File size: 14,921 Bytes
384e4ac |
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 |
# model_management.py
"""
Model management utilities for MagentaRT API.
This module handles checkpoint discovery, asset loading, and model selection logic.
It is designed to work with the global state managed in app.py without interfering
with the critical JAX/XLA initialization sequence.
"""
import os
import re
import logging
from pathlib import Path
from typing import Optional, Union, Literal, Tuple, List
import tarfile
import numpy as np
from pydantic import BaseModel
from huggingface_hub import snapshot_download, HfApi, hf_hub_download
# ---- Constants and Patterns ----
_FINETUNE_REPO_DEFAULT = os.getenv("MRT_ASSETS_REPO", "thepatch/magenta-ft")
_STEP_RE = re.compile(r"(?:^|/)checkpoint_(\d+)(?:/|\.tar\.gz|\.tgz)?$")
# ---- Pydantic Models ----
class ModelSelect(BaseModel):
size: Optional[Literal["base","large"]] = None
repo_id: Optional[str] = None
revision: Optional[str] = "main"
step: Optional[Union[int, str]] = None # allow "latest"
assets_repo_id: Optional[str] = None # default: follow repo_id
sync_assets: bool = True # load mean/centroids from repo
prewarm: bool = False # call get_mrt() to build right away
stop_active: bool = True # auto-stop jams; else 409
dry_run: bool = False # validate only, don't swap
# ---- Checkpoint Discovery ----
class CheckpointManager:
"""Handles checkpoint discovery and validation without modifying global state."""
@staticmethod
def list_ckpt_steps(repo_id: str, revision: str = "main") -> List[int]:
"""
List available checkpoint steps in a HF model repo without downloading all weights.
Looks for:
checkpoint_<step>/
checkpoint_<step>.tgz | .tar.gz
archives/checkpoint_<step>.tgz | .tar.gz
"""
api = HfApi()
files = api.list_repo_files(repo_id=repo_id, repo_type="model", revision=revision)
steps = set()
for f in files:
m = _STEP_RE.search(f)
if m:
try:
steps.add(int(m.group(1)))
except:
pass
return sorted(steps)
@staticmethod
def step_exists(repo_id: str, revision: str, step: int) -> bool:
"""Check if a specific checkpoint step exists in the repo."""
return step in CheckpointManager.list_ckpt_steps(repo_id, revision)
@staticmethod
def resolve_checkpoint_dir() -> Optional[str]:
"""
Resolve the checkpoint directory from environment variables.
Downloads and extracts if necessary.
Returns the path to the checkpoint directory or None if not configured.
"""
repo_id = os.getenv("MRT_CKPT_REPO")
if not repo_id:
return None
step = os.getenv("MRT_CKPT_STEP") # e.g. "1863001"
root = Path(snapshot_download(
repo_id=repo_id,
repo_type="model",
revision=os.getenv("MRT_CKPT_REV", "main"),
local_dir="/home/appuser/.cache/mrt_ckpt/repo",
local_dir_use_symlinks=False,
))
# Prefer an archive if present (more reliable for Zarr/T5X)
arch_names = [
f"checkpoint_{step}.tgz",
f"checkpoint_{step}.tar.gz",
f"archives/checkpoint_{step}.tgz",
f"archives/checkpoint_{step}.tar.gz",
] if step else []
cache_root = Path("/home/appuser/.cache/mrt_ckpt/extracted")
cache_root.mkdir(parents=True, exist_ok=True)
for name in arch_names:
arch = root / name
if arch.is_file():
out_dir = cache_root / f"checkpoint_{step}"
marker = out_dir.with_suffix(".ok")
if not marker.exists():
out_dir.mkdir(parents=True, exist_ok=True)
with tarfile.open(arch, "r:*") as tf:
tf.extractall(out_dir)
marker.write_text("ok")
# sanity: require .zarray to exist inside the extracted tree
if not any(out_dir.rglob(".zarray")):
raise RuntimeError(f"Extracted archive missing .zarray files: {out_dir}")
return str(out_dir / f"checkpoint_{step}") if (out_dir / f"checkpoint_{step}").exists() else str(out_dir)
# No archive; try raw folder from repo and sanity check.
if step:
raw = root / f"checkpoint_{step}"
if raw.is_dir():
if not any(raw.rglob(".zarray")):
raise RuntimeError(
f"Downloaded checkpoint_{step} appears incomplete (no .zarray). "
"Upload as a .tgz or push via git from a Unix shell."
)
return str(raw)
# Pick latest if no step
step_dirs = [d for d in root.iterdir() if d.is_dir() and re.match(r"checkpoint_\d+$", d.name)]
if step_dirs:
pick = max(step_dirs, key=lambda d: int(d.name.split('_')[-1]))
if not any(pick.rglob(".zarray")):
raise RuntimeError(f"Downloaded {pick} appears incomplete (no .zarray).")
return str(pick)
return None
# ---- Asset Management ----
class AssetManager:
"""
Handles finetune asset loading and management.
This class modifies global variables in the calling module, but encapsulates
the logic for loading and validating assets.
"""
def __init__(self):
# These will be set by the calling module
self.mean_embed = None
self.centroids = None
self.assets_repo_id = None
def load_finetune_assets_from_hf(self, repo_id: Optional[str], mrt=None) -> Tuple[bool, str]:
"""
Download & load mean_style_embed.npy and cluster_centroids.npy from a HF model repo.
Safe to call multiple times; will overwrite instance vars if successful.
Args:
repo_id: HuggingFace repo ID, defaults to _FINETUNE_REPO_DEFAULT
mrt: MagentaRT instance for dimension validation (optional)
Returns:
Tuple of (success: bool, message: str)
"""
repo_id = repo_id or _FINETUNE_REPO_DEFAULT
try:
mean_path = None
cent_path = None
try:
mean_path = hf_hub_download(repo_id, filename="mean_style_embed.npy", repo_type="model")
except Exception:
pass
try:
cent_path = hf_hub_download(repo_id, filename="cluster_centroids.npy", repo_type="model")
except Exception:
pass
if mean_path is None and cent_path is None:
return False, f"No finetune asset files found in repo {repo_id}"
if mean_path is not None:
m = np.load(mean_path)
if m.ndim != 1:
return False, f"mean_style_embed.npy must be 1-D (got {m.shape})"
else:
m = None
if cent_path is not None:
c = np.load(cent_path)
if c.ndim != 2:
return False, f"cluster_centroids.npy must be 2-D (got {c.shape})"
else:
c = None
# Optional: shape check vs model embedding dim once model is alive
if mrt is not None:
try:
d = int(mrt.style_model.config.embedding_dim)
if m is not None and m.shape[0] != d:
return False, f"mean_style_embed dim {m.shape[0]} != model dim {d}"
if c is not None and c.shape[1] != d:
return False, f"cluster_centroids dim {c.shape[1]} != model dim {d}"
except Exception:
# Model not built yet; we'll trust the files and rely on runtime checks later
pass
# Update instance variables
self.mean_embed = m.astype(np.float32, copy=False) if m is not None else None
self.centroids = c.astype(np.float32, copy=False) if c is not None else None
self.assets_repo_id = repo_id
logging.info("Loaded finetune assets from %s (mean=%s, centroids=%s)",
repo_id,
"yes" if self.mean_embed is not None else "no",
f"{self.centroids.shape[0]}x{self.centroids.shape[1]}" if self.centroids is not None else "no")
return True, "ok"
except Exception as e:
logging.exception("Failed to load finetune assets: %s", e)
return False, str(e)
def ensure_assets_loaded(self, mrt=None):
"""Best-effort lazy load if nothing is loaded yet."""
if self.mean_embed is None and self.centroids is None:
self.load_finetune_assets_from_hf(self.assets_repo_id or _FINETUNE_REPO_DEFAULT, mrt)
def get_status(self, mrt=None) -> dict:
"""Get current asset status."""
d = None
if mrt is not None:
try:
d = int(mrt.style_model.config.embedding_dim)
except Exception:
pass
return {
"repo_id": self.assets_repo_id,
"mean_loaded": self.mean_embed is not None,
"centroids_loaded": self.centroids is not None,
"centroid_count": None if self.centroids is None else int(self.centroids.shape[0]),
"embedding_dim": d,
}
# ---- Model Selection Logic ----
class ModelSelector:
"""
Handles model selection and validation logic.
This class encapsulates the complex logic from the /model/select endpoint
while keeping environment variable management in the calling code.
"""
def __init__(self, checkpoint_manager: CheckpointManager, asset_manager: AssetManager):
self.checkpoint_manager = checkpoint_manager
self.asset_manager = asset_manager
def validate_selection(self, req: ModelSelect) -> Tuple[bool, dict]:
"""
Validate a model selection request without making any changes.
Returns:
Tuple of (success: bool, result_dict: dict)
"""
# Current env defaults
cur = {
"size": os.getenv("MRT_SIZE", "large"),
"repo": os.getenv("MRT_CKPT_REPO"),
"rev": os.getenv("MRT_CKPT_REV", "main"),
"step": os.getenv("MRT_CKPT_STEP"),
"assets": os.getenv("MRT_ASSETS_REPO", _FINETUNE_REPO_DEFAULT),
}
# Flags for special step values
no_ckpt = isinstance(req.step, str) and req.step.lower() == "none"
latest = isinstance(req.step, str) and req.step.lower() == "latest"
# Target selection
tgt = {
"size": req.size or cur["size"],
"repo": None if no_ckpt else (req.repo_id or cur["repo"]),
"rev": req.revision if req.revision is not None else cur["rev"],
"step": None if (no_ckpt or latest) else (str(req.step) if req.step is not None else cur["step"]),
"assets": req.assets_repo_id or req.repo_id or cur["assets"],
}
# Case 1: No checkpoint (stock model)
if no_ckpt:
return True, {
"target_size": tgt["size"],
"target_repo": None,
"target_revision": None,
"target_step": None,
"assets_repo": None,
"assets_probe": {"ok": True, "message": "skipped"},
}
# Case 2: Checkpoint selection
if not tgt["repo"]:
return False, {"error": "repo_id is required for model selection."}
# Enumerate available steps
try:
steps = self.checkpoint_manager.list_ckpt_steps(tgt["repo"], tgt["rev"])
except Exception as e:
return False, {"error": f"Failed to list checkpoints: {e}"}
if not steps:
return False, {
"error": f"No checkpoint files found in {tgt['repo']}@{tgt['rev']}",
"discovered_steps": steps
}
# Choose step (explicit or latest)
chosen_step = int(tgt["step"]) if tgt["step"] is not None else steps[-1]
if chosen_step not in steps:
return False, {
"error": f"checkpoint_{chosen_step} not present in {tgt['repo']}@{tgt['rev']}",
"discovered_steps": steps
}
# Optional finetune assets probe
assets_ok, assets_msg = True, "skipped"
if req.sync_assets:
try:
api = HfApi()
files = set(api.list_repo_files(repo_id=tgt["assets"], repo_type="model"))
if ("mean_style_embed.npy" not in files) and ("cluster_centroids.npy" not in files):
assets_ok, assets_msg = False, f"No finetune asset files in {tgt['assets']}"
else:
assets_msg = "found"
except Exception as e:
assets_ok, assets_msg = False, f"probe failed: {e}"
return True, {
"target_size": tgt["size"],
"target_repo": tgt["repo"],
"target_revision": tgt["rev"],
"target_step": chosen_step,
"assets_repo": tgt["assets"] if req.sync_assets else None,
"assets_probe": {"ok": assets_ok, "message": assets_msg},
}
def prepare_env_changes(self, req: ModelSelect, validation_result: dict) -> dict:
"""
Prepare the environment variable changes needed for a model selection.
Args:
req: The model selection request
validation_result: Result from validate_selection()
Returns:
Dictionary of environment variable changes to apply
"""
no_ckpt = isinstance(req.step, str) and req.step.lower() == "none"
if no_ckpt:
# Clear checkpoint env vars for stock model
return {
"MRT_SIZE": validation_result["target_size"],
"MRT_CKPT_REPO": None, # None means delete the env var
"MRT_CKPT_REV": None,
"MRT_CKPT_STEP": None,
"MRT_ASSETS_REPO": None,
}
else:
# Set checkpoint env vars
env_changes = {
"MRT_SIZE": validation_result["target_size"],
"MRT_CKPT_REPO": validation_result["target_repo"],
"MRT_CKPT_REV": validation_result["target_revision"],
"MRT_CKPT_STEP": str(validation_result["target_step"]),
}
if req.sync_assets:
env_changes["MRT_ASSETS_REPO"] = validation_result["assets_repo"]
return env_changes |