Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
@@ -3,7 +3,6 @@ import os
|
|
3 |
os.environ["WANDB_DISABLED"] = "true"
|
4 |
|
5 |
import numpy as np
|
6 |
-
from PIL import Image
|
7 |
|
8 |
import gradio as gr
|
9 |
import torch
|
@@ -116,25 +115,14 @@ config = model.vision_encoder.config
|
|
116 |
sf_filename = hf_hub_download("FFZG-cleopatra/M2SA", filename="model.safetensors")
|
117 |
|
118 |
load_model(model, sf_filename)
|
119 |
-
# model.load_state_dict(torch.load(model_args.model_name_or_path+"-finetuned/pytorch_model.bin"))
|
120 |
-
|
121 |
-
|
122 |
-
# model = AutoModelForSequenceClassification.from_pretrained(
|
123 |
-
# "FFZG-cleopatra/M2SA",
|
124 |
-
# num_labels=3, id2label=id2label,
|
125 |
-
# label2id=label2id
|
126 |
-
# )
|
127 |
-
|
128 |
-
|
129 |
-
|
130 |
image_processor = AutoImageProcessor.from_pretrained("openai/clip-vit-base-patch32")
|
131 |
|
132 |
def predict_sentiment(text, image):
|
133 |
print(text, image)
|
134 |
-
|
135 |
image = read_image(image, mode=ImageReadMode.RGB)
|
136 |
-
|
137 |
-
|
138 |
text_inputs = tokenizer(
|
139 |
text,
|
140 |
max_length=512,
|
@@ -152,9 +140,6 @@ def predict_sentiment(text, image):
|
|
152 |
pixel_values = image_transformations(image)
|
153 |
text_inputs["pixel_values"] = pixel_values.unsqueeze(0)
|
154 |
|
155 |
-
print(text_inputs)
|
156 |
-
print("its going in ",pixel_values)
|
157 |
-
|
158 |
prediction = None
|
159 |
with torch.no_grad():
|
160 |
outputs = model(**text_inputs)
|
|
|
3 |
os.environ["WANDB_DISABLED"] = "true"
|
4 |
|
5 |
import numpy as np
|
|
|
6 |
|
7 |
import gradio as gr
|
8 |
import torch
|
|
|
115 |
sf_filename = hf_hub_download("FFZG-cleopatra/M2SA", filename="model.safetensors")
|
116 |
|
117 |
load_model(model, sf_filename)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
118 |
image_processor = AutoImageProcessor.from_pretrained("openai/clip-vit-base-patch32")
|
119 |
|
120 |
def predict_sentiment(text, image):
|
121 |
print(text, image)
|
122 |
+
|
123 |
image = read_image(image, mode=ImageReadMode.RGB)
|
124 |
+
|
125 |
+
|
126 |
text_inputs = tokenizer(
|
127 |
text,
|
128 |
max_length=512,
|
|
|
140 |
pixel_values = image_transformations(image)
|
141 |
text_inputs["pixel_values"] = pixel_values.unsqueeze(0)
|
142 |
|
|
|
|
|
|
|
143 |
prediction = None
|
144 |
with torch.no_grad():
|
145 |
outputs = model(**text_inputs)
|