pySR查找缩放定律

定律如下：

代码如下：

import numpy as np
import pandas as pd
from pysr import PySRRegressor
from multiprocessing import cpu_count

fname = "combined_llama2.csv"
df = pd.read_csv(fname)

model = PySRRegressor(
    model_selection="best",
    populations=30,
    population_size=100,
    niterations=100,
    maxsize = 20,
    ncyclesperiteration=10000,
    weight_optimize=0.001,
    adaptive_parsimony_scaling=1000,
    binary_operators=["+","-","*","/","^"],
    constraints={"^":(-1,3),"/":(-1,5)},
    multithreading=False,
    procs=cpu_count(),
    turbo = True
)

model.fit(df[["context","params"]],df["loss"])

latex_output = "table.txt"
print("I will save a LaTeX table to table.tex")
s = model.latex_table()
with open(latex_output,"w") as f:
    f.write(s)

 结果如下：