Comments (1)
Hi @Glupapa,
Thank you for your interest in our work. The main reason of this type of fluctuation could be attributed to the dynamic nature of LLM. I would recommend reporting standard deviation by performing inference multiple times in this case.
I tried running RefSeg benchmarks multiple times on RefCOCO and following are the resutls,
# RUN 1
[
{
"model": "results/refseg",
"dataset": "refcoco|val",
"giou": "0.80515635",
"ciou": "0.7929469"
},
{
"model": "results/refseg",
"dataset": "refcoco|testA",
"giou": "0.83403975",
"ciou": "0.83168846"
},
{
"model": "results/refseg",
"dataset": "refcoco|testB",
"giou": "0.776042",
"ciou": "0.7664252"
}
]
# RUN 2
[
{
"model": "results/refseg",
"dataset": "refcoco|val",
"giou": "0.8066321",
"ciou": "0.7977847"
},
{
"model": "results/refseg",
"dataset": "refcoco|testA",
"giou": "0.8239293",
"ciou": "0.81979144"
},
{
"model": "results/refseg",
"dataset": "refcoco|testB",
"giou": "0.7788188",
"ciou": "0.761931"
}
]
# RUN 3
[
{
"model": "results/refseg",
"dataset": "refcoco|val",
"giou": "0.8075647",
"ciou": "0.80325586"
},
{
"model": "results/refseg",
"dataset": "refcoco|testA",
"giou": "0.83058774",
"ciou": "0.82961744"
},
{
"model": "results/refseg",
"dataset": "refcoco|testB",
"giou": "0.7687606",
"ciou": "0.75880194"
}
]
In summary,
ciou (val) - 79.79 with std of 0.419
ciou (testA) - 82.70 with std of 0.519
ciou (testB) - 76.23 with std of 0.312
Note that the standard deviations are ranging from 0.3 to 0.6 approx. which is quite reasonable. In some runs (for example Run # 1 & Run # 2 for testA), you may notice a fluctuation of more than 1.0 points, however running the experiment again normalize it with an overall fluctuation of around 0.6 points.
I hope this will help, please let me know if you have any questions. Thank You and Good Luck!
from groundinglmm.
Related Issues (20)
- may i ask your total parameter?
- Some bugs in the GranD_ReferringSegm_ds.py
- Online Demo Down HOT 1
- Fine-tuning Grounded Conversation Generation (GCG) Task HOT 4
- token_positives HOT 2
- assertion error cur_len == total_len HOT 1
- can not install mmcv HOT 2
- Can not find file for glamm_conda_env.zip in the given Google Drive Link HOT 4
- Training on New Data HOT 2
- training V-L and L-P projection layer HOT 1
- Can not download the train.json file for visual genome
- How can I let the model receive multiple images at once HOT 1
- How should I train on the GranD dataset
- How can I finetune on combined tasks?
- Confusing referring segmentation results. HOT 1
- mmcv failed to install HOT 1
- AssertionError when running a demo
- Offline demo error
- Why is it that during the computation of segmentation results, the model() function is used instead of model.generate()? Wouldn't this mean that when predicting the next token, the information viewed is from the actual token rather than the predicted one? HOT 1
- What are the ‘categories’ in the dataset used for? When would I use them?
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 groundinglmm.