File size: 17,292 Bytes
a103028 40b5b1f a103028 b0eedf4 a103028 307c071 a103028 307c071 a103028 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 |
import os
# List of (source, target) folder pairs
FOLDER_PAIRS = [
("/data/pretrained", "/data/pretrained"),
("/data/pretrained_models", "/data/pretrained_models"),
]
def symlink_with_fp16_rename(source_folder, target_folder):
if not os.path.exists(source_folder):
print(f"⚠️ Source folder does not exist: {source_folder}")
return
os.makedirs(target_folder, exist_ok=True)
for root, dirs, files in os.walk(source_folder):
rel_path = os.path.relpath(root, source_folder)
target_root = (
target_folder
if rel_path == "."
else os.path.join(target_folder, rel_path)
)
os.makedirs(target_root, exist_ok=True)
for file_name in files:
source_file = os.path.join(root, file_name)
# Rename "*-fp16.*" → "*.fp16.*"
if "-fp16." in file_name:
target_file_name = file_name.replace("-fp16.", ".fp16.")
else:
target_file_name = file_name
target_file = os.path.join(target_root, target_file_name)
# Remove existing file/symlink
if os.path.lexists(target_file):
os.remove(target_file)
# Create symlink
os.symlink(source_file, target_file)
print(f"🔗 {target_file} -> {source_file}")
print(f"✅ Finished symlinking {source_folder} → {target_folder}\n")
# Run for all folder pairs
for src, dst in FOLDER_PAIRS:
symlink_with_fp16_rename(src, dst)
print("🎉 All pretrained folders have been symlinked successfully.")
import uuid
import logging
import uvicorn
import zipfile
import subprocess
# ------------------------------------------------------
# Anchor Endpoint (updated to accept video file)
# ------------------------------------------------------
from typing import Optional
import os, io, zipfile, shutil, logging
from fastapi import FastAPI, HTTPException, UploadFile, File, Form
from fastapi.responses import StreamingResponse
from gradio_epic_only import anchor_generation
# from gradio_epic_only import run_epic_vsr_pipeline
from pathlib import Path
from fastapi.responses import FileResponse
import numpy as np
import pandas as pd
from gradio_crop_only import merge_crops_wrapper
from merg_crops.utils import apply_merge_crops
# ------------------------------------------------------
# Config + Logging
# ------------------------------------------------------
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
app = FastAPI()
# ------------------------------------------------------
# Routes
# ------------------------------------------------------
@app.get("/repository")
async def repository():
repo_path = "/app"
if not os.path.exists(repo_path):
raise HTTPException(status_code=404, detail="Repository folder not found.")
files = []
for root, dirs, filenames in os.walk(repo_path):
for filename in filenames:
file_path = os.path.relpath(os.path.join(root, filename), repo_path)
files.append(file_path)
return {"files": files}
@app.get("/")
def read_root():
return {"message": "Welcome to the Camera vfx Server."}
@app.get("/health")
def health():
return {"message": "EPiC Server is running."}
from fastapi.responses import Response
@app.post("/anchor")
async def anchor_endpoint(
video_file: UploadFile = File(...),
fps: int = Form(24),
num_frames: int = Form(49),
target_pose: str = Form("0 8 0 0 0"),
unique_identifier: str = Form(default_factory=lambda: str(uuid.uuid4())),
mode: str = Form("gradual"),
radius_scale: float = Form(1.0),
near_far_estimated: bool = Form(True),
anchor_incre_res_input: bool = Form(True),
sampler_name: str = Form("DDIM_Origin"),
diffusion_guidance_scale: float = Form(6.0),
diffusion_inference_steps: int = Form(50),
prompt: str = Form(""),
negative_prompt: str = Form("The video is not of a high quality, low resolution, watermark present, etc."),
refine_prompt: str = Form("High quality, best quality, ultra-detailed."),
depth_inference_steps: int = Form(5),
depth_guidance_scale: float = Form(1.0),
window_size: int = Form(64),
overlap: int = Form(25),
max_res: int = Form(1024),
load_size: str = Form("480,720"),
sample_size: str = Form("480,720"),
depth_size: str = Form("768,1152"),
seed_input: int = Form(42),
aspect_ratio_inputs: str = Form("3,2"),
init_dx: float = Form(0.0),
init_dy: float = Form(0.0),
init_dz: float = Form(0.0),
init_theta: float = Form(0.0),
init_phi: float = Form(0.0),
reverse_effect: bool = Form(False),
extended_effect: bool = Form(False)
):
"""
Anchor generation API compatible with generate_anchor client.
Saves uploaded video, runs anchor_generation(), returns ZIP file fully at once.
"""
logger.info(f"Processing request {unique_identifier}")
temp_dir = os.path.join("/app", unique_identifier)
os.makedirs(temp_dir, exist_ok=True)
input_video_path = os.path.join(temp_dir, "input.mp4")
# Save uploaded video
try:
with open(input_video_path, "wb") as f:
shutil.copyfileobj(video_file.file, f)
except Exception as e:
raise HTTPException(status_code=400, detail=f"Failed to save uploaded video: {e}")
try:
# Call your anchor generation function
from gradio_epic_only import anchor_generation
video_output_path, logs, caption, frame_shape = anchor_generation(
video_path=input_video_path,
Unique_identifier=unique_identifier,
fps=fps,
num_frames=num_frames,
target_pose=target_pose,
mode=mode,
radius_scale=radius_scale,
near_far_estimated=near_far_estimated,
anchor_incre_res_input=anchor_incre_res_input,
sampler_name=sampler_name,
diffusion_guidance_scale=diffusion_guidance_scale,
diffusion_inference_steps=diffusion_inference_steps,
prompt=prompt,
negative_prompt=negative_prompt,
refine_prompt=refine_prompt,
depth_inference_steps=depth_inference_steps,
depth_guidance_scale=depth_guidance_scale,
window_size=window_size,
overlap=overlap,
max_res=max_res,
load_size=load_size,
sample_size=sample_size,
depth_size=depth_size,
seed_input=seed_input,
aspect_ratio_inputs=aspect_ratio_inputs,
init_dx=init_dx,
init_dy=init_dy,
init_dz=init_dz,
init_theta=init_theta,
init_phi=init_phi,
reverse_effect=reverse_effect,
extended_effect=extended_effect
)
# Save logs, captions, frame_shape
with open(os.path.join(temp_dir, "logs.txt"), "w") as f:
f.write(logs or "")
with open(os.path.join(temp_dir, "caption.txt"), "w") as f:
f.write(caption or "")
with open(os.path.join(temp_dir, "frame_shape.txt"), "w") as f:
f.write(str(frame_shape or (0,0,0,0)))
# Create ZIP
zip_path = os.path.join(temp_dir, f"{unique_identifier}.zip")
with zipfile.ZipFile(zip_path, "w", compression=zipfile.ZIP_DEFLATED) as zipf:
for root, _, files in os.walk(temp_dir):
for file in files:
file_path = os.path.join(root, file)
if os.path.abspath(file_path) == os.path.abspath(zip_path):
continue
arcname = os.path.relpath(file_path, temp_dir)
zipf.write(file_path, arcname)
# Return ZIP fully at once
with open(zip_path, "rb") as f:
zip_bytes = f.read()
return Response(
content=zip_bytes,
media_type="application/zip",
headers={"Content-Disposition": f"attachment; filename={unique_identifier}.zip"}
)
except Exception as e:
logger.exception("Anchor generation failed")
raise HTTPException(status_code=500, detail=f"Anchor generation failed: {e}")
@app.post("/epic-vsr")
async def inference_epic_vsr(
zip_file: UploadFile = File(...),
fps: int = Form(24),
num_frames: int = Form(49),
vsr: bool = Form(False),
controlnet_weights: float = Form(0.5),
controlnet_guidance_start: float = Form(0.0),
controlnet_guidance_end: float = Form(0.4),
guidance_scale: float = Form(6.0),
inference_steps: int = Form(50),
dtype: str = Form("bfloat16"),
seed: int = Form(42),
height_input: int = Form(480),
width_input: int = Form(720),
downscale_coef: int = Form(8),
vae_channels: int = Form(16),
controlnet_input_channels: int = Form(6),
controlnet_layers: int = Form(8),
out_dir: str = Form("/app/out/epic_vsr_output"),
input_folder: str = Form("/app/epic_vsr_input"),
temp_vsr_input: str = Form("/app/epic_vsr_temp_vsr"),
logs_all: list = Form([]),
width: int = Form(1280)
):
input_folder = Path(input_folder)
input_folder.mkdir(parents=True, exist_ok=True)
if any(input_folder.iterdir()):
shutil.rmtree(input_folder)
input_folder.mkdir(parents=True, exist_ok=True)
zip_bytes = io.BytesIO(await zip_file.read())
zip_bytes.seek(0)
with zipfile.ZipFile(zip_bytes, "r") as zf:
zf.extractall(input_folder)
print(f"Extracted ZIP to {input_folder}, files: {list(input_folder.iterdir())}")
model_path = "/data/pretrained/CogVideoX-5b-I2V"
ckpt_path = "/app/out/EPiC_pretrained/checkpoint-500.pt"
command = [
"python", "-u",
"/app/inference/cli_demo_camera_i2v_pcd.py",
"--video_root_dir", str(input_folder),
"--base_model_path", model_path,
"--controlnet_model_path", ckpt_path,
"--output_path", out_dir,
"--controlnet_weights", str(controlnet_weights),
"--controlnet_guidance_start", str(controlnet_guidance_start),
"--controlnet_guidance_end", str(controlnet_guidance_end),
"--guidance_scale", str(guidance_scale),
"--num_inference_steps", str(inference_steps),
"--dtype", str(dtype),
"--controlnet_transformer_num_attn_heads", "4",
"--controlnet_transformer_attention_head_dim", "64",
"--controlnet_transformer_out_proj_dim_factor", "64",
"--controlnet_transformer_out_proj_dim_zero_init",
"--seed", str(seed),
"--height", str(height_input),
"--width", str(width_input),
"--infer_with_mask",
"--num_frames", str(num_frames),
"--fps", str(fps),
"--downscale_coef", str(downscale_coef),
"--vae_channels", str(vae_channels),
"--pool_style", "max",
"--controlnet_input_channels", str(controlnet_input_channels),
"--controlnet_transformer_num_layers", str(controlnet_layers),
]
logs_combined = f"[{input_folder.name}] Starting EPiC inference...\n"
print("🚀 Starting EPiC subprocess...")
process = subprocess.Popen(
command,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
text=True,
bufsize=1,
universal_newlines=True,
)
for line in process.stdout:
print("[EPiC]", line, end="")
logs_combined += line
process.wait()
if process.returncode != 0:
logs_combined += f"\n❌ EPiC failed with code {process.returncode}\n"
logs_all.append(logs_combined)
return logs_all
epic_out = f"{out_dir}/00000_{seed}_out.mp4"
if not os.path.exists(epic_out):
logs_combined += "\n❌ EPiC output missing."
logs_all.append(logs_combined)
return logs_all
final_path = f"{out_dir}/{input_folder.name}_{seed}_out.mp4"
shutil.move(epic_out, final_path)
logs_combined += f"\n✅ EPiC done → {final_path}"
temp_vsr_input = Path(temp_vsr_input)
temp_vsr_input.mkdir(parents=True, exist_ok=True)
for f in temp_vsr_input.glob("*"):
try:
if f.is_file():
f.unlink()
except Exception:
pass
temp_copy = temp_vsr_input / Path(final_path).name
shutil.copy(final_path, temp_copy)
if vsr:
try:
cap = cv2.VideoCapture(str(final_path))
ok, frame = cap.read()
cap.release()
if not ok or frame is None:
logs_combined += "\n⚠️ Could not read EPiC output for DOVE-VSR."
else:
h, w = frame.shape[:2]
logs_combined += f"\n🔍 Resolution: {w}×{h}"
if width < 1600:
upscale = 2
elif width < 2300:
upscale = 3
else:
upscale = 4
logs_combined += f"\n🚀 Starting DOVE-VSR (x{upscale})...\n"
cmd = [
"python", "-u",
"/app/Dove/inference_script.py",
"--input_dir", str(temp_vsr_input),
"--model_path", "/data/pretrained_models/DOVE",
"--output_path", "/app/Dove/output",
"--is_vae_st",
"--save_format", "yuv420p",
"--upscale", str(upscale)
]
vsr_process = subprocess.Popen(
cmd,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
text=True,
bufsize=1,
universal_newlines=True,
)
for line in vsr_process.stdout:
print("[DOVE]", line, end="")
logs_combined += line
vsr_process.wait()
if vsr_process.returncode != 0:
logs_combined += f"\n❌ DOVE-VSR failed with code {vsr_process.returncode}\n"
else:
vsr_output_path = Path(out_dir) / f"{Path(final_path).stem}_vsr.mp4"
dove_generated_path = Path("/app/Dove/output") / temp_copy.name
if dove_generated_path.exists():
shutil.move(str(dove_generated_path), str(vsr_output_path))
logs_combined += f"\n✅ DOVE-VSR done → {vsr_output_path}"
else:
logs_combined += "\n⚠️ DOVE output missing."
except Exception as e:
logs_combined += f"\n❌ DOVE stage exception: {e}"
else:
logs_combined += "\n⚠️ DOVE-VSR skipped."
logs_all.append(logs_combined)
out_dir = Path(out_dir)
out_dir.mkdir(parents=True, exist_ok=True)
logs_path = out_dir / "logs.txt"
with open(logs_path, "w") as f:
f.write("\n".join(logs_all))
zip_output_path = f"/app/{out_dir.name}.zip"
shutil.make_archive(zip_output_path.replace(".zip", ""), "zip", str(out_dir))
return FileResponse(
zip_output_path,
media_type="application/zip",
filename=f"{out_dir.name}.zip"
)
@app.post("/merge-crops")
async def merge_crops_endpoint(
zip_file: UploadFile = File(...),
shot_type: str = Form("tele"),
apply_bokeh: str = Form("No"),
number_persons: int = Form(1),
bokeh_strength: float = Form(3.5),
reverse_effect: bool = Form(False),
blending_type: str = Form("dilation"),
unique_identifier: str = Form(default_factory=lambda: str(uuid.uuid4()))
):
logger.info(f"Processing merge-crops for {unique_identifier}")
base_dir = Path(f"/app/{unique_identifier}")
base_dir.mkdir(parents=True, exist_ok=True)
# --- Read uploaded ZIP into memory and extract ---
try:
zip_bytes = io.BytesIO(await zip_file.read())
zip_bytes.seek(0)
with zipfile.ZipFile(zip_bytes, "r") as zf:
zf.extractall(base_dir)
logger.info(f"Extracted ZIP to {base_dir}")
except Exception as e:
logger.exception("Failed to extract ZIP")
raise HTTPException(status_code=400, detail=f"Failed to extract ZIP: {e}")
try:
result_path = merge_crops_wrapper(
Unique_identifier=unique_identifier,
shot_type=shot_type,
apply_bokeh=apply_bokeh,
number_persons=number_persons,
bokeh_strength=bokeh_strength,
reverse_effect=reverse_effect,
blending_type=blending_type
)
if result_path and os.path.exists(result_path):
return FileResponse(result_path, media_type="video/mp4", filename=os.path.basename(result_path))
else:
raise HTTPException(status_code=500, detail="Merge crops failed to produce output video.")
except Exception as e:
logger.exception("Merge crops failed")
raise HTTPException(status_code=500, detail=f"Merge crops failed: {str(e)}")
# ------------------------------------------------------
# Run Uvicorn
# ------------------------------------------------------
if __name__ == "__main__":
uvicorn.run(app, host="0.0.0.0", port=90)
|