Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,41 +1,74 @@
|
|
| 1 |
import torch
|
| 2 |
import torch.nn as nn
|
|
|
|
| 3 |
import gradio as gr
|
| 4 |
from huggingface_hub import hf_hub_download
|
| 5 |
|
| 6 |
REPO_ID = "SentilyticsOPJ/lstm_model"
|
| 7 |
FILENAME = "lstm_model.pt"
|
| 8 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
|
| 10 |
-
|
| 11 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
super().__init__()
|
| 13 |
-
self.
|
| 14 |
-
|
| 15 |
-
self.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
|
| 17 |
def forward(self, x):
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
|
| 23 |
path = hf_hub_download(repo_id=REPO_ID, filename=FILENAME)
|
| 24 |
obj = torch.load(path, map_location="cpu", weights_only=False)
|
| 25 |
sd = obj["state_dict"]
|
| 26 |
|
| 27 |
-
vocab_size, embed_dim = sd["embedding.weight"].shape
|
| 28 |
-
hidden_dim = sd["lstm.weight_hh_l0"].shape[1]
|
| 29 |
-
num_classes = sd["fc.weight"].shape[0]
|
| 30 |
-
|
| 31 |
-
model = TextBiLSTM(vocab_size, embed_dim, hidden_dim, num_classes)
|
| 32 |
-
model.load_state_dict(sd)
|
| 33 |
-
model.eval()
|
| 34 |
-
|
| 35 |
word2id = obj["word2id"]
|
| 36 |
labels = obj["label_classes"]
|
| 37 |
max_len = obj["max_len"]
|
| 38 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
|
| 40 |
def predict(text):
|
| 41 |
if not text.strip():
|
|
@@ -46,4 +79,5 @@ def predict(text):
|
|
| 46 |
probs = torch.softmax(model(torch.tensor(ids).unsqueeze(0)), dim=1)[0]
|
| 47 |
return {labels[i]: float(probs[i]) for i in range(len(labels))}
|
| 48 |
|
|
|
|
| 49 |
gr.Interface(fn=predict, inputs="text", outputs="text").launch()
|
|
|
|
| 1 |
import torch
|
| 2 |
import torch.nn as nn
|
| 3 |
+
import torch.nn.functional as F
|
| 4 |
import gradio as gr
|
| 5 |
from huggingface_hub import hf_hub_download
|
| 6 |
|
| 7 |
REPO_ID = "SentilyticsOPJ/lstm_model"
|
| 8 |
FILENAME = "lstm_model.pt"
|
| 9 |
|
| 10 |
+
MAX_LEN = 80
|
| 11 |
+
EMBEDDING_DIM = 300
|
| 12 |
+
LSTM_UNITS = 96
|
| 13 |
+
SPATIAL_DROPOUT = 0.3
|
| 14 |
+
DROPOUT = 0.5
|
| 15 |
|
| 16 |
+
|
| 17 |
+
class BiLSTM(nn.Module):
|
| 18 |
+
def __init__(self, vocab_size, num_classes, embedding_matrix=None,
|
| 19 |
+
embedding_dim=EMBEDDING_DIM, lstm_units=LSTM_UNITS,
|
| 20 |
+
spatial_dropout=SPATIAL_DROPOUT, dropout=DROPOUT,
|
| 21 |
+
padding_idx=0):
|
| 22 |
super().__init__()
|
| 23 |
+
self.padding_idx = padding_idx
|
| 24 |
+
|
| 25 |
+
self.embedding = nn.Embedding(vocab_size, embedding_dim, padding_idx=padding_idx)
|
| 26 |
+
if embedding_matrix is not None:
|
| 27 |
+
self.embedding.weight.data.copy_(torch.from_numpy(embedding_matrix))
|
| 28 |
+
|
| 29 |
+
self.spatial_dropout = nn.Dropout1d(spatial_dropout)
|
| 30 |
+
|
| 31 |
+
self.lstm = nn.LSTM(
|
| 32 |
+
input_size=embedding_dim,
|
| 33 |
+
hidden_size=lstm_units,
|
| 34 |
+
num_layers=1,
|
| 35 |
+
batch_first=True,
|
| 36 |
+
bidirectional=True,
|
| 37 |
+
)
|
| 38 |
+
self.attn = nn.Linear(lstm_units * 2, 1)
|
| 39 |
+
self.dropout = nn.Dropout(dropout)
|
| 40 |
+
self.fc = nn.Linear(lstm_units * 2, num_classes)
|
| 41 |
|
| 42 |
def forward(self, x):
|
| 43 |
+
emb = self.embedding(x)
|
| 44 |
+
|
| 45 |
+
emb = emb.transpose(1, 2)
|
| 46 |
+
emb = self.spatial_dropout(emb)
|
| 47 |
+
emb = emb.transpose(1, 2)
|
| 48 |
+
|
| 49 |
+
out, _ = self.lstm(emb)
|
| 50 |
+
|
| 51 |
+
scores = self.attn(out).squeeze(-1)
|
| 52 |
+
mask = (x == self.padding_idx)
|
| 53 |
+
mask[:, 0] = False
|
| 54 |
+
scores = scores.masked_fill(mask, float("-inf"))
|
| 55 |
+
weights = F.softmax(scores, dim=1).unsqueeze(-1)
|
| 56 |
+
context = (out * weights).sum(dim=1)
|
| 57 |
+
|
| 58 |
+
return self.fc(self.dropout(context))
|
| 59 |
|
| 60 |
path = hf_hub_download(repo_id=REPO_ID, filename=FILENAME)
|
| 61 |
obj = torch.load(path, map_location="cpu", weights_only=False)
|
| 62 |
sd = obj["state_dict"]
|
| 63 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 64 |
word2id = obj["word2id"]
|
| 65 |
labels = obj["label_classes"]
|
| 66 |
max_len = obj["max_len"]
|
| 67 |
|
| 68 |
+
model = BiLSTM(len(word2id), len(labels))
|
| 69 |
+
model.load_state_dict(sd)
|
| 70 |
+
model.eval()
|
| 71 |
+
|
| 72 |
|
| 73 |
def predict(text):
|
| 74 |
if not text.strip():
|
|
|
|
| 79 |
probs = torch.softmax(model(torch.tensor(ids).unsqueeze(0)), dim=1)[0]
|
| 80 |
return {labels[i]: float(probs[i]) for i in range(len(labels))}
|
| 81 |
|
| 82 |
+
|
| 83 |
gr.Interface(fn=predict, inputs="text", outputs="text").launch()
|