Spaces:
Running
on
A100
Running
on
A100
using transformers to handle the model
Browse files- app.py +226 -61
- requirements.txt +1 -1
app.py
CHANGED
@@ -2,23 +2,15 @@ import os
|
|
2 |
import json
|
3 |
import gradio as gr
|
4 |
import tempfile
|
5 |
-
from PIL import Image, ImageDraw, ImageFont
|
6 |
-
import cv2
|
7 |
-
from typing import Tuple, Optional
|
8 |
-
import torch
|
9 |
-
from pathlib import Path
|
10 |
-
import time
|
11 |
import torch
|
12 |
import spaces
|
13 |
-
import
|
14 |
-
|
|
|
|
|
15 |
|
16 |
-
|
17 |
-
|
18 |
-
BatchedVideoHighlightDetector,
|
19 |
-
get_video_duration_seconds,
|
20 |
-
get_fixed_30s_segments
|
21 |
-
)
|
22 |
|
23 |
def load_examples(json_path: str) -> dict:
|
24 |
with open(json_path, 'r') as f:
|
@@ -32,8 +24,161 @@ def format_duration(seconds: int) -> str:
|
|
32 |
return f"{hours}:{minutes:02d}:{secs:02d}"
|
33 |
return f"{minutes}:{secs:02d}"
|
34 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
35 |
|
36 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
37 |
examples_data = load_examples(examples_path)
|
38 |
|
39 |
with gr.Blocks() as app:
|
@@ -54,7 +199,6 @@ def create_ui(examples_path: str):
|
|
54 |
gr.Markdown(f"### {example['title']}")
|
55 |
|
56 |
with gr.Column():
|
57 |
-
|
58 |
gr.Video(
|
59 |
value=example["highlights"]["url"],
|
60 |
label=f"Highlights ({format_duration(example['highlights']['duration_seconds'])})",
|
@@ -102,6 +246,7 @@ def create_ui(examples_path: str):
|
|
102 |
gr.update(value=None, visible=False), # Clear video
|
103 |
gr.update(visible=False) # Hide accordion
|
104 |
]
|
|
|
105 |
if not video:
|
106 |
yield [
|
107 |
"Please upload a video",
|
@@ -124,19 +269,16 @@ def create_ui(examples_path: str):
|
|
124 |
]
|
125 |
return
|
126 |
|
127 |
-
# Make accordion visible as soon as processing starts
|
128 |
yield [
|
129 |
-
"
|
130 |
"",
|
131 |
"",
|
132 |
gr.update(visible=False),
|
133 |
gr.update(visible=False)
|
134 |
]
|
135 |
|
136 |
-
|
137 |
-
|
138 |
-
model,
|
139 |
-
processor,
|
140 |
batch_size=8
|
141 |
)
|
142 |
|
@@ -161,18 +303,21 @@ def create_ui(examples_path: str):
|
|
161 |
|
162 |
highlights = detector.determine_highlights(video_desc)
|
163 |
formatted_highlights = f"### Highlights to search for:\n {highlights[:500] + '...' if len(highlights) > 500 else highlights}"
|
|
|
|
|
|
|
|
|
164 |
|
165 |
-
|
166 |
-
|
167 |
-
total_segments = len(segments)
|
168 |
kept_segments = []
|
169 |
-
|
170 |
-
|
171 |
-
|
172 |
-
|
|
|
|
|
173 |
|
174 |
-
# Update progress
|
175 |
-
progress = int((i / total_segments) * 100)
|
176 |
yield [
|
177 |
f"Processing segments... {progress}% complete",
|
178 |
formatted_desc,
|
@@ -180,35 +325,56 @@ def create_ui(examples_path: str):
|
|
180 |
gr.update(visible=False),
|
181 |
gr.update(visible=True)
|
182 |
]
|
|
|
|
|
|
|
|
|
183 |
|
184 |
-
|
185 |
-
|
186 |
-
|
187 |
-
|
188 |
-
|
189 |
-
|
190 |
-
|
191 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
192 |
|
193 |
-
#
|
194 |
-
|
195 |
-
|
196 |
-
|
|
|
197 |
|
198 |
# Create final video
|
199 |
-
|
200 |
-
|
201 |
-
|
202 |
-
|
203 |
-
|
204 |
-
|
205 |
-
|
206 |
-
|
207 |
-
|
208 |
-
|
209 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
210 |
|
211 |
except Exception as e:
|
|
|
212 |
yield [
|
213 |
f"Error processing video: {str(e)}",
|
214 |
"",
|
@@ -217,10 +383,8 @@ def create_ui(examples_path: str):
|
|
217 |
gr.update(visible=False)
|
218 |
]
|
219 |
finally:
|
220 |
-
|
221 |
-
|
222 |
-
torch.cuda.empty_cache()
|
223 |
-
|
224 |
|
225 |
process_btn.click(
|
226 |
on_process,
|
@@ -240,7 +404,8 @@ def create_ui(examples_path: str):
|
|
240 |
if __name__ == "__main__":
|
241 |
# Initialize CUDA
|
242 |
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
|
243 |
-
|
244 |
-
|
245 |
-
|
|
|
246 |
app.launch()
|
|
|
2 |
import json
|
3 |
import gradio as gr
|
4 |
import tempfile
|
|
|
|
|
|
|
|
|
|
|
|
|
5 |
import torch
|
6 |
import spaces
|
7 |
+
from pathlib import Path
|
8 |
+
from transformers import AutoProcessor, AutoModelForVision2Seq
|
9 |
+
import subprocess
|
10 |
+
import logging
|
11 |
|
12 |
+
logging.basicConfig(level=logging.INFO)
|
13 |
+
logger = logging.getLogger(__name__)
|
|
|
|
|
|
|
|
|
14 |
|
15 |
def load_examples(json_path: str) -> dict:
|
16 |
with open(json_path, 'r') as f:
|
|
|
24 |
return f"{hours}:{minutes:02d}:{secs:02d}"
|
25 |
return f"{minutes}:{secs:02d}"
|
26 |
|
27 |
+
def get_video_duration_seconds(video_path: str) -> float:
|
28 |
+
"""Use ffprobe to get video duration in seconds."""
|
29 |
+
cmd = [
|
30 |
+
"ffprobe",
|
31 |
+
"-v", "quiet",
|
32 |
+
"-print_format", "json",
|
33 |
+
"-show_format",
|
34 |
+
video_path
|
35 |
+
]
|
36 |
+
result = subprocess.run(cmd, capture_output=True, text=True)
|
37 |
+
info = json.loads(result.stdout)
|
38 |
+
return float(info["format"]["duration"])
|
39 |
|
40 |
+
class VideoHighlightDetector:
|
41 |
+
def __init__(
|
42 |
+
self,
|
43 |
+
model_path: str,
|
44 |
+
device: str = "cuda",
|
45 |
+
batch_size: int = 8
|
46 |
+
):
|
47 |
+
self.device = device
|
48 |
+
self.batch_size = batch_size
|
49 |
+
|
50 |
+
# Initialize model and processor
|
51 |
+
self.processor = AutoProcessor.from_pretrained(model_path)
|
52 |
+
self.model = AutoModelForVision2Seq.from_pretrained(
|
53 |
+
model_path,
|
54 |
+
torch_dtype=torch.bfloat16,
|
55 |
+
attn_implementation="flash_attention_2"
|
56 |
+
).to(device)
|
57 |
+
|
58 |
+
def analyze_video_content(self, video_path: str) -> str:
|
59 |
+
"""Analyze video content to determine its type and description."""
|
60 |
+
messages = [
|
61 |
+
{
|
62 |
+
"role": "user",
|
63 |
+
"content": [
|
64 |
+
{"type": "video", "path": video_path},
|
65 |
+
{"type": "text", "text": "What type of video is this and what's happening in it? Be specific about the content type and general activities you observe."}
|
66 |
+
]
|
67 |
+
}
|
68 |
+
]
|
69 |
+
|
70 |
+
inputs = self.processor.apply_chat_template(
|
71 |
+
messages,
|
72 |
+
add_generation_prompt=True,
|
73 |
+
tokenize=True,
|
74 |
+
return_dict=True,
|
75 |
+
return_tensors="pt"
|
76 |
+
).to(self.device)
|
77 |
+
|
78 |
+
outputs = self.model.generate(**inputs, max_new_tokens=512, do_sample=True, temperature=0.7)
|
79 |
+
return self.processor.decode(outputs[0], skip_special_tokens=True)
|
80 |
+
|
81 |
+
def determine_highlights(self, video_description: str) -> str:
|
82 |
+
"""Determine what constitutes highlights based on video description."""
|
83 |
+
messages = [
|
84 |
+
{
|
85 |
+
"role": "system",
|
86 |
+
"content": [{"type": "text", "text": "You are a professional video editor specializing in creating viral highlight reels."}]
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"role": "user",
|
90 |
+
"content": [{"type": "text", "text": f"""Based on this video description:
|
91 |
+
|
92 |
+
{video_description}
|
93 |
+
|
94 |
+
List which rare segments should be included in a best of the best highlight."""}]
|
95 |
+
}
|
96 |
+
]
|
97 |
+
|
98 |
+
inputs = self.processor.apply_chat_template(
|
99 |
+
messages,
|
100 |
+
add_generation_prompt=True,
|
101 |
+
tokenize=True,
|
102 |
+
return_dict=True,
|
103 |
+
return_tensors="pt"
|
104 |
+
).to(self.device)
|
105 |
+
|
106 |
+
outputs = self.model.generate(**inputs, max_new_tokens=256, do_sample=True, temperature=0.7)
|
107 |
+
return self.processor.decode(outputs[0], skip_special_tokens=True)
|
108 |
+
|
109 |
+
def process_segment(self, video_path: str, highlight_types: str) -> bool:
|
110 |
+
"""Process a video segment and determine if it contains highlights."""
|
111 |
+
messages = [
|
112 |
+
{
|
113 |
+
"role": "user",
|
114 |
+
"content": [
|
115 |
+
{"type": "video", "path": video_path},
|
116 |
+
{"type": "text", "text": f"""Do you see any of the following types of highlight moments in this video segment?
|
117 |
+
|
118 |
+
Potential highlights to look for:
|
119 |
+
{highlight_types}
|
120 |
+
|
121 |
+
Only answer yes if you see any of those moments and answer no if you don't."""}
|
122 |
+
]
|
123 |
+
}
|
124 |
+
]
|
125 |
+
|
126 |
+
inputs = self.processor.apply_chat_template(
|
127 |
+
messages,
|
128 |
+
add_generation_prompt=True,
|
129 |
+
tokenize=True,
|
130 |
+
return_dict=True,
|
131 |
+
return_tensors="pt"
|
132 |
+
).to(self.device)
|
133 |
+
|
134 |
+
outputs = self.model.generate(**inputs, max_new_tokens=64, do_sample=False)
|
135 |
+
response = self.processor.decode(outputs[0], skip_special_tokens=True).lower()
|
136 |
+
|
137 |
+
return "yes" in response
|
138 |
+
|
139 |
+
def _concatenate_scenes(
|
140 |
+
self,
|
141 |
+
video_path: str,
|
142 |
+
scene_times: list,
|
143 |
+
output_path: str
|
144 |
+
):
|
145 |
+
"""Concatenate selected scenes into final video."""
|
146 |
+
if not scene_times:
|
147 |
+
logger.warning("No scenes to concatenate, skipping.")
|
148 |
+
return
|
149 |
+
|
150 |
+
filter_complex_parts = []
|
151 |
+
concat_inputs = []
|
152 |
+
for i, (start_sec, end_sec) in enumerate(scene_times):
|
153 |
+
filter_complex_parts.append(
|
154 |
+
f"[0:v]trim=start={start_sec}:end={end_sec},"
|
155 |
+
f"setpts=PTS-STARTPTS[v{i}];"
|
156 |
+
)
|
157 |
+
filter_complex_parts.append(
|
158 |
+
f"[0:a]atrim=start={start_sec}:end={end_sec},"
|
159 |
+
f"asetpts=PTS-STARTPTS[a{i}];"
|
160 |
+
)
|
161 |
+
concat_inputs.append(f"[v{i}][a{i}]")
|
162 |
+
|
163 |
+
concat_filter = f"{''.join(concat_inputs)}concat=n={len(scene_times)}:v=1:a=1[outv][outa]"
|
164 |
+
filter_complex = "".join(filter_complex_parts) + concat_filter
|
165 |
+
|
166 |
+
cmd = [
|
167 |
+
"ffmpeg",
|
168 |
+
"-y",
|
169 |
+
"-i", video_path,
|
170 |
+
"-filter_complex", filter_complex,
|
171 |
+
"-map", "[outv]",
|
172 |
+
"-map", "[outa]",
|
173 |
+
"-c:v", "libx264",
|
174 |
+
"-c:a", "aac",
|
175 |
+
output_path
|
176 |
+
]
|
177 |
+
|
178 |
+
logger.info(f"Running ffmpeg command: {' '.join(cmd)}")
|
179 |
+
subprocess.run(cmd, check=True)
|
180 |
+
|
181 |
+
def create_ui(examples_path: str, model_path: str):
|
182 |
examples_data = load_examples(examples_path)
|
183 |
|
184 |
with gr.Blocks() as app:
|
|
|
199 |
gr.Markdown(f"### {example['title']}")
|
200 |
|
201 |
with gr.Column():
|
|
|
202 |
gr.Video(
|
203 |
value=example["highlights"]["url"],
|
204 |
label=f"Highlights ({format_duration(example['highlights']['duration_seconds'])})",
|
|
|
246 |
gr.update(value=None, visible=False), # Clear video
|
247 |
gr.update(visible=False) # Hide accordion
|
248 |
]
|
249 |
+
|
250 |
if not video:
|
251 |
yield [
|
252 |
"Please upload a video",
|
|
|
269 |
]
|
270 |
return
|
271 |
|
|
|
272 |
yield [
|
273 |
+
"Initializing video highlight detector...",
|
274 |
"",
|
275 |
"",
|
276 |
gr.update(visible=False),
|
277 |
gr.update(visible=False)
|
278 |
]
|
279 |
|
280 |
+
detector = VideoHighlightDetector(
|
281 |
+
model_path=model_path,
|
|
|
|
|
282 |
batch_size=8
|
283 |
)
|
284 |
|
|
|
303 |
|
304 |
highlights = detector.determine_highlights(video_desc)
|
305 |
formatted_highlights = f"### Highlights to search for:\n {highlights[:500] + '...' if len(highlights) > 500 else highlights}"
|
306 |
+
|
307 |
+
# Split video into segments
|
308 |
+
temp_dir = "temp_segments"
|
309 |
+
os.makedirs(temp_dir, exist_ok=True)
|
310 |
|
311 |
+
segment_length = 10.0
|
312 |
+
duration = get_video_duration_seconds(video)
|
|
|
313 |
kept_segments = []
|
314 |
+
segments_processed = 0
|
315 |
+
total_segments = int(duration / segment_length)
|
316 |
+
|
317 |
+
for start_time in range(0, int(duration), int(segment_length)):
|
318 |
+
segments_processed += 1
|
319 |
+
progress = int((segments_processed / total_segments) * 100)
|
320 |
|
|
|
|
|
321 |
yield [
|
322 |
f"Processing segments... {progress}% complete",
|
323 |
formatted_desc,
|
|
|
325 |
gr.update(visible=False),
|
326 |
gr.update(visible=True)
|
327 |
]
|
328 |
+
|
329 |
+
# Create segment
|
330 |
+
segment_path = f"{temp_dir}/segment_{start_time}.mp4"
|
331 |
+
end_time = min(start_time + segment_length, duration)
|
332 |
|
333 |
+
cmd = [
|
334 |
+
"ffmpeg",
|
335 |
+
"-y",
|
336 |
+
"-i", video,
|
337 |
+
"-ss", str(start_time),
|
338 |
+
"-t", str(segment_length),
|
339 |
+
"-c", "copy",
|
340 |
+
segment_path
|
341 |
+
]
|
342 |
+
subprocess.run(cmd, check=True)
|
343 |
+
|
344 |
+
# Process segment
|
345 |
+
if detector.process_segment(segment_path, highlights):
|
346 |
+
kept_segments.append((start_time, end_time))
|
347 |
|
348 |
+
# Clean up segment file
|
349 |
+
os.remove(segment_path)
|
350 |
+
|
351 |
+
# Remove temp directory
|
352 |
+
os.rmdir(temp_dir)
|
353 |
|
354 |
# Create final video
|
355 |
+
if kept_segments:
|
356 |
+
with tempfile.NamedTemporaryFile(suffix='.mp4', delete=False) as tmp_file:
|
357 |
+
temp_output = tmp_file.name
|
358 |
+
detector._concatenate_scenes(video, kept_segments, temp_output)
|
359 |
+
|
360 |
+
yield [
|
361 |
+
"Processing complete!",
|
362 |
+
formatted_desc,
|
363 |
+
formatted_highlights,
|
364 |
+
gr.update(value=temp_output, visible=True),
|
365 |
+
gr.update(visible=True)
|
366 |
+
]
|
367 |
+
else:
|
368 |
+
yield [
|
369 |
+
"No highlights detected in the video.",
|
370 |
+
formatted_desc,
|
371 |
+
formatted_highlights,
|
372 |
+
gr.update(visible=False),
|
373 |
+
gr.update(visible=True)
|
374 |
+
]
|
375 |
|
376 |
except Exception as e:
|
377 |
+
logger.exception("Error processing video")
|
378 |
yield [
|
379 |
f"Error processing video: {str(e)}",
|
380 |
"",
|
|
|
383 |
gr.update(visible=False)
|
384 |
]
|
385 |
finally:
|
386 |
+
# Clean up
|
387 |
+
torch.cuda.empty_cache()
|
|
|
|
|
388 |
|
389 |
process_btn.click(
|
390 |
on_process,
|
|
|
404 |
if __name__ == "__main__":
|
405 |
# Initialize CUDA
|
406 |
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
|
407 |
+
|
408 |
+
MODEL_PATH = os.getenv("MODEL_PATH", "HuggingFaceTB/SmolVLM2-2.2B-Instruct")
|
409 |
+
|
410 |
+
app = create_ui("video_spec.json", MODEL_PATH)
|
411 |
app.launch()
|
requirements.txt
CHANGED
@@ -2,6 +2,6 @@ Pillow
|
|
2 |
opencv-python
|
3 |
num2words
|
4 |
ffmpeg-python
|
5 |
-
transformers
|
6 |
accelerate>=0.26.0
|
7 |
decord==0.6.0
|
|
|
2 |
opencv-python
|
3 |
num2words
|
4 |
ffmpeg-python
|
5 |
+
transformers @ git+https://github.com/huggingface/transformers.git@refs/pull/36126/head
|
6 |
accelerate>=0.26.0
|
7 |
decord==0.6.0
|