sable-sphynx 0.7.0.dev0__py3-none-any.whl

sable_sphynx-0.7.0.dev0.dist-info/METADATA

@@ -0,0 +1,152 @@
+Metadata-Version: 2.4
+Name: sable-sphynx
+Version: 0.7.0.dev0
+Summary: Sphynx offline evaluation SDK
+License: Proprietary
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+Provides-Extra: testing
+Requires-Dist: pytest>=8.4.2; extra == "testing"
+Provides-Extra: storage
+Requires-Dist: google-cloud-storage>=2.18; extra == "storage"
+
+# Sphynx
+
+Sphynx is a Python hub for engineering evaluations. Define a **suite**, select a
+**candidate** in a **plan**, execute a **run**, assess saved **evidence**, and
+compare **reports** against a compatible **baseline**. Criteria state what
+matters. Evaluators produce assessments. Metrics describe numeric measurements.
+
+## Current source workflow
+
+The current `0.7.0.dev0` source adds a shared Python workflow and explicit extension contracts.
+These APIs are **unpublished candidate work**; the frozen private 0.5.0 wheel does
+not contain them. Use Python **3.11 or later**. The package name is
+`sable-sphynx`; Python imports use `sphynx`.
+
+```python
+import sys
+from sphynx import CandidateSpec, INTERFACE_VERSION, Suite
+
+suite = Suite.from_pack("rag.rerank.v1")
+candidate = CandidateSpec(
+    "supplied-order-example", "example-v1",
+    (sys.executable, "-I", "-m", "sphynx.demo_candidates.dense"),
+    INTERFACE_VERSION,
+)
+report = suite.plan(candidate).run("results/first-run").assess()
+print(report.to_markdown())
+```
+
+This built-in backend executes six synthetic reranking cases in **local component
+scope**. It preserves native run and grading-pass identities. Saved evidence can
+be assessed again without executing the candidate. A completed run does not mean
+that the candidate meets its criteria.
+
+Start with the [shared workflow guide](docs/shared-workflow.md), then the
+[concepts and naming guide](docs/sdk-concepts-and-naming.md). The
+[capability inventory](docs/engineering-capabilities.md) distinguishes executable
+paths, contracts and missing prerequisites. Custom datasets, evaluators, criteria,
+metrics, structured assessments and execution backends enter through registered,
+versioned contracts. Other native RAG and speech paths remain available under
+their own documented APIs; they do not all use this shared execution path yet.
+
+Configured backends add exact-revision hosted submission, pipeline replacement
+and exact stage views. The versioned lifecycle path adds repeats, separate
+execution/grading retries, limits, cancellation, resume and verified reuse.
+Integrated retrieval-to-answer execution and installed extension libraries use
+the same registration boundaries. See [advanced workflows](docs/advanced-workflows.md)
+for the flow, configuration and support limits. Live hosted verification still
+requires the authoritative service source and deployed revision.
+
+## Frozen private 0.5.0 release
+
+The following instructions apply to the **existing private release**, not the
+new shared API above. PyPI publication is pending. Install the prepared local
+wheel. There is no public Sphynx CLI.
+
+The release includes [paper-named RAG methods](docs/rag-paper-methods.md) in
+`rag.papers.v1`. The post-release [Kido speech extensions](docs/speech-kido-benchmarks.md)
+and [speech coverage guide](docs/speech-complete-coverage.md) require the
+continuation source checkout. See [extensions and explorations](docs/explorations.md)
+for remaining data and calibration decisions.
+
+1. Use Python **3.11 or later**. Download the matching wheel from the signed-in Sphynx documentation in Evals Lab. Release maintainers can also download these files from the private GitHub release:
+
+   - `sable_sphynx-0.5.0-py3-none-any.whl`
+   - `sable_sphynx-0.5.0.tar.gz` (source distribution)
+   - `sable-sphynx-0.5.0-source.zip` (reviewed source bundle)
+
+   These are private release files; PyPI publication is separate. Check the wheel against the release SHA256SUMS before installation. The ZIP extracts to `sable-sphynx-0.5.0-source`. It contains the two starter files.
+
+2. Open a terminal in the folder with the wheel. Create your own project and a new virtual environment. Install the wheel.
+
+   PowerShell:
+
+   ```powershell
+   python --version
+   mkdir my-sphynx-eval
+   cd my-sphynx-eval
+   python -m venv .venv
+   .\.venv\Scripts\python.exe -m pip install --no-index --no-deps ..\sable_sphynx-0.5.0-py3-none-any.whl
+   ```
+
+   POSIX shell:
+
+   ```sh
+   python3 --version
+   mkdir my-sphynx-eval
+   cd my-sphynx-eval
+   python3 -m venv .venv
+   ./.venv/bin/python -m pip install --no-index --no-deps ../sable_sphynx-0.5.0-py3-none-any.whl
+   ```
+
+3. Save this as `my_first_eval.py` in your project:
+
+   ```python
+   from sphynx.demo import run_demo
+
+   result = run_demo("results/demo")
+   print("Reports:", result["reports_dir"])
+   print("Candidate summary:", result["summary_path"])
+   print("Baseline summary:", result["baseline_summary_path"])
+   print("Comparison:", result["comparison_path"])
+   ```
+
+4. Run the file with your new interpreter.
+
+   PowerShell:
+
+   ```powershell
+   .\.venv\Scripts\python.exe my_first_eval.py
+   ```
+
+   POSIX shell:
+
+   ```sh
+   ./.venv/bin/python my_first_eval.py
+   ```
+
+5. Read `results/demo/reports`. The demo saves both full scorecards as JSON and Markdown, both eight-metric summaries as text and Markdown, and the comparison as JSON and Markdown. It keeps earlier runs. Calling `run_demo()` without an output directory saves under your home folder at `.sphynx/beginner-quickstart`.
+
+6. Read the labels before the numbers. A is the supplied-score baseline. B is the word-overlap example. Each comparison delta is **B minus A**, or candidate minus baseline. Quality uses five eligible cases. Response validity, failure rate, and latency use all six synthetic cases. A ranking score does not establish answer correctness, embedding quality, or production performance.
+
+7. Use the [beginner quickstart](docs/beginner-quickstart.md) for source installation and data requirements. Use the [VS Code tutorial](docs/vscode-tutorial.md) to select your interpreter. Use [terminal batches](docs/terminal-experiments.md) to copy and edit `candidate.py` and `experiment.py`. Use [metric views](docs/metric-views.md) to read the eight metrics and comparison gates.
+
+Local results remain **`local_unverified`**. Local labels are readable by the computer owner. Subprocess execution is **not an OS sandbox**. Run trusted candidate code. The bundled demo needs no model, GPU, API key, storage account, or cloud access. Customer banks require authorized local assets. Storage, provider, and model dependencies are separate optional setup.
+
+Validation receipts apply to exact release hashes. The research integration has its own installed SDK and seven-platform gates; previous releases do not validate new bytes. See [research parity](docs/research-parity.md) for methods, matching results, asset prerequisites and remaining benchmark gates.
+
+## Hosted PR evaluation pilot
+
+The separate internal hosted CPU pilot can evaluate an unmerged registered GitHub PR and
+compare it with `supplied-score:v1` on the six-case synthetic reranking suite.
+The current source exposes `Client.submit(...)` and
+`Client.get_submission(...)`, returning a `HostedSubmission`. The private 0.5.0
+client retains `Client.evaluate(...)` / `get_experiment(...)` and its native
+`Experiment` identity. Use `submission.wait().summary()` to read a completed
+submission with either equivalent handle. No candidate runs on your computer.
+See [the hosted control guide](docs/hosted-control-plane.md) and
+[the research SDK guide](docs/research-parity.md). This pilot is not connected to `Suite.plan(..., mode="hosted")`.
+Configure the current operator-provided API URL and token. The reviewed private client wheel is in
+`dist/sable-sphynx-0.5.0/`; this update is not published on PyPI.

sable_sphynx-0.7.0.dev0.dist-info/RECORD

@@ -0,0 +1,174 @@
+sphynx/__init__.py,sha256=WLP8wbd0VHr9HXkNr1O8Ynp5Q0ixmU4erce9wceDpUo,22468
+sphynx/__init__.pyi,sha256=sQkarH-L_TRV3rK10RNa_S672CobQWA3IV-g68FUXGc,25112
+sphynx/_analytical_client.py,sha256=0kJhKPudNEDKOXYspOUIel3UuvOeERLzZfAiZwBHEQk,3859
+sphynx/_analytical_values.py,sha256=QhFECb8lpy9fjQ7xfEjLQWrAMi_qwRPVvBi44xVZo4E,11521
+sphynx/_audit_client.py,sha256=vNWeP18y-CCu34yHQoCo-7PfXsdo8jL6HAPgiBO8-F8,9266
+sphynx/_comparison_adapters.py,sha256=tf9X2NI7fKaPVZdYsL_lA2m7Vj2KSK0GOfhkwpiQtes,6998
+sphynx/_componentized_full_rag.py,sha256=nezypiwzWlX6CAtGBmngJVFaHdon6ggXSemm1WZcYHs,2463
+sphynx/_composed_pipeline.py,sha256=m3TbgztbZFmvopcau68mTNR3bs0c424-16YexQgAJqo,7602
+sphynx/_full_rag_grading.py,sha256=cpNrn_jTthglkGIqJTvKxal_cihCai_OHxJKz2MOuqY,8909
+sphynx/_full_rag_trace.py,sha256=f_RCvRPCgiqFlskJYSnmpJQd9GJ3bCvhVlTccLWtHWw,9393
+sphynx/_local_eval.py,sha256=xSXyvxDWonPpbke0uFv7HrMdfazUClRxfSJ5UsBzhHU,12415
+sphynx/_pipeline_core.py,sha256=5uf-_D-O4oIEkc0G43JkTCjI0VjlAfCc4IRR9uz9nFM,11009
+sphynx/_pr_client.py,sha256=C7-q3Ftxg5AKJ-Yt3Vw83I5z4t6MOeFpANo5ofEzRGo,16199
+sphynx/_rag_diagnostics_calibration.py,sha256=wjaBpSGkia_sc3Id8JCohdpdS2KBqYPH5NjZO0aNYS0,14086
+sphynx/_rag_diagnostics_catalog.py,sha256=nbEL-5mNR9ksIQxTBy0cqEnZnibpe_sEHu87krK58Hc,6427
+sphynx/_rag_diagnostics_common.py,sha256=YaHvxcKbXyG14PnrX23U39mcDWPYfNpWJaSTrlrlRbQ,1186
+sphynx/_rag_diagnostics_contracts.py,sha256=bOR7pGI-X-ewB50oKkvI6TRmtv8jJRDQJf9G9jnPuow,11180
+sphynx/_rag_diagnostics_engine.py,sha256=Sbp7rAdqDAWRyFLFTlWhbGEs0w1dhV6vcAbt5gkTh3E,16325
+sphynx/_rag_diagnostics_importers.py,sha256=VLbN1DM35kvIZf60GQ9tMfFZDHeuYGfT7mGC3ljUTjs,12929
+sphynx/_rag_diagnostics_ranking.py,sha256=XysT58GXoXRQxpMIe5lL6LjF1TOjmTrXxtH0BIIil-E,5676
+sphynx/_rag_diagnostics_semantic.py,sha256=rQsG5XpDa-1niLXZbvtgpOwabXdGmfcON0RV8mO8D2k,19682
+sphynx/_rag_ranking.py,sha256=EauccWnCj76oZNC922PVa_-sn3gQqiKhNS4nWDhtYVk,5135
+sphynx/_speech_benchmark_scoring.py,sha256=hfT-OaTkCZWnp1yEoUR77sNxJPC7AjgsE8REl1nx5XI,9223
+sphynx/_speech_benchmark_sources.py,sha256=XYXnxlbrRUEkstOTpM3lSM3od3RKj6fKaw6hkXz5mE0,19364
+sphynx/_speech_complete_scoring.py,sha256=78h1RhDRHGdr9YTFKdKITa-GvUnpWPorwTekx16yMwI,16240
+sphynx/_speech_complete_source.py,sha256=ZmvtU5APxc_XQ7KuSbDVCeI2woaJEFGZ31fOKlc0SDA,7714
+sphynx/_speech_kido.py,sha256=N6kHFWYPXKN7GjNnzQMU2gJvNw-85ZlI23rr0Hk91eM,17863
+sphynx/_speech_scoring.py,sha256=9gCm_AR6gjLIgpP1Gug9uos0Ghhl2FWNdmp-iXgEtaQ,3412
+sphynx/_track_capture.py,sha256=_MJqHro5vGQO_3JJK4n4UuG5uDY0c3DqNBJDb0MxuLQ,9689
+sphynx/analytical_views.py,sha256=ova6QM6S_WQJ43lDzXUFI0McQc1Hc9TcvKqSnPfgweA,26228
+sphynx/assessment_contracts.py,sha256=uZGsi0KIGskeeBta75COyEoqvNGBjc0gDI3mh5wErnY,18959
+sphynx/assessment_selection.py,sha256=BSx-YM_Tz-MtGONt0xMiVsqddACvgNb1MyvEqupU9fU,2502
+sphynx/audit_access.py,sha256=zZnitAWcAnJ_IsRDs5Zu0kIcmRZd6rcg0z0iLXZIxXE,13201
+sphynx/audit_contracts.py,sha256=dUZYBSgAQ08YI0-VWUxjKzOor03rTsp09K3TS0S3ueQ,9268
+sphynx/audit_requests.py,sha256=iGyJ8lhFCE7IMctM9E-gmJPoX3AhXWSyhXaOsTTPQNQ,8461
+sphynx/authoring.py,sha256=8SR8QK0Ihg5errYuwWuHuNtfMMyWfRZ5DTP2gxtMvEw,11837
+sphynx/benchmarks.py,sha256=560oNsZOmjwuqE3Jjva5L6GZlQIhlHwq_eBtGz_UgaU,31230
+sphynx/candidates.py,sha256=rVlo_nEA471DUp13Pupf9tG51hM-GpQvnF14QsB_454,2727
+sphynx/catalog.py,sha256=vrWPOBHpoQTBGEDqlUUEik-LNZNb-4olK5dNNrrEVG8,13623
+sphynx/chunking.py,sha256=ogTeWvPGfiYwbhppilKzZ1fJkH0xNaNWSmHzS06bae8,23614
+sphynx/clickhouse_reader.py,sha256=iwjoFFBQMvPAeQ-G3XXamqdyDwvqHUoyftTAVVxEq6E,9994
+sphynx/client.py,sha256=VBCpbB-AJVnKTrj68xs67EZdK4zI9XDwPEFsvlhmHew,29495
+sphynx/comparison_core.py,sha256=TEezWBaDKHNrdtDiWJ02Dv3beSK9PkFcA-iXiy52pwU,7436
+sphynx/comparison_views.py,sha256=mRSn70kR4FQWmIgJYuI2ZwPQZs29w3S47HgSMtO18fg,11627
+sphynx/components.py,sha256=LFvdnQfyDutLdIYU2uCJ1aa8fFzvjzwZF_T64TUpMQU,97849
+sphynx/contracts.py,sha256=wLpXUo189Dmnqd13J_7ogHNZ2NpvLWIRSatRmtN9N4s,21791
+sphynx/customer_rerank.py,sha256=hKEEqlXrmn_L42ethaSgnT_FBaRGzW90ybka_7aD-bw,17720
+sphynx/customer_retrieval.py,sha256=w2muO_JcKTt3jjZ4HpJCLhRxyIc90dtYtzv76Y-w1kE,8559
+sphynx/dataset_catalog_data.py,sha256=Txr5Ggv-pAgjD1nRvtqYAhMNr84UGvZKziVZrDxNYnY,56248
+sphynx/dataset_readiness.py,sha256=4woItZA5PM7EJnnksLDBmm4q4gaZc7O6qB-R9yMhCBI,6651
+sphynx/dataset_v3.py,sha256=Xo5LdKFDAEJ5qJ8p4AqDjy-6Wjy8FZJL_oowOSDwI0w,60845
+sphynx/demo.py,sha256=QSBB2rLkA5E7OwQSy3vAQns1eWbaUwRg0rWPnmMLGK8,5780
+sphynx/diagnostics.py,sha256=vg0v_8hajZJbrmNJJ8vzTMcBVIYVC7dNCEc9KLmK3VE,11680
+sphynx/documentation.py,sha256=JWO2LN_l146Sx-RymTsvTqGF698nvIPh3u6DYY6j8bg,54202
+sphynx/documentation_candidate.py,sha256=ur2ltqohtZjcdZ7BeUaKcylFowQUHmmOBTemCBntQ78,60305
+sphynx/evals.py,sha256=v3XS_kEO9JSlw2aYMQ9jF0wMIekvGRDqKQYiP_nLy8o,30289
+sphynx/evaluation_protocols.py,sha256=5SERS4Nm9ZCx2fQAD5_Mfaqk2GNnfw88fEg5D5j4p44,5460
+sphynx/evaluator_bindings.py,sha256=4yufCrfgLGfVtLJYxSQdQF76URXJC9mAzKrN3zU-pwU,20102
+sphynx/evidence.py,sha256=HVGPENXJHG5nhsCLdNMKG0uRqG6umb61sVeuTdWXEH8,17654
+sphynx/execution_capabilities.py,sha256=f86VA4uOUBvzQAAcu3DBmDqgU5TfdATIxlGh1w886wQ,5229
+sphynx/fast_search.py,sha256=stX5CVo46DIBhi8Yd7qNqgdHhXoOkixYYPzw69E86X4,11536
+sphynx/full_rag.py,sha256=WTrtqkFFr9DYEYSWFgKsT1wRN8Y2GSrpykLfHO6mFc0,27800
+sphynx/hosted_workflow.py,sha256=1UlPW3AYOdrgoB7xNTG_eKsgBw5scdy0olI7aPF3TXY,24776
+sphynx/http_transport.py,sha256=0N6ld3e7OSVXHPXzKYuT9mkWLmEJKj6Znewm0drY13k,4230
+sphynx/integrated_rag.py,sha256=a6Ab2iAgSX0OdUcNn-t1n-E8HOsuEFD4v6aYqzhsSvs,20729
+sphynx/integrated_workflow.py,sha256=WLhkivAflNqFX0o7nT1QVj6KVkEuh6rHAQljk4b8dEc,37395
+sphynx/integration_contracts.py,sha256=h-phl91VYORuGjHa5g_OtYgRQJOJci67qL88NMXlKcI,21277
+sphynx/invocation.py,sha256=02K3ay9cUWvhJQCje3RHFW9Iv6mr6tpd7uWenwhKOjY,42629
+sphynx/invocation_candidates.py,sha256=G1pjHPof1ZDsR70W7I4PNvTX62yetVqDD-cNF_d42ho,23479
+sphynx/invocation_catalog_data.py,sha256=ghuPFqPI4tYQEvpuEKhlS9NcgIJm6LwcPZhYDxaNNDU,7473
+sphynx/learned_catalog_data.py,sha256=v0e2t0FJkR6gIGWxqsAfGrZf4rfcLHdrlQeoNyxxR6I,2436
+sphynx/learned_rerank.py,sha256=sM2BOF98Hl8aRMQHiLQXSlne8F0qlcXr5t8_0kjL1o8,22038
+sphynx/legacy_full_rag.py,sha256=b9aJEJ1mrO2R7wxl9oR5GMEMHGQmITywdsBZYDIUb54,5558
+sphynx/lifecycle.py,sha256=GjGsY2CNbaFKpeyJpT5TMF5F4cBQ4XN6m_Thv-ftk0E,47761
+sphynx/lifecycle_workflow.py,sha256=uCRzEYcAQdg20z-T0ZvmL4LqIdWlKP2q0xfrRTRXe1E,32150
+sphynx/metric_views.py,sha256=W6Dd8mzjSEWi8nadzi15mbuodwxtNl-WCWKNXeP9MmU,27976
+sphynx/metrics.py,sha256=MQZjGErVXG_n1mDK1pY77mjcvhC5Aw9xNeScy4BejOk,13750
+sphynx/model_adapters.py,sha256=YZcp9FCiRDC74sFkGm_kdyPMn2Fjhex4FPP8tS4iScE,25134
+sphynx/model_catalog_data.py,sha256=Hnycm4HwiqIFXDBXaWB2GQOh_WIjlaYyl1T9F_fusAs,2730
+sphynx/official_benchmarks.py,sha256=0ysnsM-KXd0rV-e5lOlbH85w1ei9-_SlcAdVXvVAiIY,21416
+sphynx/packs.py,sha256=QfHXNfSCdsthft63qi2oc_MiA-CqpiP5Zco-oxlPZzs,12859
+sphynx/pipeline.py,sha256=rkP8tEDJIQCMwFUW8cDLLSwFjpfUKhfjJH5WCAueNQE,21637
+sphynx/pipeline_workflow.py,sha256=CLBuwsA1-WPm3k8Da7EAYRCnb8tCM8p7h7mNYbtUDMw,33643
+sphynx/pr_evaluations.py,sha256=0FMSta_HktZZyMaQioDTy7g3JYIhqc-U93sh71K40IQ,32134
+sphynx/progress.py,sha256=K5zL2qY8OUQePHIjUDiZZKyoSMzkAP2tE-cSiC4V3M0,4552
+sphynx/public_bundle.py,sha256=pkqocT1OVdeY3qOOWbvbE--6VPGyKfc0Pa22sbZuA6c,10258
+sphynx/public_eval.py,sha256=ZzXZD-YUocoZC5cTYX8mGsjQqiKaBN3hfm9ZDkBr7Vw,2898
+sphynx/public_rerank.py,sha256=arGl3endzFyNcLcyhy40ZKk2Mcm4qZ6I_NE1Q4hpJOE,22630
+sphynx/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+sphynx/pytest_plugin.py,sha256=bJfEeffUSpajrl5X5BPoRlTqqYwPTYG_vzNvKwXy3us,2648
+sphynx/question_robustness.py,sha256=xlM3tAEmKxPqwhqDxTg_epj0cdBxCZFxsHdJUjs5BIE,31994
+sphynx/rag.py,sha256=VKHJiOezfheflll5P52-7DbBQPgcZp8Z76fFh_o9Ab8,17941
+sphynx/rag_diagnostics.py,sha256=_1aLaam_S7v-EZgbT2Z57NPN18kTRLqLxLT6QjyM8_8,15781
+sphynx/recorded_scores.py,sha256=_Fgv6qbL0_jSJv3Fa_lSlwiJd00kFzKRPgJ8ex8YtSQ,23001
+sphynx/records.py,sha256=nJ7mvEd3K5aqzyFK6h3lFOkHpCjqfv-4dbhNuGSL9Jk,4894
+sphynx/reflex_traces.py,sha256=3UY0bNsgbE9fY9YSMT64YrCBspPDgwGpoGkP5W2UdtI,13857
+sphynx/report_rendering.py,sha256=ixJwEv4y9Wv7TNXTxS6I1CApnU5pVISLrlHqLPtrrsw,3759
+sphynx/rerank_family_contracts.py,sha256=h3RYTGkpkMje0lHLKGYg_62oklty-gf4HaEeA4XVdxg,8660
+sphynx/rerank_workflow.py,sha256=EIv5JZjL_232UE1BL9o5Qi0vElnNd82GfiQ-U3Ki-0Q,24424
+sphynx/research_candidate.py,sha256=sPQ8qjoU8warUU0J54sXVWhAF9RMjTn5dEW-mkeBUZM,10537
+sphynx/results.py,sha256=TfSM3pS_JiOsEAQAk4STC-ZvDknFZmmmg-Lcdqk8MaY,46755
+sphynx/retrieval.py,sha256=3v1UIOsqXTwvrJhn748cABZ5pDzy1wMeG46Td-fe4SI,45362
+sphynx/retrieval_candidates.py,sha256=IR4yrHDhOH4UahBVirHjfg7ADLrQ7OodehLaLahVdq0,17936
+sphynx/retrieval_catalog_data.py,sha256=7aBjdzPzpoQQvjHUHGh1JUNvlPP_nKEkJ06LZZru0yQ,1415
+sphynx/retrieval_recordings.py,sha256=D5uUNrmk0tbz7FsMaGTAt2PTUaBMlXSRHXvW-GPcsM8,8185
+sphynx/reviewer_audit.py,sha256=UDnc8L66ngkDCvaCq03J25j5y1yQ3aosy6aWhepbEWI,14580
+sphynx/routing_policies.py,sha256=Xny_mS96RRbFtmE4Gp3OvWdu4MY8vIlLij8j6jKV6Ms,24207
+sphynx/runner.py,sha256=iCtMPrb4CC6TGMK7OABQSUu8Oci4GXhEO4svypYADO8,22983
+sphynx/sable_rerank.py,sha256=Tb_CIWHf-qsx0HNiX5awh2zg-BkRqGG6J6Jp9h-HNwo,26602
+sphynx/score_reliability.py,sha256=WdrzaYl4ESSUUkHLZ20KgGmVqSNFRBxbsnQCV_44ZKI,29134
+sphynx/speech_benchmarks.py,sha256=FM6si_fFpDMg0xtw1EehVxaC6vFHiW7JcryVvM6-kls,13872
+sphynx/speech_campaigns.py,sha256=0j1T1LN-Gzy2icuIKmTXD43G9A1FvyNOF1Ly4TKS338,30306
+sphynx/speech_complete.py,sha256=Oi3mM77PpJOOkQDCBrAvVS2pEIIMqZ8HSpdEpT9lh6k,16432
+sphynx/speech_contracts.py,sha256=lXHSvo3xm99tHZ_go66aSo8O2eDiSsoP7ks3vDYXvGw,12275
+sphynx/speech_evidence.py,sha256=q1Na2nWOzgMELGLVwNGQtqLhnCVpwsKireHui_19Tok,24476
+sphynx/speech_listening.py,sha256=_6kghNGkGUDfSagmnHRkzCDXgWswiydHI7SnODQe8a8,9669
+sphynx/speech_multilingual.py,sha256=up5MHgzdL4VROJxywgeJVrAn0NFjHY6yBMSWILKro4k,5994
+sphynx/speech_tts.py,sha256=cSLWx-xco2vRx00LXqBOTeJ09ii0wLokQG4aZX4L6xk,29394
+sphynx/stage_contracts.py,sha256=OB6NpAR7nCai2R3obka9v3hDQqJI1QUi3dFdL9cxX28,9108
+sphynx/statistical_catalog_data.py,sha256=OoJx16yRPpUQ2TuUWRho7tEtSWN-KjN-Iqup_-DKGlY,3787
+sphynx/statistical_rerank.py,sha256=ulu1a0Ft6B7ZiDtf7UE35UXSCJyPfoiV5vKKLU_GKn0,17896
+sphynx/sync.py,sha256=OdYxsaxZW7Kqx5mOAQDBHs-gLFKZl_Hjy5Ih-YMbCDA,41589
+sphynx/workbench.py,sha256=zvUH9mV7bJdayqu9KV4e3m4Kd2yezX49zfPBW52NyRM,24133
+sphynx/workbench_catalog_data.py,sha256=-CfFJMuaopJ4A9gOadPPWEN3FpMVgNvc7g29Lzmrk9I,915
+sphynx/workbench_tracks.py,sha256=GIU_OqrY8Kz11LZi_We-UvkLYAriwYTKJv8SI67vkJM,5308
+sphynx/workflow_async.py,sha256=YVBPdRZHR2U3nnDS5lPBU6glUu2G_7vj9gGikVwF5x0,8509
+sphynx/workflow_contracts.py,sha256=iOBJioEYlXm9rBWL8ioen2Mcx4lLU2aZ7uYbcfd6Zx4,27054
+sphynx/workflow_decisions.py,sha256=pVIdDu56Ojw7b_SLTtcgMRv_a6sR1OKGK6sMs68TZqk,9669
+sphynx/workflow_defaults.py,sha256=od16ylvOKtk_Nw10kAAfHBPCnJiGICCKtXFoEEzNoo4,621
+sphynx/workflow_extensions.py,sha256=ogXkYbLJzxDztkPwdFfUgXErZwuslQPjxFVLb0x0Vkw,10891
+sphynx/workflow_registry.py,sha256=roqcOP2RtR04d2b3hCfl52lCs3zHMoNhEi06dGGz4Cw,27868
+sphynx/workflows.py,sha256=tcMEPywg2GGnD_UFM4U-oJ4ObsvkgP8ly1GPB2gb1sQ,37360
+sphynx/data/benchmark-releases.v1.json,sha256=9taPisMkJ80TTtSnrL30TiQQffFOxOtlPVr23rDCugE,6270
+sphynx/data/public-benchmark-release.v1.json,sha256=41ClxDNiBcjgmsTu37PupfSnjnH2CUqYHDDLvglK1bQ,13204
+sphynx/data/rag-diagnostics-synthetic.v1.json,sha256=dXtJB6O0TDFmbBVH_1coFC6D1KUvu_2ckSzzSVb0-Go,202891
+sphynx/data/rag-paper-references.v1.json,sha256=n5aH3JM2UDUGEcoOE-Gu9MpGet8Z0peoEp3pN9Eez90,7911
+sphynx/data/rag-paper-schemas.v1.json,sha256=WmSZQsR9EWhpCzu4pShJihPQTY9WP1PtYK7VRn-XC7E,26476
+sphynx/data/rag-rerank-v1.json,sha256=dIhKAXZzhknYXMi2gTIxA5YZOaqg89Xx3RVl7S5T2Gs,18661
+sphynx/demo_candidates/__init__.py,sha256=jA-hAXlZUCtTFwNijU-bG0lV5RO8gRYWxGYWDO32DOc,82
+sphynx/demo_candidates/dense.py,sha256=QoIls-NWmhzXNKbInIBwBbjJMxWLtamKKWGoi3oNke4,289
+sphynx/demo_candidates/lexical.py,sha256=WBARMZ8etDjBC8vymHX9lQUubPA5_g6Uf58yhMAovD8,630
+sphynx/hosted/__init__.py,sha256=hrnvkyoBcDDgCaPHEOq5GMDg_kaV6Z8AFVPSPlevR4U,834
+sphynx/hosted/benchmark_intake.py,sha256=s6daYl04rS7292D9wqNOeJRf8P1ymcv44YIj2XU4G-w,13236
+sphynx/hosted/contracts.py,sha256=xhg9_bWdXgkHwSdtjsu5YhL0pLptgesTqKO-TnYornY,20536
+sphynx/hosted/model_admission.py,sha256=HbLnHVB-Sq737sTtLNOJemdkmIJao1eLc3tYGEonxx4,6969
+sphynx/hosted/model_artifacts.py,sha256=Ou6CImHaJQ2jDhPi347MvdLMpsfhJ9y9EU_f4WmW240,15280
+sphynx/hosted/model_contracts.py,sha256=UXf39iNJE83nPGD_4H6NBnbwJ_O1Y_5YTl3_YkeC6pA,3401
+sphynx/hosted/model_gcs.py,sha256=U-yIfdBCLNliupM3USCwjeshWuX76He0ndY3FJ2YB98,9763
+sphynx/hosted/model_history.py,sha256=2p8YPkzaJA6djZB1Z5d-gz4Kj2vgSIn_xLK7Esc0snM,5603
+sphynx/hosted/model_session.py,sha256=oxEjJ2z-lsRStjPGb0I2jLdoE_OvTPZf4AQeB6Ld5ew,12012
+sphynx/hosted/public_suite.py,sha256=y39qjS3z29am92StuTeSu5MqpLcIK-PbsmCTXQK3EcI,3757
+sphynx/hosted/registry.py,sha256=CggYpLkhrhrzNhEs58bKDyvEE4W3U-LyfOInnNFzddA,5089
+sphynx/hosted/revisions.py,sha256=TBvHfmX30qSUhBItLkkw8HzIZnp2e20ALEUBb-ESl30,7538
+sphynx/hosted/sandbox.py,sha256=0S4wY_kPJMHpYTmuewTAbnwVkampQYGF961LWk2EX-0,6309
+sphynx/hosted/submission.py,sha256=x0XxLRgb_hzadDXqIJT04UtcT4jhXsBpMvm-uU2Tkk0,15864
+sphynx/rag_papers/__init__.py,sha256=P__RP9K2xr3nVYHso1wsW6E5GaXEzssPXorCSnmz2to,2245
+sphynx/rag_papers/__main__.py,sha256=7oAObAOi2GxYjNtWpRJLwvS5pzsfq-Gtq9qZxrbMc44,2904
+sphynx/rag_papers/candidates.py,sha256=k5gJll1t8DfzJTG18ZjlnBUBKJe0BFqt12XNARF7UsY,1302
+sphynx/rag_papers/contracts.py,sha256=yUwo6uLW2aW1Rqhby0feuHDgHaWKkjDxsfV_GxlmcpQ,14670
+sphynx/rag_papers/datasets.py,sha256=3jx-VpEMy92suXhnKEP2_STilyffhT-7s2vB8xs8FqU,17694
+sphynx/rag_papers/experiments.py,sha256=hiuJI3yR_uk5z6AU94aY-62M0NqnbYDaG77eotNyteY,7484
+sphynx/rag_papers/judges.py,sha256=OnJjIgdFD_QIH03fexH0eUcu3SsW_51zXuscRFVXUU4,23141
+sphynx/rag_papers/registry.py,sha256=yUP8HJo75eQ2UVz44zmqXPm7kqk6z6Butbh36ZQa5MQ,2875
+sphynx/rag_papers/runner.py,sha256=IiYJVwpBN5syirjyO0MH_ZemP7n1fm9D2Qy735bKpAQ,10443
+sphynx/rag_papers/sable.py,sha256=OH47cFprLIjfKN6NprhgjYtni3NDSdEYcKmKJa7sYzw,9405
+sphynx/rag_papers/schemas.py,sha256=Lfi2MPO1rDrKDgyba7XHq9xYCkgLSQekk4CJVHYxYgk,476
+sphynx/rag_papers/scoring.py,sha256=BEkaWRfi_Xbpvmt5W-6G1aF_dyTJilWCM5qa-wvvsEI,14484
+sphynx/rag_papers/semantic.py,sha256=uyHFIEpdnZBFLgHmMklxqIjNAHEy3s5MpWvRTfY2vpw,19346
+sphynx/rag_papers/trec.py,sha256=pQNoHWxIXpOHaO5oljJCvRRFQYkQfbu_ACf8KUZtU6c,4748
+sphynx/sql/001_sync_v1.sql,sha256=r_mMqteXGEN7toEjLnfHXVAPkkJ6fE0C4G-bj3fVHQM,878
+sable_sphynx-0.7.0.dev0.dist-info/METADATA,sha256=4GD0N9jC7io5x1ZXL-6T3oA1uZ9BbZ7AqsL5b8s9srg,7891
+sable_sphynx-0.7.0.dev0.dist-info/WHEEL,sha256=16fc_9Rp854SLcdhaDEjr0u65p0e3FhZxihNgcnCE98,91
+sable_sphynx-0.7.0.dev0.dist-info/top_level.txt,sha256=pCekHUyY_dfDahq2D4uz6vYt24lFf-dJWxbTCqafD5o,7
+sable_sphynx-0.7.0.dev0.dist-info/RECORD,,

sable_sphynx-0.7.0.dev0.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (84.1.0)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

sable_sphynx-0.7.0.dev0.dist-info/top_level.txt

@@ -0,0 +1 @@
+sphynx