b3649d40cc
Notebook pip/uv shim (docker/unsloth_pip_shim.py), all active only under
UNSLOTH_NB_SHIM=1:
- Parse a bare wheel filename (torch-*.whl in the CWD, no ./ or / prefix) so
it is matched against _KEEP instead of passing through as an opaque
positional and reinstalling the baked torch.
- Infer the distribution from an egg-less VCS URL by repo basename
(git+https://github.com/huggingface/transformers.git -> transformers) so
the egg-less form the repo itself recommends cannot clobber the baked stack.
- Refuse remote (URL) -r/-c requirement/constraint files -- top-level and
nested includes -- since their pins cannot be inspected before the real
tool would fetch and install them.
- Strip resolver-wide reinstall/ignore-installed switches
(--force-reinstall, --ignore-installed, -I, uv --reinstall) so they cannot
rebuild already-satisfied baked deps pulled in by a kept target.
- Route uv --reinstall-package through the same _KEEP handling as
-P/--upgrade-package (both attached and separated forms; no dangling flag).
Entrypoint (docker/entrypoint.sh): select_cuda_jit_tools() now scans every
visible GPU's compute_cap instead of only the first, so a datacenter Blackwell
(sm_103/sm_121) behind an H100/B200 still enables the cu13 JIT tools it needs.
Adds regression tests for each case (tests/python/test_unsloth_pip_shim.py,
tests/sh/test_select_cuda_jit_tools.sh).
221 lines
10 KiB
Bash
Executable file
221 lines
10 KiB
Bash
Executable file
#!/usr/bin/env bash
|
|
# Container startup checks for Unsloth.
|
|
#
|
|
# Fails fast with actionable error messages when the host GPU isn't reachable,
|
|
# instead of letting torch crash deep with cryptic CUDA errors. Catches the
|
|
# three failure modes that cover ~95% of "it doesn't work" tickets:
|
|
#
|
|
# 1. nvidia-smi inside the container can't see any GPU
|
|
# - User forgot --gpus all
|
|
# - Host missing nvidia-container-toolkit
|
|
# 2. nvidia-smi works but torch.cuda.is_available() is False
|
|
# - Host driver too old for CUDA 12.8
|
|
# 3. GPU attaches but is older than Ampere (sm < 80)
|
|
# - Unsloth requires sm_80+
|
|
#
|
|
# Bypass for offline tooling / docs / CI:
|
|
# docker run -e UNSLOTH_SKIP_GPU_CHECK=1 ...
|
|
set -euo pipefail
|
|
|
|
# --- CUDA JIT toolchain selection (device-gated) ----------------------------
|
|
# The image bakes CUDA 13 ptxas + NVRTC ONLY so the two Blackwell datacenter
|
|
# arches the cu12.8 tools cannot target -- sm_103 (B300 / GB300) and sm_121
|
|
# (GB10 / DGX Spark) -- can JIT Triton and torch/NVRTC kernels. Both launched
|
|
# AFTER cu12.8, so any host carrying them runs a >= 580 driver, which is exactly
|
|
# what a cu13-produced cubin needs to LOAD.
|
|
#
|
|
# Every OTHER supported arch (Turing..sm_120) works with the bundled cu12.8
|
|
# tools and is allowed on a 570-579 driver (the documented floor). A cu13 cubin
|
|
# CANNOT load on a 570-579 driver even when it targets an old arch like sm_80
|
|
# (CUDA has forward, not backward, driver compatibility across major versions),
|
|
# so routing those hosts' JIT through the cu13 tools would break ordinary
|
|
# training. ptxas/NVRTC are host-side compilers (they never link libcuda), so
|
|
# they RUN under any driver -- it is only their OUTPUT the older driver rejects.
|
|
#
|
|
# Pick per DEVICE at boot (the compute capability is unknown at build time):
|
|
# activate cu13 only for sm_103 / sm_121, and otherwise keep Triton on its
|
|
# bundled cu12.8 ptxas and restore the wheel-bundled cu12.8 NVRTC the build
|
|
# swapped for cu13. Runs before every early-exit below so the selection always
|
|
# applies. Best-effort: a read-only / --user-dropped rootfs that cannot
|
|
# re-point the NVRTC symlink is left unchanged.
|
|
select_cuda_jit_tools() {
|
|
local caps="" cc nvrtc_dir orig need_cu13=0
|
|
if command -v nvidia-smi >/dev/null 2>&1; then
|
|
caps="$( { nvidia-smi --query-gpu=compute_cap --format=csv,noheader 2>/dev/null || true; } )"
|
|
fi
|
|
# Scan EVERY visible GPU, not just the first: a Blackwell datacenter part
|
|
# (sm_103 B300/GB300 or sm_121 GB10/DGX Spark) can sit behind an H100/B200 in
|
|
# the nvidia-smi ordering, so keying off only the first compute_cap would
|
|
# restore cu12.8 and leave that later device unable to JIT. If ANY visible
|
|
# GPU needs cu13, enable it for the whole process -- those parts only ship on
|
|
# >= 580 drivers, so the host tolerates cu13 cubins for every arch present.
|
|
while IFS= read -r cc || [[ -n "${cc}" ]]; do
|
|
cc="$(printf '%s' "${cc}" | tr -d '[:space:]')"
|
|
case "${cc}" in
|
|
10.3|12.1) need_cu13=1 ;;
|
|
esac
|
|
done <<< "${caps}"
|
|
if [[ "${need_cu13}" -eq 1 ]]; then
|
|
# Blackwell datacenter present: the build already points each venv's
|
|
# libnvrtc.so.12 at cu13, so only Triton's ptxas needs redirecting.
|
|
# -z guard leaves an explicit `docker run -e TRITON_PTXAS_PATH` win.
|
|
if [[ -x /usr/local/cuda-13.0/bin/ptxas && -z "${TRITON_PTXAS_PATH:-}" ]]; then
|
|
export TRITON_PTXAS_PATH=/usr/local/cuda-13.0/bin/ptxas
|
|
fi
|
|
else
|
|
# No datacenter Blackwell present (or an undetectable / CPU host): leave
|
|
# TRITON_PTXAS_PATH unset so Triton uses its bundled cu12.8 ptxas, and
|
|
# restore the cu12.8 NVRTC in each venv that saved the original, so a
|
|
# 570-579 driver never sees a cu13 cubin. Covers the base venv and, on
|
|
# the Studio image, the Studio venv.
|
|
for nvrtc_dir in \
|
|
/opt/unsloth-venv/lib/python*/site-packages/nvidia/cuda_nvrtc/lib \
|
|
"${UNSLOTH_STUDIO_HOME:-/opt/unsloth-studio}"/unsloth_studio/lib/python*/site-packages/nvidia/cuda_nvrtc/lib; do
|
|
orig="${nvrtc_dir}/libnvrtc.so.12.cu128.orig"
|
|
[[ -e "${orig}" ]] || continue
|
|
ln -sf libnvrtc.so.12.cu128.orig "${nvrtc_dir}/libnvrtc.so.12" 2>/dev/null || true
|
|
done
|
|
fi
|
|
}
|
|
# Best-effort: never let JIT-tool selection block container startup.
|
|
select_cuda_jit_tools || true
|
|
|
|
# Make the unslothai/notebooks collection available under /workspace before the
|
|
# user command runs (JupyterLab, unsloth-run, or a shell). Best-effort: it is
|
|
# fully gated by UNSLOTH_SKIP_NOTEBOOK_SYNC and never blocks or fails the
|
|
# container (see unsloth_sync_notebooks.sh).
|
|
sync_notebooks() {
|
|
if [[ -x /usr/local/bin/unsloth-sync-notebooks ]]; then
|
|
/usr/local/bin/unsloth-sync-notebooks || true
|
|
fi
|
|
}
|
|
|
|
if [[ "${UNSLOTH_SKIP_GPU_CHECK:-0}" == "1" ]]; then
|
|
sync_notebooks
|
|
exec "$@"
|
|
fi
|
|
|
|
err() { printf "\033[1;31mERROR:\033[0m %s\n" "$*" >&2; }
|
|
warn() { printf "\033[1;33mWARN:\033[0m %s\n" "$*" >&2; }
|
|
|
|
# CPU mode for hosts that cannot pass a GPU into a Linux container at all:
|
|
# Docker Desktop on macOS (no Metal passthrough), Docker Desktop on Windows
|
|
# without WSL2 GPU support, plain CPU Linux boxes, and CI runners. CPU mode
|
|
# covers Jupyter, the GGUF tooling and llama.cpp-backed Studio chat (llama.cpp
|
|
# runs on CPU), and Data Recipes. It does NOT cover training or loading an
|
|
# Unsloth model for chat (FastLanguageModel.from_pretrained runs CUDA probes
|
|
# like torch.cuda.get_device_properties and raises without a GPU). With
|
|
# UNSLOTH_ALLOW_CPU=1 a missing GPU degrades to a warning instead of the hard
|
|
# pre-flight failure; when a GPU IS visible the normal checks below still run so
|
|
# a broken GPU setup is not silently ignored.
|
|
if [[ "${UNSLOTH_ALLOW_CPU:-0}" == "1" ]]; then
|
|
if ! command -v nvidia-smi >/dev/null 2>&1 || ! nvidia-smi -L 2>/dev/null | grep -q '^GPU'; then
|
|
warn "UNSLOTH_ALLOW_CPU=1 and no GPU visible -- continuing on CPU."
|
|
warn "CPU mode covers Jupyter, GGUF tooling and llama.cpp (GGUF) Studio chat."
|
|
warn "Training and loading Unsloth models (FastLanguageModel) still require an NVIDIA GPU."
|
|
sync_notebooks
|
|
exec "$@"
|
|
fi
|
|
fi
|
|
|
|
# --- Check 1: nvidia-smi present and can enumerate at least one GPU ---------
|
|
# nvidia-smi is injected by nvidia-container-toolkit when the container is
|
|
# started with a GPU request; it is NOT baked into the image. A missing
|
|
# binary therefore means "no GPU was attached", the same failure class as
|
|
# an empty -L listing, not a broken image.
|
|
if ! command -v nvidia-smi >/dev/null 2>&1 || ! nvidia-smi -L 2>/dev/null | grep -q '^GPU'; then
|
|
err "No GPU visible inside the container."
|
|
cat >&2 <<'MSG'
|
|
|
|
Likely causes (in order of frequency):
|
|
|
|
1. You started the container without --gpus all.
|
|
Re-launch with:
|
|
docker run --gpus all <other-flags> unsloth/unsloth:latest <cmd>
|
|
Or use the bundled wrapper:
|
|
bash docker/run.sh <cmd>
|
|
|
|
2. Host is missing nvidia-container-toolkit.
|
|
Install: https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/install-guide.html
|
|
Then: sudo systemctl restart docker
|
|
|
|
3. nvidia-container-toolkit is installed but the Docker daemon was not
|
|
restarted after install. Run:
|
|
sudo systemctl restart docker
|
|
|
|
4. You are using Podman / Kubernetes / a managed container service that
|
|
needs a different GPU flag than --gpus all. See the relevant docs:
|
|
podman: --device nvidia.com/gpu=all
|
|
k8s: nvidia.com/gpu resource request + GPU operator
|
|
|
|
5. This host has no NVIDIA GPU at all (Docker Desktop on macOS, Windows
|
|
without WSL2 GPU support, CPU-only Linux). Training and loading Unsloth
|
|
models need a GPU, but Jupyter, GGUF tooling and llama.cpp (GGUF) Studio
|
|
chat work on CPU:
|
|
docker run -e UNSLOTH_ALLOW_CPU=1 ...
|
|
|
|
To bypass this check entirely (e.g. offline tooling), set UNSLOTH_SKIP_GPU_CHECK=1.
|
|
MSG
|
|
exit 1
|
|
fi
|
|
|
|
# --- Check 2: torch can actually use the GPU --------------------------------
|
|
# This catches host-driver-too-old (the GPU enumerates via nvidia-smi but
|
|
# the kernel module rejects CUDA contexts).
|
|
python - >&2 <<'PY' || exit 1
|
|
import sys
|
|
import torch
|
|
if torch.cuda.is_available():
|
|
sys.exit(0)
|
|
print("ERROR: torch.cuda.is_available() is False despite nvidia-smi working.")
|
|
print()
|
|
print("This image bakes in CUDA 12.8, so the host driver MUST be:")
|
|
print(" >= 570.26 (toolkit floor for cu128, applies to every GPU)")
|
|
print()
|
|
print("Two GPUs need an even newer driver because their launch driver was")
|
|
print("released after cu128's:")
|
|
print(" >= 580 B300 / GB300 (sm_103)")
|
|
print(" >= 580 GB10 / DGX Spark (sm_121)")
|
|
print()
|
|
print("Check the host (NOT the container) with: nvidia-smi")
|
|
print("Then upgrade the driver to match.")
|
|
sys.exit(1)
|
|
PY
|
|
|
|
# --- Check 3: compute capability is supported -------------------------------
|
|
python - >&2 <<'PY' || exit 1
|
|
import sys
|
|
import torch
|
|
major, minor = torch.cuda.get_device_capability(0)
|
|
name = torch.cuda.get_device_name(0)
|
|
n = torch.cuda.device_count()
|
|
print(f"Unsloth container: {n} GPU(s). Primary: {name} sm_{major}{minor} bf16={torch.cuda.is_bf16_supported()}")
|
|
|
|
# Image targets every current x86_64 NVIDIA arch from Turing onward, per
|
|
# https://developer.nvidia.com/cuda/gpus.
|
|
SUPPORTED = (
|
|
("sm_75", "Turing", "T4, RTX 20-series, Quadro RTX"),
|
|
("sm_80", "Ampere DC", "A100, A30"),
|
|
("sm_86", "Ampere", "A40, RTX A6000, RTX 30-series"),
|
|
("sm_89", "Ada", "L4, L40, L40S, RTX 40-series"),
|
|
("sm_90", "Hopper", "H100, H200, GH200"),
|
|
("sm_100", "Blackwell DC", "B100, B200, GB200"),
|
|
("sm_103", "Blackwell DC", "B300, GB300"),
|
|
("sm_120", "Blackwell", "RTX 50-series, RTX PRO 6000 Blackwell"),
|
|
("sm_121", "Blackwell", "GB10 (DGX Spark)"),
|
|
)
|
|
if major < 7 or (major == 7 and minor < 5):
|
|
print()
|
|
print(f"ERROR: Unsloth image requires Turing or newer (sm_75+). Got {name} sm_{major}{minor}.")
|
|
print()
|
|
print("Supported architectures in this image:")
|
|
for arch, fam, ex in SUPPORTED:
|
|
print(f" {arch:7s} {fam:13s} ({ex})")
|
|
sys.exit(1)
|
|
if major < 8:
|
|
print(f"NOTE: {name} is Turing (sm_{major}{minor}) -- bfloat16 is not supported.")
|
|
print(" Unsloth will fall back to fp16. Training works but is slightly slower.")
|
|
PY
|
|
|
|
sync_notebooks
|
|
exec "$@"
|