Hugging Face's logo

genbio-ai
/
AIDO.DNA-7B

PyTorch
rnabert
biology
Model card Files Community
2
DianLiI commited on
Commit
9dcaa43
•
1 Parent(s): f78869b

Update instructions

Browse files
Files changed (1) hide show
  1. README.md +9 -9
README.md CHANGED
@@ -50,8 +50,8 @@ We evaluate the benefits of pretraining DNA FM 7B by conducting a comprehensive
50
  #### Embedding
51
  ```python
52
  from genbio_finetune.tasks import Embed
53
- model = Embed.from_config({"model.backbone": "dnafm"})
54
- collated_batch = model.collate({"sequences": ["ACGT", "ACGT"]})
55
  embedding = model(collated_batch)
56
  print(embedding.shape)
57
  print(embedding)
@@ -60,8 +60,8 @@ print(embedding)
60
  ```python
61
  import torch
62
  from genbio_finetune.tasks import SequenceClassification
63
- model = SequenceClassification.from_config({"model.backbone": "dnafm", "model.n_classes": 2})
64
- collated_batch = model.collate({"sequences": ["ACGT", "ACGT"]})
65
  logits = model(collated_batch)
66
  print(logits)
67
  print(torch.argmax(logits, dim=-1))
@@ -70,8 +70,8 @@ print(torch.argmax(logits, dim=-1))
70
  ```python
71
  import torch
72
  from genbio_finetune.tasks import TokenClassification
73
- model = TokenClassification.from_config({"model.backbone": "dnafm", "model.n_classes": 3})
74
- collated_batch = model.collate({"sequences": ["ACGT", "ACGT"]})
75
  logits = model(collated_batch)
76
  print(logits)
77
  print(torch.argmax(logits, dim=-1))
@@ -79,8 +79,8 @@ print(torch.argmax(logits, dim=-1))
79
  #### Regression
80
  ```python
81
  from genbio_finetune.tasks import SequenceRegression
82
- model = SequenceRegression.from_config({"model.backbone": "dnafm"})
83
- collated_batch = model.collate({"sequences": ["ACGT", "ACGT"]})
84
  logits = model(collated_batch)
85
  print(logits)
86
  ```
@@ -89,7 +89,7 @@ print(logits)
89
  gbft fit --model SequenceClassification --model.backbone dnafm --data SequenceClassification --data.path <hf_or_local_path_to_your_dataset>
90
  gbft test --model SequenceClassification --model.backbone dnafm --data SequenceClassification --data.path <hf_or_local_path_to_your_dataset>
91
  ```
92
- For more information, visit: [Model Generator](https://github.com/genbio-ai/test)
93
 
94
 
95
  ## Citation
 
50
  #### Embedding
51
  ```python
52
  from genbio_finetune.tasks import Embed
53
+ model = Embed.from_config({"model.backbone": "dnafm"}).eval()
54
+ collated_batch = model.collate({"sequences": ["ACGT", "AGCT"]})
55
  embedding = model(collated_batch)
56
  print(embedding.shape)
57
  print(embedding)
 
60
  ```python
61
  import torch
62
  from genbio_finetune.tasks import SequenceClassification
63
+ model = SequenceClassification.from_config({"model.backbone": "dnafm", "model.n_classes": 2}).eval()
64
+ collated_batch = model.collate({"sequences": ["ACGT", "AGCT"]})
65
  logits = model(collated_batch)
66
  print(logits)
67
  print(torch.argmax(logits, dim=-1))
 
70
  ```python
71
  import torch
72
  from genbio_finetune.tasks import TokenClassification
73
+ model = TokenClassification.from_config({"model.backbone": "dnafm", "model.n_classes": 3}).eval()
74
+ collated_batch = model.collate({"sequences": ["ACGT", "AGCT"]})
75
  logits = model(collated_batch)
76
  print(logits)
77
  print(torch.argmax(logits, dim=-1))
 
79
  #### Regression
80
  ```python
81
  from genbio_finetune.tasks import SequenceRegression
82
+ model = SequenceRegression.from_config({"model.backbone": "dnafm"}).eval()
83
+ collated_batch = model.collate({"sequences": ["ACGT", "AGCT"]})
84
  logits = model(collated_batch)
85
  print(logits)
86
  ```
 
89
  gbft fit --model SequenceClassification --model.backbone dnafm --data SequenceClassification --data.path <hf_or_local_path_to_your_dataset>
90
  gbft test --model SequenceClassification --model.backbone dnafm --data SequenceClassification --data.path <hf_or_local_path_to_your_dataset>
91
  ```
92
+ For more information, visit: [Model Generator](https://github.com/genbio-ai/modelgenerator)
93
 
94
 
95
  ## Citation