Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

为什么我在 kaggle.com 上训练的 LoRA 模型效果比较不错，模型下载到本地进行推理效果却很差？ #217

Closed
jianghushinian opened this issue Jun 5, 2023 · 2 comments
Closed

为什么我在 kaggle.com 上训练的 LoRA 模型效果比较不错，模型下载到本地进行推理效果却很差？ #217

jianghushinian opened this issue Jun 5, 2023 · 2 comments

Comments

@jianghushinian
Copy link

  1. 如下是 kaggle 机器的配置。
image

两张 T4 显卡。

  1. !pip freeze
absl-py==1.4.0
accelerate==0.15.0
access==1.1.9
affine==2.4.0
aiobotocore==2.5.0
aiofiles==22.1.0
aiohttp @ file:///home/conda/feedstock_root/build_artifacts/aiohttp_1676292661248/work
aiohttp-cors==0.7.0
aioitertools==0.11.0
aiorwlock==1.3.0
aiosignal @ file:///home/conda/feedstock_root/build_artifacts/aiosignal_1667935791922/work
aiosqlite==0.19.0
albumentations==1.3.0
alembic==1.11.1
altair==5.0.0
annoy==1.17.2
ansiwrap==0.8.4
anyio @ file:///home/conda/feedstock_root/build_artifacts/anyio_1666191106763/work/dist
apache-beam==2.46.0
aplus==0.11.0
appdirs==1.4.4
argon2-cffi @ file:///home/conda/feedstock_root/build_artifacts/argon2-cffi_1640817743617/work
argon2-cffi-bindings @ file:///home/conda/feedstock_root/build_artifacts/argon2-cffi-bindings_1666850768662/work
array-record==0.2.0
arrow==1.2.3
arviz==0.12.1
astroid==2.15.5
astropy==5.3
asttokens @ file:///home/conda/feedstock_root/build_artifacts/asttokens_1670263926556/work
astunparse==1.6.3
async-timeout @ file:///home/conda/feedstock_root/build_artifacts/async-timeout_1640026696943/work
atpublic==3.1.1
attrs @ file:///home/conda/feedstock_root/build_artifacts/attrs_1683424013410/work
audioread==3.0.0
autopep8==2.0.2
Babel==2.12.1
backcall @ file:///home/conda/feedstock_root/build_artifacts/backcall_1592338393461/work
backoff==2.2.1
backports.functools-lru-cache @ file:///home/conda/feedstock_root/build_artifacts/backports.functools_lru_cache_1618230623929/work
bayesian-optimization==1.4.3
bayespy==0.5.25
beatrix-jupyterlab @ file:///home/kbuilder/miniconda3/conda-bld/dlenv-tf-2-12-gpu_1683597552195/work/packages/beatrix_jupyterlab-2023.58.190319.tar.gz#sha256=5d0d9c77a86fcdd097390e863c1c12fc410cc84ab98b6ee16a43b6a84735e57e
beautifulsoup4 @ file:///home/conda/feedstock_root/build_artifacts/beautifulsoup4_1680888073205/work
bidict==0.22.1
biopython==1.81
bitsandbytes==0.37.0
blake3==0.2.1
bleach @ file:///home/conda/feedstock_root/build_artifacts/bleach_1674535352125/work
blessed==1.20.0
blinker==1.6.2
blis @ file:///home/conda/feedstock_root/build_artifacts/cython-blis_1668499088869/work
blosc2==2.0.0
bokeh @ file:///home/conda/feedstock_root/build_artifacts/bokeh_1660586590972/work
boltons @ file:///home/conda/feedstock_root/build_artifacts/boltons_1677499911949/work
Boruta==0.3
boto3==1.26.100
botocore==1.29.76
-e git+https://github.com/SohierDane/BigQuery_Helper@8615a7f6c1663e7f2d48aa2b32c2dbcb600a440f#egg=bq_helper
bqplot==0.12.39
branca==0.6.0
brewer2mpl==1.4.1
brotlipy==0.7.0
cached-property==1.5.2
cachetools==4.2.4
Cartopy @ file:///home/conda/feedstock_root/build_artifacts/cartopy_1679097974681/work
catalogue @ file:///home/conda/feedstock_root/build_artifacts/catalogue_1666891892909/work
catalyst==22.4
catboost==1.2
category-encoders==2.6.1
certifi==2023.5.7
cesium==0.12.1
cffi @ file:///home/conda/feedstock_root/build_artifacts/cffi_1671179353105/work
cftime==1.6.2
charset-normalizer @ file:///home/conda/feedstock_root/build_artifacts/charset-normalizer_1661170624537/work
chex==0.1.7
cleverhans==4.0.0
click @ file:///home/conda/feedstock_root/build_artifacts/click_1666798198223/work
click-plugins==1.1.1
cligj==0.7.2
cloud-tpu-client==0.10
cloud-tpu-profiler==2.4.0
cloudpickle @ file:///home/conda/feedstock_root/build_artifacts/cloudpickle_1674202310934/work
cmaes==0.9.1
cmake==3.26.3
cmdstanpy==1.1.0
cmudict==1.0.13
colorama @ file:///home/conda/feedstock_root/build_artifacts/colorama_1666700638685/work
colorcet==3.0.1
colorful==0.5.5
colorlog==6.7.0
colorlover==0.3.0
comm @ file:///home/conda/feedstock_root/build_artifacts/comm_1679481329611/work
commonmark==0.9.1
conda==23.3.1
conda-content-trust @ file:///tmp/build/80754af9/conda-content-trust_1617045594566/work
conda-package-handling @ file:///home/conda/feedstock_root/build_artifacts/conda-package-handling_1669907009957/work
conda_package_streaming @ file:///home/conda/feedstock_root/build_artifacts/conda-package-streaming_1669733752472/work
confection @ file:///home/conda/feedstock_root/build_artifacts/confection_1673621475775/work
contextily==1.3.0
contourpy @ file:///home/conda/feedstock_root/build_artifacts/contourpy_1673633665736/work
convertdate==2.4.0
crcmod==1.7
cryptography @ file:///home/conda/feedstock_root/build_artifacts/cryptography-split_1681508581703/work
cubinlinker @ file:///home/conda/feedstock_root/build_artifacts/cubinlinker_1669932549674/work
cuda-python @ file:///opt/conda/conda-bld/cuda-python_1669949846864/work
cudf==23.4.1
cufflinks==0.17.3
cuml==23.4.1
cupy @ file:///home/conda/feedstock_root/build_artifacts/cupy_1677786697638/work
CVXcanon==0.1.2
cycler @ file:///home/conda/feedstock_root/build_artifacts/cycler_1635519461629/work
cymem @ file:///home/conda/feedstock_root/build_artifacts/cymem_1666909672496/work
cysignals==1.11.2
Cython==0.29.34
cytoolz @ file:///home/conda/feedstock_root/build_artifacts/cytoolz_1666829662037/work
daal==2023.1.1
daal4py==2023.1.1
dask==2023.5.0
dask-cuda @ file:///opt/conda/conda-bld/work
dask-cudf==23.4.1
dataclasses @ file:///home/conda/feedstock_root/build_artifacts/dataclasses_1628958434797/work
dataclasses-json==0.5.7
datasets==2.8.0
datashader==0.14.4
datashape==0.5.2
datatile==1.0.3
db-dtypes==1.1.1
deap==1.3.3
debugpy @ file:///home/conda/feedstock_root/build_artifacts/debugpy_1680755465990/work
decorator @ file:///home/conda/feedstock_root/build_artifacts/decorator_1641555617451/work
deepspeed==0.8.3
defusedxml @ file:///home/conda/feedstock_root/build_artifacts/defusedxml_1615232257335/work
Delorean==1.0.0
deprecat==2.1.1
Deprecated==1.2.13
deprecation==2.1.0
descartes==1.1.0
dill==0.3.6
dipy==1.7.0
distlib==0.3.6
distributed @ file:///home/conda/feedstock_root/build_artifacts/distributed_1680715567006/work
dm-tree==0.1.8
docker==6.1.1
docker-pycreds==0.4.0
docopt==0.6.2
docstring-parser==0.15
docstring-to-markdown==0.12
docutils==0.20.1
earthengine-api==0.1.354
easydict==1.10
easyocr==1.6.2
ecos==2.0.12
einops==0.6.1
eli5==0.13.0
emoji==2.2.0
en-core-web-lg @ https://github.com/explosion/spacy-models/releases/download/en_core_web_lg-3.5.0/en_core_web_lg-3.5.0-py3-none-any.whl#sha256=c8ac64840c1eb3e3ca7bd38bd1e1c48fb0faeb2449d54d01d5ce629af4595775
en-core-web-sm @ https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-3.5.0/en_core_web_sm-3.5.0-py3-none-any.whl#sha256=0964370218b7e1672a30ac50d72cdc6b16f7c867496f1d60925691188f4d2510
entrypoints @ file:///home/conda/feedstock_root/build_artifacts/entrypoints_1643888246732/work
ephem==4.1.4
esda==2.4.3
essentia==2.1b6.dev1034
et-xmlfile==1.1.0
etils==1.2.0
evaluate==0.4.0
executing @ file:///home/conda/feedstock_root/build_artifacts/executing_1667317341051/work
explainable-ai-sdk==1.3.3
fairscale==0.4.13
fastai==2.7.12
fastapi==0.95.1
fastavro @ file:///home/conda/feedstock_root/build_artifacts/fastavro_1683226640635/work
fastcore==1.5.29
fastdownload==0.0.7
fasteners==0.18
fastjsonschema @ file:///home/conda/feedstock_root/build_artifacts/python-fastjsonschema_1677336799617/work/dist
fastprogress==1.0.3
fastrlock==0.8
fasttext==0.9.2
fbpca==1.0
feather-format==0.4.1
featuretools==1.26.0
ffmpy==0.3.0
filelock==3.12.0
Fiona==1.8.22
fire==0.5.0
fitter==1.5.2
flake8==6.0.0
flashtext==2.7
Flask==2.3.2
flatbuffers==23.3.3
flax==0.6.10
flit_core @ file:///home/conda/feedstock_root/build_artifacts/flit-core_1667734568827/work/source/flit_core
folium==0.14.0
fonttools==4.39.3
fqdn==1.5.1
frozendict==2.3.8
frozenlist @ file:///home/conda/feedstock_root/build_artifacts/frozenlist_1667935435842/work
fsspec @ file:///home/conda/feedstock_root/build_artifacts/fsspec_1683494881189/work
funcy==2.0
fury==0.9.0
future @ file:///home/conda/feedstock_root/build_artifacts/future_1673596611778/work
fuzzywuzzy==0.18.0
gast==0.4.0
gatspy==0.3
gcsfs==2023.5.0
gensim==4.3.1
geographiclib==2.0
Geohash==1.0
geojson==3.0.1
geopandas==0.13.0
geoplot==0.5.1
geopy==2.3.0
geoviews==1.9.6
ggplot @ https://github.com/hbasria/ggpy/archive/0.11.5.zip#sha256=7df947ba3fd86d3757686afec264785ad8df38dc50ffb2d2d31064fb355f69b1
giddy==2.3.4
gitdb==4.0.10
GitPython==3.1.31
google-api-core==1.33.2
google-api-python-client==2.86.0
google-apitools==0.5.31
google-auth==2.17.3
google-auth-httplib2==0.1.0
google-auth-oauthlib==0.4.6
google-cloud-aiplatform==0.6.0a1
google-cloud-artifact-registry==1.8.1
google-cloud-automl==1.0.1
google-cloud-bigquery==2.34.4
google-cloud-bigtable==1.7.3
google-cloud-core==2.3.2
google-cloud-datastore==2.15.2
google-cloud-dlp==3.12.1
google-cloud-language==2.6.1
google-cloud-monitoring==2.14.2
google-cloud-pubsub==2.16.1
google-cloud-pubsublite==1.8.1
google-cloud-recommendations-ai==0.7.1
google-cloud-resource-manager==1.10.0
google-cloud-spanner==3.33.0
google-cloud-storage==1.44.0
google-cloud-translate==3.8.4
google-cloud-videointelligence==2.8.3
google-cloud-vision==2.8.0
google-crc32c==1.5.0
google-pasta==0.2.0
google-resumable-media==2.5.0
googleapis-common-protos==1.57.1
gplearn==0.4.2
gpustat==1.0.0
gpxpy==1.5.0
gradio==3.20.0
graphviz==0.20.1
greenlet==2.0.2
grpc-google-iam-v1==0.12.6
grpcio @ file:///home/conda/feedstock_root/build_artifacts/grpc-split_1677499296072/work
grpcio-status @ file:///home/conda/feedstock_root/build_artifacts/grpcio-status_1662108958711/work
gviz-api==1.10.0
gym==0.26.2
gym-notices==0.0.8
Gymnasium==0.26.3
gymnasium-notices==0.0.1
h11==0.14.0
h2o==3.40.0.4
h5py==3.8.0
haversine==2.8.0
hdfs==2.7.0
hep-ml==0.7.2
hijri-converter==2.3.1
hjson==3.1.0
hmmlearn==0.3.0
holidays==0.24
holoviews==1.16.0
hpsklearn==0.1.0
html5lib==1.1
htmlmin==0.1.12
httpcore==0.17.2
httplib2==0.21.0
httptools==0.5.0
httpx==0.24.1
huggingface-hub==0.13.3
humanize==4.6.0
hunspell==0.5.5
husl==4.0.3
hydra-slayer==0.4.1
hyperopt==0.2.7
hypertools==0.8.0
ibis-framework==5.1.0
idna @ file:///home/conda/feedstock_root/build_artifacts/idna_1663625384323/work
igraph==0.10.4
imagecodecs==2023.3.16
ImageHash==4.3.1
imageio==2.28.1
imbalanced-learn==0.10.1
imgaug==0.4.0
implicit @ file:///home/conda/feedstock_root/build_artifacts/implicit_1643471607379/work
importlib-metadata==5.2.0
importlib-resources @ file:///home/conda/feedstock_root/build_artifacts/importlib_resources_1676919000169/work
inequality==1.0.0
ipydatawidgets==4.3.3
ipykernel @ file:///home/conda/feedstock_root/build_artifacts/ipykernel_1683553336538/work
ipyleaflet==0.17.2
ipympl==0.7.0
ipython @ file:///home/conda/feedstock_root/build_artifacts/ipython_1683225895562/work
ipython-genutils==0.2.0
ipython-sql==0.5.0
ipyvolume==0.6.1
ipyvue==1.9.0
ipyvuetify==1.8.10
ipywebrtc==0.6.0
ipywidgets==7.7.1
isoduration==20.11.0
isort==5.12.0
isoweek==1.3.3
itsdangerous==2.1.2
Janome==0.4.2
jaraco.classes==3.2.3
jax==0.4.10
jaxlib==0.4.7+cuda11.cudnn86
jedi @ file:///home/conda/feedstock_root/build_artifacts/jedi_1669134318875/work
jeepney==0.8.0
jieba==0.42.1
Jinja2 @ file:///home/conda/feedstock_root/build_artifacts/jinja2_1654302431367/work
jmespath==1.0.1
joblib @ file:///home/conda/feedstock_root/build_artifacts/joblib_1663332044897/work
json5==0.9.11
jsonpatch @ file:///home/conda/feedstock_root/build_artifacts/jsonpatch_1632759296524/work
jsonpointer==2.0
jsonschema @ file:///home/conda/feedstock_root/build_artifacts/jsonschema-meta_1669810440410/work
jupyter-console==6.6.3
jupyter-events @ file:///home/conda/feedstock_root/build_artifacts/jupyter_events_1673559782596/work
jupyter-http-over-ws==0.0.8
jupyter-lsp==1.5.1
jupyter-server-mathjax==0.2.6
jupyter-ydoc==0.2.4
jupyter_client==7.4.9
jupyter_core @ file:///home/conda/feedstock_root/build_artifacts/jupyter_core_1678994169527/work
jupyter_server @ file:///home/conda/feedstock_root/build_artifacts/jupyter_server_1679073341944/work
jupyter_server_fileid==0.9.0
jupyter_server_proxy==4.0.0
jupyter_server_terminals @ file:///home/conda/feedstock_root/build_artifacts/jupyter_server_terminals_1673491454549/work
jupyter_server_ydoc==0.8.0
jupyterlab==3.6.3
jupyterlab-git==0.41.0
jupyterlab-lsp==4.1.0
jupyterlab-pygments @ file:///home/conda/feedstock_root/build_artifacts/jupyterlab_pygments_1649936611996/work
jupyterlab-widgets==3.0.7
jupyterlab_server==2.22.1
jupytext==1.14.5
kaggle==1.5.13
kaggle-environments==1.12.0
keras==2.12.0
keras-tuner==1.3.5
keyring==23.13.1
keyrings.google-artifactregistry-auth==1.1.2
kfp==1.8.21
kfp-pipeline-spec==0.1.16
kfp-server-api==1.8.5
kiwisolver @ file:///home/conda/feedstock_root/build_artifacts/kiwisolver_1666805701884/work
kmapper==2.0.1
kmodes==0.12.2
korean-lunar-calendar==0.3.1
kornia==0.6.12
kt-legacy==1.0.5
kubernetes==25.3.0
langcodes @ file:///home/conda/feedstock_root/build_artifacts/langcodes_1636741340529/work
langid==1.1.6
lazy-object-proxy==1.9.0
lazy_loader==0.2
learntools @ git+https://github.com/Kaggle/learntools@69bc6daec79619690e758841dc2df35708d226c8
leven==1.0.4
Levenshtein==0.21.0
libclang==16.0.0
libmambapy @ file:///home/conda/feedstock_root/build_artifacts/mamba-split_1680791035685/work/libmambapy
libpysal==4.7.0
librosa==0.10.0.post2
lightgbm @ file:///tmp/lightgbm/lightgbm-3.3.2-py3-none-any.whl#sha256=54af6814e8e82596cb886f2025b8b020c2ead19b7b956525285565b101b8cd51
lightning-utilities==0.8.0
lime==0.2.0.1
line-profiler==4.0.3
linkify-it-py==2.0.2
lit==16.0.5.post0
llvmlite==0.39.1
lml==0.1.0
locket @ file:///home/conda/feedstock_root/build_artifacts/locket_1650660393415/work
loralib==0.1.1
LunarCalendar==0.0.9
lxml==4.9.2
lz4 @ file:///home/conda/feedstock_root/build_artifacts/lz4_1675806673645/work
Mako==1.2.4
mamba @ file:///home/conda/feedstock_root/build_artifacts/mamba-split_1680791035685/work/mamba
mapclassify==2.5.0
marisa-trie==0.8.0
Markdown==3.4.3
markdown-it-py==2.2.0
markovify==0.9.4
MarkupSafe @ file:///home/conda/feedstock_root/build_artifacts/markupsafe_1674135787083/work
marshmallow==3.19.0
marshmallow-enum==1.5.1
matplotlib==3.6.3
matplotlib-inline @ file:///home/conda/feedstock_root/build_artifacts/matplotlib-inline_1660814786464/work
matplotlib-venn==0.11.9
mccabe==0.7.0
mdit-py-plugins==0.3.3
mdurl==0.1.2
memory-profiler==0.61.0
mercantile==1.2.1
mgwr==2.1.2
missingno==0.5.2
mistune==0.8.4
mizani==0.9.1
ml-dtypes==0.1.0
mlcrate==0.2.0
mlens==0.2.3
mlxtend==0.22.0
mmh3==4.0.0
mne==1.4.0
mnist==0.2.2
mock==5.0.2
momepy==0.6.0
more-itertools==9.1.0
mpld3==0.5.9
mpmath==1.3.0
msgpack @ file:///home/conda/feedstock_root/build_artifacts/msgpack-python_1678312712169/work
msgpack-numpy==0.4.8
multidict @ file:///home/conda/feedstock_root/build_artifacts/multidict_1672339403932/work
multimethod==1.9.1
multipledispatch==0.6.0
multiprocess==0.70.14
munch==3.0.0
munkres==1.1.4
murmurhash @ file:///home/conda/feedstock_root/build_artifacts/murmurhash_1666946151787/work
mypy-extensions==1.0.0
nb-conda @ file:///home/conda/feedstock_root/build_artifacts/nb_conda_1654442778977/work
nb-conda-kernels @ file:///home/conda/feedstock_root/build_artifacts/nb_conda_kernels_1667060632461/work
nbclassic @ file:///home/conda/feedstock_root/build_artifacts/nbclassic_1683202085119/work
nbclient==0.5.13
nbconvert==6.4.5
nbdime==3.2.0
nbformat @ file:///home/conda/feedstock_root/build_artifacts/nbformat_1679336765223/work
nest-asyncio @ file:///home/conda/feedstock_root/build_artifacts/nest-asyncio_1664684991461/work
netCDF4==1.6.3
networkx==3.1
nibabel==5.1.0
nilearn==0.10.1
ninja==1.11.1
nltk==3.2.4
nose==1.3.7
notebook @ file:///home/conda/feedstock_root/build_artifacts/notebook_1680870634737/work
notebook-executor @ file:///home/kbuilder/miniconda3/conda-bld/dlenv-tf-2-12-gpu_1683597552195/work/packages/notebook_executor
notebook_shim @ file:///home/conda/feedstock_root/build_artifacts/notebook-shim_1682360583588/work
numba @ file:///home/conda/feedstock_root/build_artifacts/numba_1680825379968/work
numexpr==2.8.4
numpy==1.23.5
nvidia-cublas-cu11==11.10.3.66
nvidia-cuda-nvrtc-cu11==11.7.99
nvidia-cuda-runtime-cu11==11.7.99
nvidia-cudnn-cu11==8.5.0.96
nvidia-ml-py==11.495.46
nvitop==1.0.0
nvtx @ file:///home/conda/feedstock_root/build_artifacts/nvtx_1682005264204/work
oauth2client==4.1.3
oauthlib==3.2.2
objsize==0.6.1
odfpy==1.4.1
olefile==0.46
onnx==1.14.0
opencensus==0.11.2
opencensus-context==0.1.3
opencv-contrib-python==4.5.4.60
opencv-python==4.5.4.60
opencv-python-headless==4.5.4.60
openpyxl==3.1.2
openslide-python==1.2.0
opentelemetry-api==1.17.0
opentelemetry-exporter-otlp==1.17.0
opentelemetry-exporter-otlp-proto-grpc==1.17.0
opentelemetry-exporter-otlp-proto-http==1.17.0
opentelemetry-proto==1.17.0
opentelemetry-sdk==1.17.0
opentelemetry-semantic-conventions==0.38b0
opt-einsum==3.3.0
optax==0.1.5
optuna==3.1.1
orbax-checkpoint==0.2.3
orderedmultidict==1.0.1
orjson==3.8.12
ortools==9.4.1874
osmnx==1.1.1
overrides==6.5.0
packaging==21.3
pandas==1.5.3
pandas-datareader==0.10.0
pandas-profiling==3.6.6
pandas-summary==0.2.0
pandasql==0.7.3
pandocfilters @ file:///home/conda/feedstock_root/build_artifacts/pandocfilters_1631603243851/work
panel==0.14.4
papermill==2.4.0
param==1.13.0
parso @ file:///home/conda/feedstock_root/build_artifacts/parso_1638334955874/work
parsy==2.1
partd @ file:///home/conda/feedstock_root/build_artifacts/partd_1681246756246/work
path==16.6.0
path.py==12.5.0
pathos==0.3.0
pathtools==0.1.2
pathy @ file:///croot/pathy_1674585914374/work
patsy @ file:///home/conda/feedstock_root/build_artifacts/patsy_1665356157073/work
pdf2image==1.16.3
peft @ git+https://github.com/huggingface/peft.git@13e53fc7ee5d89d59b16523051006dddf0fb7a49
pexpect @ file:///home/conda/feedstock_root/build_artifacts/pexpect_1667297516076/work
phik==0.12.3
pickleshare @ file:///home/conda/feedstock_root/build_artifacts/pickleshare_1602536217715/work
Pillow @ file:///home/conda/feedstock_root/build_artifacts/pillow_1680694272008/work
pkgutil_resolve_name @ file:///home/conda/feedstock_root/build_artifacts/pkgutil-resolve-name_1633981968097/work
platformdirs @ file:///home/conda/feedstock_root/build_artifacts/platformdirs_1682644429438/work
plotly==5.14.1
plotly-express==0.4.1
plotnine==0.10.1
pluggy @ file:///home/conda/feedstock_root/build_artifacts/pluggy_1667232663820/work
pointpats==2.3.0
polars==0.17.15
polyglot==16.7.4
pooch==1.6.0
pox==0.3.2
ppca==0.0.4
ppft==1.7.6.6
preprocessing==0.1.13
preshed @ file:///home/conda/feedstock_root/build_artifacts/preshed_1666991224827/work
prettytable==3.7.0
progressbar2==4.2.0
prometheus-client @ file:///home/conda/feedstock_root/build_artifacts/prometheus_client_1674535637125/work
promise==2.3
prompt-toolkit @ file:///home/conda/feedstock_root/build_artifacts/prompt-toolkit_1677600924538/work
pronouncing==0.2.0
prophet==1.1.1
proto-plus @ file:///home/conda/feedstock_root/build_artifacts/proto-plus_1673334163294/work
protobuf==3.20.3
psutil==5.9.3
ptxcompiler @ file:///home/conda/feedstock_root/build_artifacts/ptxcompiler_1684528370140/work
ptyprocess @ file:///home/conda/feedstock_root/build_artifacts/ptyprocess_1609419310487/work/dist/ptyprocess-0.7.0-py2.py3-none-any.whl
pudb==2022.1.3
PuLP==2.7.0
pure-eval @ file:///home/conda/feedstock_root/build_artifacts/pure_eval_1642875951954/work
py-cpuinfo==9.0.0
py-lz4framed==0.14.0
py-spy==0.3.14
py4j==0.10.9.7
pyaml==23.5.9
PyArabic==0.6.15
pyarrow==10.0.1
pyasn1==0.4.8
pyasn1-modules==0.2.7
PyAstronomy==0.19.0
pybind11==2.10.4
pyclipper==1.3.0.post4
pycodestyle==2.10.0
pycolmap @ file:///home/conda/feedstock_root/build_artifacts/pycolmap_1684621279577/work
pycosat @ file:///home/conda/feedstock_root/build_artifacts/pycosat_1666836542287/work
pycparser @ file:///tmp/build/80754af9/pycparser_1636541352034/work
pycryptodome==3.18.0
pyct==0.5.0
pycuda==2022.2.2
pydantic @ file:///home/conda/feedstock_root/build_artifacts/pydantic_1679565261911/work
pydegensac==0.1.2
pydicom==2.3.1
pydocstyle==6.3.0
pydot==1.4.2
pydub==0.25.1
pyemd==1.0.0
pyerfa==2.0.0.3
pyexcel-io==0.6.6
pyexcel-ods==0.6.0
pyfasttext==0.4.6
pyflakes==3.0.1
pygltflib==1.15.6
Pygments @ file:///home/conda/feedstock_root/build_artifacts/pygments_1681904169130/work
PyJWT==2.6.0
pykalman==0.9.5
pyLDAvis==3.2.2
pylibraft==23.4.1
pylint==2.17.4
pymc3==3.11.5
PyMeeus==0.5.12
pymongo==3.13.0
Pympler==1.0.1
pynndescent==0.5.10
pynvml @ file:///home/conda/feedstock_root/build_artifacts/pynvml_1639061605391/work
pynvrtc==9.2
pyocr==0.8.3
pyOpenSSL @ file:///home/conda/feedstock_root/build_artifacts/pyopenssl_1680037383858/work
pyparsing @ file:///home/conda/feedstock_root/build_artifacts/pyparsing_1652235407899/work
pypdf==3.9.0
pyproj @ file:///home/conda/feedstock_root/build_artifacts/pyproj_1680061961999/work
pyrsistent @ file:///home/conda/feedstock_root/build_artifacts/pyrsistent_1672681463845/work
pysal==23.1
pyshp @ file:///home/conda/feedstock_root/build_artifacts/pyshp_1659002966020/work
PySocks @ file:///home/builder/ci_310/pysocks_1640793678128/work
pytesseract==0.3.10
python-bidi==0.4.2
python-dateutil @ file:///home/conda/feedstock_root/build_artifacts/python-dateutil_1626286286081/work
python-dotenv==1.0.0
python-igraph==0.10.4
python-json-logger @ file:///home/conda/feedstock_root/build_artifacts/python-json-logger_1677079630776/work
python-Levenshtein==0.21.0
python-louvain==0.16
python-lsp-jsonrpc==1.0.0
python-lsp-server==1.7.3
python-multipart==0.0.6
python-slugify==8.0.1
python-utils==3.5.2
pythreejs==2.4.2
pytoolconfig==1.2.5
pytools==2022.1.14
pytorch-ignite==0.4.12
pytorch-lightning==2.0.2
pytz @ file:///home/conda/feedstock_root/build_artifacts/pytz_1680088766131/work
pyu2f @ file:///home/conda/feedstock_root/build_artifacts/pyu2f_1604248910016/work
PyUpSet==0.1.1.post7
pyviz-comms==2.2.1
PyWavelets==1.4.1
PyYAML==5.4.1
pyzmq @ file:///home/conda/feedstock_root/build_artifacts/pyzmq_1679316826707/work
qgrid==1.3.1
qtconsole==5.4.3
QtPy==2.3.1
quantecon==0.7.0
quantities==0.14.1
qudida==0.0.4
raft-dask==23.4.1
randomgen==1.23.1
rapidfuzz==3.0.0
rasterio==1.3.7
rasterstats==0.18.0
ray==2.4.0
ray-cpp==2.4.0
regex==2023.5.5
requests==2.28.2
requests-oauthlib==1.3.1
requests-toolbelt==0.10.1
responses==0.18.0
retrying==1.3.3
rfc3339-validator @ file:///home/conda/feedstock_root/build_artifacts/rfc3339-validator_1638811747357/work
rfc3986-validator @ file:///home/conda/feedstock_root/build_artifacts/rfc3986-validator_1598024191506/work
rgf-python==3.12.0
rich @ file:///home/conda/feedstock_root/build_artifacts/rich_1664752510089/work
rmm==23.4.1
rope==1.8.0
rsa @ file:///home/conda/feedstock_root/build_artifacts/rsa_1658328885051/work
Rtree==1.0.1
ruamel-yaml-conda @ file:///home/builder/ci_310/ruamel_yaml_1640794439226/work
ruamel.yaml @ file:///home/conda/feedstock_root/build_artifacts/ruamel.yaml_1683392649173/work
ruamel.yaml.clib @ file:///home/conda/feedstock_root/build_artifacts/ruamel.yaml.clib_1670412719074/work
s2sphere==0.2.5
s3fs==2023.5.0
s3transfer==0.6.1
safetensors==0.3.1
scattertext==0.1.19
scikit-image==0.20.0
scikit-learn==1.2.2
scikit-learn-intelex==2023.1.1
scikit-multilearn==0.2.0
scikit-optimize==0.9.0
scikit-plot==0.3.7
scikit-surprise==1.1.3
scipy==1.10.1
seaborn @ file:///home/conda/feedstock_root/build_artifacts/seaborn-split_1672497695270/work
SecretStorage==3.3.3
segment-anything @ git+https://github.com/facebookresearch/segment-anything.git@6fdee8f2727f4506cfbbe553e23b895e27956588
segregation==2.4.2
semver==3.0.0
Send2Trash @ file:///home/conda/feedstock_root/build_artifacts/send2trash_1682601222253/work
sentencepiece==0.1.96
sentry-sdk==1.24.0
setproctitle==1.3.2
setuptools-git==1.2
setuptools-scm==7.1.0
shap==0.41.0
Shapely==1.8.5.post1
shellingham @ file:///home/conda/feedstock_root/build_artifacts/shellingham_1676292972954/work
simpervisor==0.4
SimpleITK==2.2.1
simplejson==3.19.1
six @ file:///tmp/build/80754af9/six_1644875935023/work
sklearn-pandas==2.2.0
slicer==0.0.7
smart-open @ file:///home/conda/feedstock_root/build_artifacts/smart_open_split_1673202927732/work/dist
smhasher==0.150.1
smmap==5.0.0
sniffio @ file:///home/conda/feedstock_root/build_artifacts/sniffio_1662051266223/work
snowballstemmer==2.2.0
snuggs==1.4.7
sortedcontainers @ file:///home/conda/feedstock_root/build_artifacts/sortedcontainers_1621217038088/work
soundfile==0.12.1
soupsieve @ file:///home/conda/feedstock_root/build_artifacts/soupsieve_1658207591808/work
soxr==0.3.5
spacy @ file:///home/conda/feedstock_root/build_artifacts/spacy_1684226337030/work
spacy-legacy @ file:///home/conda/feedstock_root/build_artifacts/spacy-legacy_1674550301837/work
spacy-loggers @ file:///home/conda/feedstock_root/build_artifacts/spacy-loggers_1672303484730/work
spaghetti==1.7.2
spectral==0.23.1
spglm==1.0.8
sphinx-rtd-theme==0.2.4
spint==1.0.7
splot==1.1.5.post1
spopt==0.5.0
spreg==1.3.2
spvcm==0.3.0
SQLAlchemy==2.0.12
sqlglot==11.7.1
sqlparse==0.4.4
squarify==0.4.3
srsly @ file:///home/conda/feedstock_root/build_artifacts/srsly_1677657434449/work
stack-data @ file:///home/conda/feedstock_root/build_artifacts/stack_data_1669632077133/work
starlette==0.26.1
statsmodels==0.13.5
stemming==1.0.1
stop-words==2018.7.23
stopit==1.1.2
strip-hints==0.1.10
stumpy==1.11.1
sympy==1.12
tables==3.8.0
tabulate==0.9.0
tangled-up-in-unicode==0.2.0
tbb==2021.9.0
tblib @ file:///home/conda/feedstock_root/build_artifacts/tblib_1616261298899/work
tenacity==8.2.2
tensorboard==2.12.0
tensorboard-data-server==0.7.0
tensorboard-plugin-profile==2.11.2
tensorboard-plugin-wit==1.8.1
tensorboardX==2.6
tensorflow==2.12.0
tensorflow-addons==0.20.0
tensorflow-cloud==0.1.16
tensorflow-datasets==4.9.2
tensorflow-decision-forests==1.3.0
tensorflow-estimator==2.12.0
tensorflow-gcs-config==2.12.0
tensorflow-hub==0.12.0
tensorflow-io==0.31.0
tensorflow-io-gcs-filesystem==0.31.0
tensorflow-metadata==0.14.0
tensorflow-probability==0.20.0
tensorflow-serving-api==2.12.1
tensorflow-text==2.12.1
tensorflow-transform==0.14.0
tensorflowjs==3.15.0
tensorpack==0.11
tensorstore==0.1.36
termcolor==2.3.0
terminado @ file:///home/conda/feedstock_root/build_artifacts/terminado_1670253674810/work
testpath==0.6.0
text-unidecode==1.3
textblob==0.17.1
texttable==1.6.7
textwrap3==0.9.2
Theano==1.0.5
Theano-PyMC==1.1.2
thinc @ file:///home/conda/feedstock_root/build_artifacts/thinc_1683130983739/work
threadpoolctl==3.1.0
tifffile==2023.4.12
timm==0.9.2
tinycss2 @ file:///home/conda/feedstock_root/build_artifacts/tinycss2_1666100256010/work
tobler==0.10
tokenizers==0.13.2
toml==0.10.2
tomli==2.0.1
tomlkit==0.11.8
toolz @ file:///home/conda/feedstock_root/build_artifacts/toolz_1657485559105/work
torch==1.13.1
torchaudio @ file:///tmp/torch/torchaudio-2.0.1-cp310-cp310-linux_x86_64.whl#sha256=83e258b68459f1ff64301c19c2fc791a692fd372271abeef8414854aafd03b06
torchdata==0.6.0
torchinfo==1.8.0
torchmetrics==0.11.4
torchtext @ file:///tmp/torch/torchtext-0.15.1-cp310-cp310-linux_x86_64.whl#sha256=110ca71f44e505c040ea2f41dcaf798cd7de1b55cbedaa6687b9e21eec759844
torchtyping==0.1.4
torchvision==0.14.1
tornado==6.3.1
TPOT==0.11.7
tqdm @ file:///home/conda/feedstock_root/build_artifacts/tqdm_1677948868469/work
traceml==1.0.8
traitlets @ file:///home/conda/feedstock_root/build_artifacts/traitlets_1675110562325/work
traittypes==0.2.1
transformers @ git+https://github.com/huggingface/transformers.git@0dcb46e7a4a9e587ba84ff35778ab4233a184c11
treelite==3.2.0
treelite-runtime==3.2.0
triton==2.0.0
trlx @ git+https://github.com/CarperAI/trlx.git@b91da7b03d8e9fa0c0d6dce10a8f2611aca3013f
trueskill==0.4.5
tsfresh==0.20.0
typeguard==2.13.3
typer @ file:///home/conda/feedstock_root/build_artifacts/typer_1667832226065/work
typing-inspect==0.8.0
typing_extensions @ file:///home/conda/feedstock_root/build_artifacts/typing_extensions_1678559861143/work
tzlocal==5.0.1
uc-micro-py==1.0.2
ucx-py @ file:///opt/conda/conda-bld/work
ujson==5.7.0
umap-learn==0.5.3
unicodedata2 @ file:///home/conda/feedstock_root/build_artifacts/unicodedata2_1667239886688/work
Unidecode==1.3.6
update-checker==0.18.0
uri-template==1.2.0
uritemplate==3.0.1
urllib3 @ file:///home/conda/feedstock_root/build_artifacts/urllib3_1678635778344/work
urwid==2.1.2
urwid-readline==0.13
uvicorn==0.22.0
uvloop==0.17.0
vaex==4.16.0
vaex-astro==0.9.3
vaex-core==4.16.1
vaex-hdf5==0.14.1
vaex-jupyter==0.8.1
vaex-ml==0.18.1
vaex-server==0.8.1
vaex-viz==0.5.4
vecstack==0.4.0
virtualenv==20.21.0
visions==0.7.5
vowpalwabbit==9.8.0
vtk==9.2.6
Wand==0.6.11
wandb==0.13.10
wasabi @ file:///home/conda/feedstock_root/build_artifacts/wasabi_1673945962927/work
watchfiles==0.19.0
wavio==0.0.7
wcwidth @ file:///home/conda/feedstock_root/build_artifacts/wcwidth_1673864653149/work
webcolors==1.13
webencodings==0.5.1
websocket-client @ file:///home/conda/feedstock_root/build_artifacts/websocket-client_1675567828044/work
websockets==11.0.3
Werkzeug==2.3.4
wfdb==4.1.1
whatthepatch==1.0.5
widgetsnbextension==3.6.4
witwidget==1.8.1
woodwork==0.23.0
Wordbatch==1.4.9
wordcloud==1.9.2
wordsegment==1.3.1
wrapt==1.14.1
wurlitzer==3.0.3
xarray==2023.5.0
xarray-einstats==0.5.1
xgboost==1.7.5
xvfbwrapper==0.2.9
xxhash==3.2.0
xyzservices==2023.5.0
y-py==0.5.9
yapf==0.33.0
yarl @ file:///home/conda/feedstock_root/build_artifacts/yarl_1682426574163/work
ydata-profiling==4.1.2
yellowbrick==1.5
ypy-websocket==0.8.2
zict @ file:///home/conda/feedstock_root/build_artifacts/zict_1681770155528/work
zipp @ file:///home/conda/feedstock_root/build_artifacts/zipp_1677313463193/work
zstandard==0.19.0
  1. 在 kaggle 上面进行推理
image

我训练的是一个「对对子」的模型，训练命令如下：

!python ./Chinese-Vicuna/finetune.py --model_path yahma/llama-7b-hf --data_path ./test-data/couplet-10k.json --test_size 2000

即共 1w 条数据，2k 条测试集。

  1. kaggle 上测试没问题以后，我将 lora-Vicuna 目录下的 LoRA 模型拷贝下来：
image
  1. 本地运行这个模型推理，效果很差：
python ./Chinese-Vicuna/generate.py --model_path decapoda-research/llama-7b-hf --lora_path ./lora-Vicuna --use_local 0

本地显卡：NVIDIA GeForce GTX 1070

image
  1. 本地 pip freeze
absl-py==1.4.0
accelerate==0.15.0
aiofiles==23.1.0
aiohttp==3.8.4
aiosignal==1.3.1
altair==5.0.1
anyio==3.7.0
appdirs==1.4.4
async-timeout==4.0.2
attrs==23.1.0
bitsandbytes==0.37.1
cachetools==5.3.1
certifi==2023.5.7
charset-normalizer==3.1.0
click==8.1.3
cmake==3.26.3
contourpy==1.0.7
cycler==0.11.0
datasets==2.8.0
deepspeed==0.8.3
dill==0.3.6
distlib==0.3.6
docker-pycreds==0.4.0
einops==0.6.1
evaluate==0.4.0
exceptiongroup==1.1.1
fairscale==0.4.13
fastapi==0.96.0
ffmpy==0.3.0
filelock==3.12.0
fonttools==4.39.4
frozenlist==1.3.3
fsspec==2023.5.0
gitdb==4.0.10
GitPython==3.1.31
google-auth==2.19.1
google-auth-oauthlib==0.4.6
gradio==3.20.0
grpcio==1.51.3
h11==0.14.0
hjson==3.1.0
httpcore==0.17.2
httpx==0.24.1
huggingface-hub==0.13.3
idna==3.4
importlib-metadata==6.6.0
importlib-resources==5.12.0
Jinja2==3.1.2
jsonschema==4.17.3
kiwisolver==1.4.4
linkify-it-py==2.0.2
lit==16.0.5.post0
loralib==0.1.1
Markdown==3.4.3
markdown-it-py==2.2.0
MarkupSafe==2.1.3
matplotlib==3.7.1
mdit-py-plugins==0.3.3
mdurl==0.1.2
msgpack==1.0.5
multidict==6.0.4
multiprocess==0.70.14
networkx==3.1
ninja==1.11.1
numpy==1.24.3
nvidia-cublas-cu11==11.10.3.66
nvidia-cuda-nvrtc-cu11==11.7.99
nvidia-cuda-runtime-cu11==11.7.99
nvidia-cudnn-cu11==8.5.0.96
nvidia-ml-py==11.525.112
nvitop==1.0.0
oauthlib==3.2.2
orjson==3.9.0
packaging==23.1
pandas==2.0.2
pathtools==0.1.2
peft @ git+https://github.com/huggingface/peft.git@13e53fc7ee5d89d59b16523051006dddf0fb7a49
pi==0.1.2
Pillow==9.5.0
pkgutil_resolve_name==1.3.10
platformdirs==3.5.1
protobuf==4.23.2
psutil==5.9.5
py-cpuinfo==9.0.0
pyarrow==12.0.0
pyasn1==0.5.0
pyasn1-modules==0.3.0
pycryptodome==3.18.0
pydantic==1.10.8
pydub==0.25.1
Pygments==2.15.1
pyparsing==3.0.9
pyrsistent==0.19.3
python-dateutil==2.8.2
python-multipart==0.0.6
pytz==2023.3
PyYAML==6.0
ray==2.4.0
regex==2023.6.3
requests==2.31.0
requests-oauthlib==1.3.1
responses==0.18.0
rich==13.4.1
rsa==4.9
sentencepiece==0.1.96
sentry-sdk==1.25.0
setproctitle==1.3.2
six==1.16.0
smmap==5.0.0
sniffio==1.3.0
starlette==0.27.0
tabulate==0.9.0
tensorboard==2.12.0
tensorboard-data-server==0.7.0
tensorboard-plugin-wit==1.8.1
termcolor==2.3.0
texttable==1.6.7
tokenizers==0.13.2
toolz==0.12.0
torch==1.13.1
torchtyping==0.1.4
torchvision==0.14.1
tqdm==4.65.0
transformers @ git+https://github.com/huggingface/transformers.git@0dcb46e7a4a9e587ba84ff35778ab4233a184c11
triton==2.0.0
trlx @ git+https://github.com/CarperAI/trlx.git@b91da7b03d8e9fa0c0d6dce10a8f2611aca3013f
typeguard==4.0.0
typing_extensions==4.6.3
tzdata==2023.3
uc-micro-py==1.0.2
urllib3==1.26.16
uvicorn==0.22.0
virtualenv==20.21.0
wandb==0.13.10
websockets==11.0.3
Werkzeug==2.3.4
xxhash==3.2.0
yarl==1.9.2
zipp==3.15.0

其中因为各种报错，bitsandbytes 版本参考这两个 issue：https://github.com/TimDettmers/bitsandbytes/issues/134，https://github.com/TimDettmers/bitsandbytes/issues/179

请问这个可能跟什么有关系？
@jianghushinian
Copy link
Author

jianghushinian commented Jun 5, 2023 •
edited
Loading

GTX 1070 的机器中间有报错：RuntimeError: expected scalar type Half but found Float

我参考了这位网友的回答：https://github.com/Facico/Chinese-Vicuna/issues/210#issuecomment-1575984136，还是不行，切换了 bitsandbytes 版本就不报错了，但是效果不好。

此外，我在另一台 v100 机器上测试这个 LoRA 模型，效果依然不好。

@jianghushinian
Copy link
Author

找到了，应该还是显卡问题，参考这个：#39

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@jianghushinian