vedantM/16lens_89M_gt_lr_1e-04_genus

seqSight_4096_512_89M-at-base-multi_16s_gt_lr_1e-04_genus

This model uses a custom GuidedTokenizer with motif-aware tokenization for genomic sequences.

Usage

from transformers import AutoModelForSequenceClassification
from guidedTokenizer import GuidedTokenizer  # Make sure to have the custom tokenizer class

# Load model and tokenizer
model = AutoModelForSequenceClassification.from_pretrained("vedantM/seqSight_4096_512_89M-at-base-multi_16s_gt_lr_1e-04_genus")
tokenizer = GuidedTokenizer.from_pretrained("vedantM/seqSight_4096_512_89M-at-base-multi_16s_gt_lr_1e-04_genus")

# The tokenizer automatically loads the motifs from motif_config.json
print(f"Loaded {len(tokenizer.motifs)} motifs")

Files included

• Standard model files (config.json, pytorch_model.bin, etc.)
• Standard tokenizer files (tokenizer.json, tokenizer_config.json, etc.)
• motif_config.json: Contains the motifs and configuration for the GuidedTokenizer

Note

This model requires the custom GuidedTokenizer class to work properly. Make sure you have access to the guidedTokenizer.py file.