Spaces:
Runtime error
Runtime error
Update app.py
Browse filesAdd seed and inference steps inputs.
app.py
CHANGED
@@ -6,30 +6,38 @@ pipeline, pipeline_params = FlaxStableDiffusionPipeline.from_pretrained(
|
|
6 |
"bguisard/stable-diffusion-nano",
|
7 |
)
|
8 |
|
9 |
-
prng_seed = jax.random.PRNGKey(0)
|
10 |
-
inference_steps = 50
|
11 |
|
|
|
|
|
12 |
|
13 |
-
def generate_image(prompt: str):
|
14 |
prompt_ids = pipeline.prepare_inputs(prompt)
|
15 |
images = pipeline(
|
16 |
prompt_ids=prompt_ids,
|
17 |
params=pipeline_params,
|
18 |
-
prng_seed=
|
19 |
height=128,
|
20 |
width=128,
|
21 |
-
num_inference_steps=inference_steps,
|
22 |
jit=False,
|
23 |
).images
|
|
|
24 |
pil_imgs = pipeline.numpy_to_pil(images)
|
25 |
return pil_imgs[0]
|
26 |
|
27 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
28 |
app = gr.Interface(
|
29 |
fn=generate_image,
|
30 |
-
inputs=
|
31 |
outputs=gr.Image(shape=(128, 128)),
|
32 |
examples=[["A watercolor painting of a bird"]],
|
33 |
)
|
34 |
|
35 |
-
app.launch()
|
|
|
6 |
"bguisard/stable-diffusion-nano",
|
7 |
)
|
8 |
|
|
|
|
|
9 |
|
10 |
+
def generate_image(prompt: str, inference_steps: int, prng_seed: int):
|
11 |
+
rng = jax.random.PRNGKey(int(prng_seed))
|
12 |
|
|
|
13 |
prompt_ids = pipeline.prepare_inputs(prompt)
|
14 |
images = pipeline(
|
15 |
prompt_ids=prompt_ids,
|
16 |
params=pipeline_params,
|
17 |
+
prng_seed=rng,
|
18 |
height=128,
|
19 |
width=128,
|
20 |
+
num_inference_steps=int(inference_steps),
|
21 |
jit=False,
|
22 |
).images
|
23 |
+
|
24 |
pil_imgs = pipeline.numpy_to_pil(images)
|
25 |
return pil_imgs[0]
|
26 |
|
27 |
|
28 |
+
prompt_input = gr.inputs.Textbox(
|
29 |
+
label="Prompt", placeholder="A watercolor painting of a bird"
|
30 |
+
)
|
31 |
+
inf_steps_input = gr.inputs.Slider(
|
32 |
+
minimum=1, maximum=100, default=30, step=1, label="Inference Steps"
|
33 |
+
)
|
34 |
+
seed_input = gr.inputs.Number(default=0, label="Seed")
|
35 |
+
|
36 |
app = gr.Interface(
|
37 |
fn=generate_image,
|
38 |
+
inputs=[prompt_input, inf_steps_input, seed_input],
|
39 |
outputs=gr.Image(shape=(128, 128)),
|
40 |
examples=[["A watercolor painting of a bird"]],
|
41 |
)
|
42 |
|
43 |
+
app.launch()
|