Buckets:
| { | |
| "schema": "kvnloo/shadow-slm MANIFEST v1", | |
| "generated": "2026-11-05", | |
| "purpose": "Small models used as fast decision/routing/guard models beside a large agent model. Weights are NOT mirrored here; fetch them from the Hub at the pinned revision.", | |
| "revisions": "Hub commit SHAs read from the Hub API on the generation date; every listed model also has a local copy at that revision or the Hub head.", | |
| "adapters": [], | |
| "adapters_note": "No adapters are published yet: the decision adapters trained so far used private agent traces and stay private. Public-data adapters (e.g. BFCL tool-choice) will be added here with their training config and input datasets.", | |
| "models": [ | |
| { | |
| "repo": "Qwen/Qwen3-0.6B", | |
| "revision": "c1899de289a04d12100db370d81485cdf75e47ca", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "text-generation", | |
| "base_model": "Qwen/Qwen3-0.6B-Base", | |
| "gated": false, | |
| "role": "shadow tool-choice scorer (yes/no log-odds per candidate tool, served as Q8 GGUF via llama.cpp)", | |
| "used_by": "z0intelligence shadow network; SLM optimisation; edge probes", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "unsloth/Qwen3-0.6B", | |
| "revision": "db4e90b8359f816f09bf5c944797ce116f26da3c", | |
| "license": "apache-2.0", | |
| "license_note": "no licence on card; inherits base Qwen/Qwen3-0.6B", | |
| "pipeline": null, | |
| "base_model": "Qwen/Qwen3-0.6B", | |
| "gated": false, | |
| "role": "fine-tuning base (unsloth) for the 0.6B decision model", | |
| "used_by": "shadow network / jev decision model training", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "unsloth/Qwen3-0.6B-unsloth-bnb-4bit", | |
| "revision": "e86f5289a2ca40ab7114071959dc2079cf97cb3c", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "text-generation", | |
| "base_model": "Qwen/Qwen3-0.6B", | |
| "gated": false, | |
| "role": "4-bit QLoRA training base", | |
| "used_by": "shadow network / jev decision model training", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "Qwen/Qwen3-1.7B", | |
| "revision": "70d244cc86ccca08cf5af4e1e306ecf908b1ad5e", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "text-generation", | |
| "base_model": "Qwen/Qwen3-1.7B-Base", | |
| "gated": false, | |
| "role": "larger decision/edge model; reference arm for edge probes", | |
| "used_by": "edge probes; AutoKernel mobile LLM tracks", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "unsloth/Qwen3-1.7B", | |
| "revision": "6262b50d6c1f8ee5e4ac750d710c33603bfc2a0c", | |
| "license": "apache-2.0", | |
| "license_note": "no licence on card; inherits base Qwen/Qwen3-1.7B", | |
| "pipeline": null, | |
| "base_model": "Qwen/Qwen3-1.7B", | |
| "gated": false, | |
| "role": "fine-tuning base (unsloth) for the 1.7B decision model", | |
| "used_by": "shadow network / jev decision model training", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "unsloth/Qwen3-1.7B-unsloth-bnb-4bit", | |
| "revision": "a972f4304f86a0a705d35b378d7a7675d58063a7", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "text-generation", | |
| "base_model": "Qwen/Qwen3-1.7B", | |
| "gated": false, | |
| "role": "4-bit QLoRA training base", | |
| "used_by": "shadow network / jev decision model training", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "Qwen/Qwen3.5-0.8B", | |
| "revision": "2fc06364715b967f1860aea9cf38778875588b17", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "image-text-to-text", | |
| "base_model": "Qwen/Qwen3.5-0.8B-Base", | |
| "gated": false, | |
| "role": "candidate intake/intent classifier", | |
| "used_by": "intent analyzer; CI sizing", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "Qwen/Qwen3.5-4B", | |
| "revision": "851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "image-text-to-text", | |
| "base_model": "Qwen/Qwen3.5-4B-Base", | |
| "gated": false, | |
| "role": "upper-bound local decision arm in shadow evals", | |
| "used_by": "shadow evals; tool-rollout audit", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "divagr1925/Qwen3.5-4B-Qwenputer-Filesystem", | |
| "revision": "9112e41f03516ca4d0125f5bfa975008487a8458", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "text-generation", | |
| "base_model": "Qwen/Qwen3.5-4B", | |
| "gated": false, | |
| "role": "filesystem/computer-use specialist candidate", | |
| "used_by": "computer-use lanes (evaluation only)", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "Qwen/Qwen2.5-0.5B", | |
| "revision": "060db6499f32faf8b98477b0a26969ef7d8b9987", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "text-generation", | |
| "base_model": null, | |
| "gated": false, | |
| "role": "small baseline for SLM optimisation and kernel experiments", | |
| "used_by": "SLM optimisation; GEPA/JEPA baselines", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "Qwen/Qwen2.5-1.5B-Instruct", | |
| "revision": "989aa7980e4cf806f80c7fef2b1adb7bc71aa306", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "text-generation", | |
| "base_model": "Qwen/Qwen2.5-1.5B", | |
| "gated": false, | |
| "role": "instruct baseline for promotion/CI sizing evals", | |
| "used_by": "OSS promotion; CI sizing", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "Mapika/decider-2b", | |
| "revision": "7614a6c7ea5a047aa7d7bcfb1a0015469381c3e3", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "text-classification", | |
| "base_model": "Qwen/Qwen3.5-2B-Base", | |
| "gated": false, | |
| "role": "one-pass typed-probability decider (Qwen3.5-2B base)", | |
| "used_by": "tool-rollout audit; jev decision candidates", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "unsloth/functiongemma-270m-it", | |
| "revision": "79eaaa84125b897f5a0ad6638362b864fec813ad", | |
| "license": "gemma", | |
| "license_note": null, | |
| "pipeline": "text-generation", | |
| "base_model": "google/functiongemma-270m-it", | |
| "gated": false, | |
| "role": "function-calling SLM; tool-choice classifier/embedder candidate", | |
| "used_by": "tool-rollout audit; OSS factory; weight-space SLM benchmarks (BFCL v3 tool choice)", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "ggml-org/functiongemma-270m-it-GGUF", | |
| "revision": "2566ce14aedfc14fdd0de955ba67346425e67126", | |
| "license": "gemma", | |
| "license_note": null, | |
| "pipeline": "text-generation", | |
| "base_model": "google/functiongemma-270m-it", | |
| "gated": false, | |
| "role": "GGUF build of functiongemma for llama.cpp serving", | |
| "used_by": "tool-rollout audit", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "google/gemma-3-270m-it", | |
| "revision": "ac82b4e820549b854eebf28ce6dedaf9fdfa17b3", | |
| "license": "gemma", | |
| "license_note": null, | |
| "pipeline": "text-generation", | |
| "base_model": "google/gemma-3-270m", | |
| "gated": "manual", | |
| "role": "tiny generalist baseline", | |
| "used_by": "CI sizing", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "unsloth/gemma-3-270m-it", | |
| "revision": "23cf460f6bb16954176b3ddcc8d4f250501458a9", | |
| "license": "gemma", | |
| "license_note": null, | |
| "pipeline": "text-generation", | |
| "base_model": "google/gemma-3-270m-it", | |
| "gated": false, | |
| "role": "fine-tuning base for tiny generalist experiments", | |
| "used_by": "tiny SLM gym", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "pipizhao/SkillRouter-Embedding-0.6B", | |
| "revision": "c03c9bcee9fce92ab0262bb6dcf54d174a8ba558", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "feature-extraction", | |
| "base_model": "Qwen/Qwen3-Embedding-0.6B", | |
| "gated": false, | |
| "role": "skill retrieval embedder (arm vs BM25 skill ranking)", | |
| "used_by": "skill routing / token-thrift tools", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "pipizhao/SkillRouter-Reranker-0.6B", | |
| "revision": "78986e1142d12857cfd85b8005e62902cd42d858", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "text-ranking", | |
| "base_model": "Qwen/Qwen3-Reranker-0.6B", | |
| "gated": false, | |
| "role": "skill retrieval reranker", | |
| "used_by": "skill routing / token-thrift tools", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "sentence-transformers/all-MiniLM-L6-v2", | |
| "revision": "1110a243fdf4706b3f48f1d95db1a4f5529b4d41", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "sentence-similarity", | |
| "base_model": "nreimers/MiniLM-L6-H384-uncased", | |
| "gated": false, | |
| "role": "general text embedder for retrieval/offload", | |
| "used_by": "shadow offload; CI sizing", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "SupersonicLabs/Julia-1", | |
| "revision": "a85b127321d580d65176c89ced8273f305745d85", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "text-classification", | |
| "base_model": "jhu-clsp/mmBERT-small", | |
| "gated": false, | |
| "role": "small text-classification router candidate (mmBERT-small base)", | |
| "used_by": "tool-rollout audit; Hermes routing dogfood", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "convaiinnovations/laya", | |
| "revision": "7b928d828b7b0e022f929d9bd2e44165aa270148", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "text-classification", | |
| "base_model": null, | |
| "gated": false, | |
| "role": "~421M calibrated classifier candidate, CPU-friendly", | |
| "used_by": "tool-rollout audit; routing evals", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "fastino/gliner2-privacy-filter-PII-multi", | |
| "revision": "1cb4166094dc58fa8d836429f060d6c95f62b495", | |
| "license": "apache-2.0", | |
| "license_note": null, | |
| "pipeline": "token-classification", | |
| "base_model": null, | |
| "gated": false, | |
| "role": "PII detector used as the pre-publication privacy gate", | |
| "used_by": "publication gates (this bucket set)", | |
| "local_copy": true | |
| }, | |
| { | |
| "repo": "LiquidAI/LFM3-Audio-2B-GGUF", | |
| "revision": "39670cb5c42e9c4bfe046519ba7ffbe47f9552ad", | |
| "license": "lfm1.0", | |
| "license_note": null, | |
| "pipeline": "any-to-any", | |
| "base_model": "LiquidAI/LFM3-Audio-2B", | |
| "gated": false, | |
| "role": "on-device speech-to-speech model under AutoKernel mobile tuning", | |
| "used_by": "z0live voice on Android (AutoKernel mobile s2s/lfm3 track)", | |
| "local_copy": true | |
| } | |
| ] | |
| } |
Xet Storage Details
- Size:
- 9.84 kB
- Xet hash:
- 1022f09631dc8015d07c18c3e2862f56ddc789b1c7f35f93643d7ef500a5871e
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.