Bram Vanroy
commited on
Commit
·
d081f4d
1
Parent(s):
d64c180
rm llama/bloom evaluations
Browse files- evals/arc/arc_nl-bloom-7b1.json +0 -23
- evals/arc/arc_nl-llama-7B.json +0 -23
- evals/hellaswag/hellaswag_nl_bloom-7b1.json +0 -23
- evals/hellaswag/hellaswag_nl_llama-7B.json +0 -23
- evals/mmlu/mmlu_nl-bloom-7b1.json +0 -23
- evals/mmlu/mmlu_nl-llama-7B.json +0 -23
- evals/{dutch_models.json → models.json} +0 -0
- evals/truthfulqa/truthfulqa_nl-bloom-7b1.json +0 -23
- evals/truthfulqa/truthfulqa_nl-llama-7B.json +0 -23
- generate_overview_json.py +0 -0
evals/arc/arc_nl-bloom-7b1.json
DELETED
@@ -1,23 +0,0 @@
|
|
1 |
-
{
|
2 |
-
"results": {
|
3 |
-
"arc_nl": {
|
4 |
-
"acc": 0.1881950384944397,
|
5 |
-
"acc_stderr": 0.011436905010368727,
|
6 |
-
"acc_norm": 0.2309666381522669,
|
7 |
-
"acc_norm_stderr": 0.012331780770152612
|
8 |
-
}
|
9 |
-
},
|
10 |
-
"versions": {
|
11 |
-
"arc_nl": 0
|
12 |
-
},
|
13 |
-
"config": {
|
14 |
-
"model": "hf-auto",
|
15 |
-
"model_args": "pretrained=bigscience/bloom-7b1",
|
16 |
-
"batch_size": 1,
|
17 |
-
"device": "cuda",
|
18 |
-
"no_cache": false,
|
19 |
-
"limit": null,
|
20 |
-
"bootstrap_iters": 100000,
|
21 |
-
"description_dict": {}
|
22 |
-
}
|
23 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
evals/arc/arc_nl-llama-7B.json
DELETED
@@ -1,23 +0,0 @@
|
|
1 |
-
{
|
2 |
-
"results": {
|
3 |
-
"arc_nl": {
|
4 |
-
"acc": 0.32677502138579984,
|
5 |
-
"acc_stderr": 0.013724076021999824,
|
6 |
-
"acc_norm": 0.3361847733105218,
|
7 |
-
"acc_norm_stderr": 0.013822646555385164
|
8 |
-
}
|
9 |
-
},
|
10 |
-
"versions": {
|
11 |
-
"arc_nl": 0
|
12 |
-
},
|
13 |
-
"config": {
|
14 |
-
"model": "hf-auto",
|
15 |
-
"model_args": "pretrained=/sensei-fs/users/daclai/uoChatGPT/llama-7B",
|
16 |
-
"batch_size": 1,
|
17 |
-
"device": "cuda",
|
18 |
-
"no_cache": false,
|
19 |
-
"limit": null,
|
20 |
-
"bootstrap_iters": 100000,
|
21 |
-
"description_dict": {}
|
22 |
-
}
|
23 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
evals/hellaswag/hellaswag_nl_bloom-7b1.json
DELETED
@@ -1,23 +0,0 @@
|
|
1 |
-
{
|
2 |
-
"results": {
|
3 |
-
"hellaswag_nl": {
|
4 |
-
"acc": 0.28667026443604965,
|
5 |
-
"acc_stderr": 0.004698261813459453,
|
6 |
-
"acc_norm": 0.3172153264975715,
|
7 |
-
"acc_norm_stderr": 0.004835258421184045
|
8 |
-
}
|
9 |
-
},
|
10 |
-
"versions": {
|
11 |
-
"hellaswag_nl": 1
|
12 |
-
},
|
13 |
-
"config": {
|
14 |
-
"model": "hf-auto",
|
15 |
-
"model_args": "pretrained=bigscience/bloom-7b1",
|
16 |
-
"batch_size": "1",
|
17 |
-
"device": "cuda",
|
18 |
-
"no_cache": false,
|
19 |
-
"limit": null,
|
20 |
-
"bootstrap_iters": 100000,
|
21 |
-
"description_dict": {}
|
22 |
-
}
|
23 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
evals/hellaswag/hellaswag_nl_llama-7B.json
DELETED
@@ -1,23 +0,0 @@
|
|
1 |
-
{
|
2 |
-
"results": {
|
3 |
-
"hellaswag_nl": {
|
4 |
-
"acc": 0.38117850205050724,
|
5 |
-
"acc_stderr": 0.0050457320519523,
|
6 |
-
"acc_norm": 0.48748111374919056,
|
7 |
-
"acc_norm_stderr": 0.00519291390537233
|
8 |
-
}
|
9 |
-
},
|
10 |
-
"versions": {
|
11 |
-
"hellaswag_nl": 1
|
12 |
-
},
|
13 |
-
"config": {
|
14 |
-
"model": "hf-auto",
|
15 |
-
"model_args": "pretrained=/sensei-fs/users/daclai/uoChatGPT/llama-7B",
|
16 |
-
"batch_size": "1",
|
17 |
-
"device": "cuda",
|
18 |
-
"no_cache": false,
|
19 |
-
"limit": null,
|
20 |
-
"bootstrap_iters": 100000,
|
21 |
-
"description_dict": {}
|
22 |
-
}
|
23 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
evals/mmlu/mmlu_nl-bloom-7b1.json
DELETED
@@ -1,23 +0,0 @@
|
|
1 |
-
{
|
2 |
-
"results": {
|
3 |
-
"mmlu_nl": {
|
4 |
-
"acc": 0.25931547393185095,
|
5 |
-
"acc_stderr": 0.0038180275621108187,
|
6 |
-
"acc_norm": 0.2749487743796008,
|
7 |
-
"acc_norm_stderr": 0.003889720954246996
|
8 |
-
}
|
9 |
-
},
|
10 |
-
"versions": {
|
11 |
-
"mmlu_nl": 0
|
12 |
-
},
|
13 |
-
"config": {
|
14 |
-
"model": "hf-auto",
|
15 |
-
"model_args": "pretrained=bigscience/bloom-7b1",
|
16 |
-
"batch_size": 1,
|
17 |
-
"device": "cuda",
|
18 |
-
"no_cache": false,
|
19 |
-
"limit": null,
|
20 |
-
"bootstrap_iters": 100000,
|
21 |
-
"description_dict": {}
|
22 |
-
}
|
23 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
evals/mmlu/mmlu_nl-llama-7B.json
DELETED
@@ -1,23 +0,0 @@
|
|
1 |
-
{
|
2 |
-
"results": {
|
3 |
-
"mmlu_nl": {
|
4 |
-
"acc": 0.3053046975791151,
|
5 |
-
"acc_stderr": 0.004012103530956046,
|
6 |
-
"acc_norm": 0.2983987250512256,
|
7 |
-
"acc_norm_stderr": 0.003986133809323066
|
8 |
-
}
|
9 |
-
},
|
10 |
-
"versions": {
|
11 |
-
"mmlu_nl": 0
|
12 |
-
},
|
13 |
-
"config": {
|
14 |
-
"model": "hf-auto",
|
15 |
-
"model_args": "pretrained=/sensei-fs/users/daclai/uoChatGPT/llama-7B",
|
16 |
-
"batch_size": 1,
|
17 |
-
"device": "cuda",
|
18 |
-
"no_cache": false,
|
19 |
-
"limit": null,
|
20 |
-
"bootstrap_iters": 100000,
|
21 |
-
"description_dict": {}
|
22 |
-
}
|
23 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
evals/{dutch_models.json → models.json}
RENAMED
File without changes
|
evals/truthfulqa/truthfulqa_nl-bloom-7b1.json
DELETED
@@ -1,23 +0,0 @@
|
|
1 |
-
{
|
2 |
-
"results": {
|
3 |
-
"truthfulqa_nl": {
|
4 |
-
"mc1": 0.25477707006369427,
|
5 |
-
"mc1_stderr": 0.01556199397314563,
|
6 |
-
"mc2": 0.42677675918475044,
|
7 |
-
"mc2_stderr": 0.016186878668566846
|
8 |
-
}
|
9 |
-
},
|
10 |
-
"versions": {
|
11 |
-
"truthfulqa_nl": 1
|
12 |
-
},
|
13 |
-
"config": {
|
14 |
-
"model": "hf-auto",
|
15 |
-
"model_args": "pretrained=bigscience/bloom-7b1",
|
16 |
-
"batch_size": 1,
|
17 |
-
"device": "cuda",
|
18 |
-
"no_cache": false,
|
19 |
-
"limit": null,
|
20 |
-
"bootstrap_iters": 100000,
|
21 |
-
"description_dict": {}
|
22 |
-
}
|
23 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
evals/truthfulqa/truthfulqa_nl-llama-7B.json
DELETED
@@ -1,23 +0,0 @@
|
|
1 |
-
{
|
2 |
-
"results": {
|
3 |
-
"truthfulqa_nl": {
|
4 |
-
"mc1": 0.24331210191082803,
|
5 |
-
"mc1_stderr": 0.015324355488601135,
|
6 |
-
"mc2": 0.40023342153314656,
|
7 |
-
"mc2_stderr": 0.014679036703865582
|
8 |
-
}
|
9 |
-
},
|
10 |
-
"versions": {
|
11 |
-
"truthfulqa_nl": 1
|
12 |
-
},
|
13 |
-
"config": {
|
14 |
-
"model": "hf-auto",
|
15 |
-
"model_args": "pretrained=/sensei-fs/users/daclai/uoChatGPT/llama-7B",
|
16 |
-
"batch_size": 1,
|
17 |
-
"device": "cuda",
|
18 |
-
"no_cache": false,
|
19 |
-
"limit": null,
|
20 |
-
"bootstrap_iters": 100000,
|
21 |
-
"description_dict": {}
|
22 |
-
}
|
23 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
generate_overview_json.py
ADDED
File without changes
|