Spaces:
Running
Running
File size: 4,870 Bytes
e70c280 | 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 | from __future__ import annotations
import os
import subprocess
import sys
import time
from pathlib import Path
SENTENCE_TRANSFORMER_MODELS = (
"sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2",
"intfloat/e5-small",
)
MAX_ATTEMPTS = 10
BASE_DELAY_SECONDS = 10
MAX_DELAY_SECONDS = 120
def _exception_chain(exc: BaseException):
seen: set[int] = set()
current: BaseException | None = exc
while current is not None and id(current) not in seen:
seen.add(id(current))
yield current
current = current.__cause__ or current.__context__
def _is_rate_limited(exc: BaseException) -> bool:
for item in _exception_chain(exc):
response = getattr(item, "response", None)
if getattr(response, "status_code", None) == 429:
return True
text = str(item).lower()
if (
"429" in text
or "too many requests" in text
or "rate limit" in text
or "ratelimit" in text
):
return True
return False
def _is_transient_hub_error(exc: BaseException) -> bool:
if _is_rate_limited(exc):
return True
for item in _exception_chain(exc):
text = str(item).lower()
if (
"couldn't connect to 'https://huggingface.co'" in text
or "localentrynotfounderror" in type(item).__name__.lower()
or "temporarily unavailable" in text
or "service unavailable" in text
or "gateway timeout" in text
):
return True
return False
def _retry_after_seconds(exc: BaseException) -> float | None:
for item in _exception_chain(exc):
response = getattr(item, "response", None)
headers = getattr(response, "headers", None) or {}
raw = headers.get("Retry-After") or headers.get("retry-after")
if raw is None:
continue
try:
return float(raw)
except (TypeError, ValueError):
continue
return None
def _retry(action, label: str) -> None:
for attempt in range(1, MAX_ATTEMPTS + 1):
try:
action()
return
except Exception as exc:
if attempt == MAX_ATTEMPTS or not _is_transient_hub_error(exc):
raise
retry_after = _retry_after_seconds(exc)
exponential = min(
BASE_DELAY_SECONDS * (2 ** (attempt - 1)),
MAX_DELAY_SECONDS,
)
delay = max(retry_after or 0.0, float(exponential))
kind = "rate limited" if _is_rate_limited(exc) else "hub error"
print(
f"[preload] {label}: {kind}, "
f"retry {attempt}/{MAX_ATTEMPTS} in {delay:.0f}s "
f"({type(exc).__name__}: {exc})",
flush=True,
)
time.sleep(delay)
def _configure_hf_token() -> None:
token = (
os.environ.get("HF_TOKEN")
or os.environ.get("HUGGING_FACE_HUB_TOKEN")
or ""
).strip()
if not token:
print(
"[preload] warning: no HF_TOKEN/HUGGING_FACE_HUB_TOKEN set; "
"anonymous Hugging Face downloads are frequently rate-limited in CI",
flush=True,
)
return
os.environ["HF_TOKEN"] = token
os.environ["HUGGING_FACE_HUB_TOKEN"] = token
print("[preload] Hugging Face token configured", flush=True)
try:
from huggingface_hub import login
login(token=token, add_to_git_credential=False)
except Exception as exc:
print(f"[preload] huggingface_hub.login skipped: {exc}", flush=True)
def _preload_sentence_transformers() -> None:
from sentence_transformers import SentenceTransformer
for model in SENTENCE_TRANSFORMER_MODELS:
print(f"[preload] loading sentence-transformers model: {model}", flush=True)
_retry(lambda model=model: SentenceTransformer(model), model)
def _preload_spacy() -> None:
root = Path(__file__).resolve().parents[1]
if str(root) not in sys.path:
sys.path.insert(0, str(root))
from src.constants.spacy_models import SPACY_MODEL_NAMES
for model in sorted(set(SPACY_MODEL_NAMES.values())):
def download(model_name: str = model) -> None:
result = subprocess.run(
[sys.executable, "-m", "spacy", "download", model_name],
check=False,
)
if result.returncode != 0:
raise RuntimeError(
f"spacy download {model_name} failed with exit code {result.returncode}"
)
print(f"[preload] downloading spaCy model: {model}", flush=True)
_retry(download, model)
def main() -> None:
_configure_hf_token()
_preload_sentence_transformers()
_preload_spacy()
if __name__ == "__main__":
main()
|