Spaces:
Sleeping
Sleeping
Add BiLSTMConfig and update BiLSTMAttentionBERT for improved model configuration and loading
Browse files- utils/model.py +12 -2
- utils/prediction.py +16 -1
utils/model.py
CHANGED
@@ -3,6 +3,8 @@ import torch.nn as nn
|
|
3 |
from transformers import PreTrainedModel, AutoModel, PretrainedConfig
|
4 |
|
5 |
class BiLSTMConfig(PretrainedConfig):
|
|
|
|
|
6 |
def __init__(self, hidden_dim=128, num_classes=22, num_layers=2, dropout=0.5, **kwargs):
|
7 |
super().__init__(**kwargs)
|
8 |
self.hidden_dim = hidden_dim
|
@@ -11,12 +13,20 @@ class BiLSTMConfig(PretrainedConfig):
|
|
11 |
self.dropout = dropout
|
12 |
|
13 |
class BiLSTMAttentionBERT(PreTrainedModel):
|
|
|
|
|
|
|
14 |
def __init__(self, config):
|
15 |
super().__init__(config)
|
16 |
self.config = config
|
17 |
self.bert = AutoModel.from_pretrained('dmis-lab/biobert-base-cased-v1.2')
|
18 |
-
self.lstm = nn.LSTM(
|
19 |
-
|
|
|
|
|
|
|
|
|
|
|
20 |
self.dropout = nn.Dropout(config.dropout)
|
21 |
self.fc = nn.Linear(config.hidden_dim * 2, config.num_classes)
|
22 |
|
|
|
3 |
from transformers import PreTrainedModel, AutoModel, PretrainedConfig
|
4 |
|
5 |
class BiLSTMConfig(PretrainedConfig):
|
6 |
+
model_type = "bilstm_attention"
|
7 |
+
|
8 |
def __init__(self, hidden_dim=128, num_classes=22, num_layers=2, dropout=0.5, **kwargs):
|
9 |
super().__init__(**kwargs)
|
10 |
self.hidden_dim = hidden_dim
|
|
|
13 |
self.dropout = dropout
|
14 |
|
15 |
class BiLSTMAttentionBERT(PreTrainedModel):
|
16 |
+
config_class = BiLSTMConfig
|
17 |
+
base_model_prefix = "bilstm_attention"
|
18 |
+
|
19 |
def __init__(self, config):
|
20 |
super().__init__(config)
|
21 |
self.config = config
|
22 |
self.bert = AutoModel.from_pretrained('dmis-lab/biobert-base-cased-v1.2')
|
23 |
+
self.lstm = nn.LSTM(
|
24 |
+
768,
|
25 |
+
config.hidden_dim,
|
26 |
+
config.num_layers,
|
27 |
+
batch_first=True,
|
28 |
+
bidirectional=True
|
29 |
+
)
|
30 |
self.dropout = nn.Dropout(config.dropout)
|
31 |
self.fc = nn.Linear(config.hidden_dim * 2, config.num_classes)
|
32 |
|
utils/prediction.py
CHANGED
@@ -1,16 +1,31 @@
|
|
1 |
-
from utils.model import BiLSTMAttentionBERT
|
2 |
import torch
|
3 |
from transformers import AutoTokenizer
|
4 |
from sklearn.preprocessing import LabelEncoder
|
5 |
import numpy as np
|
6 |
import streamlit as st
|
7 |
import requests
|
|
|
8 |
|
9 |
|
10 |
|
11 |
def load_model_for_prediction():
|
12 |
try:
|
13 |
st.write("Starting model loading...")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
14 |
|
15 |
# Test Hugging Face connectivity
|
16 |
st.write("Testing connection to Hugging Face...")
|
|
|
1 |
+
from utils.model import BiLSTMAttentionBERT, BiLSTMConfig
|
2 |
import torch
|
3 |
from transformers import AutoTokenizer
|
4 |
from sklearn.preprocessing import LabelEncoder
|
5 |
import numpy as np
|
6 |
import streamlit as st
|
7 |
import requests
|
8 |
+
from huggingface_hub import hf_hub_download
|
9 |
|
10 |
|
11 |
|
12 |
def load_model_for_prediction():
|
13 |
try:
|
14 |
st.write("Starting model loading...")
|
15 |
+
config = BiLSTMConfig(
|
16 |
+
hidden_dim=128,
|
17 |
+
num_classes=22,
|
18 |
+
num_layers=2,
|
19 |
+
dropout=0.5
|
20 |
+
)
|
21 |
+
|
22 |
+
model = BiLSTMAttentionBERT(config)
|
23 |
+
model_path = hf_hub_download(
|
24 |
+
repo_id="joko333/BiLSTM_v01",
|
25 |
+
filename="model_epoch8_acc72.53.pt"
|
26 |
+
)
|
27 |
+
state_dict = torch.load(model_path, map_location='cpu')
|
28 |
+
model.load_state_dict(state_dict)
|
29 |
|
30 |
# Test Hugging Face connectivity
|
31 |
st.write("Testing connection to Hugging Face...")
|