Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4.1, MiniMax-H3, Gemma 4, FLUX and more.
Python agentaichatgptdeepseekfine-tuninggemmaimage-generationllamallmllmsopenaipythonqwenreinforcement-learningself-hostedstable-diffusiontext-to-speechttsuiunsloth
# SPDX-License-Identifier: AGPL-3.0-only
# Copyright 2026-present the Unsloth AI Inc. team. All rights reserved.
"""On no-bf16 GPUs, full finetunes use float16 autocast, or pure float32 for FORCE_FLOAT32 models."""
from __future__ import annotations
import os
import sys
import types
from pathlib import Path
import pytest
torch = pytest.importorskip("torch")
RL_PY = Path(__file__).resolve().parents[2] / "unsloth" / "models" / "rl.py"
def _extract_mixed_precision_code() -> str:
lines = RL_PY.read_text(encoding = "utf-8").split("\n")
try:
start = next(i for i, l in enumerate(lines) if "mixed_precision = (" in l)
except StopIteration:
pytest.skip("mixed_precision template not found in rl.py")
body, k = [], start + 1
while lines[k].strip() != ")":
body.append(lines[k])
k += 1
return eval("(\n" + "\n".join(body) + "\n)") # only string literals + comments
CODE = _extract_mixed_precision_code()
def _restore(mapping, saved):
"""Restore a dict-like to its saved snapshot: pop keys that were absent."""
for k, v in saved.items():
if v is None:
mapping.pop(k, None)
else:
mapping[k] = v
def _decide(dtype, *, bf16_supported, force_float32, full_finetuning, mixed_precision, fp16, bf16):
"""Runs the real template block; stubs are restored on exit so no decision leaks into later tests."""
uzu = types.ModuleType("unsloth_zoo.utils")
uzu._get_dtype = lambda x: x
uzd = types.ModuleType("unsloth_zoo.device_type")
uzd.device_is_bf16_supported = lambda: bf16_supported
env_keys = (
"UNSLOTH_FORCE_FLOAT32",
"UNSLOTH_ENABLE_FULL_FINETUNING",
"UNSLOTH_MIXED_PRECISION",
"ACCELERATE_MIXED_PRECISION",
)
mod_keys = ("unsloth_zoo", "unsloth_zoo.utils", "unsloth_zoo.device_type")
saved_env = {k: os.environ.get(k) for k in env_keys}
saved_mods = {k: sys.modules.get(k) for k in mod_keys}
orig_bf16 = torch.cuda.is_bf16_supported
try:
sys.modules.setdefault("unsloth_zoo", types.ModuleType("unsloth_zoo"))
sys.modules["unsloth_zoo.utils"] = uzu
sys.modules["unsloth_zoo.device_type"] = uzd
for k in env_keys:
os.environ.pop(k, None)
os.environ["UNSLOTH_FORCE_FLOAT32"] = "1" if force_float32 else "0"
os.environ["UNSLOTH_ENABLE_FULL_FINETUNING"] = "1" if full_finetuning else "0"
os.environ["UNSLOTH_MIXED_PRECISION"] = mixed_precision
torch.cuda.is_bf16_supported = lambda *a, **k: bf16_supported
args = types.SimpleNamespace(fp16 = fp16, bf16 = bf16, mixed_precision = None)
emb = types.SimpleNamespace(weight = types.SimpleNamespace(dtype = dtype))
model = types.SimpleNamespace(
config = types.SimpleNamespace(dtype = dtype, torch_dtype = dtype),
get_input_embeddings = lambda: emb,
)
raised = None
try:
# One namespace, as in the trainer's function scope: with separate locals, a
# generator or comprehension in the template cannot see `args` (NameError).
exec(CODE, {"torch": torch, "os": os, "args": args, "model": model})
except TypeError:
raised = "TypeError"
return args.fp16, args.bf16, os.environ.get("ACCELERATE_MIXED_PRECISION"), raised
finally:
torch.cuda.is_bf16_supported = orig_bf16
_restore(os.environ, saved_env)
_restore(sys.modules, saved_mods)
def test_v100_normal_fullft_fp16_explicit():
fp16, bf16, amp, raised = _decide(
torch.float32,
bf16_supported = False,
force_float32 = False,
full_finetuning = True,
mixed_precision = "float32",
fp16 = True,
bf16 = False,
)
assert raised is None
assert (fp16, bf16) == (True, False)
def test_v100_normal_fullft_precision_unset():
fp16, bf16, amp, raised = _decide(
torch.float32,
bf16_supported = False,
force_float32 = False,
full_finetuning = True,
mixed_precision = "float32",
fp16 = False,
bf16 = False,
)
assert raised is None
assert (fp16, bf16) == (True, False)
assert amp == "fp16"
def test_force_float32_model_fullft_is_pure_float32():
fp16, bf16, amp, raised = _decide(
torch.float32,
bf16_supported = False,
force_float32 = True,
full_finetuning = True,
mixed_precision = "float32",
fp16 = True,
bf16 = False,
)
assert raised is None
assert (fp16, bf16) == (False, False)
assert amp in (None, "no")
def test_no_bf16_on_volta_in_auto_branch():
fp16, bf16, amp, raised = _decide(
torch.bfloat16,
bf16_supported = False,
force_float32 = False,
full_finetuning = False,
mixed_precision = "float32",
fp16 = False,
bf16 = False,
)
assert bf16 is False
def test_bf16_gpu_unchanged_auto_branch():
# On bf16 hardware a float32 model with unset precision must still pick bf16 autocast.
fp16, bf16, amp, raised = _decide(
torch.float32,
bf16_supported = True,
force_float32 = False,
full_finetuning = True,
mixed_precision = "float32",
fp16 = False,
bf16 = False,
)
assert raised is None
assert (fp16, bf16) == (False, True)
def test_genuine_bf16_model_with_fp16_still_raises():
_, _, _, raised = _decide(
torch.bfloat16,
bf16_supported = True,
force_float32 = False,
full_finetuning = False,
mixed_precision = "float32",
fp16 = True,
bf16 = False,
)
assert raised == "TypeError"
def test_explicit_bf16_exports_bf16_env():
# Issue #4891: downstream readers default ACCELERATE_MIXED_PRECISION to fp16 when unset,
# which crashes GRPO on a bf16 model, so the env var must still be exported.
fp16, bf16, amp, raised = _decide(
torch.bfloat16,
bf16_supported = True,
force_float32 = False,
full_finetuning = False,
mixed_precision = "float32",
fp16 = False,
bf16 = True,
)
assert raised is None
assert (fp16, bf16) == (False, True)
assert amp == "bf16"
def test_explicit_fp16_exports_fp16_env():
fp16, bf16, amp, raised = _decide(
torch.float16,
bf16_supported = False,
force_float32 = False,
full_finetuning = False,
mixed_precision = "float32",
fp16 = True,
bf16 = False,
)
assert raised is None
assert (fp16, bf16) == (True, False)
assert amp == "fp16"
def test_force_float32_beats_explicit_bf16():
fp16, bf16, amp, raised = _decide(
torch.float32,
bf16_supported = True,
force_float32 = True,
full_finetuning = False,
mixed_precision = "float32",
fp16 = False,
bf16 = True,
)
assert raised is None
assert (fp16, bf16) == (False, False)
assert amp == "no"
def test_unsloth_mixed_precision_bfloat16_beats_explicit_bf16():
fp16, bf16, amp, raised = _decide(
torch.bfloat16,
bf16_supported = True,
force_float32 = False,
full_finetuning = True,
mixed_precision = "bfloat16",
fp16 = False,
bf16 = True,
)
assert raised is None
assert (fp16, bf16) == (False, False)
assert amp == "no"