\n| Peak VRAM while generating<\/td>\n | 32.3GB<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n Generation times are for a rented cloud NVIDIA L40S and are not a guide to your own GPU; the text encoder runs first and is released before the video model loads, so the VRAM above does not include it. The \u201cwithout the LoRA\u201d clips use the same 4 steps for comparison only; the base model’s usual settings use many more steps with guidance, so they do not show what the base model can do.<\/p>\n Setup, Download and Run Steps (What We Actually Ran)<\/h4>\nWe split the work into two containers on Modal, both built from the Debian (slim) image with Python 3.12. The downloading container fetches only the listed files (configs, safetensors weights and the tokenizer) from Hugging Face at pinned revisions; the generating container has no network access. These are the packages we installed:<\/p>\n # 1. Environment for downloading (no GPU)\npython -m pip install huggingface_hub\n\n# 2. Environment for generating (GPU, network blocked)\npython -m pip install torch diffusers transformers accelerate safetensors sentencepiece ftfy imageio-ffmpeg numpy\n<\/code><\/pre>\nWe then called download(job)<\/code> in the first container and run(job)<\/code> in the second one from the script below (video_remote.py<\/code>, written by us; full text). Files go under \/models\/<dir><\/code>:<\/p>\n"""\nModal \u306e\u30b3\u30f3\u30c6\u30ca\u306e\u4e2d\u3067\u52d5\u304f\u3001\u52d5\u753b\u751f\u6210\u30e2\u30c7\u30eb\u306e\u751f\u6210(2026-10-01\u3001\u30e6\u30fc\u30b6\u30fc\u306e\u5224\u65ad)\u3002lmw\/lab\/gpu_modal.py \u306e\nvideo_download \/ video_run \u304b\u3089\u547c\u3070\u308c\u308b\u3002**\u6a19\u6e96\u30e9\u30a4\u30d6\u30e9\u30ea\u3060\u3051\u3092\u5148\u982d\u3067\u8aad\u307f**\u3001huggingface_hub\u30fbtorch\u30fbdiffusers \u306f\n\u5404\u95a2\u6570\u306e\u4e2d\u3067\u8aad\u3080(\u30b3\u30f3\u30c6\u30ca\u3054\u3068\u306b\u5165\u3063\u3066\u3044\u308b\u3082\u306e\u304c\u9055\u3046)\u3002\n\n\u3010\u914d\u5e03\u5143\u306e\u30b3\u30fc\u30c9\u306f\u52d5\u304b\u3055\u306a\u3044\u3011diffusers\u30fbtransformers \u306e\u516c\u958b\u30d1\u30c3\u30b1\u30fc\u30b8\u3060\u3051\u3067\u8aad\u307f\u8fbc\u3080\u3002\u91cd\u307f\u306f safetensors \u3060\u3051\u3067\u3001\nHugging Face \u306e\u30ea\u30dd\u30b8\u30c8\u30ea\u306e\u30b3\u30fc\u30c9\u3092\u8aad\u307f\u8fbc\u3080\u8a2d\u5b9a\u306f\u4f7f\u308f\u306a\u3044\u3002LoRA \u306f\u5f53\u30b5\u30a4\u30c8\u306e\u30b3\u30fc\u30c9(merge_lora)\u304c\u3001\u540d\u524d\u306e\u5bfe\u5fdc\u8868\u3068\n\u5f62\u3092\u7167\u5408\u3057\u306a\u304c\u3089\u5143\u306e\u91cd\u307f\u306b\u8db3\u3057\u8fbc\u3080(\u914d\u5e03\u5143\u306e\u8aad\u307f\u8fbc\u307f\u51e6\u7406\u306f\u4f7f\u308f\u306a\u3044)\u3002\n\n\u30102\u3064\u306e\u30b3\u30f3\u30c6\u30ca\u306b\u5206\u3051\u308b\u3011(lmw\/lab\/remote_code_remote.py \u3068\u540c\u3058)\n- download: GPU \u306e\u7121\u3044\u30b3\u30f3\u30c6\u30ca\u30fb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u3042\u308a\u3002Hugging Face \u304b\u3089\u3001\u7248\u3092\u56fa\u5b9a\u3057\u3066\u6c7a\u307e\u3063\u305f\u30d5\u30a1\u30a4\u30eb\u3060\u3051\u3092\u843d\u3068\u3059\u3002\n- run: GPU \u306e\u30b3\u30f3\u30c6\u30ca\u30fb**\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u906e\u65ad**\u30fbModal \u306e\u6a5f\u80fd\u3082\u4f7f\u3048\u306a\u3044\u30fbVolume \u306f\u8aad\u307f\u53d6\u308a\u5c02\u7528\u3002\n\n\u3010VRAM \u3092\u7bc0\u7d04\u3059\u308b\u9806\u756a\u3011\u30c6\u30ad\u30b9\u30c8\u30a8\u30f3\u30b3\u30fc\u30c0(UMT5-XXL \u306f 11GB)\u3067\u30d7\u30ed\u30f3\u30d7\u30c8\u3092\u5148\u306b\u5168\u90e8\u57cb\u3081\u8fbc\u307f\u306b\u3057\u3066\u304b\u3089\u6368\u3066\u3001\n\u305d\u306e\u5f8c\u3067\u52d5\u753b\u306e\u672c\u4f53(transformer)\u3068 VAE \u3092\u8aad\u3080\u30022\u3064\u3092\u540c\u6642\u306b GPU \u306b\u8f09\u305b\u306a\u3044\u306e\u3067\u3001\u5b89\u3044 GPU \u306b\u53ce\u307e\u308b\u3002\n\n\u3010\u7d50\u679c\u306e\u8fd4\u3057\u65b9\u3011run \u306f\u751f\u6210\u5668\u3067\u3001\u52d5\u753b\u306e\u30d0\u30a4\u30c8\u5217\u3092 CHUNK_BYTES \u305a\u3064(kind: "chunk")\u3001\u6700\u5f8c\u306b\u8a18\u9332(kind: "record")\u3092\u8fd4\u3059\u3002\nModal \u306f 2MiB \u3092\u8d85\u3048\u308b\u623b\u308a\u5024\u3092\u81ea\u5206\u306e API \u3067\u9810\u3051\u308b\u304c\u3001\u751f\u6210\u306e\u30b3\u30f3\u30c6\u30ca\u306f Modal \u306e\u6a5f\u80fd\u3092\u906e\u65ad\u3057\u3066\u3044\u308b\n(restrict_modal_access)\u306e\u3067\u65ad\u3089\u308c\u308b(2026-10-02\u300114B \u306e4\u672c\u3092\u4f5c\u308a\u7d42\u3048\u305f\u5f8c\u306b 401 \u3067\u5168\u3066\u5931\u3063\u305f)\u30021\u56de\u306e\u623b\u308a\u5024\u3092\u5c0f\u3055\u304f\u4fdd\u3064\u3002\n\n\u3010\u8cbb\u7528\u3092\u6291\u3048\u308b\u3011\u540c\u3058\u8aad\u307f\u8fbc\u307f\u306e\u307e\u307e\u3001LoRA \u3092\u8db3\u3059\u524d(\u5143\u306e\u30e2\u30c7\u30eb)\u3068\u8db3\u3057\u305f\u5f8c\u3092\u7d9a\u3051\u3066\u751f\u6210\u3059\u308b(\u8aad\u307f\u8fbc\u307f\u306f1\u56de)\u3002\n\u52d5\u753b\u306f WordPress \u306e\u30a2\u30c3\u30d7\u30ed\u30fc\u30c9\u306e\u4e0a\u9650\u306b\u53ce\u307e\u308b\u3088\u3046 H.264 \u3067\u5727\u7e2e\u3059\u308b(MAX_BYTES \u3092\u8d85\u3048\u305f\u3089\u753b\u8cea\u3092\u4e0b\u3052\u3066\u4f5c\u308a\u76f4\u3059)\u3002\n\n\u8a18\u4e8b\u57f7\u7b46\u7528\u306e LLM \u306f\u547c\u3070\u306a\u3044\u3002\n"""\nimport hashlib\nimport json\nimport os\nimport platform\nimport subprocess\nimport time\nfrom datetime import datetime, timezone\n\nMOUNT = "\/models"\nPYTHON_VERSION = "3.12"\nDOWNLOAD_PACKAGES = ("huggingface_hub",)\n# diffusers \u306e Wan \u306e\u30d1\u30a4\u30d7\u30e9\u30a4\u30f3\u304c\u4f7f\u3046\u3082\u306e(ftfy \u306f\u30d7\u30ed\u30f3\u30d7\u30c8\u306e\u6574\u5f62\u3001imageio-ffmpeg \u306f H.264 \u306e\u66f8\u304d\u51fa\u3057\u306b\u4f7f\u3046)\nRUN_PACKAGES = ("torch", "diffusers", "transformers", "accelerate", "safetensors", "sentencepiece", "ftfy",\n "imageio-ffmpeg", "numpy")\n# \u843d\u3068\u3057\u3066\u3088\u3044\u30d5\u30a1\u30a4\u30eb(pickle \u306e .bin \/ .pt \/ .ckpt \u306f\u843d\u3068\u3055\u306a\u3044)\nALLOWED_EXTS = (".json", ".safetensors", ".model", ".txt")\n\n# \u5143\u306e Wan \u306e\u5c64\u306e\u540d\u524d(LoRA \u304c\u4f7f\u3046)\u2192 diffusers \u306e WanTransformer3DModel \u306e\u5c64\u306e\u540d\u524d\u3002\n# diffusers \u306e\u5909\u63db\u30b9\u30af\u30ea\u30d7\u30c8(scripts\/convert_wan_to_diffusers.py)\u3068\u540c\u3058\u5bfe\u5fdc\nWAN_LORA_MODULES = {\n "self_attn.q": "attn1.to_q", "self_attn.k": "attn1.to_k", "self_attn.v": "attn1.to_v",\n "self_attn.o": "attn1.to_out.0",\n "cross_attn.q": "attn2.to_q", "cross_attn.k": "attn2.to_k", "cross_attn.v": "attn2.to_v",\n "cross_attn.o": "attn2.to_out.0",\n "ffn.0": "ffn.net.0.proj", "ffn.2": "ffn.net.2",\n}\n# 1\u56de\u306b\u8fd4\u3059\u30d0\u30a4\u30c8\u6570(Modal \u304c\u81ea\u5206\u306e API \u306b\u9810\u3051\u305a\u306b\u305d\u306e\u307e\u307e\u8fd4\u305b\u308b 2MiB \u3088\u308a\u5341\u5206\u5c0f\u3055\u304f)\nCHUNK_BYTES = 1024 * 1024\nLORA_PREFIXES = ("base_model.model.", "diffusion_model.", "model.")\n# \u751f\u6210\u306b\u4f7f\u3046\u30c7\u30d0\u30a4\u30b9\u3002Modal \u3067\u306f\u5e38\u306b cuda(\u5f53\u30b5\u30a4\u30c8\u306e\u8a55\u4fa1\u3067\u3001\u5c0f\u3055\u306a\u30e2\u30c7\u30eb\u3092 CPU \u3067\u901a\u3059\u3068\u304d\u3060\u3051 cpu \u306b\u3059\u308b)\nDEVICE = "cuda"\n\ndef installed_packages() -> list[str]:\n from importlib import metadata\n seen = {}\n for dist in metadata.distributions():\n name = dist.metadata["Name"]\n if name and name.lower() not in seen:\n seen[name.lower()] = f"{name}=={dist.version}"\n return sorted(seen.values(), key=str.lower)\n\ndef _log(t0: float, message: str) -> None:\n print(f" {message}", flush=True)\n\ndef _sha256(path: str) -> str:\n h = hashlib.sha256()\n with open(path, "rb") as f:\n for chunk in iter(lambda: f.read(1 << 20), b""):\n h.update(chunk)\n return h.hexdigest()\n\ndef download(job: dict) -> dict:\n """\n job: {"dir", "base": {"repo", "revision", "files"}, "loras": [{"repo", "revision", "file"}]}\u3002\n \u5143\u306e\u30e2\u30c7\u30eb\u306f <dir>\/base\u3001LoRA \u306f <dir>\/lora<i>\/ \u306b\u7f6e\u304f\u3002\u623b\u308a\u5024: {"bytes", "seconds", "packages", "lora_sha256"}\u3002\n """\n from huggingface_hub import hf_hub_download\n t0 = time.time()\n base_dir = os.path.join(MOUNT, job["dir"])\n total = 0\n targets = [(job["base"]["repo"], job["base"]["revision"], name, os.path.join(base_dir, "base"))\n for name in job["base"]["files"]]\n targets += [(lora["repo"], lora["revision"], lora["file"], os.path.join(base_dir, f"lora{i}"))\n for i, lora in enumerate(job.get("loras") or [])]\n for repo, revision, name, dest in targets:\n if not name.endswith(ALLOWED_EXTS) or ".." in name.split("\/"):\n raise ValueError(f"\u843d\u3068\u3055\u306a\u3044\u7a2e\u985e\u306e\u30d5\u30a1\u30a4\u30eb\u3067\u3059: {name}")\n path = hf_hub_download(repo, name, revision=revision, local_dir=dest)\n total += os.path.getsize(path)\n lora_sha = [_sha256(os.path.join(base_dir, f"lora{i}", lora["file"]))\n for i, lora in enumerate(job.get("loras") or [])]\n seconds = round(time.time() - t0)\n _log(t0, f"\u30c0\u30a6\u30f3\u30ed\u30fc\u30c9 {total \/ 1024 ** 3:.1f}GB\u30fb{seconds}\u79d2")\n return {"bytes": int(total), "seconds": int(seconds), "packages": installed_packages(), "lora_sha256": lora_sha}\n\ndef _lora_pairs(state: dict) -> dict:\n """{\u5143\u306e\u5c64\u306e\u540d\u524d: {"A", "B", "alpha"?}}\u3002\u540d\u524d\u306e\u982d(base_model.model. \u7b49)\u3092\u5916\u3057\u3001A \u3068 B \u306e\u63c3\u308f\u306a\u3044\u5c64\u306f\u65ad\u308b\u3002"""\n pairs: dict = {}\n for key, tensor in state.items():\n name = key\n for prefix in LORA_PREFIXES:\n if name.startswith(prefix):\n name = name[len(prefix):]\n break\n if name.endswith(".lora_A.weight"):\n pairs.setdefault(name[:-len(".lora_A.weight")], {})["A"] = tensor\n elif name.endswith(".lora_B.weight"):\n pairs.setdefault(name[:-len(".lora_B.weight")], {})["B"] = tensor\n elif name.endswith(".alpha"):\n pairs.setdefault(name[:-len(".alpha")], {})["alpha"] = float(tensor.item())\n else:\n raise ValueError(f"LoRA \u306e\u91cd\u307f\u3067\u306f\u306a\u3044\u540d\u524d\u3067\u3059: {key}")\n broken = [n for n, p in pairs.items() if "A" not in p or "B" not in p]\n if broken:\n raise ValueError(f"A \u3068 B \u304c\u63c3\u308f\u306a\u3044\u5c64\u304c\u3042\u308a\u307e\u3059: {broken[:5]}")\n return pairs\n\ndef diffusers_module_name(name: str) -> str:\n """blocks.12.self_attn.q \u2192 blocks.12.attn1.to_q\u3002\u5bfe\u5fdc\u8868\u306b\u7121\u3044\u5c64\u306f\u65ad\u308b(\u63a8\u6e2c\u3067\u5f53\u3066\u306a\u3044)\u3002"""\n parts = name.split(".")\n if len(parts) < 4 or parts[0] != "blocks" or not parts[1].isdigit():\n raise ValueError(f"\u5bfe\u5fdc\u8868\u306b\u7121\u3044\u5c64\u3067\u3059: {name}")\n tail = ".".join(parts[2:])\n if tail not in WAN_LORA_MODULES:\n raise ValueError(f"\u5bfe\u5fdc\u8868\u306b\u7121\u3044\u5c64\u3067\u3059: {name}")\n return f"blocks.{parts[1]}.{WAN_LORA_MODULES[tail]}"\n\ndef merge_lora(transformer, path: str, weight: float, alpha: float | None, rank: int | None) -> dict:\n """\n LoRA \u3092\u5143\u306e\u91cd\u307f\u306b\u8db3\u3057\u8fbc\u3080: W \u2190 W + weight \u00d7 (alpha \/ r) \u00d7 (B @ A)\u3002alpha \u3068 r \u306f\u914d\u5e03\u5143\u306e adapter_config.json \u306e\u5024\u3092\n \u30db\u30b9\u30c8\u5074\u304c\u6e21\u3059(\u91cd\u307f\u306e\u30d5\u30a1\u30a4\u30eb\u306b alpha \u304c\u3042\u308c\u3070\u305d\u3061\u3089\u3092\u4f7f\u3046)\u3002\u8db3\u3057\u8fbc\u307f\u306f float32 \u3067\u8a08\u7b97\u3057\u3066\u304b\u3089\u5143\u306e\u578b\u306b\u623b\u3059\u3002\n \u5168\u3066\u306e\u5c64\u304c\u5bfe\u5fdc\u8868\u3067\u898b\u3064\u304b\u308a\u3001\u5f62\u304c\u4e00\u81f4\u3059\u308b\u3053\u3068\u3092\u78ba\u304b\u3081\u308b(1\u3064\u3067\u3082\u5916\u308c\u305f\u3089\u65ad\u308b)\u3002\n """\n import torch\n from safetensors.torch import load_file\n pairs = _lora_pairs(load_file(path))\n merged = 0\n for name, p in sorted(pairs.items()):\n module = transformer.get_submodule(diffusers_module_name(name))\n a, b = p["A"].float(), p["B"].float()\n r = a.shape[0]\n if rank and r != rank:\n raise ValueError(f"{name}: \u30e9\u30f3\u30af {r} \u304c adapter_config.json \u306e {rank} \u3068\u9055\u3044\u307e\u3059")\n scale = weight * ((p.get("alpha") or alpha or r) \/ r)\n w = module.weight\n delta = (b.to(w.device) @ a.to(w.device)) * scale\n if tuple(delta.shape) != tuple(w.shape):\n raise ValueError(f"{name}: \u5f62 {tuple(delta.shape)} \u304c\u5143\u306e\u91cd\u307f {tuple(w.shape)} \u3068\u9055\u3044\u307e\u3059")\n with torch.no_grad():\n w.copy_((w.float() + delta).to(w.dtype))\n merged += 1\n return {"modules": merged, "weight": weight, "scale": round(weight * ((alpha or rank or 1) \/ (rank or 1)), 6)}\n\ndef encode_mp4(frames, fps: int, path: str, max_bytes: int) -> dict:\n """\n 0\u301c1 \u306e RGB \u306e\u30b3\u30de(F, H, W, 3)\u3092 H.264(yuv420p)\u306e mp4 \u306b\u3059\u308b\u3002max_bytes \u3092\u8d85\u3048\u305f\u3089\u3001\u30d3\u30c3\u30c8\u30ec\u30fc\u30c8\u306e\u4e0a\u9650\u3092\n \u4e0b\u3052\u3066\u4f5c\u308a\u76f4\u3059(3\u56de\u307e\u3067)\u3002\u623b\u308a\u5024: \u5b9f\u969b\u306b\u4f7f\u3063\u305f ffmpeg \u306e\u5f15\u6570\u3068\u5927\u304d\u3055\u3002\n """\n import imageio_ffmpeg\n import numpy as np\n pixels = (np.clip(frames, 0.0, 1.0) * 255).round().astype(np.uint8)\n count, height, width = pixels.shape[0], pixels.shape[1], pixels.shape[2]\n duration = count \/ fps\n # \u4e0a\u9650\u306e\u5927\u304d\u3055\u306e 85% \u306b\u53ce\u307e\u308b\u30d3\u30c3\u30c8\u30ec\u30fc\u30c8(\u97f3\u58f0\u306a\u3057)\n kbps = int(max_bytes * 8 * 0.85 \/ duration \/ 1000)\n for attempt, crf in enumerate((18, 22, 26)):\n rate = int(kbps * (0.8 ** attempt))\n args = ["-y", "-f", "rawvideo", "-pix_fmt", "rgb24", "-s", f"{width}x{height}", "-r", str(fps), "-i", "-",\n "-an", "-c:v", "libx264", "-preset", "slow", "-crf", str(crf), "-maxrate", f"{rate}k",\n "-bufsize", f"{rate * 2}k", "-pix_fmt", "yuv420p", "-movflags", "+faststart", path]\n proc = subprocess.run([imageio_ffmpeg.get_ffmpeg_exe(), *args], input=pixels.tobytes(),\n capture_output=True, timeout=600)\n if proc.returncode != 0:\n raise RuntimeError(f"ffmpeg \u304c\u5931\u6557\u3057\u307e\u3057\u305f: {proc.stderr.decode(errors='replace')[-400:]}")\n size = os.path.getsize(path)\n if size <= max_bytes:\n return {"ffmpeg_args": ["ffmpeg", *args[:-1], os.path.basename(path)], "bytes": size,\n "frames": int(count), "width": int(width), "height": int(height), "fps": int(fps),\n "duration": round(duration, 2)}\n raise ValueError(f"\u52d5\u753b\u3092 {max_bytes} \u30d0\u30a4\u30c8\u4ee5\u4e0b\u306b\u5727\u7e2e\u3067\u304d\u307e\u305b\u3093\u3067\u3057\u305f")\n\ndef run(job: dict):\n """\n job: {"dir", "pipeline", "settings": {height, width, num_frames, fps, steps, guidance, flow_shift?},\n "seed", "prompts": [{"id", "en"}], "loras": [{"file", "weight", "alpha", "rank"}], "compare_base",\n "max_bytes"}\u3002\n \u751f\u6210\u5668: {"kind": "chunk", "id", "index", "data": bytes} \u3092\u52d5\u753b\u3054\u3068\u306b\u9806\u306b\u8fd4\u3057\u3001\u6700\u5f8c\u306b\n {"kind": "record", ...\u8a18\u9332\u306e\u4e00\u90e8(\u79d2\u6570\u30fbVRAM\u30fb\u7248\u3002\u52d5\u753b\u306e\u4e2d\u8eab\u306f\u542b\u307e\u306a\u3044)} \u3092\u8fd4\u3059\u3002\n """\n import diffusers\n import numpy as np\n import torch\n import transformers\n os.environ["HF_HUB_OFFLINE"] = "1"\n os.environ["TRANSFORMERS_OFFLINE"] = "1"\n os.environ["HF_HUB_DISABLE_TELEMETRY"] = "1"\n t0 = time.time()\n base_dir = os.path.join(MOUNT, job["dir"])\n model_dir = os.path.join(base_dir, "base")\n settings = job["settings"]\n guidance = float(settings["guidance"])\n cuda = DEVICE == "cuda"\n gpu = torch.cuda.get_device_name(0) if cuda else "cpu"\n _log(t0, f"GPU: {gpu}\u30fbdiffusers {diffusers.__version__}\u30fbtorch {torch.__version__}")\n record = {"gpu": gpu, "vram_total": int(torch.cuda.get_device_properties(0).total_memory) if cuda else 0,\n "diffusers": str(diffusers.__version__), "transformers": str(transformers.__version__),\n "torch": str(torch.__version__), "dtype": "bfloat16", "vae_dtype": "float32",\n "packages": installed_packages(), "python": platform.python_version(),\n "measured_at": datetime.now(timezone.utc).isoformat()}\n pipeline_cls = getattr(diffusers, job["pipeline"])\n\n # 1. \u30d7\u30ed\u30f3\u30d7\u30c8\u3092\u57cb\u3081\u8fbc\u307f\u306b\u3059\u308b(\u30c6\u30ad\u30b9\u30c8\u30a8\u30f3\u30b3\u30fc\u30c0\u3060\u3051\u3092 GPU \u306b\u8f09\u305b\u3001\u7d42\u308f\u3063\u305f\u3089\u6368\u3066\u308b)\n with open(os.path.join(model_dir, "model_index.json"), encoding="utf-8") as f:\n index = json.load(f)\n skip_models = {name: None for name, spec in index.items()\n if isinstance(spec, list) and name not in ("text_encoder", "tokenizer", "scheduler")}\n text_pipe = pipeline_cls.from_pretrained(model_dir, torch_dtype=torch.bfloat16, use_safetensors=True,\n local_files_only=True, **skip_models)\n text_pipe.text_encoder.to(DEVICE)\n embeds = {}\n with torch.no_grad():\n for p in job["prompts"]:\n # max_sequence_length \u306f\u30d1\u30a4\u30d7\u30e9\u30a4\u30f3\u306e __call__ \u306e\u65e2\u5b9a(512\u3002\u5143\u306e Wan \u3082 512)\u306b\u5408\u308f\u305b\u308b\u3002\n # encode_prompt \u306e\u65e2\u5b9a\u306f 226 \u3067\u3001\u305d\u306e\u307e\u307e\u3060\u3068\u6a19\u6e96\u306e\u547c\u3073\u51fa\u3057\u3068\u9055\u3046\u57cb\u3081\u8fbc\u307f\u306b\u306a\u308b\n pos, neg = text_pipe.encode_prompt(prompt=p["en"], negative_prompt="", device=DEVICE,\n do_classifier_free_guidance=guidance > 1.0,\n max_sequence_length=int(settings.get("max_sequence_length", 512)))\n embeds[p["id"]] = (pos.cpu(), None if neg is None else neg.cpu())\n del text_pipe\n if cuda:\n torch.cuda.empty_cache()\n record["encode_sec"] = round(time.time() - t0)\n\n # 2. \u672c\u4f53\u3068 VAE \u3092\u8aad\u3080(VAE \u306f float32\u3002diffusers \u306e Wan \u306e\u4f8b\u3068\u540c\u3058)\n vae_cls = getattr(diffusers, index["vae"][1])\n vae = vae_cls.from_pretrained(model_dir, subfolder="vae", torch_dtype=torch.float32, use_safetensors=True,\n local_files_only=True)\n pipe = pipeline_cls.from_pretrained(model_dir, vae=vae, text_encoder=None, tokenizer=None,\n torch_dtype=torch.bfloat16, use_safetensors=True, local_files_only=True)\n if settings.get("flow_shift") is not None:\n pipe.scheduler = type(pipe.scheduler).from_config(pipe.scheduler.config, flow_shift=float(settings["flow_shift"]))\n record["scheduler"] = {"class": type(pipe.scheduler).__name__,\n **{k: v for k, v in dict(pipe.scheduler.config).items()\n if k in ("flow_shift", "use_flow_sigmas", "prediction_type", "solver_order")}}\n pipe.to(DEVICE)\n if cuda:\n torch.cuda.synchronize()\n record["vram_loaded"] = int(torch.cuda.memory_allocated()) if cuda else 0\n record["load_sec"] = round(time.time() - t0)\n _log(t0, f"\u8aad\u307f\u8fbc\u307f: VRAM {record['vram_loaded'] \/ 1024 ** 3:.1f}GB\u30fb{record['load_sec']}\u79d2")\n if cuda:\n torch.cuda.reset_peak_memory_stats()\n\n def generate(group: str) -> list[dict]:\n out = []\n for p in job["prompts"]:\n pos, neg = embeds[p["id"]]\n started = time.time()\n generator = torch.Generator("cpu").manual_seed(int(job["seed"]))\n frames = pipe(prompt_embeds=pos.to(DEVICE),\n negative_prompt_embeds=None if neg is None else neg.to(DEVICE),\n height=int(settings["height"]), width=int(settings["width"]),\n num_frames=int(settings["num_frames"]), num_inference_steps=int(settings["steps"]),\n guidance_scale=guidance, generator=generator, output_type="np").frames[0]\n if cuda:\n torch.cuda.synchronize()\n seconds = time.time() - started\n frames = np.asarray(frames, dtype=np.float32)\n if not np.isfinite(frames).all():\n raise FloatingPointError("\u751f\u6210\u3057\u305f\u30b3\u30de\u306b\u6570\u3067\u306a\u3044\u5024\u304c\u3042\u308a\u307e\u3059")\n # \u51e6\u7406\u306e\u8aa4\u308a(\u771f\u3063\u9ed2\u30fb\u4e00\u8272\u306e\u52d5\u753b)\u3092\u53d6\u308a\u8fbc\u307f\u5f79\u306e\u6a5f\u68b0\u306e\u691c\u67fb(lmw\/lab\/sanity.py)\u304c\u898b\u3064\u3051\u308b\u305f\u3081\u306e\u5024\n stats = {"pixel_mean": round(float(frames.mean()) * 255, 2), "pixel_std": round(float(frames.std()) * 255, 2)}\n path = os.path.join("\/tmp", f"{group}-{p['id']}.mp4")\n encoded = encode_mp4(frames, int(settings["fps"]), path, int(job["max_bytes"]))\n with open(path, "rb") as f:\n raw = f.read()\n blobs[f"{group}-{p['id']}"] = raw\n out.append({"id": f"{group}-{p['id']}", "prompt_id": p["id"], "group": group,\n "seconds": round(seconds, 1), **encoded, **stats,\n "sha256": hashlib.sha256(raw).hexdigest()})\n _log(t0, f"{group}\/{p['id']}: {encoded['frames']}\u30b3\u30de\u30fb\u751f\u6210 {seconds:.0f}\u79d2\u30fb{encoded['bytes'] \/ 1e6:.2f}MB")\n return out\n\n videos = []\n blobs: dict[str, bytes] = {}\n loras = job.get("loras") or []\n # 3. LoRA \u3092\u8db3\u3059\u524d\u306e\u5143\u306e\u30e2\u30c7\u30eb(\u6bd4\u3079\u308b\u305f\u3081\u3002\u540c\u3058\u30d7\u30ed\u30f3\u30d7\u30c8\u30fb\u30b7\u30fc\u30c9\u30fb\u30b9\u30c6\u30c3\u30d7\u6570)\n if loras and job.get("compare_base"):\n videos += generate("base")\n # 4. LoRA \u3092\u8db3\u3057\u8fbc\u3093\u3067\u304b\u3089\u751f\u6210\u3059\u308b\n merges = []\n for i, lora in enumerate(loras):\n path = os.path.join(base_dir, f"lora{i}", lora["file"])\n merges.append(merge_lora(pipe.transformer, path, float(lora["weight"]), lora.get("alpha"), lora.get("rank")))\n _log(t0, f"LoRA {i}: {merges[-1]['modules']}\u5c64\u3092\u8db3\u3057\u8fbc\u307f\u307e\u3057\u305f(\u91cd\u307f {lora['weight']})")\n if merges:\n record["merges"] = merges\n videos += generate("lora" if loras else "model")\n record["vram_peak"] = int(torch.cuda.max_memory_allocated()) if cuda else 0\n record["videos"] = videos\n for vid, raw in blobs.items():\n for index in range(0, len(raw), CHUNK_BYTES):\n yield {"kind": "chunk", "id": vid, "index": index \/\/ CHUNK_BYTES, "data": raw[index:index + CHUNK_BYTES]}\n yield {"kind": "record", **plain(record)}\n\ndef plain(value):\n """JSON \u306b\u901a\u3057\u3066\u3001\u6a19\u6e96\u306e\u578b\u3060\u3051\u306b\u3059\u308b(\u30db\u30b9\u30c8\u306b\u306f torch \u3082 diffusers \u3082\u7121\u3044)\u3002"""\n return json.loads(json.dumps(value, default=str))\n<\/code><\/pre>\nThe job<\/code> passed to download<\/code> (as used):<\/p>\n{\n "dir": "video-1790915190",\n "base": {\n "repo": "Wan-AI\/Wan2.2-TI2V-5B-Diffusers",\n "revision": "b8fff7315c768468a5333511427288870b2e9635",\n "files": [\n "model_index.json",\n "scheduler\/scheduler_config.json",\n "text_encoder\/config.json",\n "text_encoder\/model-00001-of-00003.safetensors",\n "text_encoder\/model-00002-of-00003.safetensors",\n "text_encoder\/model-00003-of-00003.safetensors",\n "text_encoder\/model.safetensors.index.json",\n "tokenizer\/special_tokens_map.json",\n "tokenizer\/spiece.model",\n "tokenizer\/tokenizer.json",\n "tokenizer\/tokenizer_config.json",\n "transformer\/config.json",\n "transformer\/diffusion_pytorch_model-00001-of-00005.safetensors",\n "transformer\/diffusion_pytorch_model-00002-of-00005.safetensors",\n "transformer\/diffusion_pytorch_model-00003-of-00005.safetensors",\n "transformer\/diffusion_pytorch_model-00004-of-00005.safetensors",\n "transformer\/diffusion_pytorch_model-00005-of-00005.safetensors",\n "transformer\/diffusion_pytorch_model.safetensors.index.json",\n "vae\/config.json",\n "vae\/diffusion_pytorch_model.safetensors"\n ]\n },\n "loras": [\n {\n "repo": "Efficient-Large-Model\/LongLive-Plug-Wan2.2-TI2V-5B-few-step",\n "revision": "38a6ec4e8f5b644749d3072bfe1ac383d00951c0",\n "file": "adapter_model.safetensors"\n },\n {\n "repo": "Efficient-Large-Model\/LongLive-Plug-Wan2.2-TI2V-5B-cfg",\n "revision": "fa1f9280778aa27c8038fe06e29fb5f2ed6be451",\n "file": "adapter_model.safetensors"\n }\n ]\n}\n<\/code><\/pre>\nThe job<\/code> passed to run<\/code> (as used):<\/p>\n{\n "dir": "video-1790915190",\n "pipeline": "WanPipeline",\n "settings": {\n "height": 704,\n "width": 1280,\n "num_frames": 121,\n "fps": 24,\n "steps": 4,\n "guidance": 1.0,\n "flow_shift": 5.0\n },\n "seed": 0,\n "prompts": [\n {\n "id": "boat",\n "en": "a small red paper boat drifting down a clear forest stream, sunlight flickering through the leaves, gentle ripples, realistic"\n },\n {\n "id": "clouds",\n "en": "time-lapse of white clouds rolling over snowy mountain peaks at sunrise, golden light, wide shot"\n }\n ],\n "loras": [\n {\n "file": "adapter_model.safetensors",\n "weight": 1.0,\n "alpha": 128,\n "rank": 128\n },\n {\n "file": "adapter_model.safetensors",\n "weight": 0.5,\n "alpha": 64,\n "rank": 64\n }\n ],\n "compare_base": true,\n "max_bytes": 12000000\n}\n<\/code><\/pre>\n\nPackage versions in the generating container (full list)<\/summary>\naccelerate==1.15.0 aiohappyeyeballs==2.6.1 aiohttp==3.12.7 aiosignal==1.3.2 annotated-doc==0.0.5 anyio==4.15.1 attrs==25.3.0 cbor2==5.7.0 certifi==2026.7.22 charset-normalizer==3.5.2 click==8.5.0 cuda-bindings==13.4.3 cuda-pathfinder==1.8.2 cuda-toolkit==13.0.3.0 diffusers==0.40.0 filelock==4.0.9 frozenlist==1.6.0 fsspec==2026.9.0 ftfy==6.3.1 grpclib==0.4.8 h11==0.16.0 h2==4.2.0 hf-xet==1.6.0 hpack==4.1.0 httpcore==1.0.9 httpx==0.28.1 huggingface_hub==1.33.0 hyperframe==6.1.0 idna==3.20 imageio-ffmpeg==0.6.0 importlib_metadata==9.0.1 Jinja2==3.1.6 markdown-it-py==4.2.0 MarkupSafe==3.0.3 mdurl==0.1.2 mpmath==1.3.0 multidict==6.4.4 networkx==3.7 numpy==2.5.3 nvidia-cublas==13.1.1.3 nvidia-cuda-cupti==13.0.85 nvidia-cuda-nvrtc==13.0.88 nvidia-cuda-runtime==13.0.96 nvidia-cudnn-cu13==9.24.0.43 nvidia-cufft==12.0.0.61 nvidia-cufile==1.15.1.6 nvidia-curand==10.4.0.35 nvidia-cusolver==12.0.4.66 nvidia-cusparse==12.6.3.3 nvidia-cusparselt-cu13==0.8.1 nvidia-nccl-cu13==2.30.7 nvidia-nvjitlink==13.4.92 nvidia-nvshmem-cu13==3.4.5 nvidia-nvtx==13.0.85 packaging==26.3 pillow==12.3.0 pip==25.1.1 propcache==0.3.1 protobuf==6.31.1 psutil==7.2.2 Pygments==2.21.0 PyYAML==6.0.3 regex==2026.9.29 requests==2.34.2 rich==15.0.0 safetensors==0.8.0 sentencepiece==0.2.2 setuptools==84.0.0 shellingham==1.5.4 sympy==1.14.0 tokenizers==0.23.2 torch==2.14.1 tqdm==4.70.1 transformers==5.18.0 triton==3.8.0 typer==0.27.2 typing_extensions==4.16.0 urllib3==2.8.0 uv==0.7.19 wcwidth==0.9.1 wheel==0.45.1 yarl==1.20.0 zipp==4.1.0<\/code><\/pre>\n<\/details>\n\nPackage versions in the downloading container (full list)<\/summary>\naiohappyeyeballs==2.6.1 aiohttp==3.12.7 aiosignal==1.3.2 anyio==4.15.1 attrs==25.3.0 cbor2==5.7.0 certifi==2025.4.26 click==8.5.0 filelock==4.0.7 frozenlist==1.6.0 fsspec==2026.9.0 grpclib==0.4.8 h11==0.16.0 h2==4.2.0 hf-xet==1.6.0 hpack==4.1.0 httpcore2==2.13.1 httpx2==2.13.1 huggingface_hub==2.0.0 hyperframe==6.1.0 idna==3.20 multidict==6.4.4 packaging==26.3 pip==25.1.1 propcache==0.3.1 protobuf==6.31.1 PyYAML==6.0.3 tqdm==4.70.1 truststore==0.10.4 typing_extensions==4.16.0 uv==0.7.19 wheel==0.45.1 yarl==1.20.0<\/code><\/pre>\n<\/details>\n<\/p>\n <\/p>\n
|