File size: 1,433 Bytes
97d0285 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 | import json
import math
import matplotlib.pyplot as plt
from matplotlib.ticker import ScalarFormatter, LogLocator
losses = []
with open("models/losses.json", "r") as f:
losses += json.load(f)
print(losses[8])
avglosses = []
iters = []
chunksize = 1
iter = 100
while iter + chunksize < len(losses):
avg = 0
for i in range(round(chunksize)):
avg += losses[iter + i]
avg = avg / round(chunksize)
avglosses.append(avg) # riktig loss, inte log(avg)
iters.append(iter) # riktigt iterationsnummer
iter += round(chunksize)
chunksize *= 1.01
print(iter, avg)
fig, ax = plt.subplots()
ax.plot(iters, avglosses)
ax.set_xscale('log')
ax.set_yscale('log')
ax.set_xlabel("Iteration")
ax.set_ylabel("Träningsloss")
import numpy as np
# Sätt egna y-ticks baserat på datans faktiska min/max
ymin, ymax = min(avglosses), max(avglosses)
yticks = np.arange(math.floor(ymin*10)/10, math.ceil(ymax*10)/10 + 0.1, 0.1)
ax.set_yticks(yticks)
ax.yaxis.set_major_formatter(ScalarFormatter())
ax.yaxis.set_minor_formatter(plt.NullFormatter()) # slipp rörig minor-text
# Grid för både major och minor
ax.grid(True, which='major', linestyle='-', linewidth=0.7, alpha=0.7)
ax.grid(True, which='minor', linestyle=':', linewidth=0.5, alpha=0.4)
plt.tight_layout()
plt.title("Träningsloss för microbatch")
plt.show()
plt.close() |