diff --git a/README.md b/README.md
index 1f34fcb..06524f5 100644
--- a/README.md
+++ b/README.md
@@ -22,8 +22,8 @@ Abstract: **The current discourse around Large Language Models (LLMs) tends to f
| Chapter 4: Safety | [html](https://www.souzatharsis.com/tamingLLMs/notebooks/safety.html) | [ipynb](https://github.com/souzatharsis/tamingLLMs/blob/master/tamingllms/notebooks/safety.ipynb) | *Ready for Review* |
| Chapter 5: Preference-Based Alignment | [html](https://www.souzatharsis.com/tamingLLMs/notebooks/alignment.html) | [ipynb](https://github.com/souzatharsis/tamingLLMs/blob/master/tamingllms/notebooks/alignment.ipynb) | *Ready for Review* |
| Chapter 6: Local LLMs in Practice | [html](https://www.souzatharsis.com/tamingLLMs/notebooks/local.html) | [ipynb](https://github.com/souzatharsis/tamingLLMs/blob/master/tamingllms/notebooks/local.ipynb) | *Ready for Review* |
-| Chapter 7: The Cost Factor | | | |
-| Chapter 8: Frontiers | | | WIP |
+| Chapter 7: The Falling Cost Paradox | | | WIP |
+| Chapter 8: Frontiers | | | |
| Appendix A: Tools and Resources | | | |
diff --git a/poetry.lock b/poetry.lock
index 69cc5dc..9067dd2 100644
--- a/poetry.lock
+++ b/poetry.lock
@@ -13,13 +13,13 @@ files = [
[[package]]
name = "accelerate"
-version = "1.1.1"
+version = "1.2.1"
description = "Accelerate"
optional = false
python-versions = ">=3.9.0"
files = [
- {file = "accelerate-1.1.1-py3-none-any.whl", hash = "sha256:61edd81762131b8d4bede008643fa1e1f3bf59bec710ebda9771443e24feae02"},
- {file = "accelerate-1.1.1.tar.gz", hash = "sha256:0d39dfac557052bc735eb2703a0e87742879e1e40b88af8a2f9a93233d4cd7db"},
+ {file = "accelerate-1.2.1-py3-none-any.whl", hash = "sha256:be1cbb958cf837e7cdfbde46b812964b1b8ae94c9c7d94d921540beafcee8ddf"},
+ {file = "accelerate-1.2.1.tar.gz", hash = "sha256:03e161fc69d495daf2b9b5c8d5b43d06e2145520c04727b5bda56d49f1a43ab5"},
]
[package.dependencies]
@@ -60,18 +60,6 @@ pygments = ">=1.5"
dev = ["pillow", "pkginfo (>=1.10)", "playwright", "pre-commit", "setuptools", "twine (>=5.0)"]
tests = ["hypothesis", "pytest"]
-[[package]]
-name = "aenum"
-version = "3.1.15"
-description = "Advanced Enumerations (compatible with Python's stdlib Enum), NamedTuples, and NamedConstants"
-optional = false
-python-versions = "*"
-files = [
- {file = "aenum-3.1.15-py2-none-any.whl", hash = "sha256:27b1710b9d084de6e2e695dab78fe9f269de924b51ae2850170ee7e1ca6288a5"},
- {file = "aenum-3.1.15-py3-none-any.whl", hash = "sha256:e0dfaeea4c2bd362144b87377e2c61d91958c5ed0b4daf89cb6f45ae23af6288"},
- {file = "aenum-3.1.15.tar.gz", hash = "sha256:8cbd76cd18c4f870ff39b24284d3ea028fbe8731a58df3aa581e434c575b9559"},
-]
-
[[package]]
name = "aiohappyeyeballs"
version = "2.4.3"
@@ -590,17 +578,6 @@ files = [
[package.dependencies]
pycparser = "*"
-[[package]]
-name = "chardet"
-version = "5.2.0"
-description = "Universal encoding detector for Python 3"
-optional = false
-python-versions = ">=3.7"
-files = [
- {file = "chardet-5.2.0-py3-none-any.whl", hash = "sha256:e1cf59446890a00105fe7b7912492ea04b6e6f06d4b742b2c788469e34c82970"},
- {file = "chardet-5.2.0.tar.gz", hash = "sha256:1b3b6ff479a8c414bc3fa2c0852995695c4a026dcd6d0633b2dd092ca39c1cf7"},
-]
-
[[package]]
name = "charset-normalizer"
version = "3.4.0"
@@ -953,25 +930,6 @@ diagnostics = ["bokeh (>=3.1.0)", "jinja2 (>=2.10.3)"]
distributed = ["distributed (==2024.11.2)"]
test = ["pandas[test]", "pre-commit", "pytest", "pytest-cov", "pytest-rerunfailures", "pytest-timeout", "pytest-xdist"]
-[[package]]
-name = "dataproperty"
-version = "1.0.1"
-description = "Python library for extract property from data."
-optional = false
-python-versions = ">=3.7"
-files = [
- {file = "DataProperty-1.0.1-py3-none-any.whl", hash = "sha256:0b8b07d4fb6453fcf975b53d35dea41f3cfd69c9d79b5010c3cf224ff0407a7a"},
- {file = "DataProperty-1.0.1.tar.gz", hash = "sha256:723e5729fa6e885e127a771a983ee1e0e34bb141aca4ffe1f0bfa7cde34650a4"},
-]
-
-[package.dependencies]
-mbstrdecoder = ">=1.0.0,<2"
-typepy = {version = ">=1.2.0,<2", extras = ["datetime"]}
-
-[package.extras]
-logging = ["loguru (>=0.4.1,<1)"]
-test = ["pytest (>=6.0.1)", "pytest-md-report (>=0.3)", "tcolorpy (>=0.1.2)"]
-
[[package]]
name = "datasets"
version = "3.1.0"
@@ -1508,6 +1466,23 @@ files = [
[package.dependencies]
attrs = ">=19.3"
+[[package]]
+name = "gguf"
+version = "0.13.0"
+description = "Read and write ML models in GGUF for GGML"
+optional = false
+python-versions = ">=3.8"
+files = [
+ {file = "gguf-0.13.0-py3-none-any.whl", hash = "sha256:4e92a73b827c4618b55a615547b5120a904c2b34d02d5179dad5093c680623f6"},
+ {file = "gguf-0.13.0.tar.gz", hash = "sha256:9f29ccbb21fc6b6cf8b4741e88aaa563f0a1c748c26b5f7e304bb48612bf41b8"},
+]
+
+[package.dependencies]
+numpy = ">=1.17"
+pyyaml = ">=5.1"
+sentencepiece = ">=0.1.98,<=0.2.0"
+tqdm = ">=4.27"
+
[[package]]
name = "gitdb"
version = "4.0.11"
@@ -2538,54 +2513,6 @@ files = [
{file = "latexcodec-3.0.0.tar.gz", hash = "sha256:917dc5fe242762cc19d963e6548b42d63a118028cdd3361d62397e3b638b6bc5"},
]
-[[package]]
-name = "lighteval"
-version = "0.6.2"
-description = "A lightweight and configurable evaluation package"
-optional = false
-python-versions = ">=3.10"
-files = [
- {file = "lighteval-0.6.2-py3-none-any.whl", hash = "sha256:1832fff4ca76d4ec617b5242c60e5dcaa1df8966f9b8352af105386fb6c910ba"},
- {file = "lighteval-0.6.2.tar.gz", hash = "sha256:e48caf17c4136f973b5b9ee0692171b797692e068bd6c8efed14657b81500956"},
-]
-
-[package.dependencies]
-accelerate = {version = "*", optional = true, markers = "extra == \"accelerate\""}
-aenum = "3.1.15"
-colorama = "*"
-datasets = ">=2.14.0"
-fsspec = ">=2023.12.2"
-GitPython = ">=3.1.41"
-huggingface-hub = ">=0.23.0"
-nltk = "3.9.1"
-protobuf = "==3.20.*"
-pycountry = "*"
-pytablewriter = "*"
-rouge-score = "0.1.2"
-sacrebleu = "*"
-scikit-learn = "*"
-sentencepiece = ">=0.1.99"
-spacy = "3.7.2"
-termcolor = "2.3.0"
-torch = ">=2.0,<2.5"
-transformers = ">=4.38.0"
-
-[package.extras]
-accelerate = ["accelerate"]
-adapters = ["peft (==0.3.0)"]
-dev = ["lighteval[accelerate,multilingual,quality,tests]"]
-extended-tasks = ["langdetect", "openai", "tiktoken"]
-multilingual = ["jieba", "pyvi", "spacy[ja,ko,th]", "stanza"]
-nanotron = ["nanotron", "tensorboardX"]
-optimum = ["optimum (==1.12.0)"]
-quality = ["pre-commit", "ruff (==v0.2.2)"]
-quantization = ["auto-gptq (>=0.4.2)", "bitsandbytes (>=0.41.0)"]
-s3 = ["s3fs"]
-tensorboardx = ["tensorboardX"]
-tests = ["pytest (==7.4.0)"]
-tgi = ["text-generation (==0.6.0)"]
-vllm = ["more-itertools", "ray", "vllm"]
-
[[package]]
name = "linkify-it-py"
version = "2.0.3"
@@ -2699,160 +2626,6 @@ files = [
{file = "locket-1.0.0.tar.gz", hash = "sha256:5c0d4c052a8bbbf750e056a8e65ccd309086f4f0f18a2eac306a8dfa4112a632"},
]
-[[package]]
-name = "lxml"
-version = "5.3.0"
-description = "Powerful and Pythonic XML processing library combining libxml2/libxslt with the ElementTree API."
-optional = false
-python-versions = ">=3.6"
-files = [
- {file = "lxml-5.3.0-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:dd36439be765e2dde7660212b5275641edbc813e7b24668831a5c8ac91180656"},
- {file = "lxml-5.3.0-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:ae5fe5c4b525aa82b8076c1a59d642c17b6e8739ecf852522c6321852178119d"},
- {file = "lxml-5.3.0-cp310-cp310-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:501d0d7e26b4d261fca8132854d845e4988097611ba2531408ec91cf3fd9d20a"},
- {file = "lxml-5.3.0-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:fb66442c2546446944437df74379e9cf9e9db353e61301d1a0e26482f43f0dd8"},
- {file = "lxml-5.3.0-cp310-cp310-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:9e41506fec7a7f9405b14aa2d5c8abbb4dbbd09d88f9496958b6d00cb4d45330"},
- {file = "lxml-5.3.0-cp310-cp310-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:f7d4a670107d75dfe5ad080bed6c341d18c4442f9378c9f58e5851e86eb79965"},
- {file = "lxml-5.3.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:41ce1f1e2c7755abfc7e759dc34d7d05fd221723ff822947132dc934d122fe22"},
- {file = "lxml-5.3.0-cp310-cp310-manylinux_2_28_aarch64.whl", hash = "sha256:44264ecae91b30e5633013fb66f6ddd05c006d3e0e884f75ce0b4755b3e3847b"},
- {file = "lxml-5.3.0-cp310-cp310-manylinux_2_28_ppc64le.whl", hash = "sha256:3c174dc350d3ec52deb77f2faf05c439331d6ed5e702fc247ccb4e6b62d884b7"},
- {file = "lxml-5.3.0-cp310-cp310-manylinux_2_28_s390x.whl", hash = "sha256:2dfab5fa6a28a0b60a20638dc48e6343c02ea9933e3279ccb132f555a62323d8"},
- {file = "lxml-5.3.0-cp310-cp310-manylinux_2_28_x86_64.whl", hash = "sha256:b1c8c20847b9f34e98080da785bb2336ea982e7f913eed5809e5a3c872900f32"},
- {file = "lxml-5.3.0-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:2c86bf781b12ba417f64f3422cfc302523ac9cd1d8ae8c0f92a1c66e56ef2e86"},
- {file = "lxml-5.3.0-cp310-cp310-musllinux_1_2_ppc64le.whl", hash = "sha256:c162b216070f280fa7da844531169be0baf9ccb17263cf5a8bf876fcd3117fa5"},
- {file = "lxml-5.3.0-cp310-cp310-musllinux_1_2_s390x.whl", hash = "sha256:36aef61a1678cb778097b4a6eeae96a69875d51d1e8f4d4b491ab3cfb54b5a03"},
- {file = "lxml-5.3.0-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:f65e5120863c2b266dbcc927b306c5b78e502c71edf3295dfcb9501ec96e5fc7"},
- {file = "lxml-5.3.0-cp310-cp310-win32.whl", hash = "sha256:ef0c1fe22171dd7c7c27147f2e9c3e86f8bdf473fed75f16b0c2e84a5030ce80"},
- {file = "lxml-5.3.0-cp310-cp310-win_amd64.whl", hash = "sha256:052d99051e77a4f3e8482c65014cf6372e61b0a6f4fe9edb98503bb5364cfee3"},
- {file = "lxml-5.3.0-cp311-cp311-macosx_10_9_universal2.whl", hash = "sha256:74bcb423462233bc5d6066e4e98b0264e7c1bed7541fff2f4e34fe6b21563c8b"},
- {file = "lxml-5.3.0-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:a3d819eb6f9b8677f57f9664265d0a10dd6551d227afb4af2b9cd7bdc2ccbf18"},
- {file = "lxml-5.3.0-cp311-cp311-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:5b8f5db71b28b8c404956ddf79575ea77aa8b1538e8b2ef9ec877945b3f46442"},
- {file = "lxml-5.3.0-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:2c3406b63232fc7e9b8783ab0b765d7c59e7c59ff96759d8ef9632fca27c7ee4"},
- {file = "lxml-5.3.0-cp311-cp311-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:2ecdd78ab768f844c7a1d4a03595038c166b609f6395e25af9b0f3f26ae1230f"},
- {file = "lxml-5.3.0-cp311-cp311-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:168f2dfcfdedf611eb285efac1516c8454c8c99caf271dccda8943576b67552e"},
- {file = "lxml-5.3.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:aa617107a410245b8660028a7483b68e7914304a6d4882b5ff3d2d3eb5948d8c"},
- {file = "lxml-5.3.0-cp311-cp311-manylinux_2_28_aarch64.whl", hash = "sha256:69959bd3167b993e6e710b99051265654133a98f20cec1d9b493b931942e9c16"},
- {file = "lxml-5.3.0-cp311-cp311-manylinux_2_28_ppc64le.whl", hash = "sha256:bd96517ef76c8654446fc3db9242d019a1bb5fe8b751ba414765d59f99210b79"},
- {file = "lxml-5.3.0-cp311-cp311-manylinux_2_28_s390x.whl", hash = "sha256:ab6dd83b970dc97c2d10bc71aa925b84788c7c05de30241b9e96f9b6d9ea3080"},
- {file = "lxml-5.3.0-cp311-cp311-manylinux_2_28_x86_64.whl", hash = "sha256:eec1bb8cdbba2925bedc887bc0609a80e599c75b12d87ae42ac23fd199445654"},
- {file = "lxml-5.3.0-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:6a7095eeec6f89111d03dabfe5883a1fd54da319c94e0fb104ee8f23616b572d"},
- {file = "lxml-5.3.0-cp311-cp311-musllinux_1_2_ppc64le.whl", hash = "sha256:6f651ebd0b21ec65dfca93aa629610a0dbc13dbc13554f19b0113da2e61a4763"},
- {file = "lxml-5.3.0-cp311-cp311-musllinux_1_2_s390x.whl", hash = "sha256:f422a209d2455c56849442ae42f25dbaaba1c6c3f501d58761c619c7836642ec"},
- {file = "lxml-5.3.0-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:62f7fdb0d1ed2065451f086519865b4c90aa19aed51081979ecd05a21eb4d1be"},
- {file = "lxml-5.3.0-cp311-cp311-win32.whl", hash = "sha256:c6379f35350b655fd817cd0d6cbeef7f265f3ae5fedb1caae2eb442bbeae9ab9"},
- {file = "lxml-5.3.0-cp311-cp311-win_amd64.whl", hash = "sha256:9c52100e2c2dbb0649b90467935c4b0de5528833c76a35ea1a2691ec9f1ee7a1"},
- {file = "lxml-5.3.0-cp312-cp312-macosx_10_9_universal2.whl", hash = "sha256:e99f5507401436fdcc85036a2e7dc2e28d962550afe1cbfc07c40e454256a859"},
- {file = "lxml-5.3.0-cp312-cp312-macosx_10_9_x86_64.whl", hash = "sha256:384aacddf2e5813a36495233b64cb96b1949da72bef933918ba5c84e06af8f0e"},
- {file = "lxml-5.3.0-cp312-cp312-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:874a216bf6afaf97c263b56371434e47e2c652d215788396f60477540298218f"},
- {file = "lxml-5.3.0-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:65ab5685d56914b9a2a34d67dd5488b83213d680b0c5d10b47f81da5a16b0b0e"},
- {file = "lxml-5.3.0-cp312-cp312-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:aac0bbd3e8dd2d9c45ceb82249e8bdd3ac99131a32b4d35c8af3cc9db1657179"},
- {file = "lxml-5.3.0-cp312-cp312-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:b369d3db3c22ed14c75ccd5af429086f166a19627e84a8fdade3f8f31426e52a"},
- {file = "lxml-5.3.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:c24037349665434f375645fa9d1f5304800cec574d0310f618490c871fd902b3"},
- {file = "lxml-5.3.0-cp312-cp312-manylinux_2_28_aarch64.whl", hash = "sha256:62d172f358f33a26d6b41b28c170c63886742f5b6772a42b59b4f0fa10526cb1"},
- {file = "lxml-5.3.0-cp312-cp312-manylinux_2_28_ppc64le.whl", hash = "sha256:c1f794c02903c2824fccce5b20c339a1a14b114e83b306ff11b597c5f71a1c8d"},
- {file = "lxml-5.3.0-cp312-cp312-manylinux_2_28_s390x.whl", hash = "sha256:5d6a6972b93c426ace71e0be9a6f4b2cfae9b1baed2eed2006076a746692288c"},
- {file = "lxml-5.3.0-cp312-cp312-manylinux_2_28_x86_64.whl", hash = "sha256:3879cc6ce938ff4eb4900d901ed63555c778731a96365e53fadb36437a131a99"},
- {file = "lxml-5.3.0-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:74068c601baff6ff021c70f0935b0c7bc528baa8ea210c202e03757c68c5a4ff"},
- {file = "lxml-5.3.0-cp312-cp312-musllinux_1_2_ppc64le.whl", hash = "sha256:ecd4ad8453ac17bc7ba3868371bffb46f628161ad0eefbd0a855d2c8c32dd81a"},
- {file = "lxml-5.3.0-cp312-cp312-musllinux_1_2_s390x.whl", hash = "sha256:7e2f58095acc211eb9d8b5771bf04df9ff37d6b87618d1cbf85f92399c98dae8"},
- {file = "lxml-5.3.0-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:e63601ad5cd8f860aa99d109889b5ac34de571c7ee902d6812d5d9ddcc77fa7d"},
- {file = "lxml-5.3.0-cp312-cp312-win32.whl", hash = "sha256:17e8d968d04a37c50ad9c456a286b525d78c4a1c15dd53aa46c1d8e06bf6fa30"},
- {file = "lxml-5.3.0-cp312-cp312-win_amd64.whl", hash = "sha256:c1a69e58a6bb2de65902051d57fde951febad631a20a64572677a1052690482f"},
- {file = "lxml-5.3.0-cp313-cp313-macosx_10_13_universal2.whl", hash = "sha256:8c72e9563347c7395910de6a3100a4840a75a6f60e05af5e58566868d5eb2d6a"},
- {file = "lxml-5.3.0-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:e92ce66cd919d18d14b3856906a61d3f6b6a8500e0794142338da644260595cd"},
- {file = "lxml-5.3.0-cp313-cp313-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:1d04f064bebdfef9240478f7a779e8c5dc32b8b7b0b2fc6a62e39b928d428e51"},
- {file = "lxml-5.3.0-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:5c2fb570d7823c2bbaf8b419ba6e5662137f8166e364a8b2b91051a1fb40ab8b"},
- {file = "lxml-5.3.0-cp313-cp313-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:0c120f43553ec759f8de1fee2f4794452b0946773299d44c36bfe18e83caf002"},
- {file = "lxml-5.3.0-cp313-cp313-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:562e7494778a69086f0312ec9689f6b6ac1c6b65670ed7d0267e49f57ffa08c4"},
- {file = "lxml-5.3.0-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:423b121f7e6fa514ba0c7918e56955a1d4470ed35faa03e3d9f0e3baa4c7e492"},
- {file = "lxml-5.3.0-cp313-cp313-manylinux_2_28_aarch64.whl", hash = "sha256:c00f323cc00576df6165cc9d21a4c21285fa6b9989c5c39830c3903dc4303ef3"},
- {file = "lxml-5.3.0-cp313-cp313-manylinux_2_28_ppc64le.whl", hash = "sha256:1fdc9fae8dd4c763e8a31e7630afef517eab9f5d5d31a278df087f307bf601f4"},
- {file = "lxml-5.3.0-cp313-cp313-manylinux_2_28_s390x.whl", hash = "sha256:658f2aa69d31e09699705949b5fc4719cbecbd4a97f9656a232e7d6c7be1a367"},
- {file = "lxml-5.3.0-cp313-cp313-manylinux_2_28_x86_64.whl", hash = "sha256:1473427aff3d66a3fa2199004c3e601e6c4500ab86696edffdbc84954c72d832"},
- {file = "lxml-5.3.0-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:a87de7dd873bf9a792bf1e58b1c3887b9264036629a5bf2d2e6579fe8e73edff"},
- {file = "lxml-5.3.0-cp313-cp313-musllinux_1_2_ppc64le.whl", hash = "sha256:0d7b36afa46c97875303a94e8f3ad932bf78bace9e18e603f2085b652422edcd"},
- {file = "lxml-5.3.0-cp313-cp313-musllinux_1_2_s390x.whl", hash = "sha256:cf120cce539453ae086eacc0130a324e7026113510efa83ab42ef3fcfccac7fb"},
- {file = "lxml-5.3.0-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:df5c7333167b9674aa8ae1d4008fa4bc17a313cc490b2cca27838bbdcc6bb15b"},
- {file = "lxml-5.3.0-cp313-cp313-win32.whl", hash = "sha256:c802e1c2ed9f0c06a65bc4ed0189d000ada8049312cfeab6ca635e39c9608957"},
- {file = "lxml-5.3.0-cp313-cp313-win_amd64.whl", hash = "sha256:406246b96d552e0503e17a1006fd27edac678b3fcc9f1be71a2f94b4ff61528d"},
- {file = "lxml-5.3.0-cp36-cp36m-macosx_10_9_x86_64.whl", hash = "sha256:8f0de2d390af441fe8b2c12626d103540b5d850d585b18fcada58d972b74a74e"},
- {file = "lxml-5.3.0-cp36-cp36m-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:1afe0a8c353746e610bd9031a630a95bcfb1a720684c3f2b36c4710a0a96528f"},
- {file = "lxml-5.3.0-cp36-cp36m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:56b9861a71575f5795bde89256e7467ece3d339c9b43141dbdd54544566b3b94"},
- {file = "lxml-5.3.0-cp36-cp36m-manylinux_2_28_x86_64.whl", hash = "sha256:9fb81d2824dff4f2e297a276297e9031f46d2682cafc484f49de182aa5e5df99"},
- {file = "lxml-5.3.0-cp36-cp36m-manylinux_2_5_x86_64.manylinux1_x86_64.whl", hash = "sha256:2c226a06ecb8cdef28845ae976da407917542c5e6e75dcac7cc33eb04aaeb237"},
- {file = "lxml-5.3.0-cp36-cp36m-musllinux_1_2_x86_64.whl", hash = "sha256:7d3d1ca42870cdb6d0d29939630dbe48fa511c203724820fc0fd507b2fb46577"},
- {file = "lxml-5.3.0-cp36-cp36m-win32.whl", hash = "sha256:094cb601ba9f55296774c2d57ad68730daa0b13dc260e1f941b4d13678239e70"},
- {file = "lxml-5.3.0-cp36-cp36m-win_amd64.whl", hash = "sha256:eafa2c8658f4e560b098fe9fc54539f86528651f61849b22111a9b107d18910c"},
- {file = "lxml-5.3.0-cp37-cp37m-macosx_10_9_x86_64.whl", hash = "sha256:cb83f8a875b3d9b458cada4f880fa498646874ba4011dc974e071a0a84a1b033"},
- {file = "lxml-5.3.0-cp37-cp37m-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:25f1b69d41656b05885aa185f5fdf822cb01a586d1b32739633679699f220391"},
- {file = "lxml-5.3.0-cp37-cp37m-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:23e0553b8055600b3bf4a00b255ec5c92e1e4aebf8c2c09334f8368e8bd174d6"},
- {file = "lxml-5.3.0-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:9ada35dd21dc6c039259596b358caab6b13f4db4d4a7f8665764d616daf9cc1d"},
- {file = "lxml-5.3.0-cp37-cp37m-manylinux_2_28_aarch64.whl", hash = "sha256:81b4e48da4c69313192d8c8d4311e5d818b8be1afe68ee20f6385d0e96fc9512"},
- {file = "lxml-5.3.0-cp37-cp37m-manylinux_2_28_x86_64.whl", hash = "sha256:2bc9fd5ca4729af796f9f59cd8ff160fe06a474da40aca03fcc79655ddee1a8b"},
- {file = "lxml-5.3.0-cp37-cp37m-musllinux_1_2_aarch64.whl", hash = "sha256:07da23d7ee08577760f0a71d67a861019103e4812c87e2fab26b039054594cc5"},
- {file = "lxml-5.3.0-cp37-cp37m-musllinux_1_2_x86_64.whl", hash = "sha256:ea2e2f6f801696ad7de8aec061044d6c8c0dd4037608c7cab38a9a4d316bfb11"},
- {file = "lxml-5.3.0-cp37-cp37m-win32.whl", hash = "sha256:5c54afdcbb0182d06836cc3d1be921e540be3ebdf8b8a51ee3ef987537455f84"},
- {file = "lxml-5.3.0-cp37-cp37m-win_amd64.whl", hash = "sha256:f2901429da1e645ce548bf9171784c0f74f0718c3f6150ce166be39e4dd66c3e"},
- {file = "lxml-5.3.0-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:c56a1d43b2f9ee4786e4658c7903f05da35b923fb53c11025712562d5cc02753"},
- {file = "lxml-5.3.0-cp38-cp38-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:6ee8c39582d2652dcd516d1b879451500f8db3fe3607ce45d7c5957ab2596040"},
- {file = "lxml-5.3.0-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:0fdf3a3059611f7585a78ee10399a15566356116a4288380921a4b598d807a22"},
- {file = "lxml-5.3.0-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:146173654d79eb1fc97498b4280c1d3e1e5d58c398fa530905c9ea50ea849b22"},
- {file = "lxml-5.3.0-cp38-cp38-manylinux_2_28_aarch64.whl", hash = "sha256:0a7056921edbdd7560746f4221dca89bb7a3fe457d3d74267995253f46343f15"},
- {file = "lxml-5.3.0-cp38-cp38-manylinux_2_28_x86_64.whl", hash = "sha256:9e4b47ac0f5e749cfc618efdf4726269441014ae1d5583e047b452a32e221920"},
- {file = "lxml-5.3.0-cp38-cp38-musllinux_1_2_aarch64.whl", hash = "sha256:f914c03e6a31deb632e2daa881fe198461f4d06e57ac3d0e05bbcab8eae01945"},
- {file = "lxml-5.3.0-cp38-cp38-musllinux_1_2_x86_64.whl", hash = "sha256:213261f168c5e1d9b7535a67e68b1f59f92398dd17a56d934550837143f79c42"},
- {file = "lxml-5.3.0-cp38-cp38-win32.whl", hash = "sha256:218c1b2e17a710e363855594230f44060e2025b05c80d1f0661258142b2add2e"},
- {file = "lxml-5.3.0-cp38-cp38-win_amd64.whl", hash = "sha256:315f9542011b2c4e1d280e4a20ddcca1761993dda3afc7a73b01235f8641e903"},
- {file = "lxml-5.3.0-cp39-cp39-macosx_10_9_universal2.whl", hash = "sha256:1ffc23010330c2ab67fac02781df60998ca8fe759e8efde6f8b756a20599c5de"},
- {file = "lxml-5.3.0-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:2b3778cb38212f52fac9fe913017deea2fdf4eb1a4f8e4cfc6b009a13a6d3fcc"},
- {file = "lxml-5.3.0-cp39-cp39-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:4b0c7a688944891086ba192e21c5229dea54382f4836a209ff8d0a660fac06be"},
- {file = "lxml-5.3.0-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:747a3d3e98e24597981ca0be0fd922aebd471fa99d0043a3842d00cdcad7ad6a"},
- {file = "lxml-5.3.0-cp39-cp39-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:86a6b24b19eaebc448dc56b87c4865527855145d851f9fc3891673ff97950540"},
- {file = "lxml-5.3.0-cp39-cp39-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:b11a5d918a6216e521c715b02749240fb07ae5a1fefd4b7bf12f833bc8b4fe70"},
- {file = "lxml-5.3.0-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:68b87753c784d6acb8a25b05cb526c3406913c9d988d51f80adecc2b0775d6aa"},
- {file = "lxml-5.3.0-cp39-cp39-manylinux_2_28_aarch64.whl", hash = "sha256:109fa6fede314cc50eed29e6e56c540075e63d922455346f11e4d7a036d2b8cf"},
- {file = "lxml-5.3.0-cp39-cp39-manylinux_2_28_ppc64le.whl", hash = "sha256:02ced472497b8362c8e902ade23e3300479f4f43e45f4105c85ef43b8db85229"},
- {file = "lxml-5.3.0-cp39-cp39-manylinux_2_28_s390x.whl", hash = "sha256:6b038cc86b285e4f9fea2ba5ee76e89f21ed1ea898e287dc277a25884f3a7dfe"},
- {file = "lxml-5.3.0-cp39-cp39-manylinux_2_28_x86_64.whl", hash = "sha256:7437237c6a66b7ca341e868cda48be24b8701862757426852c9b3186de1da8a2"},
- {file = "lxml-5.3.0-cp39-cp39-musllinux_1_2_aarch64.whl", hash = "sha256:7f41026c1d64043a36fda21d64c5026762d53a77043e73e94b71f0521939cc71"},
- {file = "lxml-5.3.0-cp39-cp39-musllinux_1_2_ppc64le.whl", hash = "sha256:482c2f67761868f0108b1743098640fbb2a28a8e15bf3f47ada9fa59d9fe08c3"},
- {file = "lxml-5.3.0-cp39-cp39-musllinux_1_2_s390x.whl", hash = "sha256:1483fd3358963cc5c1c9b122c80606a3a79ee0875bcac0204149fa09d6ff2727"},
- {file = "lxml-5.3.0-cp39-cp39-musllinux_1_2_x86_64.whl", hash = "sha256:2dec2d1130a9cda5b904696cec33b2cfb451304ba9081eeda7f90f724097300a"},
- {file = "lxml-5.3.0-cp39-cp39-win32.whl", hash = "sha256:a0eabd0a81625049c5df745209dc7fcef6e2aea7793e5f003ba363610aa0a3ff"},
- {file = "lxml-5.3.0-cp39-cp39-win_amd64.whl", hash = "sha256:89e043f1d9d341c52bf2af6d02e6adde62e0a46e6755d5eb60dc6e4f0b8aeca2"},
- {file = "lxml-5.3.0-pp310-pypy310_pp73-macosx_10_15_x86_64.whl", hash = "sha256:7b1cd427cb0d5f7393c31b7496419da594fe600e6fdc4b105a54f82405e6626c"},
- {file = "lxml-5.3.0-pp310-pypy310_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:51806cfe0279e06ed8500ce19479d757db42a30fd509940b1701be9c86a5ff9a"},
- {file = "lxml-5.3.0-pp310-pypy310_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:ee70d08fd60c9565ba8190f41a46a54096afa0eeb8f76bd66f2c25d3b1b83005"},
- {file = "lxml-5.3.0-pp310-pypy310_pp73-manylinux_2_28_aarch64.whl", hash = "sha256:8dc2c0395bea8254d8daebc76dcf8eb3a95ec2a46fa6fae5eaccee366bfe02ce"},
- {file = "lxml-5.3.0-pp310-pypy310_pp73-manylinux_2_28_x86_64.whl", hash = "sha256:6ba0d3dcac281aad8a0e5b14c7ed6f9fa89c8612b47939fc94f80b16e2e9bc83"},
- {file = "lxml-5.3.0-pp310-pypy310_pp73-win_amd64.whl", hash = "sha256:6e91cf736959057f7aac7adfc83481e03615a8e8dd5758aa1d95ea69e8931dba"},
- {file = "lxml-5.3.0-pp37-pypy37_pp73-macosx_10_9_x86_64.whl", hash = "sha256:94d6c3782907b5e40e21cadf94b13b0842ac421192f26b84c45f13f3c9d5dc27"},
- {file = "lxml-5.3.0-pp37-pypy37_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:c300306673aa0f3ed5ed9372b21867690a17dba38c68c44b287437c362ce486b"},
- {file = "lxml-5.3.0-pp37-pypy37_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:78d9b952e07aed35fe2e1a7ad26e929595412db48535921c5013edc8aa4a35ce"},
- {file = "lxml-5.3.0-pp37-pypy37_pp73-manylinux_2_28_aarch64.whl", hash = "sha256:01220dca0d066d1349bd6a1726856a78f7929f3878f7e2ee83c296c69495309e"},
- {file = "lxml-5.3.0-pp37-pypy37_pp73-manylinux_2_28_x86_64.whl", hash = "sha256:2d9b8d9177afaef80c53c0a9e30fa252ff3036fb1c6494d427c066a4ce6a282f"},
- {file = "lxml-5.3.0-pp37-pypy37_pp73-win_amd64.whl", hash = "sha256:20094fc3f21ea0a8669dc4c61ed7fa8263bd37d97d93b90f28fc613371e7a875"},
- {file = "lxml-5.3.0-pp38-pypy38_pp73-macosx_10_9_x86_64.whl", hash = "sha256:ace2c2326a319a0bb8a8b0e5b570c764962e95818de9f259ce814ee666603f19"},
- {file = "lxml-5.3.0-pp38-pypy38_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:92e67a0be1639c251d21e35fe74df6bcc40cba445c2cda7c4a967656733249e2"},
- {file = "lxml-5.3.0-pp38-pypy38_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:dd5350b55f9fecddc51385463a4f67a5da829bc741e38cf689f38ec9023f54ab"},
- {file = "lxml-5.3.0-pp38-pypy38_pp73-manylinux_2_28_aarch64.whl", hash = "sha256:4c1fefd7e3d00921c44dc9ca80a775af49698bbfd92ea84498e56acffd4c5469"},
- {file = "lxml-5.3.0-pp38-pypy38_pp73-manylinux_2_28_x86_64.whl", hash = "sha256:71a8dd38fbd2f2319136d4ae855a7078c69c9a38ae06e0c17c73fd70fc6caad8"},
- {file = "lxml-5.3.0-pp38-pypy38_pp73-win_amd64.whl", hash = "sha256:97acf1e1fd66ab53dacd2c35b319d7e548380c2e9e8c54525c6e76d21b1ae3b1"},
- {file = "lxml-5.3.0-pp39-pypy39_pp73-macosx_10_15_x86_64.whl", hash = "sha256:68934b242c51eb02907c5b81d138cb977b2129a0a75a8f8b60b01cb8586c7b21"},
- {file = "lxml-5.3.0-pp39-pypy39_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:b710bc2b8292966b23a6a0121f7a6c51d45d2347edcc75f016ac123b8054d3f2"},
- {file = "lxml-5.3.0-pp39-pypy39_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:18feb4b93302091b1541221196a2155aa296c363fd233814fa11e181adebc52f"},
- {file = "lxml-5.3.0-pp39-pypy39_pp73-manylinux_2_28_aarch64.whl", hash = "sha256:3eb44520c4724c2e1a57c0af33a379eee41792595023f367ba3952a2d96c2aab"},
- {file = "lxml-5.3.0-pp39-pypy39_pp73-manylinux_2_28_x86_64.whl", hash = "sha256:609251a0ca4770e5a8768ff902aa02bf636339c5a93f9349b48eb1f606f7f3e9"},
- {file = "lxml-5.3.0-pp39-pypy39_pp73-win_amd64.whl", hash = "sha256:516f491c834eb320d6c843156440fe7fc0d50b33e44387fcec5b02f0bc118a4c"},
- {file = "lxml-5.3.0.tar.gz", hash = "sha256:4e109ca30d1edec1ac60cdbe341905dc3b8f55b16855e03a54aaf59e51ec8c6f"},
-]
-
-[package.extras]
-cssselect = ["cssselect (>=0.7)"]
-html-clean = ["lxml-html-clean"]
-html5 = ["html5lib"]
-htmlsoup = ["BeautifulSoup4"]
-source = ["Cython (>=3.0.11)"]
-
[[package]]
name = "marisa-trie"
version = "1.2.1"
@@ -3115,23 +2888,6 @@ files = [
[package.dependencies]
traitlets = "*"
-[[package]]
-name = "mbstrdecoder"
-version = "1.1.3"
-description = "mbstrdecoder is a Python library for multi-byte character string decoder"
-optional = false
-python-versions = ">=3.7"
-files = [
- {file = "mbstrdecoder-1.1.3-py3-none-any.whl", hash = "sha256:d66c1ed3f2dc4e7c5d87cd44a75be10bc5af4250f95b38bbaedd7851308ce938"},
- {file = "mbstrdecoder-1.1.3.tar.gz", hash = "sha256:dcfd2c759322eb44fe193a9e0b1b86c5b87f3ec5ea8e1bb43b3e9ae423f1e8fe"},
-]
-
-[package.dependencies]
-chardet = ">=3.0.4,<6"
-
-[package.extras]
-test = ["Faker (>=1.0.2)", "pytest (>=6.0.1)", "pytest-md-report (>=0.1)"]
-
[[package]]
name = "mdit-py-plugins"
version = "0.4.2"
@@ -3722,46 +3478,50 @@ files = [
[[package]]
name = "nvidia-cublas-cu12"
-version = "12.1.3.1"
+version = "12.4.5.8"
description = "CUBLAS native runtime libraries"
optional = false
python-versions = ">=3"
files = [
- {file = "nvidia_cublas_cu12-12.1.3.1-py3-none-manylinux1_x86_64.whl", hash = "sha256:ee53ccca76a6fc08fb9701aa95b6ceb242cdaab118c3bb152af4e579af792728"},
- {file = "nvidia_cublas_cu12-12.1.3.1-py3-none-win_amd64.whl", hash = "sha256:2b964d60e8cf11b5e1073d179d85fa340c120e99b3067558f3cf98dd69d02906"},
+ {file = "nvidia_cublas_cu12-12.4.5.8-py3-none-manylinux2014_aarch64.whl", hash = "sha256:0f8aa1706812e00b9f19dfe0cdb3999b092ccb8ca168c0db5b8ea712456fd9b3"},
+ {file = "nvidia_cublas_cu12-12.4.5.8-py3-none-manylinux2014_x86_64.whl", hash = "sha256:2fc8da60df463fdefa81e323eef2e36489e1c94335b5358bcb38360adf75ac9b"},
+ {file = "nvidia_cublas_cu12-12.4.5.8-py3-none-win_amd64.whl", hash = "sha256:5a796786da89203a0657eda402bcdcec6180254a8ac22d72213abc42069522dc"},
]
[[package]]
name = "nvidia-cuda-cupti-cu12"
-version = "12.1.105"
+version = "12.4.127"
description = "CUDA profiling tools runtime libs."
optional = false
python-versions = ">=3"
files = [
- {file = "nvidia_cuda_cupti_cu12-12.1.105-py3-none-manylinux1_x86_64.whl", hash = "sha256:e54fde3983165c624cb79254ae9818a456eb6e87a7fd4d56a2352c24ee542d7e"},
- {file = "nvidia_cuda_cupti_cu12-12.1.105-py3-none-win_amd64.whl", hash = "sha256:bea8236d13a0ac7190bd2919c3e8e6ce1e402104276e6f9694479e48bb0eb2a4"},
+ {file = "nvidia_cuda_cupti_cu12-12.4.127-py3-none-manylinux2014_aarch64.whl", hash = "sha256:79279b35cf6f91da114182a5ce1864997fd52294a87a16179ce275773799458a"},
+ {file = "nvidia_cuda_cupti_cu12-12.4.127-py3-none-manylinux2014_x86_64.whl", hash = "sha256:9dec60f5ac126f7bb551c055072b69d85392b13311fcc1bcda2202d172df30fb"},
+ {file = "nvidia_cuda_cupti_cu12-12.4.127-py3-none-win_amd64.whl", hash = "sha256:5688d203301ab051449a2b1cb6690fbe90d2b372f411521c86018b950f3d7922"},
]
[[package]]
name = "nvidia-cuda-nvrtc-cu12"
-version = "12.1.105"
+version = "12.4.127"
description = "NVRTC native runtime libraries"
optional = false
python-versions = ">=3"
files = [
- {file = "nvidia_cuda_nvrtc_cu12-12.1.105-py3-none-manylinux1_x86_64.whl", hash = "sha256:339b385f50c309763ca65456ec75e17bbefcbbf2893f462cb8b90584cd27a1c2"},
- {file = "nvidia_cuda_nvrtc_cu12-12.1.105-py3-none-win_amd64.whl", hash = "sha256:0a98a522d9ff138b96c010a65e145dc1b4850e9ecb75a0172371793752fd46ed"},
+ {file = "nvidia_cuda_nvrtc_cu12-12.4.127-py3-none-manylinux2014_aarch64.whl", hash = "sha256:0eedf14185e04b76aa05b1fea04133e59f465b6f960c0cbf4e37c3cb6b0ea198"},
+ {file = "nvidia_cuda_nvrtc_cu12-12.4.127-py3-none-manylinux2014_x86_64.whl", hash = "sha256:a178759ebb095827bd30ef56598ec182b85547f1508941a3d560eb7ea1fbf338"},
+ {file = "nvidia_cuda_nvrtc_cu12-12.4.127-py3-none-win_amd64.whl", hash = "sha256:a961b2f1d5f17b14867c619ceb99ef6fcec12e46612711bcec78eb05068a60ec"},
]
[[package]]
name = "nvidia-cuda-runtime-cu12"
-version = "12.1.105"
+version = "12.4.127"
description = "CUDA Runtime native Libraries"
optional = false
python-versions = ">=3"
files = [
- {file = "nvidia_cuda_runtime_cu12-12.1.105-py3-none-manylinux1_x86_64.whl", hash = "sha256:6e258468ddf5796e25f1dc591a31029fa317d97a0a94ed93468fc86301d61e40"},
- {file = "nvidia_cuda_runtime_cu12-12.1.105-py3-none-win_amd64.whl", hash = "sha256:dfb46ef84d73fababab44cf03e3b83f80700d27ca300e537f85f636fac474344"},
+ {file = "nvidia_cuda_runtime_cu12-12.4.127-py3-none-manylinux2014_aarch64.whl", hash = "sha256:961fe0e2e716a2a1d967aab7caee97512f71767f852f67432d572e36cb3a11f3"},
+ {file = "nvidia_cuda_runtime_cu12-12.4.127-py3-none-manylinux2014_x86_64.whl", hash = "sha256:64403288fa2136ee8e467cdc9c9427e0434110899d07c779f25b5c068934faa5"},
+ {file = "nvidia_cuda_runtime_cu12-12.4.127-py3-none-win_amd64.whl", hash = "sha256:09c2e35f48359752dfa822c09918211844a3d93c100a715d79b59591130c5e1e"},
]
[[package]]
@@ -3780,35 +3540,41 @@ nvidia-cublas-cu12 = "*"
[[package]]
name = "nvidia-cufft-cu12"
-version = "11.0.2.54"
+version = "11.2.1.3"
description = "CUFFT native runtime libraries"
optional = false
python-versions = ">=3"
files = [
- {file = "nvidia_cufft_cu12-11.0.2.54-py3-none-manylinux1_x86_64.whl", hash = "sha256:794e3948a1aa71fd817c3775866943936774d1c14e7628c74f6f7417224cdf56"},
- {file = "nvidia_cufft_cu12-11.0.2.54-py3-none-win_amd64.whl", hash = "sha256:d9ac353f78ff89951da4af698f80870b1534ed69993f10a4cf1d96f21357e253"},
+ {file = "nvidia_cufft_cu12-11.2.1.3-py3-none-manylinux2014_aarch64.whl", hash = "sha256:5dad8008fc7f92f5ddfa2101430917ce2ffacd86824914c82e28990ad7f00399"},
+ {file = "nvidia_cufft_cu12-11.2.1.3-py3-none-manylinux2014_x86_64.whl", hash = "sha256:f083fc24912aa410be21fa16d157fed2055dab1cc4b6934a0e03cba69eb242b9"},
+ {file = "nvidia_cufft_cu12-11.2.1.3-py3-none-win_amd64.whl", hash = "sha256:d802f4954291101186078ccbe22fc285a902136f974d369540fd4a5333d1440b"},
]
+[package.dependencies]
+nvidia-nvjitlink-cu12 = "*"
+
[[package]]
name = "nvidia-curand-cu12"
-version = "10.3.2.106"
+version = "10.3.5.147"
description = "CURAND native runtime libraries"
optional = false
python-versions = ">=3"
files = [
- {file = "nvidia_curand_cu12-10.3.2.106-py3-none-manylinux1_x86_64.whl", hash = "sha256:9d264c5036dde4e64f1de8c50ae753237c12e0b1348738169cd0f8a536c0e1e0"},
- {file = "nvidia_curand_cu12-10.3.2.106-py3-none-win_amd64.whl", hash = "sha256:75b6b0c574c0037839121317e17fd01f8a69fd2ef8e25853d826fec30bdba74a"},
+ {file = "nvidia_curand_cu12-10.3.5.147-py3-none-manylinux2014_aarch64.whl", hash = "sha256:1f173f09e3e3c76ab084aba0de819c49e56614feae5c12f69883f4ae9bb5fad9"},
+ {file = "nvidia_curand_cu12-10.3.5.147-py3-none-manylinux2014_x86_64.whl", hash = "sha256:a88f583d4e0bb643c49743469964103aa59f7f708d862c3ddb0fc07f851e3b8b"},
+ {file = "nvidia_curand_cu12-10.3.5.147-py3-none-win_amd64.whl", hash = "sha256:f307cc191f96efe9e8f05a87096abc20d08845a841889ef78cb06924437f6771"},
]
[[package]]
name = "nvidia-cusolver-cu12"
-version = "11.4.5.107"
+version = "11.6.1.9"
description = "CUDA solver native runtime libraries"
optional = false
python-versions = ">=3"
files = [
- {file = "nvidia_cusolver_cu12-11.4.5.107-py3-none-manylinux1_x86_64.whl", hash = "sha256:8a7ec542f0412294b15072fa7dab71d31334014a69f953004ea7a118206fe0dd"},
- {file = "nvidia_cusolver_cu12-11.4.5.107-py3-none-win_amd64.whl", hash = "sha256:74e0c3a24c78612192a74fcd90dd117f1cf21dea4822e66d89e8ea80e3cd2da5"},
+ {file = "nvidia_cusolver_cu12-11.6.1.9-py3-none-manylinux2014_aarch64.whl", hash = "sha256:d338f155f174f90724bbde3758b7ac375a70ce8e706d70b018dd3375545fc84e"},
+ {file = "nvidia_cusolver_cu12-11.6.1.9-py3-none-manylinux2014_x86_64.whl", hash = "sha256:19e33fa442bcfd085b3086c4ebf7e8debc07cfe01e11513cc6d332fd918ac260"},
+ {file = "nvidia_cusolver_cu12-11.6.1.9-py3-none-win_amd64.whl", hash = "sha256:e77314c9d7b694fcebc84f58989f3aa4fb4cb442f12ca1a9bde50f5e8f6d1b9c"},
]
[package.dependencies]
@@ -3818,13 +3584,14 @@ nvidia-nvjitlink-cu12 = "*"
[[package]]
name = "nvidia-cusparse-cu12"
-version = "12.1.0.106"
+version = "12.3.1.170"
description = "CUSPARSE native runtime libraries"
optional = false
python-versions = ">=3"
files = [
- {file = "nvidia_cusparse_cu12-12.1.0.106-py3-none-manylinux1_x86_64.whl", hash = "sha256:f3b50f42cf363f86ab21f720998517a659a48131e8d538dc02f8768237bd884c"},
- {file = "nvidia_cusparse_cu12-12.1.0.106-py3-none-win_amd64.whl", hash = "sha256:b798237e81b9719373e8fae8d4f091b70a0cf09d9d85c95a557e11df2d8e9a5a"},
+ {file = "nvidia_cusparse_cu12-12.3.1.170-py3-none-manylinux2014_aarch64.whl", hash = "sha256:9d32f62896231ebe0480efd8a7f702e143c98cfaa0e8a76df3386c1ba2b54df3"},
+ {file = "nvidia_cusparse_cu12-12.3.1.170-py3-none-manylinux2014_x86_64.whl", hash = "sha256:ea4f11a2904e2a8dc4b1833cc1b5181cde564edd0d5cd33e3c168eff2d1863f1"},
+ {file = "nvidia_cusparse_cu12-12.3.1.170-py3-none-win_amd64.whl", hash = "sha256:9bc90fb087bc7b4c15641521f31c0371e9a612fc2ba12c338d3ae032e6b6797f"},
]
[package.dependencies]
@@ -3832,13 +3599,12 @@ nvidia-nvjitlink-cu12 = "*"
[[package]]
name = "nvidia-nccl-cu12"
-version = "2.20.5"
+version = "2.21.5"
description = "NVIDIA Collective Communication Library (NCCL) Runtime"
optional = false
python-versions = ">=3"
files = [
- {file = "nvidia_nccl_cu12-2.20.5-py3-none-manylinux2014_aarch64.whl", hash = "sha256:1fc150d5c3250b170b29410ba682384b14581db722b2531b0d8d33c595f33d01"},
- {file = "nvidia_nccl_cu12-2.20.5-py3-none-manylinux2014_x86_64.whl", hash = "sha256:057f6bf9685f75215d0c53bf3ac4a10b3e6578351de307abad9e18a99182af56"},
+ {file = "nvidia_nccl_cu12-2.21.5-py3-none-manylinux2014_x86_64.whl", hash = "sha256:8579076d30a8c24988834445f8d633c697d42397e92ffc3f63fa26766d25e0a0"},
]
[[package]]
@@ -3855,13 +3621,14 @@ files = [
[[package]]
name = "nvidia-nvtx-cu12"
-version = "12.1.105"
+version = "12.4.127"
description = "NVIDIA Tools Extension"
optional = false
python-versions = ">=3"
files = [
- {file = "nvidia_nvtx_cu12-12.1.105-py3-none-manylinux1_x86_64.whl", hash = "sha256:dc21cf308ca5691e7c04d962e213f8a4aa9bbfa23d95412f452254c2caeb09e5"},
- {file = "nvidia_nvtx_cu12-12.1.105-py3-none-win_amd64.whl", hash = "sha256:65f4d98982b31b60026e0e6de73fbdfc09d08a96f4656dd3665ca616a11e1e82"},
+ {file = "nvidia_nvtx_cu12-12.4.127-py3-none-manylinux2014_aarch64.whl", hash = "sha256:7959ad635db13edf4fc65c06a6e9f9e55fc2f92596db928d169c0bb031e88ef3"},
+ {file = "nvidia_nvtx_cu12-12.4.127-py3-none-manylinux2014_x86_64.whl", hash = "sha256:781e950d9b9f60d8241ccea575b32f5105a5baf4c2351cab5256a24869f12a1a"},
+ {file = "nvidia_nvtx_cu12-12.4.127-py3-none-win_amd64.whl", hash = "sha256:641dccaaa1139f3ffb0d3164b4b84f9d253397e38246a4f2f36728b48566d485"},
]
[[package]]
@@ -4293,22 +4060,6 @@ toolz = "*"
[package.extras]
complete = ["blosc", "numpy (>=1.20.0)", "pandas (>=1.3)", "pyzmq"]
-[[package]]
-name = "pathvalidate"
-version = "3.2.1"
-description = "pathvalidate is a Python library to sanitize/validate a string such as filenames/file-paths/etc."
-optional = false
-python-versions = ">=3.7"
-files = [
- {file = "pathvalidate-3.2.1-py3-none-any.whl", hash = "sha256:9a6255eb8f63c9e2135b9be97a5ce08f10230128c4ae7b3e935378b82b22c4c9"},
- {file = "pathvalidate-3.2.1.tar.gz", hash = "sha256:f5d07b1e2374187040612a1fcd2bcb2919f8db180df254c9581bb90bf903377d"},
-]
-
-[package.extras]
-docs = ["Sphinx (>=2.4)", "sphinx-rtd-theme (>=1.2.2)", "urllib3 (<2)"]
-readme = ["path (>=13,<17)", "readmemaker (>=1.1.0)"]
-test = ["Faker (>=1.0.8)", "allpairspy (>=2)", "click (>=6.2)", "pytest (>=6.0.1)", "pytest-md-report (>=0.6.2)"]
-
[[package]]
name = "pexpect"
version = "4.9.0"
@@ -4442,25 +4193,6 @@ docs = ["furo (>=2024.8.6)", "proselint (>=0.14)", "sphinx (>=8.0.2)", "sphinx-a
test = ["appdirs (==1.4.4)", "covdefaults (>=2.3)", "pytest (>=8.3.2)", "pytest-cov (>=5)", "pytest-mock (>=3.14)"]
type = ["mypy (>=1.11.2)"]
-[[package]]
-name = "portalocker"
-version = "3.0.0"
-description = "Wraps the portalocker recipe for easy usage"
-optional = false
-python-versions = ">=3.8"
-files = [
- {file = "portalocker-3.0.0-py3-none-any.whl", hash = "sha256:211916b539a0dc3c128a3d9e86893ecfefec5379c4ff684e798f0a00f99db406"},
- {file = "portalocker-3.0.0.tar.gz", hash = "sha256:21f535de2e7a82c94c130c054adb5c7421d480d5619d61073996e2f89bcb879b"},
-]
-
-[package.dependencies]
-pywin32 = {version = ">=226", markers = "platform_system == \"Windows\""}
-
-[package.extras]
-docs = ["sphinx (>=1.7.1)"]
-redis = ["redis"]
-tests = ["pytest (>=5.4.1)", "pytest-cov (>=2.8.1)", "pytest-mypy (>=0.8.0)", "pytest-timeout (>=2.1.0)", "redis", "sphinx (>=6.0.0)", "types-redis"]
-
[[package]]
name = "preshed"
version = "3.0.9"
@@ -4677,37 +4409,6 @@ files = [
{file = "propcache-0.2.0.tar.gz", hash = "sha256:df81779732feb9d01e5d513fad0122efb3d53bbc75f61b2a4f29a020bc985e70"},
]
-[[package]]
-name = "protobuf"
-version = "3.20.3"
-description = "Protocol Buffers"
-optional = false
-python-versions = ">=3.7"
-files = [
- {file = "protobuf-3.20.3-cp310-cp310-manylinux2014_aarch64.whl", hash = "sha256:f4bd856d702e5b0d96a00ec6b307b0f51c1982c2bf9c0052cf9019e9a544ba99"},
- {file = "protobuf-3.20.3-cp310-cp310-manylinux_2_12_x86_64.manylinux2010_x86_64.whl", hash = "sha256:9aae4406ea63d825636cc11ffb34ad3379335803216ee3a856787bcf5ccc751e"},
- {file = "protobuf-3.20.3-cp310-cp310-win32.whl", hash = "sha256:28545383d61f55b57cf4df63eebd9827754fd2dc25f80c5253f9184235db242c"},
- {file = "protobuf-3.20.3-cp310-cp310-win_amd64.whl", hash = "sha256:67a3598f0a2dcbc58d02dd1928544e7d88f764b47d4a286202913f0b2801c2e7"},
- {file = "protobuf-3.20.3-cp36-cp36m-manylinux_2_5_x86_64.manylinux1_x86_64.whl", hash = "sha256:899dc660cd599d7352d6f10d83c95df430a38b410c1b66b407a6b29265d66469"},
- {file = "protobuf-3.20.3-cp37-cp37m-macosx_10_9_x86_64.whl", hash = "sha256:e64857f395505ebf3d2569935506ae0dfc4a15cb80dc25261176c784662cdcc4"},
- {file = "protobuf-3.20.3-cp37-cp37m-manylinux2014_aarch64.whl", hash = "sha256:d9e4432ff660d67d775c66ac42a67cf2453c27cb4d738fc22cb53b5d84c135d4"},
- {file = "protobuf-3.20.3-cp37-cp37m-manylinux_2_5_x86_64.manylinux1_x86_64.whl", hash = "sha256:74480f79a023f90dc6e18febbf7b8bac7508420f2006fabd512013c0c238f454"},
- {file = "protobuf-3.20.3-cp37-cp37m-win32.whl", hash = "sha256:b6cc7ba72a8850621bfec987cb72623e703b7fe2b9127a161ce61e61558ad905"},
- {file = "protobuf-3.20.3-cp37-cp37m-win_amd64.whl", hash = "sha256:8c0c984a1b8fef4086329ff8dd19ac77576b384079247c770f29cc8ce3afa06c"},
- {file = "protobuf-3.20.3-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:de78575669dddf6099a8a0f46a27e82a1783c557ccc38ee620ed8cc96d3be7d7"},
- {file = "protobuf-3.20.3-cp38-cp38-manylinux2014_aarch64.whl", hash = "sha256:f4c42102bc82a51108e449cbb32b19b180022941c727bac0cfd50170341f16ee"},
- {file = "protobuf-3.20.3-cp38-cp38-manylinux_2_5_x86_64.manylinux1_x86_64.whl", hash = "sha256:44246bab5dd4b7fbd3c0c80b6f16686808fab0e4aca819ade6e8d294a29c7050"},
- {file = "protobuf-3.20.3-cp38-cp38-win32.whl", hash = "sha256:c02ce36ec760252242a33967d51c289fd0e1c0e6e5cc9397e2279177716add86"},
- {file = "protobuf-3.20.3-cp38-cp38-win_amd64.whl", hash = "sha256:447d43819997825d4e71bf5769d869b968ce96848b6479397e29fc24c4a5dfe9"},
- {file = "protobuf-3.20.3-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:398a9e0c3eaceb34ec1aee71894ca3299605fa8e761544934378bbc6c97de23b"},
- {file = "protobuf-3.20.3-cp39-cp39-manylinux2014_aarch64.whl", hash = "sha256:bf01b5720be110540be4286e791db73f84a2b721072a3711efff6c324cdf074b"},
- {file = "protobuf-3.20.3-cp39-cp39-manylinux_2_5_x86_64.manylinux1_x86_64.whl", hash = "sha256:daa564862dd0d39c00f8086f88700fdbe8bc717e993a21e90711acfed02f2402"},
- {file = "protobuf-3.20.3-cp39-cp39-win32.whl", hash = "sha256:819559cafa1a373b7096a482b504ae8a857c89593cf3a25af743ac9ecbd23480"},
- {file = "protobuf-3.20.3-cp39-cp39-win_amd64.whl", hash = "sha256:03038ac1cfbc41aa21f6afcbcd357281d7521b4157926f30ebecc8d4ea59dcb7"},
- {file = "protobuf-3.20.3-py2.py3-none-any.whl", hash = "sha256:a7ca6d488aa8ff7f329d4c545b2dbad8ac31464f1d8b1c87ad1346717731e4db"},
- {file = "protobuf-3.20.3.tar.gz", hash = "sha256:2e3427429c9cffebf259491be0af70189607f365c2f41c7c3764af6f337105f2"},
-]
-
[[package]]
name = "psutil"
version = "6.1.0"
@@ -5090,42 +4791,6 @@ files = [
[package.extras]
diagrams = ["jinja2", "railroad-diagrams"]
-[[package]]
-name = "pytablewriter"
-version = "1.2.0"
-description = "pytablewriter is a Python library to write a table in various formats: AsciiDoc / CSV / Elasticsearch / HTML / JavaScript / JSON / LaTeX / LDJSON / LTSV / Markdown / MediaWiki / NumPy / Excel / Pandas / Python / reStructuredText / SQLite / TOML / TSV / YAML."
-optional = false
-python-versions = ">=3.7"
-files = [
- {file = "pytablewriter-1.2.0-py3-none-any.whl", hash = "sha256:4a30e2bb4bf5bc1069b1d2b2bc41947577c4517ab0875b23a5b194d296f543d8"},
- {file = "pytablewriter-1.2.0.tar.gz", hash = "sha256:0204a4bb684a22140d640f2599f09e137bcdc18b3dd49426f4a555016e246b46"},
-]
-
-[package.dependencies]
-DataProperty = ">=1.0.1,<2"
-mbstrdecoder = ">=1.0.0,<2"
-pathvalidate = ">=2.3.0,<4"
-setuptools = ">=38.3.0"
-tabledata = ">=1.3.1,<2"
-tcolorpy = ">=0.0.5,<1"
-typepy = {version = ">=1.3.2,<2", extras = ["datetime"]}
-
-[package.extras]
-all = ["PyYAML (>=3.11,<7)", "SimpleSQLite (>=1.3.2,<2)", "XlsxWriter (>=0.9.6,<4)", "dominate (>=2.1.5,<3)", "elasticsearch (>=8.0.1,<9)", "loguru (>=0.4.1,<1)", "pandas (>=0.25.3,<3)", "pytablereader (>=0.31.3,<2)", "pytablewriter-altcol-theme (>=0.1.0,<1)", "pytablewriter-altrow-theme (>=0.2.0,<1)", "simplejson (>=3.8.1,<4)", "toml (>=0.9.3,<1)", "xlwt"]
-docs = ["PyYAML (>=3.11,<7)", "SimpleSQLite (>=1.3.2,<2)", "Sphinx (>=2.4)", "XlsxWriter (>=0.9.6,<4)", "dominate (>=2.1.5,<3)", "elasticsearch (>=8.0.1,<9)", "loguru (>=0.4.1,<1)", "pandas (>=0.25.3,<3)", "pytablereader (>=0.31.3,<2)", "pytablewriter-altcol-theme (>=0.1.0,<1)", "pytablewriter-altrow-theme (>=0.2.0,<1)", "simplejson (>=3.8.1,<4)", "sphinx-rtd-theme (>=1.2.2)", "toml (>=0.9.3,<1)", "xlwt"]
-es = ["elasticsearch (>=8.0.1,<9)"]
-es8 = ["elasticsearch (>=8.0.1,<9)"]
-excel = ["XlsxWriter (>=0.9.6,<4)", "xlwt"]
-from = ["pytablereader (>=0.31.3,<2)"]
-html = ["dominate (>=2.1.5,<3)"]
-logging = ["loguru (>=0.4.1,<1)"]
-pandas = ["pandas (>=0.25.3,<3)"]
-sqlite = ["SimpleSQLite (>=1.3.2,<2)"]
-test = ["PyYAML (>=3.11,<7)", "SimpleSQLite (>=1.3.2,<2)", "XlsxWriter (>=0.9.6,<4)", "beautifulsoup4 (>=4.10)", "dominate (>=2.1.5,<3)", "elasticsearch (>=8.0.1,<9)", "loguru (>=0.4.1,<1)", "pandas (>=0.25.3,<3)", "pytablereader (>=0.31.3,<2)", "pytablereader[excel,sqlite] (>=0.31.3)", "pytablewriter-altcol-theme (>=0.1.0,<1)", "pytablewriter-altrow-theme (>=0.2.0,<1)", "pytest (>=6.0.1)", "pytest-md-report (>=0.4.1)", "simplejson (>=3.8.1,<4)", "sqliteschema (>=1.3.0)", "tablib (>=3.2.0)", "toml (>=0.9.3,<1)", "xlwt"]
-theme = ["pytablewriter-altcol-theme (>=0.1.0,<1)", "pytablewriter-altrow-theme (>=0.2.0,<1)"]
-toml = ["toml (>=0.9.3,<1)"]
-yaml = ["PyYAML (>=3.11,<7)"]
-
[[package]]
name = "python-dateutil"
version = "2.9.0.post0"
@@ -5693,30 +5358,6 @@ files = [
{file = "rpds_py-0.21.0.tar.gz", hash = "sha256:ed6378c9d66d0de903763e7706383d60c33829581f0adff47b6535f1802fa6db"},
]
-[[package]]
-name = "sacrebleu"
-version = "2.4.3"
-description = "Hassle-free computation of shareable, comparable, and reproducible BLEU, chrF, and TER scores"
-optional = false
-python-versions = ">=3.8"
-files = [
- {file = "sacrebleu-2.4.3-py3-none-any.whl", hash = "sha256:a976fd6998d8ced267a722120ec7fc47083c8e9745d8808ccee6424464a0aa31"},
- {file = "sacrebleu-2.4.3.tar.gz", hash = "sha256:e734b1e0baeaea6ade0fefc9d23bac3df50bf15775d8b78edc108db63654192a"},
-]
-
-[package.dependencies]
-colorama = "*"
-lxml = "*"
-numpy = ">=1.17"
-portalocker = "*"
-regex = "*"
-tabulate = ">=0.8.9"
-
-[package.extras]
-dev = ["lxml-stubs", "mypy", "pytest", "setuptools", "types-tabulate", "wheel"]
-ja = ["ipadic (>=1.0,<2.0)", "mecab-python3 (>=1.0.9,<2.0.0)"]
-ko = ["mecab-ko (>=1.0.0,<=1.0.1)", "mecab-ko-dic (>=1.0,<2.0)"]
-
[[package]]
name = "safetensors"
version = "0.4.5"
@@ -5849,106 +5490,6 @@ tensorflow = ["safetensors[numpy]", "tensorflow (>=2.11.0)"]
testing = ["h5py (>=3.7.0)", "huggingface-hub (>=0.12.1)", "hypothesis (>=6.70.2)", "pytest (>=7.2.0)", "pytest-benchmark (>=4.0.0)", "safetensors[numpy]", "setuptools-rust (>=1.5.2)"]
torch = ["safetensors[numpy]", "torch (>=1.10)"]
-[[package]]
-name = "scikit-learn"
-version = "1.5.2"
-description = "A set of python modules for machine learning and data mining"
-optional = false
-python-versions = ">=3.9"
-files = [
- {file = "scikit_learn-1.5.2-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:299406827fb9a4f862626d0fe6c122f5f87f8910b86fe5daa4c32dcd742139b6"},
- {file = "scikit_learn-1.5.2-cp310-cp310-macosx_12_0_arm64.whl", hash = "sha256:2d4cad1119c77930b235579ad0dc25e65c917e756fe80cab96aa3b9428bd3fb0"},
- {file = "scikit_learn-1.5.2-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:8c412ccc2ad9bf3755915e3908e677b367ebc8d010acbb3f182814524f2e5540"},
- {file = "scikit_learn-1.5.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:3a686885a4b3818d9e62904d91b57fa757fc2bed3e465c8b177be652f4dd37c8"},
- {file = "scikit_learn-1.5.2-cp310-cp310-win_amd64.whl", hash = "sha256:c15b1ca23d7c5f33cc2cb0a0d6aaacf893792271cddff0edbd6a40e8319bc113"},
- {file = "scikit_learn-1.5.2-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:03b6158efa3faaf1feea3faa884c840ebd61b6484167c711548fce208ea09445"},
- {file = "scikit_learn-1.5.2-cp311-cp311-macosx_12_0_arm64.whl", hash = "sha256:1ff45e26928d3b4eb767a8f14a9a6efbf1cbff7c05d1fb0f95f211a89fd4f5de"},
- {file = "scikit_learn-1.5.2-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:f763897fe92d0e903aa4847b0aec0e68cadfff77e8a0687cabd946c89d17e675"},
- {file = "scikit_learn-1.5.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:f8b0ccd4a902836493e026c03256e8b206656f91fbcc4fde28c57a5b752561f1"},
- {file = "scikit_learn-1.5.2-cp311-cp311-win_amd64.whl", hash = "sha256:6c16d84a0d45e4894832b3c4d0bf73050939e21b99b01b6fd59cbb0cf39163b6"},
- {file = "scikit_learn-1.5.2-cp312-cp312-macosx_10_9_x86_64.whl", hash = "sha256:f932a02c3f4956dfb981391ab24bda1dbd90fe3d628e4b42caef3e041c67707a"},
- {file = "scikit_learn-1.5.2-cp312-cp312-macosx_12_0_arm64.whl", hash = "sha256:3b923d119d65b7bd555c73be5423bf06c0105678ce7e1f558cb4b40b0a5502b1"},
- {file = "scikit_learn-1.5.2-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:f60021ec1574e56632be2a36b946f8143bf4e5e6af4a06d85281adc22938e0dd"},
- {file = "scikit_learn-1.5.2-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:394397841449853c2290a32050382edaec3da89e35b3e03d6cc966aebc6a8ae6"},
- {file = "scikit_learn-1.5.2-cp312-cp312-win_amd64.whl", hash = "sha256:57cc1786cfd6bd118220a92ede80270132aa353647684efa385a74244a41e3b1"},
- {file = "scikit_learn-1.5.2-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:e9a702e2de732bbb20d3bad29ebd77fc05a6b427dc49964300340e4c9328b3f5"},
- {file = "scikit_learn-1.5.2-cp313-cp313-macosx_12_0_arm64.whl", hash = "sha256:b0768ad641981f5d3a198430a1d31c3e044ed2e8a6f22166b4d546a5116d7908"},
- {file = "scikit_learn-1.5.2-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:178ddd0a5cb0044464fc1bfc4cca5b1833bfc7bb022d70b05db8530da4bb3dd3"},
- {file = "scikit_learn-1.5.2-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:f7284ade780084d94505632241bf78c44ab3b6f1e8ccab3d2af58e0e950f9c12"},
- {file = "scikit_learn-1.5.2-cp313-cp313-win_amd64.whl", hash = "sha256:b7b0f9a0b1040830d38c39b91b3a44e1b643f4b36e36567b80b7c6bd2202a27f"},
- {file = "scikit_learn-1.5.2-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:757c7d514ddb00ae249832fe87100d9c73c6ea91423802872d9e74970a0e40b9"},
- {file = "scikit_learn-1.5.2-cp39-cp39-macosx_12_0_arm64.whl", hash = "sha256:52788f48b5d8bca5c0736c175fa6bdaab2ef00a8f536cda698db61bd89c551c1"},
- {file = "scikit_learn-1.5.2-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:643964678f4b5fbdc95cbf8aec638acc7aa70f5f79ee2cdad1eec3df4ba6ead8"},
- {file = "scikit_learn-1.5.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:ca64b3089a6d9b9363cd3546f8978229dcbb737aceb2c12144ee3f70f95684b7"},
- {file = "scikit_learn-1.5.2-cp39-cp39-win_amd64.whl", hash = "sha256:3bed4909ba187aca80580fe2ef370d9180dcf18e621a27c4cf2ef10d279a7efe"},
- {file = "scikit_learn-1.5.2.tar.gz", hash = "sha256:b4237ed7b3fdd0a4882792e68ef2545d5baa50aca3bb45aa7df468138ad8f94d"},
-]
-
-[package.dependencies]
-joblib = ">=1.2.0"
-numpy = ">=1.19.5"
-scipy = ">=1.6.0"
-threadpoolctl = ">=3.1.0"
-
-[package.extras]
-benchmark = ["matplotlib (>=3.3.4)", "memory_profiler (>=0.57.0)", "pandas (>=1.1.5)"]
-build = ["cython (>=3.0.10)", "meson-python (>=0.16.0)", "numpy (>=1.19.5)", "scipy (>=1.6.0)"]
-docs = ["Pillow (>=7.1.2)", "matplotlib (>=3.3.4)", "memory_profiler (>=0.57.0)", "numpydoc (>=1.2.0)", "pandas (>=1.1.5)", "plotly (>=5.14.0)", "polars (>=0.20.30)", "pooch (>=1.6.0)", "pydata-sphinx-theme (>=0.15.3)", "scikit-image (>=0.17.2)", "seaborn (>=0.9.0)", "sphinx (>=7.3.7)", "sphinx-copybutton (>=0.5.2)", "sphinx-design (>=0.5.0)", "sphinx-design (>=0.6.0)", "sphinx-gallery (>=0.16.0)", "sphinx-prompt (>=1.4.0)", "sphinx-remove-toctrees (>=1.0.0.post1)", "sphinxcontrib-sass (>=0.3.4)", "sphinxext-opengraph (>=0.9.1)"]
-examples = ["matplotlib (>=3.3.4)", "pandas (>=1.1.5)", "plotly (>=5.14.0)", "pooch (>=1.6.0)", "scikit-image (>=0.17.2)", "seaborn (>=0.9.0)"]
-install = ["joblib (>=1.2.0)", "numpy (>=1.19.5)", "scipy (>=1.6.0)", "threadpoolctl (>=3.1.0)"]
-maintenance = ["conda-lock (==2.5.6)"]
-tests = ["black (>=24.3.0)", "matplotlib (>=3.3.4)", "mypy (>=1.9)", "numpydoc (>=1.2.0)", "pandas (>=1.1.5)", "polars (>=0.20.30)", "pooch (>=1.6.0)", "pyamg (>=4.0.0)", "pyarrow (>=12.0.0)", "pytest (>=7.1.2)", "pytest-cov (>=2.9.0)", "ruff (>=0.2.1)", "scikit-image (>=0.17.2)"]
-
-[[package]]
-name = "scipy"
-version = "1.14.1"
-description = "Fundamental algorithms for scientific computing in Python"
-optional = false
-python-versions = ">=3.10"
-files = [
- {file = "scipy-1.14.1-cp310-cp310-macosx_10_13_x86_64.whl", hash = "sha256:b28d2ca4add7ac16ae8bb6632a3c86e4b9e4d52d3e34267f6e1b0c1f8d87e389"},
- {file = "scipy-1.14.1-cp310-cp310-macosx_12_0_arm64.whl", hash = "sha256:d0d2821003174de06b69e58cef2316a6622b60ee613121199cb2852a873f8cf3"},
- {file = "scipy-1.14.1-cp310-cp310-macosx_14_0_arm64.whl", hash = "sha256:8bddf15838ba768bb5f5083c1ea012d64c9a444e16192762bd858f1e126196d0"},
- {file = "scipy-1.14.1-cp310-cp310-macosx_14_0_x86_64.whl", hash = "sha256:97c5dddd5932bd2a1a31c927ba5e1463a53b87ca96b5c9bdf5dfd6096e27efc3"},
- {file = "scipy-1.14.1-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:2ff0a7e01e422c15739ecd64432743cf7aae2b03f3084288f399affcefe5222d"},
- {file = "scipy-1.14.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:8e32dced201274bf96899e6491d9ba3e9a5f6b336708656466ad0522d8528f69"},
- {file = "scipy-1.14.1-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:8426251ad1e4ad903a4514712d2fa8fdd5382c978010d1c6f5f37ef286a713ad"},
- {file = "scipy-1.14.1-cp310-cp310-win_amd64.whl", hash = "sha256:a49f6ed96f83966f576b33a44257d869756df6cf1ef4934f59dd58b25e0327e5"},
- {file = "scipy-1.14.1-cp311-cp311-macosx_10_13_x86_64.whl", hash = "sha256:2da0469a4ef0ecd3693761acbdc20f2fdeafb69e6819cc081308cc978153c675"},
- {file = "scipy-1.14.1-cp311-cp311-macosx_12_0_arm64.whl", hash = "sha256:c0ee987efa6737242745f347835da2cc5bb9f1b42996a4d97d5c7ff7928cb6f2"},
- {file = "scipy-1.14.1-cp311-cp311-macosx_14_0_arm64.whl", hash = "sha256:3a1b111fac6baec1c1d92f27e76511c9e7218f1695d61b59e05e0fe04dc59617"},
- {file = "scipy-1.14.1-cp311-cp311-macosx_14_0_x86_64.whl", hash = "sha256:8475230e55549ab3f207bff11ebfc91c805dc3463ef62eda3ccf593254524ce8"},
- {file = "scipy-1.14.1-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:278266012eb69f4a720827bdd2dc54b2271c97d84255b2faaa8f161a158c3b37"},
- {file = "scipy-1.14.1-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:fef8c87f8abfb884dac04e97824b61299880c43f4ce675dd2cbeadd3c9b466d2"},
- {file = "scipy-1.14.1-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:b05d43735bb2f07d689f56f7b474788a13ed8adc484a85aa65c0fd931cf9ccd2"},
- {file = "scipy-1.14.1-cp311-cp311-win_amd64.whl", hash = "sha256:716e389b694c4bb564b4fc0c51bc84d381735e0d39d3f26ec1af2556ec6aad94"},
- {file = "scipy-1.14.1-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:631f07b3734d34aced009aaf6fedfd0eb3498a97e581c3b1e5f14a04164a456d"},
- {file = "scipy-1.14.1-cp312-cp312-macosx_12_0_arm64.whl", hash = "sha256:af29a935803cc707ab2ed7791c44288a682f9c8107bc00f0eccc4f92c08d6e07"},
- {file = "scipy-1.14.1-cp312-cp312-macosx_14_0_arm64.whl", hash = "sha256:2843f2d527d9eebec9a43e6b406fb7266f3af25a751aa91d62ff416f54170bc5"},
- {file = "scipy-1.14.1-cp312-cp312-macosx_14_0_x86_64.whl", hash = "sha256:eb58ca0abd96911932f688528977858681a59d61a7ce908ffd355957f7025cfc"},
- {file = "scipy-1.14.1-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:30ac8812c1d2aab7131a79ba62933a2a76f582d5dbbc695192453dae67ad6310"},
- {file = "scipy-1.14.1-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:8f9ea80f2e65bdaa0b7627fb00cbeb2daf163caa015e59b7516395fe3bd1e066"},
- {file = "scipy-1.14.1-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:edaf02b82cd7639db00dbff629995ef185c8df4c3ffa71a5562a595765a06ce1"},
- {file = "scipy-1.14.1-cp312-cp312-win_amd64.whl", hash = "sha256:2ff38e22128e6c03ff73b6bb0f85f897d2362f8c052e3b8ad00532198fbdae3f"},
- {file = "scipy-1.14.1-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:1729560c906963fc8389f6aac023739ff3983e727b1a4d87696b7bf108316a79"},
- {file = "scipy-1.14.1-cp313-cp313-macosx_12_0_arm64.whl", hash = "sha256:4079b90df244709e675cdc8b93bfd8a395d59af40b72e339c2287c91860deb8e"},
- {file = "scipy-1.14.1-cp313-cp313-macosx_14_0_arm64.whl", hash = "sha256:e0cf28db0f24a38b2a0ca33a85a54852586e43cf6fd876365c86e0657cfe7d73"},
- {file = "scipy-1.14.1-cp313-cp313-macosx_14_0_x86_64.whl", hash = "sha256:0c2f95de3b04e26f5f3ad5bb05e74ba7f68b837133a4492414b3afd79dfe540e"},
- {file = "scipy-1.14.1-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:b99722ea48b7ea25e8e015e8341ae74624f72e5f21fc2abd45f3a93266de4c5d"},
- {file = "scipy-1.14.1-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:5149e3fd2d686e42144a093b206aef01932a0059c2a33ddfa67f5f035bdfe13e"},
- {file = "scipy-1.14.1-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:e4f5a7c49323533f9103d4dacf4e4f07078f360743dec7f7596949149efeec06"},
- {file = "scipy-1.14.1-cp313-cp313-win_amd64.whl", hash = "sha256:baff393942b550823bfce952bb62270ee17504d02a1801d7fd0719534dfb9c84"},
- {file = "scipy-1.14.1.tar.gz", hash = "sha256:5a275584e726026a5699459aa72f828a610821006228e841b94275c4a7c08417"},
-]
-
-[package.dependencies]
-numpy = ">=1.23.5,<2.3"
-
-[package.extras]
-dev = ["cython-lint (>=0.12.2)", "doit (>=0.36.0)", "mypy (==1.10.0)", "pycodestyle", "pydevtool", "rich-click", "ruff (>=0.0.292)", "types-psutil", "typing_extensions"]
-doc = ["jupyterlite-pyodide-kernel", "jupyterlite-sphinx (>=0.13.1)", "jupytext", "matplotlib (>=3.5)", "myst-nb", "numpydoc", "pooch", "pydata-sphinx-theme (>=0.15.2)", "sphinx (>=5.0.0,<=7.3.7)", "sphinx-design (>=0.4.0)"]
-test = ["Cython", "array-api-strict (>=2.0)", "asv", "gmpy2", "hypothesis (>=6.30)", "meson", "mpmath", "ninja", "pooch", "pytest", "pytest-cov", "pytest-timeout", "pytest-xdist", "scikit-umfpack", "threadpoolctl"]
-
[[package]]
name = "seaborn"
version = "0.13.2"
@@ -6844,25 +6385,6 @@ mpmath = ">=1.1.0,<1.4"
[package.extras]
dev = ["hypothesis (>=6.70.0)", "pytest (>=7.1.0)"]
-[[package]]
-name = "tabledata"
-version = "1.3.3"
-description = "tabledata is a Python library to represent tabular data. Used for pytablewriter/pytablereader/SimpleSQLite/etc."
-optional = false
-python-versions = ">=3.7"
-files = [
- {file = "tabledata-1.3.3-py3-none-any.whl", hash = "sha256:4abad1c996d8607e23b045b44dc0c5f061668f3c37585302c5f6c84c93a89962"},
- {file = "tabledata-1.3.3.tar.gz", hash = "sha256:c90daaba9a408e4397934b3ff2f6c06797d5289676420bf520c741ad43e6ff91"},
-]
-
-[package.dependencies]
-DataProperty = ">=1.0.1,<2"
-typepy = ">=1.2.0,<2"
-
-[package.extras]
-logging = ["loguru (>=0.4.1,<1)"]
-test = ["pytablewriter (>=0.46)", "pytest"]
-
[[package]]
name = "tabulate"
version = "0.9.0"
@@ -6888,34 +6410,6 @@ files = [
{file = "tblib-3.0.0.tar.gz", hash = "sha256:93622790a0a29e04f0346458face1e144dc4d32f493714c6c3dff82a4adb77e6"},
]
-[[package]]
-name = "tcolorpy"
-version = "0.1.6"
-description = "tcolopy is a Python library to apply true color for terminal text."
-optional = false
-python-versions = ">=3.7"
-files = [
- {file = "tcolorpy-0.1.6-py3-none-any.whl", hash = "sha256:8c15cb3167f30b0a433d72297e9d68667c825bd9e2af41c8dd7dfbd3d7f7e207"},
- {file = "tcolorpy-0.1.6.tar.gz", hash = "sha256:8cea0bf5f8cf03f77528a9acfbf312df935573892ba5ea3b2516e61fa54de9a5"},
-]
-
-[package.extras]
-test = ["pytest (>=6.0.1)", "pytest-md-report (>=0.5)"]
-
-[[package]]
-name = "termcolor"
-version = "2.3.0"
-description = "ANSI color formatting for output in terminal"
-optional = false
-python-versions = ">=3.7"
-files = [
- {file = "termcolor-2.3.0-py3-none-any.whl", hash = "sha256:3afb05607b89aed0ffe25202399ee0867ad4d3cb4180d98aaf8eefa6a5f7d475"},
- {file = "termcolor-2.3.0.tar.gz", hash = "sha256:b5b08f68937f138fe92f6c089b99f1e2da0ae56c52b78bf7075fd95420fd9a5a"},
-]
-
-[package.extras]
-tests = ["pytest", "pytest-cov"]
-
[[package]]
name = "terminado"
version = "0.18.1"
@@ -7022,17 +6516,6 @@ mxnet = ["mxnet (>=1.5.1,<1.6.0)"]
tensorflow = ["tensorflow (>=2.0.0,<2.6.0)"]
torch = ["torch (>=1.6.0)"]
-[[package]]
-name = "threadpoolctl"
-version = "3.5.0"
-description = "threadpoolctl"
-optional = false
-python-versions = ">=3.8"
-files = [
- {file = "threadpoolctl-3.5.0-py3-none-any.whl", hash = "sha256:56c1e26c150397e58c4926da8eeee87533b1e32bef131bd4bf6a2f45f3185467"},
- {file = "threadpoolctl-3.5.0.tar.gz", hash = "sha256:082433502dd922bf738de0d8bcc4fdcbf0979ff44c42bd40f5af8a282f6fa107"},
-]
-
[[package]]
name = "tiktoken"
version = "0.7.0"
@@ -7266,31 +6749,28 @@ files = [
[[package]]
name = "torch"
-version = "2.4.1"
+version = "2.5.1"
description = "Tensors and Dynamic neural networks in Python with strong GPU acceleration"
optional = false
python-versions = ">=3.8.0"
files = [
- {file = "torch-2.4.1-cp310-cp310-manylinux1_x86_64.whl", hash = "sha256:362f82e23a4cd46341daabb76fba08f04cd646df9bfaf5da50af97cb60ca4971"},
- {file = "torch-2.4.1-cp310-cp310-manylinux2014_aarch64.whl", hash = "sha256:e8ac1985c3ff0f60d85b991954cfc2cc25f79c84545aead422763148ed2759e3"},
- {file = "torch-2.4.1-cp310-cp310-win_amd64.whl", hash = "sha256:91e326e2ccfb1496e3bee58f70ef605aeb27bd26be07ba64f37dcaac3d070ada"},
- {file = "torch-2.4.1-cp310-none-macosx_11_0_arm64.whl", hash = "sha256:d36a8ef100f5bff3e9c3cea934b9e0d7ea277cb8210c7152d34a9a6c5830eadd"},
- {file = "torch-2.4.1-cp311-cp311-manylinux1_x86_64.whl", hash = "sha256:0b5f88afdfa05a335d80351e3cea57d38e578c8689f751d35e0ff36bce872113"},
- {file = "torch-2.4.1-cp311-cp311-manylinux2014_aarch64.whl", hash = "sha256:ef503165f2341942bfdf2bd520152f19540d0c0e34961232f134dc59ad435be8"},
- {file = "torch-2.4.1-cp311-cp311-win_amd64.whl", hash = "sha256:092e7c2280c860eff762ac08c4bdcd53d701677851670695e0c22d6d345b269c"},
- {file = "torch-2.4.1-cp311-none-macosx_11_0_arm64.whl", hash = "sha256:ddddbd8b066e743934a4200b3d54267a46db02106876d21cf31f7da7a96f98ea"},
- {file = "torch-2.4.1-cp312-cp312-manylinux1_x86_64.whl", hash = "sha256:fdc4fe11db3eb93c1115d3e973a27ac7c1a8318af8934ffa36b0370efe28e042"},
- {file = "torch-2.4.1-cp312-cp312-manylinux2014_aarch64.whl", hash = "sha256:18835374f599207a9e82c262153c20ddf42ea49bc76b6eadad8e5f49729f6e4d"},
- {file = "torch-2.4.1-cp312-cp312-win_amd64.whl", hash = "sha256:ebea70ff30544fc021d441ce6b219a88b67524f01170b1c538d7d3ebb5e7f56c"},
- {file = "torch-2.4.1-cp312-none-macosx_11_0_arm64.whl", hash = "sha256:72b484d5b6cec1a735bf3fa5a1c4883d01748698c5e9cfdbeb4ffab7c7987e0d"},
- {file = "torch-2.4.1-cp38-cp38-manylinux1_x86_64.whl", hash = "sha256:c99e1db4bf0c5347107845d715b4aa1097e601bdc36343d758963055e9599d93"},
- {file = "torch-2.4.1-cp38-cp38-manylinux2014_aarch64.whl", hash = "sha256:b57f07e92858db78c5b72857b4f0b33a65b00dc5d68e7948a8494b0314efb880"},
- {file = "torch-2.4.1-cp38-cp38-win_amd64.whl", hash = "sha256:f18197f3f7c15cde2115892b64f17c80dbf01ed72b008020e7da339902742cf6"},
- {file = "torch-2.4.1-cp38-none-macosx_11_0_arm64.whl", hash = "sha256:5fc1d4d7ed265ef853579caf272686d1ed87cebdcd04f2a498f800ffc53dab71"},
- {file = "torch-2.4.1-cp39-cp39-manylinux1_x86_64.whl", hash = "sha256:40f6d3fe3bae74efcf08cb7f8295eaddd8a838ce89e9d26929d4edd6d5e4329d"},
- {file = "torch-2.4.1-cp39-cp39-manylinux2014_aarch64.whl", hash = "sha256:c9299c16c9743001ecef515536ac45900247f4338ecdf70746f2461f9e4831db"},
- {file = "torch-2.4.1-cp39-cp39-win_amd64.whl", hash = "sha256:6bce130f2cd2d52ba4e2c6ada461808de7e5eccbac692525337cfb4c19421846"},
- {file = "torch-2.4.1-cp39-none-macosx_11_0_arm64.whl", hash = "sha256:a38de2803ee6050309aac032676536c3d3b6a9804248537e38e098d0e14817ec"},
+ {file = "torch-2.5.1-cp310-cp310-manylinux1_x86_64.whl", hash = "sha256:71328e1bbe39d213b8721678f9dcac30dfc452a46d586f1d514a6aa0a99d4744"},
+ {file = "torch-2.5.1-cp310-cp310-manylinux2014_aarch64.whl", hash = "sha256:34bfa1a852e5714cbfa17f27c49d8ce35e1b7af5608c4bc6e81392c352dbc601"},
+ {file = "torch-2.5.1-cp310-cp310-win_amd64.whl", hash = "sha256:32a037bd98a241df6c93e4c789b683335da76a2ac142c0973675b715102dc5fa"},
+ {file = "torch-2.5.1-cp310-none-macosx_11_0_arm64.whl", hash = "sha256:23d062bf70776a3d04dbe74db950db2a5245e1ba4f27208a87f0d743b0d06e86"},
+ {file = "torch-2.5.1-cp311-cp311-manylinux1_x86_64.whl", hash = "sha256:de5b7d6740c4b636ef4db92be922f0edc425b65ed78c5076c43c42d362a45457"},
+ {file = "torch-2.5.1-cp311-cp311-manylinux2014_aarch64.whl", hash = "sha256:340ce0432cad0d37f5a31be666896e16788f1adf8ad7be481196b503dad675b9"},
+ {file = "torch-2.5.1-cp311-cp311-win_amd64.whl", hash = "sha256:603c52d2fe06433c18b747d25f5c333f9c1d58615620578c326d66f258686f9a"},
+ {file = "torch-2.5.1-cp311-none-macosx_11_0_arm64.whl", hash = "sha256:31f8c39660962f9ae4eeec995e3049b5492eb7360dd4f07377658ef4d728fa4c"},
+ {file = "torch-2.5.1-cp312-cp312-manylinux1_x86_64.whl", hash = "sha256:ed231a4b3a5952177fafb661213d690a72caaad97d5824dd4fc17ab9e15cec03"},
+ {file = "torch-2.5.1-cp312-cp312-manylinux2014_aarch64.whl", hash = "sha256:3f4b7f10a247e0dcd7ea97dc2d3bfbfc90302ed36d7f3952b0008d0df264e697"},
+ {file = "torch-2.5.1-cp312-cp312-win_amd64.whl", hash = "sha256:73e58e78f7d220917c5dbfad1a40e09df9929d3b95d25e57d9f8558f84c9a11c"},
+ {file = "torch-2.5.1-cp312-none-macosx_11_0_arm64.whl", hash = "sha256:8c712df61101964eb11910a846514011f0b6f5920c55dbf567bff8a34163d5b1"},
+ {file = "torch-2.5.1-cp313-cp313-manylinux1_x86_64.whl", hash = "sha256:9b61edf3b4f6e3b0e0adda8b3960266b9009d02b37555971f4d1c8f7a05afed7"},
+ {file = "torch-2.5.1-cp39-cp39-manylinux1_x86_64.whl", hash = "sha256:1f3b7fb3cf7ab97fae52161423f81be8c6b8afac8d9760823fd623994581e1a3"},
+ {file = "torch-2.5.1-cp39-cp39-manylinux2014_aarch64.whl", hash = "sha256:7974e3dce28b5a21fb554b73e1bc9072c25dde873fa00d54280861e7a009d7dc"},
+ {file = "torch-2.5.1-cp39-cp39-win_amd64.whl", hash = "sha256:46c817d3ea33696ad3b9df5e774dba2257e9a4cd3c4a3afbf92f6bb13ac5ce2d"},
+ {file = "torch-2.5.1-cp39-none-macosx_11_0_arm64.whl", hash = "sha256:8046768b7f6d35b85d101b4b38cba8aa2f3cd51952bc4c06a49580f2ce682291"},
]
[package.dependencies]
@@ -7298,25 +6778,26 @@ filelock = "*"
fsspec = "*"
jinja2 = "*"
networkx = "*"
-nvidia-cublas-cu12 = {version = "12.1.3.1", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
-nvidia-cuda-cupti-cu12 = {version = "12.1.105", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
-nvidia-cuda-nvrtc-cu12 = {version = "12.1.105", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
-nvidia-cuda-runtime-cu12 = {version = "12.1.105", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
+nvidia-cublas-cu12 = {version = "12.4.5.8", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
+nvidia-cuda-cupti-cu12 = {version = "12.4.127", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
+nvidia-cuda-nvrtc-cu12 = {version = "12.4.127", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
+nvidia-cuda-runtime-cu12 = {version = "12.4.127", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
nvidia-cudnn-cu12 = {version = "9.1.0.70", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
-nvidia-cufft-cu12 = {version = "11.0.2.54", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
-nvidia-curand-cu12 = {version = "10.3.2.106", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
-nvidia-cusolver-cu12 = {version = "11.4.5.107", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
-nvidia-cusparse-cu12 = {version = "12.1.0.106", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
-nvidia-nccl-cu12 = {version = "2.20.5", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
-nvidia-nvtx-cu12 = {version = "12.1.105", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
-setuptools = "*"
-sympy = "*"
-triton = {version = "3.0.0", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\" and python_version < \"3.13\""}
+nvidia-cufft-cu12 = {version = "11.2.1.3", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
+nvidia-curand-cu12 = {version = "10.3.5.147", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
+nvidia-cusolver-cu12 = {version = "11.6.1.9", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
+nvidia-cusparse-cu12 = {version = "12.3.1.170", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
+nvidia-nccl-cu12 = {version = "2.21.5", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
+nvidia-nvjitlink-cu12 = {version = "12.4.127", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
+nvidia-nvtx-cu12 = {version = "12.4.127", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\""}
+setuptools = {version = "*", markers = "python_version >= \"3.12\""}
+sympy = {version = "1.13.1", markers = "python_version >= \"3.9\""}
+triton = {version = "3.1.0", markers = "platform_system == \"Linux\" and platform_machine == \"x86_64\" and python_version < \"3.13\""}
typing-extensions = ">=4.8.0"
[package.extras]
opt-einsum = ["opt-einsum (>=3.3)"]
-optree = ["optree (>=0.11.0)"]
+optree = ["optree (>=0.12.0)"]
[[package]]
name = "tornado"
@@ -7445,16 +6926,16 @@ vision = ["Pillow (>=10.0.1,<=15.0)"]
[[package]]
name = "triton"
-version = "3.0.0"
+version = "3.1.0"
description = "A language and compiler for custom Deep Learning operations"
optional = false
python-versions = "*"
files = [
- {file = "triton-3.0.0-1-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:e1efef76935b2febc365bfadf74bcb65a6f959a9872e5bddf44cc9e0adce1e1a"},
- {file = "triton-3.0.0-1-cp311-cp311-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:5ce8520437c602fb633f1324cc3871c47bee3b67acf9756c1a66309b60e3216c"},
- {file = "triton-3.0.0-1-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:34e509deb77f1c067d8640725ef00c5cbfcb2052a1a3cb6a6d343841f92624eb"},
- {file = "triton-3.0.0-1-cp38-cp38-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:bcbf3b1c48af6a28011a5c40a5b3b9b5330530c3827716b5fbf6d7adcc1e53e9"},
- {file = "triton-3.0.0-1-cp39-cp39-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:6e5727202f7078c56f91ff13ad0c1abab14a0e7f2c87e91b12b6f64f3e8ae609"},
+ {file = "triton-3.1.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:6b0dd10a925263abbe9fa37dcde67a5e9b2383fc269fdf59f5657cac38c5d1d8"},
+ {file = "triton-3.1.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:0f34f6e7885d1bf0eaaf7ba875a5f0ce6f3c13ba98f9503651c1e6dc6757ed5c"},
+ {file = "triton-3.1.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:c8182f42fd8080a7d39d666814fa36c5e30cc00ea7eeeb1a2983dbb4c99a0fdc"},
+ {file = "triton-3.1.0-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:6dadaca7fc24de34e180271b5cf864c16755702e9f63a16f62df714a8099126a"},
+ {file = "triton-3.1.0-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:aafa9a20cd0d9fee523cd4504aa7131807a864cd77dcf6efe7e981f18b8c6c11"},
]
[package.dependencies]
@@ -7465,27 +6946,6 @@ build = ["cmake (>=3.20)", "lit"]
tests = ["autopep8", "flake8", "isort", "llnl-hatchet", "numpy", "pytest", "scipy (>=1.7.1)"]
tutorials = ["matplotlib", "pandas", "tabulate"]
-[[package]]
-name = "typepy"
-version = "1.3.2"
-description = "typepy is a Python library for variable type checker/validator/converter at a run time."
-optional = false
-python-versions = ">=3.7"
-files = [
- {file = "typepy-1.3.2-py3-none-any.whl", hash = "sha256:d5d1022a424132622993800f1d2cd16cfdb691ac4e3b9c325f0fcb37799db1ae"},
- {file = "typepy-1.3.2.tar.gz", hash = "sha256:b69fd48b9f50cdb3809906eef36b855b3134ff66c8893a4f8580abddb0b39517"},
-]
-
-[package.dependencies]
-mbstrdecoder = ">=1.0.0,<2"
-packaging = {version = "*", optional = true, markers = "extra == \"datetime\""}
-python-dateutil = {version = ">=2.8.0,<3.0.0", optional = true, markers = "extra == \"datetime\""}
-pytz = {version = ">=2018.9", optional = true, markers = "extra == \"datetime\""}
-
-[package.extras]
-datetime = ["packaging", "python-dateutil (>=2.8.0,<3.0.0)", "pytz (>=2018.9)"]
-test = ["packaging", "pytest (>=6.0.1)", "python-dateutil (>=2.8.0,<3.0.0)", "pytz (>=2018.9)", "tcolorpy"]
-
[[package]]
name = "typer"
version = "0.9.4"
@@ -7982,4 +7442,4 @@ type = ["pytest-mypy"]
[metadata]
lock-version = "2.0"
python-versions = ">=3.11,<3.13"
-content-hash = "a4c8ab84bda79bbf8fb3f92fc685ea17b00105cc729a39973e0fedcf81462441"
+content-hash = "d9bc55f679878efa255457b264b6c142d779c1f2d9f1d0233ed562eaed4c195a"
diff --git a/pyproject.toml b/pyproject.toml
index ff4574f..219e756 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -30,7 +30,6 @@ evaluate = "^0.4.3"
absl-py = "^2.1.0"
rouge-score = "^0.1.2"
matplotlib = "^3.9.2"
-lighteval = {extras = ["accelerate"], version = "^0.6.2"}
outlines = "^0.1.7"
datasets = "^3.1.0"
text-generation = "^0.7.0"
@@ -43,6 +42,9 @@ mistralai = "^1.2.5"
llm-guard = "^0.3.15"
pygments = "^2.18.0"
llama-cpp-python = "^0.3.5"
+torch = "^2.5.1"
+gguf = "^0.13.0"
+accelerate = "^1.2.1"
[build-system]
diff --git a/tamingllms/_build/.doctrees/environment.pickle b/tamingllms/_build/.doctrees/environment.pickle
index db81a7e..42959ee 100644
Binary files a/tamingllms/_build/.doctrees/environment.pickle and b/tamingllms/_build/.doctrees/environment.pickle differ
diff --git a/tamingllms/_build/.doctrees/markdown/preface.doctree b/tamingllms/_build/.doctrees/markdown/preface.doctree
index d4a62be..9ba8a92 100644
Binary files a/tamingllms/_build/.doctrees/markdown/preface.doctree and b/tamingllms/_build/.doctrees/markdown/preface.doctree differ
diff --git a/tamingllms/_build/.doctrees/markdown/toc.doctree b/tamingllms/_build/.doctrees/markdown/toc.doctree
index 376e410..50ce0d7 100644
Binary files a/tamingllms/_build/.doctrees/markdown/toc.doctree and b/tamingllms/_build/.doctrees/markdown/toc.doctree differ
diff --git a/tamingllms/_build/.doctrees/notebooks/alignment.doctree b/tamingllms/_build/.doctrees/notebooks/alignment.doctree
index d7908d9..15d23b1 100644
Binary files a/tamingllms/_build/.doctrees/notebooks/alignment.doctree and b/tamingllms/_build/.doctrees/notebooks/alignment.doctree differ
diff --git a/tamingllms/_build/.doctrees/notebooks/cost.doctree b/tamingllms/_build/.doctrees/notebooks/cost.doctree
new file mode 100644
index 0000000..181b54c
Binary files /dev/null and b/tamingllms/_build/.doctrees/notebooks/cost.doctree differ
diff --git a/tamingllms/_build/.doctrees/notebooks/evals.doctree b/tamingllms/_build/.doctrees/notebooks/evals.doctree
index 5593861..b6a64e7 100644
Binary files a/tamingllms/_build/.doctrees/notebooks/evals.doctree and b/tamingllms/_build/.doctrees/notebooks/evals.doctree differ
diff --git a/tamingllms/_build/.doctrees/notebooks/local.doctree b/tamingllms/_build/.doctrees/notebooks/local.doctree
index 6014cf8..08986fc 100644
Binary files a/tamingllms/_build/.doctrees/notebooks/local.doctree and b/tamingllms/_build/.doctrees/notebooks/local.doctree differ
diff --git a/tamingllms/_build/.doctrees/notebooks/safety.doctree b/tamingllms/_build/.doctrees/notebooks/safety.doctree
index f0ec824..6199fc9 100644
Binary files a/tamingllms/_build/.doctrees/notebooks/safety.doctree and b/tamingllms/_build/.doctrees/notebooks/safety.doctree differ
diff --git a/tamingllms/_build/.doctrees/notebooks/structured_output.doctree b/tamingllms/_build/.doctrees/notebooks/structured_output.doctree
index 053c0bc..f5c9be8 100644
Binary files a/tamingllms/_build/.doctrees/notebooks/structured_output.doctree and b/tamingllms/_build/.doctrees/notebooks/structured_output.doctree differ
diff --git a/tamingllms/_build/html/_images/bitnet.png b/tamingllms/_build/html/_images/bitnet.png
new file mode 100644
index 0000000..5d0d74e
Binary files /dev/null and b/tamingllms/_build/html/_images/bitnet.png differ
diff --git a/tamingllms/_build/html/_images/llmflation.png b/tamingllms/_build/html/_images/llmflation.png
new file mode 100644
index 0000000..5061149
Binary files /dev/null and b/tamingllms/_build/html/_images/llmflation.png differ
diff --git a/tamingllms/_build/html/_images/quantized.png b/tamingllms/_build/html/_images/quantized.png
new file mode 100644
index 0000000..2dc6d44
Binary files /dev/null and b/tamingllms/_build/html/_images/quantized.png differ
diff --git a/tamingllms/_build/html/_sources/markdown/toc.md b/tamingllms/_build/html/_sources/markdown/toc.md
index 83c6895..1578091 100644
--- a/tamingllms/_build/html/_sources/markdown/toc.md
+++ b/tamingllms/_build/html/_sources/markdown/toc.md
@@ -32,7 +32,7 @@ Abstract: *The current discourse around Large Language Models (LLMs) tends to fo
## [Chapter 6: Local LLMs in Practice](https://www.souzatharsis.com/tamingLLMs/notebooks/local.html)
-## Chapter 7: The Cost Factor
+## Chapter 7: The Falling Cost Paradox
## Chapter 8: Frontiers
diff --git a/tamingllms/_build/html/_sources/notebooks/alignment.ipynb b/tamingllms/_build/html/_sources/notebooks/alignment.ipynb
index 552ad7f..9eeeffa 100644
--- a/tamingllms/_build/html/_sources/notebooks/alignment.ipynb
+++ b/tamingllms/_build/html/_sources/notebooks/alignment.ipynb
@@ -2537,7 +2537,7 @@
"source": [
"## Discussion and Conclusions\n",
"\n",
- "LLMs are complex systems and alignment is a challenging problem. In this chapter, we discussed how post-training alignment techniques can be used to align a language model to human preferences. In the case study, we demonstrated how to use DPO to align a language model to a user-provider policy further automating the process via synthetic data generation and LLM-as-judge evaluation. Our approach does serve as a proof of concept, however, several considerations should be taken into account when using this methodology in practice.\n",
+ "LLMs are complex systems and alignment is a challenging problem. In this chapter, we discussed how post-training techniques can be used to align a language model to human preferences. In the case study, we demonstrated how to use DPO to align a language model to a user-provider policy further automating the process via synthetic data generation and LLM-as-judge evaluation. Our approach serves as a proof of concept and several considerations should be taken into account when using this methodology in practice.\n",
"\n",
"**Synthetic Data Generation**\n",
"\n",
diff --git a/tamingllms/_build/html/_sources/notebooks/cost.ipynb b/tamingllms/_build/html/_sources/notebooks/cost.ipynb
new file mode 100644
index 0000000..45e3ee6
--- /dev/null
+++ b/tamingllms/_build/html/_sources/notebooks/cost.ipynb
@@ -0,0 +1,450 @@
+{
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "(cost)=\n",
+ "# The Falling Cost Paradox\n",
+ "```{epigraph}\n",
+ "It is a confusion of ideas to suppose that the economical use of fuel is equivalent to diminished consumption.
\n",
+ "The very contrary is the truth. \n",
+ "\n",
+ "-- William Stanley Jevons\n",
+ "```\n",
+ "```{contents}\n",
+ "```"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Why Optimization Matters More Than Ever\n",
+ "\n",
+ "According to recent analysis from a16z {cite}`a16z2024llmflation`, the cost of LLM inference is decreasing by approximately 10x every year - a rate that outpaces even Moore's Law in the PC revolution or Edholm's Law during the bandwidth explosion of the dot-com era.\n",
+ "\n",
+ "```{figure} ../_static/cost/llmflation.png\n",
+ "---\n",
+ "name: llmflation\n",
+ "alt: LLMflation\n",
+ "scale: 70%\n",
+ "align: center\n",
+ "---\n",
+ "LLMflation {cite}`a16z2024llmflation`: The cost of LLM inference is decreasing by approximately 10x every year.\n",
+ "```\n",
+ "\n",
+ "A model achieving an MMLU score of 42 that cost \\$60 per million tokens in late 2021 can now be run for just \\$0.06 per million tokens. For higher-capability models scoring 83 on MMLU, prices have fallen by a factor of 62 since GPT-4's introduction in March 2023. \n",
+ "\n",
+ "This dramatic decline stems from multiple compounding factors including:\n",
+ "\n",
+ "- Improved GPU efficiency through architectural advances and Moore's Law\n",
+ "- Model quantization progress, moving from 16-bit to 4-bit or lower precision\n",
+ "- Software optimizations reducing compute and memory bandwidth requirements\n",
+ "- Emergence of smaller yet similarly capable models\n",
+ "- Better instruction tuning techniques like RLHF and DPO\n",
+ "- Competition from open-source models and low-cost providers\n",
+ "\n",
+ "This trend raises a critical question: If LLM costs are plummeting so rapidly, why should businesses and developers invest precious time and resources in optimizing their LLM usage? Wouldn't it make more sense to simply wait for the next wave of cost improvements rather than optimize today? In two words: **Jevons Paradox**. \n",
+ "\n",
+ "The Jevons Paradox was first observed by English economist William Stanley Jevons in 1865. Studying coal consumption during the Industrial Revolution, Jevons made a counterintuitive discovery: as steam engines became more efficient and coal use became more economical, total coal consumption increased rather than decreased driving the (Industrial Revolution) and the total spending up.\n",
+ "\n",
+ "This pattern has repeated throughout technological history:\n",
+ "\n",
+ "- Computing Power: As cost per computation plummeted, we didn't spend less on computing, instead we found new creative uses for computers, from smartphones to cloud servers\n",
+ "- Network Bandwidth: As data transmission got cheaper, we shifted from text messaging to HD video streaming and real-time gaming\n",
+ "- Data Storage: As cost per gigabyte fell, we moved from storing documents to hosting entire media libraries and training massive AI models\n",
+ "\n",
+ "One could argue that LLMs and Generative AI more broadly are following a similar trajectory. As costs decline, we're seeing the emergence of new applications:\n",
+ "- Embedding AI capabilities into every application and workflow\n",
+ "- Real-time analysis of audio transcripts and conversations\n",
+ "- Running AI models directly on edge devices and smartphones\n",
+ "- Multimodal applications combining text, images, audio and video \n",
+ "\n",
+ "In this environment of rapidly falling costs but potential for exponential growth in usage, optimizing LLM costs becomes more, not less, important. Here's why:\n",
+ "\n",
+ "**A) Scale Magnifies Everything**. When operating at billions of tokens per day, even small inefficiencies have major effects:\n",
+ "- A single digit improvement in efficiency can save millions of dollars annually at scale\n",
+ "- Every 100 milliseconds of latency is about 8% difference in engagement rates (30% on mobile) [^groklatency]\n",
+ "[^groklatency]: Quote from Jonathan Ross, CEO of Groq, a company that specializes in AI Inference services.\n",
+ "\n",
+ "**B) Tiered Pricing Persists**. While average costs are declining, the market maintains a tiered structure:\n",
+ "- Different models offer varying price-performance tradeoffs\n",
+ "- ChatGPT Pro at \\$200 per month breaks the price drop trend perhaps triggering a new wave of premium models\n",
+ "- Cost optimization is still required to select the right model for each specific use case\n",
+ "\n",
+ "**C) Competition Drives Innovation**. Companies that master LLM efficiency gain significant advantages:\n",
+ "- Ability to offer more competitive pricing\n",
+ "- Capacity to handle larger scale operations\n",
+ "- Resources to invest in product improvement\n",
+ "\n",
+ "**D) Performance and Cost Are Linked**. Cost optimization often yields performance benefits:\n",
+ "- Resource efficiency enables handling larger user loads\n",
+ "- More efficiency and reduced latency leads to improved user experience\n",
+ "\n",
+ "In this environment, companies that master efficient LLM usage while exploiting new capabilities opened up by falling costs will be best positioned to innovate and scale. This dual focus - efficiency and innovation - will likely characterize successful AI companies in the years ahead.\n",
+ "\n",
+ "Motivated by this insight, in the next sections we will dive into the factors that drive LLM cost decay and how to optimize LLM usage in practical applications. The discussion will explore key optimization areas including inference optimization through techniques like Flash Attention and cached prompts, model compression via quantization and distillation, and practical implementation patterns such as response caching, batch processing, and early stopping - all aimed at achieving efficient usage and cost reductions while maintaining model performance and reliability.\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Right-Sizing LLMs: A Strategic Approach\n",
+ "\n",
+ "Before implementing cost optimization strategies for LLMs, organizations must develop a comprehensive understanding of their own requirements and constraints. This systematic approach prevents both over-engineering and under-provisioning, leading to more efficient and cost-effective implementations.\n",
+ "\n",
+ "In this section, we define key performance and cost related metrics that will guide our discussion. Then we propose a set of requirements practitioners should consider before we dive into cost optimization techniques.\n",
+ "\n",
+ "\n",
+ "### Metrics"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "### Requirements\n",
+ "\n",
+ "#### Business Requirements\n",
+ "\n",
+ "First, one needs to define the problem to be solved and to what extent it is worth to be solved. Use case requirements form the foundation of any LLM implementation project. A clear definition of the specific business problema and task to be accomplished must be established upfront, along with concrete performance metrics covering accuracy, latency and throughput. This should be accompanied by well-defined cost-per-transaction targets, clear ROI expectations, and a strategic allocation of budgets across different use cases to ensure resources are optimally distributed.\n",
+ "\n",
+ "Budget and ROI considerations are critical for ensuring the long-term viability of LLM implementations. Organizations must establish clear spending limits that align with their financial capabilities while defining realistic cost-per-transaction targets. ROI expectations need to be carefully established through detailed analysis, followed by a strategic allocation of budgets across various use cases based on their business impact and priority.\n",
+ "\n",
+ "Compliance and security requirements cannot be overlooked. This involves a thorough identification of all applicable regulatory requirements and the establishment of robust data handling standards. Organizations must specify comprehensive audit requirements to maintain transparency and accountability, while implementing appropriate security controls to protect sensitive data and system access.\n",
+ "\n",
+ "Future-proofing considerations help ensure the longevity and adaptability of LLM implementations. This requires careful planning for scale to accommodate future growth, along with the evaluation of multi-model strategies to reduce dependency on single solutions. Organizations should carefully assess vendor lock-in risks and explore open-source alternatives to maintain flexibility and control over their AI infrastructure.\n",
+ "\n",
+ "Chapter {ref}`local` provides a detailed discussion on relevant considerations when {ref}`local-model-selection`.\n",
+ "\n",
+ "#### Performance Requirements\n",
+ "\n",
+ "Accuracy and quality form the foundation of any LLM deployment's performance requirements. At its core, this involves determining the minimum level of accuracy that the model must achieve to be considered successful. This serves as a critical baseline for evaluating model performance and making deployment decisions. Establishing clear evaluation metrics, whether through automated measures or human evaluation processes, provides concrete ways to assess if these thresholds are being met. Continuous monitoring of these accuracy metrics ensures the system maintains its performance over time as usage patterns and data distributions evolve. Chapter {ref}`evals` provides a detailed discussion on how to evaluate the performance of LLM-based applications.\n",
+ "\n",
+ "Latency and throughput requirements are equally crucial for ensuring a positive user experience and system reliability. These specifications define how quickly the system must respond to requests and how many concurrent users it can handle. Response time requirements must be carefully balanced against the computational resources available, while peak load capabilities need to account for usage spikes and growth patterns. The decision between real-time processing for immediate responses versus batch processing for efficiency depends heavily on the use case and user expectations. \n",
+ "\n",
+ "\n",
+ "#### Operational Requirements\n",
+ "\n",
+ "Scale and capacity planning forms the foundation of operational requirements for LLM deployments. This involves a comprehensive analysis of expected system usage and growth patterns to ensure the infrastructure can handle both current and future demands. Organizations must carefully project their daily and monthly API call volumes while calculating the average number of tokens per request to accurately estimate resource needs. Understanding usage patterns, including seasonal variations, enables proper capacity planning. Additionally, developing 12-24 month growth projections helps ensure the infrastructure can scale appropriately as demand increases.\n",
+ "\n",
+ "Reliability and availability requirements are equally critical for maintaining consistent service quality. These specifications define the expected uptime percentage that the system must maintain, typically expressed as a percentage of total operational time. Organizations need to establish clear maintenance windows that minimize disruption to users while ensuring necessary system updates and optimizations can be performed. Comprehensive backup and failover requirements must be specified to ensure business continuity in case of failures. High availability needs should be clearly defined, including redundancy levels and recovery time objectives, to maintain service quality even during unexpected events.\n",
+ "\n",
+ "#### Technical Requirements\n",
+ "\n",
+ "System integration requirements define how the LLM system will interact and communicate with existing infrastructure and applications. This involves carefully mapping all integration points where the LLM system needs to connect with other systems, establishing standardized data formats and interfaces for seamless communication, implementing robust security measures to protect data in transit, and identifying any technical constraints that could impact integration. Getting these integration requirements right is crucial for ensuring the LLM system can function effectively within the broader technical ecosystem.\n",
+ "\n",
+ "Data management requirements address how information will be stored, processed, and maintained within the LLM system. This encompasses determining appropriate storage solutions for maintaining conversation context and history, selecting and configuring vector databases to enable efficient retrieval-augmented generation (RAG), creating comprehensive data retention policies that balance operational needs with resource constraints, and ensuring all data handling practices comply with relevant privacy regulations. Proper data management is essential for both system performance and regulatory compliance, making it a critical consideration in any LLM implementation.\n",
+ "\n",
+ "\n",
+ "This structured approach to requirements analysis enables organizations to:\n",
+ "1. Select appropriate models aligned with specific needs\n",
+ "2. Identify targeted optimization opportunities\n",
+ "3. Scale efficiently while controlling costs\n",
+ "4. Develop realistic resource allocation strategies\n",
+ "\n",
+ "The following sections explore specific optimization techniques, but their implementation should always be guided by these foundational requirements.\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Quantization\n",
+ "\n",
+ "Quantization is a common and relevant technique in making LLMs more efficient and accessible. At a high level, quantization reduces the number of bits used to represent a model's parameters. The most common form of quantization is to represent model's weights at lower precision at post-training phase. It has become a standard technique to generate a series of quantized models given a large pre-trained base model.\n",
+ "\n",
+ "While a standard pre-trained LLM might use 32-bit floating-point (FP32) or 16-bit floating-point (FP16) numbers to store its weights, quantized versions can operate at lower precision levels such as 8, 4 or even 2 bits per parameter, reducing memory footprint without proportional losses in performance, necessarily. For instance, for a model of 30 billion parameters, using FP32 means 4 bytes per weight or 120 GB for the whole weights. If the model is quantized such that weights are represented in 1 byte, the memory needed for the model’s weights decreases to 30 GB, hence potentially fitting into consumer grade hardware. This is done at the cost of precision loss, but the trade-off is often worth it though require careful analysis."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Let's take a look at model weights of a language model (`SmolLM2-135M-Instruct`) that has been quantized to 2-bit and 16-bit precisions. We will use an utility function `load_gguf` from the `taming_utils` package to load model weights of the quantized models directly from Hugging Face."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "from taming_utils import load_gguf\n",
+ "\n",
+ "MODEL_NAME = \"bartowski/SmolLM2-135M-Instruct-GGUF\"\n",
+ "GGUF_FILE_Q2_K = \"SmolLM2-135M-Instruct-Q2_K.gguf\"\n",
+ "GGUF_FILE_F16 = \"SmolLM2-135M-Instruct-F16.gguf\"\n",
+ "\n",
+ "model_q2_k = load_gguf(model_name=MODEL_NAME, \n",
+ " gguf_file=GGUF_FILE_Q2_K)\n",
+ "\n",
+ "model_f16 = load_gguf(model_name=MODEL_NAME, \n",
+ " gguf_file=GGUF_FILE_F16)\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "We extract the MLP weights from the first layer of the model as a proxy."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "mlp_weights_q2_k = model_q2_k.model.layers[0].mlp.gate_proj.weight\n",
+ "mlp_weights_f16 = model_f16.model.layers[0].mlp.gate_proj.weight"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Original weights at 16-bit precision:"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 13,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "Parameter containing:\n",
+ "tensor([[-0.0145, 0.1826, 0.1377, ..., 0.1719, -0.1387, -0.0298],\n",
+ " [-0.1631, 0.0781, -0.2051, ..., -0.2070, -0.0334, 0.2891],\n",
+ " [-0.1768, -0.0488, -0.2393, ..., -0.0396, -0.1348, -0.1533],\n",
+ " ...,\n",
+ " [ 0.0771, 0.0845, -0.0232, ..., 0.0178, -0.1040, -0.0771],\n",
+ " [ 0.1582, 0.1167, -0.0474, ..., 0.0845, 0.0359, -0.2500],\n",
+ " [ 0.0432, 0.0972, 0.0933, ..., 0.2188, 0.0776, 0.0674]],\n",
+ " requires_grad=True)"
+ ]
+ },
+ "execution_count": 13,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "mlp_weights_f16"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Quantized weights at 2-bit precision:"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 10,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "Parameter containing:\n",
+ "tensor([[-0.0028, 0.1852, 0.1396, ..., 0.1506, -0.1635, -0.0043],\n",
+ " [-0.1768, 0.0680, -0.2257, ..., -0.1890, -0.0464, 0.2960],\n",
+ " [-0.1840, -0.0451, -0.2395, ..., -0.0413, -0.1446, -0.1446],\n",
+ " ...,\n",
+ " [ 0.0621, 0.0621, -0.0478, ..., 0.0038, -0.0830, -0.0830],\n",
+ " [ 0.1473, 0.0926, -0.0547, ..., 0.0824, 0.0429, -0.2737],\n",
+ " [ 0.0355, 0.0782, 0.0782, ..., 0.2043, 0.0740, 0.0740]],\n",
+ " requires_grad=True)"
+ ]
+ },
+ "execution_count": 10,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "mlp_weights_q2_k"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "How do they compare? We arrive at a Pearson correlation of 99.7% between the two sets of weights."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 14,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Pearson correlation: 0.9970\n"
+ ]
+ }
+ ],
+ "source": [
+ "import numpy as np\n",
+ "\n",
+ "# Convert tensors to numpy arrays (detach from computation graph if needed)\n",
+ "weights_f16 = mlp_weights_f16.detach().cpu().numpy()\n",
+ "weights_q2_k = mlp_weights_q2_k.detach().cpu().numpy()\n",
+ "\n",
+ "flat_f16 = weights_f16.flatten()\n",
+ "flat_q2_k = weights_q2_k.flatten()\n",
+ "\n",
+ "# Calculate correlation\n",
+ "correlation = np.corrcoef(flat_f16, flat_q2_k)[0,1]\n",
+ "print(f\"Pearson correlation: {correlation:.4f}\")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Quantization is a powerful technique for reducing the memory footprint of LLMs. This can be exemplified by the case of LLaMa 3.3 70B as quantized by {cite}`unsloth2024llama3` [^unsloth]. The model's memory requirements vary significantly based on the quantization level used as demonstrated in {numref}`quantized`.\n",
+ "\n",
+ "[^unsloth]: Unsloth runs a business of making LLMs fine-tuning streamlined. Check them out at [unsloth.ai](https://unsloth.ai).\n",
+ "\n",
+ "```{figure} ../_static/cost/quantized.png\n",
+ "---\n",
+ "name: quantized\n",
+ "alt: Quantized Model Size\n",
+ "scale: 50%\n",
+ "align: center\n",
+ "---\n",
+ "Quantized Model Size: `unsloth/Llama-3.3-70B-Instruct-GGUF`\n",
+ "```\n",
+ "\n",
+ "We observe that the quantization process yields remarkable reductions in model size, demonstrating a clear trade-off between precision and memory requirements. The transition from F16 (141.1 GB) to Q8_0 (75 GB) achieves a dramatic 47% reduction in model size while maintaining relatively high numerical precision. Further quantization levels reveal an interesting pattern of diminishing returns - each step down in precision yields progressively smaller absolute size reductions, though the cumulative effect remains significant. At the extreme end, the Q2_K model (26.4 GB) requires only 19% of the storage space of its F16 counterpart [^quantization-levels].\n",
+ "\n",
+ "[^quantization-levels]: You may have noticed quantization levels have a special notation. Including the bit width in the name of the model but also quantization types (e.g. _K, _0). You can find more information about the quantization levels in {cite}`huggingface2024quantization`.\n",
+ "\n",
+ "This wide spectrum of model sizes enables deployment across diverse hardware environments. The lightweight Q2_K variant opens possibilities for running inference on consumer-grade hardware like high-end laptops or desktop computers. In contrast, the full-precision F16 model demands enterprise-grade computing resources with substantial memory capacity. This flexibility in deployment options makes quantization a powerful tool for democratizing access to large language models while managing computational costs.\n",
+ " \n",
+ "While quantization has proven highly effective, there is a limit to how far it can be pushed - specifically, the 1-bit ceiling. A notable advancement in this space is BitNet {cite}`wang20241bitaiinfra11` which pushes the boundaries of extreme quantization.\n",
+ "\n",
+ "BitNet's implementation, bitnet.cpp, has demonstrated significant performance improvements across both ARM and x86 architectures (see {numref}`bitnet`). When compared to llama.cpp, the framework achieves speedups ranging from 1.37x to 5.07x on ARM processors and 2.37x to 6.17x on x86 systems. These performance gains scale with model size - larger models benefit more substantially from BitNet's optimizations. The efficiency improvements extend beyond raw speed: energy consumption drops by 55-70% on ARM and 71-82% on x86 processors. Perhaps most impressively, bitnet.cpp enables running a 100B parameter BitNet b1.58 model on a single CPU at speeds matching human reading pace (5-7 tokens per second).\n",
+ "\n",
+ "```{figure} ../_static/cost/bitnet.png\n",
+ "---\n",
+ "name: bitnet\n",
+ "alt: BitNet\n",
+ "scale: 30%\n",
+ "align: center\n",
+ "---\n",
+ "BitNet: {cite}`wang20241bitaiinfra11`\n",
+ "```\n",
+ "\n",
+ "The framework's initial release focused on CPU inference optimization, with particular emphasis on 1-bit LLM architectures (BitNet b1.58). While initial testing shows promising results, these findings are specific to the tested models and kernels (its specialized kernels are carefully crafted to exploit the unique characteristics of these extremely quantized models). Further validation is needed before generalizing these results across different architectures and use cases.\n",
+ "\n",
+ "As a relatively recent innovation, 1-bit LLMs represent an exciting frontier in model compression. However, their full potential and limitations require additional research and real-world validation. The technology demonstrates how creative approaches to quantization can continue pushing the boundaries of efficient AI deployment."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Beyond its memory footprint reduction, quantization delivers several compelling advantages: it accelerates computation through faster arithmetic operations and larger batch sizes, reduces costs by enabling deployment on less expensive hardware and making LLMs more accessible to smaller organizations, and improves energy efficiency by lowering memory bandwidth usage and power consumption - particularly beneficial for mobile and edge devices, ultimately contributing to more sustainable AI deployment.\n",
+ "\n",
+ "Each reduction in precision risks performance degradation. Finding optimal quantization schemes remains an active research area. See Case Study on Quantization for Local Models in Chapter {ref}`local` for more details.\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Check-list\n",
+ "\n",
+ "**Planning and Requirements**\n",
+ "- [ ] Start with a clear understanding of your application's needs and the factors that contribute to LLM costs\n",
+ "- [ ] Choose the right model for your task, balancing performance and cost\n",
+ "- [ ] Be aware of the potential challenges and limitations of open-source LLMs and take appropriate measures to mitigate them\n",
+ "\n",
+ "**Model Optimization**\n",
+ "- [ ] Explore model compression and quantization to reduce model size and computational demands\n",
+ "- [ ] Fine-tune pre-trained models on domain-specific data to improve accuracy and efficiency\n",
+ "- [ ] Consider using RAG to enhance performance and reduce reliance on purely generative processes\n",
+ "\n",
+ "**Prompt Engineering**\n",
+ "- [ ] Optimize prompts and utilize prompt engineering techniques to minimize token usage\n",
+ "- [ ] Experiment with different prompting strategies to unlock the full potential of open-source LLMs\n",
+ "\n",
+ "**Infrastructure and Operations**\n",
+ "- [ ] Implement caching and batching strategies to optimize resource utilization\n",
+ "- [ ] Monitor LLM usage patterns and costs to identify areas for optimization\n",
+ "- [ ] Set up observability and logging to track model performance and costs\n",
+ "- [ ] Establish automated testing and evaluation pipelines\n",
+ "\n",
+ "**Cost Management**\n",
+ "- [ ] Track and analyze inference costs across different model variants\n",
+ "- [ ] Implement cost allocation and chargeback mechanisms\n",
+ "- [ ] Set up cost alerts and budgeting controls\n",
+ "- [ ] Regularly review and optimize resource utilization\n",
+ "\n",
+ "## Conclusion"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "[![CC BY-NC-SA 4.0][cc-by-nc-sa-image]][cc-by-nc-sa]\n",
+ "\n",
+ "[cc-by-nc-sa]: http://creativecommons.org/licenses/by-nc-sa/4.0/\n",
+ "[cc-by-nc-sa-image]: https://licensebuttons.net/l/by-nc-sa/4.0/88x31.png\n",
+ "[cc-by-nc-sa-shield]: https://img.shields.io/badge/License-CC-BY--NC--SA-4.0-lightgrey.svg\n",
+ "\n",
+ "```\n",
+ "@misc{tharsistpsouza2024tamingllms,\n",
+ " author = {Tharsis T. P. Souza},\n",
+ " title = {Taming LLMs: A Practical Guide to LLM Pitfalls with Open Source Software},\n",
+ " year = {2024},\n",
+ " chapter = {The Falling Cost Paradox},\n",
+ " journal = {GitHub repository},\n",
+ " url = {https://github.com/souzatharsis/tamingLLMs)\n",
+ "}\n",
+ "```\n",
+ "## References\n",
+ "```{bibliography}\n",
+ ":filter: docname in docnames\n",
+ "```"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": []
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": ".venv",
+ "language": "python",
+ "name": "python3"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.11.11"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}
diff --git a/tamingllms/_build/html/_sources/notebooks/local.ipynb b/tamingllms/_build/html/_sources/notebooks/local.ipynb
index b451331..fde2739 100644
--- a/tamingllms/_build/html/_sources/notebooks/local.ipynb
+++ b/tamingllms/_build/html/_sources/notebooks/local.ipynb
@@ -4,6 +4,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
+ "(local)=\n",
"# Local LLMs in Practice\n",
"```{epigraph}\n",
"Freedom is something that dies unless it's used.\n",
@@ -40,7 +41,8 @@
"cell_type": "markdown",
"metadata": {},
"source": [
- "## Models Considerations\n",
+ "(local-model-selection)=\n",
+ "## Choosing your Model\n",
"\n",
"The landscape of open source LLMs is rapidly evolving, with new models emerging by the day. While proprietary LLMs have garnered significant attention, open source LLMs are gaining traction due to their flexibility, customization options, and cost-effectiveness. \n",
"\n",
@@ -1352,7 +1354,6 @@
"cell_type": "markdown",
"metadata": {},
"source": [
- "## Citation\n",
"[![CC BY-NC-SA 4.0][cc-by-nc-sa-image]][cc-by-nc-sa]\n",
"\n",
"[cc-by-nc-sa]: http://creativecommons.org/licenses/by-nc-sa/4.0/\n",
diff --git a/tamingllms/_build/html/_sources/notebooks/structured_output.ipynb b/tamingllms/_build/html/_sources/notebooks/structured_output.ipynb
index 4bc64db..64359b4 100644
--- a/tamingllms/_build/html/_sources/notebooks/structured_output.ipynb
+++ b/tamingllms/_build/html/_sources/notebooks/structured_output.ipynb
@@ -467,9 +467,18 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 1,
"metadata": {},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+ "/home/tobias/src/tamingLLMs/tamingllms/.venv/lib/python3.11/site-packages/torch/cuda/__init__.py:654: UserWarning: Can't initialize NVML\n",
+ " warnings.warn(\"Can't initialize NVML\")\n"
+ ]
+ }
+ ],
"source": [
"MODEL_NAME = \"HuggingFaceTB/SmolLM2-1.7B-Instruct\"\n",
"PROMPT = \"Is Enzo a good name for a baby?\"\n",
@@ -1384,7 +1393,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
- "version": "3.12.3"
+ "version": "3.11.11"
}
},
"nbformat": 4,
diff --git a/tamingllms/_build/html/_static/cost/bitnet.png b/tamingllms/_build/html/_static/cost/bitnet.png
new file mode 100644
index 0000000..5d0d74e
Binary files /dev/null and b/tamingllms/_build/html/_static/cost/bitnet.png differ
diff --git a/tamingllms/_build/html/_static/cost/llmflation.png b/tamingllms/_build/html/_static/cost/llmflation.png
new file mode 100644
index 0000000..5061149
Binary files /dev/null and b/tamingllms/_build/html/_static/cost/llmflation.png differ
diff --git a/tamingllms/_build/html/_static/cost/quantized.png b/tamingllms/_build/html/_static/cost/quantized.png
new file mode 100644
index 0000000..2dc6d44
Binary files /dev/null and b/tamingllms/_build/html/_static/cost/quantized.png differ
diff --git a/tamingllms/_build/html/_static/cost/quantized.tsx b/tamingllms/_build/html/_static/cost/quantized.tsx
new file mode 100644
index 0000000..aef322c
--- /dev/null
+++ b/tamingllms/_build/html/_static/cost/quantized.tsx
@@ -0,0 +1,58 @@
+import React from 'react';
+import { LineChart, Line, XAxis, YAxis, CartesianGrid, Tooltip, ResponsiveContainer } from 'recharts';
+
+const MemoryUsageChart = () => {
+ const data = [
+ { name: 'F16', value: 141.1 },
+ { name: 'Q8_0', value: 75.0 },
+ { name: 'Q6_K', value: 59.9 },
+ { name: 'Q5_K_M', value: 49.9 },
+ { name: 'Q4_K_M', value: 42.5 },
+ { name: 'Q3_K_M', value: 34.3 },
+ { name: 'Q2_K', value: 26.4 }
+ ];
+
+ return (
+
+
+
+
+
+
+ [`${value} GB`, 'Model Size']}
+ contentStyle={{
+ backgroundColor: '#fff',
+ border: '1px solid #ccc',
+ fontWeight: 'bold'
+ }}
+ />
+
+
+
+
+ );
+};
+
+export default MemoryUsageChart;
\ No newline at end of file
diff --git a/tamingllms/_build/html/genindex.html b/tamingllms/_build/html/genindex.html
index 92c3679..18858b4 100644
--- a/tamingllms/_build/html/genindex.html
+++ b/tamingllms/_build/html/genindex.html
@@ -190,6 +190,15 @@
+
+
+ The Falling Cost Paradox
+
+
+
+
+
+
diff --git a/tamingllms/_build/html/markdown/intro.html b/tamingllms/_build/html/markdown/intro.html
index a77a538..484ac75 100644
--- a/tamingllms/_build/html/markdown/intro.html
+++ b/tamingllms/_build/html/markdown/intro.html
@@ -208,6 +208,15 @@
+
+
+ The Falling Cost Paradox
+
+
+
+
+
+
diff --git a/tamingllms/_build/html/markdown/preface.html b/tamingllms/_build/html/markdown/preface.html
index 36b808e..2f54208 100644
--- a/tamingllms/_build/html/markdown/preface.html
+++ b/tamingllms/_build/html/markdown/preface.html
@@ -190,6 +190,15 @@
+
+
+ The Falling Cost Paradox
+
+
+
+
+
+
@@ -227,7 +236,7 @@ 1. Preface
