AutoModelForSeq2SeqLM.from_pretrained('facebook/nllb-200-1.3B') loads M2M100

Models
1

When I run this code

from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline
model = AutoModelForSeq2SeqLM.from_pretrained("facebook/nllb-200-distilled-600M")

where transformers version == 4.44.2

then I print the model object, I get

M2M100ForConditionalGeneration(
  (model): M2M100Model(
    (shared): M2M100ScaledWordEmbedding(256206, 1024, padding_idx=1)
    (encoder): M2M100Encoder(
      (embed_tokens): M2M100ScaledWordEmbedding(256206, 1024, padding_idx=1)
      (embed_positions): M2M100SinusoidalPositionalEmbedding()
      (layers): ModuleList(
        (0-11): 12 x M2M100EncoderLayer(
          (self_attn): M2M100Attention(
            (k_proj): Linear(in_features=1024, out_features=1024, bias=True)
            (v_proj): Linear(in_features=1024, out_features=1024, bias=True)
            (q_proj): Linear(in_features=1024, out_features=1024, bias=True)
            (out_proj): Linear(in_features=1024, out_features=1024, bias=True)
          )
.........

Any reason why my object is returning the M2M100 model instead of the NLLB model? I tried this inside a Jupyter notebook and my local env, and I get the same result.