@synsci/cli-darwin-x64 1.1.97 → 1.1.99

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (1549) hide show
  1. package/bin/synsc +0 -0
  2. package/package.json +1 -1
  3. package/bin/skills/accelerate/SKILL.md +0 -332
  4. package/bin/skills/accelerate/references/custom-plugins.md +0 -453
  5. package/bin/skills/accelerate/references/megatron-integration.md +0 -489
  6. package/bin/skills/accelerate/references/performance.md +0 -525
  7. package/bin/skills/adaptyv/SKILL.md +0 -114
  8. package/bin/skills/adaptyv/reference/api_reference.md +0 -308
  9. package/bin/skills/adaptyv/reference/examples.md +0 -913
  10. package/bin/skills/adaptyv/reference/experiments.md +0 -360
  11. package/bin/skills/adaptyv/reference/protein_optimization.md +0 -637
  12. package/bin/skills/aeon/SKILL.md +0 -374
  13. package/bin/skills/aeon/references/anomaly_detection.md +0 -154
  14. package/bin/skills/aeon/references/classification.md +0 -144
  15. package/bin/skills/aeon/references/clustering.md +0 -123
  16. package/bin/skills/aeon/references/datasets_benchmarking.md +0 -387
  17. package/bin/skills/aeon/references/distances.md +0 -256
  18. package/bin/skills/aeon/references/forecasting.md +0 -140
  19. package/bin/skills/aeon/references/networks.md +0 -289
  20. package/bin/skills/aeon/references/regression.md +0 -118
  21. package/bin/skills/aeon/references/segmentation.md +0 -163
  22. package/bin/skills/aeon/references/similarity_search.md +0 -187
  23. package/bin/skills/aeon/references/transformations.md +0 -246
  24. package/bin/skills/alphafold-database/SKILL.md +0 -513
  25. package/bin/skills/alphafold-database/references/api_reference.md +0 -423
  26. package/bin/skills/anndata/SKILL.md +0 -400
  27. package/bin/skills/anndata/references/best_practices.md +0 -525
  28. package/bin/skills/anndata/references/concatenation.md +0 -396
  29. package/bin/skills/anndata/references/data_structure.md +0 -314
  30. package/bin/skills/anndata/references/io_operations.md +0 -404
  31. package/bin/skills/anndata/references/manipulation.md +0 -516
  32. package/bin/skills/arboreto/SKILL.md +0 -243
  33. package/bin/skills/arboreto/references/algorithms.md +0 -138
  34. package/bin/skills/arboreto/references/basic_inference.md +0 -151
  35. package/bin/skills/arboreto/references/distributed_computing.md +0 -242
  36. package/bin/skills/arboreto/scripts/basic_grn_inference.py +0 -97
  37. package/bin/skills/astropy/SKILL.md +0 -331
  38. package/bin/skills/astropy/references/coordinates.md +0 -273
  39. package/bin/skills/astropy/references/cosmology.md +0 -307
  40. package/bin/skills/astropy/references/fits.md +0 -396
  41. package/bin/skills/astropy/references/tables.md +0 -489
  42. package/bin/skills/astropy/references/time.md +0 -404
  43. package/bin/skills/astropy/references/units.md +0 -178
  44. package/bin/skills/astropy/references/wcs_and_other_modules.md +0 -373
  45. package/bin/skills/audiocraft/SKILL.md +0 -564
  46. package/bin/skills/audiocraft/references/advanced-usage.md +0 -666
  47. package/bin/skills/audiocraft/references/troubleshooting.md +0 -504
  48. package/bin/skills/autogpt/SKILL.md +0 -403
  49. package/bin/skills/autogpt/references/advanced-usage.md +0 -535
  50. package/bin/skills/autogpt/references/troubleshooting.md +0 -420
  51. package/bin/skills/awq/SKILL.md +0 -310
  52. package/bin/skills/awq/references/advanced-usage.md +0 -324
  53. package/bin/skills/awq/references/troubleshooting.md +0 -344
  54. package/bin/skills/axolotl/SKILL.md +0 -158
  55. package/bin/skills/axolotl/references/api.md +0 -5548
  56. package/bin/skills/axolotl/references/dataset-formats.md +0 -1029
  57. package/bin/skills/axolotl/references/index.md +0 -15
  58. package/bin/skills/axolotl/references/other.md +0 -3563
  59. package/bin/skills/benchling-integration/SKILL.md +0 -480
  60. package/bin/skills/benchling-integration/references/api_endpoints.md +0 -883
  61. package/bin/skills/benchling-integration/references/authentication.md +0 -379
  62. package/bin/skills/benchling-integration/references/sdk_reference.md +0 -774
  63. package/bin/skills/bigcode-evaluation-harness/SKILL.md +0 -405
  64. package/bin/skills/bigcode-evaluation-harness/references/benchmarks.md +0 -393
  65. package/bin/skills/bigcode-evaluation-harness/references/custom-tasks.md +0 -424
  66. package/bin/skills/bigcode-evaluation-harness/references/issues.md +0 -394
  67. package/bin/skills/biopython/SKILL.md +0 -443
  68. package/bin/skills/biopython/references/advanced.md +0 -577
  69. package/bin/skills/biopython/references/alignment.md +0 -362
  70. package/bin/skills/biopython/references/blast.md +0 -455
  71. package/bin/skills/biopython/references/databases.md +0 -484
  72. package/bin/skills/biopython/references/phylogenetics.md +0 -566
  73. package/bin/skills/biopython/references/sequence_io.md +0 -285
  74. package/bin/skills/biopython/references/structure.md +0 -564
  75. package/bin/skills/biorxiv-database/SKILL.md +0 -483
  76. package/bin/skills/biorxiv-database/references/api_reference.md +0 -280
  77. package/bin/skills/biorxiv-database/scripts/biorxiv_search.py +0 -445
  78. package/bin/skills/bioservices/SKILL.md +0 -361
  79. package/bin/skills/bioservices/references/identifier_mapping.md +0 -685
  80. package/bin/skills/bioservices/references/services_reference.md +0 -636
  81. package/bin/skills/bioservices/references/workflow_patterns.md +0 -811
  82. package/bin/skills/bioservices/scripts/batch_id_converter.py +0 -347
  83. package/bin/skills/bioservices/scripts/compound_cross_reference.py +0 -378
  84. package/bin/skills/bioservices/scripts/pathway_analysis.py +0 -309
  85. package/bin/skills/bioservices/scripts/protein_analysis_workflow.py +0 -408
  86. package/bin/skills/bitsandbytes/SKILL.md +0 -411
  87. package/bin/skills/bitsandbytes/references/memory-optimization.md +0 -521
  88. package/bin/skills/bitsandbytes/references/qlora-training.md +0 -521
  89. package/bin/skills/bitsandbytes/references/quantization-formats.md +0 -447
  90. package/bin/skills/blip-2/SKILL.md +0 -564
  91. package/bin/skills/blip-2/references/advanced-usage.md +0 -680
  92. package/bin/skills/blip-2/references/troubleshooting.md +0 -526
  93. package/bin/skills/brenda-database/SKILL.md +0 -719
  94. package/bin/skills/brenda-database/references/api_reference.md +0 -537
  95. package/bin/skills/brenda-database/scripts/brenda_queries.py +0 -844
  96. package/bin/skills/brenda-database/scripts/brenda_visualization.py +0 -772
  97. package/bin/skills/brenda-database/scripts/enzyme_pathway_builder.py +0 -1053
  98. package/bin/skills/cellxgene-census/SKILL.md +0 -511
  99. package/bin/skills/cellxgene-census/references/census_schema.md +0 -182
  100. package/bin/skills/cellxgene-census/references/common_patterns.md +0 -351
  101. package/bin/skills/chembl-database/SKILL.md +0 -389
  102. package/bin/skills/chembl-database/references/api_reference.md +0 -272
  103. package/bin/skills/chembl-database/scripts/example_queries.py +0 -278
  104. package/bin/skills/chroma/SKILL.md +0 -406
  105. package/bin/skills/chroma/references/integration.md +0 -38
  106. package/bin/skills/cirq/SKILL.md +0 -346
  107. package/bin/skills/cirq/references/building.md +0 -307
  108. package/bin/skills/cirq/references/experiments.md +0 -572
  109. package/bin/skills/cirq/references/hardware.md +0 -515
  110. package/bin/skills/cirq/references/noise.md +0 -515
  111. package/bin/skills/cirq/references/simulation.md +0 -350
  112. package/bin/skills/cirq/references/transformation.md +0 -416
  113. package/bin/skills/citation-management/SKILL.md +0 -1109
  114. package/bin/skills/citation-management/assets/bibtex_template.bib +0 -264
  115. package/bin/skills/citation-management/assets/citation_checklist.md +0 -386
  116. package/bin/skills/citation-management/references/bibtex_formatting.md +0 -908
  117. package/bin/skills/citation-management/references/citation_validation.md +0 -794
  118. package/bin/skills/citation-management/references/google_scholar_search.md +0 -725
  119. package/bin/skills/citation-management/references/metadata_extraction.md +0 -870
  120. package/bin/skills/citation-management/references/pubmed_search.md +0 -839
  121. package/bin/skills/citation-management/scripts/doi_to_bibtex.py +0 -182
  122. package/bin/skills/citation-management/scripts/extract_metadata.py +0 -570
  123. package/bin/skills/citation-management/scripts/format_bibtex.py +0 -349
  124. package/bin/skills/citation-management/scripts/search_google_scholar.py +0 -251
  125. package/bin/skills/citation-management/scripts/search_pubmed.py +0 -348
  126. package/bin/skills/citation-management/scripts/validate_citations.py +0 -494
  127. package/bin/skills/clinical-decision-support/README.md +0 -129
  128. package/bin/skills/clinical-decision-support/SKILL.md +0 -506
  129. package/bin/skills/clinical-decision-support/assets/biomarker_report_template.tex +0 -380
  130. package/bin/skills/clinical-decision-support/assets/clinical_pathway_template.tex +0 -222
  131. package/bin/skills/clinical-decision-support/assets/cohort_analysis_template.tex +0 -359
  132. package/bin/skills/clinical-decision-support/assets/color_schemes.tex +0 -149
  133. package/bin/skills/clinical-decision-support/assets/example_gbm_cohort.md +0 -208
  134. package/bin/skills/clinical-decision-support/assets/recommendation_strength_guide.md +0 -328
  135. package/bin/skills/clinical-decision-support/assets/treatment_recommendation_template.tex +0 -529
  136. package/bin/skills/clinical-decision-support/references/biomarker_classification.md +0 -719
  137. package/bin/skills/clinical-decision-support/references/clinical_decision_algorithms.md +0 -604
  138. package/bin/skills/clinical-decision-support/references/evidence_synthesis.md +0 -840
  139. package/bin/skills/clinical-decision-support/references/outcome_analysis.md +0 -640
  140. package/bin/skills/clinical-decision-support/references/patient_cohort_analysis.md +0 -427
  141. package/bin/skills/clinical-decision-support/references/treatment_recommendations.md +0 -521
  142. package/bin/skills/clinical-decision-support/scripts/biomarker_classifier.py +0 -383
  143. package/bin/skills/clinical-decision-support/scripts/build_decision_tree.py +0 -417
  144. package/bin/skills/clinical-decision-support/scripts/create_cohort_tables.py +0 -509
  145. package/bin/skills/clinical-decision-support/scripts/generate_survival_analysis.py +0 -441
  146. package/bin/skills/clinical-decision-support/scripts/validate_cds_document.py +0 -326
  147. package/bin/skills/clinical-reports/IMPLEMENTATION_SUMMARY.md +0 -641
  148. package/bin/skills/clinical-reports/README.md +0 -236
  149. package/bin/skills/clinical-reports/SKILL.md +0 -1127
  150. package/bin/skills/clinical-reports/assets/case_report_template.md +0 -352
  151. package/bin/skills/clinical-reports/assets/clinical_trial_csr_template.md +0 -353
  152. package/bin/skills/clinical-reports/assets/clinical_trial_sae_template.md +0 -359
  153. package/bin/skills/clinical-reports/assets/consult_note_template.md +0 -305
  154. package/bin/skills/clinical-reports/assets/discharge_summary_template.md +0 -453
  155. package/bin/skills/clinical-reports/assets/hipaa_compliance_checklist.md +0 -395
  156. package/bin/skills/clinical-reports/assets/history_physical_template.md +0 -305
  157. package/bin/skills/clinical-reports/assets/lab_report_template.md +0 -309
  158. package/bin/skills/clinical-reports/assets/pathology_report_template.md +0 -249
  159. package/bin/skills/clinical-reports/assets/quality_checklist.md +0 -338
  160. package/bin/skills/clinical-reports/assets/radiology_report_template.md +0 -318
  161. package/bin/skills/clinical-reports/assets/soap_note_template.md +0 -253
  162. package/bin/skills/clinical-reports/references/case_report_guidelines.md +0 -570
  163. package/bin/skills/clinical-reports/references/clinical_trial_reporting.md +0 -693
  164. package/bin/skills/clinical-reports/references/data_presentation.md +0 -530
  165. package/bin/skills/clinical-reports/references/diagnostic_reports_standards.md +0 -629
  166. package/bin/skills/clinical-reports/references/medical_terminology.md +0 -588
  167. package/bin/skills/clinical-reports/references/patient_documentation.md +0 -744
  168. package/bin/skills/clinical-reports/references/peer_review_standards.md +0 -585
  169. package/bin/skills/clinical-reports/references/regulatory_compliance.md +0 -577
  170. package/bin/skills/clinical-reports/scripts/check_deidentification.py +0 -332
  171. package/bin/skills/clinical-reports/scripts/compliance_checker.py +0 -78
  172. package/bin/skills/clinical-reports/scripts/extract_clinical_data.py +0 -97
  173. package/bin/skills/clinical-reports/scripts/format_adverse_events.py +0 -97
  174. package/bin/skills/clinical-reports/scripts/generate_report_template.py +0 -149
  175. package/bin/skills/clinical-reports/scripts/terminology_validator.py +0 -126
  176. package/bin/skills/clinical-reports/scripts/validate_case_report.py +0 -323
  177. package/bin/skills/clinical-reports/scripts/validate_trial_report.py +0 -88
  178. package/bin/skills/clinicaltrials-database/SKILL.md +0 -507
  179. package/bin/skills/clinicaltrials-database/references/api_reference.md +0 -358
  180. package/bin/skills/clinicaltrials-database/scripts/query_clinicaltrials.py +0 -215
  181. package/bin/skills/clinpgx-database/SKILL.md +0 -638
  182. package/bin/skills/clinpgx-database/references/api_reference.md +0 -757
  183. package/bin/skills/clinpgx-database/scripts/query_clinpgx.py +0 -518
  184. package/bin/skills/clinvar-database/SKILL.md +0 -362
  185. package/bin/skills/clinvar-database/references/api_reference.md +0 -227
  186. package/bin/skills/clinvar-database/references/clinical_significance.md +0 -218
  187. package/bin/skills/clinvar-database/references/data_formats.md +0 -358
  188. package/bin/skills/clip/SKILL.md +0 -253
  189. package/bin/skills/clip/references/applications.md +0 -207
  190. package/bin/skills/cobrapy/SKILL.md +0 -463
  191. package/bin/skills/cobrapy/references/api_quick_reference.md +0 -655
  192. package/bin/skills/cobrapy/references/workflows.md +0 -593
  193. package/bin/skills/colab-finetuning/SKILL.md +0 -153
  194. package/bin/skills/colab-finetuning/references/bridge-setup.md +0 -68
  195. package/bin/skills/colab-finetuning/references/gpu-tiers.md +0 -54
  196. package/bin/skills/colab-finetuning/references/troubleshooting.md +0 -79
  197. package/bin/skills/constitutional-ai/SKILL.md +0 -290
  198. package/bin/skills/cosmic-database/SKILL.md +0 -336
  199. package/bin/skills/cosmic-database/references/cosmic_data_reference.md +0 -220
  200. package/bin/skills/cosmic-database/scripts/download_cosmic.py +0 -231
  201. package/bin/skills/crewai/SKILL.md +0 -498
  202. package/bin/skills/crewai/references/flows.md +0 -438
  203. package/bin/skills/crewai/references/tools.md +0 -429
  204. package/bin/skills/crewai/references/troubleshooting.md +0 -480
  205. package/bin/skills/dask/SKILL.md +0 -456
  206. package/bin/skills/dask/references/arrays.md +0 -497
  207. package/bin/skills/dask/references/bags.md +0 -468
  208. package/bin/skills/dask/references/best-practices.md +0 -277
  209. package/bin/skills/dask/references/dataframes.md +0 -368
  210. package/bin/skills/dask/references/futures.md +0 -541
  211. package/bin/skills/dask/references/schedulers.md +0 -504
  212. package/bin/skills/datacommons-client/SKILL.md +0 -255
  213. package/bin/skills/datacommons-client/references/getting_started.md +0 -417
  214. package/bin/skills/datacommons-client/references/node.md +0 -250
  215. package/bin/skills/datacommons-client/references/observation.md +0 -185
  216. package/bin/skills/datacommons-client/references/resolve.md +0 -246
  217. package/bin/skills/datamol/SKILL.md +0 -706
  218. package/bin/skills/datamol/references/conformers_module.md +0 -131
  219. package/bin/skills/datamol/references/core_api.md +0 -130
  220. package/bin/skills/datamol/references/descriptors_viz.md +0 -195
  221. package/bin/skills/datamol/references/fragments_scaffolds.md +0 -174
  222. package/bin/skills/datamol/references/io_module.md +0 -109
  223. package/bin/skills/datamol/references/reactions_data.md +0 -218
  224. package/bin/skills/deepchem/SKILL.md +0 -597
  225. package/bin/skills/deepchem/references/api_reference.md +0 -303
  226. package/bin/skills/deepchem/references/workflows.md +0 -491
  227. package/bin/skills/deepchem/scripts/graph_neural_network.py +0 -338
  228. package/bin/skills/deepchem/scripts/predict_solubility.py +0 -224
  229. package/bin/skills/deepchem/scripts/transfer_learning.py +0 -375
  230. package/bin/skills/deepspeed/SKILL.md +0 -141
  231. package/bin/skills/deepspeed/references/08.md +0 -17
  232. package/bin/skills/deepspeed/references/09.md +0 -173
  233. package/bin/skills/deepspeed/references/2020.md +0 -378
  234. package/bin/skills/deepspeed/references/2023.md +0 -279
  235. package/bin/skills/deepspeed/references/assets.md +0 -179
  236. package/bin/skills/deepspeed/references/index.md +0 -35
  237. package/bin/skills/deepspeed/references/mii.md +0 -118
  238. package/bin/skills/deepspeed/references/other.md +0 -1191
  239. package/bin/skills/deepspeed/references/tutorials.md +0 -6554
  240. package/bin/skills/deeptools/SKILL.md +0 -531
  241. package/bin/skills/deeptools/assets/quick_reference.md +0 -58
  242. package/bin/skills/deeptools/references/effective_genome_sizes.md +0 -116
  243. package/bin/skills/deeptools/references/normalization_methods.md +0 -410
  244. package/bin/skills/deeptools/references/tools_reference.md +0 -533
  245. package/bin/skills/deeptools/references/workflows.md +0 -474
  246. package/bin/skills/deeptools/scripts/validate_files.py +0 -195
  247. package/bin/skills/deeptools/scripts/workflow_generator.py +0 -454
  248. package/bin/skills/denario/SKILL.md +0 -215
  249. package/bin/skills/denario/references/examples.md +0 -494
  250. package/bin/skills/denario/references/installation.md +0 -213
  251. package/bin/skills/denario/references/llm_configuration.md +0 -265
  252. package/bin/skills/denario/references/research_pipeline.md +0 -471
  253. package/bin/skills/diffdock/SKILL.md +0 -483
  254. package/bin/skills/diffdock/assets/batch_template.csv +0 -4
  255. package/bin/skills/diffdock/assets/custom_inference_config.yaml +0 -90
  256. package/bin/skills/diffdock/references/confidence_and_limitations.md +0 -182
  257. package/bin/skills/diffdock/references/parameters_reference.md +0 -163
  258. package/bin/skills/diffdock/references/workflows_examples.md +0 -392
  259. package/bin/skills/diffdock/scripts/analyze_results.py +0 -334
  260. package/bin/skills/diffdock/scripts/prepare_batch_csv.py +0 -254
  261. package/bin/skills/diffdock/scripts/setup_check.py +0 -278
  262. package/bin/skills/dnanexus-integration/SKILL.md +0 -383
  263. package/bin/skills/dnanexus-integration/references/app-development.md +0 -247
  264. package/bin/skills/dnanexus-integration/references/configuration.md +0 -646
  265. package/bin/skills/dnanexus-integration/references/data-operations.md +0 -400
  266. package/bin/skills/dnanexus-integration/references/job-execution.md +0 -412
  267. package/bin/skills/dnanexus-integration/references/python-sdk.md +0 -523
  268. package/bin/skills/document-skills/docx/LICENSE.txt +0 -30
  269. package/bin/skills/document-skills/docx/SKILL.md +0 -233
  270. package/bin/skills/document-skills/docx/docx-js.md +0 -350
  271. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/dml-chart.xsd +0 -1499
  272. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/dml-chartDrawing.xsd +0 -146
  273. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/dml-diagram.xsd +0 -1085
  274. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/dml-lockedCanvas.xsd +0 -11
  275. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/dml-main.xsd +0 -3081
  276. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/dml-picture.xsd +0 -23
  277. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/dml-spreadsheetDrawing.xsd +0 -185
  278. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/dml-wordprocessingDrawing.xsd +0 -287
  279. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/pml.xsd +0 -1676
  280. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/shared-additionalCharacteristics.xsd +0 -28
  281. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/shared-bibliography.xsd +0 -144
  282. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/shared-commonSimpleTypes.xsd +0 -174
  283. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/shared-customXmlDataProperties.xsd +0 -25
  284. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/shared-customXmlSchemaProperties.xsd +0 -18
  285. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesCustom.xsd +0 -59
  286. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesExtended.xsd +0 -56
  287. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesVariantTypes.xsd +0 -195
  288. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/shared-math.xsd +0 -582
  289. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/shared-relationshipReference.xsd +0 -25
  290. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/sml.xsd +0 -4439
  291. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/vml-main.xsd +0 -570
  292. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/vml-officeDrawing.xsd +0 -509
  293. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/vml-presentationDrawing.xsd +0 -12
  294. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/vml-spreadsheetDrawing.xsd +0 -108
  295. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/vml-wordprocessingDrawing.xsd +0 -96
  296. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/wml.xsd +0 -3646
  297. package/bin/skills/document-skills/docx/ooxml/schemas/ISO-IEC29500-4_2016/xml.xsd +0 -116
  298. package/bin/skills/document-skills/docx/ooxml/schemas/ecma/fouth-edition/opc-contentTypes.xsd +0 -42
  299. package/bin/skills/document-skills/docx/ooxml/schemas/ecma/fouth-edition/opc-coreProperties.xsd +0 -50
  300. package/bin/skills/document-skills/docx/ooxml/schemas/ecma/fouth-edition/opc-digSig.xsd +0 -49
  301. package/bin/skills/document-skills/docx/ooxml/schemas/ecma/fouth-edition/opc-relationships.xsd +0 -33
  302. package/bin/skills/document-skills/docx/ooxml/schemas/mce/mc.xsd +0 -75
  303. package/bin/skills/document-skills/docx/ooxml/schemas/microsoft/wml-2010.xsd +0 -560
  304. package/bin/skills/document-skills/docx/ooxml/schemas/microsoft/wml-2012.xsd +0 -67
  305. package/bin/skills/document-skills/docx/ooxml/schemas/microsoft/wml-2018.xsd +0 -14
  306. package/bin/skills/document-skills/docx/ooxml/schemas/microsoft/wml-cex-2018.xsd +0 -20
  307. package/bin/skills/document-skills/docx/ooxml/schemas/microsoft/wml-cid-2016.xsd +0 -13
  308. package/bin/skills/document-skills/docx/ooxml/schemas/microsoft/wml-sdtdatahash-2020.xsd +0 -4
  309. package/bin/skills/document-skills/docx/ooxml/schemas/microsoft/wml-symex-2015.xsd +0 -8
  310. package/bin/skills/document-skills/docx/ooxml/scripts/pack.py +0 -159
  311. package/bin/skills/document-skills/docx/ooxml/scripts/unpack.py +0 -29
  312. package/bin/skills/document-skills/docx/ooxml/scripts/validate.py +0 -69
  313. package/bin/skills/document-skills/docx/ooxml/scripts/validation/__init__.py +0 -15
  314. package/bin/skills/document-skills/docx/ooxml/scripts/validation/base.py +0 -951
  315. package/bin/skills/document-skills/docx/ooxml/scripts/validation/docx.py +0 -274
  316. package/bin/skills/document-skills/docx/ooxml/scripts/validation/pptx.py +0 -315
  317. package/bin/skills/document-skills/docx/ooxml/scripts/validation/redlining.py +0 -279
  318. package/bin/skills/document-skills/docx/ooxml.md +0 -610
  319. package/bin/skills/document-skills/docx/scripts/__init__.py +0 -1
  320. package/bin/skills/document-skills/docx/scripts/document.py +0 -1276
  321. package/bin/skills/document-skills/docx/scripts/templates/comments.xml +0 -3
  322. package/bin/skills/document-skills/docx/scripts/templates/commentsExtended.xml +0 -3
  323. package/bin/skills/document-skills/docx/scripts/templates/commentsExtensible.xml +0 -3
  324. package/bin/skills/document-skills/docx/scripts/templates/commentsIds.xml +0 -3
  325. package/bin/skills/document-skills/docx/scripts/templates/people.xml +0 -3
  326. package/bin/skills/document-skills/docx/scripts/utilities.py +0 -374
  327. package/bin/skills/document-skills/pdf/LICENSE.txt +0 -30
  328. package/bin/skills/document-skills/pdf/SKILL.md +0 -330
  329. package/bin/skills/document-skills/pdf/forms.md +0 -205
  330. package/bin/skills/document-skills/pdf/reference.md +0 -612
  331. package/bin/skills/document-skills/pdf/scripts/check_bounding_boxes.py +0 -70
  332. package/bin/skills/document-skills/pdf/scripts/check_bounding_boxes_test.py +0 -226
  333. package/bin/skills/document-skills/pdf/scripts/check_fillable_fields.py +0 -12
  334. package/bin/skills/document-skills/pdf/scripts/convert_pdf_to_images.py +0 -35
  335. package/bin/skills/document-skills/pdf/scripts/create_validation_image.py +0 -41
  336. package/bin/skills/document-skills/pdf/scripts/extract_form_field_info.py +0 -152
  337. package/bin/skills/document-skills/pdf/scripts/fill_fillable_fields.py +0 -114
  338. package/bin/skills/document-skills/pdf/scripts/fill_pdf_form_with_annotations.py +0 -108
  339. package/bin/skills/document-skills/pptx/LICENSE.txt +0 -30
  340. package/bin/skills/document-skills/pptx/SKILL.md +0 -520
  341. package/bin/skills/document-skills/pptx/html2pptx.md +0 -625
  342. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/dml-chart.xsd +0 -1499
  343. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/dml-chartDrawing.xsd +0 -146
  344. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/dml-diagram.xsd +0 -1085
  345. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/dml-lockedCanvas.xsd +0 -11
  346. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/dml-main.xsd +0 -3081
  347. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/dml-picture.xsd +0 -23
  348. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/dml-spreadsheetDrawing.xsd +0 -185
  349. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/dml-wordprocessingDrawing.xsd +0 -287
  350. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/pml.xsd +0 -1676
  351. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/shared-additionalCharacteristics.xsd +0 -28
  352. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/shared-bibliography.xsd +0 -144
  353. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/shared-commonSimpleTypes.xsd +0 -174
  354. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/shared-customXmlDataProperties.xsd +0 -25
  355. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/shared-customXmlSchemaProperties.xsd +0 -18
  356. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesCustom.xsd +0 -59
  357. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesExtended.xsd +0 -56
  358. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesVariantTypes.xsd +0 -195
  359. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/shared-math.xsd +0 -582
  360. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/shared-relationshipReference.xsd +0 -25
  361. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/sml.xsd +0 -4439
  362. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/vml-main.xsd +0 -570
  363. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/vml-officeDrawing.xsd +0 -509
  364. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/vml-presentationDrawing.xsd +0 -12
  365. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/vml-spreadsheetDrawing.xsd +0 -108
  366. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/vml-wordprocessingDrawing.xsd +0 -96
  367. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/wml.xsd +0 -3646
  368. package/bin/skills/document-skills/pptx/ooxml/schemas/ISO-IEC29500-4_2016/xml.xsd +0 -116
  369. package/bin/skills/document-skills/pptx/ooxml/schemas/ecma/fouth-edition/opc-contentTypes.xsd +0 -42
  370. package/bin/skills/document-skills/pptx/ooxml/schemas/ecma/fouth-edition/opc-coreProperties.xsd +0 -50
  371. package/bin/skills/document-skills/pptx/ooxml/schemas/ecma/fouth-edition/opc-digSig.xsd +0 -49
  372. package/bin/skills/document-skills/pptx/ooxml/schemas/ecma/fouth-edition/opc-relationships.xsd +0 -33
  373. package/bin/skills/document-skills/pptx/ooxml/schemas/mce/mc.xsd +0 -75
  374. package/bin/skills/document-skills/pptx/ooxml/schemas/microsoft/wml-2010.xsd +0 -560
  375. package/bin/skills/document-skills/pptx/ooxml/schemas/microsoft/wml-2012.xsd +0 -67
  376. package/bin/skills/document-skills/pptx/ooxml/schemas/microsoft/wml-2018.xsd +0 -14
  377. package/bin/skills/document-skills/pptx/ooxml/schemas/microsoft/wml-cex-2018.xsd +0 -20
  378. package/bin/skills/document-skills/pptx/ooxml/schemas/microsoft/wml-cid-2016.xsd +0 -13
  379. package/bin/skills/document-skills/pptx/ooxml/schemas/microsoft/wml-sdtdatahash-2020.xsd +0 -4
  380. package/bin/skills/document-skills/pptx/ooxml/schemas/microsoft/wml-symex-2015.xsd +0 -8
  381. package/bin/skills/document-skills/pptx/ooxml/scripts/pack.py +0 -159
  382. package/bin/skills/document-skills/pptx/ooxml/scripts/unpack.py +0 -29
  383. package/bin/skills/document-skills/pptx/ooxml/scripts/validate.py +0 -69
  384. package/bin/skills/document-skills/pptx/ooxml/scripts/validation/__init__.py +0 -15
  385. package/bin/skills/document-skills/pptx/ooxml/scripts/validation/base.py +0 -951
  386. package/bin/skills/document-skills/pptx/ooxml/scripts/validation/docx.py +0 -274
  387. package/bin/skills/document-skills/pptx/ooxml/scripts/validation/pptx.py +0 -315
  388. package/bin/skills/document-skills/pptx/ooxml/scripts/validation/redlining.py +0 -279
  389. package/bin/skills/document-skills/pptx/ooxml.md +0 -427
  390. package/bin/skills/document-skills/pptx/scripts/html2pptx.js +0 -979
  391. package/bin/skills/document-skills/pptx/scripts/inventory.py +0 -1020
  392. package/bin/skills/document-skills/pptx/scripts/rearrange.py +0 -231
  393. package/bin/skills/document-skills/pptx/scripts/replace.py +0 -385
  394. package/bin/skills/document-skills/pptx/scripts/thumbnail.py +0 -450
  395. package/bin/skills/document-skills/xlsx/LICENSE.txt +0 -30
  396. package/bin/skills/document-skills/xlsx/SKILL.md +0 -325
  397. package/bin/skills/document-skills/xlsx/recalc.py +0 -178
  398. package/bin/skills/drugbank-database/SKILL.md +0 -190
  399. package/bin/skills/drugbank-database/references/chemical-analysis.md +0 -590
  400. package/bin/skills/drugbank-database/references/data-access.md +0 -242
  401. package/bin/skills/drugbank-database/references/drug-queries.md +0 -386
  402. package/bin/skills/drugbank-database/references/interactions.md +0 -425
  403. package/bin/skills/drugbank-database/references/targets-pathways.md +0 -518
  404. package/bin/skills/drugbank-database/scripts/drugbank_helper.py +0 -350
  405. package/bin/skills/dspy/SKILL.md +0 -590
  406. package/bin/skills/dspy/references/examples.md +0 -663
  407. package/bin/skills/dspy/references/modules.md +0 -475
  408. package/bin/skills/dspy/references/optimizers.md +0 -566
  409. package/bin/skills/ena-database/SKILL.md +0 -204
  410. package/bin/skills/ena-database/references/api_reference.md +0 -490
  411. package/bin/skills/ensembl-database/SKILL.md +0 -311
  412. package/bin/skills/ensembl-database/references/api_endpoints.md +0 -346
  413. package/bin/skills/ensembl-database/scripts/ensembl_query.py +0 -427
  414. package/bin/skills/esm/SKILL.md +0 -306
  415. package/bin/skills/esm/references/esm-c-api.md +0 -583
  416. package/bin/skills/esm/references/esm3-api.md +0 -452
  417. package/bin/skills/esm/references/forge-api.md +0 -657
  418. package/bin/skills/esm/references/workflows.md +0 -685
  419. package/bin/skills/etetoolkit/SKILL.md +0 -623
  420. package/bin/skills/etetoolkit/references/api_reference.md +0 -583
  421. package/bin/skills/etetoolkit/references/visualization.md +0 -783
  422. package/bin/skills/etetoolkit/references/workflows.md +0 -774
  423. package/bin/skills/etetoolkit/scripts/quick_visualize.py +0 -214
  424. package/bin/skills/etetoolkit/scripts/tree_operations.py +0 -229
  425. package/bin/skills/exploratory-data-analysis/SKILL.md +0 -446
  426. package/bin/skills/exploratory-data-analysis/assets/report_template.md +0 -196
  427. package/bin/skills/exploratory-data-analysis/references/bioinformatics_genomics_formats.md +0 -664
  428. package/bin/skills/exploratory-data-analysis/references/chemistry_molecular_formats.md +0 -664
  429. package/bin/skills/exploratory-data-analysis/references/general_scientific_formats.md +0 -518
  430. package/bin/skills/exploratory-data-analysis/references/microscopy_imaging_formats.md +0 -620
  431. package/bin/skills/exploratory-data-analysis/references/proteomics_metabolomics_formats.md +0 -517
  432. package/bin/skills/exploratory-data-analysis/references/spectroscopy_analytical_formats.md +0 -633
  433. package/bin/skills/exploratory-data-analysis/scripts/eda_analyzer.py +0 -547
  434. package/bin/skills/faiss/SKILL.md +0 -221
  435. package/bin/skills/faiss/references/index_types.md +0 -280
  436. package/bin/skills/fda-database/SKILL.md +0 -518
  437. package/bin/skills/fda-database/references/animal_veterinary.md +0 -377
  438. package/bin/skills/fda-database/references/api_basics.md +0 -687
  439. package/bin/skills/fda-database/references/devices.md +0 -632
  440. package/bin/skills/fda-database/references/drugs.md +0 -468
  441. package/bin/skills/fda-database/references/foods.md +0 -374
  442. package/bin/skills/fda-database/references/other.md +0 -472
  443. package/bin/skills/fda-database/scripts/fda_examples.py +0 -335
  444. package/bin/skills/fda-database/scripts/fda_query.py +0 -440
  445. package/bin/skills/fireworks-ai/SKILL.md +0 -665
  446. package/bin/skills/flash-attention/SKILL.md +0 -367
  447. package/bin/skills/flash-attention/references/benchmarks.md +0 -215
  448. package/bin/skills/flash-attention/references/transformers-integration.md +0 -293
  449. package/bin/skills/flowio/SKILL.md +0 -608
  450. package/bin/skills/flowio/references/api_reference.md +0 -372
  451. package/bin/skills/fluidsim/SKILL.md +0 -349
  452. package/bin/skills/fluidsim/references/advanced_features.md +0 -398
  453. package/bin/skills/fluidsim/references/installation.md +0 -68
  454. package/bin/skills/fluidsim/references/output_analysis.md +0 -283
  455. package/bin/skills/fluidsim/references/parameters.md +0 -198
  456. package/bin/skills/fluidsim/references/simulation_workflow.md +0 -172
  457. package/bin/skills/fluidsim/references/solvers.md +0 -94
  458. package/bin/skills/fred-economic-data/SKILL.md +0 -433
  459. package/bin/skills/fred-economic-data/references/api_basics.md +0 -212
  460. package/bin/skills/fred-economic-data/references/categories.md +0 -442
  461. package/bin/skills/fred-economic-data/references/geofred.md +0 -588
  462. package/bin/skills/fred-economic-data/references/releases.md +0 -642
  463. package/bin/skills/fred-economic-data/references/series.md +0 -584
  464. package/bin/skills/fred-economic-data/references/sources.md +0 -423
  465. package/bin/skills/fred-economic-data/references/tags.md +0 -485
  466. package/bin/skills/fred-economic-data/scripts/fred_examples.py +0 -354
  467. package/bin/skills/fred-economic-data/scripts/fred_query.py +0 -590
  468. package/bin/skills/gene-database/SKILL.md +0 -179
  469. package/bin/skills/gene-database/references/api_reference.md +0 -404
  470. package/bin/skills/gene-database/references/common_workflows.md +0 -428
  471. package/bin/skills/gene-database/scripts/batch_gene_lookup.py +0 -298
  472. package/bin/skills/gene-database/scripts/fetch_gene_data.py +0 -277
  473. package/bin/skills/gene-database/scripts/query_gene.py +0 -251
  474. package/bin/skills/generate-image/SKILL.md +0 -178
  475. package/bin/skills/generate-image/scripts/generate_image.py +0 -254
  476. package/bin/skills/geniml/SKILL.md +0 -318
  477. package/bin/skills/geniml/references/bedspace.md +0 -127
  478. package/bin/skills/geniml/references/consensus_peaks.md +0 -238
  479. package/bin/skills/geniml/references/region2vec.md +0 -90
  480. package/bin/skills/geniml/references/scembed.md +0 -197
  481. package/bin/skills/geniml/references/utilities.md +0 -385
  482. package/bin/skills/geo-database/SKILL.md +0 -815
  483. package/bin/skills/geo-database/references/geo_reference.md +0 -829
  484. package/bin/skills/geopandas/SKILL.md +0 -251
  485. package/bin/skills/geopandas/references/crs-management.md +0 -243
  486. package/bin/skills/geopandas/references/data-io.md +0 -165
  487. package/bin/skills/geopandas/references/data-structures.md +0 -70
  488. package/bin/skills/geopandas/references/geometric-operations.md +0 -221
  489. package/bin/skills/geopandas/references/spatial-analysis.md +0 -184
  490. package/bin/skills/geopandas/references/visualization.md +0 -243
  491. package/bin/skills/get-available-resources/SKILL.md +0 -277
  492. package/bin/skills/get-available-resources/scripts/detect_resources.py +0 -401
  493. package/bin/skills/gget/SKILL.md +0 -871
  494. package/bin/skills/gget/references/database_info.md +0 -300
  495. package/bin/skills/gget/references/module_reference.md +0 -467
  496. package/bin/skills/gget/references/workflows.md +0 -814
  497. package/bin/skills/gget/scripts/batch_sequence_analysis.py +0 -191
  498. package/bin/skills/gget/scripts/enrichment_pipeline.py +0 -235
  499. package/bin/skills/gget/scripts/gene_analysis.py +0 -161
  500. package/bin/skills/gguf/SKILL.md +0 -427
  501. package/bin/skills/gguf/references/advanced-usage.md +0 -504
  502. package/bin/skills/gguf/references/troubleshooting.md +0 -442
  503. package/bin/skills/gptq/SKILL.md +0 -450
  504. package/bin/skills/gptq/references/calibration.md +0 -337
  505. package/bin/skills/gptq/references/integration.md +0 -129
  506. package/bin/skills/gptq/references/troubleshooting.md +0 -95
  507. package/bin/skills/groq/SKILL.md +0 -347
  508. package/bin/skills/grpo-rl-training/README.md +0 -97
  509. package/bin/skills/grpo-rl-training/SKILL.md +0 -572
  510. package/bin/skills/grpo-rl-training/examples/reward_functions_library.py +0 -393
  511. package/bin/skills/grpo-rl-training/templates/basic_grpo_training.py +0 -228
  512. package/bin/skills/gtars/SKILL.md +0 -285
  513. package/bin/skills/gtars/references/cli.md +0 -222
  514. package/bin/skills/gtars/references/coverage.md +0 -172
  515. package/bin/skills/gtars/references/overlap.md +0 -156
  516. package/bin/skills/gtars/references/python-api.md +0 -211
  517. package/bin/skills/gtars/references/refget.md +0 -147
  518. package/bin/skills/gtars/references/tokenizers.md +0 -103
  519. package/bin/skills/guidance/SKILL.md +0 -572
  520. package/bin/skills/guidance/references/backends.md +0 -554
  521. package/bin/skills/guidance/references/constraints.md +0 -674
  522. package/bin/skills/guidance/references/examples.md +0 -767
  523. package/bin/skills/gwas-database/SKILL.md +0 -608
  524. package/bin/skills/gwas-database/references/api_reference.md +0 -793
  525. package/bin/skills/histolab/SKILL.md +0 -678
  526. package/bin/skills/histolab/references/filters_preprocessing.md +0 -514
  527. package/bin/skills/histolab/references/slide_management.md +0 -172
  528. package/bin/skills/histolab/references/tile_extraction.md +0 -421
  529. package/bin/skills/histolab/references/tissue_masks.md +0 -251
  530. package/bin/skills/histolab/references/visualization.md +0 -547
  531. package/bin/skills/hmdb-database/SKILL.md +0 -196
  532. package/bin/skills/hmdb-database/references/hmdb_data_fields.md +0 -267
  533. package/bin/skills/hqq/SKILL.md +0 -445
  534. package/bin/skills/hqq/references/advanced-usage.md +0 -528
  535. package/bin/skills/hqq/references/troubleshooting.md +0 -503
  536. package/bin/skills/hugging-face-cli/SKILL.md +0 -191
  537. package/bin/skills/hugging-face-cli/references/commands.md +0 -954
  538. package/bin/skills/hugging-face-cli/references/examples.md +0 -374
  539. package/bin/skills/hugging-face-datasets/SKILL.md +0 -547
  540. package/bin/skills/hugging-face-datasets/examples/diverse_training_examples.json +0 -239
  541. package/bin/skills/hugging-face-datasets/examples/system_prompt_template.txt +0 -196
  542. package/bin/skills/hugging-face-datasets/examples/training_examples.json +0 -176
  543. package/bin/skills/hugging-face-datasets/scripts/dataset_manager.py +0 -522
  544. package/bin/skills/hugging-face-datasets/scripts/sql_manager.py +0 -844
  545. package/bin/skills/hugging-face-datasets/templates/chat.json +0 -55
  546. package/bin/skills/hugging-face-datasets/templates/classification.json +0 -62
  547. package/bin/skills/hugging-face-datasets/templates/completion.json +0 -51
  548. package/bin/skills/hugging-face-datasets/templates/custom.json +0 -75
  549. package/bin/skills/hugging-face-datasets/templates/qa.json +0 -54
  550. package/bin/skills/hugging-face-datasets/templates/tabular.json +0 -81
  551. package/bin/skills/hugging-face-evaluation/SKILL.md +0 -656
  552. package/bin/skills/hugging-face-evaluation/examples/.env.example +0 -7
  553. package/bin/skills/hugging-face-evaluation/examples/USAGE_EXAMPLES.md +0 -382
  554. package/bin/skills/hugging-face-evaluation/examples/artificial_analysis_to_hub.py +0 -141
  555. package/bin/skills/hugging-face-evaluation/examples/example_readme_tables.md +0 -135
  556. package/bin/skills/hugging-face-evaluation/examples/metric_mapping.json +0 -50
  557. package/bin/skills/hugging-face-evaluation/requirements.txt +0 -20
  558. package/bin/skills/hugging-face-evaluation/scripts/evaluation_manager.py +0 -1374
  559. package/bin/skills/hugging-face-evaluation/scripts/inspect_eval_uv.py +0 -104
  560. package/bin/skills/hugging-face-evaluation/scripts/inspect_vllm_uv.py +0 -317
  561. package/bin/skills/hugging-face-evaluation/scripts/lighteval_vllm_uv.py +0 -303
  562. package/bin/skills/hugging-face-evaluation/scripts/run_eval_job.py +0 -98
  563. package/bin/skills/hugging-face-evaluation/scripts/run_vllm_eval_job.py +0 -331
  564. package/bin/skills/hugging-face-evaluation/scripts/test_extraction.py +0 -206
  565. package/bin/skills/hugging-face-jobs/SKILL.md +0 -1040
  566. package/bin/skills/hugging-face-jobs/index.html +0 -216
  567. package/bin/skills/hugging-face-jobs/references/hardware_guide.md +0 -336
  568. package/bin/skills/hugging-face-jobs/references/hub_saving.md +0 -352
  569. package/bin/skills/hugging-face-jobs/references/token_usage.md +0 -546
  570. package/bin/skills/hugging-face-jobs/references/troubleshooting.md +0 -475
  571. package/bin/skills/hugging-face-jobs/scripts/cot-self-instruct.py +0 -718
  572. package/bin/skills/hugging-face-jobs/scripts/finepdfs-stats.py +0 -546
  573. package/bin/skills/hugging-face-jobs/scripts/generate-responses.py +0 -587
  574. package/bin/skills/hugging-face-model-trainer/SKILL.md +0 -710
  575. package/bin/skills/hugging-face-model-trainer/references/gguf_conversion.md +0 -296
  576. package/bin/skills/hugging-face-model-trainer/references/hardware_guide.md +0 -283
  577. package/bin/skills/hugging-face-model-trainer/references/hub_saving.md +0 -364
  578. package/bin/skills/hugging-face-model-trainer/references/reliability_principles.md +0 -371
  579. package/bin/skills/hugging-face-model-trainer/references/trackio_guide.md +0 -189
  580. package/bin/skills/hugging-face-model-trainer/references/training_methods.md +0 -150
  581. package/bin/skills/hugging-face-model-trainer/references/training_patterns.md +0 -203
  582. package/bin/skills/hugging-face-model-trainer/references/troubleshooting.md +0 -282
  583. package/bin/skills/hugging-face-model-trainer/scripts/convert_to_gguf.py +0 -424
  584. package/bin/skills/hugging-face-model-trainer/scripts/dataset_inspector.py +0 -417
  585. package/bin/skills/hugging-face-model-trainer/scripts/estimate_cost.py +0 -150
  586. package/bin/skills/hugging-face-model-trainer/scripts/train_dpo_example.py +0 -106
  587. package/bin/skills/hugging-face-model-trainer/scripts/train_grpo_example.py +0 -89
  588. package/bin/skills/hugging-face-model-trainer/scripts/train_sft_example.py +0 -122
  589. package/bin/skills/hugging-face-paper-publisher/SKILL.md +0 -627
  590. package/bin/skills/hugging-face-paper-publisher/examples/example_usage.md +0 -327
  591. package/bin/skills/hugging-face-paper-publisher/references/quick_reference.md +0 -216
  592. package/bin/skills/hugging-face-paper-publisher/scripts/paper_manager.py +0 -508
  593. package/bin/skills/hugging-face-paper-publisher/templates/arxiv.md +0 -299
  594. package/bin/skills/hugging-face-paper-publisher/templates/ml-report.md +0 -358
  595. package/bin/skills/hugging-face-paper-publisher/templates/modern.md +0 -319
  596. package/bin/skills/hugging-face-paper-publisher/templates/standard.md +0 -201
  597. package/bin/skills/hugging-face-tool-builder/SKILL.md +0 -115
  598. package/bin/skills/hugging-face-tool-builder/references/baseline_hf_api.py +0 -57
  599. package/bin/skills/hugging-face-tool-builder/references/baseline_hf_api.sh +0 -40
  600. package/bin/skills/hugging-face-tool-builder/references/baseline_hf_api.tsx +0 -57
  601. package/bin/skills/hugging-face-tool-builder/references/find_models_by_paper.sh +0 -230
  602. package/bin/skills/hugging-face-tool-builder/references/hf_enrich_models.sh +0 -96
  603. package/bin/skills/hugging-face-tool-builder/references/hf_model_card_frontmatter.sh +0 -188
  604. package/bin/skills/hugging-face-tool-builder/references/hf_model_papers_auth.sh +0 -171
  605. package/bin/skills/hugging-face-trackio/.claude-plugin/plugin.json +0 -19
  606. package/bin/skills/hugging-face-trackio/SKILL.md +0 -65
  607. package/bin/skills/hugging-face-trackio/references/logging_metrics.md +0 -206
  608. package/bin/skills/hugging-face-trackio/references/retrieving_metrics.md +0 -223
  609. package/bin/skills/huggingface-tokenizers/SKILL.md +0 -516
  610. package/bin/skills/huggingface-tokenizers/references/algorithms.md +0 -653
  611. package/bin/skills/huggingface-tokenizers/references/integration.md +0 -637
  612. package/bin/skills/huggingface-tokenizers/references/pipeline.md +0 -723
  613. package/bin/skills/huggingface-tokenizers/references/training.md +0 -565
  614. package/bin/skills/hypogenic/SKILL.md +0 -655
  615. package/bin/skills/hypogenic/references/config_template.yaml +0 -150
  616. package/bin/skills/hypothesis-generation/SKILL.md +0 -293
  617. package/bin/skills/hypothesis-generation/assets/FORMATTING_GUIDE.md +0 -672
  618. package/bin/skills/hypothesis-generation/assets/hypothesis_generation.sty +0 -307
  619. package/bin/skills/hypothesis-generation/assets/hypothesis_report_template.tex +0 -572
  620. package/bin/skills/hypothesis-generation/references/experimental_design_patterns.md +0 -329
  621. package/bin/skills/hypothesis-generation/references/hypothesis_quality_criteria.md +0 -198
  622. package/bin/skills/hypothesis-generation/references/literature_search_strategies.md +0 -622
  623. package/bin/skills/imaging-data-commons/SKILL.md +0 -1182
  624. package/bin/skills/imaging-data-commons/references/bigquery_guide.md +0 -556
  625. package/bin/skills/imaging-data-commons/references/cli_guide.md +0 -272
  626. package/bin/skills/imaging-data-commons/references/cloud_storage_guide.md +0 -333
  627. package/bin/skills/imaging-data-commons/references/dicomweb_guide.md +0 -399
  628. package/bin/skills/infographics/SKILL.md +0 -563
  629. package/bin/skills/infographics/references/color_palettes.md +0 -496
  630. package/bin/skills/infographics/references/design_principles.md +0 -636
  631. package/bin/skills/infographics/references/infographic_types.md +0 -907
  632. package/bin/skills/infographics/scripts/generate_infographic.py +0 -234
  633. package/bin/skills/infographics/scripts/generate_infographic_ai.py +0 -1290
  634. package/bin/skills/instructor/SKILL.md +0 -740
  635. package/bin/skills/instructor/references/examples.md +0 -107
  636. package/bin/skills/instructor/references/providers.md +0 -70
  637. package/bin/skills/instructor/references/validation.md +0 -606
  638. package/bin/skills/iso-13485-certification/SKILL.md +0 -680
  639. package/bin/skills/iso-13485-certification/assets/templates/procedures/CAPA-procedure-template.md +0 -453
  640. package/bin/skills/iso-13485-certification/assets/templates/procedures/document-control-procedure-template.md +0 -567
  641. package/bin/skills/iso-13485-certification/assets/templates/quality-manual-template.md +0 -521
  642. package/bin/skills/iso-13485-certification/references/gap-analysis-checklist.md +0 -568
  643. package/bin/skills/iso-13485-certification/references/iso-13485-requirements.md +0 -610
  644. package/bin/skills/iso-13485-certification/references/mandatory-documents.md +0 -606
  645. package/bin/skills/iso-13485-certification/references/quality-manual-guide.md +0 -688
  646. package/bin/skills/iso-13485-certification/scripts/gap_analyzer.py +0 -440
  647. package/bin/skills/kegg-database/SKILL.md +0 -377
  648. package/bin/skills/kegg-database/references/kegg_reference.md +0 -326
  649. package/bin/skills/kegg-database/scripts/kegg_api.py +0 -251
  650. package/bin/skills/knowledge-distillation/SKILL.md +0 -458
  651. package/bin/skills/knowledge-distillation/references/minillm.md +0 -334
  652. package/bin/skills/labarchive-integration/SKILL.md +0 -268
  653. package/bin/skills/labarchive-integration/references/api_reference.md +0 -342
  654. package/bin/skills/labarchive-integration/references/authentication_guide.md +0 -357
  655. package/bin/skills/labarchive-integration/references/integrations.md +0 -425
  656. package/bin/skills/labarchive-integration/scripts/entry_operations.py +0 -334
  657. package/bin/skills/labarchive-integration/scripts/notebook_operations.py +0 -269
  658. package/bin/skills/labarchive-integration/scripts/setup_config.py +0 -205
  659. package/bin/skills/lambda-labs/SKILL.md +0 -545
  660. package/bin/skills/lambda-labs/references/advanced-usage.md +0 -611
  661. package/bin/skills/lambda-labs/references/troubleshooting.md +0 -530
  662. package/bin/skills/lamindb/SKILL.md +0 -390
  663. package/bin/skills/lamindb/references/annotation-validation.md +0 -513
  664. package/bin/skills/lamindb/references/core-concepts.md +0 -380
  665. package/bin/skills/lamindb/references/data-management.md +0 -433
  666. package/bin/skills/lamindb/references/integrations.md +0 -642
  667. package/bin/skills/lamindb/references/ontologies.md +0 -497
  668. package/bin/skills/lamindb/references/setup-deployment.md +0 -733
  669. package/bin/skills/langchain/SKILL.md +0 -480
  670. package/bin/skills/langchain/references/agents.md +0 -499
  671. package/bin/skills/langchain/references/integration.md +0 -562
  672. package/bin/skills/langchain/references/rag.md +0 -600
  673. package/bin/skills/langsmith/SKILL.md +0 -422
  674. package/bin/skills/langsmith/references/advanced-usage.md +0 -548
  675. package/bin/skills/langsmith/references/troubleshooting.md +0 -537
  676. package/bin/skills/latchbio-integration/SKILL.md +0 -353
  677. package/bin/skills/latchbio-integration/references/data-management.md +0 -427
  678. package/bin/skills/latchbio-integration/references/resource-configuration.md +0 -429
  679. package/bin/skills/latchbio-integration/references/verified-workflows.md +0 -487
  680. package/bin/skills/latchbio-integration/references/workflow-creation.md +0 -254
  681. package/bin/skills/latex-posters/README.md +0 -417
  682. package/bin/skills/latex-posters/SKILL.md +0 -1602
  683. package/bin/skills/latex-posters/assets/baposter_template.tex +0 -257
  684. package/bin/skills/latex-posters/assets/beamerposter_template.tex +0 -244
  685. package/bin/skills/latex-posters/assets/poster_quality_checklist.md +0 -358
  686. package/bin/skills/latex-posters/assets/tikzposter_template.tex +0 -251
  687. package/bin/skills/latex-posters/references/latex_poster_packages.md +0 -745
  688. package/bin/skills/latex-posters/references/poster_content_guide.md +0 -748
  689. package/bin/skills/latex-posters/references/poster_design_principles.md +0 -806
  690. package/bin/skills/latex-posters/references/poster_layout_design.md +0 -900
  691. package/bin/skills/latex-posters/scripts/review_poster.sh +0 -214
  692. package/bin/skills/literature-review/SKILL.md +0 -641
  693. package/bin/skills/literature-review/assets/review_template.md +0 -412
  694. package/bin/skills/literature-review/references/citation_styles.md +0 -166
  695. package/bin/skills/literature-review/references/database_strategies.md +0 -455
  696. package/bin/skills/literature-review/scripts/generate_pdf.py +0 -184
  697. package/bin/skills/literature-review/scripts/search_databases.py +0 -310
  698. package/bin/skills/literature-review/scripts/verify_citations.py +0 -218
  699. package/bin/skills/litgpt/SKILL.md +0 -469
  700. package/bin/skills/litgpt/references/custom-models.md +0 -568
  701. package/bin/skills/litgpt/references/distributed-training.md +0 -451
  702. package/bin/skills/litgpt/references/supported-models.md +0 -336
  703. package/bin/skills/litgpt/references/training-recipes.md +0 -619
  704. package/bin/skills/llama-cpp/SKILL.md +0 -258
  705. package/bin/skills/llama-cpp/references/optimization.md +0 -89
  706. package/bin/skills/llama-cpp/references/quantization.md +0 -213
  707. package/bin/skills/llama-cpp/references/server.md +0 -125
  708. package/bin/skills/llama-factory/SKILL.md +0 -80
  709. package/bin/skills/llama-factory/references/_images.md +0 -23
  710. package/bin/skills/llama-factory/references/advanced.md +0 -1055
  711. package/bin/skills/llama-factory/references/getting_started.md +0 -349
  712. package/bin/skills/llama-factory/references/index.md +0 -19
  713. package/bin/skills/llama-factory/references/other.md +0 -31
  714. package/bin/skills/llamaguard/SKILL.md +0 -337
  715. package/bin/skills/llamaindex/SKILL.md +0 -569
  716. package/bin/skills/llamaindex/references/agents.md +0 -83
  717. package/bin/skills/llamaindex/references/data_connectors.md +0 -108
  718. package/bin/skills/llamaindex/references/query_engines.md +0 -406
  719. package/bin/skills/llava/SKILL.md +0 -304
  720. package/bin/skills/llava/references/training.md +0 -197
  721. package/bin/skills/llm-as-judge-evaluation/SKILL.md +0 -385
  722. package/bin/skills/llm-as-judge-evaluation/references/pairwise-comparison.md +0 -95
  723. package/bin/skills/llm-as-judge-evaluation/references/scoring-rubrics.md +0 -169
  724. package/bin/skills/lm-evaluation-harness/SKILL.md +0 -490
  725. package/bin/skills/lm-evaluation-harness/references/api-evaluation.md +0 -490
  726. package/bin/skills/lm-evaluation-harness/references/benchmark-guide.md +0 -488
  727. package/bin/skills/lm-evaluation-harness/references/custom-tasks.md +0 -602
  728. package/bin/skills/lm-evaluation-harness/references/distributed-eval.md +0 -519
  729. package/bin/skills/long-context/SKILL.md +0 -536
  730. package/bin/skills/long-context/references/extension_methods.md +0 -468
  731. package/bin/skills/long-context/references/fine_tuning.md +0 -611
  732. package/bin/skills/long-context/references/rope.md +0 -402
  733. package/bin/skills/mamba/SKILL.md +0 -260
  734. package/bin/skills/mamba/references/architecture-details.md +0 -206
  735. package/bin/skills/mamba/references/benchmarks.md +0 -255
  736. package/bin/skills/mamba/references/training-guide.md +0 -388
  737. package/bin/skills/market-research-reports/SKILL.md +0 -904
  738. package/bin/skills/market-research-reports/assets/FORMATTING_GUIDE.md +0 -428
  739. package/bin/skills/market-research-reports/assets/market_report_template.tex +0 -1380
  740. package/bin/skills/market-research-reports/assets/market_research.sty +0 -564
  741. package/bin/skills/market-research-reports/references/data_analysis_patterns.md +0 -548
  742. package/bin/skills/market-research-reports/references/report_structure_guide.md +0 -999
  743. package/bin/skills/market-research-reports/references/visual_generation_guide.md +0 -1077
  744. package/bin/skills/market-research-reports/scripts/generate_market_visuals.py +0 -472
  745. package/bin/skills/markitdown/INSTALLATION_GUIDE.md +0 -318
  746. package/bin/skills/markitdown/LICENSE.txt +0 -22
  747. package/bin/skills/markitdown/OPENROUTER_INTEGRATION.md +0 -359
  748. package/bin/skills/markitdown/QUICK_REFERENCE.md +0 -309
  749. package/bin/skills/markitdown/README.md +0 -184
  750. package/bin/skills/markitdown/SKILL.md +0 -486
  751. package/bin/skills/markitdown/SKILL_SUMMARY.md +0 -307
  752. package/bin/skills/markitdown/assets/example_usage.md +0 -463
  753. package/bin/skills/markitdown/references/api_reference.md +0 -399
  754. package/bin/skills/markitdown/references/file_formats.md +0 -542
  755. package/bin/skills/markitdown/scripts/batch_convert.py +0 -195
  756. package/bin/skills/markitdown/scripts/convert_literature.py +0 -262
  757. package/bin/skills/markitdown/scripts/convert_with_ai.py +0 -224
  758. package/bin/skills/matchms/SKILL.md +0 -203
  759. package/bin/skills/matchms/references/filtering.md +0 -288
  760. package/bin/skills/matchms/references/importing_exporting.md +0 -416
  761. package/bin/skills/matchms/references/similarity.md +0 -380
  762. package/bin/skills/matchms/references/workflows.md +0 -647
  763. package/bin/skills/matlab/SKILL.md +0 -376
  764. package/bin/skills/matlab/references/data-import-export.md +0 -479
  765. package/bin/skills/matlab/references/executing-scripts.md +0 -444
  766. package/bin/skills/matlab/references/graphics-visualization.md +0 -579
  767. package/bin/skills/matlab/references/mathematics.md +0 -553
  768. package/bin/skills/matlab/references/matrices-arrays.md +0 -349
  769. package/bin/skills/matlab/references/octave-compatibility.md +0 -544
  770. package/bin/skills/matlab/references/programming.md +0 -672
  771. package/bin/skills/matlab/references/python-integration.md +0 -433
  772. package/bin/skills/matplotlib/SKILL.md +0 -361
  773. package/bin/skills/matplotlib/references/api_reference.md +0 -412
  774. package/bin/skills/matplotlib/references/common_issues.md +0 -563
  775. package/bin/skills/matplotlib/references/plot_types.md +0 -476
  776. package/bin/skills/matplotlib/references/styling_guide.md +0 -589
  777. package/bin/skills/matplotlib/scripts/plot_template.py +0 -401
  778. package/bin/skills/matplotlib/scripts/style_configurator.py +0 -409
  779. package/bin/skills/medchem/SKILL.md +0 -406
  780. package/bin/skills/medchem/references/api_guide.md +0 -600
  781. package/bin/skills/medchem/references/rules_catalog.md +0 -604
  782. package/bin/skills/medchem/scripts/filter_molecules.py +0 -418
  783. package/bin/skills/megatron-core/SKILL.md +0 -366
  784. package/bin/skills/megatron-core/references/benchmarks.md +0 -249
  785. package/bin/skills/megatron-core/references/parallelism-guide.md +0 -404
  786. package/bin/skills/megatron-core/references/production-examples.md +0 -473
  787. package/bin/skills/megatron-core/references/training-recipes.md +0 -547
  788. package/bin/skills/metabolomics-workbench-database/SKILL.md +0 -259
  789. package/bin/skills/metabolomics-workbench-database/references/api_reference.md +0 -494
  790. package/bin/skills/miles/SKILL.md +0 -315
  791. package/bin/skills/miles/references/api-reference.md +0 -141
  792. package/bin/skills/miles/references/troubleshooting.md +0 -352
  793. package/bin/skills/ml-paper-writing/SKILL.md +0 -937
  794. package/bin/skills/ml-paper-writing/references/checklists.md +0 -361
  795. package/bin/skills/ml-paper-writing/references/citation-workflow.md +0 -562
  796. package/bin/skills/ml-paper-writing/references/reviewer-guidelines.md +0 -367
  797. package/bin/skills/ml-paper-writing/references/sources.md +0 -159
  798. package/bin/skills/ml-paper-writing/references/writing-guide.md +0 -476
  799. package/bin/skills/ml-paper-writing/templates/README.md +0 -251
  800. package/bin/skills/ml-paper-writing/templates/aaai2026/README.md +0 -534
  801. package/bin/skills/ml-paper-writing/templates/aaai2026/aaai2026-unified-supp.tex +0 -144
  802. package/bin/skills/ml-paper-writing/templates/aaai2026/aaai2026-unified-template.tex +0 -952
  803. package/bin/skills/ml-paper-writing/templates/aaai2026/aaai2026.bib +0 -111
  804. package/bin/skills/ml-paper-writing/templates/aaai2026/aaai2026.bst +0 -1493
  805. package/bin/skills/ml-paper-writing/templates/aaai2026/aaai2026.sty +0 -315
  806. package/bin/skills/ml-paper-writing/templates/acl/README.md +0 -50
  807. package/bin/skills/ml-paper-writing/templates/acl/acl.sty +0 -312
  808. package/bin/skills/ml-paper-writing/templates/acl/acl_latex.tex +0 -377
  809. package/bin/skills/ml-paper-writing/templates/acl/acl_lualatex.tex +0 -101
  810. package/bin/skills/ml-paper-writing/templates/acl/acl_natbib.bst +0 -1940
  811. package/bin/skills/ml-paper-writing/templates/acl/anthology.bib.txt +0 -26
  812. package/bin/skills/ml-paper-writing/templates/acl/custom.bib +0 -70
  813. package/bin/skills/ml-paper-writing/templates/acl/formatting.md +0 -326
  814. package/bin/skills/ml-paper-writing/templates/colm2025/README.md +0 -3
  815. package/bin/skills/ml-paper-writing/templates/colm2025/colm2025_conference.bib +0 -11
  816. package/bin/skills/ml-paper-writing/templates/colm2025/colm2025_conference.bst +0 -1440
  817. package/bin/skills/ml-paper-writing/templates/colm2025/colm2025_conference.pdf +0 -0
  818. package/bin/skills/ml-paper-writing/templates/colm2025/colm2025_conference.sty +0 -218
  819. package/bin/skills/ml-paper-writing/templates/colm2025/colm2025_conference.tex +0 -305
  820. package/bin/skills/ml-paper-writing/templates/colm2025/fancyhdr.sty +0 -485
  821. package/bin/skills/ml-paper-writing/templates/colm2025/math_commands.tex +0 -508
  822. package/bin/skills/ml-paper-writing/templates/colm2025/natbib.sty +0 -1246
  823. package/bin/skills/ml-paper-writing/templates/iclr2026/fancyhdr.sty +0 -485
  824. package/bin/skills/ml-paper-writing/templates/iclr2026/iclr2026_conference.bib +0 -24
  825. package/bin/skills/ml-paper-writing/templates/iclr2026/iclr2026_conference.bst +0 -1440
  826. package/bin/skills/ml-paper-writing/templates/iclr2026/iclr2026_conference.pdf +0 -0
  827. package/bin/skills/ml-paper-writing/templates/iclr2026/iclr2026_conference.sty +0 -246
  828. package/bin/skills/ml-paper-writing/templates/iclr2026/iclr2026_conference.tex +0 -414
  829. package/bin/skills/ml-paper-writing/templates/iclr2026/math_commands.tex +0 -508
  830. package/bin/skills/ml-paper-writing/templates/iclr2026/natbib.sty +0 -1246
  831. package/bin/skills/ml-paper-writing/templates/icml2026/algorithm.sty +0 -79
  832. package/bin/skills/ml-paper-writing/templates/icml2026/algorithmic.sty +0 -201
  833. package/bin/skills/ml-paper-writing/templates/icml2026/example_paper.bib +0 -75
  834. package/bin/skills/ml-paper-writing/templates/icml2026/example_paper.pdf +0 -0
  835. package/bin/skills/ml-paper-writing/templates/icml2026/example_paper.tex +0 -662
  836. package/bin/skills/ml-paper-writing/templates/icml2026/fancyhdr.sty +0 -864
  837. package/bin/skills/ml-paper-writing/templates/icml2026/icml2026.bst +0 -1443
  838. package/bin/skills/ml-paper-writing/templates/icml2026/icml2026.sty +0 -767
  839. package/bin/skills/ml-paper-writing/templates/icml2026/icml_numpapers.pdf +0 -0
  840. package/bin/skills/ml-paper-writing/templates/neurips2025/Makefile +0 -36
  841. package/bin/skills/ml-paper-writing/templates/neurips2025/extra_pkgs.tex +0 -53
  842. package/bin/skills/ml-paper-writing/templates/neurips2025/main.tex +0 -38
  843. package/bin/skills/ml-paper-writing/templates/neurips2025/neurips.sty +0 -382
  844. package/bin/skills/mlflow/SKILL.md +0 -704
  845. package/bin/skills/mlflow/references/deployment.md +0 -744
  846. package/bin/skills/mlflow/references/model-registry.md +0 -770
  847. package/bin/skills/mlflow/references/tracking.md +0 -680
  848. package/bin/skills/modal/SKILL.md +0 -418
  849. package/bin/skills/modal/references/advanced-patterns.md +0 -695
  850. package/bin/skills/modal/references/examples-catalog.md +0 -423
  851. package/bin/skills/modal/references/troubleshooting.md +0 -494
  852. package/bin/skills/modal-research-gpu/SKILL.md +0 -238
  853. package/bin/skills/model-economics/SKILL.md +0 -238
  854. package/bin/skills/model-merging/SKILL.md +0 -539
  855. package/bin/skills/model-merging/references/evaluation.md +0 -462
  856. package/bin/skills/model-merging/references/examples.md +0 -428
  857. package/bin/skills/model-merging/references/methods.md +0 -352
  858. package/bin/skills/model-pruning/SKILL.md +0 -495
  859. package/bin/skills/model-pruning/references/wanda.md +0 -347
  860. package/bin/skills/moe-training/SKILL.md +0 -526
  861. package/bin/skills/moe-training/references/architectures.md +0 -432
  862. package/bin/skills/moe-training/references/inference.md +0 -348
  863. package/bin/skills/moe-training/references/training.md +0 -425
  864. package/bin/skills/molfeat/SKILL.md +0 -511
  865. package/bin/skills/molfeat/references/api_reference.md +0 -428
  866. package/bin/skills/molfeat/references/available_featurizers.md +0 -333
  867. package/bin/skills/molfeat/references/examples.md +0 -723
  868. package/bin/skills/nanogpt/SKILL.md +0 -290
  869. package/bin/skills/nanogpt/references/architecture.md +0 -382
  870. package/bin/skills/nanogpt/references/data.md +0 -476
  871. package/bin/skills/nanogpt/references/training.md +0 -564
  872. package/bin/skills/nemo-curator/SKILL.md +0 -383
  873. package/bin/skills/nemo-curator/references/deduplication.md +0 -87
  874. package/bin/skills/nemo-curator/references/filtering.md +0 -102
  875. package/bin/skills/nemo-evaluator/SKILL.md +0 -494
  876. package/bin/skills/nemo-evaluator/references/adapter-system.md +0 -340
  877. package/bin/skills/nemo-evaluator/references/configuration.md +0 -447
  878. package/bin/skills/nemo-evaluator/references/custom-benchmarks.md +0 -315
  879. package/bin/skills/nemo-evaluator/references/execution-backends.md +0 -361
  880. package/bin/skills/nemo-guardrails/SKILL.md +0 -297
  881. package/bin/skills/networkx/SKILL.md +0 -437
  882. package/bin/skills/networkx/references/algorithms.md +0 -383
  883. package/bin/skills/networkx/references/generators.md +0 -378
  884. package/bin/skills/networkx/references/graph-basics.md +0 -283
  885. package/bin/skills/networkx/references/io.md +0 -441
  886. package/bin/skills/networkx/references/visualization.md +0 -529
  887. package/bin/skills/neurokit2/SKILL.md +0 -356
  888. package/bin/skills/neurokit2/references/bio_module.md +0 -417
  889. package/bin/skills/neurokit2/references/complexity.md +0 -715
  890. package/bin/skills/neurokit2/references/ecg_cardiac.md +0 -355
  891. package/bin/skills/neurokit2/references/eda.md +0 -497
  892. package/bin/skills/neurokit2/references/eeg.md +0 -506
  893. package/bin/skills/neurokit2/references/emg.md +0 -408
  894. package/bin/skills/neurokit2/references/eog.md +0 -407
  895. package/bin/skills/neurokit2/references/epochs_events.md +0 -471
  896. package/bin/skills/neurokit2/references/hrv.md +0 -480
  897. package/bin/skills/neurokit2/references/ppg.md +0 -413
  898. package/bin/skills/neurokit2/references/rsp.md +0 -510
  899. package/bin/skills/neurokit2/references/signal_processing.md +0 -648
  900. package/bin/skills/neuropixels-analysis/SKILL.md +0 -350
  901. package/bin/skills/neuropixels-analysis/assets/analysis_template.py +0 -271
  902. package/bin/skills/neuropixels-analysis/references/AI_CURATION.md +0 -345
  903. package/bin/skills/neuropixels-analysis/references/ANALYSIS.md +0 -392
  904. package/bin/skills/neuropixels-analysis/references/AUTOMATED_CURATION.md +0 -358
  905. package/bin/skills/neuropixels-analysis/references/MOTION_CORRECTION.md +0 -323
  906. package/bin/skills/neuropixels-analysis/references/PREPROCESSING.md +0 -273
  907. package/bin/skills/neuropixels-analysis/references/QUALITY_METRICS.md +0 -359
  908. package/bin/skills/neuropixels-analysis/references/SPIKE_SORTING.md +0 -339
  909. package/bin/skills/neuropixels-analysis/references/api_reference.md +0 -415
  910. package/bin/skills/neuropixels-analysis/references/plotting_guide.md +0 -454
  911. package/bin/skills/neuropixels-analysis/references/standard_workflow.md +0 -385
  912. package/bin/skills/neuropixels-analysis/scripts/compute_metrics.py +0 -178
  913. package/bin/skills/neuropixels-analysis/scripts/explore_recording.py +0 -168
  914. package/bin/skills/neuropixels-analysis/scripts/export_to_phy.py +0 -79
  915. package/bin/skills/neuropixels-analysis/scripts/neuropixels_pipeline.py +0 -432
  916. package/bin/skills/neuropixels-analysis/scripts/preprocess_recording.py +0 -122
  917. package/bin/skills/neuropixels-analysis/scripts/run_sorting.py +0 -98
  918. package/bin/skills/nnsight/SKILL.md +0 -436
  919. package/bin/skills/nnsight/references/README.md +0 -78
  920. package/bin/skills/nnsight/references/api.md +0 -344
  921. package/bin/skills/nnsight/references/tutorials.md +0 -300
  922. package/bin/skills/offer-k-dense-web/SKILL.md +0 -21
  923. package/bin/skills/omero-integration/SKILL.md +0 -251
  924. package/bin/skills/omero-integration/references/advanced.md +0 -631
  925. package/bin/skills/omero-integration/references/connection.md +0 -369
  926. package/bin/skills/omero-integration/references/data_access.md +0 -544
  927. package/bin/skills/omero-integration/references/image_processing.md +0 -665
  928. package/bin/skills/omero-integration/references/metadata.md +0 -688
  929. package/bin/skills/omero-integration/references/rois.md +0 -648
  930. package/bin/skills/omero-integration/references/scripts.md +0 -637
  931. package/bin/skills/omero-integration/references/tables.md +0 -532
  932. package/bin/skills/openalex-database/SKILL.md +0 -494
  933. package/bin/skills/openalex-database/references/api_guide.md +0 -371
  934. package/bin/skills/openalex-database/references/common_queries.md +0 -381
  935. package/bin/skills/openalex-database/scripts/openalex_client.py +0 -337
  936. package/bin/skills/openalex-database/scripts/query_helpers.py +0 -306
  937. package/bin/skills/openrlhf/SKILL.md +0 -249
  938. package/bin/skills/openrlhf/references/algorithm-comparison.md +0 -404
  939. package/bin/skills/openrlhf/references/custom-rewards.md +0 -530
  940. package/bin/skills/openrlhf/references/hybrid-engine.md +0 -287
  941. package/bin/skills/openrlhf/references/multi-node-training.md +0 -454
  942. package/bin/skills/opentargets-database/SKILL.md +0 -373
  943. package/bin/skills/opentargets-database/references/api_reference.md +0 -249
  944. package/bin/skills/opentargets-database/references/evidence_types.md +0 -306
  945. package/bin/skills/opentargets-database/references/target_annotations.md +0 -401
  946. package/bin/skills/opentargets-database/scripts/query_opentargets.py +0 -403
  947. package/bin/skills/opentrons-integration/SKILL.md +0 -573
  948. package/bin/skills/opentrons-integration/references/api_reference.md +0 -366
  949. package/bin/skills/opentrons-integration/scripts/basic_protocol_template.py +0 -67
  950. package/bin/skills/opentrons-integration/scripts/pcr_setup_template.py +0 -154
  951. package/bin/skills/opentrons-integration/scripts/serial_dilution_template.py +0 -96
  952. package/bin/skills/outlines/SKILL.md +0 -652
  953. package/bin/skills/outlines/references/backends.md +0 -615
  954. package/bin/skills/outlines/references/examples.md +0 -773
  955. package/bin/skills/outlines/references/json_generation.md +0 -652
  956. package/bin/skills/paper-2-web/SKILL.md +0 -491
  957. package/bin/skills/paper-2-web/references/installation.md +0 -141
  958. package/bin/skills/paper-2-web/references/paper2poster.md +0 -346
  959. package/bin/skills/paper-2-web/references/paper2video.md +0 -305
  960. package/bin/skills/paper-2-web/references/paper2web.md +0 -187
  961. package/bin/skills/paper-2-web/references/usage_examples.md +0 -436
  962. package/bin/skills/pathml/SKILL.md +0 -166
  963. package/bin/skills/pathml/references/data_management.md +0 -742
  964. package/bin/skills/pathml/references/graphs.md +0 -653
  965. package/bin/skills/pathml/references/image_loading.md +0 -448
  966. package/bin/skills/pathml/references/machine_learning.md +0 -725
  967. package/bin/skills/pathml/references/multiparametric.md +0 -686
  968. package/bin/skills/pathml/references/preprocessing.md +0 -722
  969. package/bin/skills/pdb-database/SKILL.md +0 -309
  970. package/bin/skills/pdb-database/references/api_reference.md +0 -617
  971. package/bin/skills/peer-review/SKILL.md +0 -702
  972. package/bin/skills/peer-review/references/calibration_guidelines.md +0 -196
  973. package/bin/skills/peer-review/references/common_issues.md +0 -552
  974. package/bin/skills/peer-review/references/paper_mechanics.md +0 -269
  975. package/bin/skills/peer-review/references/reporting_standards.md +0 -290
  976. package/bin/skills/peer-review/references/scoring_rubric.md +0 -239
  977. package/bin/skills/peft/SKILL.md +0 -431
  978. package/bin/skills/peft/references/advanced-usage.md +0 -514
  979. package/bin/skills/peft/references/troubleshooting.md +0 -480
  980. package/bin/skills/pennylane/SKILL.md +0 -226
  981. package/bin/skills/pennylane/references/advanced_features.md +0 -667
  982. package/bin/skills/pennylane/references/devices_backends.md +0 -596
  983. package/bin/skills/pennylane/references/getting_started.md +0 -227
  984. package/bin/skills/pennylane/references/optimization.md +0 -671
  985. package/bin/skills/pennylane/references/quantum_chemistry.md +0 -567
  986. package/bin/skills/pennylane/references/quantum_circuits.md +0 -437
  987. package/bin/skills/pennylane/references/quantum_ml.md +0 -571
  988. package/bin/skills/perplexity-search/SKILL.md +0 -448
  989. package/bin/skills/perplexity-search/assets/.env.example +0 -16
  990. package/bin/skills/perplexity-search/references/model_comparison.md +0 -386
  991. package/bin/skills/perplexity-search/references/openrouter_setup.md +0 -454
  992. package/bin/skills/perplexity-search/references/search_strategies.md +0 -258
  993. package/bin/skills/perplexity-search/scripts/perplexity_search.py +0 -277
  994. package/bin/skills/perplexity-search/scripts/setup_env.py +0 -171
  995. package/bin/skills/phoenix/SKILL.md +0 -475
  996. package/bin/skills/phoenix/references/advanced-usage.md +0 -619
  997. package/bin/skills/phoenix/references/troubleshooting.md +0 -538
  998. package/bin/skills/pinecone/SKILL.md +0 -358
  999. package/bin/skills/pinecone/references/deployment.md +0 -181
  1000. package/bin/skills/plotly/SKILL.md +0 -267
  1001. package/bin/skills/plotly/references/chart-types.md +0 -488
  1002. package/bin/skills/plotly/references/export-interactivity.md +0 -453
  1003. package/bin/skills/plotly/references/graph-objects.md +0 -302
  1004. package/bin/skills/plotly/references/layouts-styling.md +0 -457
  1005. package/bin/skills/plotly/references/plotly-express.md +0 -213
  1006. package/bin/skills/polars/SKILL.md +0 -387
  1007. package/bin/skills/polars/references/best_practices.md +0 -649
  1008. package/bin/skills/polars/references/core_concepts.md +0 -378
  1009. package/bin/skills/polars/references/io_guide.md +0 -557
  1010. package/bin/skills/polars/references/operations.md +0 -602
  1011. package/bin/skills/polars/references/pandas_migration.md +0 -417
  1012. package/bin/skills/polars/references/transformations.md +0 -549
  1013. package/bin/skills/pptx-posters/SKILL.md +0 -410
  1014. package/bin/skills/pptx-posters/assets/poster_html_template.html +0 -257
  1015. package/bin/skills/pptx-posters/assets/poster_quality_checklist.md +0 -358
  1016. package/bin/skills/pptx-posters/references/poster_content_guide.md +0 -748
  1017. package/bin/skills/pptx-posters/references/poster_design_principles.md +0 -806
  1018. package/bin/skills/pptx-posters/references/poster_layout_design.md +0 -900
  1019. package/bin/skills/prime-intellect-lab/README.md +0 -69
  1020. package/bin/skills/prime-intellect-lab/SKILL.md +0 -598
  1021. package/bin/skills/prime-intellect-lab/templates/basic_rl_training.toml +0 -82
  1022. package/bin/skills/protocolsio-integration/SKILL.md +0 -421
  1023. package/bin/skills/protocolsio-integration/references/additional_features.md +0 -387
  1024. package/bin/skills/protocolsio-integration/references/authentication.md +0 -100
  1025. package/bin/skills/protocolsio-integration/references/discussions.md +0 -225
  1026. package/bin/skills/protocolsio-integration/references/file_manager.md +0 -412
  1027. package/bin/skills/protocolsio-integration/references/protocols_api.md +0 -294
  1028. package/bin/skills/protocolsio-integration/references/workspaces.md +0 -293
  1029. package/bin/skills/pubchem-database/SKILL.md +0 -574
  1030. package/bin/skills/pubchem-database/references/api_reference.md +0 -440
  1031. package/bin/skills/pubchem-database/scripts/bioactivity_query.py +0 -367
  1032. package/bin/skills/pubchem-database/scripts/compound_search.py +0 -297
  1033. package/bin/skills/pubmed-database/SKILL.md +0 -460
  1034. package/bin/skills/pubmed-database/references/api_reference.md +0 -298
  1035. package/bin/skills/pubmed-database/references/common_queries.md +0 -453
  1036. package/bin/skills/pubmed-database/references/search_syntax.md +0 -436
  1037. package/bin/skills/pufferlib/SKILL.md +0 -436
  1038. package/bin/skills/pufferlib/references/environments.md +0 -508
  1039. package/bin/skills/pufferlib/references/integration.md +0 -621
  1040. package/bin/skills/pufferlib/references/policies.md +0 -653
  1041. package/bin/skills/pufferlib/references/training.md +0 -360
  1042. package/bin/skills/pufferlib/references/vectorization.md +0 -557
  1043. package/bin/skills/pufferlib/scripts/env_template.py +0 -340
  1044. package/bin/skills/pufferlib/scripts/train_template.py +0 -239
  1045. package/bin/skills/pydeseq2/SKILL.md +0 -559
  1046. package/bin/skills/pydeseq2/references/api_reference.md +0 -228
  1047. package/bin/skills/pydeseq2/references/workflow_guide.md +0 -582
  1048. package/bin/skills/pydeseq2/scripts/run_deseq2_analysis.py +0 -353
  1049. package/bin/skills/pydicom/SKILL.md +0 -434
  1050. package/bin/skills/pydicom/references/common_tags.md +0 -228
  1051. package/bin/skills/pydicom/references/transfer_syntaxes.md +0 -352
  1052. package/bin/skills/pydicom/scripts/anonymize_dicom.py +0 -137
  1053. package/bin/skills/pydicom/scripts/dicom_to_image.py +0 -172
  1054. package/bin/skills/pydicom/scripts/extract_metadata.py +0 -173
  1055. package/bin/skills/pyhealth/SKILL.md +0 -491
  1056. package/bin/skills/pyhealth/references/datasets.md +0 -178
  1057. package/bin/skills/pyhealth/references/medical_coding.md +0 -284
  1058. package/bin/skills/pyhealth/references/models.md +0 -594
  1059. package/bin/skills/pyhealth/references/preprocessing.md +0 -638
  1060. package/bin/skills/pyhealth/references/tasks.md +0 -379
  1061. package/bin/skills/pyhealth/references/training_evaluation.md +0 -648
  1062. package/bin/skills/pylabrobot/SKILL.md +0 -185
  1063. package/bin/skills/pylabrobot/references/analytical-equipment.md +0 -464
  1064. package/bin/skills/pylabrobot/references/hardware-backends.md +0 -480
  1065. package/bin/skills/pylabrobot/references/liquid-handling.md +0 -403
  1066. package/bin/skills/pylabrobot/references/material-handling.md +0 -620
  1067. package/bin/skills/pylabrobot/references/resources.md +0 -489
  1068. package/bin/skills/pylabrobot/references/visualization.md +0 -532
  1069. package/bin/skills/pymatgen/SKILL.md +0 -691
  1070. package/bin/skills/pymatgen/references/analysis_modules.md +0 -530
  1071. package/bin/skills/pymatgen/references/core_classes.md +0 -318
  1072. package/bin/skills/pymatgen/references/io_formats.md +0 -469
  1073. package/bin/skills/pymatgen/references/materials_project_api.md +0 -517
  1074. package/bin/skills/pymatgen/references/transformations_workflows.md +0 -591
  1075. package/bin/skills/pymatgen/scripts/phase_diagram_generator.py +0 -233
  1076. package/bin/skills/pymatgen/scripts/structure_analyzer.py +0 -266
  1077. package/bin/skills/pymatgen/scripts/structure_converter.py +0 -169
  1078. package/bin/skills/pymc/SKILL.md +0 -572
  1079. package/bin/skills/pymc/assets/hierarchical_model_template.py +0 -333
  1080. package/bin/skills/pymc/assets/linear_regression_template.py +0 -241
  1081. package/bin/skills/pymc/references/distributions.md +0 -320
  1082. package/bin/skills/pymc/references/sampling_inference.md +0 -424
  1083. package/bin/skills/pymc/references/workflows.md +0 -526
  1084. package/bin/skills/pymc/scripts/model_comparison.py +0 -387
  1085. package/bin/skills/pymc/scripts/model_diagnostics.py +0 -350
  1086. package/bin/skills/pymoo/SKILL.md +0 -571
  1087. package/bin/skills/pymoo/references/algorithms.md +0 -180
  1088. package/bin/skills/pymoo/references/constraints_mcdm.md +0 -417
  1089. package/bin/skills/pymoo/references/operators.md +0 -345
  1090. package/bin/skills/pymoo/references/problems.md +0 -265
  1091. package/bin/skills/pymoo/references/visualization.md +0 -353
  1092. package/bin/skills/pymoo/scripts/custom_problem_example.py +0 -181
  1093. package/bin/skills/pymoo/scripts/decision_making_example.py +0 -161
  1094. package/bin/skills/pymoo/scripts/many_objective_example.py +0 -72
  1095. package/bin/skills/pymoo/scripts/multi_objective_example.py +0 -63
  1096. package/bin/skills/pymoo/scripts/single_objective_example.py +0 -59
  1097. package/bin/skills/pyopenms/SKILL.md +0 -217
  1098. package/bin/skills/pyopenms/references/data_structures.md +0 -497
  1099. package/bin/skills/pyopenms/references/feature_detection.md +0 -410
  1100. package/bin/skills/pyopenms/references/file_io.md +0 -349
  1101. package/bin/skills/pyopenms/references/identification.md +0 -422
  1102. package/bin/skills/pyopenms/references/metabolomics.md +0 -482
  1103. package/bin/skills/pyopenms/references/signal_processing.md +0 -433
  1104. package/bin/skills/pysam/SKILL.md +0 -265
  1105. package/bin/skills/pysam/references/alignment_files.md +0 -280
  1106. package/bin/skills/pysam/references/common_workflows.md +0 -520
  1107. package/bin/skills/pysam/references/sequence_files.md +0 -407
  1108. package/bin/skills/pysam/references/variant_files.md +0 -365
  1109. package/bin/skills/pytdc/SKILL.md +0 -460
  1110. package/bin/skills/pytdc/references/datasets.md +0 -246
  1111. package/bin/skills/pytdc/references/oracles.md +0 -400
  1112. package/bin/skills/pytdc/references/utilities.md +0 -684
  1113. package/bin/skills/pytdc/scripts/benchmark_evaluation.py +0 -327
  1114. package/bin/skills/pytdc/scripts/load_and_split_data.py +0 -214
  1115. package/bin/skills/pytdc/scripts/molecular_generation.py +0 -404
  1116. package/bin/skills/pytorch-fsdp/SKILL.md +0 -126
  1117. package/bin/skills/pytorch-fsdp/references/index.md +0 -7
  1118. package/bin/skills/pytorch-fsdp/references/other.md +0 -4249
  1119. package/bin/skills/pytorch-lightning/SKILL.md +0 -346
  1120. package/bin/skills/pytorch-lightning/references/callbacks.md +0 -436
  1121. package/bin/skills/pytorch-lightning/references/distributed.md +0 -490
  1122. package/bin/skills/pytorch-lightning/references/hyperparameter-tuning.md +0 -556
  1123. package/bin/skills/pyvene/SKILL.md +0 -473
  1124. package/bin/skills/pyvene/references/README.md +0 -73
  1125. package/bin/skills/pyvene/references/api.md +0 -383
  1126. package/bin/skills/pyvene/references/tutorials.md +0 -376
  1127. package/bin/skills/qdrant/SKILL.md +0 -493
  1128. package/bin/skills/qdrant/references/advanced-usage.md +0 -648
  1129. package/bin/skills/qdrant/references/troubleshooting.md +0 -631
  1130. package/bin/skills/qiskit/SKILL.md +0 -275
  1131. package/bin/skills/qiskit/references/algorithms.md +0 -607
  1132. package/bin/skills/qiskit/references/backends.md +0 -433
  1133. package/bin/skills/qiskit/references/circuits.md +0 -197
  1134. package/bin/skills/qiskit/references/patterns.md +0 -533
  1135. package/bin/skills/qiskit/references/primitives.md +0 -277
  1136. package/bin/skills/qiskit/references/setup.md +0 -99
  1137. package/bin/skills/qiskit/references/transpilation.md +0 -286
  1138. package/bin/skills/qiskit/references/visualization.md +0 -415
  1139. package/bin/skills/qutip/SKILL.md +0 -318
  1140. package/bin/skills/qutip/references/advanced.md +0 -555
  1141. package/bin/skills/qutip/references/analysis.md +0 -523
  1142. package/bin/skills/qutip/references/core_concepts.md +0 -293
  1143. package/bin/skills/qutip/references/time_evolution.md +0 -348
  1144. package/bin/skills/qutip/references/visualization.md +0 -431
  1145. package/bin/skills/ray-data/SKILL.md +0 -326
  1146. package/bin/skills/ray-data/references/integration.md +0 -82
  1147. package/bin/skills/ray-data/references/transformations.md +0 -83
  1148. package/bin/skills/ray-train/SKILL.md +0 -406
  1149. package/bin/skills/ray-train/references/multi-node.md +0 -628
  1150. package/bin/skills/rdkit/SKILL.md +0 -780
  1151. package/bin/skills/rdkit/references/api_reference.md +0 -432
  1152. package/bin/skills/rdkit/references/descriptors_reference.md +0 -595
  1153. package/bin/skills/rdkit/references/smarts_patterns.md +0 -668
  1154. package/bin/skills/rdkit/scripts/molecular_properties.py +0 -243
  1155. package/bin/skills/rdkit/scripts/similarity_search.py +0 -297
  1156. package/bin/skills/rdkit/scripts/substructure_filter.py +0 -386
  1157. package/bin/skills/reactome-database/SKILL.md +0 -278
  1158. package/bin/skills/reactome-database/references/api_reference.md +0 -465
  1159. package/bin/skills/reactome-database/scripts/reactome_query.py +0 -286
  1160. package/bin/skills/research-grants/README.md +0 -285
  1161. package/bin/skills/research-grants/SKILL.md +0 -938
  1162. package/bin/skills/research-grants/assets/budget_justification_template.md +0 -453
  1163. package/bin/skills/research-grants/assets/nih_specific_aims_template.md +0 -166
  1164. package/bin/skills/research-grants/assets/nsf_project_summary_template.md +0 -92
  1165. package/bin/skills/research-grants/references/broader_impacts.md +0 -392
  1166. package/bin/skills/research-grants/references/darpa_guidelines.md +0 -636
  1167. package/bin/skills/research-grants/references/doe_guidelines.md +0 -586
  1168. package/bin/skills/research-grants/references/nih_guidelines.md +0 -851
  1169. package/bin/skills/research-grants/references/nsf_guidelines.md +0 -570
  1170. package/bin/skills/research-grants/references/specific_aims_guide.md +0 -458
  1171. package/bin/skills/research-lookup/README.md +0 -156
  1172. package/bin/skills/research-lookup/SKILL.md +0 -606
  1173. package/bin/skills/research-lookup/examples.py +0 -174
  1174. package/bin/skills/research-lookup/lookup.py +0 -187
  1175. package/bin/skills/research-lookup/research_lookup.py +0 -483
  1176. package/bin/skills/research-lookup/scripts/research_lookup.py +0 -483
  1177. package/bin/skills/rowan/SKILL.md +0 -427
  1178. package/bin/skills/rowan/references/api_reference.md +0 -413
  1179. package/bin/skills/rowan/references/molecule_handling.md +0 -429
  1180. package/bin/skills/rowan/references/proteins_and_organization.md +0 -499
  1181. package/bin/skills/rowan/references/rdkit_native.md +0 -438
  1182. package/bin/skills/rowan/references/results_interpretation.md +0 -481
  1183. package/bin/skills/rowan/references/workflow_types.md +0 -591
  1184. package/bin/skills/rwkv/SKILL.md +0 -260
  1185. package/bin/skills/rwkv/references/architecture-details.md +0 -344
  1186. package/bin/skills/rwkv/references/rwkv7.md +0 -386
  1187. package/bin/skills/rwkv/references/state-management.md +0 -369
  1188. package/bin/skills/saelens/SKILL.md +0 -386
  1189. package/bin/skills/saelens/references/README.md +0 -70
  1190. package/bin/skills/saelens/references/api.md +0 -333
  1191. package/bin/skills/saelens/references/tutorials.md +0 -318
  1192. package/bin/skills/scanpy/SKILL.md +0 -386
  1193. package/bin/skills/scanpy/assets/analysis_template.py +0 -295
  1194. package/bin/skills/scanpy/references/api_reference.md +0 -251
  1195. package/bin/skills/scanpy/references/plotting_guide.md +0 -352
  1196. package/bin/skills/scanpy/references/standard_workflow.md +0 -206
  1197. package/bin/skills/scanpy/scripts/qc_analysis.py +0 -200
  1198. package/bin/skills/scholar-evaluation/SKILL.md +0 -289
  1199. package/bin/skills/scholar-evaluation/references/evaluation_framework.md +0 -663
  1200. package/bin/skills/scholar-evaluation/scripts/calculate_scores.py +0 -366
  1201. package/bin/skills/scientific-brainstorming/SKILL.md +0 -191
  1202. package/bin/skills/scientific-brainstorming/references/brainstorming_methods.md +0 -326
  1203. package/bin/skills/scientific-critical-thinking/SKILL.md +0 -566
  1204. package/bin/skills/scientific-critical-thinking/references/common_biases.md +0 -364
  1205. package/bin/skills/scientific-critical-thinking/references/evidence_hierarchy.md +0 -484
  1206. package/bin/skills/scientific-critical-thinking/references/experimental_design.md +0 -496
  1207. package/bin/skills/scientific-critical-thinking/references/logical_fallacies.md +0 -478
  1208. package/bin/skills/scientific-critical-thinking/references/scientific_method.md +0 -169
  1209. package/bin/skills/scientific-critical-thinking/references/statistical_pitfalls.md +0 -506
  1210. package/bin/skills/scientific-schematics/QUICK_REFERENCE.md +0 -207
  1211. package/bin/skills/scientific-schematics/README.md +0 -327
  1212. package/bin/skills/scientific-schematics/SKILL.md +0 -615
  1213. package/bin/skills/scientific-schematics/example_usage.sh +0 -89
  1214. package/bin/skills/scientific-schematics/references/best_practices.md +0 -559
  1215. package/bin/skills/scientific-schematics/scripts/generate_schematic.py +0 -135
  1216. package/bin/skills/scientific-schematics/scripts/generate_schematic_ai.py +0 -837
  1217. package/bin/skills/scientific-schematics/test_ai_generation.py +0 -243
  1218. package/bin/skills/scientific-slides/SKILL.md +0 -942
  1219. package/bin/skills/scientific-slides/assets/timing_guidelines.md +0 -597
  1220. package/bin/skills/scientific-slides/references/data_visualization_slides.md +0 -708
  1221. package/bin/skills/scientific-slides/references/presentation_structure.md +0 -642
  1222. package/bin/skills/scientific-slides/references/slide_design_principles.md +0 -849
  1223. package/bin/skills/scientific-slides/references/talk_types_guide.md +0 -687
  1224. package/bin/skills/scientific-slides/references/visual_review_workflow.md +0 -775
  1225. package/bin/skills/scientific-slides/scripts/generate_slide_image.py +0 -143
  1226. package/bin/skills/scientific-slides/scripts/generate_slide_image_ai.py +0 -748
  1227. package/bin/skills/scientific-slides/scripts/pdf_to_images.py +0 -201
  1228. package/bin/skills/scientific-slides/scripts/slides_to_pdf.py +0 -220
  1229. package/bin/skills/scientific-slides/scripts/validate_presentation.py +0 -367
  1230. package/bin/skills/scientific-visualization/SKILL.md +0 -779
  1231. package/bin/skills/scientific-visualization/assets/color_palettes.py +0 -197
  1232. package/bin/skills/scientific-visualization/assets/nature.mplstyle +0 -63
  1233. package/bin/skills/scientific-visualization/assets/presentation.mplstyle +0 -61
  1234. package/bin/skills/scientific-visualization/assets/publication.mplstyle +0 -68
  1235. package/bin/skills/scientific-visualization/references/color_palettes.md +0 -348
  1236. package/bin/skills/scientific-visualization/references/journal_requirements.md +0 -320
  1237. package/bin/skills/scientific-visualization/references/matplotlib_examples.md +0 -620
  1238. package/bin/skills/scientific-visualization/references/publication_guidelines.md +0 -205
  1239. package/bin/skills/scientific-visualization/scripts/figure_export.py +0 -343
  1240. package/bin/skills/scientific-visualization/scripts/style_presets.py +0 -416
  1241. package/bin/skills/scientific-writing/SKILL.md +0 -714
  1242. package/bin/skills/scientific-writing/assets/REPORT_FORMATTING_GUIDE.md +0 -574
  1243. package/bin/skills/scientific-writing/assets/scientific_report.sty +0 -606
  1244. package/bin/skills/scientific-writing/assets/scientific_report_template.tex +0 -449
  1245. package/bin/skills/scientific-writing/references/citation_styles.md +0 -720
  1246. package/bin/skills/scientific-writing/references/figures_tables.md +0 -806
  1247. package/bin/skills/scientific-writing/references/imrad_structure.md +0 -686
  1248. package/bin/skills/scientific-writing/references/professional_report_formatting.md +0 -664
  1249. package/bin/skills/scientific-writing/references/reporting_guidelines.md +0 -748
  1250. package/bin/skills/scientific-writing/references/writing_principles.md +0 -824
  1251. package/bin/skills/scikit-bio/SKILL.md +0 -437
  1252. package/bin/skills/scikit-bio/references/api_reference.md +0 -749
  1253. package/bin/skills/scikit-learn/SKILL.md +0 -521
  1254. package/bin/skills/scikit-learn/references/model_evaluation.md +0 -592
  1255. package/bin/skills/scikit-learn/references/pipelines_and_composition.md +0 -612
  1256. package/bin/skills/scikit-learn/references/preprocessing.md +0 -606
  1257. package/bin/skills/scikit-learn/references/quick_reference.md +0 -433
  1258. package/bin/skills/scikit-learn/references/supervised_learning.md +0 -378
  1259. package/bin/skills/scikit-learn/references/unsupervised_learning.md +0 -505
  1260. package/bin/skills/scikit-learn/scripts/classification_pipeline.py +0 -257
  1261. package/bin/skills/scikit-learn/scripts/clustering_analysis.py +0 -386
  1262. package/bin/skills/scikit-survival/SKILL.md +0 -399
  1263. package/bin/skills/scikit-survival/references/competing-risks.md +0 -397
  1264. package/bin/skills/scikit-survival/references/cox-models.md +0 -182
  1265. package/bin/skills/scikit-survival/references/data-handling.md +0 -494
  1266. package/bin/skills/scikit-survival/references/ensemble-models.md +0 -327
  1267. package/bin/skills/scikit-survival/references/evaluation-metrics.md +0 -378
  1268. package/bin/skills/scikit-survival/references/svm-models.md +0 -411
  1269. package/bin/skills/scvi-tools/SKILL.md +0 -190
  1270. package/bin/skills/scvi-tools/references/differential-expression.md +0 -581
  1271. package/bin/skills/scvi-tools/references/models-atac-seq.md +0 -321
  1272. package/bin/skills/scvi-tools/references/models-multimodal.md +0 -367
  1273. package/bin/skills/scvi-tools/references/models-scrna-seq.md +0 -330
  1274. package/bin/skills/scvi-tools/references/models-spatial.md +0 -438
  1275. package/bin/skills/scvi-tools/references/models-specialized.md +0 -408
  1276. package/bin/skills/scvi-tools/references/theoretical-foundations.md +0 -438
  1277. package/bin/skills/scvi-tools/references/workflows.md +0 -546
  1278. package/bin/skills/seaborn/SKILL.md +0 -673
  1279. package/bin/skills/seaborn/references/examples.md +0 -822
  1280. package/bin/skills/seaborn/references/function_reference.md +0 -770
  1281. package/bin/skills/seaborn/references/objects_interface.md +0 -964
  1282. package/bin/skills/segment-anything/SKILL.md +0 -500
  1283. package/bin/skills/segment-anything/references/advanced-usage.md +0 -589
  1284. package/bin/skills/segment-anything/references/troubleshooting.md +0 -484
  1285. package/bin/skills/sentence-transformers/SKILL.md +0 -255
  1286. package/bin/skills/sentence-transformers/references/models.md +0 -123
  1287. package/bin/skills/sentencepiece/SKILL.md +0 -235
  1288. package/bin/skills/sentencepiece/references/algorithms.md +0 -200
  1289. package/bin/skills/sentencepiece/references/training.md +0 -304
  1290. package/bin/skills/sglang/SKILL.md +0 -442
  1291. package/bin/skills/sglang/references/deployment.md +0 -490
  1292. package/bin/skills/sglang/references/radix-attention.md +0 -413
  1293. package/bin/skills/sglang/references/structured-generation.md +0 -541
  1294. package/bin/skills/shap/SKILL.md +0 -566
  1295. package/bin/skills/shap/references/explainers.md +0 -339
  1296. package/bin/skills/shap/references/plots.md +0 -507
  1297. package/bin/skills/shap/references/theory.md +0 -449
  1298. package/bin/skills/shap/references/workflows.md +0 -605
  1299. package/bin/skills/simpo/SKILL.md +0 -219
  1300. package/bin/skills/simpo/references/datasets.md +0 -478
  1301. package/bin/skills/simpo/references/hyperparameters.md +0 -452
  1302. package/bin/skills/simpo/references/loss-functions.md +0 -350
  1303. package/bin/skills/simpy/SKILL.md +0 -429
  1304. package/bin/skills/simpy/references/events.md +0 -374
  1305. package/bin/skills/simpy/references/monitoring.md +0 -475
  1306. package/bin/skills/simpy/references/process-interaction.md +0 -424
  1307. package/bin/skills/simpy/references/real-time.md +0 -395
  1308. package/bin/skills/simpy/references/resources.md +0 -275
  1309. package/bin/skills/simpy/scripts/basic_simulation_template.py +0 -193
  1310. package/bin/skills/simpy/scripts/resource_monitor.py +0 -345
  1311. package/bin/skills/skypilot/SKILL.md +0 -509
  1312. package/bin/skills/skypilot/references/advanced-usage.md +0 -491
  1313. package/bin/skills/skypilot/references/troubleshooting.md +0 -570
  1314. package/bin/skills/slime/SKILL.md +0 -464
  1315. package/bin/skills/slime/references/api-reference.md +0 -392
  1316. package/bin/skills/slime/references/troubleshooting.md +0 -386
  1317. package/bin/skills/speculative-decoding/SKILL.md +0 -467
  1318. package/bin/skills/speculative-decoding/references/lookahead.md +0 -309
  1319. package/bin/skills/speculative-decoding/references/medusa.md +0 -350
  1320. package/bin/skills/stable-baselines3/SKILL.md +0 -299
  1321. package/bin/skills/stable-baselines3/references/algorithms.md +0 -333
  1322. package/bin/skills/stable-baselines3/references/callbacks.md +0 -556
  1323. package/bin/skills/stable-baselines3/references/custom_environments.md +0 -526
  1324. package/bin/skills/stable-baselines3/references/vectorized_envs.md +0 -568
  1325. package/bin/skills/stable-baselines3/scripts/custom_env_template.py +0 -314
  1326. package/bin/skills/stable-baselines3/scripts/evaluate_agent.py +0 -245
  1327. package/bin/skills/stable-baselines3/scripts/train_rl_agent.py +0 -165
  1328. package/bin/skills/stable-diffusion/SKILL.md +0 -519
  1329. package/bin/skills/stable-diffusion/references/advanced-usage.md +0 -716
  1330. package/bin/skills/stable-diffusion/references/troubleshooting.md +0 -555
  1331. package/bin/skills/statistical-analysis/SKILL.md +0 -632
  1332. package/bin/skills/statistical-analysis/references/assumptions_and_diagnostics.md +0 -369
  1333. package/bin/skills/statistical-analysis/references/bayesian_statistics.md +0 -661
  1334. package/bin/skills/statistical-analysis/references/effect_sizes_and_power.md +0 -581
  1335. package/bin/skills/statistical-analysis/references/reporting_standards.md +0 -469
  1336. package/bin/skills/statistical-analysis/references/test_selection_guide.md +0 -129
  1337. package/bin/skills/statistical-analysis/scripts/assumption_checks.py +0 -539
  1338. package/bin/skills/statsmodels/SKILL.md +0 -614
  1339. package/bin/skills/statsmodels/references/discrete_choice.md +0 -669
  1340. package/bin/skills/statsmodels/references/glm.md +0 -619
  1341. package/bin/skills/statsmodels/references/linear_models.md +0 -447
  1342. package/bin/skills/statsmodels/references/stats_diagnostics.md +0 -859
  1343. package/bin/skills/statsmodels/references/time_series.md +0 -716
  1344. package/bin/skills/string-database/SKILL.md +0 -534
  1345. package/bin/skills/string-database/references/string_reference.md +0 -455
  1346. package/bin/skills/string-database/scripts/string_api.py +0 -369
  1347. package/bin/skills/sympy/SKILL.md +0 -500
  1348. package/bin/skills/sympy/references/advanced-topics.md +0 -635
  1349. package/bin/skills/sympy/references/code-generation-printing.md +0 -599
  1350. package/bin/skills/sympy/references/core-capabilities.md +0 -348
  1351. package/bin/skills/sympy/references/matrices-linear-algebra.md +0 -526
  1352. package/bin/skills/sympy/references/physics-mechanics.md +0 -592
  1353. package/bin/skills/tensorboard/SKILL.md +0 -629
  1354. package/bin/skills/tensorboard/references/integrations.md +0 -638
  1355. package/bin/skills/tensorboard/references/profiling.md +0 -545
  1356. package/bin/skills/tensorboard/references/visualization.md +0 -620
  1357. package/bin/skills/tensorpool/SKILL.md +0 -519
  1358. package/bin/skills/tensorrt-llm/SKILL.md +0 -187
  1359. package/bin/skills/tensorrt-llm/references/multi-gpu.md +0 -298
  1360. package/bin/skills/tensorrt-llm/references/optimization.md +0 -242
  1361. package/bin/skills/tensorrt-llm/references/serving.md +0 -470
  1362. package/bin/skills/tinker/SKILL.md +0 -466
  1363. package/bin/skills/tinker/references/api-reference.md +0 -168
  1364. package/bin/skills/tinker/references/dpo-and-preference.md +0 -174
  1365. package/bin/skills/tinker/references/evaluations.md +0 -183
  1366. package/bin/skills/tinker/references/getting-started.md +0 -157
  1367. package/bin/skills/tinker/references/loss-functions.md +0 -163
  1368. package/bin/skills/tinker/references/models-and-lora.md +0 -148
  1369. package/bin/skills/tinker/references/recipes.md +0 -326
  1370. package/bin/skills/tinker/references/reinforcement-learning.md +0 -357
  1371. package/bin/skills/tinker/references/rendering.md +0 -255
  1372. package/bin/skills/tinker/references/supervised-learning.md +0 -256
  1373. package/bin/skills/tinker-training-cost/SKILL.md +0 -187
  1374. package/bin/skills/tinker-training-cost/scripts/calculate_cost.py +0 -123
  1375. package/bin/skills/together-ai/SKILL.md +0 -722
  1376. package/bin/skills/torch_geometric/SKILL.md +0 -676
  1377. package/bin/skills/torch_geometric/references/datasets_reference.md +0 -574
  1378. package/bin/skills/torch_geometric/references/layers_reference.md +0 -485
  1379. package/bin/skills/torch_geometric/references/transforms_reference.md +0 -679
  1380. package/bin/skills/torch_geometric/scripts/benchmark_model.py +0 -309
  1381. package/bin/skills/torch_geometric/scripts/create_gnn_template.py +0 -529
  1382. package/bin/skills/torch_geometric/scripts/visualize_graph.py +0 -313
  1383. package/bin/skills/torchdrug/SKILL.md +0 -450
  1384. package/bin/skills/torchdrug/references/core_concepts.md +0 -565
  1385. package/bin/skills/torchdrug/references/datasets.md +0 -380
  1386. package/bin/skills/torchdrug/references/knowledge_graphs.md +0 -320
  1387. package/bin/skills/torchdrug/references/models_architectures.md +0 -541
  1388. package/bin/skills/torchdrug/references/molecular_generation.md +0 -352
  1389. package/bin/skills/torchdrug/references/molecular_property_prediction.md +0 -169
  1390. package/bin/skills/torchdrug/references/protein_modeling.md +0 -272
  1391. package/bin/skills/torchdrug/references/retrosynthesis.md +0 -436
  1392. package/bin/skills/torchforge/SKILL.md +0 -433
  1393. package/bin/skills/torchforge/references/api-reference.md +0 -327
  1394. package/bin/skills/torchforge/references/troubleshooting.md +0 -409
  1395. package/bin/skills/torchtitan/SKILL.md +0 -358
  1396. package/bin/skills/torchtitan/references/checkpoint.md +0 -181
  1397. package/bin/skills/torchtitan/references/custom-models.md +0 -258
  1398. package/bin/skills/torchtitan/references/float8.md +0 -133
  1399. package/bin/skills/torchtitan/references/fsdp.md +0 -126
  1400. package/bin/skills/training-data-pipeline/SKILL.md +0 -427
  1401. package/bin/skills/training-data-pipeline/references/data-quality.md +0 -136
  1402. package/bin/skills/training-data-pipeline/references/frontier-distillation.md +0 -129
  1403. package/bin/skills/training-data-pipeline/references/production-data-formatting.md +0 -126
  1404. package/bin/skills/transformer-lens/SKILL.md +0 -346
  1405. package/bin/skills/transformer-lens/references/README.md +0 -54
  1406. package/bin/skills/transformer-lens/references/api.md +0 -362
  1407. package/bin/skills/transformer-lens/references/tutorials.md +0 -339
  1408. package/bin/skills/transformers/SKILL.md +0 -164
  1409. package/bin/skills/transformers/references/generation.md +0 -467
  1410. package/bin/skills/transformers/references/models.md +0 -361
  1411. package/bin/skills/transformers/references/pipelines.md +0 -335
  1412. package/bin/skills/transformers/references/tokenizers.md +0 -447
  1413. package/bin/skills/transformers/references/training.md +0 -500
  1414. package/bin/skills/treatment-plans/README.md +0 -488
  1415. package/bin/skills/treatment-plans/SKILL.md +0 -1579
  1416. package/bin/skills/treatment-plans/assets/STYLING_QUICK_REFERENCE.md +0 -185
  1417. package/bin/skills/treatment-plans/assets/chronic_disease_management_plan.tex +0 -665
  1418. package/bin/skills/treatment-plans/assets/general_medical_treatment_plan.tex +0 -547
  1419. package/bin/skills/treatment-plans/assets/medical_treatment_plan.sty +0 -222
  1420. package/bin/skills/treatment-plans/assets/mental_health_treatment_plan.tex +0 -774
  1421. package/bin/skills/treatment-plans/assets/one_page_treatment_plan.tex +0 -193
  1422. package/bin/skills/treatment-plans/assets/pain_management_plan.tex +0 -799
  1423. package/bin/skills/treatment-plans/assets/perioperative_care_plan.tex +0 -753
  1424. package/bin/skills/treatment-plans/assets/quality_checklist.md +0 -471
  1425. package/bin/skills/treatment-plans/assets/rehabilitation_treatment_plan.tex +0 -756
  1426. package/bin/skills/treatment-plans/references/goal_setting_frameworks.md +0 -411
  1427. package/bin/skills/treatment-plans/references/intervention_guidelines.md +0 -507
  1428. package/bin/skills/treatment-plans/references/regulatory_compliance.md +0 -476
  1429. package/bin/skills/treatment-plans/references/specialty_specific_guidelines.md +0 -655
  1430. package/bin/skills/treatment-plans/references/treatment_plan_standards.md +0 -485
  1431. package/bin/skills/treatment-plans/scripts/check_completeness.py +0 -322
  1432. package/bin/skills/treatment-plans/scripts/generate_template.py +0 -233
  1433. package/bin/skills/treatment-plans/scripts/timeline_generator.py +0 -385
  1434. package/bin/skills/treatment-plans/scripts/validate_treatment_plan.py +0 -369
  1435. package/bin/skills/trl-fine-tuning/SKILL.md +0 -455
  1436. package/bin/skills/trl-fine-tuning/references/dpo-variants.md +0 -227
  1437. package/bin/skills/trl-fine-tuning/references/online-rl.md +0 -82
  1438. package/bin/skills/trl-fine-tuning/references/reward-modeling.md +0 -122
  1439. package/bin/skills/trl-fine-tuning/references/sft-training.md +0 -168
  1440. package/bin/skills/umap-learn/SKILL.md +0 -479
  1441. package/bin/skills/umap-learn/references/api_reference.md +0 -532
  1442. package/bin/skills/uniprot-database/SKILL.md +0 -195
  1443. package/bin/skills/uniprot-database/references/api_examples.md +0 -413
  1444. package/bin/skills/uniprot-database/references/api_fields.md +0 -275
  1445. package/bin/skills/uniprot-database/references/id_mapping_databases.md +0 -285
  1446. package/bin/skills/uniprot-database/references/query_syntax.md +0 -256
  1447. package/bin/skills/uniprot-database/scripts/uniprot_client.py +0 -341
  1448. package/bin/skills/unsloth/SKILL.md +0 -635
  1449. package/bin/skills/unsloth/docs/advanced-rl.md +0 -222
  1450. package/bin/skills/unsloth/docs/chat-templates.md +0 -141
  1451. package/bin/skills/unsloth/docs/datasets.md +0 -489
  1452. package/bin/skills/unsloth/docs/docker-extended.md +0 -99
  1453. package/bin/skills/unsloth/docs/dynamic-ggufs-2.0.md +0 -116
  1454. package/bin/skills/unsloth/docs/dynamic-ggufs-aider.md +0 -118
  1455. package/bin/skills/unsloth/docs/faq.md +0 -91
  1456. package/bin/skills/unsloth/docs/fp16-vs-bf16.md +0 -61
  1457. package/bin/skills/unsloth/docs/fp8-rl.md +0 -224
  1458. package/bin/skills/unsloth/docs/glm-4.7-flash.md +0 -997
  1459. package/bin/skills/unsloth/docs/inference-deployment-overview.md +0 -17
  1460. package/bin/skills/unsloth/docs/inference.md +0 -27
  1461. package/bin/skills/unsloth/docs/installation-docker.md +0 -155
  1462. package/bin/skills/unsloth/docs/installation-pip.md +0 -148
  1463. package/bin/skills/unsloth/docs/kernels-packing.md +0 -190
  1464. package/bin/skills/unsloth/docs/kimi-k2.5.md +0 -634
  1465. package/bin/skills/unsloth/docs/lm-studio.md +0 -235
  1466. package/bin/skills/unsloth/docs/lora-hot-swapping.md +0 -75
  1467. package/bin/skills/unsloth/docs/lora-hyperparameters.md +0 -363
  1468. package/bin/skills/unsloth/docs/memory-efficient-rl.md +0 -267
  1469. package/bin/skills/unsloth/docs/model-selection.md +0 -70
  1470. package/bin/skills/unsloth/docs/models.md +0 -532
  1471. package/bin/skills/unsloth/docs/multi-gpu-ddp.md +0 -90
  1472. package/bin/skills/unsloth/docs/notebooks.md +0 -223
  1473. package/bin/skills/unsloth/docs/overview.md +0 -110
  1474. package/bin/skills/unsloth/docs/qwen3-coder-next-extended.md +0 -900
  1475. package/bin/skills/unsloth/docs/qwen3-coder-next.md +0 -900
  1476. package/bin/skills/unsloth/docs/requirements.md +0 -45
  1477. package/bin/skills/unsloth/docs/reward-hacking.md +0 -25
  1478. package/bin/skills/unsloth/docs/saving-to-gguf.md +0 -138
  1479. package/bin/skills/unsloth/docs/saving-to-ollama.md +0 -46
  1480. package/bin/skills/unsloth/docs/sglang-guide.md +0 -278
  1481. package/bin/skills/unsloth/docs/speculative-decoding.md +0 -70
  1482. package/bin/skills/unsloth/docs/tool-calling.md +0 -334
  1483. package/bin/skills/unsloth/docs/troubleshooting-faq.md +0 -204
  1484. package/bin/skills/unsloth/docs/troubleshooting-inference.md +0 -26
  1485. package/bin/skills/unsloth/docs/tts-fine-tuning.md +0 -149
  1486. package/bin/skills/unsloth/docs/tutorial-grpo.md +0 -273
  1487. package/bin/skills/unsloth/docs/tutorial-llama3-ollama.md +0 -356
  1488. package/bin/skills/unsloth/docs/vision-fine-tuning.md +0 -135
  1489. package/bin/skills/unsloth/docs/vision-rl.md +0 -170
  1490. package/bin/skills/unsloth/docs/vllm-engine-arguments.md +0 -43
  1491. package/bin/skills/unsloth/docs/vllm-guide.md +0 -98
  1492. package/bin/skills/uspto-database/SKILL.md +0 -607
  1493. package/bin/skills/uspto-database/references/additional_apis.md +0 -394
  1494. package/bin/skills/uspto-database/references/patentsearch_api.md +0 -266
  1495. package/bin/skills/uspto-database/references/peds_api.md +0 -212
  1496. package/bin/skills/uspto-database/references/trademark_api.md +0 -358
  1497. package/bin/skills/uspto-database/scripts/patent_search.py +0 -290
  1498. package/bin/skills/uspto-database/scripts/peds_client.py +0 -285
  1499. package/bin/skills/uspto-database/scripts/trademark_client.py +0 -311
  1500. package/bin/skills/vaex/SKILL.md +0 -182
  1501. package/bin/skills/vaex/references/core_dataframes.md +0 -367
  1502. package/bin/skills/vaex/references/data_processing.md +0 -555
  1503. package/bin/skills/vaex/references/io_operations.md +0 -703
  1504. package/bin/skills/vaex/references/machine_learning.md +0 -728
  1505. package/bin/skills/vaex/references/performance.md +0 -571
  1506. package/bin/skills/vaex/references/visualization.md +0 -613
  1507. package/bin/skills/venue-templates/SKILL.md +0 -686
  1508. package/bin/skills/venue-templates/assets/examples/cell_summary_example.md +0 -247
  1509. package/bin/skills/venue-templates/assets/examples/medical_structured_abstract.md +0 -313
  1510. package/bin/skills/venue-templates/assets/examples/nature_abstract_examples.md +0 -213
  1511. package/bin/skills/venue-templates/assets/examples/neurips_introduction_example.md +0 -245
  1512. package/bin/skills/venue-templates/assets/grants/nih_specific_aims.tex +0 -235
  1513. package/bin/skills/venue-templates/assets/grants/nsf_proposal_template.tex +0 -375
  1514. package/bin/skills/venue-templates/assets/journals/nature_article.tex +0 -171
  1515. package/bin/skills/venue-templates/assets/journals/neurips_article.tex +0 -283
  1516. package/bin/skills/venue-templates/assets/journals/plos_one.tex +0 -317
  1517. package/bin/skills/venue-templates/assets/posters/beamerposter_academic.tex +0 -311
  1518. package/bin/skills/venue-templates/references/cell_press_style.md +0 -483
  1519. package/bin/skills/venue-templates/references/conferences_formatting.md +0 -564
  1520. package/bin/skills/venue-templates/references/cs_conference_style.md +0 -463
  1521. package/bin/skills/venue-templates/references/grants_requirements.md +0 -787
  1522. package/bin/skills/venue-templates/references/journals_formatting.md +0 -486
  1523. package/bin/skills/venue-templates/references/medical_journal_styles.md +0 -535
  1524. package/bin/skills/venue-templates/references/ml_conference_style.md +0 -556
  1525. package/bin/skills/venue-templates/references/nature_science_style.md +0 -405
  1526. package/bin/skills/venue-templates/references/posters_guidelines.md +0 -628
  1527. package/bin/skills/venue-templates/references/reviewer_expectations.md +0 -417
  1528. package/bin/skills/venue-templates/references/venue_writing_styles.md +0 -321
  1529. package/bin/skills/venue-templates/scripts/customize_template.py +0 -195
  1530. package/bin/skills/venue-templates/scripts/query_template.py +0 -266
  1531. package/bin/skills/venue-templates/scripts/validate_format.py +0 -250
  1532. package/bin/skills/verl/SKILL.md +0 -391
  1533. package/bin/skills/verl/references/api-reference.md +0 -301
  1534. package/bin/skills/verl/references/troubleshooting.md +0 -391
  1535. package/bin/skills/vllm/SKILL.md +0 -364
  1536. package/bin/skills/vllm/references/optimization.md +0 -226
  1537. package/bin/skills/vllm/references/quantization.md +0 -284
  1538. package/bin/skills/vllm/references/server-deployment.md +0 -255
  1539. package/bin/skills/vllm/references/troubleshooting.md +0 -447
  1540. package/bin/skills/weights-and-biases/SKILL.md +0 -590
  1541. package/bin/skills/weights-and-biases/references/artifacts.md +0 -584
  1542. package/bin/skills/weights-and-biases/references/integrations.md +0 -700
  1543. package/bin/skills/weights-and-biases/references/sweeps.md +0 -847
  1544. package/bin/skills/whisper/SKILL.md +0 -317
  1545. package/bin/skills/whisper/references/languages.md +0 -189
  1546. package/bin/skills/zarr-python/SKILL.md +0 -779
  1547. package/bin/skills/zarr-python/references/api_reference.md +0 -515
  1548. package/bin/skills/zinc-database/SKILL.md +0 -404
  1549. package/bin/skills/zinc-database/references/api_reference.md +0 -692
@@ -1,1055 +0,0 @@
1
- # Llama-Factory - Advanced
2
-
3
- **Pages:** 14
4
-
5
- ---
6
-
7
- ## GPT-OSS¶
8
-
9
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/best_practice/gpt-oss.html
10
-
11
- **Contents:**
12
- - GPT-OSS¶
13
- - 3 Steps to LoRA Fine-tuning for GPT-OSS¶
14
- - 1. Install LLaMA-Factory and transformers¶
15
- - 2. Train GPT-OSS on a single GPU (requires VRAM > 44 GB, multi-GPU supported)¶
16
- - 3. Merge LoRA Weights¶
17
- - Chat with the Fine-tuned Model¶
18
- - Full Fine-tuning Script¶
19
-
20
- Fine-tune the Model via Web UI:
21
-
22
- ---
23
-
24
- ## NPU 推理¶
25
-
26
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/npu_inference.html
27
-
28
- **Contents:**
29
- - NPU 推理¶
30
- - 环境安装¶
31
- - 版本需求¶
32
- - 硬件环境¶
33
- - 软件环境¶
34
- - vLLM-Ascend安装¶
35
- - LLaMA-Factory安装¶
36
- - 推理测试¶
37
- - 可视化界面¶
38
- - 性能对比¶
39
-
40
- Python:>= 3.10, < 3.12
41
-
42
- CANN >= 8.1.RC1,包括 toolkit、kernels、nnal。
43
-
44
- 使用下述命令安装 vLLM-Ascend 。
45
-
46
- 使用下述命令安装 LLaMA-Factory 。
47
-
48
- 使用下述命令启动LLaMA-Factory的可视化界面。
49
-
50
- 选择模型并切换到chat模式并将推理引擎修改为vLLM,然后点击加载模型。
51
-
52
- 在推理性能上。vLLM框架比huggingface的推理速度提升了超过一倍。
53
-
54
- ---
55
-
56
- ## Trainers¶
57
-
58
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/trainers.html
59
-
60
- **Contents:**
61
- - Trainers¶
62
- - Pre-training¶
63
- - Post-training¶
64
- - Supervised Fine-Tuning¶
65
- - RLHF¶
66
- - Reward model¶
67
- - PPO¶
68
- - DPO¶
69
- - KTO¶
70
-
71
- 大语言模型通过在一个大型的通用数据集上通过无监督学习的方式进行预训练来学习语言的表征/初始化模型权重/学习概率分布。 我们期望在预训练后模型能够处理大量、多种类的数据集,进而可以通过监督学习的方式来微调模型使其适应特定的任务。
72
-
73
- 预训练时,请将 stage 设置为 pt ,并确保使用的数据集符合 预训练数据集 格式 。
74
-
75
- 在预训练结束后,模型的参数得到初始化,模型能够理解语义、语法以及识别上下文关系,在处理一般性任务时有着不错的表现。 尽管模型涌现出的零样本学习,少样本学习的特性使其能在一定程度上完成特定任务, 但仅通过提示(prompt)并不一定能使其表现令人满意。因此,我们需要后训练(post-training)来使得模型在特定任务上也有足够好的表现。
76
-
77
- Supervised Fine-Tuning(监督微调)是一种在预训练模型上使用小规模有标签数据集进行训练的方法。 相比于预训练一个全新的模型,对已有的预训练模型进行监督微调是更快速更节省成本的途径。
78
-
79
- 监督微调时,请将 stage 设置为 sft 。 下面提供监督微调的配置示例:
80
-
81
- 由于在监督微调中语言模型学习的数据来自互联网,所以模型可能无法很好地遵循用户指令,甚至可能输出非法、暴力的内容,因此我们需要将模型行为与用户需求对齐(alignment)。 通过 RLHF(Reinforcement Learning from Human Feedback) 方法,我们可以通过人类反馈来进一步微调模型,使得模型能够更好更安全地遵循用户指令。
82
-
83
- 但是,获取真实的人类数据是十分耗时且昂贵的。一个自然的想法是我们可以训练一个奖励模型(reward model)来代替人类对语言模型的输出进行评价。 为了训练这个奖励模型,我们需要让奖励模型获知人类偏好,而这通常通过输入经过人类标注的偏好数据集来实现。 在偏好数据集中,数据由三部分组成:输入、好的回答、坏的回答。奖励模型在偏好数据集上训练,从而可以更符合人类偏好地评价语言模型的输出。
84
-
85
- 在训练奖励模型时,请将 stage 设置为 rm ,确保使用的数据集符合 偏好数据集 格式并且指定奖励模型的保存路径。 以下提供一个示例:
86
-
87
- 在训练奖励完模型之后,我们可以开始进行模型的强化学习部分。与监督学习不同,在强化学习中我们没有标注好的数据。语言模型接受prompt作为输入,其输出作为奖励模型的输入。奖励模型评价语言模型的输出,并将评价返回给语言模型。确保两个模型都能良好运行是一个具有挑战性的任务。 一种实现方式是使用近端策略优化(PPO,Proximal Policy Optimization)。其主要思想是:我们既希望语言模型的输出能够尽可能地获得奖励模型的高评价,又不希望语言模型的变化过于“激进”。 通过这种方法,我们可以使得模型在学习趋近人类偏好的同时不过多地丢失其原有的解决问题的能力。
88
-
89
- 在使用 PPO 进行强化学习时,请将 stage 设置为 ppo,并且指定所使用奖励模型的路径。 下面是一个示例:
90
-
91
- 既然同时保证语言模型与奖励模型的良好运行是有挑战性的,一种想法是我们可以丢弃奖励模型, 进而直接基于人类偏好训练我们的语言模型,这大大简化了训练过程。
92
-
93
- 在使用 DPO 时,请将 stage 设置为 dpo,确保使用的数据集符合 偏好数据集 格式并且设置偏好优化相关参数。 以下是一个示例:
94
-
95
- KTO(Kahneman-Taversky Optimization) 的出现是为了解决成对的偏好数据难以获得的问题。 KTO使用了一种新的损失函数使其只需二元的标记数据, 即只需标注回答的好坏即可训练,并取得与 DPO 相似甚至更好的效果。
96
-
97
- 在使用 KTO 时,请将 stage 设置为 kto ,设置偏好优化相关参数并使用 KTO 数据集。
98
-
99
- ---
100
-
101
- ## 模型支持¶
102
-
103
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/model_support.html
104
-
105
- **Contents:**
106
- - 模型支持¶
107
- - 注册 template¶
108
- - 多模态数据构建¶
109
- - 提供模型路径¶
110
-
111
- LLaMA-Factory 允许用户添加自定义模型支持。我们将以 LLaMA-4 多模态模型为例,详细介绍如何为新模型添加支持。对于多模态模型,我们需要完成两个主要任务:
112
-
113
- 首先,我们可以通过以下方法获取 LLaMA-4 模型的 template
114
-
115
- 输出如下。通过观察输出我们可以得到模型的 chat_template。除此以外也可以通过 huggingface repo 来获取模型的 template.
116
-
117
- 通过观察输出,我们可以得知 LLaMA-4 的 chat_template 主要由以下几部分组成:
118
-
119
- 用户消息: <|header_start|>user<|header_end|>\n\n{{content}}<|eot|>
120
-
121
- 助手消息: <|header_start|>assistant<|header_end|>\n\n{{content}}<|eot|>
122
-
123
- 系统消息: <|header_start|>system<|header_end|>\n\n{{content}}<|eot|>
124
-
125
- 工具消息: <|header_start|>ipython<|header_end|>\n\n"{{content}}"<|eot|>
126
-
127
- 我们可以在 src/llamafactory/data/template.py 中使用 register_template 方法为自定义模型注册 chat_template。 在实际应用中,我们往往会在用户输入的信息后添加助手回复模板的头部 <|header_start|>assistant<|header_end|> 来引导模型进行回复。 因此我们可以看到,用户消息和工具输出的模板中都附有了助手回复的头部,而助手消息格式 format_assitant 也因此省略了助手回复的头部, 只保留其内容部分 {{content}}<|eot|>
128
-
129
- 我们可以根据上面的输出完成 name, format_user, format_assistant, format_system 与 format_observation 字段的填写。
130
-
131
- format_prefix 字段用于指定模型的开头部分,通常可以在 tokenizer_config.json 中找到。
132
-
133
- stop_words 字段用于指定模型的停止词,可以在 generation_config.json 中找到 eos_token_id,再把 eos_token_id 对应的 token 填入。
134
-
135
- 对于多模态模型,我们还需要在 mm_plugin 字段中指定多模态插件。
136
-
137
- 对于多模态模型,我们参照原始模型在 LLaMA-Factory 中实现多模态数据的解析。
138
-
139
- 我们可以在 src/llamafactory/data/mm_plugin.py 中实现 Llama4Plugin 类来解析多模态数据。
140
-
141
- Llama4Plugin 类继承自 BasePlugin 类,并实现了 get_mm_inputs 和 process_messages 方法来解析多模态数据。
142
-
143
- get_mm_inputs 的作用是将图像、视频等多模态数据转化为模型可以接收的输入,如 pixel_values。为实现 get_mm_inputs,首先我们需要检查 llama4 的 processor 是否可以与 已有实现 兼容。 模型官方仓库中的 processing_llama4.py 表明 llama4 的 processor 返回数据包含字段 pixel_values,这与 LLaMA-Factory 中的已有实现兼容。因此,我们只需要参照已有的 get_mm_inputs 方法实现即可。
144
-
145
- process_messages 的作用是根据输入图片/视频的大小,数量等信息在 messages 中插入相应数量的占位符,以便模型可以正确解析多模态数据。 我们需要参考 原仓库实现 以及 LLaMA-Factory 中的规范返回 list[dict[str, str]] 类型的 messages 。
146
-
147
- 最后, 在 src/llamafactory/extras/constants.py 中提供模型的下载路径。 例如:
148
-
149
- **Examples:**
150
-
151
- Example 1 (python):
152
- ```python
153
- ========== Template ==========
154
- <|begin_of_text|><|header_start|>user<|header_end|>
155
-
156
- {{content}}<|eot|><|header_start|>assistant<|header_end|>
157
-
158
- {{content}}<|eot|><|header_start|>system<|header_end|>
159
-
160
- {{content}}<|eot|><|header_start|>ipython<|header_end|>
161
-
162
- "{{content}}"<|eot|><|header_start|>assistant<|header_end|>
163
- ```
164
-
165
- Example 2 (python):
166
- ```python
167
- register_template(
168
- # 模板名称
169
- name="llama4",
170
- # 用户消息格式,结尾附有 generation prompt 的模板
171
- format_user=StringFormatter(
172
- slots=["<|header_start|>user<|header_end|>\n\n{{content}}<|eot|><|header_start|>assistant<|header_end|>\n\n"]
173
- ),
174
- # 助手消息格式
175
- format_assistant=StringFormatter(slots=["{{content}}<|eot|>"]),
176
- # 系统消息格式
177
- format_system=StringFormatter(slots=["<|header_start|>system<|header_end|>\n\n{{content}}<|eot|>"]),
178
- # 函数调用格式
179
- format_function=FunctionFormatter(slots=["{{content}}<|eot|>"], tool_format="llama3"),
180
- # 工具输出格式,结尾附有 generation prompt 的模板
181
- format_observation=StringFormatter(
182
- slots=[
183
- "<|header_start|>ipython<|header_end|>\n\n{{content}}<|eot|><|header_start|>assistant<|header_end|>\n\n"
184
- ]
185
- ),
186
- # 工具调用格式
187
- format_tools=ToolFormatter(tool_format="llama3"),
188
- format_prefix=EmptyFormatter(slots=[{"bos_token"}]),
189
- stop_words=["<|eot|>", "<|eom|>"],
190
- mm_plugin=get_mm_plugin(name="llama4", image_token="<|image|>"),
191
- )
192
- ```
193
-
194
- ---
195
-
196
- ## Quantization¶
197
-
198
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/quantization.html
199
-
200
- **Contents:**
201
- - Quantization¶
202
- - PTQ¶
203
- - GPTQ¶
204
- - QAT¶
205
- - AWQ¶
206
- - AQLM¶
207
- - OFTQ¶
208
- - bitsandbytes¶
209
- - HQQ¶
210
- - EETQ¶
211
-
212
- 随着语言模型规模的不断增大,其训练的难度和成本已成为共识。 而随着用户数量的增加,模型推理的成本也在不断攀升,甚至可能成为限制模型部署的首要因素。 因此,我们需要对模型进行压缩以加速推理过程,而模型量化是其中一种有效的方法。
213
-
214
- 大语言模型的参数通常以高精度浮点数存储,这导致模型推理需要大量计算资源。 量化技术通过将高精度数据类型存储的参数转换为低精度数据类型存储, 可以在不改变模型参数量和架构的前提下加速推理过程。这种方法使得模型的部署更加经济高效,也更具可行性。
215
-
216
- 浮点数一般由3部分组成:符号位、指数位和尾数位。指数位越大,可表示的数字范围越大。尾数位越大、数字的精度越高。
217
-
218
- 量化可以根据何时量化分为:后训练量化和训练感知量化,也可以根据量化参数的确定方式分为:静态量化和动态量化。
219
-
220
- 后训练量化(PTQ, Post-Training Quantization)一般是指在模型预训练完成后,基于校准数据集(calibration dataset)确定量化参数进而对模型进行量化。
221
-
222
- GPTQ(Group-wise Precision Tuning Quantization)是一种静态的后训练量化技术。”静态”指的是预训练模型一旦确定,经过量化后量化参数不再更改。GPTQ 量化技术将 fp16 精度的模型量化为 4-bit ,在节省了约 75% 的显存的同时大幅提高了推理速度。 为了使用GPTQ量化模型,您需要指定量化模型名称或路径,例如 model_name_or_path: TechxGenus/Meta-Llama-3-8B-Instruct-GPTQ
223
-
224
- 在训练感知量化(QAT, Quantization-Aware Training)中,模型一般在预训练过程中被量化,然后又在训练数据上再次微调,得到最后的量化模型。
225
-
226
- AWQ(Activation-Aware Layer Quantization)是一种静态的后训练量化技术。其思想基于:有很小一部分的权重十分重要,为了保持性能这些权重不会被量化。 AWQ 的优势在于其需要的校准数据集更小,且在指令微调和多模态模型上表现良好。 为了使用 AWQ 量化模型,您需要指定量化模型名称或路径,例如 model_name_or_path: TechxGenus/Meta-Llama-3-8B-Instruct-AWQ
227
-
228
- AQLM(Additive Quantization of Language Models)作为一种只对模型权重进行量化的PTQ方法,在 2-bit 量化下达到了当时的最佳表现,并且在 3-bit 和 4-bit 量化下也展示了性能的提升。 尽管 AQLM 在模型推理速度方面的提升并不是最显著的,但其在 2-bit 量化下的优异表现意味着您可以以极低的显存占用来部署大模型。
229
-
230
- OFTQ(On-the-fly Quantization)指的是模型无需校准数据集,直接在推理阶段进行量化。OFTQ是一种动态的后训练量化技术. OFTQ在保持性能的同时。 因此,在使用OFTQ量化方法时,您需要指定预训练模型、指定量化方法 quantization_method 和指定量化位数 quantization_bit 下面提供了一个使用bitsandbytes量化方法的配置示例:
231
-
232
- 区别于 GPTQ, bitsandbytes 是一种动态的后训练量化技术。bitsandbytes 使得大于 1B 的语言模型也能在 8-bit 量化后不过多地损失性能。 经过bitsandbytes 8-bit 量化的模型能够在保持性能的情况下节省约50%的显存。
233
-
234
- 依赖校准数据集的方法往往准确度较高,不依赖校准数据集的方法往往速度较快。HQQ(Half-Quadratic Quantization)希望能在准确度和速度之间取得较好的平衡。作为一种动态的后训练量化方法,HQQ无需校准阶段, 但能够取得与需要校准数据集的方法相当的准确度,并且有着极快的推理速度。
235
-
236
- EETQ(Easy and Efficient Quantization for Transformers)是一种只对模型权重进行量化的PTQ方法。具有较快的速度和简单易用的特性。
237
-
238
- ---
239
-
240
- ## NPU 训练¶
241
-
242
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/npu_training.html
243
-
244
- **Contents:**
245
- - NPU 训练¶
246
- - 支持设备¶
247
- - 单机微调¶
248
- - 多机微调¶
249
-
250
- Atlas A2训练系列(Atlas 800T A2, Atlas 900 A2 PoD, Atlas 200T A2 Box16, Atlas 300T A2)
251
-
252
- Atlas 800I A2推理系列(Atlas 800I A2)
253
-
254
- 以 davinci0 单卡为例,下载并使用ascend llamafactory镜像。
255
-
256
- 首先在环境当前目录下执行如下命令,进入容器。
257
-
258
- 如果在单机上使用多卡微调时,可使用 --device /dev/davinci1, --device /dev/davinci2, ... 来增加 NPU 卡。
259
-
260
- 昇腾 NPU 卡从 0 开始编号,docker 容器内也是如此;
261
-
262
- 如映射物理机上的 davinci6,davinci7 NPU 卡到容器内使用,其对应的卡号分别为 0,1
263
-
264
- 进入docker后安装相关依赖、设置环境变量、配置 LoRA 微调参数文件(qwen1_5_lora_sft_ds.yaml)
265
-
266
- ASCEND_RT_VISIBLE_DEVICES=0指定使用容器内卡号
267
-
268
- USE_MODELSCOPE_HUB=1使用modelscope
269
-
270
- 在 LLAMA-Factory 目录下,创建如下 qwen1_5_lora_sft_ds.yaml:
271
-
272
- 使用 torchrun 启动 LoRA 微调,如正常输出模型加载、损失 loss 等日志,即说明成功微调。
273
-
274
- 经 LoRA 微调后,通过 llamafactory-cli chat 使用微调后的模型进行交互对话,使用 Ctrl+C 或输入 exit 退出该问答聊天。
275
-
276
- 多机微调时,不建议使用容器部署方式(单机都不够用的情况下,起多个容器资源更加紧张),请直接在每个节点安装 llamafactory(请参考 NPU 中的安装步骤),同时仍需要安装 DeepSpeed 和 ModelScope:
277
-
278
- 安装成功后,请在每个节点上使用 export ASCEND_RT_VISIBLE_DEVICES=0,1,2,3 显式指定所需的 NPU 卡号,不指定时默认使用当前节点的所有 NPU 卡。
279
-
280
- 然后,必须在每个节点上使用 export HCCL_SOCKET_IFNAME=eth0 来指定当前节点的 HCCL 通信网卡(请使用目标网卡名替换 eth0)。
281
-
282
- 以两机环境为例,分别在主、从节点(机器)上执行如下两条命令即可启动多机训练:
283
-
284
- ---
285
-
286
- ## NPU¶
287
-
288
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/npu.html
289
-
290
- **Contents:**
291
- - NPU¶
292
- - Install By Docker¶
293
- - 使用 docker-compose 构建并启动 docker 容器¶
294
- - 不使用 docker-compose¶
295
- - Install By pip¶
296
- - 依赖1: NPU 驱动¶
297
- - 依赖2: NPU 开发包¶
298
- - 依赖3: torch-npu¶
299
- - 依赖校验¶
300
- - Verification¶
301
-
302
- 目前LLaMA-Factory 通过 torch-npu 库完成了对华为昇腾 910b 系列芯片的支持, 包含 32GB 和 64GB 两个版本。跟其他使用相比,会需要额外3个前置条件
303
-
304
- CANN Toolkit 和 Kernels库正常安装
305
-
306
- 为方便昇腾用户使用,LLaMA-Factory 提供已预装昇腾环境的 Install By Docker 及自行安装昇腾环境,Install By pip 两种方式,可按需自行选择:
307
-
308
- 请确保宿主机已根据昇腾卡型号成功安装对应的固件和驱动,可参考 快速安装昇腾环境 指引。
309
-
310
- LLaMA-Factory 提供 使用 docker-compose 构建并启动 docker 容器 和 不使用 docker-compose 两种构建方式,请根据需求选择其一。
311
-
312
- 进入 LLaMA-Factory 项目中存放 Dockerfile 及 docker-compose.yaml 的 docker-npu 目录:
313
-
314
- 构建 docker 镜像并启动 docker 容器:
315
-
316
- 使用 docker build 直接构建 docker 镜像:
317
-
318
- 自行 pip 安装时, python 版本建议使用3.10, 目前该版本对于 NPU 的使用情况会相对稳定,其他版本可能会遇到一些未知的情况
319
-
320
- 可以按照 快速安装昇腾环境 指引,或者使用以下命令完成快速安装:
321
-
322
- 依赖3建议在安装 LLaMA-Factory 的时候一起选配安装, 把 torch-npu 一起加入安装目标,命令如下
323
-
324
- 3个依赖都安装后,可以通过如下的 python 脚本对 torch_npu 的可用情况做一下校验
325
-
326
- 使用以下指令对 LLaMA-Factory × 昇腾的安装进行校验:
327
-
328
- 如下所示,正确显示 LLaMA-Factory、PyTorch NPU 和 CANN 版本号及 NPU 型号等信息即说明安装成功。
329
-
330
- 前面依赖安装完毕和完成校验后,即可像文档的其他部分一样正常使用 llamafactory-cli 的相关功能, NPU 的使用是无侵入的。主要的区别是需要修改一下命令行中 设备变量使用 将原来的 Nvidia 卡的变量 CUDA_VISIBLE_DEVICES 替换为 ASCEND_RT_VISIBLE_DEVICES, 类似如下命令
331
-
332
- 通过 ASCEND_RT_VISIBLE_DEVICES 环境变量指定昇腾 NPU 卡,如 ASCEND_RT_VISIBLE_DEVICES=0,1,2,3 指定使用 0,1,2,3四张 NPU 卡进行微调/推理。
333
-
334
- 昇腾 NPU 卡从 0 开始编号,docker 容器内也是如此; 如映射物理机上的 6,7 号 NPU 卡到容器内使用,其对应的卡号分别为 0,1
335
-
336
- 检查是否安装 torch-npu,建议通过 pip install -e '.[torch-npu,metrics]' 安装 LLaMA-Factory。
337
-
338
- Q:使用昇腾 NPU 推理报错 RuntimeError: ACL stream synchronize failed, error code:507018
339
-
340
- A: 设置 do_sample: false,取消随机抽样策略。
341
-
342
- https://github.com/hiyouga/LLaMA-Factory/issues/3840
343
-
344
- Q:使用 ChatGLM 系列模型微调/训练模型时,报错 NotImplementedError: Unknown device for graph fuser
345
-
346
- A: 在 modelscope 或 huggingface 下载的 repo 里修改 modeling_chatglm.py 代码,取消 torch.jit 装饰器注释
347
-
348
- https://github.com/hiyouga/LLaMA-Factory/issues/3788
349
-
350
- https://github.com/hiyouga/LLaMA-Factory/issues/4228
351
-
352
- Q:微调/训练启动后,HCCL 报错,包含如下关键信息:
353
-
354
- A: 杀掉 device 侧所有进程,等待 10s 后重新启动训练。
355
-
356
- https://github.com/hiyouga/LLaMA-Factory/issues/3839
357
-
358
- Q:使用 TeleChat 模型在昇腾 NPU 推理时,报错 AssertionError: Torch not compiled with CUDA enabled
359
-
360
- A: 此问题一般由代码中包含 cuda 相关硬编码造成,根据报错信息,找到 cuda 硬编码所在位置,对应修改为 NPU 代码。如 .cuda() 替换为 .npu() ; .to("cuda") 替换为 .to("npu")
361
-
362
- Q:模型微调遇到报错 DeviceType must be NPU. Actual DeviceType is: cpu,例如下列报错信息
363
-
364
- A: 此类报错通常为部分 Tensor 未放到 NPU 上,请确保报错中算子所涉及的操作数均在 NPU 上。如上面的报错中,MulKernelNpuOpApi 算子为乘法算子,应确保 next_tokens 和 unfinished_sequences 均已放在 NPU 上。
365
-
366
- 如需更多 LLaMA-Factory × 昇腾实践指引,可参考 全流程昇腾实践 。
367
-
368
- ---
369
-
370
- ## Monitors¶
371
-
372
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/monitor.html
373
-
374
- **Contents:**
375
- - Monitors¶
376
- - LlamaBoard¶
377
- - SwanLab¶
378
- - TensorBoard¶
379
- - Wandb¶
380
- - MLflow¶
381
-
382
- LLaMA-Factory 支持多种训练可视化工具,包括:LlamaBoard 、 SwanLab、TensorBoard 、 Wandb 、 MLflow 。
383
-
384
- LlamaBoard 是指 WebUI 中自带的Loss曲线看板,可以方便的查看训练过程中的Loss变化情况。
385
-
386
- 如果你想使用 LlamaBoard,只需使用 WebUI 启动训练即可。
387
-
388
- SwanLab 是一个开源的训练跟踪与可视化工具,云端和离线均可使用,支持超参数记录、指标记录、多实验对比、硬件监控、实验环境记录等功能,可以有效地帮助开发者管理实验。
389
-
390
- 如果你想使用 SwanLab,请在启动训练时在训练配置文件中添加以下参数:
391
-
392
- 或者,在WebUI的 SwanLab 模块中开启 SwanLab 记录:
393
-
394
- TensorBoard 是 TensorFlow 开源的离线训练跟踪工具,可以用于记录与可视化训练过程。
395
-
396
- 如果你想使用 TensorBoard,请在启动训练时在训练配置文件中添加以下参数:
397
-
398
- 或者,在WebUI的 其他参数设置 模块中的 启用外部记录面板 中开启 TensorBoard 记录:
399
-
400
- Wandb(Weights and Biases)是一个云端的训练跟踪工具,可以用于记录与可视化训练过程。
401
-
402
- 如果你想使用 Wandb,请在启动训练时在训练配置文件中添加以下参数:
403
-
404
- 或者,在WebUI的 其他参数设置 模块中的 启用外部记录面板 中开启 Wandb 记录:
405
-
406
- MLflow 是Databricks开源的离线训练跟踪工具,用于记录与可视化训练过程。
407
-
408
- 如果你想使用 MLflow,请在启动训练时在训练配置文件中添加以下参数:
409
-
410
- 或者,在WebUI的 其他参数设置 模块中的 启用外部记录面板 中开启 MLflow 记录:
411
-
412
- ---
413
-
414
- ## Acceleration¶
415
-
416
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/acceleration.html
417
-
418
- **Contents:**
419
- - Acceleration¶
420
- - FlashAttention¶
421
- - Unsloth¶
422
- - Liger Kernel¶
423
-
424
- LLaMA-Factory 支持多种加速技术,包括:FlashAttention 、 Unsloth 、 Liger Kernel 。
425
-
426
- FlashAttention 能够加快注意力机制的运算速度,同时减少对内存的使用。
427
-
428
- 如果您想使用 FlashAttention,请在启动训练时在训练配置文件中添加以下参数:
429
-
430
- Unsloth 框架支持 Llama, Mistral, Phi-3, Gemma, Yi, DeepSeek, Qwen等大语言模型并且支持 4-bit 和 16-bit 的 QLoRA/LoRA 微调,该框架在提高运算速度的同时还减少了显存占用。
431
-
432
- 如果您想使用 Unsloth, 请在启动训练时在训练配置文件中添加以下参数:
433
-
434
- Liger Kernel 是一个大语言模型训练的性能优化框架, 可有效地提高吞吐量并减少内存占用。
435
-
436
- 如果您想使用 Liger Kernel,请在启动训练时在训练配置文件中添加以下参数:
437
-
438
- ---
439
-
440
- ## Distributed Training¶
441
-
442
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/distributed.html
443
-
444
- **Contents:**
445
- - Distributed Training¶
446
- - NativeDDP¶
447
- - 单机多卡¶
448
- - llamafactory-cli¶
449
- - torchrun¶
450
- - accelerate¶
451
- - 多机多卡¶
452
- - llamafactory-cli¶
453
- - torchrun¶
454
- - accelerate¶
455
-
456
- LLaMA-Factory 支持单机多卡和多机多卡分布式训练。同时也支持 DDP , DeepSpeed 和 FSDP 三种分布式引擎。
457
-
458
- DDP (DistributedDataParallel) 通过实现模型并行和数据并行实现训练加速。 使用 DDP 的程序需要生成多个进程并且为每个进程创建一个 DDP 实例,他们之间通过 torch.distributed 库同步。
459
-
460
- DeepSpeed 是微软开发的分布式训练引擎,并提供ZeRO(Zero Redundancy Optimizer)、offload、Sparse Attention、1 bit Adam、流水线并行等优化技术。 您可以根据任务需求与设备选择使用。
461
-
462
- FSDP 通过全切片数据并行技术(Fully Sharded Data Parallel)来处理更多更大的模型。在 DDP 中,每张 GPU 都各自保留了一份完整的模型参数和优化器参数。而 FSDP 切分了模型参数、梯度与优化器参数,使得每张 GPU 只保留这些参数的一部分。 除了并行技术之外,FSDP 还支持将模型参数卸载至CPU,从而进一步降低显存需求。
463
-
464
- NativeDDP 是 PyTorch 提供的一种分布式训练方式,您可以通过以下命令启动训练:
465
-
466
- 您可以使用 llamafactory-cli 启动 NativeDDP 引擎。
467
-
468
- 如果 CUDA_VISIBLE_DEVICES 没有指定,则默认使用所有GPU。如果需要指定GPU,例如第0、1个GPU,可以使用:
469
-
470
- 您也可以使用 torchrun 指令启动 NativeDDP 引擎进行单机多卡训练。下面提供一个示例:
471
-
472
- 您还可以使用 accelerate 指令启动进行单机多卡训练。
473
-
474
- 首先运行以下命令,根据需求回答一系列问题后生成配置文件:
475
-
476
- 您也可以使用 torchrun 指令启动 NativeDDP 引擎进行多机多卡训练。
477
-
478
- 您还可以使用 accelerate 指令启动进行多机多卡训练。
479
-
480
- 首先运行以下命令,根据需求回答一系列问题后生成配置文件:
481
-
482
- DeepSpeed 是由微软开发的一个开源深度学习优化库,旨在提高大模型训练的效率和速度。在使用 DeepSpeed 之前,您需要先估计训练任务的显存大小,再根据任务需求与资源情况选择合适的 ZeRO 阶段。
483
-
484
- ZeRO-1: 仅划分优化器参数,每个GPU各有一份完整的模型参数与梯度。
485
-
486
- ZeRO-2: 划分优化器参数与梯度,每个GPU各有一份完整的模型参数。
487
-
488
- ZeRO-3: 划分优化器参数、梯度与模型参数。
489
-
490
- 简单来说:从 ZeRO-1 到 ZeRO-3,阶段数越高,显存需求越小,但是训练速度也依次变慢。此外,设置 offload_param=cpu 参数会大幅减小显存需求,但会极大地使训练速度减慢。因此,如果您有足够的显存, 应当使用 ZeRO-1,并且确保 offload_param=none。
491
-
492
- LLaMA-Factory提供了使用不同阶段的 DeepSpeed 配置文件的示例。包括:
493
-
494
- https://huggingface.co/docs/transformers/deepspeed 提供了更为详细的介绍。
495
-
496
- 您可以使用 llamafactory-cli 启动 DeepSpeed 引擎进行单机多卡训练。
497
-
498
- 为了启动 DeepSpeed 引擎,配置文件中 deepspeed 参数指定了 DeepSpeed 配置文件的路径:
499
-
500
- 您也可以使用 deepspeed 指令启动 DeepSpeed 引擎进行单机多卡训练。
501
-
502
- 使用 deepspeed 指令启动 DeepSpeed 引擎时您无法使用 CUDA_VISIBLE_DEVICES 指定GPU。而需要:
503
-
504
- --include localhost:1 表示只是用本节点的gpu1。
505
-
506
- LLaMA-Factory 支持使用 DeepSpeed 的多机多卡训练,您可以通过以下命令启动:
507
-
508
- 您也可以使用 deepspeed 指令来启动多机多卡训练。
509
-
510
- hostfile的每一行指定一个节点,每行的格式为 <hostname> slots=<num_slots> , 其中 <hostname> 是节点的主机名, <num_slots> 是该节点上的GPU数量。下面是一个例子: .. code-block:
511
-
512
- 请在 https://www.deepspeed.ai/getting-started/ 了解更多。
513
-
514
- 如果没有指定 hostfile 变量, DeepSpeed 会搜索 /job/hostfile 文件。如果仍未找到,那么 DeepSpeed 会使用本机上所有可用的GPU。
515
-
516
- 您还可以使用 accelerate 指令启动 DeepSpeed 引擎。 首先通过以下命令生成 DeepSpeed 配置文件:
517
-
518
- 只需在 ZeRO-0 的基础上修改 zero_optimization 中的 stage 参数即可。
519
-
520
- 只需在 ZeRO-0 的基础上在 zero_optimization 中添加 offload_optimizer 参数即可。
521
-
522
- 只需在 ZeRO-0 的基础上修改 zero_optimization 中的参数。
523
-
524
- 只需在 ZeRO-3 的基础上添加 zero_optimization 中的 offload_optimizer 和 offload_param 参数即可。
525
-
526
- https://www.deepspeed.ai/docs/config-json/ 提供了关于deepspeed配置文件的更详细的介绍。
527
-
528
- PyTorch 的全切片数据并行技术 FSDP (Fully Sharded Data Parallel)能让我们处理更多更大的模型。LLaMA-Factory支持使用 FSDP 引擎进行分布式训练。
529
-
530
- FSDP 的参数 ShardingStrategy 的不同取值决定了模型的划分方式:
531
-
532
- FULL_SHARD: 将模型参数、梯度和优化器状态都切分到不同的GPU上,类似ZeRO-3。
533
-
534
- SHARD_GRAD_OP: 将梯度、优化器状态切分到不同的GPU上,每个GPU仍各自保留一份完整的模型参数。类似ZeRO-2。
535
-
536
- NO_SHARD: 不切分任何参数。类似ZeRO-0。
537
-
538
- 您只需根据需要修改 examples/accelerate/fsdp_config.yaml 以及 examples/extras/fsdp_qlora/llama3_lora_sft.yaml ,文件然后运行以下命令即可启动 FSDP+QLoRA 微调:
539
-
540
- 此外,您也可以使用 accelerate 启动 FSDP 引擎, 节点数与 GPU 数可以通过 num_machines 和 num_processes 指定。对此,Huggingface 提供了便捷的配置功能。 只需运行:
541
-
542
- 根据提示回答一系列问题后,我们就可以生成 FSDP 所需的配置文件。
543
-
544
- 当然您也可以根据需求自行配置 fsdp_config.yaml 。
545
-
546
- 请确保 num_processes 和实际使用的总GPU数量一致
547
-
548
- 不要在 FSDP+QLoRA 中使用 GPTQ/AWQ 模型
549
-
550
- ---
551
-
552
- ## Arguments¶
553
-
554
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/arguments.html
555
-
556
- **Contents:**
557
- - Arguments¶
558
- - Finetuning Arguments¶
559
- - 基本参数¶
560
- - LoRA¶
561
- - RLHF¶
562
- - Freeze¶
563
- - Apollo¶
564
- - BAdam¶
565
- - GaLore¶
566
- - Data Arguments¶
567
-
568
- 是否以纯 bf16 精度训练模型(不使用 AMP)。
569
-
570
- Literal[“pt”, “sft”, “rm”, “ppo”, “dpo”, “kto”]
571
-
572
- Literal[“lora”, “freeze”, “full”]
573
-
574
- 是否仅训练扩展块中的参数(LLaMA Pro 模式)。
575
-
576
- freeze_multi_modal_projector
577
-
578
- 是否在评估时计算 token 级别的准确率。
579
-
580
- include_effective_tokens_per_second
581
-
582
- 除 LoRA 层之外设置为可训练并保存在最终检查点中的模块名称。使用逗号分隔多个模块。
583
-
584
- LoRA 缩放系数。一般情况下为 lora_rank * 2。
585
-
586
- LoRA 微调的本征维数 r,r 越大可训练的参数越多。
587
-
588
- 应用 LoRA 方法的模块名称。使用逗号分隔多个模块,使用 all 指定所有模块。
589
-
590
- LoRA+ 学习率比例(λ = ηB/ηA)。 ηA, ηB 分别是 adapter matrices A 与 B 的学习率。
591
-
592
- loraplus_lr_embedding
593
-
594
- 是否使用秩稳定 LoRA (Rank-Stabilized LoRA)。
595
-
596
- 是否使用权重分解 LoRA(Weight-Decomposed LoRA)。
597
-
598
- PiSSA 中 FSVD 执行的迭代步数。使用 -1 将其禁用。
599
-
600
- 是否将 PiSSA 适配器转换为正常的 LoRA 适配器。
601
-
602
- 是否创建一个具有随机初始化权重的新适配器。
603
-
604
- DPO 训练中的 sft loss 系数。
605
-
606
- Literal[“sigmoid”, “hinge”, “ipo”, “kto_pair”, “orpo”, “simpo”]
607
-
608
- DPO 训练中使用的偏好损失类型。可选值为: sigmoid, hinge, ipo, kto_pair, orpo, simpo。
609
-
610
- 标签平滑系数,取值范围为 [0,0.5]。
611
-
612
- KTO 训练中 chosen 标签 loss 的权重。
613
-
614
- KTO 训练中 rejected 标签 loss 的权重。
615
-
616
- SimPO 损失中的 reward margin。
617
-
618
- PPO 训练中的 mini-batch 大小。
619
-
620
- PPO 训练中自适应 KL 控制的目标 KL 值。
621
-
622
- PPO 或 DPO 训练中使用的参考模型路径。
623
-
624
- ref_model_quantization_bit
625
-
626
- 参考模型的量化位数,支持 4 位或 8 位量化。
627
-
628
- reward_model_adapters
629
-
630
- reward_model_quantization_bit
631
-
632
- Literal[“lora”, “full”, “api”]
633
-
634
- PPO 训练中使用的奖励模型类型。可选值为: lora, full, api。
635
-
636
- freeze_trainable_layers
637
-
638
- 可训练层的数量。正数表示最后 n 层被设置为可训练的,负数表示前 n 层被设置为可训练的。
639
-
640
- freeze_trainable_modules
641
-
642
- 可训练层的名称。使用 all 来指定所有模块。
643
-
644
- 除了隐藏层外可以被训练的模块名称,被指定的模块将会被设置为可训练的。使用逗号分隔多个模块。
645
-
646
- 适用 APOLLO 的模块名称。使用逗号分隔多个模块,使用 all 指定所有线性模块。
647
-
648
- apollo_update_interval
649
-
650
- Literal[“svd”, “random”]
651
-
652
- APOLLO 低秩投影算法类型(svd 或 random)。
653
-
654
- Literal[“std”, “right”, “left”]
655
-
656
- Literal[“channel”, “tensor”]
657
-
658
- APOLLO 缩放类型(channel 或 tensor)。
659
-
660
- BAdam 的使用模式,可选值为 layer 或 ratio。
661
-
662
- layer-wise BAdam 的起始块索引。
663
-
664
- layer-wise BAdam 中块更新策略,可选值有: ascending, descending, random, fixed。
665
-
666
- badam_switch_interval
667
-
668
- layer-wise BAdam 中块更新步数间隔。使用 -1 禁用块更新。
669
-
670
- ratio-wise BAdam 中的更新比例。
671
-
672
- BAdam 优化器的掩码模式,可选值为 adjacent 或 scatter。
673
-
674
- BAdam 优化器的详细输出级别,0 表示无输出,1 表示输出块前缀,2 表示输出可训练参数。
675
-
676
- 应用 GaLore 的模块名称。使用逗号分隔多个模块,使用 all 指定所有线性模块。
677
-
678
- galore_update_interval
679
-
680
- GaLore 投影的类型,可选值有: std, reverse_std, right, left, full。
681
-
682
- 用于训练的数据集名称。使用逗号分隔多个数据集。
683
-
684
- 用于评估的数据集名称。使用逗号分隔多个数据集。
685
-
686
- 是否在每个评估数据集上分开计算loss,默认concate后为整体计算。
687
-
688
- Union[str, Dict[str, Any]]
689
-
690
- 存储数据集的文件夹路径,可以是字符串或字典。 类型:str 或 dict(需符合 dataset_info.json 的格式)
691
-
692
- 当为字符串时,表示数据集目录的路径,例如:data 。
693
-
694
- 当为字典时,将覆盖默认从本地 dataset_info.json 加载的行为。应具有以下结构:
695
-
696
- 存储图像、视频或音频的文件夹路径。如果未指定,默认为 dataset_dir。
697
-
698
- data_shared_file_system
699
-
700
- 多机多卡时,不同机器存放数据集的路径是否是共享文件系统。数据集处理在该值为true时只在第一个node发生,为false时在每个node都处理一次。
701
-
702
- 输入的最大 token 数,超过该长度会被截断。
703
-
704
- 启用 streaming 时用于随机选择样本的 buffer 大小。
705
-
706
- Literal[“concat”, “interleave_under”, “interleave_over”]
707
-
708
- 数据集混合策略,支持 concat、 interleave_under、 interleave_over。
709
-
710
- 使用 interleave 策略时,指定从多个数据集中采样的概率。多个数据集的概率用逗号分隔。
711
-
712
- preprocessing_batch_size
713
-
714
- preprocessing_num_workers
715
-
716
- 每个数据集的最大样本数:设置后,每个数据集的样本数将被截断至指定的 max_samples。
717
-
718
- ignore_pad_token_for_loss
719
-
720
- 计算 loss 时是否忽略 pad token。
721
-
722
- 验证集相对所使用的训练数据集的大小。取值在 [0,1) 之间。启用 streaming 时 val_size 应是整数。
723
-
724
- 是否启用 sequences packing。预训练时默认启用。
725
-
726
- 是否启用不使用 cross-attention 的 sequences packing。
727
-
728
- Tokenized datasets的保存或加载路径。如果路径存在,会加载已有的 tokenized datasets;如果路径不存在,则会在分词后将 tokenized datasets 保存在此路径中。
729
-
730
- 模型路径(本地路径或 Huggingface/ModelScope 路径)。
731
-
732
- 适配器路径(本地路径或 Huggingface/ModelScope 路径)。使用逗号分隔多个适配器路径。
733
-
734
- 保存从 Hugging Face 或 ModelScope 下载的模型的本地路径。
735
-
736
- 是否使用 fast_tokenizer 。
737
-
738
- 是否在分词时将 special token 分割。
739
-
740
- 要添加到 tokenizer 中的 special token。多个 special token 用逗号分隔。
741
-
742
- Optional[Literal[“linear”, “dynamic”, “yarn”, “llama3”]]
743
-
744
- RoPE Embedding 的缩放策略,支持 linear、dynamic、yarn 或 llama3。
745
-
746
- Literal[“auto”, “disabled”, “sdpa”, “fa2”]
747
-
748
- 是否启用 FlashAttention 来加速训练和推理。可选值为 auto, disabled, sdpa, fa2。
749
-
750
- 是否启用 Shift Short Attention (S^2-Attn)。
751
-
752
- Optional[Literal[“convert”, “load”]]
753
-
754
- 需要将模型转换为 mixture_of_depths(MoD)模型时指定: convert 需要加载 mixture_of_depths(MoD)模型时指定: load。
755
-
756
- 是否使用 unsloth 优化 LoRA 微调。
757
-
758
- MoE 架构中 aux_loss 系数。数值越大,各个专家负载越均衡。
759
-
760
- disable_gradient_checkpointing
761
-
762
- 是否将 layernorm 层权重精度提高至 fp32。
763
-
764
- 是否将 lm_head 输出精度提高至 fp32。
765
-
766
- Literal[“huggingface”, “vllm”]
767
-
768
- 推理时使用的后端引擎,支持 huggingface 或 vllm。
769
-
770
- Literal[“auto”, “float16”, “bfloat16”, “float32”]
771
-
772
- 推理时使用的模型权重和激活值的数据类型。支持 auto, float16, bfloat16, float32。
773
-
774
- 用于登录 HuggingFace 的验证 token。
775
-
776
- 用于登录 ModelScope Hub 的验证 token。
777
-
778
- 用于登录 Modelers Hub 的验证 token。
779
-
780
- 是否信任来自 Hub 上数据集/模型的代码执行。
781
-
782
- Optional[torch.dtype]
783
-
784
- 用于计算模型输出的数据类型,无需手动指定。
785
-
786
- Optional[Union[str, Dict[str, Any]]]
787
-
788
- 是否禁用 vLLM 中的 CUDA graph。
789
-
790
- Optional[Union[dict, str]]
791
-
792
- vLLM引擎初始化配置。以字典或JSON字符串输入。
793
-
794
- Literal[“bitsandbytes”, “hqq”, “eetq”]
795
-
796
- 指定用于量化的算法,支持 “bitsandbytes”, “hqq” 和 “eetq”。
797
-
798
- 指定在量化过程中使用的位数,通常是4位、8位等。
799
-
800
- Literal[“fp4”, “nf4”]
801
-
802
- 量化时使用的数据类型,支持 “fp4” 和 “nf4”。
803
-
804
- 是否在量化过程中使用 double quantization,通常用于 “bitsandbytes” int4 量化训练。
805
-
806
- quantization_device_map
807
-
808
- Optional[Literal[“auto”]]
809
-
810
- 用于推理 4-bit 量化模型的设备映射。需要 “bitsandbytes >= 0.43.0”。
811
-
812
- Literal[“cpu”, “auto”]
813
-
814
- 导出模型时使用的设备,auto 可自动加速导出。
815
-
816
- export_quantization_bit
817
-
818
- export_quantization_dataset
819
-
820
- 用于量化导出模型的数据集路径或数据集名称。
821
-
822
- export_quantization_nsamples
823
-
824
- export_quantization_maxlen
825
-
826
- True: .bin 格式保存。 False: .safetensors 格式保存。
827
-
828
- 模型上传至 Huggingface 的仓库名称。
829
-
830
- 评估任务的名称,可选项有 mmlu_test, ceval_validation, cmmlu_test
831
-
832
- 保存评估结果的路径。 如果该路径已经存在则会抛出错误。
833
-
834
- 评估数据集的下载模式,如果数据集已经存在则重复使用,否则则下载。
835
-
836
- DownloadMode.REUSE_DATASET_IF_EXISTS
837
-
838
- 是否使用采样策略生成文本。如果设置为 False,将使用 greedy decoding。
839
-
840
- 用于调整生成文本的随机性。temperature 越高,生成的文本越随机;temperature 越低,生成的文本越确定。
841
-
842
- 用于控制生成时候选 token 集合大小的参数。例如:top_p = 0.7 意味着模型会先选择概率最高的若干个 token 直到其累积概率之和大于 0.7,然后在这些 token 组成的集合中进行采样。
843
-
844
- 用于控制生成时候选 token 集合大小的参数。例如:top_k = 50 意味着模型会在概率最高的50个 token 组成的集合中进行采样。
845
-
846
- 用于 beam_search 的束宽度。值为 1 表示不使用 beam_search。
847
-
848
- 文本最大长度(包括输入文本和生成文本的长度)。
849
-
850
- 生成文本的最大长度。设置 max_new_tokens 会覆盖 max_length。
851
-
852
- 对生成重复 token 的惩罚系数。对于已经生成过的 token 生成概率乘以 1/repetition_penalty。值小于 1.0 会提高重复 token 的生成概率,大于 1.0 则会降低重复 token 的生成概率。
853
-
854
- 在使用 beam_search 时对生成文本长度的惩罚系数。length_penalty > 0 鼓励模型生成更长的序列,length_penalty < 0 会鼓励模型生成更短的序列。
855
-
856
- 默认的 system_message,例如: “You are a helpful assistant.”
857
-
858
- Literal[“cloud”, “local”]
859
-
860
- 训练结果将保存在 <ray_storage_path>/ray_run_name 路径下。
861
-
862
- 每个工作进程分配的资源。默认使用 1 GPU。
863
-
864
- Literal[“SPREAD”, “PACK”, “STRICT_SPREAD”, “STRICT_PACK”]
865
-
866
- Ray 训练的资源调度策略。可选值包括 SPREAD、PACK、STRICT_SPREAD 和 STRICT_PACK。
867
-
868
- DISABLE_VERSION_CHECK
869
-
870
- LLAMAFACTORY_VERBOSITY
871
-
872
- 设置 LLaMA-Factory 的日志级别(“DEBUG”,”INFO”,”WARN”)
873
-
874
- 优先使用 ModelScope 下载模型/数据集或使用缓存路径中的模型/数据集
875
-
876
- 优先使用 Openmind 下载模型/数据集或使用缓存路径中的模型/数据集
877
-
878
- 是否使用 Ray 进行分布式执行或任务管理。
879
-
880
- 是否表示启用特定的 PyTorch 优化。
881
-
882
- ASCEND_RT_VISIBLE_DEVICES
883
-
884
- Torchrun部署中主节点 (master node) 的网络地址
885
-
886
- Torchrun部署中主节点用于通信的端口号
887
-
888
- 当前节点在所有节点中的 rank,通常从 0 到 NNODES-1。
889
-
890
- 设置 Gradio 服务器 IP 地址(例如 0.0.0.0)
891
-
892
- 启用 Gradio 服务器的 IPv6 支持
893
-
894
- 支持使用 lmf 表示 llamafactory-cli
895
-
896
- ---
897
-
898
- ## Adapters¶
899
-
900
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/adapters.html
901
-
902
- **Contents:**
903
- - Adapters¶
904
- - Full Parameter Fine-tuning¶
905
- - Freeze¶
906
- - LoRA¶
907
- - LoRA+¶
908
- - rsLoRA¶
909
- - DoRA¶
910
- - PiSSA¶
911
- - Galore¶
912
- - BAdam¶
913
-
914
- LLaMA-Factory 支持多种调优算法,包括: Full Parameter Fine-tuning 、 Freeze 、 LoRA 、 Galore 、 BAdam 。
915
-
916
- 全参微调指的是在训练过程中对于预训练模型的所有权重都进行更新,但其对显存的要求是巨大的。
917
-
918
- 如果您需要进行全参微调,请将 finetuning_type 设置为 full 。 下面是一个例子:
919
-
920
- Freeze(冻结微调)指的是在训练过程中只对模型的小部分权重进行更新,这样可以降低对显存的要求。
921
-
922
- 如果您需要进行冻结微调,请将 finetuning_type 设置为 freeze 并且设置相关参数, 例如冻结的层数 freeze_trainable_layers 、可训练的模块名称 freeze_trainable_modules 等。
923
-
924
- freeze_trainable_layers
925
-
926
- 可训练层的数量。正数表示最后 n 层被设置为可训练的,负数表示前 n 层被设置为可训练的。默认值为 2
927
-
928
- freeze_trainable_modules
929
-
930
- 可训练层的名称。使用 all 来指定所有模块。默认值为 all
931
-
932
- freeze_extra_modules[非必须]
933
-
934
- 除了隐藏层外可以被训练的模块名称,被指定的模块将会被设置为可训练的。使用逗号分隔多个模块。默认值为 None
935
-
936
- 如果您需要进行 LoRA 微调,请将 finetuning_type 设置为 lora 并且设置相关参数。 下面是一个例子:
937
-
938
- additional_target[非必须]
939
-
940
- 除 LoRA 层之外设置为可训练并保存在最终检查点中的模块名称。使用逗号分隔多个模块。默认值为 None
941
-
942
- LoRA 缩放系数。一般情况下为 lora_rank * 2, 默认值为 None
943
-
944
- LoRA 微调中的 dropout 率。默认值为 0
945
-
946
- LoRA 微调的本征维数 r, r 越大可训练的参数越多。默认值为 8
947
-
948
- 应用 LoRA 方法的模块名称。使用逗号分隔多个模块,使用 all 指定所有模块。默认值为 all
949
-
950
- loraplus_lr_ratio[非必须]
951
-
952
- LoRA+ 学习率比例(λ = ηB/ηA)。 ηA, ηB 分别是 adapter matrices A 与 B 的学习率。LoRA+ 的理想取值与所选择的模型和任务有关。默认值为 None
953
-
954
- loraplus_lr_embedding[非必须]
955
-
956
- LoRA+ 嵌入层的学习率, 默认值为 1e-6
957
-
958
- 是否使用秩稳定 LoRA(Rank-Stabilized LoRA),默认值为 False。
959
-
960
- 是否使用权重分解 LoRA(Weight-Decomposed LoRA),默认值为 False
961
-
962
- 是否初始化 PiSSA 适配器,默认值为 False
963
-
964
- PiSSA 中 FSVD 执行的迭代步数。使用 -1 将其禁用,默认值为 16
965
-
966
- 是否将 PiSSA 适配器转换为正常的 LoRA 适配器,默认值为 False
967
-
968
- 是否创建一个具有随机初始化权重的新适配器,默认值为 False
969
-
970
- 在LoRA中,适配器矩阵 A 和 B 的学习率相同。您可以通过设置 loraplus_lr_ratio 来调整学习率比例。在 LoRA+ 中,适配器矩阵 A 的学习率 ηA 即为优化器学习率。适配器矩阵 B 的学习率 ηB 为 λ * ηA。 其中 λ 为 loraplus_lr_ratio 的值。
971
-
972
- LoRA 通过添加低秩适配器进行微调,然而 lora_rank 的增大往往会导致梯度塌陷,使得训练变得不稳定。这使得在使用较大的 lora_rank 进行 LoRA 微调时较难取得令人满意的效果。rsLoRA(Rank-Stabilized LoRA) 通过修改缩放因子使得模型训练更加稳定。 使用 rsLoRA 时, 您只需要将 use_rslora 设置为 True 并设置所需的 lora_rank。
973
-
974
- DoRA (Weight-Decomposed Low-Rank Adaptation)提出尽管 LoRA 大幅降低了推理成本,但这种方式取得的性能与全量微调之间仍有差距。
975
-
976
- DoRA 将权重矩阵分解为大小与单位方向矩阵的乘积,并进一步微调二者(对方向矩阵则进一步使用 LoRA 分解),从而实现 LoRA 与 Full Fine-tuning 之间的平衡。
977
-
978
- 如果您需要使用 DoRA,请将 use_dora 设置为 True 。
979
-
980
- 在 LoRA 中,适配器矩阵 A 由 kaiming_uniform 初始化,而适配器矩阵 B 则全初始化为0。这导致一开始的输入并不会改变模型输出并且使得梯度较小,收敛较慢。 PiSSA 通过奇异值分解直接分解原权重矩阵进行初始化,其优势在于它可以更快更好地收敛。
981
-
982
- 如果您需要使用 PiSSA,请将 pissa_init 设置为 True 。
983
-
984
- 当您需要在训练中使用 GaLore(Gradient Low-Rank Projection)算法时,可以通过设置 GaloreArguments 中的参数进行配置。
985
-
986
- 不要将 LoRA 和 GaLore/BAdam 一起使用。
987
-
988
- ``galore_layerwise``为 ``true``时请不要设置 ``gradient_accumulation``参数。
989
-
990
- 是否使用 GaLore 算法,默认值为 False。
991
-
992
- 应用 GaLore 的模块名称。使用逗号分隔多个模块,使用 all 指定所有线性模块。默认值为 all。
993
-
994
- galore_update_interval
995
-
996
- 更新 GaLore 投影的步数间隔,默认值为 200。
997
-
998
- GaLore 的缩放系数,默认值为 0.25。
999
-
1000
- GaLore 投影的类型,可选值有: std , reverse_std, right, left, full。默认值为 std。
1001
-
1002
- 是否启用逐层更新以进一步节省内存,默认值为 False。
1003
-
1004
- BAdam 是一种内存高效的全参优化方法,您通过配置 BAdamArgument 中的参数可以对其进行详细设置。 下面是一个例子:
1005
-
1006
- 不要将 LoRA 和 GaLore/BAdam 一起使用。
1007
-
1008
- 使用 BAdam 时请设置 finetuning_type 为 full 且 pure_bf16 为 True 。
1009
-
1010
- badam_mode = layer 时仅支持使用 DeepSpeed ZeRO3 进行 单卡 或 多卡 训练。
1011
-
1012
- badam_mode = ratio 时仅支持 单卡 训练。
1013
-
1014
- 是否使用 BAdam 优化器,默认值为 False。
1015
-
1016
- BAdam 的使用模式,可选值为 layer 或 ratio,默认值为 layer。
1017
-
1018
- layer-wise BAdam 的起始块索引,默认值为 None。
1019
-
1020
- layer-wise BAdam 中块更新策略,可选值有: ascending, descending, random, fixed。默认值为 ascending。
1021
-
1022
- badam_switch_interval
1023
-
1024
- layer-wise BAdam 中块更新步数间隔。使用 -1 禁用块更新,默认值为 50。
1025
-
1026
- ratio-wise BAdam 中的更新比例,默认值为 0.05。
1027
-
1028
- BAdam 优化器的掩码模式,可选值为 adjacent 或 scatter,默认值为 adjacent。
1029
-
1030
- BAdam 优化器的详细输出级别,0 表示无输出,1 表示输出块前缀,2 表示输出可训练参数。默认值为 0。
1031
-
1032
- ---
1033
-
1034
- ## Extras¶
1035
-
1036
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/extras.html
1037
-
1038
- **Contents:**
1039
- - Extras¶
1040
- - LLaMA Pro¶
1041
-
1042
- 为了解决大语言模型的遗忘问题, LLaMA Pro 通过在原有模型上增加新模块以适应新的任务,使其在多个新任务上的表现均优于原始模型。 LLaMA-Factory 支持 LLaMA Pro 的使用。 您可以使用运行 expand.sh 将 Meta-Llama-3-8B-Instruct 扩展为 llama3-8b-instruct-pro。
1043
-
1044
- 对于 LLaMA Pro 模型进行训练时,您需要指定 use_llama_pro 为 true。
1045
-
1046
- ---
1047
-
1048
- ## Fine-tuning Best Practices¶
1049
-
1050
- **URL:** https://llamafactory.readthedocs.io/en/latest/advanced/best_practice/index.html
1051
-
1052
- **Contents:**
1053
- - Fine-tuning Best Practices¶
1054
-
1055
- ---