- 03 Jul, 2023 1 commit
-
-
Hailey Schoelkopf authored
-
- 31 May, 2023 1 commit
-
-
Wang, Yi authored
Signed-off-by:Wang, Yi <yi.a.wang@intel.com>
-
- 30 May, 2023 1 commit
-
-
Sam Passaglia authored
-
- 27 May, 2023 1 commit
-
-
Stella Biderman authored
Add support for loading GPTQ models via AutoGPTQ
-
- 26 May, 2023 1 commit
-
-
gk authored
-
- 25 May, 2023 2 commits
-
-
Hailey Schoelkopf authored
* allow for hf-causal to take dtype arg * document this change
-
gk authored
-
- 23 May, 2023 4 commits
-
-
Lintang Sutawika authored
-
lintangsutawika authored
-
lintangsutawika authored
-
gk authored
-
- 21 May, 2023 4 commits
-
-
Stella Biderman authored
Add perplexity task on arbitrary JSON data
-
Stella Biderman authored
-
Stella Biderman authored
Add option to dump prompts and completions to a JSON file
-
Stella Biderman authored
Evaluation Against Portion of Benchmark Data
-
- 19 May, 2023 2 commits
-
-
Stella Biderman authored
Add results of various models in json and md format
-
Stella Biderman authored
Create output path directory if necessary
-
- 18 May, 2023 1 commit
-
-
Stella Biderman authored
-
- 12 May, 2023 1 commit
-
-
Julen Etxaniz authored
-
- 11 May, 2023 3 commits
-
-
Julen Etxaniz authored
* add xcopa dataset * add xstory_cloze dataset and run pre-commit * fix xcopa validation and test sets * add xwinograd dataset * add pawsx task * add xnli task * update task table with recently added tasks * remove unused metrics from paws-x * add mgsm task and fix gsm8k * fix gsm8k until * update task table
-
Julen Etxaniz authored
-
Julen Etxaniz authored
-
- 10 May, 2023 4 commits
-
-
Stella Biderman authored
fix adaptive batch crash when there are no new requests
-
Stella Biderman authored
-
Julen Etxaniz authored
-
Jeffrey Quesnelle authored
-
- 09 May, 2023 2 commits
-
-
Stella Biderman authored
-
Stella Biderman authored
-
- 08 May, 2023 3 commits
-
-
janEbert authored
-
janEbert authored
-
Julen Etxaniz authored
-
- 07 May, 2023 2 commits
-
-
Stella Biderman authored
Set PAD token to EOS token
-
Ken Tsui authored
When `limit` is <1, limit represents the percentage of the total number of examples. If it is >=1, then it means the number of examples per task (only use this for testing).
-
- 06 May, 2023 5 commits
-
-
Julen Etxaniz authored
-
Julen Etxaniz authored
-
Julen Etxaniz authored
-
Stella Biderman authored
-
Julen Etxaniz authored
-
- 05 May, 2023 2 commits
-
-
Julen Etxaniz authored
This makes comparing the results of different models easier because tasks are ordered in the same way.
-
Nikhil Pinnaparaju authored
-