question about gene2idx order

[hongruhu]

to extract gene embeddings, I was trying to use three models: model_types = ['whole_human', 'human_brain', 'human_blood'], which means, I have 3 vocab.json files

I briefly checked the files themselves by looking at their last few lines:
(1) for whole_human:

  "RP11-390P2.2": 42033,
  "SGSM3-AS1": 31087,
  "RP11-812I20.2": 42036,
  "AC022154.7": 828,
  "RBMXP1": 42041
}

(2) for human_brain:

  "SETP7": 53181,
  "RP11-175P13.3": 53185,
  "HSPA8P14": 53196,
  "BTG1P1": 53065,
  "OR7E47P": 53197
}

(3) for human_blood:

  "PXT1": 20738,
  "BAG3": 3005,
  "PYCR3": 20744,
  "TAB3": 32782,
  "PYDC1": 20745
}

when I checked the gene2idx using

vocab_file = model_dir / "vocab.json"
vocab = GeneVocab.from_file(vocab_file)
gene2idx = vocab.get_stoi()

I was assuming they would return the consistent order with their corresponding json file order,
but both (1) and (2) return the identical gene2idx which matches the (1) whole_human's gene order:

... 'RP11-390P2.2': 42033, 'SGSM3-AS1': 31087, 'RP11-812I20.2': 42036, 'AC022154.7': 828, 'RBMXP1': 42041}

while (3) is consistent with its own json file:

... 'PXT1': 20738, 'BAG3': 3005, 'PYCR3': 20744, 'TAB3': 32782, 'PYDC1': 20745}

I was just confused why the order of (2) is not consistent with its own json file but identical to (1)'s? Did I miss something? Just wanted to make sure the order of gene index of each model's extracted gene embeddings is correct.

Thanks!