Spaces:

---

Dianor

/

AiTradingCrypto

like 8

Running

Hello, it's a beautiful visual analysis AI, but I'm still not clear on the results. I fed your AI a screenshot of your demo: the value of the result obtained is not at all similar to the one shown in the demo, why?

hi, the model in the demo video was trained on a slightly different architecture. lets say its old model which was here before. i just didnt update the demo video.

and just in case, if u will send screenshot here and enter trade its not a good idea. for good trading need to use second pretrained model for higher timeframe and many other things.

this space is just for showing technology.

currently i am trying to use this model to train an RL agent which will be fully autonomous trading system. there is a big roadmap ahead

Hello, it's a beautiful visual analysis AI, but I'm still not clear on the results. I fed your AI a screenshot of your demo: the value of the result obtained is not at all similar to the one shown in the demo, why?

hi, the model in the demo video was trained on a slightly different architecture. lets say its old model which was here before. i just didnt update the demo video.

and just in case, if u will send screenshot here and enter trade its not a good idea. for good trading need to use second pretrained model for higher timeframe and many other things.

this space is just for showing technology.

currently i am trying to use this model to train an RL agent which will be fully autonomous trading system. there is a big roadmap ahead

But is the training you do always visual on the new model?

I am testing this system, on TF at 1 minute to identify the cryptocurrency pump. I take screenshots from trading view and feed it to your AI. But I saw that the values ​​of the signals were always low and never decisive (long/short were always the same as values).

Look at this image, he gives low values, and indecision, but within 4 candles later the system made a + 9% in a few subsequent candles...is there a chance I can help you?

Look at this image, he gives low values, and indecision, but within 4 candles later the system made a + 9% in a few subsequent candles...is there a chance I can help you?

actually this chart should work, the situation seems kinda same when model have to show high confidence in long
let me check

Look at this image, he gives low values, and indecision, but within 4 candles later the system made a + 9% in a few subsequent candles...is there a chance I can help you?

yes it works. i did augmentation when were training so some noise on chart (your grid for example) should not affect on results so much. but it still cant give perfect results if maybe u upload with white boarders or if there is empty space on the right part of the chart and so on.
chart should looks similar as examples attached on main page

your idea is interesting actually

but still be careful pls :)
i did not train for exactly this way. it can make false signals

check i cropped your chart how it should be and upload to model to get a prediction

Look at this image, he gives low values, and indecision, but within 4 candles later the system made a + 9% in a few subsequent candles...is there a chance I can help you?

yes it works. i did augmentation when were training so some noise on chart (your grid for example) should not affect on results so much. but it still cant give perfect results if maybe u upload with white boarders or if there is empty space on the right part of the chart and so on.
chart should looks similar as examples attached on main page

your idea is interesting actually

but still be careful pls :)
i did not train for exactly this way. it can make false signals

check i cropped your chart how it should be and upload to model to get a prediction

Ok, thanks. I will adjust the chart to have as little noise as possible. If you want I can give you many images. I created a program that takes an image from tradingview when a 1 minute tf candle exceeds 2% (SPIKE or FOMO)

do you have a discord account? can we hear them?