diff --git a/ai_worker/main.py b/ai_worker/main.py index b65171d..fb50daa 100644 --- a/ai_worker/main.py +++ b/ai_worker/main.py @@ -20,6 +20,7 @@ from httpx_sse import aconnect_sse from llama_cpp.server.app import Settings as LlamaSettings, create_app as create_llama_app import llama_cpp.server.app +import whisper from pydantic import BaseModel, Field from pydantic_settings import BaseSettings, SettingsConfigDict from pynvml.smi import nvidia_smi @@ -153,9 +154,11 @@ def __init__(self, conf: Config): self.slug = "" self.stopped = False self.llama = None + self.whisper = None self.llama_model = None + self.whisper_model = None self.llama_cli: Optional[AsyncClient] = None - + if FineTuner: self.fine_tuner = FineTuner(self.conf) else: @@ -270,6 +273,18 @@ def clear_llama_model(self): self.llama_cli = None self.llama_model = None + def clear_whisper_model(self): + self.whisper = None + self.whisper_model = None + + async def load_whisper_model(self, name): + assert name, "No model name" + if name == self.whisper_model: + return + log.debug("loading model: %s", name) + + self.whisper = whisper.load_model(name) + self.whisper_model = name async def load_model(self, name): assert name, "No model name" @@ -295,6 +310,7 @@ async def load_model(self, name): assert self.llama, "Load llama failed. Try lowering layers." self.llama_cli = AsyncClient(app=self.llama, base_url="http://test") self.llama_model = name + def _get_connect_info(self) -> ConnectMessage: disk_space = get_free_space_mb(".") @@ -302,6 +318,7 @@ def _get_connect_info(self) -> ConnectMessage: caps = [] caps += ['llama-infer'] + caps += ["whisper"] if self.fine_tuner: caps += ["llama-fine-tune"] @@ -412,6 +429,19 @@ async def run_ws(self): await asyncio.sleep(1) self.stopped = True + async def download_tmp_file(self, url: str) -> str: + import urllib.parse + res = urllib.parse.urlparse(self.conf.queen_url) + scheme = "https" if res.scheme == "wss" else "http" + async with AsyncClient() as cli: + async with cli.stream('GET', f"{scheme}://{res.netloc}/storage/?filename={url}") as response: + with tempfile.NamedTemporaryFile("wb", delete=False) as download_file: + async for chunk in response.aiter_bytes(): + download_file.write(chunk) + download_file.close() + return download_file.name + + async def run_one(self): event = None req_str = None @@ -419,7 +449,6 @@ async def run_one(self): req_str = await self.ws_recv() req = Req.model_validate_json(req_str) model = req.openai_req.get("model") - log.debug("loading %s", model) st = time.monotonic() @@ -435,6 +464,13 @@ async def run_one(self): elif req.openai_url == "/v1/embeddings" and model.startswith(MODEL_PREFIX): res = self.fast_embed.embed(req.openai_req) await self.ws_send(json.dumps(res), True) + elif req.openai_url == "/v1/audio/transcriptions": + await self.load_whisper_model(model) + filename = await self.download_tmp_file(req.openai_req["file"]) + # whisper.transcribe is not async wrap it to make it works + loop = asyncio.get_running_loop() + result = await loop.run_in_executor(None, lambda: self.whisper.transcribe(filename)) + await self.ws_send(json.dumps({"text": result["text"]}), True) elif req.openai_req.get("stream"): await self.load_model(model) async with aconnect_sse(self.llama_cli, "POST", req.openai_url, json=req.openai_req) as sse: @@ -469,7 +505,7 @@ async def handle_image_generation(self, request_data): async def get_model(self, name): return await self.download_model(name) - + async def download_file(self, url: str) -> str: output_file = url_to_tempfile(self.conf, url) if not os.path.exists(output_file): diff --git a/build-bin.sh b/build-bin.sh index 8ec20ad..ef87077 100755 --- a/build-bin.sh +++ b/build-bin.sh @@ -3,7 +3,7 @@ gpu="$1" arch="$2" cmake="$3" -opts="--onefile" +opts="--onedir" if [ -z "$cmake" -o -z "$gpu" ]; then echo usage build-bin.sh gpu arch "cmake-args" @@ -36,6 +36,8 @@ rm -f ~/.cache/pypoetry/artifacts/*/*/*/*/llama* CMAKE_ARGS="$cmake" FORCE_CMAKE=1 poetry install $with_torch +poetry run pip install openai-whisper + if [ "$gpu" == "cuda-torch" ]; then # annoying hack because fastembed is probably a bad idea pip install transformers==4.35.0 diff --git a/hooks/hook-whisper.py b/hooks/hook-whisper.py new file mode 100644 index 0000000..2e4aafe --- /dev/null +++ b/hooks/hook-whisper.py @@ -0,0 +1,26 @@ +from PyInstaller.utils.hooks import collect_data_files, get_package_paths +import os + +# Get the package path +package_path = get_package_paths('nvidia')[0] + +# Collect data files +datas = collect_data_files('nvidia') + +# Append the additional .dll or .so file +if os.name == 'nt': # Windows + # need to check because i dont have a gpu with windows to test + dll_path = os.path.join(package_path, 'nvidia', 'cudnn', 'lib', 'libcudnn_ops_infer.dll') + datas.append((dll_path, '.')) + dll_path = os.path.join(package_path, 'nvidia', 'cudnn', 'lib', 'libcudnn_cnn_infer.dll') + datas.append((dll_path, '.')) +elif os.name == 'posix': # Linux/Mac + so_path = os.path.join(package_path, 'nvidia', 'cudnn', 'lib', 'libcudnn_ops_infer.so.8') + datas.append((so_path, '.')) + so_path = os.path.join(package_path, 'nvidia', 'cudnn', 'lib', 'libcudnn_cnn_infer.so.8') + datas.append((so_path, '.')) + + +assets = os.path.join(package_path, "whisper", 'assets') +datas.append((assets, './whisper/assets')) + \ No newline at end of file diff --git a/poetry.lock b/poetry.lock index c3f9353..9640caf 100644 --- a/poetry.lock +++ b/poetry.lock @@ -1,4 +1,4 @@ -# This file is automatically @generated by Poetry 1.6.1 and should not be changed by hand. +# This file is automatically @generated by Poetry 1.7.1 and should not be changed by hand. [[package]] name = "accelerate" @@ -1125,6 +1125,39 @@ dev = ["black (>=23.3.0)", "httpx (>=0.24.1)", "mkdocs (>=1.4.3)", "mkdocs-mater server = ["fastapi (>=0.100.0)", "pydantic-settings (>=2.0.1)", "sse-starlette (>=1.6.1)", "starlette-context (>=0.3.6,<0.4)", "uvicorn (>=0.22.0)"] test = ["httpx (>=0.24.1)", "pytest (>=7.4.0)"] +[[package]] +name = "llvmlite" +version = "0.41.1" +description = "lightweight wrapper around basic LLVM functionality" +optional = false +python-versions = ">=3.8" +files = [ + {file = "llvmlite-0.41.1-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:c1e1029d47ee66d3a0c4d6088641882f75b93db82bd0e6178f7bd744ebce42b9"}, + {file = "llvmlite-0.41.1-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:150d0bc275a8ac664a705135e639178883293cf08c1a38de3bbaa2f693a0a867"}, + {file = "llvmlite-0.41.1-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:1eee5cf17ec2b4198b509272cf300ee6577229d237c98cc6e63861b08463ddc6"}, + {file = "llvmlite-0.41.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:0dd0338da625346538f1173a17cabf21d1e315cf387ca21b294ff209d176e244"}, + {file = "llvmlite-0.41.1-cp310-cp310-win32.whl", hash = "sha256:fa1469901a2e100c17eb8fe2678e34bd4255a3576d1a543421356e9c14d6e2ae"}, + {file = "llvmlite-0.41.1-cp310-cp310-win_amd64.whl", hash = "sha256:2b76acee82ea0e9304be6be9d4b3840208d050ea0dcad75b1635fa06e949a0ae"}, + {file = "llvmlite-0.41.1-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:210e458723436b2469d61b54b453474e09e12a94453c97ea3fbb0742ba5a83d8"}, + {file = "llvmlite-0.41.1-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:855f280e781d49e0640aef4c4af586831ade8f1a6c4df483fb901cbe1a48d127"}, + {file = "llvmlite-0.41.1-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:b67340c62c93a11fae482910dc29163a50dff3dfa88bc874872d28ee604a83be"}, + {file = "llvmlite-0.41.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:2181bb63ef3c607e6403813421b46982c3ac6bfc1f11fa16a13eaafb46f578e6"}, + {file = "llvmlite-0.41.1-cp311-cp311-win_amd64.whl", hash = "sha256:9564c19b31a0434f01d2025b06b44c7ed422f51e719ab5d24ff03b7560066c9a"}, + {file = "llvmlite-0.41.1-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:5940bc901fb0325970415dbede82c0b7f3e35c2d5fd1d5e0047134c2c46b3281"}, + {file = "llvmlite-0.41.1-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:8b0a9a47c28f67a269bb62f6256e63cef28d3c5f13cbae4fab587c3ad506778b"}, + {file = "llvmlite-0.41.1-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:f8afdfa6da33f0b4226af8e64cfc2b28986e005528fbf944d0a24a72acfc9432"}, + {file = "llvmlite-0.41.1-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:8454c1133ef701e8c050a59edd85d238ee18bb9a0eb95faf2fca8b909ee3c89a"}, + {file = "llvmlite-0.41.1-cp38-cp38-win32.whl", hash = "sha256:2d92c51e6e9394d503033ffe3292f5bef1566ab73029ec853861f60ad5c925d0"}, + {file = "llvmlite-0.41.1-cp38-cp38-win_amd64.whl", hash = "sha256:df75594e5a4702b032684d5481db3af990b69c249ccb1d32687b8501f0689432"}, + {file = "llvmlite-0.41.1-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:04725975e5b2af416d685ea0769f4ecc33f97be541e301054c9f741003085802"}, + {file = "llvmlite-0.41.1-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:bf14aa0eb22b58c231243dccf7e7f42f7beec48970f2549b3a6acc737d1a4ba4"}, + {file = "llvmlite-0.41.1-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:92c32356f669e036eb01016e883b22add883c60739bc1ebee3a1cc0249a50828"}, + {file = "llvmlite-0.41.1-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:24091a6b31242bcdd56ae2dbea40007f462260bc9bdf947953acc39dffd54f8f"}, + {file = "llvmlite-0.41.1-cp39-cp39-win32.whl", hash = "sha256:880cb57ca49e862e1cd077104375b9d1dfdc0622596dfa22105f470d7bacb309"}, + {file = "llvmlite-0.41.1-cp39-cp39-win_amd64.whl", hash = "sha256:92f093986ab92e71c9ffe334c002f96defc7986efda18397d0f08534f3ebdc4d"}, + {file = "llvmlite-0.41.1.tar.gz", hash = "sha256:f19f767a018e6ec89608e1f6b13348fa2fcde657151137cb64e56d48598a92db"}, +] + [[package]] name = "macholib" version = "1.16.3" @@ -1208,6 +1241,17 @@ files = [ {file = "MarkupSafe-2.1.3.tar.gz", hash = "sha256:af598ed32d6ae86f1b747b82783958b1a4ab8f617b06fe68795c7f026abbdcad"}, ] +[[package]] +name = "more-itertools" +version = "10.1.0" +description = "More routines for operating on iterables, beyond itertools" +optional = false +python-versions = ">=3.8" +files = [ + {file = "more-itertools-10.1.0.tar.gz", hash = "sha256:626c369fa0eb37bac0291bce8259b332fd59ac792fa5497b59837309cd5b114a"}, + {file = "more_itertools-10.1.0-py3-none-any.whl", hash = "sha256:64e0735fcfdc6f3464ea133afe8ea4483b1c5fe3a3d69852e6503b43a0b222e6"}, +] + [[package]] name = "mpmath" version = "1.3.0" @@ -1354,6 +1398,40 @@ doc = ["nb2plots (>=0.7)", "nbconvert (<7.9)", "numpydoc (>=1.6)", "pillow (>=9. extra = ["lxml (>=4.6)", "pydot (>=1.4.2)", "pygraphviz (>=1.11)", "sympy (>=1.10)"] test = ["pytest (>=7.2)", "pytest-cov (>=4.0)"] +[[package]] +name = "numba" +version = "0.58.1" +description = "compiling Python code using LLVM" +optional = false +python-versions = ">=3.8" +files = [ + {file = "numba-0.58.1-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:07f2fa7e7144aa6f275f27260e73ce0d808d3c62b30cff8906ad1dec12d87bbe"}, + {file = "numba-0.58.1-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:7bf1ddd4f7b9c2306de0384bf3854cac3edd7b4d8dffae2ec1b925e4c436233f"}, + {file = "numba-0.58.1-cp310-cp310-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:bc2d904d0319d7a5857bd65062340bed627f5bfe9ae4a495aef342f072880d50"}, + {file = "numba-0.58.1-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:4e79b6cc0d2bf064a955934a2e02bf676bc7995ab2db929dbbc62e4c16551be6"}, + {file = "numba-0.58.1-cp310-cp310-win_amd64.whl", hash = "sha256:81fe5b51532478149b5081311b0fd4206959174e660c372b94ed5364cfb37c82"}, + {file = "numba-0.58.1-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:bcecd3fb9df36554b342140a4d77d938a549be635d64caf8bd9ef6c47a47f8aa"}, + {file = "numba-0.58.1-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:a1eaa744f518bbd60e1f7ccddfb8002b3d06bd865b94a5d7eac25028efe0e0ff"}, + {file = "numba-0.58.1-cp311-cp311-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:bf68df9c307fb0aa81cacd33faccd6e419496fdc621e83f1efce35cdc5e79cac"}, + {file = "numba-0.58.1-cp311-cp311-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:55a01e1881120e86d54efdff1be08381886fe9f04fc3006af309c602a72bc44d"}, + {file = "numba-0.58.1-cp311-cp311-win_amd64.whl", hash = "sha256:811305d5dc40ae43c3ace5b192c670c358a89a4d2ae4f86d1665003798ea7a1a"}, + {file = "numba-0.58.1-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:ea5bfcf7d641d351c6a80e8e1826eb4a145d619870016eeaf20bbd71ef5caa22"}, + {file = "numba-0.58.1-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:e63d6aacaae1ba4ef3695f1c2122b30fa3d8ba039c8f517784668075856d79e2"}, + {file = "numba-0.58.1-cp38-cp38-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:6fe7a9d8e3bd996fbe5eac0683227ccef26cba98dae6e5cee2c1894d4b9f16c1"}, + {file = "numba-0.58.1-cp38-cp38-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:898af055b03f09d33a587e9425500e5be84fc90cd2f80b3fb71c6a4a17a7e354"}, + {file = "numba-0.58.1-cp38-cp38-win_amd64.whl", hash = "sha256:d3e2fe81fe9a59fcd99cc572002101119059d64d31eb6324995ee8b0f144a306"}, + {file = "numba-0.58.1-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:5c765aef472a9406a97ea9782116335ad4f9ef5c9f93fc05fd44aab0db486954"}, + {file = "numba-0.58.1-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:9e9356e943617f5e35a74bf56ff6e7cc83e6b1865d5e13cee535d79bf2cae954"}, + {file = "numba-0.58.1-cp39-cp39-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:240e7a1ae80eb6b14061dc91263b99dc8d6af9ea45d310751b780888097c1aaa"}, + {file = "numba-0.58.1-cp39-cp39-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:45698b995914003f890ad839cfc909eeb9c74921849c712a05405d1a79c50f68"}, + {file = "numba-0.58.1-cp39-cp39-win_amd64.whl", hash = "sha256:bd3dda77955be03ff366eebbfdb39919ce7c2620d86c906203bed92124989032"}, + {file = "numba-0.58.1.tar.gz", hash = "sha256:487ded0633efccd9ca3a46364b40006dbdaca0f95e99b8b83e778d1195ebcbaa"}, +] + +[package.dependencies] +llvmlite = "==0.41.*" +numpy = ">=1.22,<1.27" + [[package]] name = "numpy" version = "1.26.1" @@ -1645,6 +1723,28 @@ packaging = "*" protobuf = "*" sympy = "*" +[[package]] +name = "openai-whisper" +version = "20231117" +description = "Robust Speech Recognition via Large-Scale Weak Supervision" +optional = false +python-versions = ">=3.8" +files = [ + {file = "openai-whisper-20231117.tar.gz", hash = "sha256:7af424181436f1800cc0b7d75cf40ede34e9ddf1ba4983a910832fcf4aade4a4"}, +] + +[package.dependencies] +more-itertools = "*" +numba = "*" +numpy = "*" +tiktoken = "*" +torch = "*" +tqdm = "*" +triton = ">=2.0.0,<3" + +[package.extras] +dev = ["black", "flake8", "isort", "pytest", "scipy"] + [[package]] name = "packaging" version = "23.2" @@ -2313,6 +2413,20 @@ files = [ [package.extras] cli = ["click (>=5.0)"] +[[package]] +name = "python-multipart" +version = "0.0.6" +description = "A streaming multipart parser for Python" +optional = false +python-versions = ">=3.7" +files = [ + {file = "python_multipart-0.0.6-py3-none-any.whl", hash = "sha256:ee698bab5ef148b0a760751c261902cd096e57e10558e11aca17646b74ee1c18"}, + {file = "python_multipart-0.0.6.tar.gz", hash = "sha256:e9925a80bb668529f1b67c7fdb0a5dacdd7cbfc6fb0bff3ea443fe22bdd62132"}, +] + +[package.extras] +dev = ["atomicwrites (==1.2.1)", "attrs (==19.2.0)", "coverage (==6.5.0)", "hatch", "invoke (==1.7.3)", "more-itertools (==4.3.0)", "pbr (==4.3.0)", "pluggy (==1.0.0)", "py (==1.11.0)", "pytest (==7.2.0)", "pytest-cov (==4.0.0)", "pytest-timeout (==2.1.0)", "pyyaml (==5.1)"] + [[package]] name = "pytools" version = "2023.1.1" @@ -2868,6 +2982,58 @@ files = [ [package.dependencies] mpmath = ">=0.19" +[[package]] +name = "tiktoken" +version = "0.5.2" +description = "tiktoken is a fast BPE tokeniser for use with OpenAI's models" +optional = false +python-versions = ">=3.8" +files = [ + {file = "tiktoken-0.5.2-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:8c4e654282ef05ec1bd06ead22141a9a1687991cef2c6a81bdd1284301abc71d"}, + {file = "tiktoken-0.5.2-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:7b3134aa24319f42c27718c6967f3c1916a38a715a0fa73d33717ba121231307"}, + {file = "tiktoken-0.5.2-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:6092e6e77730929c8c6a51bb0d7cfdf1b72b63c4d033d6258d1f2ee81052e9e5"}, + {file = "tiktoken-0.5.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:72ad8ae2a747622efae75837abba59be6c15a8f31b4ac3c6156bc56ec7a8e631"}, + {file = "tiktoken-0.5.2-cp310-cp310-musllinux_1_1_aarch64.whl", hash = "sha256:51cba7c8711afa0b885445f0637f0fcc366740798c40b981f08c5f984e02c9d1"}, + {file = "tiktoken-0.5.2-cp310-cp310-musllinux_1_1_x86_64.whl", hash = "sha256:3d8c7d2c9313f8e92e987d585ee2ba0f7c40a0de84f4805b093b634f792124f5"}, + {file = "tiktoken-0.5.2-cp310-cp310-win_amd64.whl", hash = "sha256:692eca18c5fd8d1e0dde767f895c17686faaa102f37640e884eecb6854e7cca7"}, + {file = "tiktoken-0.5.2-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:138d173abbf1ec75863ad68ca289d4da30caa3245f3c8d4bfb274c4d629a2f77"}, + {file = "tiktoken-0.5.2-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:7388fdd684690973fdc450b47dfd24d7f0cbe658f58a576169baef5ae4658607"}, + {file = "tiktoken-0.5.2-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:a114391790113bcff670c70c24e166a841f7ea8f47ee2fe0e71e08b49d0bf2d4"}, + {file = "tiktoken-0.5.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:ca96f001e69f6859dd52926d950cfcc610480e920e576183497ab954e645e6ac"}, + {file = "tiktoken-0.5.2-cp311-cp311-musllinux_1_1_aarch64.whl", hash = "sha256:15fed1dd88e30dfadcdd8e53a8927f04e1f6f81ad08a5ca824858a593ab476c7"}, + {file = "tiktoken-0.5.2-cp311-cp311-musllinux_1_1_x86_64.whl", hash = "sha256:93f8e692db5756f7ea8cb0cfca34638316dcf0841fb8469de8ed7f6a015ba0b0"}, + {file = "tiktoken-0.5.2-cp311-cp311-win_amd64.whl", hash = "sha256:bcae1c4c92df2ffc4fe9f475bf8148dbb0ee2404743168bbeb9dcc4b79dc1fdd"}, + {file = "tiktoken-0.5.2-cp312-cp312-macosx_10_9_x86_64.whl", hash = "sha256:b76a1e17d4eb4357d00f0622d9a48ffbb23401dcf36f9716d9bd9c8e79d421aa"}, + {file = "tiktoken-0.5.2-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:01d8b171bb5df4035580bc26d4f5339a6fd58d06f069091899d4a798ea279d3e"}, + {file = "tiktoken-0.5.2-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:42adf7d4fb1ed8de6e0ff2e794a6a15005f056a0d83d22d1d6755a39bffd9e7f"}, + {file = "tiktoken-0.5.2-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:4c3f894dbe0adb44609f3d532b8ea10820d61fdcb288b325a458dfc60fefb7db"}, + {file = "tiktoken-0.5.2-cp312-cp312-musllinux_1_1_aarch64.whl", hash = "sha256:58ccfddb4e62f0df974e8f7e34a667981d9bb553a811256e617731bf1d007d19"}, + {file = "tiktoken-0.5.2-cp312-cp312-musllinux_1_1_x86_64.whl", hash = "sha256:58902a8bad2de4268c2a701f1c844d22bfa3cbcc485b10e8e3e28a050179330b"}, + {file = "tiktoken-0.5.2-cp312-cp312-win_amd64.whl", hash = "sha256:5e39257826d0647fcac403d8fa0a474b30d02ec8ffc012cfaf13083e9b5e82c5"}, + {file = "tiktoken-0.5.2-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:8bde3b0fbf09a23072d39c1ede0e0821f759b4fa254a5f00078909158e90ae1f"}, + {file = "tiktoken-0.5.2-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:2ddee082dcf1231ccf3a591d234935e6acf3e82ee28521fe99af9630bc8d2a60"}, + {file = "tiktoken-0.5.2-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:35c057a6a4e777b5966a7540481a75a31429fc1cb4c9da87b71c8b75b5143037"}, + {file = "tiktoken-0.5.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:4c4a049b87e28f1dc60509f8eb7790bc8d11f9a70d99b9dd18dfdd81a084ffe6"}, + {file = "tiktoken-0.5.2-cp38-cp38-musllinux_1_1_aarch64.whl", hash = "sha256:5bf5ce759089f4f6521ea6ed89d8f988f7b396e9f4afb503b945f5c949c6bec2"}, + {file = "tiktoken-0.5.2-cp38-cp38-musllinux_1_1_x86_64.whl", hash = "sha256:0c964f554af1a96884e01188f480dad3fc224c4bbcf7af75d4b74c4b74ae0125"}, + {file = "tiktoken-0.5.2-cp38-cp38-win_amd64.whl", hash = "sha256:368dd5726d2e8788e47ea04f32e20f72a2012a8a67af5b0b003d1e059f1d30a3"}, + {file = "tiktoken-0.5.2-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:a2deef9115b8cd55536c0a02c0203512f8deb2447f41585e6d929a0b878a0dd2"}, + {file = "tiktoken-0.5.2-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:2ed7d380195affbf886e2f8b92b14edfe13f4768ff5fc8de315adba5b773815e"}, + {file = "tiktoken-0.5.2-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:c76fce01309c8140ffe15eb34ded2bb94789614b7d1d09e206838fc173776a18"}, + {file = "tiktoken-0.5.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:60a5654d6a2e2d152637dd9a880b4482267dfc8a86ccf3ab1cec31a8c76bfae8"}, + {file = "tiktoken-0.5.2-cp39-cp39-musllinux_1_1_aarch64.whl", hash = "sha256:41d4d3228e051b779245a8ddd21d4336f8975563e92375662f42d05a19bdff41"}, + {file = "tiktoken-0.5.2-cp39-cp39-musllinux_1_1_x86_64.whl", hash = "sha256:a5c1cdec2c92fcde8c17a50814b525ae6a88e8e5b02030dc120b76e11db93f13"}, + {file = "tiktoken-0.5.2-cp39-cp39-win_amd64.whl", hash = "sha256:84ddb36faedb448a50b246e13d1b6ee3437f60b7169b723a4b2abad75e914f3e"}, + {file = "tiktoken-0.5.2.tar.gz", hash = "sha256:f54c581f134a8ea96ce2023ab221d4d4d81ab614efa0b2fbce926387deb56c80"}, +] + +[package.dependencies] +regex = ">=2022.1.18" +requests = ">=2.26.0" + +[package.extras] +blobfile = ["blobfile (>=2)"] + [[package]] name = "tokenizers" version = "0.13.3" @@ -3457,4 +3623,4 @@ testing = ["big-O", "jaraco.functools", "jaraco.itertools", "more-itertools", "p [metadata] lock-version = "2.0" python-versions = ">=3.10,<3.12" -content-hash = "bf5bc222fc88579f906ace5295144bd3dfe352c562e2968214a63c8f7ce3e69d" +content-hash = "e62a9f60124eeb975678c7d87fbaed70732db7a4f48cf050e75bb267ca9c724a" diff --git a/pyproject.toml b/pyproject.toml index 316d8e6..ccc19fb 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -26,6 +26,8 @@ pyopencl = "^2023.1.2" coincurve = "^18.0.0" # https://github.com/huggingface/datasets/issues/6330 fsspec = "2023.9.2" +python-multipart = "^0.0.6" +openai-whisper = "^20231117" [tool.poetry.group.torch] optional = true