Comments (6)
I also encountered similar problems
I ran the following command multiple times:
alpaca_eval --model_outputs "Mistral-7B+RAHF-DUAL+LoRA.json" --output_path "Mistral-7B+RAHF-DUAL+LoRA_gpt4" ----annotators_config "weighted_alpaca_eval_gpt4_turbo" --reference_outputs "reference_outputs.json" --is_reapply_parsing True
But the result is that only more than 200 examples are evaluated.
from alpaca_eval.
I also encountered similar problems I ran the following command multiple times:
alpaca_eval --model_outputs "Mistral-7B+RAHF-DUAL+LoRA.json" --output_path "Mistral-7B+RAHF-DUAL+LoRA_gpt4" ----annotators_config "weighted_alpaca_eval_gpt4_turbo" --reference_outputs "reference_outputs.json" --is_reapply_parsing True
But the result is that only more than 200 examples are evaluated.
Can you access OpenAI successfully?
Unfortunately, I only have intermittent access to it.
"INFO:root:Sleeping 2 before retrying to call openai API..."
I don't know if this is the reason for the incomplete evaluation.
from alpaca_eval.
Can either of you reproduce this result using OpenAI rather than through a proxy server? In my experience (from previous gh issues) the problem always comes from the proxy server and I can't help for that.
For example, many proxy servers don't return logprobs. Can you share the warnings/errors that you see during annotation?
from alpaca_eval.
E.g. @LiuAmber the fact that you could perform AlpacaEval 1 here: #307, strongly suggests to me that the error comes from the logprobs (necessary for AlapcaEval 2 but not 1) and that this comes from your proxy server.
from alpaca_eval.
I'm assuming that the issue is the fact that your proxy doesn't have access to logprobs. If so try:
alpaca_eval evaluate_from_model --model_configs '...' --annotators_config 'alpaca_eval_clf_gpt4_turbo'
Feel free to reopen if your issue is different
from alpaca_eval.
Can either of you reproduce this result using OpenAI rather than through a proxy server? In my experience (from previous gh issues) the problem always comes from the proxy server and I can't help for that.
For example, many proxy servers don't return logprobs. Can you share the warnings/errors that you see during annotation?
Thank you. I solved this problem. The problem is caused by proxy servers and the setting of annotators_config.
from alpaca_eval.
Related Issues (20)
- With unstable GPT-4 API, I encounterd a tricky problem HOT 1
- With unstable GPT-4 API, I encounterd a tricky problem HOT 2
- Question about the GPT-4 API HOT 11
- openai_configs.yaml when using Azure only HOT 7
- Confusion in Model Evaluation Results Due to GPT Updates HOT 2
- How instruction_difficulty feature is obtained HOT 1
- Llama-3-Instruct not using official prompt template? HOT 1
- Unable to reproduce results HOT 1
- Missing item in results/llama-2-70b-chat-hf HOT 1
- Missing result file in notebook HOT 1
- How to solve the problem of null appearing in the evaluation results HOT 4
- What to do if the log prob is not returned? HOT 7
- How to get the LC Win Rate in AlpacaEval 1.0 version? HOT 4
- How to change cache path when evaluating multi-models HOT 1
- Overly High Win Rate for Alpaca v2 on mistral 7b orpo HOT 12
- How to use AE1 to evaluate model HOT 2
- possibility of adding llama3-70b as the evaluator? HOT 3
- Alpaca Evaluation Instruction Difficulty used also for Custom Evaluation Dataset HOT 3
- Add Phi 3 models HOT 3
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from alpaca_eval.