Muennighoff
commited on
Commit
•
52322d5
1
Parent(s):
f3c3380
Add babi
Browse files- 2b855b50oscarpy/evaluation/2b855b50oscarpy_0_babi.json +22 -0
- 2b855b50oscarpy/evaluation/2b855b50oscarpy_1_babi.json +22 -0
- 2b855b50oscarpy/evaluation/2b855b50oscarpy_2_babi.json +22 -0
- 2b855b50oscarpy/evaluation/2b855b50oscarpy_3_babi.json +22 -0
- 2b855b50oscarpy/evaluation/2b855b50oscarpy_4_babi.json +22 -0
- 2b855b50oscarpy/evaluation/2b855b50oscarpy_5_babi.json +4 -4
- 2b855b60oscarpy/evaluation/2b855b60oscarpy_0_babi.json +22 -0
- 2b855b60oscarpy/evaluation/2b855b60oscarpy_1_babi.json +22 -0
- 2b855b60oscarpy/evaluation/2b855b60oscarpy_2_babi.json +22 -0
- 2b855b60oscarpy/evaluation/2b855b60oscarpy_3_babi.json +22 -0
- 2b855b60oscarpy/evaluation/2b855b60oscarpy_4_babi.json +22 -0
- 2b855b60oscarpy/evaluation/2b855b60oscarpy_5_babi.json +4 -4
- 2b855b70oscarpy/evaluation/2b855b70oscarpy_0_babi.json +22 -0
- 2b855b70oscarpy/evaluation/2b855b70oscarpy_1_babi.json +22 -0
- 2b855b70oscarpy/evaluation/2b855b70oscarpy_2_babi.json +22 -0
- 2b855b70oscarpy/evaluation/2b855b70oscarpy_3_babi.json +22 -0
- 2b855b70oscarpy/evaluation/2b855b70oscarpy_4_babi.json +22 -0
- 2b855b70oscarpy/evaluation/2b855b70oscarpy_5_babi.json +4 -4
- 2b855b80oscarpy/evaluation/2b855b80oscarpy_0_babi.json +22 -0
- 2b855b80oscarpy/evaluation/2b855b80oscarpy_1_babi.json +22 -0
- 2b855b80oscarpy/evaluation/2b855b80oscarpy_2_babi.json +22 -0
- 2b855b80oscarpy/evaluation/2b855b80oscarpy_3_babi.json +22 -0
- 2b855b80oscarpy/evaluation/2b855b80oscarpy_4_babi.json +22 -0
- 2b855b80oscarpy/evaluation/2b855b80oscarpy_5_babi.json +4 -4
- 2b855b90oscarpy/evaluation/2b855b90oscarpy_0_babi.json +22 -0
- 2b855b90oscarpy/evaluation/2b855b90oscarpy_1_babi.json +22 -0
- 2b855b90oscarpy/evaluation/2b855b90oscarpy_2_babi.json +22 -0
- 2b855b90oscarpy/evaluation/2b855b90oscarpy_3_babi.json +22 -0
- 2b855b90oscarpy/evaluation/2b855b90oscarpy_4_babi.json +22 -0
- 2b855b90oscarpy/evaluation/2b855b90oscarpy_5_babi.json +4 -4
2b855b50oscarpy/evaluation/2b855b50oscarpy_0_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.0,
|
5 |
+
"em_stderr": 0.0
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b50oscarpy/transformers",
|
14 |
+
"num_fewshot": 0,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b50oscarpy/evaluation/2b855b50oscarpy_1_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.07333333333333333,
|
5 |
+
"em_stderr": 0.0047601895956827945
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b50oscarpy/transformers",
|
14 |
+
"num_fewshot": 1,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b50oscarpy/evaluation/2b855b50oscarpy_2_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.15933333333333333,
|
5 |
+
"em_stderr": 0.00668308526183433
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b50oscarpy/transformers",
|
14 |
+
"num_fewshot": 2,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b50oscarpy/evaluation/2b855b50oscarpy_3_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.20766666666666667,
|
5 |
+
"em_stderr": 0.007407115668615664
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b50oscarpy/transformers",
|
14 |
+
"num_fewshot": 3,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b50oscarpy/evaluation/2b855b50oscarpy_4_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.24566666666666667,
|
5 |
+
"em_stderr": 0.007860796135919237
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b50oscarpy/transformers",
|
14 |
+
"num_fewshot": 4,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b50oscarpy/evaluation/2b855b50oscarpy_5_babi.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"results": {
|
3 |
"babi": {
|
4 |
-
"em": 0.
|
5 |
-
"em_stderr": 0.
|
6 |
}
|
7 |
},
|
8 |
"versions": {
|
@@ -14,8 +14,8 @@
|
|
14 |
"num_fewshot": 5,
|
15 |
"batch_size": null,
|
16 |
"device": null,
|
17 |
-
"no_cache":
|
18 |
-
"limit":
|
19 |
"bootstrap_iters": 100000,
|
20 |
"description_dict": {}
|
21 |
}
|
|
|
1 |
{
|
2 |
"results": {
|
3 |
"babi": {
|
4 |
+
"em": 0.2683333333333333,
|
5 |
+
"em_stderr": 0.00809106546127376
|
6 |
}
|
7 |
},
|
8 |
"versions": {
|
|
|
14 |
"num_fewshot": 5,
|
15 |
"batch_size": null,
|
16 |
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
"bootstrap_iters": 100000,
|
20 |
"description_dict": {}
|
21 |
}
|
2b855b60oscarpy/evaluation/2b855b60oscarpy_0_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.0,
|
5 |
+
"em_stderr": 0.0
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b60oscarpy/transformers",
|
14 |
+
"num_fewshot": 0,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b60oscarpy/evaluation/2b855b60oscarpy_1_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.09033333333333333,
|
5 |
+
"em_stderr": 0.0052345207494638975
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b60oscarpy/transformers",
|
14 |
+
"num_fewshot": 1,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b60oscarpy/evaluation/2b855b60oscarpy_2_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.17933333333333334,
|
5 |
+
"em_stderr": 0.007005282836938614
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b60oscarpy/transformers",
|
14 |
+
"num_fewshot": 2,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b60oscarpy/evaluation/2b855b60oscarpy_3_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.23166666666666666,
|
5 |
+
"em_stderr": 0.0077040373487692005
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b60oscarpy/transformers",
|
14 |
+
"num_fewshot": 3,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b60oscarpy/evaluation/2b855b60oscarpy_4_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.26966666666666667,
|
5 |
+
"em_stderr": 0.0081037486448691
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b60oscarpy/transformers",
|
14 |
+
"num_fewshot": 4,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b60oscarpy/evaluation/2b855b60oscarpy_5_babi.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"results": {
|
3 |
"babi": {
|
4 |
-
"em": 0.
|
5 |
-
"em_stderr": 0.
|
6 |
}
|
7 |
},
|
8 |
"versions": {
|
@@ -14,8 +14,8 @@
|
|
14 |
"num_fewshot": 5,
|
15 |
"batch_size": null,
|
16 |
"device": null,
|
17 |
-
"no_cache":
|
18 |
-
"limit":
|
19 |
"bootstrap_iters": 100000,
|
20 |
"description_dict": {}
|
21 |
}
|
|
|
1 |
{
|
2 |
"results": {
|
3 |
"babi": {
|
4 |
+
"em": 0.29333333333333333,
|
5 |
+
"em_stderr": 0.00831380394735107
|
6 |
}
|
7 |
},
|
8 |
"versions": {
|
|
|
14 |
"num_fewshot": 5,
|
15 |
"batch_size": null,
|
16 |
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
"bootstrap_iters": 100000,
|
20 |
"description_dict": {}
|
21 |
}
|
2b855b70oscarpy/evaluation/2b855b70oscarpy_0_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.0,
|
5 |
+
"em_stderr": 0.0
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b70oscarpy/transformers",
|
14 |
+
"num_fewshot": 0,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b70oscarpy/evaluation/2b855b70oscarpy_1_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.07666666666666666,
|
5 |
+
"em_stderr": 0.004858411710742635
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b70oscarpy/transformers",
|
14 |
+
"num_fewshot": 1,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b70oscarpy/evaluation/2b855b70oscarpy_2_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.18033333333333335,
|
5 |
+
"em_stderr": 0.00702050591067492
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b70oscarpy/transformers",
|
14 |
+
"num_fewshot": 2,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b70oscarpy/evaluation/2b855b70oscarpy_3_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.22833333333333333,
|
5 |
+
"em_stderr": 0.007664984762185963
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b70oscarpy/transformers",
|
14 |
+
"num_fewshot": 3,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b70oscarpy/evaluation/2b855b70oscarpy_4_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.268,
|
5 |
+
"em_stderr": 0.00808788010553647
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b70oscarpy/transformers",
|
14 |
+
"num_fewshot": 4,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b70oscarpy/evaluation/2b855b70oscarpy_5_babi.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"results": {
|
3 |
"babi": {
|
4 |
-
"em": 0.
|
5 |
-
"em_stderr": 0.
|
6 |
}
|
7 |
},
|
8 |
"versions": {
|
@@ -14,8 +14,8 @@
|
|
14 |
"num_fewshot": 5,
|
15 |
"batch_size": null,
|
16 |
"device": null,
|
17 |
-
"no_cache":
|
18 |
-
"limit":
|
19 |
"bootstrap_iters": 100000,
|
20 |
"description_dict": {}
|
21 |
}
|
|
|
1 |
{
|
2 |
"results": {
|
3 |
"babi": {
|
4 |
+
"em": 0.27666666666666667,
|
5 |
+
"em_stderr": 0.008168821862685123
|
6 |
}
|
7 |
},
|
8 |
"versions": {
|
|
|
14 |
"num_fewshot": 5,
|
15 |
"batch_size": null,
|
16 |
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
"bootstrap_iters": 100000,
|
20 |
"description_dict": {}
|
21 |
}
|
2b855b80oscarpy/evaluation/2b855b80oscarpy_0_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.0,
|
5 |
+
"em_stderr": 0.0
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b70oscarpy/transformers",
|
14 |
+
"num_fewshot": 0,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b80oscarpy/evaluation/2b855b80oscarpy_1_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.07666666666666666,
|
5 |
+
"em_stderr": 0.004858411710742635
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b70oscarpy/transformers",
|
14 |
+
"num_fewshot": 1,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b80oscarpy/evaluation/2b855b80oscarpy_2_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.18033333333333335,
|
5 |
+
"em_stderr": 0.00702050591067492
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b70oscarpy/transformers",
|
14 |
+
"num_fewshot": 2,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b80oscarpy/evaluation/2b855b80oscarpy_3_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.22833333333333333,
|
5 |
+
"em_stderr": 0.007664984762185963
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b70oscarpy/transformers",
|
14 |
+
"num_fewshot": 3,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b80oscarpy/evaluation/2b855b80oscarpy_4_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.268,
|
5 |
+
"em_stderr": 0.00808788010553647
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b70oscarpy/transformers",
|
14 |
+
"num_fewshot": 4,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b80oscarpy/evaluation/2b855b80oscarpy_5_babi.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"results": {
|
3 |
"babi": {
|
4 |
-
"em": 0.
|
5 |
-
"em_stderr": 0.
|
6 |
}
|
7 |
},
|
8 |
"versions": {
|
@@ -14,8 +14,8 @@
|
|
14 |
"num_fewshot": 5,
|
15 |
"batch_size": null,
|
16 |
"device": null,
|
17 |
-
"no_cache":
|
18 |
-
"limit":
|
19 |
"bootstrap_iters": 100000,
|
20 |
"description_dict": {}
|
21 |
}
|
|
|
1 |
{
|
2 |
"results": {
|
3 |
"babi": {
|
4 |
+
"em": 0.27666666666666667,
|
5 |
+
"em_stderr": 0.008168821862685123
|
6 |
}
|
7 |
},
|
8 |
"versions": {
|
|
|
14 |
"num_fewshot": 5,
|
15 |
"batch_size": null,
|
16 |
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
"bootstrap_iters": 100000,
|
20 |
"description_dict": {}
|
21 |
}
|
2b855b90oscarpy/evaluation/2b855b90oscarpy_0_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.0,
|
5 |
+
"em_stderr": 0.0
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b90oscarpy/transformers",
|
14 |
+
"num_fewshot": 0,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b90oscarpy/evaluation/2b855b90oscarpy_1_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.078,
|
5 |
+
"em_stderr": 0.004896937137753452
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b90oscarpy/transformers",
|
14 |
+
"num_fewshot": 1,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b90oscarpy/evaluation/2b855b90oscarpy_2_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.17166666666666666,
|
5 |
+
"em_stderr": 0.006885846283046829
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b90oscarpy/transformers",
|
14 |
+
"num_fewshot": 2,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b90oscarpy/evaluation/2b855b90oscarpy_3_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.22266666666666668,
|
5 |
+
"em_stderr": 0.0075970155734085614
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b90oscarpy/transformers",
|
14 |
+
"num_fewshot": 3,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b90oscarpy/evaluation/2b855b90oscarpy_4_babi.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"results": {
|
3 |
+
"babi": {
|
4 |
+
"em": 0.25766666666666665,
|
5 |
+
"em_stderr": 0.007986203415903146
|
6 |
+
}
|
7 |
+
},
|
8 |
+
"versions": {
|
9 |
+
"babi": 0
|
10 |
+
},
|
11 |
+
"config": {
|
12 |
+
"model": "gpt2",
|
13 |
+
"model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b8-55b-oscarpy/2b855b90oscarpy/transformers",
|
14 |
+
"num_fewshot": 4,
|
15 |
+
"batch_size": null,
|
16 |
+
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
+
"bootstrap_iters": 100000,
|
20 |
+
"description_dict": {}
|
21 |
+
}
|
22 |
+
}
|
2b855b90oscarpy/evaluation/2b855b90oscarpy_5_babi.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"results": {
|
3 |
"babi": {
|
4 |
-
"em": 0.
|
5 |
-
"em_stderr": 0.
|
6 |
}
|
7 |
},
|
8 |
"versions": {
|
@@ -14,8 +14,8 @@
|
|
14 |
"num_fewshot": 5,
|
15 |
"batch_size": null,
|
16 |
"device": null,
|
17 |
-
"no_cache":
|
18 |
-
"limit":
|
19 |
"bootstrap_iters": 100000,
|
20 |
"description_dict": {}
|
21 |
}
|
|
|
1 |
{
|
2 |
"results": {
|
3 |
"babi": {
|
4 |
+
"em": 0.26866666666666666,
|
5 |
+
"em_stderr": 0.008094244986218791
|
6 |
}
|
7 |
},
|
8 |
"versions": {
|
|
|
14 |
"num_fewshot": 5,
|
15 |
"batch_size": null,
|
16 |
"device": null,
|
17 |
+
"no_cache": true,
|
18 |
+
"limit": 3000,
|
19 |
"bootstrap_iters": 100000,
|
20 |
"description_dict": {}
|
21 |
}
|