try1995

Results 4 comments of try1995

You forget the tanh function in the last computation in the part of def bottom_data_is():

self.state.h = self.state.o * np.tanh(self.state.s)

The tanh_derivative should be : 1. + values**2

[tanh(x)]' = [sin(x)/cos(x)]' = (cos(x)^2 + sin(x)^2)/cos(x)^2 = 1 + tanx(x)^2

stockfetch.fetch_stocks_selection处理异常：'NoneType' object is not subscriptable

我也是，看过去的issue说最新的修复了，但是我每天运行还是这样，我是在linux是定时跑的

Running the model using Ollama

yes, you can set step by step. - set .env file like ```python OPENAI_API_KEY=ollama MODEL="qwen2.5:32b" BASE_URL="http://localhost:11434/v1" ``` - change the code file "src\llm\models.py" ,line 96,get model func ```python elif model_provider...