Files changed (1) hide show
  1. README.md +15 -27
README.md CHANGED
@@ -1,31 +1,6 @@
1
  ---
2
- pretty_name: Evaluation run of TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ
3
- dataset_summary: "Dataset automatically created during the evaluation run of model\
4
- \ [TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ](https://huggingface.co/TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ)\
5
- \ on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).\n\
6
- \nThe dataset is composed of 3 configuration, each one coresponding to one of the\
7
- \ evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be\
8
- \ found as a specific split in each configuration, the split being named using the\
9
- \ timestamp of the run.The \"train\" split is always pointing to the latest results.\n\
10
- \nAn additional configuration \"results\" store all the aggregated results of the\
11
- \ run (and is used to compute and display the aggregated metrics on the [Open LLM\
12
- \ Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).\n\
13
- \nTo load the details from a run, you can for instance do the following:\n```python\n\
14
- from datasets import load_dataset\ndata = load_dataset(\"open-llm-leaderboard/details_TheBloke__Wizard-Vicuna-13B-Uncensored-GPTQ_public\"\
15
- ,\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\n\
16
- These are the [latest results from run 2023-11-07T13:33:21.987098](https://huggingface.co/datasets/open-llm-leaderboard/details_TheBloke__Wizard-Vicuna-13B-Uncensored-GPTQ_public/blob/main/results_2023-11-07T13-33-21.987098.json)(note\
17
- \ that their might be results for other tasks in the repos if successive evals didn't\
18
- \ cover the same tasks. You find each in the results and the \"latest\" split for\
19
- \ each eval):\n\n```python\n{\n \"all\": {\n \"em\": 0.07581795302013423,\n\
20
- \ \"em_stderr\": 0.0027108434788949637,\n \"f1\": 0.14551698825503265,\n\
21
- \ \"f1_stderr\": 0.0029781067344765754,\n \"acc\": 0.42850650411417607,\n\
22
- \ \"acc_stderr\": 0.010140361969429381\n },\n \"harness|drop|3\": {\n\
23
- \ \"em\": 0.07581795302013423,\n \"em_stderr\": 0.0027108434788949637,\n\
24
- \ \"f1\": 0.14551698825503265,\n \"f1_stderr\": 0.0029781067344765754\n\
25
- \ },\n \"harness|gsm8k|5\": {\n \"acc\": 0.09931766489764973,\n \
26
- \ \"acc_stderr\": 0.008238371412683973\n },\n \"harness|winogrande|5\"\
27
- : {\n \"acc\": 0.7576953433307024,\n \"acc_stderr\": 0.012042352526174787\n\
28
- \ }\n}\n```"
29
  repo_url: https://huggingface.co/TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ
30
  leaderboard_url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard
31
  point_of_contact: clementine@hf.co
@@ -62,6 +37,19 @@ configs:
62
  - split: latest
63
  path:
64
  - results_2023-11-07T13-33-21.987098.parquet
 
 
 
 
 
 
 
 
 
 
 
 
 
65
  ---
66
 
67
  # Dataset Card for Evaluation run of TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ
 
1
  ---
2
+ pretty_name: red-GPTQ
3
+ dataset_summary: "Dataset automatically created during the evaluation run of model [TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ](https://huggingface.co/TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).\n\nThe dataset is composed of 3 configuration, each one coresponding to one of the evaluated task.\n\nThe dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The \"train\" split is always pointing to the latest results.\n\nAn additional configuration \"results\" store all the aggregated results of the run (and is used to compute and display the aggregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).\n\nTo load the details from a run, you can for instance do the following:\n```python\nfrom datasets import load_dataset\ndata = load_dataset(\"open-llm-leaderboard/details_TheBloke__Wizard-Vicuna-13B-Uncensored-GPTQ_public\",\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\nThese are the [latest results from run 2023-11-07T13:33:21.987098](https://huggingface.co/datasets/open-llm-leaderboard/details_TheBloke__Wizard-Vicuna-13B-Uncensored-GPTQ_public/blob/main/results_2023-11-07T13-33-21.987098.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the \"latest\" split for each eval):\n\n```python\n{\n \"all\": {\n \"em\": 0.07581795302013423,\n \"em_stderr\": 0.0027108434788949637,\n \"f1\": 0.14551698825503265,\n \"f1_stderr\": 0.0029781067344765754,\n \"acc\": 0.42850650411417607,\n \"acc_stderr\": 0.010140361969429381\n },\n \"harness|drop|3\": {\n \"em\": 0.07581795302013423,\n \"em_stderr\": 0.0027108434788949637,\n \"f1\": 0.14551698825503265,\n \"f1_stderr\": 0.0029781067344765754\n },\n \"harness|gsm8k|5\": {\n \"acc\": 0.09931766489764973,\n \"acc_stderr\": 0.008238371412683973\n },\n \"harness|winogrande|5\": {\n \"acc\": 0.7576953433307024,\n \"acc_stderr\": 0.012042352526174787\n }\n}\n```"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4
  repo_url: https://huggingface.co/TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ
5
  leaderboard_url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard
6
  point_of_contact: clementine@hf.co
 
37
  - split: latest
38
  path:
39
  - results_2023-11-07T13-33-21.987098.parquet
40
+ license: mit
41
+ task_categories:
42
+ - table-question-answering
43
+ - text-classification
44
+ - token-classification
45
+ - question-answering
46
+ - translation
47
+ - text2text-generation
48
+ - fill-mask
49
+ language:
50
+ - en
51
+ size_categories:
52
+ - n>1T
53
  ---
54
 
55
  # Dataset Card for Evaluation run of TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ