是否支持在华为昇腾910B卡上进行GKD on-policy蒸馏Qwen3-VL模型

[liaomingg]

Describe the bug
What the bug is, and how to reproduce, better with screenshots(描述bug以及复现过程，最好有截图)
报错："/***/python3.11/site-packages/transformers/generation/utils.py", in line 1599, in _validate_model_args
raise ValueError(
ValueError: The fllowing model_kwargs are not used by the model: ['text_position_ids'] (note: types in the generate arguments will also show up in this list)

Your hardware and system info
Write your system info like CUDA version/system/GPU/torch version here(在这里给出硬件信息和系统信息，如CUDA版本，系统，GPU型号和torch版本等)
昇腾910B

Additional context
Add any other context about the problem here(在这里补充其他信息)

[hjh0119]

plz provide env infos (pip list)

[liaomingg]

plz provide env infos (pip list)

absl-py==2.3.1
accelerate==1.11.0
addict==2.4.0
aiofiles==24.1.0
aiohappyeyeballs==2.6.1
aiohttp==3.13.2
aiosignal==1.4.0
aliyun-python-sdk-core==2.16.0
aliyun-python-sdk-kms==2.16.5
annotated-doc==0.0.4
annotated-types==0.7.0
antlr4-python3-runtime==4.9.3
anyio==4.11.0
asc_opc_tool @ file:///home/selfgz95298347/compiler/lib64/asc_opc_tool-0.1.0-py3-none-any.whl
astor==0.8.1
asttokens==3.0.1
attrdict==2.0.1
attrs==25.4.0
auto_tune @ file:///home/selfgz95298347/compiler/lib64/auto_tune-0.1.0-py3-none-any.whl
av==16.0.1
binpacking==1.5.2
blake3==1.0.8
blinker==1.9.0
brotli==1.2.0
cachetools==6.2.2
cbor2==5.7.1
certifi==2025.11.12
cffi==2.0.0
charset-normalizer==3.4.4
click==8.3.1
cloudpickle==3.1.2
cmake==4.2.0
compressed-tensors==0.11.0
contourpy==1.3.3
cpm-kernels==1.0.11
crcmod==1.7
cryptography==46.0.3
cycler==0.12.1
dacite==1.9.2
dataflow @ file:///home/selfgz95298347/compiler/lib64/dataflow-0.0.1-py3-none-any.whl
datasets==3.6.0
decorator==5.2.1
deepspeed==0.18.2
depyf==0.19.0
dill==0.3.8
diskcache==5.6.3
distro==1.9.0
dnspython==2.8.0
einops==0.8.1
email-validator==2.3.0
executing==2.2.1
fastapi==0.122.0
fastapi-cli==0.0.16
fastapi-cloud-cli==0.5.2
fastar==0.7.0
ffmpy==1.0.0
filelock==3.20.0
Flask==3.1.2
fonttools==4.60.1
frozendict==2.4.7
frozenlist==1.8.0
fsspec==2025.3.0
future==1.0.0
gguf==0.17.1
gradio==5.49.1
gradio_client==1.13.3
groovy==0.1.2
grpcio==1.76.0
h11==0.16.0
h2==4.3.0
hccl @ file:///home/selfgz2685628016/hccl/lib64/hccl-0.1.0-py3-none-any.whl
hccl_parser @ file:///home/Ascend/ascend-toolkit/8.3.RC1/toolkit/tools/hccl_parser-0.1-py3-none-any.whl
hf-xet==1.2.1
hjson==3.1.0
hpack==4.1.0
httpcore==1.0.9
httptools==0.7.1
httpx==0.28.1
huggingface-hub==0.36.0
Hypercorn==0.18.0
hyperframe==6.1.0
idna==3.11
importlib_metadata==8.7.0
interegular==0.3.3
ipython==9.7.0
ipython_pygments_lexers==1.1.1
isort==7.0.0
itsdangerous==2.2.0
jedi==0.19.2
jieba==0.42.1
Jinja2==3.1.6
jiter==0.12.0
jmespath==0.10.0
joblib==1.5.2
json_repair==0.53.0
jsonschema==4.25.1
jsonschema-specifications==2025.9.1
kiwisolver==1.4.9
lark==1.2.2
latex2sympy2_extended==1.0.6
lazy_import==0.2.2
llguidance==0.7.30
llm_datadist @ file:///home/selfgz95298347/compiler/lib64/llm_datadist-0.0.1-py3-none-any.whl
llm_datadist_v1 @ file:///home/selfgz95298347/compiler/lib64/llm_datadist_v1-0.0.1-py3-none-any.whl
llvmlite==0.45.1
lm-format-enforcer==0.11.3
lxml==6.0.2
Markdown==3.10
markdown-it-py==4.0.0
MarkupSafe==3.0.3
math-verify==0.5.2
matplotlib==3.10.7
matplotlib-inline==0.2.1
mdurl==0.1.2
mistral_common==1.8.5
modelscope==1.32.0
moxing-framework @ file:///home/moxing_framework-2.2.7%2Bd78e9bef-py2.py3-none-any.whl#sha256=241e284cb5d427c9f62e189823dd5ebeec3c47fc44b3d24c831d33434b0cfdda
mpmath==1.3.0
ms_swift==3.10.0
msgpack==1.1.2
msgspec==0.20.0
msobjdump @ file:///home/Ascend/ascend-toolkit/8.3.RC1/toolkit/tools/msobjdump-0.1.0-py3-none-any.whl
multidict==6.7.0
multiprocess==0.70.16
networkx==3.6
ninja==1.13.0
nltk==3.9.2
numba==0.62.1
numpy==1.26.4
omegaconf==2.3.0
op_compile_tool @ file:///home/selfgz95298347/compiler/lib64/op_compile_tool-0.1.0-py3-none-any.whl
op_gen @ file:///home/Ascend/ascend-toolkit/8.3.RC1/toolkit/tools/op_gen-0.1-py3-none-any.whl
op_test_frame @ file:///home/Ascend/ascend-toolkit/8.3.RC1/toolkit/tools/op_test_frame-0.1-py3-none-any.whl
opc_tool @ file:///home/selfgz95298347/compiler/lib64/opc_tool-0.1.0-py3-none-any.whl
openai==2.8.1
openai-harmony==0.0.8
opencv-python-headless==4.11.0.86
orjson==3.11.4
oss2==2.19.1
outlines_core==0.2.11
packaging==25.0
pandas==2.3.3
parso==0.8.5
partial-json-parser==0.2.1.1.post7
pathlib2==2.3.7.post1
peft==0.15.2
pexpect==4.9.0
pillow==11.3.0
pip2pi==0.8.2
priority==2.0.0
prometheus-fastapi-instrumentator==7.1.0
prometheus_client==0.23.1
prompt_toolkit==3.0.52
propcache==0.4.1
protobuf==3.20.3
psutil==7.1.3
ptyprocess==0.7.0
pure_eval==0.2.3
py-cpuinfo==9.0.0
pyarrow==22.0.0
pybase64==1.4.2
pybind11==3.0.1
pycountry==24.6.1
pycparser==2.23
pycryptodome==3.23.0
pydantic==2.11.10
pydantic-extra-types==2.10.6
pydantic_core==2.33.2
pydub==0.25.1
Pygments==2.19.2
pyparsing==3.2.5
python-dateutil==2.9.0.post0
python-dotenv==1.2.1
python-json-logger==4.0.0
python-multipart==0.0.20
pytz==2025.2
PyYAML==6.0.3
pyzmq==27.1.0
Quart==0.20.0
qwen-vl-utils==0.0.11
referencing==0.37.0
regex==2025.11.3
requests==2.32.5
rich==14.2.0
rich-toolkit==0.16.0
rignore==0.7.6
rouge==1.0.1
rpds-py==0.29.0
ruff==0.14.5
safehttpx==0.1.7
safetensors==0.7.0
schedule_search @ file:///home/selfgz95298347/compiler/lib64/schedule_search-0.1.0-py3-none-any.whl
scipy==1.16.3
semantic-version==2.10.0
sentencepiece==0.2.1
sentry-sdk==2.46.0
setproctitle==1.3.7
setuptools-scm==9.2.2
shellingham==1.5.4
show_kernel_debug_data @ file:///home/Ascend/ascend-toolkit/8.3.RC1/toolkit/tools/show_kernel_debug_data-0.1.0-py3-none-any.whl
simplejson==3.20.2
six==1.17.0
sniffio==1.3.1
sortedcontainers==2.4.0
soundfile==0.13.1
soxr==1.0.0
stack-data==0.6.3
starlette==0.50.0
sympy==1.14.0
te @ file:///home/selfgz95298347/compiler/lib64/te-0.4.0-py3-none-any.whl
tensorboard==2.20.0
tensorboard-data-server==0.7.2
tiktoken==0.12.0
tokenizers==0.22.1
tomlkit==0.13.3
torch==2.7.1+cpu
torch_npu==2.7.1
torchvision==0.22.1
tqdm==4.67.1
traitlets==5.14.3
transformers==4.57.0
transformers-stream-generator==0.0.5
trl==0.24.0
typer==0.20.0
typing-inspection==0.4.2
typing_extensions==4.15.0
tzdata==2025.2
urllib3==2.5.0
uvicorn==0.38.0
uvloop==0.22.1
-e git+https://jiashu.1win.eu.org/https://github.com/vllm-project/vllm.git@b8b302cde434df8c9289a2b465406b47ebab1c2d#egg=vllm
-e git+https://jiashu.1win.eu.org/https://github.com/vllm-project/vllm-ascend.git@a2e4c3fe783dc8c3ce500a6e8794685bb90193e8#egg=vllm_ascend
watchfiles==1.1.1
wcwidth==0.2.14
websockets==15.0.1
Werkzeug==3.1.3
wsproto==1.3.2
xgrammar==0.1.25
xlrd==2.0.2
xlwt==1.3.0
xxhash==3.6.0
yarl==1.22.0
zipp==3.23.0
zstandard==0.25.

please check it out. thanks

[hjh0119]

plz retry with
ms_swift==3.10.3
transformers==4.57.3

[liaomingg]

plz retry with ms_swift==3.10.3 transformers==4.57.3

试过了，还是不行，还是同样的错误。
把训练的模型换成了Qwen3-32B和Qwen3-8B,就没有上述报错了，但是出现了另一个错误。
如下：
RuntimeError: element 0 of tensors does not require grad and does not have a grad_fn
查了下是模型训练的loss.requires_grad为False，这个咋回事？