MegaDNA Models

In [1]:

## Make sure you have installed the dependencies for megaDNA model
# !git clone https://github.com/lingxusb/megaDNA
# !cd megaDNA
# !uv pip install -e .

## Make sure you have installed the dependencies for megaDNA model # !git clone https://github.com/lingxusb/megaDNA # !cd megaDNA # !uv pip install -e .

In [2]:

from dnallm import load_config
from dnallm import load_model_and_tokenizer, DNAInference

from dnallm import load_config from dnallm import load_model_and_tokenizer, DNAInference

In [3]:

# Load configurations
configs = load_config("./inference_megaDNA_config.yaml")

# Load configurations configs = load_config("./inference_megaDNA_config.yaml")

In [ ]:

model_name = "lingxusb/megaDNA_updated"
model, tokenizer = load_model_and_tokenizer(model_name, task_config=configs['task'], source="huggingface")

model_name = "lingxusb/megaDNA_updated" model, tokenizer = load_model_and_tokenizer(model_name, task_config=configs['task'], source="huggingface")

2025-10-10 02:38:15,906 - modelscope - INFO - Not logged-in, you can login for uploadingor accessing controlled entities.

Downloading Model from https://www.modelscope.cn to directory: /home/liuguanqing/.cache/modelscope/hub/models/lgq12697/megaDNA_updated
02:38:16 - dnallm.models.model - INFO - Model files are stored in /home/liuguanqing/.cache/modelscope/hub/models/lgq12697/megaDNA_updated

In [5]:

# Create inference engine
inference_engine = DNAInference(
    model=model,
    tokenizer=tokenizer,
    config=configs
)

# Create inference engine inference_engine = DNAInference( model=model, tokenizer=tokenizer, config=configs )

02:38:17 - dnallm.models.model - INFO - Using device: cuda

In [6]:

output = inference_engine.generate(["ACGT"], n_tokens=1024, temperature=0.95, top_p=0.1)

output = inference_engine.generate(["ACGT"], n_tokens=1024, temperature=0.95, top_p=0.1)

  0%|          | 0/1020 [00:00<?, ?it/s]

/home/liuguanqing/miniforge3/lib/python3.12/contextlib.py:105: FutureWarning: `torch.backends.cuda.sdp_kernel()` is deprecated. In the future, this context manager will be removed. Please see `torch.nn.attention.sdpa_kernel()` for the new context manager, with updated signature.
  self.gen = func(*args, **kwds)

In [7]:

for seq in output:
    print(f"Input Sequence: {seq['Prompt']}")
    print(f"Generated Sequence: {seq['Output']}")
    print()

for seq in output: print(f"Input Sequence: {seq['Prompt']}") print(f"Generated Sequence: {seq['Output']}") print()

Input Sequence: ACGT
Generated Sequence: ACGTGTGGCTCGAATATGGGCAGATTGAGCAGCTTGAGCGGCTCGTAAATGAAAGCGGAATTGACATTGCAAATACCGACGACCACTATTACTTGAAAATTGATGAGGTGTAAATGATGGACGCTACAATGAACTATAACACGTTGCCCTATCTTTCCAAGGCGCAGAACGGAGTGACTGAGAGGCTCCAAAAGGGAGACATCACGCAGAAAGATATCTTGCAGGACATTTTGTCAGACGGTTTCTTTTTTTTGAAGATTAAAGGGCGGAGCTTTGCGGATGCGTCTGAAAAACTGGCTCATCATGGGCGAAAAAAGACAGTCGCTAATATTGACATCTCAAAGTTGTTAAATGCGGCGGAAGATAAGGGCTTTATGAACCGTTACCTAAAGCTCGTGGATTCGCTACAGAGAATAATCAAGGGAGGGGATGTTGGAATCTTGGACGGCTATCAAAACGGCACTGAATTCGACAGCGTATTGGCTAGATGCATTGACAAGTTTAATATACAAGACTTCATGAGACCTGCAAACGAGGGAAGAGGCGTGGAAAGTATATCTTTGACGGCTTTGCTTACCCGTCAAAAGCAAAAAGCCAAGTATTCCAAGATGCTTCTGCGAATGCCGGCTGCGGAATTGAAAAATGAATTGGCAAAGCTGTGCCCCGACTGCAAAGATAGAAAAGAGAAATACAAATTACCCAGATGGTTAGCTTCAAATAATAAAGATGCATGGCTTTATTCTTATATGCCTGATGTGGGCAAACGTTGGAATGTCAATAATGCGACGCTTGTAGAATGAGGAATTAACAATATCAGAGGGAGGATTCCCGTAGCCGCCTTTCTTATTTCTTTCGATACCGAATGAAATCGATGATCTCACCGGGCTTGAGCTTGACGATGCAGGACTGGATTTTGGAGATTCCCTGTCGGCGCATGATTCCGACATTGAGGCAGCGCAGTTTTTCCCGGCGCATTTCTTCAACAGGTTTGCGCCCGGCCTTGGCCTCCGCCGTAAGATAGTTC

In [8]:

scores = inference_engine.scoring(["ATCCGCATG", "ATGCGCATG"])
for res in scores:
    print(f"Input Sequence: {res['Input']}")
    print(f"Score: {res['Score']}")
    print()

scores = inference_engine.scoring(["ATCCGCATG", "ATGCGCATG"]) for res in scores: print(f"Input Sequence: {res['Input']}") print(f"Score: {res['Score']}") print()

Input Sequence: ATCCGCATG
Score: 1.4011831283569336

Input Sequence: ATGCGCATG
Score: 1.4185738563537598

In [ ]: