560
+ 522,
561
+ 1113,
562
+ 1670,
563
+ 323,
564
+ 236,
565
+ 1263,
566
+ 1562,
567
+ 2519,
568
+ 1049,
569
+ 432,
570
+ 1167,
571
+ 1394,
572
+ 2022,
573
+ 2551,
574
+ 2194,
575
+ 2187,
576
+ 2282,
577
+ 2816,
578
+ 108,
579
+ 301,
580
+ 1185,
581
+ 1315,
582
+ 1420,
583
+ 2436,
584
+ 2322,
585
+ 766
586
+ ]
587
+ }
588
+ },
589
+ "num_fewshot": 15,
590
+ "metric_list": [
591
+ {
592
+ "metric": "f1_macro",
593
+ "aggregation": "f1_macro",
594
+ "higher_is_better": true
595
+ },
596
+ {
597
+ "metric": "acc",
598
+ "aggregation": "acc",
599
+ "higher_is_better": true
600
+ }
601
+ ],
602
+ "output_type": "generate_until",
603
+ "generation_kwargs": {
604
+ "max_gen_toks": 32,
605
+ "do_sample": false,
606
+ "temperature": 0.0,
607
+ "top_k": null,
608
+ "top_p": null,
609
+ "until": [
610
+ "\n\n"
611
+ ]
612
+ },
613
+ "repeats": 1,
614
+ "filter_list": [
615
+ {
616
+ "name": "all",
617
+ "filter": [
618
+ {
619
+ "function": "find_similar_label",
620
+ "labels": [
621
+ "Sim",
622
+ "Não"
623
+ ]
624
+ },
625
+ {
626
+ "function": "take_first"
627
+ }
628
+ ]
629
+ }
630
+ ],
631
+ "should_decontaminate": false,
632
+ "metadata": {
633
+ "version": 1.1
634
+ }
635
+ },
636
+ "hatebr_offensive": {
637
+ "task": "hatebr_offensive",
638
+ "task_alias": "hatebr_offensive_binary",
639
+ "group": [
640
+ "pt_benchmark"
641
+ ],
642
+ "dataset_path": "eduagarcia/portuguese_benchmark",
643
+ "dataset_name": "HateBR_offensive_binary",
644
+ "test_split": "test",
645
+ "fewshot_split": "train",
646
+ "doc_to_text": "Texto: {{sentence}}\nPergunta: O texto é ofensivo?\nResposta:",
647
+ "doc_to_target": "{{'Sim' if label == 1 else 'Não'}}",
648
+ "description": "Abaixo contém o texto de comentários de usuários do Instagram em português, sua tarefa é classificar se o texto é ofensivo ou não. Responda apenas com \"Sim\" ou \"Não\".\n\n",
649
+ "target_delimiter": " ",
650
+ "fewshot_delimiter": "\n\n",
651
+ "fewshot_config": {
652
+ "sampler": "id_sampler",
653
+ "sampler_config": {
654
+ "id_list": [
655
+ 48,
656
+ 44,
657
+ 36,
658
+ 20,
659
+ 3511,
660
+ 88,
661
+ 3555,
662
+ 16,
663
+ 56,
664
+ 3535,
665
+ 60,
666
+ 40,
667
+ 3527,
668
+ 4,
669
+ 76,
670
+ 3579,
671
+ 3523,
672
+ 3551,
673
+ 68,
674
+ 3503,
675
+ 84,
676
+ 3539,
677
+ 64,
678
+ 3599,
679
+ 80,
680
+ 3563,
681
+ 3559,
682
+ 3543,
683
+ 3547,
684
+ 3587,
685
+ 3595,
686
+ 3575,
687
+ 3567,
688
+ 3591,
689
+ 24,
690
+ 96,
691
+ 92,
692
+ 3507,
693
+ 52,
694
+ 72,
695
+ 8,
696
+ 3571,
697
+ 3515,
698
+ 3519,
699
+ 3531,
700
+ 28,
701
+ 32,
702
+ 0,
703
+ 12,
704
+ 3583
705
+ ],
706
+ "id_column": "idx"
707
+ }
708
+ },
709
+ "num_fewshot": 25,
710
+ "metric_list": [
711
+ {
712
+ "metric": "f1_macro",
713
+ "aggregation": "f1_macro",
714
+ "higher_is_better": true
715
+ },
716
+ {
717
+ "metric": "acc",
718
+ "aggregation": "acc",
719
+ "higher_is_better": true
720
+ }
721
+ ],
722
+ "output_type": "generate_until",
723
+ "generation_kwargs": {
724
+ "max_gen_toks": 32,
725
+ "do_sample": false,
726
+ "temperature": 0.0,
727
+ "top_k": null,
728
+ "top_p": null,
729
+ "until": [
730
+ "\n\n"
731
+ ]
732
+ },
733
+ "repeats": 1,
734
+ "filter_list": [
735
+ {
736
+ "name": "all",
737
+ "filter": [
738
+ {
739
+ "function": "find_similar_label",
740
+ "labels": [
741
+ "Sim",
742
+ "Não"
743
+ ]
744
+ },
745
+ {
746
+ "function": "take_first"
747
+ }
748
+ ]
749
+ }
750
+ ],
751
+ "should_decontaminate": false,
752
+ "metadata": {
753
+ "version": 1.0
754
+ }
755
+ },
756
+ "oab_exams": {
757
+ "task": "oab_exams",
758
+ "group": [
759
+ "legal_benchmark",
760
+ "pt_benchmark"
761
+ ],
762
+ "dataset_path": "eduagarcia/oab_exams",
763
+ "test_split": "train",
764
+ "fewshot_split": "train",
765
+ "doc_to_text": "<function doc_to_text at 0x7f80d0699580>",
766
+ "doc_to_target": "{{answerKey}}",
767
+ "description": "As perguntas a seguir são questões de múltipla escolha do Exame de Ordem da Ordem dos Advogados do Brasil (OAB), selecione a única alternativa correta e responda apenas com as letras \"A\", \"B\", \"C\" ou \"D\".\n\n",
768
+ "target_delimiter": " ",
769
+ "fewshot_delimiter": "\n\n",
770
+ "fewshot_config": {
771
+ "sampler": "id_sampler",
772
+ "sampler_config": {
773
+ "id_list": [
774
+ "2010-01_1",
775
+ "2010-01_11",
776
+ "2010-01_13",
777
+ "2010-01_23",
778
+ "2010-01_26",
779
+ "2010-01_28",
780
+ "2010-01_38",
781
+ "2010-01_48",
782
+ "2010-01_58",
783
+ "2010-01_68",
784
+ "2010-01_76",
785
+ "2010-01_83",
786
+ "2010-01_85",
787
+ "2010-01_91",
788
+ "2010-01_99"
789
+ ],
790
+ "id_column": "id",
791
+ "exclude_from_task": true
792
+ }
793
+ },
794
+ "num_fewshot": 3,
795
+ "metric_list": [
796
+ {
797
+ "metric": "acc",
798
+ "aggregation": "acc",
799
+ "higher_is_better": true
800
+ }
801
+ ],
802
+ "output_type": "generate_until",
803
+ "generation_kwargs": {
804
+ "max_gen_toks": 32,
805
+ "do_sample": false,
806
+ "temperature": 0.0,
807
+ "top_k": null,
808
+ "top_p": null,
809
+ "until": [
810
+ "\n\n"
811
+ ]
812
+ },
813
+ "repeats": 1,
814
+ "filter_list": [
815
+ {
816
+ "name": "all",
817
+ "filter": [
818
+ {
819
+ "function": "normalize_spaces"
820
+ },
821
+ {
822
+ "function": "remove_accents"
823
+ },
824
+ {
825
+ "function": "find_choices",
826
+ "choices": [
827
+ "A",
828
+ "B",
829
+ "C",
830
+ "D"
831
+ ],
832
+ "regex_patterns": [
833
+ "(?:[Ll]etra|[Aa]lternativa|[Rr]esposta|[Rr]esposta [Cc]orreta|[Rr]esposta [Cc]orreta e|[Oo]pcao):? ([ABCD])\\b",
834
+ "\\b([ABCD])\\.",
835
+ "\\b([ABCD]) ?[.):-]",
836
+ "\\b([ABCD])$",
837
+ "\\b([ABCD])\\b"
838
+ ]
839
+ },
840
+ {
841
+ "function": "take_first"
842
+ }
843
+ ],
844
+ "group_by": {
845
+ "column": "exam_id"
846
+ }
847
+ }
848
+ ],
849
+ "should_decontaminate": true,
850
+ "doc_to_decontamination_query": "<function doc_to_text at 0x7f80d0699800>",
851
+ "metadata": {
852
+ "version": 1.5
853
+ }
854
+ },
855
+ "portuguese_hate_speech": {
856
+ "task": "portuguese_hate_speech",
857
+ "task_alias": "portuguese_hate_speech_binary",
858
+ "group": [
859
+ "pt_benchmark"
860
+ ],
861
+ "dataset_path": "eduagarcia/portuguese_benchmark",
862
+ "dataset_name": "Portuguese_Hate_Speech_binary",
863
+ "test_split": "test",
864
+ "fewshot_split": "train",
865
+ "doc_to_text": "Texto: {{sentence}}\nPergunta: O texto contém discurso de ódio?\nResposta:",
866
+ "doc_to_target": "{{'Sim' if label == 1 else 'Não'}}",
867
+ "description": "Abaixo contém o texto de tweets de usuários do Twitter em português, sua tarefa é classificar se o texto contém discurso de ódio ou não. Responda apenas com \"Sim\" ou \"Não\".\n\n",
868
+ "target_delimiter": " ",
869
+ "fewshot_delimiter": "\n\n",
870
+ "fewshot_config": {
871
+ "sampler": "id_sampler",
872
+ "sampler_config": {
873
+ "id_list": [
874
+ 52,
875
+ 50,
876
+ 39,
877
+ 28,
878
+ 3,
879
+ 105,
880
+ 22,
881
+ 25,
882
+ 60,
883
+ 11,
884
+ 66,
885
+ 41,
886
+ 9,
887
+ 4,
888
+ 91,
889
+ 42,
890
+ 7,
891
+ 20,
892
+ 76,
893
+ 1,
894
+ 104,
895
+ 13,
896
+ 67,
897
+ 54,
898
+ 97,
899
+ 27,
900
+ 24,
901
+ 14,
902
+ 16,
903
+ 48,
904
+ 53,
905
+ 40,
906
+ 34,
907
+ 49,
908
+ 32,
909
+ 119,
910
+ 114,
911
+ 2,
912
+ 58,
913
+ 83,
914
+ 18,
915
+ 36,
916
+ 5,
917
+ 6,
918
+ 10,
919
+ 35,
920
+ 38,
921
+ 0,
922
+ 21,
923
+ 46
924
+ ],
925
+ "id_column": "idx"
926
+ }
927
+ },
928
+ "num_fewshot": 25,
929
+ "metric_list": [
930
+ {
931
+ "metric": "f1_macro",
932
+ "aggregation": "f1_macro",
933
+ "higher_is_better": true
934
+ },
935
+ {
936
+ "metric": "acc",
937
+ "aggregation": "acc",
938
+ "higher_is_better": true
939
+ }
940
+ ],
941
+ "output_type": "generate_until",
942
+ "generation_kwargs": {
943
+ "max_gen_toks": 32,
944
+ "do_sample": false,
945
+ "temperature": 0.0,
946
+ "top_k": null,
947
+ "top_p": null,
948
+ "until": [
949
+ "\n\n"
950
+ ]
951
+ },
952
+ "repeats": 1,
953
+ "filter_list": [
954
+ {
955
+ "name": "all",
956
+ "filter": [
957
+ {
958
+ "function": "find_similar_label",
959
+ "labels": [
960
+ "Sim",
961
+ "Não"
962
+ ]
963
+ },
964
+ {
965
+ "function": "take_first"
966
+ }
967
+ ]
968
+ }
969
+ ],
970
+ "should_decontaminate": false,
971
+ "metadata": {
972
+ "version": 1.0
973
+ }
974
+ },
975
+ "tweetsentbr": {
976
+ "task": "tweetsentbr",
977
+ "group": [
978
+ "pt_benchmark"
979
+ ],
980
+ "dataset_path": "eduagarcia/tweetsentbr_fewshot",
981
+ "test_split": "test",
982
+ "fewshot_split": "train",
983
+ "doc_to_text": "Texto: {{sentence}}\nPergunta: O sentimento do texto é Positivo, Neutro ou Negativo?\nResposta:",
984
+ "doc_to_target": "{{'Positivo' if label == 'Positive' else ('Negativo' if label == 'Negative' else 'Neutro')}}",
985
+ "description": "Abaixo contém o texto de tweets de usuários do Twitter em português, sua tarefa é classificar se o sentimento do texto é Positivo, Neutro ou Negativo. Responda apenas com uma das opções.\n\n",
986
+ "target_delimiter": " ",
987
+ "fewshot_delimiter": "\n\n",
988
+ "fewshot_config": {
989
+ "sampler": "first_n"
990
+ },
991
+ "num_fewshot": 25,
992
+ "metric_list": [
993
+ {
994
+ "metric": "f1_macro",
995
+ "aggregation": "f1_macro",
996
+ "higher_is_better": true
997
+ },
998
+ {
999
+ "metric": "acc",
1000
+ "aggregation": "acc",
1001
+ "higher_is_better": true
1002
+ }
1003
+ ],
1004
+ "output_type": "generate_until",
1005
+ "generation_kwargs": {
1006
+ "max_gen_toks": 32,
1007
+ "do_sample": false,
1008
+ "temperature": 0.0,
1009
+ "top_k": null,
1010
+ "top_p": null,
1011
+ "until": [
1012
+ "\n\n"
1013
+ ]
1014
+ },
1015
+ "repeats": 1,
1016
+ "filter_list": [
1017
+ {
1018
+ "name": "all",
1019
+ "filter": [
1020
+ {
1021
+ "function": "find_similar_label",
1022
+ "labels": [
1023
+ "Positivo",
1024
+ "Neutro",
1025
+ "Negativo"
1026
+ ]
1027
+ },
1028
+ {
1029
+ "function": "take_first"
1030
+ }
1031
+ ]
1032
+ }
1033
+ ],
1034
+ "should_decontaminate": false,
1035
+ "metadata": {
1036
+ "version": 1.0
1037
+ }
1038
+ }
1039
+ },
1040
+ "versions": {
1041
+ "assin2_rte": 1.1,
1042
+ "assin2_sts": 1.1,
1043
+ "bluex": 1.1,
1044
+ "enem_challenge": 1.1,
1045
+ "faquad_nli": 1.1,
1046
+ "hatebr_offensive": 1.0,
1047
+ "oab_exams": 1.5,
1048
+ "portuguese_hate_speech": 1.0,
1049
+ "tweetsentbr": 1.0
1050
+ },
1051
+ "n-shot": {
1052
+ "assin2_rte": 15,
1053
+ "assin2_sts": 15,
1054
+ "bluex": 3,
1055
+ "enem_challenge": 3,
1056
+ "faquad_nli": 15,
1057
+ "hatebr_offensive": 25,
1058
+ "oab_exams": 3,
1059
+ "portuguese_hate_speech": 25,
1060
+ "tweetsentbr": 25
1061
+ },
1062
+ "model_meta": {
1063
+ "truncated": 0,
1064
+ "non_truncated": 14150,
1065
+ "padded": 0,
1066
+ "non_padded": 14150,
1067
+ "fewshots_truncated": 0,
1068
+ "has_chat_template": true,
1069
+ "chat_type": "system_user_assistant",
1070
+ "n_gpus": 1,
1071
+ "accelerate_num_process": null,
1072
+ "model_sha": "d0d40655b25f24afe504142cce427fc698f25239",
1073
+ "model_dtype": "torch.float16",
1074
+ "model_memory_footprint": 16060530688,
1075
+ "model_num_parameters": 8030261248,
1076
+ "model_is_loaded_in_4bit": null,
1077
+ "model_is_loaded_in_8bit": null,
1078
+ "model_is_quantized": null,
1079
+ "model_device": "cuda:0",
1080
+ "batch_size": 8,
1081
+ "max_length": 2560,
1082
+ "max_ctx_length": 2528,
1083
+ "max_gen_toks": 32
1084
+ },
1085
+ "task_model_meta": {
1086
+ "assin2_rte": {
1087
+ "sample_size": 2448,
1088
+ "truncated": 0,
1089
+ "non_truncated": 2448,
1090
+ "padded": 0,
1091
+ "non_padded": 2448,
1092
+ "fewshots_truncated": 0,
1093
+ "mean_seq_length": 1318.5322712418301,
1094
+ "min_seq_length": 1299,
1095
+ "max_seq_length": 1382,
1096
+ "max_ctx_length": 2528,
1097
+ "max_gen_toks": 32,
1098
+ "mean_original_fewshots_size": 15.0,
1099
+ "mean_effective_fewshot_size": 15.0
1100
+ },
1101
+ "assin2_sts": {
1102
+ "sample_size": 2448,
1103
+ "truncated": 0,
1104
+ "non_truncated": 2448,
1105
+ "padded": 0,
1106
+ "non_padded": 2448,
1107
+ "fewshots_truncated": 0,
1108
+ "mean_seq_length": 1509.5322712418301,
1109
+ "min_seq_length": 1490,
1110
+ "max_seq_length": 1573,
1111
+ "max_ctx_length": 2528,
1112
+ "max_gen_toks": 32,
1113
+ "mean_original_fewshots_size": 15.0,
1114
+ "mean_effective_fewshot_size": 15.0
1115
+ },
1116
+ "bluex": {
1117
+ "sample_size": 719,
1118
+ "truncated": 0,
1119
+ "non_truncated": 719,
1120
+ "padded": 0,
1121
+ "non_padded": 719,
1122
+ "fewshots_truncated": 0,
1123
+ "mean_seq_length": 1484.769123783032,
1124
+ "min_seq_length": 1165,
1125
+ "max_seq_length": 2134,
1126
+ "max_ctx_length": 2528,
1127
+ "max_gen_toks": 32,
1128
+ "mean_original_fewshots_size": 3.0,
1129
+ "mean_effective_fewshot_size": 3.0
1130
+ },
1131
+ "enem_challenge": {
1132
+ "sample_size": 1429,
1133
+ "truncated": 0,
1134
+ "non_truncated": 1429,
1135
+ "padded": 0,
1136
+ "non_padded": 1429,
1137
+ "fewshots_truncated": 0,
1138
+ "mean_seq_length": 1412.3547935619315,
1139
+ "min_seq_length": 1187,
1140
+ "max_seq_length": 2340,
1141
+ "max_ctx_length": 2528,
1142
+ "max_gen_toks": 32,
1143
+ "mean_original_fewshots_size": 3.0,
1144
+ "mean_effective_fewshot_size": 3.0
1145
+ },
1146
+ "faquad_nli": {
1147
+ "sample_size": 650,
1148
+ "truncated": 0,
1149
+ "non_truncated": 650,
1150
+ "padded": 0,
1151
+ "non_padded": 650,
1152
+ "fewshots_truncated": 0,
1153
+ "mean_seq_length": 1447.8215384615385,
1154
+ "min_seq_length": 1402,
1155
+ "max_seq_length": 1544,
1156
+ "max_ctx_length": 2528,
1157
+ "max_gen_toks": 32,
1158
+ "mean_original_fewshots_size": 15.0,
1159
+ "mean_effective_fewshot_size": 15.0
1160
+ },
1161
+ "hatebr_offensive": {
1162
+ "sample_size": 1400,
1163
+ "truncated": 0,
1164
+ "non_truncated": 1400,
1165
+ "padded": 0,
1166
+ "non_padded": 1400,
1167
+ "fewshots_truncated": 0,
1168
+ "mean_seq_length": 1279.3878571428572,
1169
+ "min_seq_length": 1259,
1170
+ "max_seq_length": 1498,
1171
+ "max_ctx_length": 2528,
1172
+ "max_gen_toks": 32,
1173
+ "mean_original_fewshots_size": 25.0,
1174
+ "mean_effective_fewshot_size": 25.0
1175
+ },
1176
+ "oab_exams": {
1177
+ "sample_size": 2195,
1178
+ "truncated": 0,
1179
+ "non_truncated": 2195,
1180
+ "padded": 0,
1181
+ "non_padded": 2195,
1182
+ "fewshots_truncated": 0,
1183
+ "mean_seq_length": 1220.3772209567198,
1184
+ "min_seq_length": 988,
1185
+ "max_seq_length": 1654,
1186
+ "max_ctx_length": 2528,
1187
+ "max_gen_toks": 32,
1188
+ "mean_original_fewshots_size": 3.0,
1189
+ "mean_effective_fewshot_size": 3.0
1190
+ },
1191
+ "portuguese_hate_speech": {
1192
+ "sample_size": 851,
1193
+ "truncated": 0,
1194
+ "non_truncated": 851,
1195
+ "padded": 0,
1196
+ "non_padded": 851,
1197
+ "fewshots_truncated": 0,
1198
+ "mean_seq_length": 1676.4195064629848,
1199
+ "min_seq_length": 1646,
1200
+ "max_seq_length": 1708,
1201
+ "max_ctx_length": 2528,
1202
+ "max_gen_toks": 32,
1203
+ "mean_original_fewshots_size": 25.0,
1204
+ "mean_effective_fewshot_size": 25.0
1205
+ },
1206
+ "tweetsentbr": {
1207
+ "sample_size": 2010,
1208
+ "truncated": 0,
1209
+ "non_truncated": 2010,
1210
+ "padded": 0,
1211
+ "non_padded": 2010,
1212
+ "fewshots_truncated": 0,
1213
+ "mean_seq_length": 1537.1537313432837,
1214
+ "min_seq_length": 1520,
1215
+ "max_seq_length": 1585,
1216
+ "max_ctx_length": 2528,
1217
+ "max_gen_toks": 32,
1218
+ "mean_original_fewshots_size": 25.0,
1219
+ "mean_effective_fewshot_size": 25.0
1220
+ }
1221
+ },
1222
+ "config": {
1223
+ "model": "huggingface",
1224
+ "model_args": "pretrained=mlabonne/Llama-3-8B-Instruct-abliterated-dpomix,dtype=float16,device=cuda:0,revision=main,trust_remote_code=True,starting_max_length=2560",
1225
+ "batch_size": "auto",
1226
+ "batch_sizes": [],
1227
+ "device": null,
1228
+ "use_cache": null,
1229
+ "limit": [
1230
+ null,
1231
+ null,
1232
+ null,
1233
+ null,
1234
+ null,
1235
+ null,
1236
+ null,
1237
+ null,
1238
+ null
1239
+ ],
1240
+ "bootstrap_iters": 0,
1241
+ "gen_kwargs": null
1242
+ },
1243
+ "git_hash": "51e0e5e"
1244
+ }
mlabonne/Llama-3-8B-Instruct-abliterated-dpomix/results_2024-05-27T14-44-42.467656.json ADDED
@@ -0,0 +1,325 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "config_general": {
3
+ "start_date": "2024-05-27T14-44-42.467656",
4
+ "start_time": 1716821083.1431003,
5
+ "end_time": 1716824910.6436794,
6
+ "total_evaluation_time_seconds": 3827.5005791187286,
7
+ "has_chat_template": true,
8
+ "chat_type": "system_user_assistant",
9
+ "n_gpus": 1,
10
+ "accelerate_num_process": null,
11
+ "model_sha": "d0d40655b25f24afe504142cce427fc698f25239",
12
+ "model_dtype": "float16",
13
+ "model_memory_footprint": 16060530688,
14
+ "model_num_parameters": 8030261248,
15
+ "model_is_loaded_in_4bit": null,
16
+ "model_is_loaded_in_8bit": null,
17
+ "model_is_quantized": null,
18
+ "model_device": "cuda:0",
19
+ "batch_size": 8,
20
+ "max_length": 2560,
21
+ "max_ctx_length": 2528,
22
+ "max_gen_toks": 32,
23
+ "model_name": "mlabonne/Llama-3-8B-Instruct-abliterated-dpomix",
24
+ "job_id": 747,
25
+ "model_id": "mlabonne/Llama-3-8B-Instruct-abliterated-dpomix_eval_request_False_float16_Original",
26
+ "model_base_model": "",
27
+ "model_weight_type": "Original",
28
+ "model_revision": "main",
29
+ "model_private": false,
30
+ "model_type": "💬 : chat (RLHF, DPO, IFT, ...)",
31
+ "model_architectures": "LlamaForCausalLM",
32
+ "submitted_time": "2024-05-27T09:26:30Z",
33
+ "lm_eval_model_type": "huggingface",
34
+ "eval_version": "1.1.0"
35
+ },
36
+ "results": {
37
+ "all_grouped_average": 0.6853042372941637,
38
+ "all_grouped_npm": 0.5283923058028247,
39
+ "all_grouped": {
40
+ "enem_challenge": 0.7053883834849545,
41
+ "bluex": 0.5730180806675939,
42
+ "oab_exams": 0.5097949886104783,
43
+ "assin2_rte": 0.9136602188080808,
44
+ "assin2_sts": 0.7676336804192703,
45
+ "faquad_nli": 0.7143293906330614,
46
+ "hatebr_offensive": 0.8645687620632841,
47
+ "portuguese_hate_speech": 0.628750468656138,
48
+ "tweetsentbr": 0.4905941623046118
49
+ },
50
+ "all": {
51
+ "harness|enem_challenge|enem_challenge|None|3": 0.7053883834849545,
52
+ "harness|bluex|bluex|None|3": 0.5730180806675939,
53
+ "harness|oab_exams|oab_exams|None|3": 0.5097949886104783,
54
+ "harness|assin2_rte|assin2_rte|None|15": 0.9136602188080808,
55
+ "harness|assin2_sts|assin2_sts|None|15": 0.7676336804192703,
56
+ "harness|faquad_nli|faquad_nli|None|15": 0.7143293906330614,
57
+ "harness|hatebr_offensive|hatebr_offensive|None|25": 0.8645687620632841,
58
+ "harness|portuguese_hate_speech|portuguese_hate_speech|None|25": 0.628750468656138,
59
+ "harness|tweetsentbr|tweetsentbr|None|25": 0.4905941623046118
60
+ },
61
+ "harness|enem_challenge|enem_challenge|None|3": {
62
+ "acc,all": 0.7053883834849545,
63
+ "acc,exam_id__2011": 0.717948717948718,
64
+ "acc,exam_id__2017": 0.6637931034482759,
65
+ "acc,exam_id__2015": 0.7310924369747899,
66
+ "acc,exam_id__2016": 0.7024793388429752,
67
+ "acc,exam_id__2016_2": 0.6585365853658537,
68
+ "acc,exam_id__2009": 0.7130434782608696,
69
+ "acc,exam_id__2012": 0.7241379310344828,
70
+ "acc,exam_id__2010": 0.7435897435897436,
71
+ "acc,exam_id__2013": 0.6944444444444444,
72
+ "acc,exam_id__2014": 0.6972477064220184,
73
+ "acc,exam_id__2022": 0.6616541353383458,
74
+ "acc,exam_id__2023": 0.7555555555555555,
75
+ "main_score": 0.7053883834849545
76
+ },
77
+ "harness|bluex|bluex|None|3": {
78
+ "acc,all": 0.5730180806675939,
79
+ "acc,exam_id__UNICAMP_2018": 0.5,
80
+ "acc,exam_id__UNICAMP_2023": 0.627906976744186,
81
+ "acc,exam_id__USP_2023": 0.6818181818181818,
82
+ "acc,exam_id__UNICAMP_2024": 0.6666666666666666,
83
+ "acc,exam_id__USP_2024": 0.6341463414634146,
84
+ "acc,exam_id__UNICAMP_2021_1": 0.5217391304347826,
85
+ "acc,exam_id__USP_2020": 0.5714285714285714,
86
+ "acc,exam_id__UNICAMP_2020": 0.5818181818181818,
87
+ "acc,exam_id__UNICAMP_2022": 0.5641025641025641,
88
+ "acc,exam_id__UNICAMP_2019": 0.54,
89
+ "acc,exam_id__UNICAMP_2021_2": 0.5686274509803921,
90
+ "acc,exam_id__USP_2018": 0.46296296296296297,
91
+ "acc,exam_id__USP_2021": 0.5961538461538461,
92
+ "acc,exam_id__USP_2019": 0.55,
93
+ "acc,exam_id__USP_2022": 0.5714285714285714,
94
+ "main_score": 0.5730180806675939
95
+ },
96
+ "harness|oab_exams|oab_exams|None|3": {
97
+ "acc,all": 0.5097949886104783,
98
+ "acc,exam_id__2011-03": 0.42424242424242425,
99
+ "acc,exam_id__2014-13": 0.45,
100
+ "acc,exam_id__2013-10": 0.45,
101
+ "acc,exam_id__2017-24": 0.4375,
102
+ "acc,exam_id__2017-22": 0.6,
103
+ "acc,exam_id__2012-06a": 0.55,
104
+ "acc,exam_id__2016-20a": 0.425,
105
+ "acc,exam_id__2012-09": 0.45454545454545453,
106
+ "acc,exam_id__2015-16": 0.5,
107
+ "acc,exam_id__2011-04": 0.4625,
108
+ "acc,exam_id__2012-07": 0.5125,
109
+ "acc,exam_id__2014-14": 0.575,
110
+ "acc,exam_id__2014-15": 0.5641025641025641,
111
+ "acc,exam_id__2010-02": 0.57,
112
+ "acc,exam_id__2015-18": 0.525,
113
+ "acc,exam_id__2016-19": 0.5512820512820513,
114
+ "acc,exam_id__2012-06": 0.5875,
115
+ "acc,exam_id__2013-12": 0.5875,
116
+ "acc,exam_id__2011-05": 0.475,
117
+ "acc,exam_id__2017-23": 0.4875,
118
+ "acc,exam_id__2013-11": 0.5375,
119
+ "acc,exam_id__2016-20": 0.5375,
120
+ "acc,exam_id__2016-21": 0.375,
121
+ "acc,exam_id__2018-25": 0.525,
122
+ "acc,exam_id__2010-01": 0.43529411764705883,
123
+ "acc,exam_id__2012-08": 0.525,
124
+ "acc,exam_id__2015-17": 0.6538461538461539,
125
+ "main_score": 0.5097949886104783
126
+ },
127
+ "harness|assin2_rte|assin2_rte|None|15": {
128
+ "f1_macro,all": 0.9136602188080808,
129
+ "acc,all": 0.9138071895424836,
130
+ "main_score": 0.9136602188080808
131
+ },
132
+ "harness|assin2_sts|assin2_sts|None|15": {
133
+ "pearson,all": 0.7676336804192703,
134
+ "mse,all": 0.6167238562091503,
135
+ "main_score": 0.7676336804192703
136
+ },
137
+ "harness|faquad_nli|faquad_nli|None|15": {
138
+ "f1_macro,all": 0.7143293906330614,
139
+ "acc,all": 0.7523076923076923,
140
+ "main_score": 0.7143293906330614
141
+ },
142
+ "harness|hatebr_offensive|hatebr_offensive|None|25": {
143
+ "f1_macro,all": 0.8645687620632841,
144
+ "acc,all": 0.865,
145
+ "main_score": 0.8645687620632841
146
+ },
147
+ "harness|portuguese_hate_speech|portuguese_hate_speech|None|25": {
148
+ "f1_macro,all": 0.628750468656138,
149
+ "acc,all": 0.6321974148061105,
150
+ "main_score": 0.628750468656138
151
+ },
152
+ "harness|tweetsentbr|tweetsentbr|None|25": {
153
+ "f1_macro,all": 0.4905941623046118,
154
+ "acc,all": 0.7174129353233831,
155
+ "main_score": 0.4905941623046118
156
+ }
157
+ },
158
+ "config_tasks": {
159
+ "harness|enem_challenge|enem_challenge": "LM Harness task",
160
+ "harness|bluex|bluex": "LM Harness task",
161
+ "harness|oab_exams|oab_exams": "LM Harness task",
162
+ "harness|assin2_rte|assin2_rte": "LM Harness task",
163
+ "harness|assin2_sts|assin2_sts": "LM Harness task",
164
+ "harness|faquad_nli|faquad_nli": "LM Harness task",
165
+ "harness|hatebr_offensive|hatebr_offensive": "LM Harness task",
166
+ "harness|portuguese_hate_speech|portuguese_hate_speech": "LM Harness task",
167
+ "harness|tweetsentbr|tweetsentbr": "LM Harness task"
168
+ },
169
+ "versions": {
170
+ "all": 0,
171
+ "harness|enem_challenge|enem_challenge": 1.1,
172
+ "harness|bluex|bluex": 1.1,
173
+ "harness|oab_exams|oab_exams": 1.5,
174
+ "harness|assin2_rte|assin2_rte": 1.1,
175
+ "harness|assin2_sts|assin2_sts": 1.1,
176
+ "harness|faquad_nli|faquad_nli": 1.1,
177
+ "harness|hatebr_offensive|hatebr_offensive": 1.0,
178
+ "harness|portuguese_hate_speech|portuguese_hate_speech": 1.0,
179
+ "harness|tweetsentbr|tweetsentbr": 1.0
180
+ },
181
+ "summary_tasks": {
182
+ "harness|enem_challenge|enem_challenge|None|3": {
183
+ "sample_size": 1429,
184
+ "truncated": 0,
185
+ "non_truncated": 1429,
186
+ "padded": 0,
187
+ "non_padded": 1429,
188
+ "fewshots_truncated": 0,
189
+ "mean_seq_length": 1412.3547935619315,
190
+ "min_seq_length": 1187,
191
+ "max_seq_length": 2340,
192
+ "max_ctx_length": 2528,
193
+ "max_gen_toks": 32,
194
+ "mean_original_fewshots_size": 3.0,
195
+ "mean_effective_fewshot_size": 3.0
196
+ },
197
+ "harness|bluex|bluex|None|3": {
198
+ "sample_size": 719,
199
+ "truncated": 0,
200
+ "non_truncated": 719,
201
+ "padded": 0,
202
+ "non_padded": 719,
203
+ "fewshots_truncated": 0,
204
+ "mean_seq_length": 1484.769123783032,
205
+ "min_seq_length": 1165,
206
+ "max_seq_length": 2134,
207
+ "max_ctx_length": 2528,
208
+ "max_gen_toks": 32,
209
+ "mean_original_fewshots_size": 3.0,
210
+ "mean_effective_fewshot_size": 3.0
211
+ },
212
+ "harness|oab_exams|oab_exams|None|3": {
213
+ "sample_size": 2195,
214
+ "truncated": 0,
215
+ "non_truncated": 2195,
216
+ "padded": 0,
217
+ "non_padded": 2195,
218
+ "fewshots_truncated": 0,
219
+ "mean_seq_length": 1220.3772209567198,
220
+ "min_seq_length": 988,
221
+ "max_seq_length": 1654,
222
+ "max_ctx_length": 2528,
223
+ "max_gen_toks": 32,
224
+ "mean_original_fewshots_size": 3.0,
225
+ "mean_effective_fewshot_size": 3.0
226
+ },
227
+ "harness|assin2_rte|assin2_rte|None|15": {
228
+ "sample_size": 2448,
229
+ "truncated": 0,
230
+ "non_truncated": 2448,
231
+ "padded": 0,
232
+ "non_padded": 2448,
233
+ "fewshots_truncated": 0,
234
+ "mean_seq_length": 1318.5322712418301,
235
+ "min_seq_length": 1299,
236
+ "max_seq_length": 1382,
237
+ "max_ctx_length": 2528,
238
+ "max_gen_toks": 32,
239
+ "mean_original_fewshots_size": 15.0,
240
+ "mean_effective_fewshot_size": 15.0
241
+ },
242
+ "harness|assin2_sts|assin2_sts|None|15": {
243
+ "sample_size": 2448,
244
+ "truncated": 0,
245
+ "non_truncated": 2448,
246
+ "padded": 0,
247
+ "non_padded": 2448,
248
+ "fewshots_truncated": 0,
249
+ "mean_seq_length": 1509.5322712418301,
250
+ "min_seq_length": 1490,
251
+ "max_seq_length": 1573,
252
+ "max_ctx_length": 2528,
253
+ "max_gen_toks": 32,
254
+ "mean_original_fewshots_size": 15.0,
255
+ "mean_effective_fewshot_size": 15.0
256
+ },
257
+ "harness|faquad_nli|faquad_nli|None|15": {
258
+ "sample_size": 650,
259
+ "truncated": 0,
260
+ "non_truncated": 650,
261
+ "padded": 0,
262
+ "non_padded": 650,
263
+ "fewshots_truncated": 0,
264
+ "mean_seq_length": 1447.8215384615385,
265
+ "min_seq_length": 1402,
266
+ "max_seq_length": 1544,
267
+ "max_ctx_length": 2528,
268
+ "max_gen_toks": 32,
269
+ "mean_original_fewshots_size": 15.0,
270
+ "mean_effective_fewshot_size": 15.0
271
+ },
272
+ "harness|hatebr_offensive|hatebr_offensive|None|25": {
273
+ "sample_size": 1400,
274
+ "truncated": 0,
275
+ "non_truncated": 1400,
276
+ "padded": 0,
277
+ "non_padded": 1400,
278
+ "fewshots_truncated": 0,
279
+ "mean_seq_length": 1279.3878571428572,
280
+ "min_seq_length": 1259,
281
+ "max_seq_length": 1498,
282
+ "max_ctx_length": 2528,
283
+ "max_gen_toks": 32,
284
+ "mean_original_fewshots_size": 25.0,
285
+ "mean_effective_fewshot_size": 25.0
286
+ },
287
+ "harness|portuguese_hate_speech|portuguese_hate_speech|None|25": {
288
+ "sample_size": 851,
289
+ "truncated": 0,
290
+ "non_truncated": 851,
291
+ "padded": 0,
292
+ "non_padded": 851,
293
+ "fewshots_truncated": 0,
294
+ "mean_seq_length": 1676.4195064629848,
295
+ "min_seq_length": 1646,
296
+ "max_seq_length": 1708,
297
+ "max_ctx_length": 2528,
298
+ "max_gen_toks": 32,
299
+ "mean_original_fewshots_size": 25.0,
300
+ "mean_effective_fewshot_size": 25.0
301
+ },
302
+ "harness|tweetsentbr|tweetsentbr|None|25": {
303
+ "sample_size": 2010,
304
+ "truncated": 0,
305
+ "non_truncated": 2010,
306
+ "padded": 0,
307
+ "non_padded": 2010,
308
+ "fewshots_truncated": 0,
309
+ "mean_seq_length": 1537.1537313432837,
310
+ "min_seq_length": 1520,
311
+ "max_seq_length": 1585,
312
+ "max_ctx_length": 2528,
313
+ "max_gen_toks": 32,
314
+ "mean_original_fewshots_size": 25.0,
315
+ "mean_effective_fewshot_size": 25.0
316
+ }
317
+ },
318
+ "summary_general": {
319
+ "truncated": 0,
320
+ "non_truncated": 14150,
321
+ "padded": 0,
322
+ "non_padded": 14150,
323
+ "fewshots_truncated": 0
324
+ }
325
+ }