Text Classification
Transformers
Safetensors
PEFT
branchone
feature-extraction
decision-model
multimodal
custom-code
custom_code
Instructions to use ndemirdemir/BranchOne with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use ndemirdemir/BranchOne with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="ndemirdemir/BranchOne", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("ndemirdemir/BranchOne", trust_remote_code=True, device_map="auto") - PEFT
How to use ndemirdemir/BranchOne with PEFT:
Task type is invalid.
- Notebooks
- Google Colab
- Kaggle
File size: 2,051 Bytes
f78d5a7 | 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 | """Select CUDA 12.8 for extension builds, independently of the runtime driver."""
import os
import re
import subprocess
from pathlib import Path
PACKAGES = [
"cuda-nvcc-12-8=12.8.93-1",
"cuda-cudart-dev-12-8=12.8.90-1",
"cuda-cccl-12-8=12.8.90-1",
]
def compiler_version(home):
compiler = Path(home) / "bin/nvcc"
if not compiler.is_file():
return None
result = subprocess.run(
[str(compiler), "--version"], capture_output=True, text=True, check=True
)
match = re.search(r"release (\d+\.\d+),", result.stdout)
return match.group(1) if match else None
def configure(*, install=False):
candidates = [os.environ.get("CUDA_HOME"), "/usr/local/cuda-12.8"]
selected = next(
(Path(p) for p in candidates if p and compiler_version(p) == "12.8"),
None,
)
if selected is None and install:
if os.geteuid() != 0:
raise RuntimeError(
"Install CUDA 12.8 development tools and set CUDA_HOME before setup"
)
# Colab provides NVIDIA's signed Ubuntu repository. Install compiler and
# development headers only; leave the runtime's GPU driver in place.
subprocess.run(["apt-get", "update", "-qq"], check=True)
subprocess.run(
["apt-get", "install", "-y", "--no-install-recommends", *PACKAGES],
check=True,
env=os.environ | {"DEBIAN_FRONTEND": "noninteractive"},
)
selected = Path("/usr/local/cuda-12.8")
if selected is None or compiler_version(selected) != "12.8":
raise RuntimeError(
"The extension requires a CUDA 12.8 compiler to match Torch cu128; "
"run setup_colab.py or set CUDA_HOME to a CUDA 12.8 toolkit"
)
os.environ["CUDA_HOME"] = str(selected)
os.environ["PATH"] = str(selected / "bin") + os.pathsep + os.environ["PATH"]
version = subprocess.check_output(
[str(selected / "bin/nvcc"), "--version"], text=True
)
return {"cuda_home": str(selected), "nvcc": version}
|