Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/Re-Align/just-eval
/ functions
Functions
19 in github.com/Re-Align/just-eval
⨍
Functions
19
◇
Types & classes
2
↓ 3 callers
Function
better_json_loads
(s)
just_eval/utils.py:176
↓ 2 callers
Function
shorten
(text, K=-1)
just_eval/evaluate.py:183
↓ 1 callers
Function
api
(ind, item, **kwargs)
just_eval/evaluate.py:127
↓ 1 callers
Function
completion_with_backoff
(**kwargs)
just_eval/utils.py:16
↓ 1 callers
Function
fix_inner_quotes
(s, filed="preference")
just_eval/utils.py:156
↓ 1 callers
Function
get_args
()
just_eval/evaluate.py:22
↓ 1 callers
Function
gpt_eval
(results, args)
just_eval/evaluate.py:95
↓ 1 callers
Function
main
()
just_eval/evaluate.py:234
↓ 1 callers
Function
openai_chat_request
Request the evaluation prompt from the OpenAI API in chat format. Args: prompt (str): The encoded prompt. messages (List[dict
just_eval/utils.py:95
↓ 1 callers
Function
report
(results, mode, args)
just_eval/evaluate.py:54
↓ 1 callers
Function
score_eval
(args)
just_eval/evaluate.py:189
Method
__init__
(self, model_name)
just_eval/utils.py:23
Method
__init__
(self, config)
just_eval/reward_model.py:8
Function
decorate
(func)
just_eval/utils.py:57
Method
forward
( # args are the same as LlamaForCausalLM self, input_ids: torch.LongTensor = None, at
just_eval/reward_model.py:14
Method
infer_generate
(self, input_text, max_tokens=2048)
just_eval/utils.py:27
Function
replacer
(match)
just_eval/utils.py:157
Function
retry_handler
This is an error handler for requests to OpenAI API. If will retry for the request for `retry_limit` times if the error is not a rate
just_eval/utils.py:42
Function
wrapper
(*args, **kwargs)
just_eval/utils.py:59