Spaces:
Running
Running
Commit
·
9c86aa3
1
Parent(s):
bb79ce7
feat: progress bat
Browse files- app.py +36 -36
- src/improved_diffusion/gaussian_diffusion.py +9 -4
app.py
CHANGED
@@ -70,43 +70,43 @@ encoder = get_encoder()
|
|
70 |
model = get_model()
|
71 |
diffusion = get_diffusion()
|
72 |
|
73 |
-
sample_fn = diffusion.ddim_sample_loop
|
74 |
-
|
75 |
st.title("Lang2mol-Diff")
|
76 |
text_input = st.text_area("Enter molecule description")
|
77 |
-
|
78 |
-
|
79 |
-
|
80 |
-
|
81 |
-
|
82 |
-
|
83 |
-
|
84 |
-
|
85 |
-
|
86 |
-
|
87 |
-
|
88 |
-
|
89 |
-
|
90 |
-
|
91 |
-
|
|
|
|
|
92 |
|
93 |
-
|
94 |
-
|
95 |
-
|
96 |
-
|
97 |
-
|
98 |
-
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
|
104 |
-
|
105 |
-
|
106 |
-
|
107 |
-
|
108 |
-
|
109 |
-
|
110 |
-
|
111 |
|
112 |
-
|
|
|
70 |
model = get_model()
|
71 |
diffusion = get_diffusion()
|
72 |
|
|
|
|
|
73 |
st.title("Lang2mol-Diff")
|
74 |
text_input = st.text_area("Enter molecule description")
|
75 |
+
button = st.button("Submit")
|
76 |
+
if button:
|
77 |
+
with st.spinner("Please wait..."):
|
78 |
+
output = tokenizer(
|
79 |
+
text_input,
|
80 |
+
max_length=256,
|
81 |
+
truncation=True,
|
82 |
+
padding="max_length",
|
83 |
+
add_special_tokens=True,
|
84 |
+
return_tensors="pt",
|
85 |
+
return_attention_mask=True,
|
86 |
+
)
|
87 |
+
caption_state = encoder(
|
88 |
+
input_ids=output["input_ids"],
|
89 |
+
attention_mask=output["attention_mask"],
|
90 |
+
).last_hidden_state
|
91 |
+
caption_mask = output["attention_mask"]
|
92 |
|
93 |
+
outputs = diffusion.p_sample_loop(
|
94 |
+
model,
|
95 |
+
(1, 256, 32),
|
96 |
+
clip_denoised=False,
|
97 |
+
denoised_fn=None,
|
98 |
+
model_kwargs={},
|
99 |
+
top_p=1.0,
|
100 |
+
progress=True,
|
101 |
+
caption=(caption_state, caption_mask),
|
102 |
+
)
|
103 |
+
logits = model.get_logits(torch.tensor(outputs))
|
104 |
+
cands = torch.topk(logits, k=1, dim=-1)
|
105 |
+
outputs = cands.indices
|
106 |
+
outputs = outputs.squeeze(-1)
|
107 |
+
outputs = tokenizer.decode(outputs)
|
108 |
+
result = sf.decoder(
|
109 |
+
outputs[0].replace("<pad>", "").replace("</s>", "").replace("\t", "")
|
110 |
+
).replace("\t", "")
|
111 |
|
112 |
+
st.write(result)
|
src/improved_diffusion/gaussian_diffusion.py
CHANGED
@@ -9,7 +9,7 @@ import enum
|
|
9 |
import math
|
10 |
import torch
|
11 |
import numpy as np
|
12 |
-
|
13 |
from .nn import mean_flat
|
14 |
from .losses import normal_kl, discretized_gaussian_log_likelihood
|
15 |
|
@@ -667,16 +667,19 @@ class GaussianDiffusion:
|
|
667 |
# print(indices[-10:])
|
668 |
if progress:
|
669 |
# Lazy import so that we don't depend on tqdm.
|
670 |
-
from tqdm.auto import tqdm
|
|
|
|
|
|
|
671 |
|
672 |
-
indices = tqdm(indices)
|
673 |
if caption is not None:
|
674 |
print("Text Guiding Generation ......")
|
675 |
caption = (
|
676 |
caption[0].to(img.device),
|
677 |
caption[1].to(img.device),
|
678 |
) # (caption_state, caption_mask)
|
679 |
-
|
|
|
680 |
t = torch.tensor([i] * shape[0], device=device)
|
681 |
with torch.no_grad():
|
682 |
out = self.p_sample(
|
@@ -691,6 +694,8 @@ class GaussianDiffusion:
|
|
691 |
)
|
692 |
yield out
|
693 |
img = out["sample"]
|
|
|
|
|
694 |
|
695 |
def p_sample_loop_langevin_progressive(
|
696 |
self,
|
|
|
9 |
import math
|
10 |
import torch
|
11 |
import numpy as np
|
12 |
+
import streamlit as st
|
13 |
from .nn import mean_flat
|
14 |
from .losses import normal_kl, discretized_gaussian_log_likelihood
|
15 |
|
|
|
667 |
# print(indices[-10:])
|
668 |
if progress:
|
669 |
# Lazy import so that we don't depend on tqdm.
|
670 |
+
# from tqdm.auto import tqdm
|
671 |
+
|
672 |
+
# indices = tqdm(indices)
|
673 |
+
pass
|
674 |
|
|
|
675 |
if caption is not None:
|
676 |
print("Text Guiding Generation ......")
|
677 |
caption = (
|
678 |
caption[0].to(img.device),
|
679 |
caption[1].to(img.device),
|
680 |
) # (caption_state, caption_mask)
|
681 |
+
my_bar = st.progress(0, text="Processing")
|
682 |
+
for pro, i in enumerate(indices):
|
683 |
t = torch.tensor([i] * shape[0], device=device)
|
684 |
with torch.no_grad():
|
685 |
out = self.p_sample(
|
|
|
694 |
)
|
695 |
yield out
|
696 |
img = out["sample"]
|
697 |
+
my_bar.progress(pro + 1, text="Processing")
|
698 |
+
my_bar.empty()
|
699 |
|
700 |
def p_sample_loop_langevin_progressive(
|
701 |
self,
|