Spaces:
Sleeping
Sleeping
mrbeliever
commited on
Update app.py
Browse files
app.py
CHANGED
@@ -20,7 +20,7 @@ DEFAULT_QUERY = (
|
|
20 |
"Avoid subjective interpretations or speculation."
|
21 |
)
|
22 |
|
23 |
-
DTYPE = torch.
|
24 |
DEVICE = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
25 |
|
26 |
tokenizer = LlamaTokenizer.from_pretrained(
|
@@ -39,12 +39,11 @@ model = model.to(device=DEVICE)
|
|
39 |
@torch.no_grad()
|
40 |
def generate_caption(
|
41 |
image: Image.Image,
|
42 |
-
query: str = DEFAULT_QUERY,
|
43 |
params: dict[str, Any] = DEFAULT_PARAMS,
|
44 |
) -> str:
|
45 |
inputs = model.build_conversation_input_ids(
|
46 |
tokenizer=tokenizer,
|
47 |
-
query=query
|
48 |
history=[],
|
49 |
images=[image],
|
50 |
)
|
@@ -56,7 +55,7 @@ def generate_caption(
|
|
56 |
}
|
57 |
|
58 |
outputs = model.generate(**inputs, **params)
|
59 |
-
outputs = outputs[:, inputs["input_ids"].shape[1]:]
|
60 |
result = tokenizer.decode(outputs[0])
|
61 |
|
62 |
result = result.replace("This image showcases", "").strip().removesuffix("</s>").strip().capitalize()
|
@@ -65,14 +64,14 @@ def generate_caption(
|
|
65 |
with gr.Blocks() as demo:
|
66 |
with gr.Row():
|
67 |
with gr.Column():
|
68 |
-
input_image = gr.Image(type="pil")
|
69 |
run_button = gr.Button(value="Generate Caption")
|
70 |
with gr.Column():
|
71 |
output_caption = gr.Textbox(label="Generated Caption", show_copy_button=True)
|
72 |
|
73 |
run_button.click(
|
74 |
fn=generate_caption,
|
75 |
-
inputs=[input_image], # Only
|
76 |
outputs=output_caption,
|
77 |
)
|
78 |
|
|
|
20 |
"Avoid subjective interpretations or speculation."
|
21 |
)
|
22 |
|
23 |
+
DTYPE = torch.bfloat16
|
24 |
DEVICE = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
25 |
|
26 |
tokenizer = LlamaTokenizer.from_pretrained(
|
|
|
39 |
@torch.no_grad()
|
40 |
def generate_caption(
|
41 |
image: Image.Image,
|
|
|
42 |
params: dict[str, Any] = DEFAULT_PARAMS,
|
43 |
) -> str:
|
44 |
inputs = model.build_conversation_input_ids(
|
45 |
tokenizer=tokenizer,
|
46 |
+
query=DEFAULT_QUERY, # Use the default query directly
|
47 |
history=[],
|
48 |
images=[image],
|
49 |
)
|
|
|
55 |
}
|
56 |
|
57 |
outputs = model.generate(**inputs, **params)
|
58 |
+
outputs = outputs[:, inputs["input_ids"].shape[1] :]
|
59 |
result = tokenizer.decode(outputs[0])
|
60 |
|
61 |
result = result.replace("This image showcases", "").strip().removesuffix("</s>").strip().capitalize()
|
|
|
64 |
with gr.Blocks() as demo:
|
65 |
with gr.Row():
|
66 |
with gr.Column():
|
67 |
+
input_image = gr.Image(type="pil") # Image input remains
|
68 |
run_button = gr.Button(value="Generate Caption")
|
69 |
with gr.Column():
|
70 |
output_caption = gr.Textbox(label="Generated Caption", show_copy_button=True)
|
71 |
|
72 |
run_button.click(
|
73 |
fn=generate_caption,
|
74 |
+
inputs=[input_image], # Only the image input is passed
|
75 |
outputs=output_caption,
|
76 |
)
|
77 |
|