ligeti commited on
Commit
9de6a1b
1 Parent(s): 2dc2d26

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -3
README.md CHANGED
@@ -11,7 +11,7 @@ tags:
11
  - promoter-prediction
12
  - phage
13
  ---
14
- ## ProkBERT-mini-long-phage Model
15
 
16
  This finetuned model is specifically designed for promoter identification and is based on the [ProkBERT-mini-c model](https://huggingface.co/neuralbioinfo/prokbert-mini-long).
17
 
@@ -37,14 +37,14 @@ The following example demonstrates how to use the ProkBERT-mini-promoter model f
37
  ```python
38
  from prokbert.prokbert_tokenizer import ProkBERTTokenizer
39
  from transformers import MegatronBertForSequenceClassification
40
- finetuned_model = "neuralbioinfo/prokbert-mini-long-phage"
41
  kmer = 1
42
  shift= 1
43
 
44
  tok_params = {'kmer' : kmer,
45
  'shift' : shift}
46
  tokenizer = ProkBERTTokenizer(tokenization_params=tok_params)
47
- model = BertForBinaryClassificationWithPooling.from_pretrained(finetuned_model)
48
  sequence = 'CACCGCATGGAGATCGGCACCTACTTCGACAAGCTGGAGGCGCTGCTGAAGGAGTGGTACGAGGCGCGCGGGGGTGAGGCATGACGGACTGGCAAGAGGAGCAGCGTCAGCGC'
49
  inputs = tokenizer(sequence, return_tensors="pt")
50
  # Ensure that inputs have a batch dimension
 
11
  - promoter-prediction
12
  - phage
13
  ---
14
+ ## ProkBERT-mini-c-phage Model
15
 
16
  This finetuned model is specifically designed for promoter identification and is based on the [ProkBERT-mini-c model](https://huggingface.co/neuralbioinfo/prokbert-mini-long).
17
 
 
37
  ```python
38
  from prokbert.prokbert_tokenizer import ProkBERTTokenizer
39
  from transformers import MegatronBertForSequenceClassification
40
+ finetuned_model = "neuralbioinfo/prokbert-mini-c-phage"
41
  kmer = 1
42
  shift= 1
43
 
44
  tok_params = {'kmer' : kmer,
45
  'shift' : shift}
46
  tokenizer = ProkBERTTokenizer(tokenization_params=tok_params)
47
+ model = MegatronBertForSequenceClassification.from_pretrained(finetuned_model)
48
  sequence = 'CACCGCATGGAGATCGGCACCTACTTCGACAAGCTGGAGGCGCTGCTGAAGGAGTGGTACGAGGCGCGCGGGGGTGAGGCATGACGGACTGGCAAGAGGAGCAGCGTCAGCGC'
49
  inputs = tokenizer(sequence, return_tensors="pt")
50
  # Ensure that inputs have a batch dimension