File size: 3,862 Bytes
06883ee 8356dae 06883ee 8356dae 06883ee 8356dae 06883ee | 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 | """One-shot helper: drop the stale httpx<0.28 pin from cell-pip.
huggingface_hub v1.x has migrated to httpx's `follow_redirects` API, so the
old `httpx<0.28` workaround (for hub 0.27 + httpx 0.28 incompatibility) no
longer applies. The assertion at the end of the cell now fires falsely on
modern Colab images where firebase-admin pulls httpx==0.28.1 back in.
Idempotent: re-running just replaces the cell source again.
"""
import json
from pathlib import Path
NB_PATH = Path(__file__).resolve().parent / "cxrvlm_colab_train.ipynb"
NEW_SRC = r'''!pip uninstall -y -q torchao transformers bitsandbytes peft accelerate
# Let pip pick latest bnb that matches Colab's CUDA 12.8 + triton 3.x
!pip install -q -U bitsandbytes
# Install everything. We DON'T pin httpx anymore — Colab's firebase-admin and
# google-genai hard-pin httpx==0.28.1, so the resolver always wins. Instead
# we monkey-patch httpx 0.28+ below to keep accepting the legacy
# `allow_redirects` kwarg that transformers ≤4.50 still passes.
!pip install -q \
'transformers>=4.46,<4.50' \
'peft>=0.13,<0.15' \
'accelerate>=1.0' \
'huggingface_hub>=0.27,<1.0' \
omegaconf sentencepiece 'protobuf>=3.20' \
nltk rouge-score bert-score sacrebleu
import torch, transformers, bitsandbytes, peft, accelerate, huggingface_hub, httpx
print('torch :', torch.__version__, '| cuda:', torch.cuda.is_available())
print('transformers :', transformers.__version__)
print('bitsandbytes :', bitsandbytes.__version__)
print('peft :', peft.__version__)
print('accelerate :', accelerate.__version__)
print('huggingface_hub:', huggingface_hub.__version__)
print('httpx :', httpx.__version__)
# ── httpx 0.28+ compat shim ───────────────────────────────────────────────
# transformers ≤4.50 calls httpx.Client.head(..., allow_redirects=True) which
# httpx 0.28 removed → "Client.head() got an unexpected keyword argument
# 'allow_redirects'". Translate the kwarg at the call site so the rest of
# the stack keeps working. No-op on httpx <0.28.
#
# The same patch is auto-applied inside the train.py subprocess via
# utils._quiet → utils._httpx_compat. Here we apply it in the NOTEBOOK
# kernel too, so the smoke test cell (which runs in-kernel) benefits.
def _patch_httpx():
if tuple(int(x) for x in httpx.__version__.split('.')[:2]) < (0, 28):
return
if getattr(httpx.Client, '_cxr_vlm_compat_patched', False):
return
def _make(orig):
def patched(self, *args, **kwargs):
if 'allow_redirects' in kwargs:
kwargs['follow_redirects'] = kwargs.pop('allow_redirects')
return orig(self, *args, **kwargs)
return patched
for cls in (httpx.Client, httpx.AsyncClient):
for m in ('request', 'get', 'head', 'post', 'put',
'patch', 'delete', 'options'):
if hasattr(cls, m):
setattr(cls, m, _make(getattr(cls, m)))
httpx.Client._cxr_vlm_compat_patched = True
print(f'httpx {httpx.__version__}: monkey-patched allow_redirects → follow_redirects')
_patch_httpx()
'''
def src_to_lines(s: str):
lines = s.split("\n")
return [ln + "\n" for ln in lines[:-1]] + ([lines[-1]] if lines[-1] else [])
def main():
nb = json.loads(NB_PATH.read_text(encoding="utf-8"))
for c in nb["cells"]:
if c.get("id") == "cell-pip":
c["source"] = src_to_lines(NEW_SRC)
c["outputs"] = []
c["execution_count"] = None
break
else:
raise RuntimeError("cell-pip not found")
NB_PATH.write_text(json.dumps(nb, indent=1, ensure_ascii=False) + "\n", encoding="utf-8")
print(f"Patched cell-pip in {NB_PATH}")
if __name__ == "__main__":
main()
|