{"model_id": "techkiyan/kiyannet-100M", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes own base pretraining followed by two SFT stages; custom constructor and 32000-vocab/12-layer/768-width configuration support own scratch backbone inference. Selected weights explicitly come from own structured_mvp1_100k_mb8/instruction_best.pt.", "provenance_sources": [{"url": "https://huggingface.co/techkiyan/kiyannet-100M/raw/7044fe6ce8054da60bb8a2723a6d24a3f58e1405/README.md", "accessed_at": "2026-10-30T12:26:35.049618+00:00", "quote": ", (2) instruction supervised fine-tuning, and (3) financial-news structured-extraction fine-tuning. Project artifacts confirm a 20,000-example structured stage followed by 100,000 additional distinct training examples. The final release weights come from str", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Own base pretraining plus instruction and financial structured-extraction SFT (20K then100K additional distinct examples).", "provenance_confidence": "medium", "architecture_lineage": "Own PyTorch implementation of standard grouped-query causal Transformer with RoPE/RMSNorm/SwiGLU and tied embeddings; no external pretrained checkpoint named.", "training_hardware_estimate": {"device": "NVIDIA RTX2060", "confidence": "low", "direct_or_inferred": "inferred_available_device", "basis": "Same project explicitly used RTX2060 for measured inference; training on it is not stated."}, "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "siliconperception/vla_v502", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/siliconperception/vla_v502/raw/c04938571a75ffe1a754602d59386e432b9a5669/README.md", "accessed_at": "2026-10-30T12:26:35.066146+00:00", "quote": "NUL> token (\"say nothing this tick\") | Training - From scratch: 220,000 steps × 250 rows, one token per row per step (55 M tokens), plain SGD. - Learning rate: a half-cosine rise from 4e-4 to 0.02 over the first 20,000 steps, then held. ", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "edeneldith/WiggleGPT", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretrained_from_scratch", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'pretrained_from_scratch', 'basis': 'Model card explicitly states pretraining from scratch on OpenWebText and documents600000 updates. Existing GPT-2 tokenizer and nanoGPT-derived training/code are reused; pretrained transformer weights are not claimed as initialization.', 'pretraining_dataset': 'OpenWebText', 'updates': 600000, 'calendar_start': '2025-10-23', 'calendar_end': '2025-11-12', 'from_scratch_evidence': 'documented_in_public_model_card'}", "provenance_sources": [{"url": "https://huggingface.co/edeneldith/WiggleGPT/raw/1ba13498a367e214496b72f202745c1846adda89/README.md", "basis": "Pinned public model card: introduction; Architecture; Training Details; final independent-lab statement."}, {"url": "https://garden-backend-three.vercel.app/", "basis": "HF-linked personal website author self-description.", "accessed_at": "2026-10-30"}, {"url": "https://github.com/Eden-Eldith/WiggleGPT", "basis": "Model card links public implementation and dated development log; card is the pinned source used for recorded training facts.", "accessed_at": "2026-10-30"}, {"url": "https://garden-backend-three.vercel.app/finalized-work/eden-s-autobiography-3/", "basis": "Author educational background; not inferred from self-taught label.", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"class": "consumer_gpu", "single_gpu_at_a_time": true, "gpus": ["NVIDIA RTX 3070", "NVIDIA RTX 5060 Ti 16GB"], "steps_by_gpu": {"RTX 3070": 240000, "RTX 5060 Ti 16GB": 360000}, "gpu_hours_estimate": 310, "basis": "Card's dated development-log reconstruction: approximately140h plus170h. Hardware was changed mid-run and resumed from240000-step checkpoint; no multi-GPU training asserted."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "reaperdoesntknow/SAGI", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original52.72M swarm/transformer architecture with own routing/memory/controller code and explicit TinyStories-subset CPU training details support own scratch-origin model.", "provenance_sources": [{"url": "https://huggingface.co/reaperdoesntknow/SAGI/raw/f342678c585a578f217a73325232feb5e78cba2f/README.md", "accessed_at": "2026-10-30T12:26:35.057631+00:00"}, {"url": "https://huggingface.co/reaperdoesntknow/SAGI/raw/f342678c585a578f217a73325232feb5e78cba2f/README.md", "read_scope": "Saved card model/task/training and source attribution", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "toonist/AnuLM-Base-v2-400M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/toonist/AnuLM-Base-v2-400M/raw/16e72e0a86d1ccb1ef6fa6b90fbed5b535657d6d/README.md", "accessed_at": "2026-10-30T12:26:35.053628+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "arthurblg1802/AtoMinima", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Frozen from-scratch tag plus explicit training description, tokens_trained config and creator linked model-specific training announcement support own from-scratch pretraining as publisher claim.", "provenance_sources": [{"url": "https://huggingface.co/arthurblg1802/AtoMinima/blob/main/README.md"}, {"url": "https://huggingface.co/arthurblg1802/AtoMinima/raw/6437258129bf6469fc3c946769ca99fa66bfd108/README.md", "accessed_at": "2026-10-30T12:26:35.061146+00:00"}, {"url": "https://huggingface.co/arthurblg1802/AtoMinima/raw/6437258129bf6469fc3c946769ca99fa66bfd108/README.md", "access_date": "2026-10-31", "scope": "complete frozen card read", "finding": "Card tags from-scratch and explicitly reports training on370Mcharacters; says420trainableparameters and provides PyTorch implementation."}, {"url": "https://huggingface.co/arthurblg1802/AtoMinima/raw/6437258129bf6469fc3c946769ca99fa66bfd108/atominima.py", "access_date": "2026-10-31", "scope": "complete4389-byte implementation read", "finding": "Custom small character-level nn.Module implementation with one pre-norm attention block, tied embeddings, fixed sinusoidal positions and GELU MLP; inference loaders load this repository own saved weights."}, {"url": "https://huggingface.co/arthurblg1802/AtoMinima/raw/6437258129bf6469fc3c946769ca99fa66bfd108/config.json", "access_date": "2026-10-31", "scope": "complete426-byte frozen config read", "finding": "420parameters, d_model4, one layer/head,42vocabulary; tokens_trained370000000."}, {"url": "https://huggingface.co/spaces/AxiomicLabs/Open_SLM_Leaderboard/discussions/276", "access_date": "2026-10-31", "event_date": "2026-10-28", "scope": "all14public comments read from full JSON payload", "finding": "Creator describes a420parameter basic decoder-only transformer trained on320Mcharacters of FineWeb; figure differs from370Min frozen card/config, and both observations retained."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Specific run device/provider, duration, physical compute country and builder training place unstated.; Card/config370Mcharacters and announcement320Mcharacters retained separately without selecting an invented reconciled count.", "upstream_checkpoint": "No external pretrained checkpoint identified; inference from_pretrained method loads this model own published checkpoint and is not evidence of reuse during training.", "creator_contribution": "Model-specific420parameter implementation and reported FineWeb character pretraining experiment.", "upstream_architecture_or_code": [{"name": "Standard decoder-only pre-norm Transformer / character GPT recipe", "implementation": "model-specific pure PyTorch AtomInima class", "source_url": "https://huggingface.co/arthurblg1802/AtoMinima/raw/6437258129bf6469fc3c946769ca99fa66bfd108/atominima.py"}], "training_characters_publisher_reported": 370000000, "other_publisher_training_count": {"characters": 320000000, "source_url": "https://huggingface.co/spaces/AxiomicLabs/Open_SLM_Leaderboard/discussions/276", "date": "2026-10-28"}, "provenance_evidence_detail": "publisher_statement_and_model_implementation", "evidence_updated_at": "2026-10-31"}
{"model_id": "wyf-yfw/MiniSeek-M0.1-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original Dense28.45M/MoE150.22M architecture family and authored training repository document base pretraining, released checkpoints, steps and losses. Best-supported reading is own scratch-origin lineage; tokenizer/data reference MiniMind rather than an imported pretrained model.", "provenance_sources": [{"url": "https://huggingface.co/wyf-yfw/MiniSeek-M0.1-base/raw/ba5d7010b84513dfb3f0f8230933aa98c140d9c5/README.md", "accessed_at": "2026-10-30T12:26:35.652064+00:00"}, {"url": "https://huggingface.co/wyf-yfw/MiniSeek-M0.1-base", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Kanz4kii/Next-Concept-Modeling", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Own bespoke hierarchical NCM model is created fresh by train.py (NCMModel(args), optional own-state resume). sft_train.py loads its own base checkpoint for Alpaca SFT. Frozen card reports 1B FineWeb-Edu pretraining tokens and 10,000 Alpaca dialogues; selected release is SFT.", "provenance_sources": [{"url": "https://huggingface.co/Kanz4kii/Next-Concept-Modeling/raw/2b64655ce8b0c3fb2a572375711e9347d792dfff/README.md"}, {"url": "https://github.com/Kanz4kii/Next-Concept-Modeling/blob/main/train.py", "scope": "Model creation and optional own-checkpoint resume paths read", "excerpt": "raw_model = NCMModel(args=args).to(device)", "access_date": "2026-11-01"}, {"url": "https://github.com/Kanz4kii/Next-Concept-Modeling/blob/main/sft_train.py", "scope": "Own base-checkpoint SFT loading path read", "excerpt": "model = NCMModel.from_safetensors(base_ckpt, device=device)", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/Kanz4kii/Next-Concept-Modeling/raw/2b64655ce8b0c3fb2a572375711e9347d792dfff/README.md", "scope": "Previously complete frozen card read reused", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "RTX 5050 8GB measured for training throughput and inference benchmarks; full-run device estimate medium", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "architecture_lineage": "Own hierarchical RVQ-4 ConceptPlanner and semi-autoregressive SurfaceDecoder implementation; GPT-2 tokenizer reused.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "hardware": "RTX 5050 8GB measured for training throughput and inference benchmarks; full-run device estimate medium", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "compute_sources": [{"url": "https://github.com/Kanz4kii/Next-Concept-Modeling/blob/main/train.py", "scope": "Model creation and optional own-checkpoint resume paths read", "excerpt": "raw_model = NCMModel(args=args).to(device)", "access_date": "2026-11-01"}, {"url": "https://github.com/Kanz4kii/Next-Concept-Modeling/blob/main/sft_train.py", "scope": "Own base-checkpoint SFT loading path read", "excerpt": "model = NCMModel.from_safetensors(base_ckpt, device=device)", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/Kanz4kii/Next-Concept-Modeling/raw/2b64655ce8b0c3fb2a572375711e9347d792dfff/README.md", "scope": "Previously complete frozen card read reused", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "KerzgrAI/Monody-Exp", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original matched Monody/Rondo base-model architectures trained side by side on88M FineWeb tokens repeated7times; released EMA weights after training support own scratch-origin lineage.", "provenance_sources": [{"url": "https://huggingface.co/KerzgrAI/Monody-Exp/raw/5c63f8f04f3e0790621dcdbfc108241ed0754aed/README.md", "accessed_at": "2026-10-30T12:26:35.656065+00:00"}, {"url": "https://huggingface.co/KerzgrAI/Monody-Exp/raw/5c63f8f04f3e0790621dcdbfc108241ed0754aed/README.md", "read_scope": "Saved card architecture, training and checkpoint description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "KerzgrAI/Rondo-Exp", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original matched Monody/Rondo base-model architectures trained side by side on88M FineWeb tokens repeated7times; released EMA weights after training support own scratch-origin lineage.", "provenance_sources": [{"url": "https://huggingface.co/KerzgrAI/Rondo-Exp/raw/02720c57348150665f52834ea0144d1a0bc0fa51/README.md", "accessed_at": "2026-10-30T12:26:35.700133+00:00"}, {"url": "https://huggingface.co/KerzgrAI/Rondo-Exp/raw/02720c57348150665f52834ea0144d1a0bc0fa51/README.md", "read_scope": "Saved card architecture, training and checkpoint description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "ShivankXD/stasis-1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/ShivankXD/stasis-1/raw/f99950814d1846b8b8aaa80056a7df5ca1f92baf/README.md", "accessed_at": "2026-10-30T12:26:35.675087+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "keytron46/Allan-LM-Quark", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/keytron46/Allan-LM-Quark/raw/33371900be1c2a00c259671c6f02cbd780da7b56/README.md", "accessed_at": "2026-10-30T12:26:35.679594+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sarathi-balakrishnan/tiny-sql-gpt", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card explicitly states trained from random weights with no pretrained weights,3,000 steps on generated SQL grammar data using a laptop CPU. Original model implementation/training is distinct from narrow synthetic task coverage.", "provenance_sources": [{"url": "https://huggingface.co/sarathi-balakrishnan/tiny-sql-gpt/raw/ded1965ef386f997450059d4145450547d667d59/README.md", "accessed_at": "2026-10-30T12:26:36.692910+00:00"}, {"url": "https://huggingface.co/sarathi-balakrishnan/tiny-sql-gpt/raw/ded1965ef386f997450059d4145450547d667d59/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "adriaflores/kimi-k3-300m-from-scratch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/adriaflores/kimi-k3-300m-from-scratch/raw/41b865fbb357b0454ba05b7fb738f4f2075b542b/README.md", "accessed_at": "2026-10-30T12:26:36.684844+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "wyf-yfw/MiniSeek-D0.1-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original Dense28.45M/MoE150.22M architecture family and authored training repository document base pretraining, released checkpoints, steps and losses. Best-supported reading is own scratch-origin lineage; tokenizer/data reference MiniMind rather than an imported pretrained model.", "provenance_sources": [{"url": "https://huggingface.co/wyf-yfw/MiniSeek-D0.1-base/raw/2b171677ea937c36bfcfaefc41024ddd580cbabc/README.md", "accessed_at": "2026-10-30T12:26:36.273208+00:00"}, {"url": "https://huggingface.co/wyf-yfw/MiniSeek-D0.1-base", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Chesi-AI-Models/Chesi-AI-stage1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/Chesi-AI-Models/Chesi-AI-stage1/raw/4a09e5abe63a3d2a8a7e65eb15160bc4ec7bb192/README.md", "accessed_at": "2026-10-30T12:26:36.675331+00:00", "quote": "s/CORPUSLEDGER.md> 9. 引用 Chesi AI (澈思), \"Stage-1: from-scratch character-level pretraining, verified\", 2026. https://huggingface.co/Chesi-AI-Models/Chesi-AI-stage1 10. 致谢与联系 一个人 + 一张显卡的项目。赞助与进展：<https://ko-fi.com/chesiai> Issues / PRs 欢迎：<https://git", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Neithanvarvas/CBT", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card documents original19M character-transformer pretraining and a matched positional-encoding twin, followed by several supervised fine-tuning rounds; the released chat/tool checkpoint is own-base post-training.", "provenance_sources": [{"url": "https://huggingface.co/Neithanvarvas/CBT/raw/8414bb48e81186268e271d6095f7085c6e314c66/README.md", "accessed_at": "2026-10-30T12:26:36.684844+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "Arko007/tiny-edu-114m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Arko007/tiny-edu-114m/raw/d186af37d44959f9349307f4d99b2863b58b9488/README.md", "accessed_at": "2026-10-30T12:26:36.704423+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ztor2/jerboa-sft", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "sft_of_own_base", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Own custom Jerboa221M architecture and1.23B-token FineWeb/Python pretraining described; sft release is best interpreted as post-training of the authors same original base family.", "provenance_sources": [{"url": "https://huggingface.co/ztor2/jerboa-sft/raw/eb446a596c5b30e767a231acd6c39c73a0c56473/README.md", "accessed_at": "2026-10-30T12:26:36.709935+00:00"}, {"url": "https://huggingface.co/ztor2/jerboa-sft/raw/eb446a596c5b30e767a231acd6c39c73a0c56473/README.md", "read_scope": "Cached model/base/architecture/training and attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Majipa/axiom", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Majipa/axiom/raw/45955a892a8448115cf2af789962ab5aa435bf97/README.md", "accessed_at": "2026-10-30T12:26:37.305473+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Bagnoli-AI/Bagnoli-Zero-v10.2-250M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "unresolved", "provenance_basis": "Merged PR#2 visible old README explicitly states developed and trained from scratch by independent researcher on RX7900XTX24GB using1001850880 cumulative training tokens. Current card replaced this alongside removed benchmark claims with evaluation-pending notice; historical training report is retained as such.", "provenance_sources": [{"url": "https://huggingface.co/Bagnoli-AI/Bagnoli-Zero-v10.2-250M/blob/main/README.md"}, {"url": "https://huggingface.co/Bagnoli-AI/Bagnoli-Zero-v10.2-250M/raw/0aa17d86c24cf0bb57cbd5385369004e011de1f0/README.md", "accessed_at": "2026-10-30T12:26:37.329004+00:00"}, {"url": "https://huggingface.co/Bagnoli-AI/Bagnoli-Zero-v10.2-250M/discussions/2/files", "accessed_at": "2026-11-01", "read_method": "Chrome_rendered_UI"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "evidence_updated_at": "2026-11-01"}
{"model_id": "solhost/gpt2-300m-climbmix-pretrain", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "unresolved", "provenance_basis": "Public card explicitly says custom architecture trained from scratch;303389696-parameter RoPE GPT2-style decoder,20BClimbMix tokens,38147steps,statusCOMPLETE. Gated weight access not required to read this report.", "provenance_sources": [{"url": "https://huggingface.co/solhost/gpt2-300m-climbmix-pretrain/raw/cc6d722098e234e6d97021ec40c2456039b05dbc/README.md", "accessed_at": "2026-10-30T12:26:37.316989+00:00"}, {"url": "https://huggingface.co/solhost/gpt2-300m-climbmix-pretrain", "accessed_at": "2026-11-01", "read_method": "Chrome_rendered_UI"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "evidence_updated_at": "2026-11-01"}
{"model_id": "KerzgrAI/PromptTercet-0.5B", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Explicitly trained from random weights on user-turn datasets; released EMA snapshot.", "provenance_sources": [{"url": "https://huggingface.co/KerzgrAI/PromptTercet-0.5B/raw/f2e0f258402cdef57e8846c54522140bafcb2ef7/README.md", "accessed_at": "2026-10-30T12:26:37.317989+00:00"}, {"url": "https://huggingface.co/KerzgrAI/PromptTercet-0.5B/raw/f2e0f258402cdef57e8846c54522140bafcb2ef7/README.md", "read_scope": "Saved card architecture, training and checkpoint description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "DedeProGames/Kiyo-65M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Kiyo-65M reports random initialization and 80B training tokens; Kiyo-230M-Preview from-scratch pretraining with 42.7B. Exact-card hardware absent. Other model cards name RTX3060/HF Jobs/Colab hardware, kept separate.", "provenance_sources": [{"url": "https://huggingface.co/DedeProGames/Kiyo-65M/raw/ec416a8a95e227fd72bcacacb0f5638237b02c48/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not stated for exact run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "DedeProGames/Kiyo-230M-Preview", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Kiyo-65M reports random initialization and 80B training tokens; Kiyo-230M-Preview from-scratch pretraining with 42.7B. Exact-card hardware absent. Other model cards name RTX3060/HF Jobs/Colab hardware, kept separate.", "provenance_sources": [{"url": "https://huggingface.co/DedeProGames/Kiyo-230M-Preview/raw/85c03f29c590a91b08d8ce2a3d8c981ad64fe348/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not stated for exact run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Ahmadbader17/tiny-gpt", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_documented", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'from_scratch_documented', 'detail': 'Card describes a custom PyTorch implementation, tokenizer trained on training split, and 9500-step/77.8M-token training run over TinyStories.'}", "provenance_sources": [{"url": "https://huggingface.co/Ahmadbader17/tiny-gpt/raw/36ee8edddcf66cf37ba6d85cd893c9b18a53e083/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:26:37.915031+00:00"}, {"url": "https://github.com/Ahmadbader17", "basis": "HF_linked_GitHub_profile", "accessed_at": "2026-10-30"}, {"url": "https://github.com/Ahmadbader17/Ahmadbader17/blob/main/README.md", "basis": "HF_linked_GitHub_profile_README", "accessed_at": "2026-10-30T14:55:30Z"}, {"url": "https://www.linkedin.com/in/ahmad-bader-8a5381193/", "basis": "user_supplied_LinkedIn_text", "accessed_at": "2026-10-30"}, {"url": "https://github.com/Ahmadbader17/tinygpt/blob/a60bae92d1e53fdea3ecd76947abad7b570b12ef/docs/ENVIRONMENT.md", "basis": "parent_read_pinned_repository_document", "accessed_at": "2026-10-30"}, {"url": "https://github.com/Ahmadbader17/tinygpt/blob/a60bae92d1e53fdea3ecd76947abad7b570b12ef/docs/RESULTS.md", "basis": "parent_read_pinned_repository_training_result", "accessed_at": "2026-10-30"}, {"url": "https://github.com/Ahmadbader17/tinygpt/blob/a60bae92d1e53fdea3ecd76947abad7b570b12ef/docs/RESUME.md", "basis": "parent_read_pinned_repository_portfolio_document", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "documented", "detail": "Personal laptop training reported: Intel Core Ultra 9 185H CPU, 12 threads; environment 15.3 GB visible RAM, 22 logical cores, CPU-only PyTorch, no CUDA/MPS. No NPU training evidenced. Laptop make and cost unknown.", "training_time_seconds_excluding_sleep": 28388, "sources": [{"url": "https://github.com/Ahmadbader17/tinygpt/blob/a60bae92d1e53fdea3ecd76947abad7b570b12ef/docs/ENVIRONMENT.md", "quote": "", "basis": "parent_read_pinned_repository_document", "source_date": null, "accessed_at": "2026-10-30", "access_outcome": "cached_source_read"}, {"url": "https://github.com/Ahmadbader17/tinygpt/blob/a60bae92d1e53fdea3ecd76947abad7b570b12ef/docs/RESULTS.md", "quote": "", "basis": "parent_read_pinned_repository_training_result", "source_date": null, "accessed_at": "2026-10-30", "access_outcome": "cached_source_read"}]}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "BrodyMakezAI/IAU-1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BrodyMakezAI/IAU-1/raw/228f8cd6b96e439fbbc04f3dec7f0d98b5124643/README.md"}, {"url": "https://huggingface.co/BrodyMakezAI/IAU-1/raw/228f8cd6b96e439fbbc04f3dec7f0d98b5124643/README.md", "access_date": "2026-10-31", "scope": "complete small source file read"}, {"url": "https://huggingface.co/BrodyMakezAI/IAU-1/raw/228f8cd6b96e439fbbc04f3dec7f0d98b5124643/model.py", "access_date": "2026-10-31", "scope": "complete small source file read"}, {"url": "https://huggingface.co/BrodyMakezAI/IAU-1/raw/228f8cd6b96e439fbbc04f3dec7f0d98b5124643/config.json", "access_date": "2026-10-31", "scope": "complete small source file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Publisher reports own from-scratch training, 4000 steps; model.py implements fresh random initialization and own-checkpoint inference loading. The training entrypoint is not present in the frozen small-file tree.", "architecture_and_code_lineage": "Own compact GPT implementation of standard learned-position, pre-LN causal self-attention and GELU MLP recipe; card calls it custom architecture, and code is not a Transformers model class. GPT-2 tokenizer is reused. No new architecture claim established merely by the custom class.", "evidence_updated_at": "2026-10-31"}
{"model_id": "arihntk/gpt-from-scratch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/arihntk/gpt-from-scratch/blob/main/README.md"}, {"url": "https://huggingface.co/arihntk/gpt-from-scratch/raw/b9cdebaac9a460ccbe489919fc7ac71af8db5e1d/README.md", "accessed_at": "2026-10-30T12:26:37.929048+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "viknesh100/Atom-1.0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/viknesh100/Atom-1.0/raw/122d5dddff0360c0d0e2b8c3657768da65df01ff/README.md", "accessed_at": "2026-10-30T12:26:37.952309+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Atlas-AI-research/Quasar", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "curriculum_training_and_own_checkpoint_continuation", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Own Axiom-1.1 implementation and constructor/new-init path; distinct original text-stage dimensions; completed text-to-media curriculum/weight-preserving expansion history; card explicitly says Emu3 is a pixel codec and not a pretrained generative language backbone. Training implementation also supports own-stage continuation. No external pretrained language base is identified.", "provenance_sources": [{"url": "https://huggingface.co/Atlas-AI-research/Quasar/blob/main/README.md"}, {"url": "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/README.md", "accessed_at": "2026-10-30T12:26:37.972621+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA GeForce RTX 5070", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder training location, physical GPU country and institutional route unstated.; Initial backbone weight ancestry is not stated explicitly; code supports new initialization and own-stage continuation but is not itself proof which initial branch was executed.", "training_duration_reported": "Approximately 95 minutes for the complete six-experiment research session, including preparation, waiting, compilation and evaluation; not pure training time or one-model runtime.", "compute_source": "publisher hardware statement in frozen model card and actual saved config meta.env.gpu", "upstream_checkpoint": "Own v5 to covered-motion to joint_repair lineage is explicit; original language-backbone initialization is not explicitly established by available saved run metadata. Pretrained Emu3/DAC codecs are separate components.", "creator_contribution": "Own hybrid autoregressive backbone, multimodal token integration, curriculum training and subsequent shared-model research updates; pretrained codecs reused.", "upstream_architecture_or_code": [{"name": "Own Axiom-1.1 hybrid recurrence/attention backbone", "source_url": "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/quasar/model/axiom.py"}, {"name": "Emu3 frozen visual tokenizer", "revision": "c81f916ad371289e205310a7539255e8a9396488", "role": "pretrained pixel codec, not language backbone", "source_url": "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/README.md"}, {"name": "Descript DAC 24 kHz", "revision": "6ba020b5ba7d9d8076fb90db7e67f27e31980f6e", "role": "pretrained frozen audio codec", "source_url": "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/README.md"}], "best_origin_estimate": {"origin": "own_from_scratch_backbone_with_pretrained_media_codecs", "confidence": "medium", "direct_or_inferred": "inferred", "basis": "Own Axiom-1.1 implementation and constructor/new-init path; distinct original text-stage dimensions; completed text-to-media curriculum/weight-preserving expansion history; card explicitly says Emu3 is a pixel codec and not a pretrained generative language backbone. Training implementation also supports own-stage continuation. No external pretrained language base is identified.", "remaining_gap": "No preserved first-stage run initialization receipt or explicit from-scratch sentence for Quasar.", "source_urls": ["https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/README.md", "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/config.json", "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/release.json", "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/quasar/model/axiom.py", "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/quasar/training/trainer.py"]}, "training_tokens_recorded": {"padded_input_tokens": 807174144, "final_stage_steps": 512, "source_url": "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/config.json", "note": "Cumulative pilot/history accounting; not unique tokens or supervised outputs."}, "hardware": "1 x NVIDIA GeForce RTX 5070", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "model_provenance_scope": "language_backbone; pretrained media codecs separately attributed", "compute_sources": [{"url": "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/README.md", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}, {"url": "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/config.json", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}, {"url": "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/release.json", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}, {"url": "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/quasar/model/axiom.py", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}, {"url": "https://huggingface.co/Atlas-AI-research/Quasar/raw/a31cef2d42cc7e7cd1e2856b7cee69d69809ce22/quasar/training/trainer.py", "access_date": "2026-10-31", "scope": "first 190 lines: constructor, resume/init paths, saved environment metadata"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "kalia-lm/kalia-v020", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/kalia-lm/kalia-v020/raw/8f8b115c5a52deb1deb1a24a1a0f6bb3fa827aa6/README.md", "accessed_at": "2026-10-30T12:26:38.020687+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "kalia-lm/kalia-v012", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/kalia-lm/kalia-v012/raw/a2ab794c7fea8c43192a6ffac7095e7b6e2117b0/README.md", "accessed_at": "2026-10-30T12:26:38.549087+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Nebulixlabs/Horizon-10M-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card states from scratch,500Mtokens.", "provenance_sources": [{"url": "https://huggingface.co/Nebulixlabs/Horizon-10M-Base/blob/main/README.md"}, {"url": "https://huggingface.co/Nebulixlabs/Horizon-10M-Base/raw/684624ddabf7631b418d59b0e9a0d053842b9db6/README.md", "accessed_at": "2026-10-30T12:26:38.586784+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"gpus": "2×NVIDIA T4", "provider": "Kaggle", "precision": "FP16", "physical_country": null}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "jfdasher/gptlab-57m-3e17", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/jfdasher/gptlab-57m-3e17/raw/d55b9afdf432b98a6f8e5e9e88424dfe4467dc5c/README.md", "accessed_at": "2026-10-30T12:26:38.599084+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Aobangaming/lightning-30m-ft-src", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Selected card names own Lightning/Aoban base, describes initial large-data training then smaller-data fine-tuning of own tiny4-layer256-width model. Own scratch ancestry is best-supported medium inference; earliest initialization source not supplied.", "provenance_sources": [{"url": "https://huggingface.co/Aobangaming/lightning-30m-ft-src/blob/main/README.md"}, {"url": "https://huggingface.co/Aobangaming/lightning-30m-ft-src/raw/36518c0f714fe08e31fd3b212ab9df7aad8aa57f/README.md", "accessed_at": "2026-10-30T12:26:38.578779+00:00"}], "uk_build_status": "no", "builder_location_country": "Thailand", "builder_location_basis": "Own local-PC training explicitly in Asia, September release; same verified creator channel identifies Thailand in October. Medium nearby-date base estimate.", "builder_location_evidence_type": "inferred", "builder_location_period": "September2026 training/release; channel country observed31October2026", "compute_used": "NVIDIA RTX3050 6GB, own computer; Windows11 and Intel i5-10400 context", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "training_duration_reported": "1.5 hours (publisher environmental/training record)", "compute_source": "Own computer", "creator_contribution": "Own Lightning/Aoban development and Booksum fine-tuning; both selected IDs cite the same project/training description and are not assumed independent training runs.", "provenance_confidence": "medium", "architecture_lineage": "Own small standard autoregressive Transformer,4 layers/width256/4-head MHA and FlashAttention;50K vocabulary.", "compute_physical_region": "Asia", "hardware": "NVIDIA RTX3050 6GB, own computer; Windows11 and Intel i5-10400 context", "builder_location_at_training": {"country": "Thailand", "confidence": "medium", "direct_or_inferred": "inferred", "period": "September2026", "basis": "Local own-PC Asia training and same creator Thailand country field October2026."}, "compute_sources": [{"url": "https://huggingface.co/Aobangaming/lightning-30m-ft-src/raw/36518c0f714fe08e31fd3b212ab9df7aad8aa57f/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}, {"url": "https://www.youtube.com/@AobanZ/about", "access_date": "2026-10-31", "scope": "Valid prior actual rendered About country reading reused"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "ztor2/jerboa-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original221M28-layer Jerboa base with49152BPE and explicit1.23B-token FineWeb/Python pretraining supports own scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/ztor2/jerboa-base/raw/57507efb08a0cc768bf4bd5f82d69c158a738111/README.md", "accessed_at": "2026-10-30T12:26:38.620126+00:00"}, {"url": "https://huggingface.co/ztor2/jerboa-base/raw/57507efb08a0cc768bf4bd5f82d69c158a738111/README.md", "read_scope": "Cached model/base/architecture/training and attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "AneekC/quipu-114m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/AneekC/quipu-114m/raw/c9bcf161d4dbadbe0e7f6df3105c249b77887eea/README.md", "accessed_at": "2026-10-30T12:26:38.650173+00:00"}, {"url": "https://sg.linkedin.com/in/aneek-chattopadhyay-1a253020a", "basis": "Public LinkedIn search excerpt read; direct access restricted or partial", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 RTX5060 Laptop8GB;45h22m", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AneekC/quipu-moe-1B-A149M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/AneekC/quipu-moe-1B-A149M/raw/70c136b10f191791f39236f4b966061dd7f421ed/README.md", "accessed_at": "2026-10-30T12:26:39.129321+00:00"}, {"url": "https://sg.linkedin.com/in/aneek-chattopadhyay-1a253020a", "basis": "Public LinkedIn search excerpt read; direct access restricted or partial", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 rented RTX5090 32GB;USD18.75 session/data-build ledger", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "kerzgrr/Senryu", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/kerzgrr/Senryu/raw/d5df53019cacc6b4662fcb6876869945259fb366/README.md", "accessed_at": "2026-10-30T12:26:39.233145+00:00", "quote": "uery. - Pretraining on 1.15B FineWeb-Edu tokens - Supervised fine-tuning at a 32,768-token sequence length on 6.7B tokens (5.39B assistant tokens) of reasoning, tool-call, instruction-following and chat data - Then APO-zero on 22,517 on-policy prefer", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Astra228/ChudGPT-Femboy", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Astra228/ChudGPT-Femboy/blob/main/README.md"}, {"url": "https://huggingface.co/Astra228/ChudGPT-Femboy/raw/40cdc610cf031d34f22874b7a8d8c800070a8820/README.md", "accessed_at": "2026-10-30T12:26:39.242396+00:00"}, {"url": "https://huggingface.co/Astra228/ChudGPT-Femboy/raw/40cdc610cf031d34f22874b7a8d8c800070a8820/README.md", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}, {"url": "https://huggingface.co/Astra228/ChudGPT-Femboy/raw/40cdc610cf031d34f22874b7a8d8c800070a8820/config.json", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}, {"url": "https://huggingface.co/Astra228/ChudGPT-Femboy/raw/40cdc610cf031d34f22874b7a8d8c800070a8820/manifest.json", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}, {"url": "https://huggingface.co/Astra228/ChudGPT-Femboy/raw/40cdc610cf031d34f22874b7a8d8c800070a8820/model.py", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Exact GPU product, training duration, physical compute location, builder country and institutional route unstated.", "compute_source": "No actual device report found in frozen model files.", "upstream_checkpoint": "Own V16 personality development checkpoint selected at step 500; no external pretrained base reported.", "creator_contribution": "Own small BPE-tokenized decoder pretraining and personality-training experiment; selected development checkpoint published.", "upstream_architecture_or_code": [{"name": "Decoder-only Transformer with RoPE and standard PyTorch components", "finding": "Model-specific implementation; Femboy code explicitly initializes modules with normal weights. Standard architecture recipe is separate from checkpoint lineage.", "source_url": "https://huggingface.co/Astra228/ChudGPT-Femboy/raw/40cdc610cf031d34f22874b7a8d8c800070a8820/model.py"}], "parameter_count_publisher_reported": 29697408, "parameter_count_inventory": "32843136", "evidence_updated_at": "2026-10-31"}
{"model_id": "Astra228/ChudGPT-Public", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "publisher_statement_and_selected_checkpoint_report", "provenance_basis": "Selected public_assistant_rebalance/latest.pt is own-lineage response-only fine-tuning from games-v2 step800 to step1400, while project ancestry is explicitly independently trained from scratch.", "provenance_sources": [{"url": "https://huggingface.co/Astra228/ChudGPT-Public/raw/29261661276b9fd8393a490dcf41a38a6dcb2db6/README.md", "accessed_at": "2026-10-30T12:26:39.222639+00:00", "quote": "antrebalance/latest.pt neural checkpoint. It is a custom, independently trained decoder-only model, not a fine-tune of an external model. The live Public V20 application combines neural generation with additional reviewed responses, exact arithmetic, normali", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}, {"url": "https://huggingface.co/Astra228/ChudGPT-Public/raw/29261661276b9fd8393a490dcf41a38a6dcb2db6/README.md", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}, {"url": "https://huggingface.co/Astra228/ChudGPT-Public/raw/29261661276b9fd8393a490dcf41a38a6dcb2db6/config.json", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}, {"url": "https://huggingface.co/Astra228/ChudGPT-Public/raw/29261661276b9fd8393a490dcf41a38a6dcb2db6/manifest.json", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}, {"url": "https://huggingface.co/Astra228/ChudGPT-Public/raw/29261661276b9fd8393a490dcf41a38a6dcb2db6/LICENSE", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}, {"url": "https://github.com/ASTRA228b/ChudGPT-Public/blob/main/README.md", "access_date": "2026-10-31", "scope": "complete project README read", "finding": "Independently trained 20,999,184-parameter decoder Transformer; current selected assistant-rebalance step 1400. Runtime reviewed responses and exact operations are separate. GPU with 8 GB is a recommendation, not an actual device report."}, {"url": "https://github.com/ASTRA228b/ChudGPT-Public/blob/main/train.py", "access_date": "2026-10-31", "scope": "complete training implementation read", "finding": "Instantiates TransformerLM then optionally resumes or initializes from a checkpoint; CUDA/CPU selection is configurable. No GPU product name or training country is reported."}, {"url": "https://github.com/ASTRA228b/ChudGPT-Public/blob/main/train_pipeline.ps1", "access_date": "2026-10-31", "scope": "complete training launcher read", "finding": "Windows local CUDA launch pipeline runs base pretraining then response-only fine-tuning; this is execution configuration, not a completed-run hardware specification."}, {"url": "https://github.com/ASTRA228b/ChudGPT-Public/blob/main/reports/ASSISTANT_REBALANCE.md", "access_date": "2026-10-31", "scope": "complete checkpoint-matched report read", "finding": "Selected public_assistant_rebalance/latest.pt at step 1400, continued from own games-v2 step 800. Reports completed 1400 CUDA fine-tuning steps on 1347 original/reviewed conversations; no exact GPU product or duration."}, {"url": "https://github.com/ASTRA228b/ChudGPT-Public/blob/main/PUBLIC_FINAL_REPORT.md", "access_date": "2026-10-31", "scope": "complete training and final-quality report read", "finding": "Own foundation and SFT experiments and rejected candidates described, distinct from selected checkpoint."}, {"url": "https://github.com/ASTRA228b/ChudGPT-Public/blob/main/BALANCED_FINETUNING_REPORT.md", "access_date": "2026-10-31", "scope": "complete training report read", "finding": "Actual fine-tuning candidates failed release criteria; retained locally rather than incorrectly counted as new selected weights."}, {"url": "https://github.com/ASTRA228b/ChudGPT-Public/blob/main/IMPROVEMENT_REPORT.md", "access_date": "2026-10-31", "scope": "complete project improvement report read", "finding": "Own checkpoint and CUDA response-only training history; substantial runtime answer routing distinguished from raw neural output."}, {"url": "https://github.com/ASTRA228b/ChudGPT-Public/blob/main/reports/PUBLIC_V20_FINAL_REPORT.md", "access_date": "2026-10-31", "scope": "complete public training/serving report read", "finding": "Earlier own Public V15 foundation and CUDA response-only SFT documented; reviewed-response runtime explicitly separate. Private underlying logs were not accessed."}, {"url": "https://github.com/ASTRA228b/ChudGPT-Public/blob/main/reports/PUBLIC_V20_RETRAINING_2026-09-08.md", "access_date": "2026-10-31", "scope": "complete retraining report read", "finding": "Dolly/TinyStories foundation-data sources and unselected experimental candidates reported; no physical training geography."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Exact GPU product, training duration, physical compute location, builder country and institutional route unstated.", "compute_source": "Checkpoint-matched report directly states completed CUDA fine-tuning; exact GPU product, provider and duration unstated.", "upstream_checkpoint": "Own games-v2 step-800 checkpoint continued into public_assistant_rebalance/latest.pt step 1400. From-scratch ancestry is claimed for the project; this released checkpoint includes subsequent own-lineage fine-tuning.", "creator_contribution": "Own small decoder pretraining lineage, original/reviewed curriculum, repeated response-only fine-tuning and app/runtime engineering.", "upstream_architecture_or_code": [{"name": "Decoder-only Transformer with RoPE and standard PyTorch components", "finding": "Model-specific implementation; Femboy code explicitly initializes modules with normal weights. Standard architecture recipe is separate from checkpoint lineage.", "source_url": "https://huggingface.co/Astra228/ChudGPT-Public/raw/29261661276b9fd8393a490dcf41a38a6dcb2db6/README.md"}], "training_execution": {"backend": "CUDA", "selected_checkpoint_steps": 1400, "stage": "response-only fine-tuning of own lineage", "source_url": "https://github.com/ASTRA228b/ChudGPT-Public/blob/main/reports/ASSISTANT_REBALANCE.md"}, "parameter_count_publisher_reported": 20999184, "parameter_count_inventory": "24144912", "evidence_updated_at": "2026-10-31"}
{"model_id": "Atlas-AI-research/Axiom-1-Chat", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Own from-scratch pretraining followed by chat tuning; this release is the chat stage.", "provenance_sources": [{"url": "https://huggingface.co/Atlas-AI-research/Axiom-1-Chat/blob/main/README.md"}, {"url": "https://huggingface.co/Atlas-AI-research/Axiom-1-Chat/raw/d09bd311735714bd7fa7c4d51324df56f7241921/README.md", "accessed_at": "2026-10-30T12:26:39.254411+00:00"}, {"url": "https://huggingface.co/Atlas-AI-research/Axiom-1-Chat/raw/d09bd311735714bd7fa7c4d51324df56f7241921/README.md", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA GeForce RTX 5070 (12 GB)", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder training location, physical GPU country and institutional route unstated.", "training_duration_reported": "22.2 hours pretraining; 1.9 hours chat tuning (publisher stage totals)", "compute_source": "publisher hardware statement in frozen model card", "upstream_checkpoint": "Own from-scratch AXIOM base then chat tuning; no external pretrained backbone stated.", "creator_contribution": "Own architecture, from-scratch 3.15B-token pretraining and 262M-token chat-tuning experiment.", "upstream_architecture_or_code": [{"name": "AXIOM recurrent working memory and Law-Cell bank", "source_url": "https://huggingface.co/Atlas-AI-research/Axiom-1-Chat/raw/d09bd311735714bd7fa7c4d51324df56f7241921/README.md", "finding": "Creator Lee Madover architecture/pipeline. Fixed two-cell result after all nine fission trials rejected; own research mechanism is distinguished from successful growth."}], "hardware": "1 x NVIDIA GeForce RTX 5070 (12 GB)", "compute_sources": [{"url": "https://huggingface.co/Atlas-AI-research/Axiom-1-Chat/raw/d09bd311735714bd7fa7c4d51324df56f7241921/README.md", "access_date": "2026-10-31", "scope": "complete relevant frozen file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "North-ML1/willow-alpha-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "creator_model_specific_statement", "provenance_basis": "Both cards identify early Forge-1V checkpoints. In the frozen Willow-base community, org-member creator arthu1 explicitly states that he pretrained from scratch before SFT/more CPT.", "provenance_sources": [{"url": "https://huggingface.co/North-ML1/willow-alpha-base/raw/687f7a605657716029e9c2442efcd05e2a2e99dd/README.md"}, {"url": "https://huggingface.co/North-ML1/willow-alpha-base/discussions/1", "access_date": "2026-11-01", "finding": "Read actual five-comment thread. Creator arthu1 identifies early Forge-1V checkpoint, states yes I pretrained from scratch, ~4–9 hours of pretraining/CPT, SFT and more CPT; hardware A10G + L40S on Modal.", "reused": false}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "North-ML1/willow-alpha", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "creator_model_specific_statement", "provenance_basis": "Both cards identify early Forge-1V checkpoints. In the frozen Willow-base community, org-member creator arthu1 explicitly states that he pretrained from scratch before SFT/more CPT.", "provenance_sources": [{"url": "https://huggingface.co/North-ML1/willow-alpha/raw/fb9d3c8bf3279fb3eb9b79cdb561184da32586ea/README.md"}, {"url": "https://huggingface.co/North-ML1/willow-alpha-base/discussions/1", "access_date": "2026-11-01", "finding": "Read actual five-comment thread. Creator arthu1 identifies early Forge-1V checkpoint, states yes I pretrained from scratch, ~4–9 hours of pretraining/CPT, SFT and more CPT; hardware A10G + L40S on Modal.", "reused": false}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "muonai/photon-x-preview", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card states from-scratch10.104M model,2500steps20.5Mtokens.", "provenance_sources": [{"url": "https://huggingface.co/muonai/photon-x-preview/blob/main/README.md"}, {"url": "https://huggingface.co/muonai/photon-x-preview/raw/374f4544141f5effdec47d465d353e2c7d00de97/README.md", "accessed_at": "2026-10-30T12:26:39.860621+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"gpus": "1×NVIDIA T4", "provider": "Kaggle", "duration": "about12minutes", "physical_country": null}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "muonai/photon-p1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Explicit no pretrained weights/base model.1.003Mparameters,~7.5Munique/~120Mseen tokens,16epochs. Card bibliography2025 differs from repository release2026; retain dates separately.", "provenance_sources": [{"url": "https://huggingface.co/muonai/photon-p1/blob/main/README.md"}, {"url": "https://huggingface.co/muonai/photon-p1/raw/dc5151951c892e05385a06af88093b80b59b46ee/README.md", "accessed_at": "2026-10-30T12:26:39.906625+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"gpus": "1×NVIDIA T4", "provider": "Kaggle", "duration": "9.46minutes", "physical_country": null}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "CinqicX/juniper1-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/CinqicX/juniper1-base/raw/cd4dce2cfa7f8005a6bd7f8f100257352abc116c/README.md", "accessed_at": "2026-10-30T12:26:39.865623+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ViuAI/Sarus-500M", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Author explicitly says developed from scratch and aligned by ViuAI; final checkpoint SFT/DPO is retained as post-training of own lineage.", "provenance_sources": [{"url": "https://huggingface.co/ViuAI/Sarus-500M/raw/436b0878ccfde0b681cf9bc9a619f110901f908c/README.md", "accessed_at": "2026-10-30T12:26:39.871135+00:00", "quote": " entirely by ViuAI. - Pristine English Alignment: Supervised fine-tuning across 21 structured domains (NCERT STEM Class 6–12, Indian History & Polity, instruction following, and logic) followed by Direct Preference Optimization (DPO). - Direc", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "evidence_updated_at": "2026-11-01"}
{"model_id": "mirgedw/tabula-nova-157m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "From-scratch tag and own custom PyTorch pretraining pipeline;166ksteps and339,967,824tokens. Last100ksteps FineWebEdu, earlier66k data provenance undocumented. Config24layers768hidden12Q/4KVheads8192vocab. Profile laptop inventory is not actual-run hardware.", "provenance_sources": [{"url": "https://huggingface.co/mirgedw/tabula-nova-157m/blob/main/README.md"}, {"url": "https://huggingface.co/mirgedw/tabula-nova-157m/raw/37880c9762dfc2e169704e7b4bfd2622c2453de9/README.md", "accessed_at": "2026-10-30T12:26:39.902114+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Own 157.55M-parameter pretraining project, 166,000 optimization steps/339,967,824 processed tokens; final 100,000 steps FineWeb-Edu.", "architecture_lineage": "Custom PyTorch implementation of standard decoder-only grouped-query Transformer with RoPE/RMSNorm/SwiGLU; no external pretrained checkpoint identified.", "training_hardware_estimate": {"device": "1 NVIDIA RTX 4050 Laptop 6GB", "confidence": "medium", "direct_or_inferred": "inferred", "basis": "Creator public hardware inventory entered 16 October 2026 on the selected model release date; own custom BF16 CUDA-capable training/inference project.", "source": "https://huggingface.co/mirgedw"}, "evidence_updated_at": "2026-11-01"}
{"model_id": "jfrick/fineweb-edu-scholarly-gpt-10m", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "continued_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "Explicit reuse of pretrained FineWeb parent, vocabulary expansion and continuation", "provenance_sources": [{"url": "https://huggingface.co/jfrick/fineweb-edu-scholarly-gpt-10m/blob/320d1fe54c611756317ad0a7bbeeb87c686325c3/README.md", "basis": "Census-pinned full model card.", "accessed_at": "2026-10-30"}, {"url": "https://www.linkedin.com/in/jakobfrick", "basis": "Self-profile excerpt, exact Oxide/Rust project match to HF-linked GitHub. Direct fetch999.", "accessed_at": "2026-10-30"}, {"url": "https://patentimages.storage.googleapis.com/0c/da/8b/b0c685ccdc8817/US11106331.pdf", "basis": "Primary patent page1 lists Jakob Frick, London(GB); filed2019-05-20, granted2021-08-31. Only city/country retained.", "accessed_at": "2026-10-30"}, {"url": "https://arda.ai/", "basis": "Current company founder biography explicitly identifies Jakob Frick, Radiant history and six years at Palantir.", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/jfrick/fineweb-edu-scholarly-gpt-10m/raw/320d1fe54c611756317ad0a7bbeeb87c686325c3/README.md", "accessed_at": "2026-10-30T12:26:40.334306+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "One NVIDIA L40S on Modal,176 seconds worker time", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "zekun-shi/FreeSolo-743M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies the authors' custom FreeSolo research base, actual FineWeb-Edu training at4096 context and checkpoint benchmark results.", "provenance_sources": [{"url": "https://huggingface.co/zekun-shi/FreeSolo-743M/raw/14d2137f3e1ab44bd7825abc775ef464015c8c13/README.md", "accessed_at": "2026-10-30T12:26:40.506309+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "zekun-shi/FreeSolo-341M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies the authors' custom FreeSolo research base, actual FineWeb-Edu training at4096 context and checkpoint benchmark results.", "provenance_sources": [{"url": "https://huggingface.co/zekun-shi/FreeSolo-341M/raw/2f9bb7e015d913bad3c847976e46384cd958c623/README.md", "accessed_at": "2026-10-30T12:26:40.523909+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "ml-resources/MobileMoE-S-QAT", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_of_quantization_aware_trained_instruction_checkpoint", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'mirror_of_quantization_aware_trained_instruction_checkpoint', 'base_family_origin': 'Meta MobileMoE pretraining and mid-training, then SFT and INT4 QAT', 'cohort_checkpoint_initialization': 'MobileMoE-S-SFT; this mirror is not a new from-scratch training run', 'stages': [{'stage': 'pretraining', 'tokens_approx': 6000000000000}, {'stage': 'mid-training', 'tokens_approx': 500000000000}, {'stage': 'SFT', 'tokens_approx': 126000000000}, {'stage': 'QAT', 'tokens_approx': 21000000000}], 'basis': 'Pinned card says the mirror credits the original authors and QAT is applied to the SFT checkpoint. Original base-family training and this released derivative are separated.'}", "provenance_sources": [{"url": "https://huggingface.co/ml-resources/MobileMoE-S-QAT/raw/bfec284daa1ba7a71168b002c1020e307ade5dda/README.md", "basis": "Pinned model card from the frozen cohort.", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2605.27358v1", "basis": "Primary report, training setup/Table1; source prints96GB for these nodes.", "accessed_at": "2026-10-30"}, {"url": "https://yanbeic.github.io/", "basis": "Historical doctoral institution; current biography identifies Meta research.", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/ml-resources/MobileMoE-S-QAT/blob/main/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"gpus": "8×H100 per node (paper prints96GB)", "scope": "Original Meta family, not archive uploader", "pretraining_midtraining_nodes": "16–32", "sft_qat_nodes": "4–8", "wall_clock": "pretraining3–4weeks; midtraining2days;SFT/QAT2–3days", "source_url": "https://arxiv.org/html/2605.27358v1"}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "jfrick/shakespeare-vanilla-gpt-10m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "Explicit model-card random initialisation/no pretrained weights", "provenance_sources": [{"url": "https://huggingface.co/jfrick/shakespeare-vanilla-gpt-10m/blob/a289931765de3e1e338164806daed5288e0ffed9/README.md", "basis": "Census-pinned full model card.", "accessed_at": "2026-10-30"}, {"url": "https://www.linkedin.com/in/jakobfrick", "basis": "Self-profile excerpt, exact Oxide/Rust project match to HF-linked GitHub. Direct fetch999.", "accessed_at": "2026-10-30"}, {"url": "https://patentimages.storage.googleapis.com/0c/da/8b/b0c685ccdc8817/US11106331.pdf", "basis": "Primary patent page1 lists Jakob Frick, London(GB); filed2019-05-20, granted2021-08-31. Only city/country retained.", "accessed_at": "2026-10-30"}, {"url": "https://arda.ai/", "basis": "Current company founder biography explicitly identifies Jakob Frick, Radiant history and six years at Palantir.", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/jfrick/shakespeare-vanilla-gpt-10m/raw/a289931765de3e1e338164806daed5288e0ffed9/README.md", "accessed_at": "2026-10-30T12:26:40.595607+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "One NVIDIA L4 on Modal", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sam-eer12/chadGPT", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/sam-eer12/chadGPT/raw/d78a6ed405b51b28f26eec31e68e88919f18ce17/README.md", "accessed_at": "2026-10-30T12:26:40.534099+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sraivante/rxgpt-india", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sraivante/rxgpt-india/raw/5d1fe3eb04e24f069ebf3631a67c0603c9b32a2c/README.md", "accessed_at": "2026-10-30T12:26:40.965644+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Can111/ShellElastic-baseline-FullyElastic-119M-285M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card documents original ShellElastic research pretraining, 25,000 steps/6.55B DCLM tokens and truncated-normal initialization. Baselines are expressly trained for this paper, with upstream MatFormer/slimmable-network ideas credited, not copied official baseline checkpoints. Builder attribution is Can Jin and Rutgers/Adobe coauthors.", "provenance_sources": [{"url": "https://huggingface.co/Can111/ShellElastic-baseline-FullyElastic-119M-285M/raw/d2420dcf56bcc41def48c7e924d28d3addece243/README.md", "accessed_at": "2026-10-30T12:26:41.156807+00:00"}, {"url": "https://huggingface.co/Can111/ShellElastic-baseline-FullyElastic-119M-285M/raw/d2420dcf56bcc41def48c7e924d28d3addece243/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "Can111/ShellElastic-baseline-MatFormer-119M-285M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card documents original ShellElastic research pretraining, 25,000 steps/6.55B DCLM tokens and truncated-normal initialization. Baselines are expressly trained for this paper, with upstream MatFormer/slimmable-network ideas credited, not copied official baseline checkpoints. Builder attribution is Can Jin and Rutgers/Adobe coauthors.", "provenance_sources": [{"url": "https://huggingface.co/Can111/ShellElastic-baseline-MatFormer-119M-285M/raw/85e4f1a598ac584c8d501daf4004da6a29344eda/README.md", "accessed_at": "2026-10-30T12:26:41.244155+00:00"}, {"url": "https://huggingface.co/Can111/ShellElastic-baseline-MatFormer-119M-285M/raw/85e4f1a598ac584c8d501daf4004da6a29344eda/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "Can111/ShellElastic-119M-285M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card documents original ShellElastic research pretraining, 25,000 steps/6.55B DCLM tokens and truncated-normal initialization. Baselines are expressly trained for this paper, with upstream MatFormer/slimmable-network ideas credited, not copied official baseline checkpoints. Builder attribution is Can Jin and Rutgers/Adobe coauthors.", "provenance_sources": [{"url": "https://huggingface.co/Can111/ShellElastic-119M-285M/raw/7515f3d925bcd9620c0db1b20dd681624ca33278/README.md", "accessed_at": "2026-10-30T12:26:41.173356+00:00"}, {"url": "https://huggingface.co/Can111/ShellElastic-119M-285M/raw/7515f3d925bcd9620c0db1b20dd681624ca33278/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "Usaidddddddddddddd/UsaidAI-500M", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/Usaidddddddddddddd/UsaidAI-500M/raw/1f312027f1160a8a40c01686b20d9b67492ca3fb/README.md", "accessed_at": "2026-10-30T12:26:41.202878+00:00", "quote": "500M Parameter Causal Language Model Architected, Pretrained & SFT Aligned from Scratch Engineered by Mohamed Usaid [![Space](https://img.shields.io/badge/%F0%9F%A4%97%20Space-Interactive%20Chat-orange.svg)](https://huggingface.co/spaces/Usaiddddd", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Manovishaal/MaVi-0.2-65M", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/Manovishaal/MaVi-0.2-65M/raw/88278842044763335cddb74836f812d608496791/README.md", "accessed_at": "2026-10-30T12:26:41.224131+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Manovishaal/MaVi-0.1-65M", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/Manovishaal/MaVi-0.1-65M/raw/33bc97fea6c7b45914e27a8958228359aafcc8c7/README.md", "accessed_at": "2026-10-30T12:26:41.511256+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "mithulaartigala/vega-1-65m-base-experimental", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author identifies the initial pretraining checkpoint of a custom65.3M Qyrou architecture with bespoke20K tokenizer,21 mixed convolution/attention layers and actual mixed-corpus training.", "provenance_sources": [{"url": "https://huggingface.co/mithulaartigala/vega-1-65m-base-experimental/raw/2ed0014129572f24b80369e5bb89aa5110e92931/README.md", "accessed_at": "2026-10-30T12:26:41.784070+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "ArchiveStudio/Muse2-125M-Base", "model_provenance": "mirror", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "not_established", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Upstream Muse Research checkpoint; no training claim by the archive publisher.", "provenance_sources": [{"url": "https://huggingface.co/ArchiveStudio/Muse2-125M-Base/blob/main/README.md"}, {"url": "https://huggingface.co/ArchiveStudio/Muse2-125M-Base/raw/6cc6b7a799e1258f5fd1447cccc9917197af4eb7/README.md", "accessed_at": "2026-10-30T12:26:41.809107+00:00"}, {"url": "https://huggingface.co/ArchiveStudio/Muse2-125M-Base/blob/main/README.md", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/ArchiveStudio/Muse2-125M-Base/raw/6cc6b7a799e1258f5fd1447cccc9917197af4eb7/README.md", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/api/models/ArchiveStudio/Muse2-125M-Base/discussions?p=0", "access_date": "2026-10-31", "scope": "complete discussion-list JSON read", "finding": "unavailable listed public threads", "http_status": 401}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Current archive model Community endpoints are inaccessible, so historical thread contents remain unreviewable.; Original-builder training hardware and geography will be followed on the separately queued Muse-research account; no country attributed to Jin from upstream work.", "upstream_checkpoint": "Muse Research Muse2 model family; original publisher Muse-research, original uploader Ill-Ness in saved history.", "creator_contribution": "ArchiveStudio redistributes Muse Research weights. The read sources do not attribute a training run to archive operator Jin.", "upstream_architecture_or_code": [{"name": "LFM2-style hybrid short-convolution / grouped-query-attention layout", "attribution_location": "saved frozen model card; upstream developer Muse Research", "source_url": "https://huggingface.co/ArchiveStudio/Muse2-125M-Base/blob/main/README.md"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "ArchiveStudio/Muse2-230M-Base", "model_provenance": "mirror", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Upstream Muse Research checkpoint; no training claim by the archive publisher.", "provenance_sources": [{"url": "https://huggingface.co/ArchiveStudio/Muse2-230M-Base/blob/main/README.md"}, {"url": "https://huggingface.co/ArchiveStudio/Muse2-230M-Base/raw/5037ac62b5bbea752d4aec8ee45c4fc250b30771/README.md", "accessed_at": "2026-10-30T12:26:41.829125+00:00"}, {"url": "https://huggingface.co/ArchiveStudio/Muse2-230M-Base/blob/main/README.md", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/ArchiveStudio/Muse2-230M-Base/raw/5037ac62b5bbea752d4aec8ee45c4fc250b30771/README.md", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/api/models/ArchiveStudio/Muse2-230M-Base/discussions?p=0", "access_date": "2026-10-31", "scope": "complete discussion-list JSON read", "finding": "unavailable listed public threads", "http_status": 401}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Current archive model Community endpoints are inaccessible, so historical thread contents remain unreviewable.; Original-builder training hardware and geography will be followed on the separately queued Muse-research account; no country attributed to Jin from upstream work.", "upstream_checkpoint": "Muse Research Muse2 model family; original publisher Muse-research, original uploader Ill-Ness in saved history.", "creator_contribution": "ArchiveStudio redistributes Muse Research weights. The read sources do not attribute a training run to archive operator Jin.", "upstream_architecture_or_code": [{"name": "LFM2-style hybrid short-convolution / grouped-query-attention layout", "attribution_location": "saved frozen model card; upstream developer Muse Research", "source_url": "https://huggingface.co/ArchiveStudio/Muse2-230M-Base/blob/main/README.md"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "n00nehere/recursive-bitnet-ngram-102M-64K-instruct-preview", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "continued_pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Own100.68M backbone initialized randomly, no external pretrained model weights; pretrained Cosmo2 tokenizer is separate. Continuation adds~1.6M n-gram parameters;4,703,125,504inputtokens,selectedstep35,882,64Kcontext,9,281additionalupdates. Own original pretraining then mixed instruction continuation, rather than external-backbone fine-tune.", "provenance_sources": [{"url": "https://huggingface.co/n00nehere/recursive-bitnet-ngram-102M-64K-instruct-preview/blob/main/README.md"}, {"url": "https://huggingface.co/n00nehere/recursive-bitnet-ngram-102M-64K-instruct-preview/raw/541a5fc330efba4df818c476d6e4a88c4d3735f9/README.md", "accessed_at": "2026-10-30T12:26:41.849958+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"backbone": "4×NVIDIA B300 ranks,3901.12seconds", "continuation": "1×NVIDIA B300 SXM6 AC,12165.55seconds", "provider": "Modal.com free credits explicitly in author's Reddit comment", "physical_country": null, "source_url": "https://www.reddit.com/r/Italia/comments/1x6t525/comment/pgz6z1f/"}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "OkeyMetaLtd/Tern-0.1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/OkeyMetaLtd/Tern-0.1/raw/bd2deb138f3e546f6391659c09c8f770516a4d53/README.md", "accessed_at": "2026-10-30T12:26:41.861483+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "solintellegence/sol-pro", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Selected card explicitly says task-adapted checkpoint. Own bespoke 138M Sol Flash/Sol Pro recurrent tensorized n-gram architecture has fresh constructor/normal initialization, own 4096 tokenizer, and same identified creator’s scratch-development lineage; medium own-backbone origin inference.", "provenance_sources": [{"url": "https://huggingface.co/solintellegence/sol-pro/raw/812d478a4af90c80b73d79c2a5cbdcc6edfb6667/README.md", "accessed_at": "2026-10-30T12:26:42.152458+00:00"}, {"url": "https://huggingface.co/solintellegence/sol-pro/raw/812d478a4af90c80b73d79c2a5cbdcc6edfb6667/README.md", "scope": "Complete saved frozen card read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/solintellegence/sol-pro/raw/812d478a4af90c80b73d79c2a5cbdcc6edfb6667/modeling_sol_pro.py", "scope": "Fresh constructor, initializers, own-artifact loader and component descriptions read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/solintellegence/sol-pro/discussions/1", "scope": "All public comments read", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "creator_contribution": "Own 138M recurrent model development plus documented benchmark-training-split adaptation; reported scores are not an independent clean benchmark result.", "best_origin_estimate": {"origin": "own_from_scratch_backbone_then_task_adaptation", "confidence": "medium", "basis": "Own bespoke configuration/constructor and creator development chain, followed by explicit card task adaptation.", "remaining_gap": "Earliest base-run initialization receipt absent."}, "provenance_confidence": "medium", "architecture_lineage": "Own recurrent GQA/XSA implementation with RoPE/RMSNorm/SwiGLU and rank-299 tensorized 2/3/4-gram memory; source calls it scaled Sol Flash architecture.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "slendermantvb/jarvis-v2.5.0", "model_provenance": "mirror", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "runtime_only_republication_of_own_checkpoint", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Project describes original Wikimedia Spanish/English pretraining and locally generated arithmetic/Python examples. Bespoke TuIA constructor creates all modules and applies normal random initialization. Best-supported own scratch core, medium confidence. Selected v2.5 card explicitly keeps the same stable neural core and changes the runtime; it is not evidence of another training run.", "provenance_sources": [{"url": "https://huggingface.co/slendermantvb/jarvis-v2.5.0/raw/a3483d418050de19c67010c70282aec27effc3ac/README.md", "accessed_at": "2026-10-30T12:26:42.539996+00:00"}, {"url": "https://huggingface.co/slendermantvb/jarvis-v2.5.0/raw/a3483d418050de19c67010c70282aec27effc3ac/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/slendermantvb/jarvis-v2.5.0/raw/a3483d418050de19c67010c70282aec27effc3ac/NOTICE", "scope": "Complete training-data/provenance notice read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/slendermantvb/jarvis-v2.5.0/raw/a3483d418050de19c67010c70282aec27effc3ac/model.py", "scope": "Fresh model constructor, initialization and architecture attribution read", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "creator_contribution": "Own compact neural-core pretraining plus application runtime/RAG/Lisp routing/code verification; application safeguards are separate from raw neural weights.", "provenance_confidence": "medium", "architecture_lineage": "Own compact PyTorch implementation using DeepSeek-style MLA-inspired latent-KV attention, shared-expert top-2 MoE, RoPE/RMSNorm/SwiGLU and multi-token prediction. Source explicitly credits MLA/DeepSeek-style component inspiration.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "MonumentalSystems/rotor-25m-annealed", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "continued_pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Annealed checkpoint belongs to the author's original RotorNative25M oscillator/Clifford byte-level model; saved step64545 and BPB document trained continuation of that custom family.", "provenance_sources": [{"url": "https://huggingface.co/MonumentalSystems/rotor-25m-annealed/raw/0503f8670df86a8f27cc2d4c66def4a17012c979/README.md", "accessed_at": "2026-10-30T12:26:42.466769+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "slendermantvb/jarvis-v2.4.1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Project describes original Wikimedia Spanish/English pretraining and locally generated arithmetic/Python examples. Bespoke TuIA constructor creates all modules and applies normal random initialization. Best-supported own scratch core, medium confidence.", "provenance_sources": [{"url": "https://huggingface.co/slendermantvb/jarvis-v2.4.1/raw/6fec68d1dd73d58a89d1bb8db1d5bef4417f5939/README.md", "accessed_at": "2026-10-30T12:26:42.602125+00:00"}, {"url": "https://huggingface.co/slendermantvb/jarvis-v2.4.1/raw/6fec68d1dd73d58a89d1bb8db1d5bef4417f5939/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/slendermantvb/jarvis-v2.4.1/raw/6fec68d1dd73d58a89d1bb8db1d5bef4417f5939/NOTICE", "scope": "Complete training-data/provenance notice read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/slendermantvb/jarvis-v2.4.1/raw/6fec68d1dd73d58a89d1bb8db1d5bef4417f5939/model.py", "scope": "Fresh model constructor, initialization and architecture attribution read", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "creator_contribution": "Own compact neural-core pretraining plus application runtime/RAG/Lisp routing/code verification; application safeguards are separate from raw neural weights.", "provenance_confidence": "medium", "architecture_lineage": "Own compact PyTorch implementation using DeepSeek-style MLA-inspired latent-KV attention, shared-expert top-2 MoE, RoPE/RMSNorm/SwiGLU and multi-token prediction. Source explicitly credits MLA/DeepSeek-style component inspiration.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "KrynexLabs/KrynexAI-1.2-25M-Mobile-TFLite", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "25M card expressly fine-tuned original Pebble25M after25B-token base,250M SFT tokens.50M attributes original Pebble50M and reports25B-token base; uploader's transformation not clearly described. Inference-compatible GPU recommendations are not actual training hardware.", "provenance_sources": [{"url": "https://huggingface.co/KrynexLabs/KrynexAI-1.2-25M-Mobile-TFLite/raw/f717d2a5e7757d1a6fb8865dac10f7216bf708de/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "longtimedevs/TimeGPT-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "TimeGPTBase explicitly says entirely from scratch without third-party weights/distillation, independent LongTimeDevs team,1xRTX5060Ti16GB,50ksteps,~820Mtokens,~27h. Predominantly Russian training text does not locate builder.", "provenance_sources": [{"url": "https://huggingface.co/longtimedevs/TimeGPT-Base/raw/4bba61547c71cfa4ad5c4f0a61331c7b87d72001/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 NVIDIA GeForce RTX5060Ti16GB", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "training_duration_reported": "~27 hours continuous training", "creator_contribution": "Own0.5B multilingual pretraining, approximately55% Russian text, plus own tokenizer.", "architecture_lineage": "Standard modern Llama3/Qwen2.5-style GQA Transformer with RoPE/RMSNorm/SwiGLU and own32K byte-level BPE; card claims no borrowed pretrained weights.", "hardware": "1 NVIDIA GeForce RTX5060Ti16GB", "compute_sources": [{"url": "https://huggingface.co/longtimedevs/TimeGPT-Base/raw/4bba61547c71cfa4ad5c4f0a61331c7b87d72001/README.md", "scope": "Complete frozen Russian card read", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "FWKV/Myosotis-1.5-base-Preview", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original110M FWKV/RWKV-style Myosotis family, factorized embedding/head and described general-language pretraining mixture/optimizer support own scratch-origin base preview. Training-regimen section names prior Myosotis1, so exact1.5run details remain family-level inference.", "provenance_sources": [{"url": "https://huggingface.co/FWKV/Myosotis-1.5-base-Preview/raw/00d9a79c066f9ed86a37cfae86c5099f198cd4ef/README.md", "accessed_at": "2026-10-30T12:26:42.718995+00:00"}, {"url": "https://huggingface.co/FWKV/Myosotis-1.5-base-Preview/raw/00d9a79c066f9ed86a37cfae86c5099f198cd4ef/README.md", "read_scope": "Cached card model/architecture/training/result and upstream description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "kerzgrr/Haiku", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/kerzgrr/Haiku/raw/1f5051c184daea2a1bf618ebea1733b173eda7c0/README.md", "accessed_at": "2026-10-30T12:26:43.049034+00:00", "quote": " Haiku family, the ~655M successor to Tercet-R. - Supervised fine-tuning at a 32,768-token sequence length on ~9.5B tokens of reasoning, tool-call, instruction-following and chat data - Then APO-zero preference optimization on the SmolTalk2 preferenc", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "kerzgrr/Haiku-base", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/kerzgrr/Haiku-base/raw/a99182ba5f6854e342de650105b09d40bc1f1c2c/README.md", "accessed_at": "2026-10-30T12:26:43.074029+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "alby13/LibraMindMini", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/alby13/LibraMindMini/raw/eb19c1c3d00fe60545db6a576133821e7edc62e0/README.md", "accessed_at": "2026-10-30T12:26:43.144595+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Kepler182ab/llama2-15m-tinystories", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "llama2-15m-tinystories explicitly republishes Karpathy stories15M checkpoint; examples point to upstream0rn0/aryandeore code. Republisher identity is separate from original training authors. No own training hardware.", "provenance_sources": [{"url": "https://huggingface.co/api/models/Kepler182ab/llama2-15m-tinystories/commits/6c7dbdeb4fbd8a7a4227629c62f38c0d3207457f"}, {"url": "https://huggingface.co/Kepler182ab/llama2-15m-tinystories/raw/6c7dbdeb4fbd8a7a4227629c62f38c0d3207457f/README.md", "basis": "Pinned cohort model card", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "YuvrajSingh9886/facebook-opt-350m-8bit-bnb", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/YuvrajSingh9886/facebook-opt-350m-8bit-bnb/raw/780b515ead9294728c470da56053759272b11f90/README.md", "accessed_at": "2026-10-30T12:26:43.283415+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "omurberaisik/NoTokenLM-Gen-4.6", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/omurberaisik/NoTokenLM-Gen-4.6/raw/2851d0008d46640872cf71a9c81dca5f30745271/README.md", "accessed_at": "2026-10-30T12:26:43.361236+00:00", "quote": ". What is this checkpoint, actually? Gen-4.6 is a from-scratch pretraining run of a 36.8M-parameter byte-level transformer (15 layers, dmodel=448). This card describes step 9,000, and every number in it comes from that checkpoint's EMA (e", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "PiotrSty/gollem-v6-250m-sft-answers", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved Polish card expressly describes two SFT phases on SlayerLab/GoLLeM-v6-250M and calls it a CC-BY-SA derivative of SlayerLab/Fabryka AI. Underlying model creator attribution remains SlayerLab.", "provenance_sources": [{"url": "https://huggingface.co/PiotrSty/gollem-v6-250m-sft-answers/raw/3a60edc280d93fb73ead0f5a5ba220986525aa76/README.md", "accessed_at": "2026-10-30T12:26:43.669861+00:00"}, {"url": "https://huggingface.co/PiotrSty/gollem-v6-250m-sft-answers/raw/3a60edc280d93fb73ead0f5a5ba220986525aa76/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "ArsenyIvanov/mdlm-moe-lm1b", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Authored E-MoE/MDLM dense and MoE research baselines explicitly trained1Msteps/65.5B LM1B tokens on4H200; new diffusion backbone and controlled training support own scratch-origin lineage.", "provenance_sources": [{"url": "https://huggingface.co/ArsenyIvanov/mdlm-moe-lm1b/raw/39d98d6c31b3653767420d246c5dd7e12bf9a41e/README.md", "accessed_at": "2026-10-30T12:26:43.715940+00:00"}, {"url": "https://huggingface.co/ArsenyIvanov/mdlm-moe-lm1b/raw/39d98d6c31b3653767420d246c5dd7e12bf9a41e/README.md", "read_scope": "Saved card architecture, training and checkpoint description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "ArsenyIvanov/emoe-lm1b", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Authored E-MoE/MDLM dense and MoE research baselines explicitly trained1Msteps/65.5B LM1B tokens on4H200; new diffusion backbone and controlled training support own scratch-origin lineage.", "provenance_sources": [{"url": "https://huggingface.co/ArsenyIvanov/emoe-lm1b/raw/499fb5e6e8fcc8f11c1e6ba93af41dad23f89b06/README.md", "accessed_at": "2026-10-30T12:26:43.782591+00:00"}, {"url": "https://huggingface.co/ArsenyIvanov/emoe-lm1b/raw/499fb5e6e8fcc8f11c1e6ba93af41dad23f89b06/README.md", "read_scope": "Saved card architecture, training and checkpoint description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "ArsenyIvanov/mdlm-lm1b", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Authored E-MoE/MDLM dense and MoE research baselines explicitly trained1Msteps/65.5B LM1B tokens on4H200; new diffusion backbone and controlled training support own scratch-origin lineage.", "provenance_sources": [{"url": "https://huggingface.co/ArsenyIvanov/mdlm-lm1b/raw/b6ffec52c23c345150abe891380701350e7dd5c8/README.md", "accessed_at": "2026-10-30T12:26:43.927297+00:00"}, {"url": "https://huggingface.co/ArsenyIvanov/mdlm-lm1b/raw/b6ffec52c23c345150abe891380701350e7dd5c8/README.md", "read_scope": "Saved card architecture, training and checkpoint description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Zangeeeeerange/causal-byte-lm-shakespeare", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Zangeeeeerange/causal-byte-lm-shakespeare/raw/dfabdcaa8c6cde2b4ff73268116724801fd00e51/README.md", "accessed_at": "2026-10-30T12:26:43.938813+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "KrynexLabs/KrynexAI-1.3-50M-Mobile-Model", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "repackaging", "provenance_evidence_level": "explicit_original_author_attribution", "provenance_basis": "Card credits Pebble(50M) as original author and reports the base model's25B training tokens. Best classification is a republished upstream model; exact uploader transformation is unspecified.", "provenance_sources": [{"url": "https://huggingface.co/KrynexLabs/KrynexAI-1.3-50M-Mobile-Model/raw/225f499a486c9ff1aa1350503542bd131e989eee/README.md"}, {"url": "https://huggingface.co/KrynexLabs/KrynexAI-1.3-50M-Mobile-Model/blob/main/README.md", "access_date": "2026-11-01", "finding": "Saved full card read: original author Pebble(50M),25B training tokens; KrynexMobile packaging. Exact community empty.", "reused": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "DavidLanz/uuu_fine_tune_gpt2", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/DavidLanz/uuu_fine_tune_gpt2/raw/9b278f8da90656ec2e8252813ceac34d4b03ecde/README.md", "accessed_at": "2026-10-30T12:26:44.305291+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Maggio33/gollem-v6-250m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Maggio33/gollem-v6-250m/raw/6dbc5289608855a3b8167f1c3baa0367407608aa/README.md", "accessed_at": "2026-10-30T12:26:44.383362+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "SlayerLab/GoLLeM-v6-250M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/SlayerLab/GoLLeM-v6-250M/raw/77d75a2b48eaab7aec59816f9849cee75a7bdfc1/README.md", "accessed_at": "2026-10-30T12:26:44.422409+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ResonatexIntegratedTechnologies/ResonateX-T1-125M-Talking", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card explicitly states random initialization and no pretrained weights;125.86M,1Btokens,18blocks,2048context.", "provenance_sources": [{"url": "https://huggingface.co/ResonatexIntegratedTechnologies/ResonateX-T1-125M-Talking/raw/a5abd9281ae7a3a9972b73f6daf875b5267ca468/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA A100", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AxiomicLabs/NCN-RW", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/AxiomicLabs/NCN-RW/raw/7b60c89bbf189e7f13726f3f6fe6073ed57118c7/README.md", "accessed_at": "2026-10-30T12:26:44.558095+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "uctnlp/mzansilm-125m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/uctnlp/mzansilm-125m/raw/dd093a0e31ffd74b861cc016b4bdfa744cb7b1af/README.md", "accessed_at": "2026-10-30T12:26:44.682191+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sraivante/LMLM_97M_1", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/sraivante/LMLM_97M_1/raw/0876b3169ba49f0716ca5aed157e2a7d96a7f0cb/README.md", "accessed_at": "2026-10-30T12:26:44.927254+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Tasty-Kiwi/KiwiLM-2", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Tasty-Kiwi/KiwiLM-2/raw/53d6abcb4d11c8a98936c19d55d2b741a99d7602/README.md", "accessed_at": "2026-10-30T12:26:45.002764+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AndrewThompson1233/maba-v1-101m-test", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original101M Maba architecture and official pretrained weights, authored16M-token TinyStories controlled training comparison and linked architecture implementation support own scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/AndrewThompson1233/maba-v1-101m-test/blob/main/README.md"}, {"url": "https://huggingface.co/AndrewThompson1233/maba-v1-101m-test/raw/b4680cb79a487d2dddf06b77c25de9bfbd002542/README.md", "accessed_at": "2026-10-30T12:26:45.059601+00:00"}, {"url": "https://huggingface.co/AndrewThompson1233/maba-v1-101m-test", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "helloboy91/tiny-base-300m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "continued_pretraining_of_own_from_scratch_model", "provenance_evidence_level": "publisher_statement", "provenance_basis": "tiny-base-300m reports from-scratch311.9M,5Btokens,RTX3060Ti8GB; tiny-english-30m from scratchTinyStories,148.91Mprocessedtokens,RTX3060TiBF16 for later stages. Config/history/source preserve no place.", "provenance_sources": [{"url": "https://huggingface.co/helloboy91/tiny-base-300m/raw/4f331d827dd7967f54a352a009419ced6cfc6110/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/helloboy91/tiny-base-300m/raw/4f331d827dd7967f54a352a009419ced6cfc6110/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-base-300m/raw/4f331d827dd7967f54a352a009419ced6cfc6110/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-base-300m/raw/4f331d827dd7967f54a352a009419ced6cfc6110/training_summary.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-base-300m/raw/4f331d827dd7967f54a352a009419ced6cfc6110/tiny_english/model.py", "access_date": "2026-10-31", "scope": "complete frozen file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA GeForce RTX3060Ti,8GB VRAM", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "compute_source": "Explicit card says trained/evaluated locally on a normalPC using RTX3060Ti8GB; physicalcountry unstated.", "creator_contribution": "Own311.9M model/tokenizer and four-stage pretraining lineage:3successive1B-token stages plus2B-token continuation; selected export5,000,265,728 recorded tokens.", "architecture_and_code_lineage": "Own small PyTorch implementation of established causal MHA/RoPE/RMSNorm/SwiGLU decoder recipe, tied embeddings/output. Complete constructor initializes new weights normally withstd0.02. Synthetic/data-source reuse is explicitly attributed; no external pretrained model weights named.", "training_steps_recorded": 15259, "hardware": "NVIDIA GeForce RTX3060Ti,8GB VRAM", "compute_sources": [{"url": "https://huggingface.co/helloboy91/tiny-base-300m/raw/4f331d827dd7967f54a352a009419ced6cfc6110/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-base-300m/raw/4f331d827dd7967f54a352a009419ced6cfc6110/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-base-300m/raw/4f331d827dd7967f54a352a009419ced6cfc6110/training_summary.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-base-300m/raw/4f331d827dd7967f54a352a009419ced6cfc6110/tiny_english/model.py", "access_date": "2026-10-31", "scope": "complete frozen file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "pihu21057w/astrallm", "model_provenance": "unresolved", "own_from_scratch_lineage": "unknown", "trained_status": "unknown", "checkpoint_operation": "not_established", "provenance_evidence_level": "unresolved", "provenance_basis": "Pinned138M AstralLM card/config supply architecture and local/astral-lm naming without a training run or initial-weight statement. Creator separate0.6B fine-tuning issue does not identify this138M frozen model.", "provenance_sources": [{"url": "https://huggingface.co/pihu21057w/astrallm/raw/50eb64c94385ce1d9734d2275f2e7ecbb045076d/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Pinned138M AstralLM card/config supply architecture and local/astral-lm naming without a training run or initial-weight statement. Creator separate0.6B fine-tuning issue does not identify this138M frozen model."}
{"model_id": "helloboy91/tiny-english-30m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "continued_pretraining_of_own_from_scratch_model", "provenance_evidence_level": "publisher_statement", "provenance_basis": "tiny-base-300m reports from-scratch311.9M,5Btokens,RTX3060Ti8GB; tiny-english-30m from scratchTinyStories,148.91Mprocessedtokens,RTX3060TiBF16 for later stages. Config/history/source preserve no place.", "provenance_sources": [{"url": "https://huggingface.co/helloboy91/tiny-english-30m/raw/3a9d10642b07438635e5785316aa332148e1d3c7/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/helloboy91/tiny-english-30m/raw/3a9d10642b07438635e5785316aa332148e1d3c7/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-english-30m/raw/3a9d10642b07438635e5785316aa332148e1d3c7/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-english-30m/raw/3a9d10642b07438635e5785316aa332148e1d3c7/training_summary.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-english-30m/raw/3a9d10642b07438635e5785316aa332148e1d3c7/tiny_english/model.py", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-english-30m/discussions/1", "access_date": "2026-10-31", "finding": "Full actual ownerreply: original release trained on EPYC CPU instance, subsequent RTX3060Ti training and earlierGPU-stage selection."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Original phase: AMD EPYC CPU instance; later selected continuation: NVIDIA RTX3060Ti", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "compute_source": "Direct owner discussion1 confirms original CPUtraining and later RTX3060Ti; card describes laterCUDA BF16. Exact EPYCmodel/provider and devicecountries unstated.", "creator_contribution": "Own29.9M model/tokenizer on prepared TinyStories; originalCPUrun then ownGPU continuation, selected checkpointstage23000 with148,910,080 totalprocessedtokens. Later200M stage overfit and is not selected artifact.", "architecture_and_code_lineage": "Own small PyTorch implementation of established causal MHA/RoPE/RMSNorm/SwiGLU decoder recipe, tied embeddings/output. Complete constructor initializes new weights normally withstd0.02. Synthetic/data-source reuse is explicitly attributed; no external pretrained model weights named.", "training_steps_recorded": 23000, "hardware": "Original phase: AMD EPYC CPU instance; later selected continuation: NVIDIA RTX3060Ti", "compute_sources": [{"url": "https://huggingface.co/helloboy91/tiny-english-30m/raw/3a9d10642b07438635e5785316aa332148e1d3c7/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-english-30m/raw/3a9d10642b07438635e5785316aa332148e1d3c7/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-english-30m/raw/3a9d10642b07438635e5785316aa332148e1d3c7/training_summary.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-english-30m/raw/3a9d10642b07438635e5785316aa332148e1d3c7/tiny_english/model.py", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/helloboy91/tiny-english-30m/discussions/1", "access_date": "2026-10-31", "finding": "Full actual ownerreply: original release trained on EPYC CPU instance, subsequent RTX3060Ti training and earlierGPU-stage selection."}], "evidence_updated_at": "2026-10-31"}
{"model_id": "SLM-Archive/AppleMind-1.0-Mini", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Archive publisher credits another model creator; no new publisher training is established.", "provenance_sources": [{"url": "https://huggingface.co/SLM-Archive/AppleMind-1.0-Mini/raw/ea31d80aee538f62d139fef4aa4e30e535806eb2/README.md", "accessed_at": "2026-10-30T12:26:45.531046+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "SLM-Archive/BananaMind-2-Pro", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Archive publisher credits another model creator; no new publisher training is established.", "provenance_sources": [{"url": "https://huggingface.co/SLM-Archive/BananaMind-2-Pro/raw/d1c77cff6064c96b0995e3b539ea6bc78f0d9d51/README.md", "accessed_at": "2026-10-30T12:26:45.616246+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Vivid86/MiniTransformer-91M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Vivid86/MiniTransformer-91M/raw/703027494f30d81118d5df6a6e59502880221927/README.md", "accessed_at": "2026-10-30T12:26:45.666812+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "VTXAI/vortex-50m", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/VTXAI/vortex-50m/raw/ea749e03ace13618ba09fbe5bf422caabd7e1411/README.md", "accessed_at": "2026-10-30T12:26:45.827717+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Maggio33/GoLLeM-v5-64M-Muon-v1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Maggio33/GoLLeM-v5-64M-Muon-v1/raw/1ed515ce81697898da12ebcf0454df0688f94148/README.md", "accessed_at": "2026-10-30T12:26:45.863836+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AMT-Studio/StellarAI-1-beta-0.05b", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/AMT-Studio/StellarAI-1-beta-0.05b/blob/main/README.md"}, {"url": "https://huggingface.co/AMT-Studio/StellarAI-1-beta-0.05b/raw/9f901d651c4c78aef8f2e08b12ce2bc74cf108e5/README.md", "accessed_at": "2026-10-30T12:26:45.945979+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "CPU; product/count unspecified", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "training_duration_reported": "~3.2 hours", "creator_contribution": "Own small bilingual model training and tokenizer/custom loading code; card says no image-text training yet for vision capability.", "architecture_lineage": "Card explicitly credits GPT-2/Llama/ViT/BLIP-2 architecture inspiration; own4-layer decoder plus CNN/ViT and fusion components.", "hardware": "CPU; product/count unspecified", "compute_sources": [{"url": "https://huggingface.co/AMT-Studio/StellarAI-1-beta-0.05b/raw/9f901d651c4c78aef8f2e08b12ce2bc74cf108e5/README.md", "scope": "Complete frozen card read", "excerpt": "Training time | ~3.2 hours", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "Maggio33/GoLLeM-v5-128M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Maggio33/GoLLeM-v5-128M/raw/aa529f94b7df73822163c8b3e09211af2659ac13/README.md", "accessed_at": "2026-10-30T12:26:46.163083+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Maggio33/GoLLeM-v5-128M-Muon-v1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Maggio33/GoLLeM-v5-128M-Muon-v1/raw/dd081fb9cac3ecfa50f10f3912c91f4e32011c24/README.md", "accessed_at": "2026-10-30T12:26:46.230692+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Maggio33/GoLLeM-v5-64M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Maggio33/GoLLeM-v5-64M/raw/cec0d0b7cc25eb0ab5eabd3f3bbbd3b055119864/README.md", "accessed_at": "2026-10-30T12:26:46.274087+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Maggio33/GoLLeM-v5-32M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Maggio33/GoLLeM-v5-32M/raw/99d084ced0db77a3927afabf42207b84a57ffac8/README.md", "accessed_at": "2026-10-30T12:26:46.457088+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Maggio33/GoLLeM-v5-16M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Maggio33/GoLLeM-v5-16M/raw/04e92a79e597e7771a1bbb9eb335b0d770f44542/README.md", "accessed_at": "2026-10-30T12:26:46.496176+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "alea-institute/kl3m-002-520m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining_with_ntp_sft_cotraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card documents original 273 Ventures clean-corpus Mixtral-architecture 520M training, November2023–January2024 on12 RTX4090, later donated to ALEA. Own original pretraining is attributed to273 Ventures; donation/publisher change is not another training run.", "provenance_sources": [{"url": "https://huggingface.co/alea-institute/kl3m-002-520m/raw/510047500f4eafa43a0463b54d8f768f307f4c36/README.md", "accessed_at": "2026-10-30T12:26:46.553350+00:00"}, {"url": "https://huggingface.co/alea-institute/kl3m-002-520m/raw/510047500f4eafa43a0463b54d8f768f307f4c36/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "BananaMind/TernaryBananaMind-10M", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "quantization_of_own_trained_ternary_model", "provenance_evidence_level": "publisher_training_statement_and_framework_artifacts", "provenance_basis": "Publisher reports own BananaAll ternary training on 2B FineWeb-Edu tokens. Frozen dataset_tokens.json is emitted by linked framework pretraining mode, which constructs a new BananaAll model; config ternary=true. Released file is then explicitly packed from its own preserved model_backup_full.safetensors by included pack_ternary.py. Own from-scratch ancestry is a strong code/artifact inference; packed frozen release is derived from own trained weights.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/TernaryBananaMind-10M/raw/b7cc3740c631301d093caef27a75b8b9d0245a00/README.md"}, {"url": "https://huggingface.co/BananaMind/TernaryBananaMind-10M/raw/b7cc3740c631301d093caef27a75b8b9d0245a00/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://github.com/BananaMind/BananaAll/blob/main/README.md", "access_date": "2026-10-31", "scope": "complete linked training framework README"}, {"url": "https://github.com/BananaMind/BananaAll/blob/main/backend/train.py", "access_date": "2026-10-31", "scope": "pretraining versus fine-tuning model construction and trainer resume sections read"}, {"url": "https://huggingface.co/BananaMind/TernaryBananaMind-10M/raw/b7cc3740c631301d093caef27a75b8b9d0245a00/dataset_tokens.json", "access_date": "2026-10-31", "scope": "complete small frozen training/packing file read"}, {"url": "https://huggingface.co/BananaMind/TernaryBananaMind-10M/raw/b7cc3740c631301d093caef27a75b8b9d0245a00/config.json", "access_date": "2026-10-31", "scope": "complete small frozen training/packing file read"}, {"url": "https://huggingface.co/BananaMind/TernaryBananaMind-10M/raw/b7cc3740c631301d093caef27a75b8b9d0245a00/pack_ternary.py", "access_date": "2026-10-31", "scope": "complete small frozen training/packing file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA RTX PRO 6000 (Molab)", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "training_duration_reported": "2 hours, publisher report.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "Own model_backup_full.safetensors, trained with BananaAll; packed into release by included converter. No external pretrained base identified.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Llama-style decoder Transformer / GQA, QK norm, SwiGLU, RoPE and RMSNorm components", "source_url": "https://huggingface.co/BananaMind/TernaryBananaMind-10M/raw/b7cc3740c631301d093caef27a75b8b9d0245a00/README.md"}], "best_origin_estimate": {"origin": "own_from_scratch_ancestry_then_own_weight_packing", "confidence": "high", "direct_or_inferred": "inferred_from_training_framework_and_saved_mode_artifacts", "basis": "Publisher reports own BananaAll ternary training on 2B FineWeb-Edu tokens. Frozen dataset_tokens.json is emitted by linked framework pretraining mode, which constructs a new BananaAll model; config ternary=true. Released file is then explicitly packed from its own preserved model_backup_full.safetensors by included pack_ternary.py. Own from-scratch ancestry is a strong code/artifact inference; packed frozen release is derived from own trained weights.", "remaining_gap": "No original notebook launch record is packaged."}, "hardware": "NVIDIA RTX PRO 6000 (Molab)", "compute_sources": [{"url": "https://huggingface.co/BananaMind/TernaryBananaMind-10M/raw/b7cc3740c631301d093caef27a75b8b9d0245a00/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://github.com/BananaMind/BananaAll/blob/main/README.md", "access_date": "2026-10-31", "scope": "complete linked training framework README"}, {"url": "https://github.com/BananaMind/BananaAll/blob/main/backend/train.py", "access_date": "2026-10-31", "scope": "pretraining versus fine-tuning model construction and trainer resume sections read"}, {"url": "https://huggingface.co/BananaMind/TernaryBananaMind-10M/raw/b7cc3740c631301d093caef27a75b8b9d0245a00/dataset_tokens.json", "access_date": "2026-10-31", "scope": "complete small frozen training/packing file read"}, {"url": "https://huggingface.co/BananaMind/TernaryBananaMind-10M/raw/b7cc3740c631301d093caef27a75b8b9d0245a00/config.json", "access_date": "2026-10-31", "scope": "complete small frozen training/packing file read"}, {"url": "https://huggingface.co/BananaMind/TernaryBananaMind-10M/raw/b7cc3740c631301d093caef27a75b8b9d0245a00/pack_ternary.py", "access_date": "2026-10-31", "scope": "complete small frozen training/packing file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "Ried-wqs/omega-sft-145m", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "sft_of_own_custom_base", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Release uses creators original legacy OmegaGPT decoder, own model implementation and tokenizer with trained SFT weights. Best-supported reading is post-training of own custom scratch-origin family; exact pretraining corpus/history not stated.", "provenance_sources": [{"url": "https://huggingface.co/Ried-wqs/omega-sft-145m/raw/a544a2cc8116407f12da88468aa894a3cd11160c/README.md", "accessed_at": "2026-10-30T12:26:46.895930+00:00"}, {"url": "https://huggingface.co/Ried-wqs/omega-sft-145m", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_original_custom_model_family", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "OthmaneW/DiffuThink-Story-512", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/OthmaneW/DiffuThink-Story-512/raw/c6f3061229beb2cb13603a10be05ac1d57e72e84/README.md", "accessed_at": "2026-10-30T12:26:46.927569+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Tulifo/chittu-story-v0-26m", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/Tulifo/chittu-story-v0-26m/raw/4f11284f8ebc19348ce2885bfea66eb13c8717cc/README.md", "accessed_at": "2026-10-30T12:26:47.110137+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "omurberaisik/NoTokenLM-Gen-3.6", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/omurberaisik/NoTokenLM-Gen-3.6/raw/e80c100d282b2e30e7c7091e610e4917cc75547f/README.md", "accessed_at": "2026-10-30T12:26:47.123157+00:00", "quote": ". What is this checkpoint, actually? Gen-3.6 is a from-scratch pretraining run of a 13.87M-parameter byte-level transformer (12 layers, dmodel=320). This card describes step 13,000, and every number in it comes from that checkpoint's EMA ", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "seyhachhorn/khmer-food-lm", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/seyhachhorn/khmer-food-lm/raw/39ff7dae259fa9ddfcd9a5343474ad86a4f9c8c0/README.md", "accessed_at": "2026-10-30T12:26:47.181274+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Batman55072/NanoChat-31M-V2", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/Batman55072/NanoChat-31M-V2/raw/9008925948dc3418f1741e648a8e9bc927d0a099/README.md", "accessed_at": "2026-10-30T12:26:47.455021+00:00", "quote": ".4M parameter GPT-style decoder-only transformer, trained fully from scratch, that generates simulated live-stream chat comments -- both regular fast-paced \"hype chat\" and rarer, longer \"superchat\"-style paid messages/questions. Built t", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sraivante/Custom-GPT-40M-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sraivante/Custom-GPT-40M-Base/raw/87842ae2297d19730ef520863082ca85fce0b510/README.md", "accessed_at": "2026-10-30T12:26:47.532712+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Melih1234/ANKA-50M-RMW3", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Melih1234/ANKA-50M-RMW3/raw/5176386744d68869bd2c033d2a2a87ff4e51f7e2/README.md", "accessed_at": "2026-10-30T12:26:47.556729+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "StrongDev2024/mini", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/StrongDev2024/mini/raw/d3166c0514fa56bcea7d6f29cc367326f0fe4b30/README.md", "accessed_at": "2026-10-30T12:26:47.753758+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Circuits-V2/Meridian-Tiny", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/Circuits-V2/Meridian-Tiny/raw/7c2347d2b791a5ebbed6167bb00dbd9c4caabc2a/README.md"}, {"url": "https://huggingface.co/Circuits-V2/Meridian-Tiny/raw/7c2347d2b791a5ebbed6167bb00dbd9c4caabc2a/README.md", "access_date": "2026-10-31", "scope": "complete card"}, {"url": "https://huggingface.co/Circuits-V2/Meridian-Tiny/raw/7c2347d2b791a5ebbed6167bb00dbd9c4caabc2a/config.json", "access_date": "2026-10-31", "scope": "complete config"}, {"url": "https://huggingface.co/Circuits-V2/Meridian-Tiny/raw/7c2347d2b791a5ebbed6167bb00dbd9c4caabc2a/presets.py", "access_date": "2026-10-31", "scope": "complete presets"}, {"url": "https://huggingface.co/Circuits-V2/Meridian-Tiny/raw/7c2347d2b791a5ebbed6167bb00dbd9c4caabc2a/model.py", "access_date": "2026-10-31", "scope": "imports, component declarations and complete MiniMeridian initialization/forward path; diagnostic source examples, not executed"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA GeForce RTX 3060 12GB", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "compute_source": "Publisher card says trained on single RTX 3060; public org describes two people working on consumer GPUs. Ownership/provider not explicitly stated.", "creator_contribution": "hashtagg1 is explicitly credited for the Meridian series in publisher org card. Own tokenizer/data/model/training implementation reported; own integration mHC with AttnRes and simplified Engram plus component ablations. Saved constructor initializes new weights with normal std 0.02. No external pretrained checkpoint established.", "architecture_and_code_lineage": "Explicitly attributed Moonshot Kimi KDA/attention residuals, DeepSeek MLA/routing/mHC/Engram and Qwen output gate. Card claims own interpretation and integration, not official-code reproduction; optional actual imported fast kernel fla.ops.kda.chunk_kda from flash-linear-attention, with own recurrent fallback.", "hardware": "1 x NVIDIA GeForce RTX 3060 12GB", "compute_sources": [{"url": "https://huggingface.co/Circuits-V2/Meridian-Tiny/raw/7c2347d2b791a5ebbed6167bb00dbd9c4caabc2a/README.md", "access_date": "2026-10-31", "scope": "complete card"}, {"url": "https://huggingface.co/Circuits-V2/Meridian-Tiny/raw/7c2347d2b791a5ebbed6167bb00dbd9c4caabc2a/config.json", "access_date": "2026-10-31", "scope": "complete config"}, {"url": "https://huggingface.co/Circuits-V2/Meridian-Tiny/raw/7c2347d2b791a5ebbed6167bb00dbd9c4caabc2a/presets.py", "access_date": "2026-10-31", "scope": "complete presets"}, {"url": "https://huggingface.co/Circuits-V2/Meridian-Tiny/raw/7c2347d2b791a5ebbed6167bb00dbd9c4caabc2a/model.py", "access_date": "2026-10-31", "scope": "imports, component declarations and complete MiniMeridian initialization/forward path; diagnostic source examples, not executed"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "FWKV/Myosotis-1-base", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/FWKV/Myosotis-1-base/raw/392bfa209da18fd749f257740359e8e008067d7b/README.md", "accessed_at": "2026-10-30T12:26:47.809322+00:00", "quote": "g training. 4. Loss Masking for Conversational AI Fine-tuned exclusively on multi-turn chat data, the loss is masked so that only assistant tokens contribute to the gradient. This forces the model to master response generation while treating user prompts p", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "superAVTR/tinizong-46M_v1.1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "export_of_own_scratch_checkpoint", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Owner describes third native experimental model built to understand full LLM stack instead of fine-tuning existing open models, and reports own 2.2B Wikipedia plus20M synthetic data training. Card documents mapping own native PyTorch checkpoint to LlamaForCausalLM with numerical parity.", "provenance_sources": [{"url": "https://huggingface.co/superAVTR/tinizong-46M_v1.1/raw/7c955a3fa52dfb72f6bb6ee13cecdb2962711300/README.md", "accessed_at": "2026-10-30T12:26:48.077454+00:00"}, {"url": "https://huggingface.co/superAVTR/tinizong-46M_v1.1/raw/7c955a3fa52dfb72f6bb6ee13cecdb2962711300/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/superAVTR/tinizong-46M_v1.1/discussions/1", "scope": "All public comments including full creator reply read", "excerpt": "my background is in sys-admin/cloud/general development/systems security", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "creator_contribution": "Native approximately46M language-model pretraining, own synthetic factCard data design and exact-parity HF export. Source developed with ChatGPT and creator code checking with Claude/Kimi, per owner.", "provenance_confidence": "high", "architecture_lineage": "Own native PyTorch implementation of standard Llama-compatible decoder (RoPE/RMSNorm/SwiGLU/MHA), exported to upstream Transformers LlamaForCausalLM; own16K SentencePiece tokenizer.", "evidence_updated_at": "2026-11-01"}
{"model_id": "FlyinGodzilla/AMOR-GatedDeltaNet-440M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly states the recurrent backbone and attention blocks were pretrained jointly from scratch. Optional distilled deployment routers are separate from the pretrained checkpoint.", "provenance_sources": [{"url": "https://huggingface.co/FlyinGodzilla/AMOR-GatedDeltaNet-440M/raw/de797220a4df4fc5cb98a31e168b29f6ed6fed95/README.md", "accessed_at": "2026-10-30T12:26:48.127555+00:00", "quote": ". The recurrent backbone and attention blocks are pretrained jointly from scratch. “Post-hoc” describes where the attention blocks sit, after the complete backbone. Keys and values come from residual-stream projections and are retained at ev", "basis": "Card explicitly states the recurrent backbone and attention blocks were pretrained jointly from scratch. Optional distilled deployment routers are separate from the pretrained checkpoint."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "FlyinGodzilla/AMOR-Mamba2-440M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly states the recurrent backbone and attention blocks were pretrained jointly from scratch. Optional distilled deployment routers are separate from the pretrained checkpoint.", "provenance_sources": [{"url": "https://huggingface.co/FlyinGodzilla/AMOR-Mamba2-440M/raw/317d316ea94b096ed63c8c0bf7d37b60a6018bce/README.md", "accessed_at": "2026-10-30T12:26:48.191265+00:00", "quote": ". The recurrent backbone and attention blocks are pretrained jointly from scratch. “Post-hoc” describes where the attention blocks sit, after the complete backbone. Keys and values come from residual-stream projections and are retained at ev", "basis": "Card explicitly states the recurrent backbone and attention blocks were pretrained jointly from scratch. Optional distilled deployment routers are separate from the pretrained checkpoint."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "FlyinGodzilla/AMOR-GatedDeltaNet-180M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly states the recurrent backbone and attention blocks were pretrained jointly from scratch. Optional distilled deployment routers are separate from the pretrained checkpoint.", "provenance_sources": [{"url": "https://huggingface.co/FlyinGodzilla/AMOR-GatedDeltaNet-180M/raw/9e348a1cd40058cbff091a2e61f5672943b83ecd/README.md", "accessed_at": "2026-10-30T12:26:48.380066+00:00", "quote": ". The recurrent backbone and attention blocks are pretrained jointly from scratch. “Post-hoc” describes where the attention blocks sit, after the complete backbone. Keys and values come from residual-stream projections and are retained at ev", "basis": "Card explicitly states the recurrent backbone and attention blocks were pretrained jointly from scratch. Optional distilled deployment routers are separate from the pretrained checkpoint."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "FlyinGodzilla/AMOR-Mamba2-180M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly states the recurrent backbone and attention blocks were pretrained jointly from scratch. Optional distilled deployment routers are separate from the pretrained checkpoint.", "provenance_sources": [{"url": "https://huggingface.co/FlyinGodzilla/AMOR-Mamba2-180M/raw/846351a2b48244553cb7ee59236f710a329e1f11/README.md", "accessed_at": "2026-10-30T12:26:48.383070+00:00", "quote": ". The recurrent backbone and attention blocks are pretrained jointly from scratch. “Post-hoc” describes where the attention blocks sit, after the complete backbone. Keys and values come from residual-stream projections and are retained at ev", "basis": "Card explicitly states the recurrent backbone and attention blocks were pretrained jointly from scratch. Optional distilled deployment routers are separate from the pretrained checkpoint."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "tetracta/tetracta-van-1b", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/tetracta/tetracta-van-1b/raw/741b3f4202538fde66d48371868c20cf5a610001/README.md", "accessed_at": "2026-10-30T12:26:48.442211+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "pratikshourabh/indAI", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned training_metadata.json explicitly names base_model=gpt2-medium, epoch4,global_step8438 and recorded train/validation loss. This resolves the generic card placeholders.", "provenance_sources": [{"url": "https://huggingface.co/pratikshourabh/indAI/raw/34c1d1b14aeff76c78d79c74ee5c7f2b48bb9512/training_metadata.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "upstream_checkpoint": "gpt2-medium", "training_steps_recorded": 8438}
{"model_id": "SOYORiN/Susu", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining_and_sft", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Custom10.2M architecture and own6144-token BPE, original language pretraining on LCCC/XDailyDialog and later teacher-response SFT support own scratch-origin lineage. Teacher responses supply supervision, not imported base weights.", "provenance_sources": [{"url": "https://huggingface.co/SOYORiN/Susu/raw/a071dd25d289cbe406f4526c51c59fcb83ff9768/README.md", "accessed_at": "2026-10-30T12:26:48.756817+00:00"}, {"url": "https://huggingface.co/SOYORiN/Susu", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Mspanz88/BananaMind-2-Nano", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card attributes BananaMind family trained from scratch:9,968,128parameters,30Btokens,55,485steps,8kvocab. Config names BananaMind/BananaMind-2-Nano; upload history includes Banaxi-Tech and Mspanz88. Original family from-scratch statement is not evidence Marc Spaniol trained it.", "provenance_sources": [{"url": "https://huggingface.co/Mspanz88/BananaMind-2-Nano/blob/main/README.md"}, {"url": "https://huggingface.co/Mspanz88/BananaMind-2-Nano/raw/30899954a05743b2523ccd1f7971b906c3037162/README.md", "accessed_at": "2026-10-30T12:26:48.846444+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Mirror publication of BananaMind model; original builder attribution remains BananaMind.", "architecture_lineage": "BananaMind-2-Nano architecture; card explicitly credits AxiomicLabs and GPT X2 125M inspiration.", "evidence_updated_at": "2026-11-01"}
{"model_id": "AvinashRicky/AViGPT", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/AvinashRicky/AViGPT/raw/5788d8957ca0b561b2e0e9c39d4e71020a7ccbdb/README.md", "accessed_at": "2026-10-30T12:26:48.999523+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Aobangaming/lightning-30m-ft", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Selected card names own Lightning/Aoban base, describes initial large-data training then smaller-data fine-tuning of own tiny4-layer256-width model. Own scratch ancestry is best-supported medium inference; earliest initialization source not supplied.", "provenance_sources": [{"url": "https://huggingface.co/Aobangaming/lightning-30m-ft/blob/main/README.md"}, {"url": "https://huggingface.co/Aobangaming/lightning-30m-ft/raw/a0fb96120e352569d92a6fc683b63de9ed9d972e/README.md", "accessed_at": "2026-10-30T12:26:48.989008+00:00"}], "uk_build_status": "no", "builder_location_country": "Thailand", "builder_location_basis": "Own local-PC training explicitly in Asia, September release; same verified creator channel identifies Thailand in October. Medium nearby-date base estimate.", "builder_location_evidence_type": "inferred", "builder_location_period": "September2026 training/release; channel country observed31October2026", "compute_used": "NVIDIA RTX3050 6GB, own computer; Windows11 and Intel i5-10400 context", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "training_duration_reported": "1.5 hours (publisher environmental/training record)", "compute_source": "Own computer", "creator_contribution": "Own Lightning/Aoban development and Booksum fine-tuning; both selected IDs cite the same project/training description and are not assumed independent training runs.", "provenance_confidence": "medium", "architecture_lineage": "Own small standard autoregressive Transformer,4 layers/width256/4-head MHA and FlashAttention;50K vocabulary.", "compute_physical_region": "Asia", "hardware": "NVIDIA RTX3050 6GB, own computer; Windows11 and Intel i5-10400 context", "builder_location_at_training": {"country": "Thailand", "confidence": "medium", "direct_or_inferred": "inferred", "period": "September2026", "basis": "Local own-PC Asia training and same creator Thailand country field October2026."}, "compute_sources": [{"url": "https://huggingface.co/Aobangaming/lightning-30m-ft/raw/a0fb96120e352569d92a6fc683b63de9ed9d972e/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}, {"url": "https://www.youtube.com/@AobanZ/about", "access_date": "2026-10-31", "scope": "Valid prior actual rendered About country reading reused"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "toonist/AnuLM-Base-400M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/toonist/AnuLM-Base-400M/raw/617d50b7cdc684914e21ec98254717d982231f1a/README.md", "accessed_at": "2026-10-30T12:26:49.102731+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "toonist/AnuLM-Coder-400M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/toonist/AnuLM-Coder-400M/raw/1892298baa78e4fcfbdf587928920d62a3cf85a5/README.md", "accessed_at": "2026-10-30T12:26:49.307885+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "toonist/AnuLM-Smoke-30B", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/toonist/AnuLM-Smoke-30B/raw/45d7d487a476b85a210499bbba7ae9c4563a3cfb/README.md", "accessed_at": "2026-10-30T12:26:49.371173+00:00", "quote": "rrors=\"replace\", which sample.py does. Provenance Trained 2026-09-18 from scratch on one machine, on public-domain text, by the pipeline in the repository. No Sarvam weights are used or redistributed; AnuLM is an independent re-implementatio", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Nebulixlabs/Nutral-GPT2-16M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "From-scratch200M FineWeb tokens followed by instruction fine-tuning. Actual config model_config.json:4layers4heads256hidden256context50304vocab,biasfalse.", "provenance_sources": [{"url": "https://huggingface.co/Nebulixlabs/Nutral-GPT2-16M/blob/main/README.md"}, {"url": "https://huggingface.co/Nebulixlabs/Nutral-GPT2-16M/raw/83ca7e3107d2afc530055b455a40432700e3e783/README.md", "accessed_at": "2026-10-30T12:26:49.441072+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"gpus": "2×NVIDIA T4", "provider": "Kaggle", "precision": "FP16 DDP", "physical_country": null}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "BananaMind/BananaMind-2-Pro", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro/raw/846592429ab5f01390f6fa7340921823603c3c3a/README.md"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro/raw/846592429ab5f01390f6fa7340921823603c3c3a/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind", "access_date": "2026-10-31", "scope": "actual rendered organisation card", "finding": "All BananaMind 2 models described as trained from scratch on a single RTX 5070 Ti; shared GQA/QK norm, SwiGLU, RoPE, RMSNorm recipe. Model-specific eight-GPU Unified report takes precedence over series generalisation."}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro/raw/846592429ab5f01390f6fa7340921823603c3c3a/LICENSE", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro/raw/846592429ab5f01390f6fa7340921823603c3c3a/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}, {"url": "https://huggingface.co/posts/Banaxi-Tech/835972853024844", "access_date": "2026-10-31", "scope": "actual rendered creator release announcement", "excerpt": "The training took 22 days on one RTX 5070 Ti."}, {"url": "https://huggingface.co/posts/Banaxi-Tech/678789079412959", "access_date": "2026-10-31", "event_date": "2026-08-02", "scope": "actual rendered full creator announcement", "excerpt": "August 3, 1PM in Austria time", "finding": "Explicit Austrian release-time reference during the local Pro pretraining run; combined with own Austrian-grid local-system power estimate supports an inferred Austrian home/training base."}], "uk_build_status": "unknown", "builder_location_country": "Austria", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA RTX 5070 Ti", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Austrian builder-base estimate is inferred; no literal residence or physical compute-country statement. University/institutional route unstated.", "training_duration_reported": "22 days, direct creator release post; card separately derives 529.72 hours from token throughput.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Llama-style decoder Transformer / GQA, QK norm, SwiGLU, RoPE and RMSNorm components", "source_url": "https://huggingface.co/BananaMind/BananaMind-2-Pro/raw/846592429ab5f01390f6fa7340921823603c3c3a/README.md"}], "hardware": "1 x NVIDIA RTX 5070 Ti", "builder_location_at_training": {"country": "Austria", "city": null, "confidence": "medium", "direct_or_inferred": "inferred", "period": "July-August 2026 local Pro pretraining run", "basis": "Creator says Pro trains locally, schedules release in Austria time, and calculates own measured-system emissions with Austrian grid intensity. October Italian vacation is separately documented.", "sources": ["https://huggingface.co/posts/Banaxi-Tech/678789079412959", "https://huggingface.co/BananaMind/BananaMind-2-Pro/raw/846592429ab5f01390f6fa7340921823603c3c3a/README.md", "https://huggingface.co/BananaMind/BananaMind-2-SLMoE/discussions/2"]}, "compute_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro/raw/846592429ab5f01390f6fa7340921823603c3c3a/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind", "access_date": "2026-10-31", "scope": "actual rendered organisation card", "finding": "All BananaMind 2 models described as trained from scratch on a single RTX 5070 Ti; shared GQA/QK norm, SwiGLU, RoPE, RMSNorm recipe. Model-specific eight-GPU Unified report takes precedence over series generalisation."}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro/raw/846592429ab5f01390f6fa7340921823603c3c3a/LICENSE", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro/raw/846592429ab5f01390f6fa7340921823603c3c3a/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}, {"url": "https://huggingface.co/posts/Banaxi-Tech/835972853024844", "access_date": "2026-10-31", "scope": "actual rendered creator release announcement", "excerpt": "The training took 22 days on one RTX 5070 Ti."}, {"url": "https://huggingface.co/posts/Banaxi-Tech/678789079412959", "access_date": "2026-10-31", "event_date": "2026-08-02", "scope": "actual rendered full creator announcement", "excerpt": "August 3, 1PM in Austria time", "finding": "Explicit Austrian release-time reference during the local Pro pretraining run; combined with own Austrian-grid local-system power estimate supports an inferred Austrian home/training base."}], "evidence_updated_at": "2026-10-31"}
{"model_id": "sovythos/Sovythos-66M-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sovythos/Sovythos-66M-Base/raw/3d267521771488200a2c311fb7b1b45ac4ce6413/README.md", "accessed_at": "2026-10-30T12:26:49.622813+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Tirth2109/legal-slm", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Tirth2109/legal-slm/raw/5b804aa43cd640542be25e6fa009326f2b52b117/README.md", "accessed_at": "2026-10-30T12:26:49.752938+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "SonexaAI/Sonexa-Beta", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_projection_edit_of_own_scratch_model", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card says custom140.9M bilingual MoE built from scratch and trained on no_robots/Russian dialogue, then calibrated through output-projection damping. Own scratch backbone is supported; selected weight-edit/calibration is kept separate from original model training.", "provenance_sources": [{"url": "https://huggingface.co/SonexaAI/Sonexa-Beta/raw/6a1ef1f20dfb645e62f175bd0a88c2b20cb47228/README.md", "accessed_at": "2026-10-30T12:26:49.923161+00:00"}, {"url": "https://huggingface.co/SonexaAI/Sonexa-Beta/raw/6a1ef1f20dfb645e62f175bd0a88c2b20cb47228/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "sivasub987/iso20022-extract-53m", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly trains the53M extraction transformer from scratch, then trains its confidence mechanism in a second stage on its own generated right/wrong outcomes. Own original lineage is preserved through post-training.", "provenance_sources": [{"url": "https://huggingface.co/sivasub987/iso20022-extract-53m/raw/554fe490338b98640cfb11cf3126623da5b95851/README.md", "accessed_at": "2026-10-30T12:26:49.979204+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "notnotsamuel/LFM2.5-350M-RLCD", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly attributes the released pretrained weights to an upstream model and describes unchanged redistribution, rather than publisher training.", "provenance_sources": [{"url": "https://huggingface.co/notnotsamuel/LFM2.5-350M-RLCD/raw/deb589d803d141cabd158ef55f6617b128529f36/README.md", "accessed_at": "2026-10-30T12:26:50.053319+00:00", "quote": "epository includes code, benchmark results, and a byte-for-byte copy of the original model weights and tokenizer. Engine below continues to load the pinned original model for reproducibility. Performance: 28 fields One synthetic input with 28 boolean fields,", "basis": "Card explicitly attributes the released pretrained weights to an upstream model and describes unchanged redistribution, rather than publisher training."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "fromziro/ZeroS-Qana-5M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Best-supported own from-scratch pretraining: custom ACSwiGLU architecture and dimensions, newly constructed modules/post_init, complete single-session40000-update recipe and trained checkpoint record. Only borrowed tokenizer is identified; no external weight base. Earliest initialization receipt/literal scratch sentence absent, so origin remains explicitly inferred.", "provenance_sources": [{"url": "https://huggingface.co/fromziro/ZeroS-Qana-5M/raw/bbad48298493a2e290488613b0c7c2a94b4fb3d4/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/fromziro/ZeroS-Qana-5M/raw/bbad48298493a2e290488613b0c7c2a94b4fb3d4/README.md", "access_date": "2026-10-31", "scope": "complete frozen card"}, {"url": "https://huggingface.co/fromziro/ZeroS-Qana-5M/raw/bbad48298493a2e290488613b0c7c2a94b4fb3d4/LICENSE", "access_date": "2026-10-31", "scope": "complete authorship and original personal release link"}, {"url": "https://huggingface.co/fromziro/ZeroS-Qana-5M/raw/bbad48298493a2e290488613b0c7c2a94b4fb3d4/config.json", "access_date": "2026-10-31", "scope": "complete selected architecture/tokenizer config"}, {"url": "https://huggingface.co/fromziro/ZeroS-Qana-5M/raw/bbad48298493a2e290488613b0c7c2a94b4fb3d4/training_config.json", "access_date": "2026-10-31", "scope": "complete training-session configuration"}, {"url": "https://huggingface.co/fromziro/ZeroS-Qana-5M/raw/bbad48298493a2e290488613b0c7c2a94b4fb3d4/modeling_ac_swiglu.py", "access_date": "2026-10-31", "scope": "component declarations/header, custom channel-mixing initialization and complete final model constructor/tying path"}, {"url": "https://huggingface.co/api/models/User01110/Qana-mini-5M", "access_date": "2026-10-31", "finding": "Old personal path returns canonical id fromziro/ZeroS-Qana-5M and identical createdAt/currentSHA, indicating same transferred/renamed repository, not an extra model or independent mirror. All exposed file blobs/hash metadata are identical."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "native BF16; GPU not stated", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "compute_source": "Native BF16 compute stated; product/count/provider unstated. Recipe tpu524k is tokens per update, not evidence of a TPU device.", "creator_contribution": "User01110 explicitly led, developed, trained, evaluated and released; additional project authors PaulCourneya and JonathonLY credited. Model-specific custom channel-mixing integration and actual40000-update run.", "best_origin_estimate": {"origin": "own_from_scratch_backbone", "confidence": "medium", "direct_or_inferred": "inferred", "basis": "Own custom channel-mixing graph/dimensions; fresh model constructor/post_init; single40000-update training-session recipe and completed checkpoint report; only external tokenizer named.", "remaining_gap": "No first-step initialization receipt or explicit from-scratch sentence in public frozen package.", "source_urls": ["https://huggingface.co/fromziro/ZeroS-Qana-5M/raw/bbad48298493a2e290488613b0c7c2a94b4fb3d4/README.md", "https://huggingface.co/fromziro/ZeroS-Qana-5M/raw/bbad48298493a2e290488613b0c7c2a94b4fb3d4/LICENSE", "https://huggingface.co/fromziro/ZeroS-Qana-5M/raw/bbad48298493a2e290488613b0c7c2a94b4fb3d4/config.json", "https://huggingface.co/fromziro/ZeroS-Qana-5M/raw/bbad48298493a2e290488613b0c7c2a94b4fb3d4/training_config.json", "https://huggingface.co/fromziro/ZeroS-Qana-5M/raw/bbad48298493a2e290488613b0c7c2a94b4fb3d4/modeling_ac_swiglu.py", "https://huggingface.co/api/models/User01110/Qana-mini-5M"]}, "provenance_confidence": "medium", "architecture_and_code_lineage": "Standard causal GQA plus own content-dependent channel mixing within expanded SwiGLU; custom PyTorch/Transformers implementation. Borrowed AxiomicLabs/GPT-S-5M tokenizer is explicitly identified with revision, not borrowed model weights.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-10-31"}
{"model_id": "specklabs/Speck2-140M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/specklabs/Speck2-140M/raw/9c3392de3b6da6ed23b4fccb3a4dd1b325f0d8f5/README.md", "accessed_at": "2026-10-30T12:26:50.236265+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 NVIDIA RTX 5090", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "training_duration_reported": "41.40 hours", "compute_source": "Rented through Vast.ai", "creator_contribution": "Own 140.65M scratch pretraining and curriculum/design implementation; base release, no instruction tuning.", "architecture_lineage": "Own hybrid of 8 global grouped-query attention and 10 gated causal-convolution blocks with SwiGLU/RoPE/RMSNorm; pretrained Mistral v0.1 SentencePiece tokenizer reused, not Mistral model weights.", "hardware": "1 NVIDIA RTX 5090", "compute_sources": [{"url": "https://huggingface.co/specklabs/Speck2-140M/raw/9c3392de3b6da6ed23b4fccb3a4dd1b325f0d8f5/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "specklabs/Speck1.5-140M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/specklabs/Speck1.5-140M/raw/f6770c6b620c31a7d25596c5f918eda745010780/README.md", "accessed_at": "2026-10-30T12:26:50.358790+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "training_duration_reported": "25.31 hours", "creator_contribution": "Own 140.65M scratch pretraining and curriculum/design implementation; base release, no instruction tuning.", "architecture_lineage": "Own hybrid of 8 global grouped-query attention and 10 gated causal-convolution blocks with SwiGLU/RoPE/RMSNorm; pretrained Mistral v0.1 SentencePiece tokenizer reused, not Mistral model weights.", "training_hardware_estimate": {"device": "NVIDIA RTX 3090", "confidence": "medium", "direct_or_inferred": "inferred_available_local_device", "basis": "Same creator project directly measures local RTX 3090 inference, while selected 25-hour training recipe does not separately name device."}, "compute_sources": [{"url": "https://huggingface.co/specklabs/Speck1.5-140M/raw/f6770c6b620c31a7d25596c5f918eda745010780/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "specklabs/Speck1-140M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/specklabs/Speck1-140M/raw/7e512faea3af59a401ab083bca3f4a7c6b0f5fc2/README.md", "accessed_at": "2026-10-30T12:26:50.518856+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "training_duration_reported": "25.35 hours", "creator_contribution": "Own 140.65M scratch pretraining and curriculum/design implementation; base release, no instruction tuning.", "architecture_lineage": "Own hybrid of 8 global grouped-query attention and 10 gated causal-convolution blocks with SwiGLU/RoPE/RMSNorm; pretrained Mistral v0.1 SentencePiece tokenizer reused, not Mistral model weights.", "training_hardware_estimate": {"device": "NVIDIA RTX 3090", "confidence": "medium", "direct_or_inferred": "inferred_available_local_device", "basis": "Same creator project directly measures local RTX 3090 inference, while selected 25-hour training recipe does not separately name device."}, "compute_sources": [{"url": "https://huggingface.co/specklabs/Speck1-140M/raw/7e512faea3af59a401ab083bca3f4a7c6b0f5fc2/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "AwakeningOS/VISTA-24M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "explicit_model_card_training_recipe", "provenance_basis": "Creator's detailed training recipe explicitly gives causal next-token prediction from scratch and completed BabyLM100M-word run; default80M raw checkpoint.", "provenance_sources": [{"url": "https://huggingface.co/AwakeningOS/VISTA-24M/blob/main/README.md"}, {"url": "https://huggingface.co/AwakeningOS/VISTA-24M/raw/9287d39faf875a15faba017877f57a81b89370ae/README.md", "accessed_at": "2026-10-30T12:26:50.619017+00:00"}, {"url": "https://huggingface.co/AwakeningOS/VISTA-24M/blob/main/README.md", "access_date": "2026-11-01", "finding": "Saved full card read: Yusuke Maeda independent experiment; objective causal next-token prediction from scratch; BabyLM Strict-Small10M corpus,100M total words; oneRTX3090 in author's local PC; default raw80M checkpoint.", "reused": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "OPENGCM/GTM-3-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact card states from scratch; new scratch training.", "provenance_sources": [{"url": "https://huggingface.co/OPENGCM/GTM-3-base/raw/89327c31d52dc5089b3ac6d132b54bacfe706c7c/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "single NVIDIA RTX Pro6000", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "vitin12/gpt-pt-270m-base-1b", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/vitin12/gpt-pt-270m-base-1b/raw/7ca6058540eded97c4ae2e3e938388a5d3bba8f9/README.md", "accessed_at": "2026-10-30T12:26:50.882476+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "OPENGCM/GTM-v2-chat", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "explicit_own_base_lineage", "provenance_basis": "Card explicitly states this is the same publisher's scratch-pretrained GTM-v2-base continued with UltraChat SFT. Released operation remains fine-tuning and original lineage is own from-scratch.", "provenance_sources": [{"url": "https://huggingface.co/OPENGCM/GTM-v2-chat/raw/2183247ea091961215541689614db1ac4c5654f4/README.md"}, {"url": "https://huggingface.co/OPENGCM/GTM-v2-chat/blob/main/README.md", "access_date": "2026-11-01", "finding": "Saved full card read: same GTM-v2-base weights, original base pretraining from scratch then UltraChat SFT, singleRTXPro6000.", "reused": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "single NVIDIA RTX Pro6000", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "CH3NDev/Nool-Alpha-100M-Chat", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original Nool hybrid GSLA/HFK-MoE architecture, authored implementation and explicit900-step checkpoint/loss support trained own scratch-origin model.", "provenance_sources": [{"url": "https://huggingface.co/CH3NDev/Nool-Alpha-100M-Chat/raw/b5302b5e2cdd4e71c0675f918379f43aa268da63/README.md", "accessed_at": "2026-10-30T12:26:50.973830+00:00"}, {"url": "https://huggingface.co/CH3NDev/Nool-Alpha-100M-Chat/raw/b5302b5e2cdd4e71c0675f918379f43aa268da63/README.md", "read_scope": "Cached card model/architecture/training/result and upstream description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Voilum/voilum-1", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "sft_and_grpo_of_own_base", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author Elise Paul reports original463.6M sparse MoE,12B-token base pretraining, followed byChatMLSFT and execution-groundedGRPO. Supports post-training of original scratch lineage; preserve Elise Paul/elisepaul model attribution.", "provenance_sources": [{"url": "https://huggingface.co/Voilum/voilum-1/raw/f9da539c05b96bb70db464cae880a07babd9a7f7/README.md", "accessed_at": "2026-10-30T12:26:51.124195+00:00"}, {"url": "https://huggingface.co/Voilum/voilum-1/raw/f9da539c05b96bb70db464cae880a07babd9a7f7/README.md", "read_scope": "Saved card architecture/base/training/result and upstream attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "OPENGCM/GTM-v2-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact card states from scratch; 4.16B tokens.", "provenance_sources": [{"url": "https://huggingface.co/OPENGCM/GTM-v2-base/raw/1e76ace314b87441ead845c706d83e5a85c6eb01/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "single NVIDIA RTX Pro6000", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "tdizhere/quotes-gpt", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Custom10.9M character-level GPT,94-symbol vocabulary and11000-step Colby/quotes pretraining with own model-class/vocab support own scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/tdizhere/quotes-gpt/raw/a0cd602c16ab529f68f6b0a10883253af930b692/README.md", "accessed_at": "2026-10-30T12:26:51.329714+00:00"}, {"url": "https://huggingface.co/tdizhere/quotes-gpt", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "OPENGCM/GTM-v1-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact card states from scratch; 3B tokens.", "provenance_sources": [{"url": "https://huggingface.co/OPENGCM/GTM-v1-base/raw/69301ffb6050783450afa4d8257347400971600b/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "single NVIDIA RTX Pro6000", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ev1lkow/token-garden-story-28m-v1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "continued_pretraining_of_own_from_scratch_model", "provenance_evidence_level": "publisher_statement", "provenance_basis": "token-garden-story-28m-v1 reports from-scratch27.8M TinyStories pretraining with Apple MPS/BF16; no exact device or country. Config/commits retain handle.", "provenance_sources": [{"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/README.md", "access_date": "2026-10-31", "scope": "complete frozen card"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/NOTICE", "access_date": "2026-10-31", "scope": "complete provenance notice"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/config.json", "access_date": "2026-10-31", "scope": "complete frozen configuration"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/training_metrics.json", "access_date": "2026-10-31", "scope": "complete selected-lineage training metrics"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/release_manifest.json", "access_date": "2026-10-31", "scope": "complete release manifest"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/verification.json", "access_date": "2026-10-31", "scope": "environment, scope and limitations only: recorded CPU parity environment, not training hardware"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/story_model.py", "access_date": "2026-10-31", "scope": "full module header/architecture reference and complete fresh initializer; no execution"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Apple GPU via PyTorch MPS; exact product/count not stated", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Exact Apple product, duration, physical compute country and model-time builder base unstated. CPU/Darwinarm64 parity environment is inference verification, not training-device identification.", "creator_contribution": "Own27.8M model/tokenizer and successive same-project pretraining runs starting random; selected49k-update lineage ends with6k-update state continuation. PublicexportFP32Safetensors, no new training for export.", "training_tokens_recorded": {"physical_target_exposures": 630363511, "nonzero_loss_target_exposures": 609561922, "zero_weight_exposures": 20801589, "note": "Repeated target exposures, not unique tokens. Zero-weight SimpleStories targets are not positive-weight training data."}, "architecture_and_code_lineage": "Own plain-PyTorch implementation of established RMSNorm/RoPE/SwiGLU/SDPA decoder recipe, with references explicitly in story_model.py. NOTICE explicitly says adapted from own TokenGarden code rather than upstream TinyStories model implementation. TinyStories synthetic data reused, not pretrained TinyStories weights.", "hardware": "Apple GPU via PyTorch MPS; exact product/count not stated", "compute_sources": [{"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/README.md", "access_date": "2026-10-31", "scope": "complete frozen card"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/NOTICE", "access_date": "2026-10-31", "scope": "complete provenance notice"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/config.json", "access_date": "2026-10-31", "scope": "complete frozen configuration"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/training_metrics.json", "access_date": "2026-10-31", "scope": "complete selected-lineage training metrics"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/release_manifest.json", "access_date": "2026-10-31", "scope": "complete release manifest"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/verification.json", "access_date": "2026-10-31", "scope": "environment, scope and limitations only: recorded CPU parity environment, not training hardware"}, {"url": "https://huggingface.co/ev1lkow/token-garden-story-28m-v1/raw/084d93b11020399a7aeb7b021facfd054ad250eb/story_model.py", "access_date": "2026-10-31", "scope": "full module header/architecture reference and complete fresh initializer; no execution"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "my0919175/Sovythos-66m-instruct", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/my0919175/Sovythos-66m-instruct/raw/5b8905f912ab785161b8d66689988e5f87afc420/README.md", "accessed_at": "2026-10-30T12:26:51.644109+00:00", "quote": "okenizer | Custom | | Training | From Scratch | | Fine-tuning | Instruction Tuning | | Status | Experimental | | Version | v1 | --- 🌍 Languages Current multilingual support includes: - 🇪🇬 Arabic - 🇪🇬 Egyptian Arabic - 🇺🇸 English - Programming", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "bugkira-ai/babylm-paranlru-19m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Matched bespoke6×384 recurrent-cell BabyLM pretraining arm on10M-word Strict-Small data, own16K BPE, seed0,3 epochs/1809 steps and full loss/run schedule. Best-supported origin is own scratch training; earliest execution script currently404.", "provenance_sources": [{"url": "https://huggingface.co/bugkira-ai/babylm-paranlru-19m/raw/47c25b825f88920ff1dc0ebedce585bde14778ce/README.md"}, {"url": "https://huggingface.co/bugkira-ai/babylm-paranlru-19m/raw/47c25b825f88920ff1dc0ebedce585bde14778ce/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 NVIDIA GeForce RTX2080Ti (CC7.5)", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "training_duration_reported": "~12.3 minutes", "creator_contribution": "Own pararnn-torch implementation and matched BabyLM pretraining/evaluation of selected recurrent-cell variant; underlying cited ParaRNN research attribution retained.", "provenance_confidence": "medium", "architecture_lineage": "ParaRNN recurrent-cell/solver approach expressly credits Federico Danieli, Miguel Sarabia, Aviv Navon, Amos Storkey and Aaron van den Oord. GRU/LSTM and Griffin/RG-LRU lineage separately acknowledged by relevant cards.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "hardware": "1 NVIDIA GeForce RTX2080Ti (CC7.5)", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "compute_sources": [{"url": "https://huggingface.co/bugkira-ai/babylm-paranlru-19m/raw/47c25b825f88920ff1dc0ebedce585bde14778ce/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "bugkira-ai/babylm-paraslstm-20m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Matched bespoke6×384 recurrent-cell BabyLM pretraining arm on10M-word Strict-Small data, own16K BPE, seed0,3 epochs/1809 steps and full loss/run schedule. Best-supported origin is own scratch training; earliest execution script currently404.", "provenance_sources": [{"url": "https://huggingface.co/bugkira-ai/babylm-paraslstm-20m/raw/d1e293cbbe9aad7f274e3ca34a023d7ce43548f9/README.md"}, {"url": "https://huggingface.co/bugkira-ai/babylm-paraslstm-20m/raw/d1e293cbbe9aad7f274e3ca34a023d7ce43548f9/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 NVIDIA GeForce RTX2080Ti (CC7.5)", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "training_duration_reported": "~29.5 minutes", "creator_contribution": "Own pararnn-torch implementation and matched BabyLM pretraining/evaluation of selected recurrent-cell variant; underlying cited ParaRNN research attribution retained.", "provenance_confidence": "medium", "architecture_lineage": "ParaRNN recurrent-cell/solver approach expressly credits Federico Danieli, Miguel Sarabia, Aviv Navon, Amos Storkey and Aaron van den Oord. GRU/LSTM and Griffin/RG-LRU lineage separately acknowledged by relevant cards.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "hardware": "1 NVIDIA GeForce RTX2080Ti (CC7.5)", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "compute_sources": [{"url": "https://huggingface.co/bugkira-ai/babylm-paraslstm-20m/raw/d1e293cbbe9aad7f274e3ca34a023d7ce43548f9/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "bugkira-ai/babylm-paragru-20m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Matched bespoke6×384 recurrent-cell BabyLM pretraining arm on10M-word Strict-Small data, own16K BPE, seed0,3 epochs/1809 steps and full loss/run schedule. Best-supported origin is own scratch training; earliest execution script currently404.", "provenance_sources": [{"url": "https://huggingface.co/bugkira-ai/babylm-paragru-20m/raw/1dd5ac2411f377000912e431ae303bec87a20ea7/README.md"}, {"url": "https://huggingface.co/bugkira-ai/babylm-paragru-20m/raw/1dd5ac2411f377000912e431ae303bec87a20ea7/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 NVIDIA GeForce RTX2080Ti (CC7.5)", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "training_duration_reported": "~16.3 minutes", "creator_contribution": "Own pararnn-torch implementation and matched BabyLM pretraining/evaluation of selected recurrent-cell variant; underlying cited ParaRNN research attribution retained.", "provenance_confidence": "medium", "architecture_lineage": "ParaRNN recurrent-cell/solver approach expressly credits Federico Danieli, Miguel Sarabia, Aviv Navon, Amos Storkey and Aaron van den Oord. GRU/LSTM and Griffin/RG-LRU lineage separately acknowledged by relevant cards.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "hardware": "1 NVIDIA GeForce RTX2080Ti (CC7.5)", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "compute_sources": [{"url": "https://huggingface.co/bugkira-ai/babylm-paragru-20m/raw/1dd5ac2411f377000912e431ae303bec87a20ea7/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "bugkira-ai/babylm-paralstm-20m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Matched bespoke6×384 recurrent-cell BabyLM pretraining arm on10M-word Strict-Small data, own16K BPE, seed0,3 epochs/1809 steps and full loss/run schedule. Best-supported origin is own scratch training; earliest execution script currently404.", "provenance_sources": [{"url": "https://huggingface.co/bugkira-ai/babylm-paralstm-20m/raw/e0e625f99512f3f86948b902787221fd2630df08/README.md"}, {"url": "https://huggingface.co/bugkira-ai/babylm-paralstm-20m/raw/e0e625f99512f3f86948b902787221fd2630df08/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 NVIDIA GeForce RTX2080Ti (CC7.5)", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "training_duration_reported": "~18.0 minutes", "creator_contribution": "Own pararnn-torch implementation and matched BabyLM pretraining/evaluation of selected recurrent-cell variant; underlying cited ParaRNN research attribution retained.", "provenance_confidence": "medium", "architecture_lineage": "ParaRNN recurrent-cell/solver approach expressly credits Federico Danieli, Miguel Sarabia, Aviv Navon, Amos Storkey and Aaron van den Oord. GRU/LSTM and Griffin/RG-LRU lineage separately acknowledged by relevant cards.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "hardware": "1 NVIDIA GeForce RTX2080Ti (CC7.5)", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "compute_sources": [{"url": "https://huggingface.co/bugkira-ai/babylm-paralstm-20m/raw/e0e625f99512f3f86948b902787221fd2630df08/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "basically-experimental/Pebble-50M-beta", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes an experimental enlarged Pebble hybrid, custom BPE and completed25B-token training, with explicit performance comparison against smaller authored Pebble models.", "provenance_sources": [{"url": "https://huggingface.co/basically-experimental/Pebble-50M-beta/raw/3a25f23677b598b8054aa075c40cd8a98f980a63/README.md", "accessed_at": "2026-10-30T12:26:52.156350+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "basically-experimental/Pebble-50M-Chat-beta", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/basically-experimental/Pebble-50M-Chat-beta/raw/b09c361b193d04b298e229cc986c1e890cc64b32/README.md", "accessed_at": "2026-10-30T12:26:52.229222+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "basically-ai/Pebble-25M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes the custom tiny Pebble Mamba2/Transformer base architecture, its25B-token pretraining mix and evaluated trained release.", "provenance_sources": [{"url": "https://huggingface.co/basically-ai/Pebble-25M/raw/9abce57d0017d9eb2f344579edf7bd09a61e6b0f/README.md", "accessed_at": "2026-10-30T12:26:52.305809+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "OliverSundaram/MoE-Study-Remastered", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/OliverSundaram/MoE-Study-Remastered/raw/98749b06980d48b5203502218104bf44a7832466/README.md", "accessed_at": "2026-10-30T12:26:52.483360+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "omurberaisik/NoTokenLM-Gen-4.5", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/omurberaisik/NoTokenLM-Gen-4.5/raw/07f2896162aa6b8d8d8252d1249f26c0dc20bcab/README.md", "accessed_at": "2026-10-30T12:26:52.559725+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "BytesTalk/Muse-525M", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card explicitly says every weight originated from random initialization;19.2B-token pretraining followed by several successive SFT rounds. Released artifact retains scratch origin and later fine-tuning; both stages preserved.", "provenance_sources": [{"url": "https://huggingface.co/BytesTalk/Muse-525M/raw/0b53297cfe50761bb79810e672cd0d869e355cf3/README.md"}, {"url": "https://huggingface.co/BytesTalk/Muse-525M/raw/0b53297cfe50761bb79810e672cd0d869e355cf3/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/BytesTalk/Muse-525M/raw/0b53297cfe50761bb79810e672cd0d869e355cf3/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Explicit 19.2B-token training from random initialization, own 48000-token tokenizer and assembled multilingual corpus, followed by several successive SFT rounds; selected artifact is the preferred SFT round.", "architecture_and_code_lineage": "Card describes own custom PyTorch Muse implementation with GQA, RoPE, RMSNorm and SwiGLU, a Llama-style recipe; actual frozen config names LlamaForCausalLM/model_type llama. Card references model_muse.py and a .pt checkpoint, but frozen file tree contains config.json and model.safetensors, without model_muse.py. Record this packaging/documentation mismatch rather than infer new architecture from the label.", "evidence_updated_at": "2026-10-31"}
{"model_id": "Se00n00/TinyLM2-50M-Reasoning", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/Se00n00/TinyLM2-50M-Reasoning/raw/ea91ffaf59a57ea2d70b9708a04d491be2912faf/README.md", "accessed_at": "2026-10-30T12:26:52.792829+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Se00n00/TinyLM2-50M-IFT", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/Se00n00/TinyLM2-50M-IFT/raw/a72a126cb2df9b89a4aa959cf3a678f696f666e6/README.md", "accessed_at": "2026-10-30T12:26:52.847867+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Se00n00/TinyLM2-50M-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes an authored custom50M ALiBi decoder and actual4B-token pretraining with training curve, corpus and model-specific evaluation.", "provenance_sources": [{"url": "https://huggingface.co/Se00n00/TinyLM2-50M-Base/raw/bcec9b0ca63ef0fc28d44291c73ec6916b6e2dee/README.md", "accessed_at": "2026-10-30T12:26:52.931293+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "hari-8/transformer-pretraining-from-scratch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/hari-8/transformer-pretraining-from-scratch/raw/bb5cb4f2ac2f37a9c89c516541a60cc1ccc25f05/README.md", "accessed_at": "2026-10-30T12:26:53.205069+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "verify01234/try", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "unknown", "checkpoint_operation": "merge", "provenance_evidence_level": "explicit_model_card_parent_lineage", "provenance_basis": "Card explicitly identifies five pretrained parent model families fused into Child_TextOnly_Fused, followed by lossless AI-DNA-to-safetensors conversion.", "provenance_sources": [{"url": "https://huggingface.co/verify01234/try/raw/161e856dfd3e49c4a41b16d6438ac69aadcfb6a6/README.md", "accessed_at": "2026-10-30T12:26:53.218620+00:00"}, {"url": "https://huggingface.co/verify01234/try/blob/main/README.md", "access_date": "2026-11-01", "finding": "Saved full card read: Child_TextOnly_Fused, fused from SmolLM2, Qwen2.5-0.5B, SmolLM2-360M, TinyLlama-1.1B and OPT125M; AI-DNA lossless conversion to safetensors.", "reused": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "AwakeningOS/ARO-BabyLM-1B-Cooldown", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Continuation of the publisher's own original pretrained parent under a cooldown schedule.", "provenance_sources": [{"url": "https://huggingface.co/AwakeningOS/ARO-BabyLM-1B-Cooldown/blob/main/README.md"}, {"url": "https://huggingface.co/AwakeningOS/ARO-BabyLM-1B-Cooldown/raw/3ee6f6a9356a05dd845a564e7264794d907a50ed/README.md", "accessed_at": "2026-10-30T12:26:53.344648+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "j0no12/Nero-XS-1.1", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "own_checkpoint_channel_expansion_and_recovery_pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Publisher explicitly reports its own original model trained from scratch. Selected release preserves 84 compatible tensors from that trained base, expands 30 FFN tensors from 531 to 582 channels with 51 new random channels, removes Engram by annealing, and continues recovery pretraining. This is a modified own checkpoint with scratch ancestry.", "provenance_sources": [{"url": "https://huggingface.co/j0no12/Nero-XS-1.1/raw/6744b8211226c598d8e98bf3ecf43a1bfc55973c/README.md"}, {"url": "https://huggingface.co/j0no12/Nero-XS-1.1/blob/6744b8211226c598d8e98bf3ecf43a1bfc55973c/README.md", "access_date": "2026-10-31", "scope": "complete frozen card read"}, {"url": "https://huggingface.co/j0no12/Nero-XS-1.1/blob/6744b8211226c598d8e98bf3ecf43a1bfc55973c/training_state.json", "access_date": "2026-10-31", "scope": "complete training state read"}, {"url": "https://huggingface.co/j0no12/Nero-XS-1.1/blob/6744b8211226c598d8e98bf3ecf43a1bfc55973c/config.json", "access_date": "2026-10-31", "scope": "complete configuration read"}, {"url": "https://huggingface.co/j0no12/Nero-XS-1.1/blob/6744b8211226c598d8e98bf3ecf43a1bfc55973c/LICENSE", "access_date": "2026-10-31", "scope": "complete rights and upstream-material notice read"}, {"url": "https://huggingface.co/j0no12/Nero-XS-1.1/blob/6744b8211226c598d8e98bf3ecf43a1bfc55973c/modeling_nero_xs_2.py", "access_date": "2026-10-31", "scope": "model constructors and loading/initialization paths read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "TPU v5e-8 through PyTorch/XLA", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Original own pretraining plus preserving/expanding trained FFN tensors, Engram annealing and recovery continued pretraining.", "architecture_lineage": "Custom recurrent grouped-query XSA language model with RoPE, RMSNorm, loop conditioning and FFN; original EngramLite hashed n-gram memory removed during selected recovery. This is architecture/component reuse and own implementation, not evidence of external pretrained weights.", "hardware": "TPU v5e-8 through PyTorch/XLA", "compute_sources": [{"url": "https://huggingface.co/j0no12/Nero-XS-1.1/blob/6744b8211226c598d8e98bf3ecf43a1bfc55973c/README.md", "access_date": "2026-10-31", "scope": "complete frozen card read"}, {"url": "https://huggingface.co/j0no12/Nero-XS-1.1/blob/6744b8211226c598d8e98bf3ecf43a1bfc55973c/training_state.json", "access_date": "2026-10-31", "scope": "complete training state read"}, {"url": "https://huggingface.co/j0no12/Nero-XS-1.1/blob/6744b8211226c598d8e98bf3ecf43a1bfc55973c/config.json", "access_date": "2026-10-31", "scope": "complete configuration read"}, {"url": "https://huggingface.co/j0no12/Nero-XS-1.1/blob/6744b8211226c598d8e98bf3ecf43a1bfc55973c/LICENSE", "access_date": "2026-10-31", "scope": "complete rights and upstream-material notice read"}, {"url": "https://huggingface.co/j0no12/Nero-XS-1.1/blob/6744b8211226c598d8e98bf3ecf43a1bfc55973c/modeling_nero_xs_2.py", "access_date": "2026-10-31", "scope": "model constructors and loading/initialization paths read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "Amr-Hegazy/grt-medium-isoparam", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author's Gated Recurrent Transformer research checkpoint reports novel shared recurrent blocks, a specific isoFLOP/isoParam architecture and about9.8B CCCC training tokens with measured validation loss.", "provenance_sources": [{"url": "https://huggingface.co/Amr-Hegazy/grt-medium-isoparam/raw/9862c91231a4b3ff289e911135959fd81de65adc/README.md", "accessed_at": "2026-10-30T12:26:53.464649+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "Amr-Hegazy/grt-medium-isoflop", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author's Gated Recurrent Transformer research checkpoint reports novel shared recurrent blocks, a specific isoFLOP/isoParam architecture and about9.8B CCCC training tokens with measured validation loss.", "provenance_sources": [{"url": "https://huggingface.co/Amr-Hegazy/grt-medium-isoflop/raw/d9b91a93e75348bc0bda3286a0e75744a7d5a0fc/README.md", "accessed_at": "2026-10-30T12:26:53.682979+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "Amr-Hegazy/grt-large-isoparam", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author's Gated Recurrent Transformer research checkpoint reports novel shared recurrent blocks, a specific isoFLOP/isoParam architecture and about9.8B CCCC training tokens with measured validation loss.", "provenance_sources": [{"url": "https://huggingface.co/Amr-Hegazy/grt-large-isoparam/raw/bf1eca2ee66219613b3044cf48d26ab6728c0a57/README.md", "accessed_at": "2026-10-30T12:26:53.903265+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "Amr-Hegazy/grt-large-isoflop", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author's Gated Recurrent Transformer research checkpoint reports novel shared recurrent blocks, a specific isoFLOP/isoParam architecture and about9.8B CCCC training tokens with measured validation loss.", "provenance_sources": [{"url": "https://huggingface.co/Amr-Hegazy/grt-large-isoflop/raw/96203022595693072f6bfcf6efb0a69517cb33e1/README.md", "accessed_at": "2026-10-30T12:26:53.931309+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "yehezkielhaganta/NimoLM-244M-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining_in_progress", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card explicitly describes NimoLM developed from scratch with own tokenizer and hybrid GLA/causal-attention architecture. Current checkpoint has10,750 completed optimizer steps/~5.64B tokens; ongoing training is actual training, not an untrained proposal.", "provenance_sources": [{"url": "https://huggingface.co/yehezkielhaganta/NimoLM-244M-base/raw/c2f12299c33bc0695ad59e533d6b283e6e43a1ad/README.md", "accessed_at": "2026-10-30T12:26:54.010051+00:00"}, {"url": "https://huggingface.co/yehezkielhaganta/NimoLM-244M-base/raw/c2f12299c33bc0695ad59e533d6b283e6e43a1ad/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "BananaMind/BananaMind-2.1-Pico-Preview", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2.1-Pico-Preview/raw/b1309d2416faadd43132b7d765fcd66e64ae60e6/README.md"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2.1-Pico-Preview/raw/b1309d2416faadd43132b7d765fcd66e64ae60e6/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2.1-Pico-Preview/raw/b1309d2416faadd43132b7d765fcd66e64ae60e6/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA RTX 5070 Ti", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "training_duration_reported": "Approximately 1.2 hours, publisher report.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Three-block decoder with shared middle-block loop, XSA refresh gates and causal hashed n-gram embedding", "source_url": "https://huggingface.co/BananaMind/BananaMind-2.1-Pico-Preview/raw/b1309d2416faadd43132b7d765fcd66e64ae60e6/README.md"}], "hardware": "1 x NVIDIA RTX 5070 Ti", "compute_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2.1-Pico-Preview/raw/b1309d2416faadd43132b7d765fcd66e64ae60e6/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2.1-Pico-Preview/raw/b1309d2416faadd43132b7d765fcd66e64ae60e6/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "AwakeningOS/ARO-BabyLM-1B", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/AwakeningOS/ARO-BabyLM-1B/blob/main/README.md"}, {"url": "https://huggingface.co/AwakeningOS/ARO-BabyLM-1B/raw/7463e99468fcd785932845ea940e488f5701feee/README.md", "accessed_at": "2026-10-30T12:26:54.097359+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "shawam/project-manager-slm-5m", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "copy_or_repackaging", "provenance_evidence_level": "exact_file_identity", "provenance_basis": "Frozen shawam54ee037 and upstream sathishphdai ec57c101 publish identical SafeTensorsSHA2569da26ae4af977344ba7a95cba8aa08322836a8461938b8f8c188f6dd97b5c9a8 and PyTorchSHA2564fec4eefcefa9b9f8c0dba3627942a55d7aaef0668cf1f8ce2e582131441d895. Config, tokenizer andREADME Git objects also match; usage explicitly downloads upstream sathishphdai weights. This is a mirror of someone else model.", "provenance_sources": [{"url": "https://huggingface.co/api/models/shawam/project-manager-slm-5m/tree/54ee0371319432463ddf0df7f319ac88c8662388?recursive=false", "access_date": "2026-11-01", "finding": "Pinned LFS SHA256 identities match for both weight formats; README/config/tokenizer Git object identities also match."}, {"url": "https://huggingface.co/api/models/sathishphdai/project-manager-slm-5m/tree/ec57c1016dd299c49ab2ebd567586913b08bd8ce?recursive=false", "access_date": "2026-11-01", "finding": "Pinned LFS SHA256 identities match for both weight formats; README/config/tokenizer Git object identities also match."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "upstream_checkpoint": "sathishphdai/project-manager-slm-5m"}
{"model_id": "basically-ai/Pebble-10M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes the custom tiny Pebble Mamba2/Transformer base architecture, its25B-token pretraining mix and evaluated trained release.", "provenance_sources": [{"url": "https://huggingface.co/basically-ai/Pebble-10M/raw/eebf5dadcf997917c35feb50b3052b3b9f44c0eb/README.md", "accessed_at": "2026-10-30T12:26:54.541623+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "nithin521/Meme_Caption_Generator", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/nithin521/Meme_Caption_Generator/raw/152b057ec6c87ac719a867e195f7d4a9b034e742/README.md", "accessed_at": "2026-10-30T12:26:54.598626+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "facebook/MobileMoE-S-QAT", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "quantization_aware_training_of_own_sft_checkpoint", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card explicitly identifies this as stage4 INT4 quantization-aware training on top of MobileMoE-S-SFT, following own family pretraining and mid-training.", "provenance_sources": [{"url": "https://huggingface.co/facebook/MobileMoE-S-QAT/blob/afda132cad380ac47da5ef055f186884d1c12f65/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "upstream_checkpoint": "facebook/MobileMoE-S-SFT"}
{"model_id": "seyhachhorn/cambodia-tourism-lm", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/seyhachhorn/cambodia-tourism-lm/raw/2db1e8db69965ad467be1b3ba09910b3b7e930c5/README.md", "accessed_at": "2026-10-30T12:26:54.750498+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "BananaMind/BananaMind-2.1-Unified", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Frozen card explicitly says trained from scratch; completed checkpoint metadata reports all 72479 steps, 37999869952 tokens, eight GPUs and elapsed runtime.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2.1-Unified/raw/3d4533034db064648704d4f4eac7d4e9b4ccf276/README.md"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2.1-Unified/raw/3d4533034db064648704d4f4eac7d4e9b4ccf276/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2.1-Unified/raw/3d4533034db064648704d4f4eac7d4e9b4ccf276/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "8 x NVIDIA RTX PRO 6000 Blackwell Server Edition", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "training_duration_reported": "27,405 seconds (7 hours 37 minutes), publisher card; metadata elapsed_seconds 27405.331552505493.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Own three-tower relay Transformer with output heads A/C and no direct output loss on relay B", "source_url": "https://huggingface.co/BananaMind/BananaMind-2.1-Unified/raw/3d4533034db064648704d4f4eac7d4e9b4ccf276/README.md"}], "hardware": "8 x NVIDIA RTX PRO 6000 Blackwell Server Edition", "compute_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2.1-Unified/raw/3d4533034db064648704d4f4eac7d4e9b4ccf276/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2.1-Unified/raw/3d4533034db064648704d4f4eac7d4e9b4ccf276/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "vedantjadhav701/SparkAI-47m-llama-10b-token", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/vedantjadhav701/SparkAI-47m-llama-10b-token/raw/83503570efa3d3ec2916e00eac98cc2acc81432a/README.md", "accessed_at": "2026-10-30T12:26:54.892436+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "kerzgrr/Tercet-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card identifies this as own pretraining-only TinyGDN base, with custom hybrid GDN-2/GQA dimensions, explicit corpus exposure, optimizer checkpoint and validation loss. Best-supported own scratch origin. flash-linear-attention/Gated DeltaNet-2 implementation credit remains upstream; no external pretrained weights named.", "provenance_sources": [{"url": "https://huggingface.co/kerzgrr/Tercet-base/raw/d8b65a5870188dd5568344729711170c6158e955/README.md", "accessed_at": "2026-10-30T12:26:55.168955+00:00"}, {"url": "https://huggingface.co/kerzgrr/Tercet-base/raw/d8b65a5870188dd5568344729711170c6158e955/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "Muse-research/Muse2-125M-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author describes entirely from-scratch family, custom hybrid architecture/tokenizer and base plus own SFT weights. Best-supported interpretation is original pretraining.", "provenance_sources": [{"url": "https://huggingface.co/Muse-research/Muse2-125M-Base/blob/main/README.md"}, {"url": "https://huggingface.co/Muse-research/Muse2-125M-Base/raw/f89f923969fb535cea6898b8e2f45288ee69aaaf/README.md", "accessed_at": "2026-10-30T12:26:55.223359+00:00"}, {"url": "https://huggingface.co/Muse-research/Muse2-125M-Base", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Muse-research/Muse2-230M-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author describes entirely from-scratch family, custom hybrid architecture/tokenizer and base plus own SFT weights. Best-supported interpretation is original pretraining.", "provenance_sources": [{"url": "https://huggingface.co/Muse-research/Muse2-230M-Base/blob/main/README.md"}, {"url": "https://huggingface.co/Muse-research/Muse2-230M-Base/raw/93750cf644683927864d09877ecd85ac6f208c4d/README.md", "accessed_at": "2026-10-30T12:26:55.303684+00:00"}, {"url": "https://huggingface.co/Muse-research/Muse2-230M-Base", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "AETHORIA-AI/TR-HASH-MoE-200M-130B", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card reports completed 130B-token exposure training of own bespoke TR-HASH MoE 201.2M base, 70B unique selection/replay plan, deterministic token-ID experts and own tokenizer/runtime links. This supports own scratch origin rather than named checkpoint adaptation.", "provenance_sources": [{"url": "https://huggingface.co/AETHORIA-AI/TR-HASH-MoE-200M-130B/raw/c6bd0f2cab456fdb8beb4465a083992ae231752d/README.md", "accessed_at": "2026-10-30T12:26:55.384076+00:00"}, {"url": "https://huggingface.co/AETHORIA-AI/TR-HASH-MoE-200M-130B/raw/c6bd0f2cab456fdb8beb4465a083992ae231752d/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "VertexResearch/Vertex-0.1-10M-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original10.9M4-layer Llama-style base,3207steps/26M FineWeb tokens and1h M5Pro training support own scratch-origin model.", "provenance_sources": [{"url": "https://huggingface.co/VertexResearch/Vertex-0.1-10M-base/raw/5b29705e453e8bab496df16f93ee96ededc9ae41/README.md", "accessed_at": "2026-10-30T12:26:55.449878+00:00"}, {"url": "https://huggingface.co/VertexResearch/Vertex-0.1-10M-base", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "AETHORIA-AI/TR-HASH-MoE-200M-160B-Refinement", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/AETHORIA-AI/TR-HASH-MoE-200M-160B-Refinement/raw/1401f5c0de1babbbd1738573ca544ff850ec2328/README.md", "accessed_at": "2026-10-30T12:26:55.679753+00:00", "quote": "on, a fresh optimizer and learning-rate schedule, initialized from the base checkpoint's weights only. | Setting | Value | |---|---:| | Base checkpoint | TR-HASH-200M-130B (post-decay) | | Training tokens | 69,997,690,880 (70B unique, exact match", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Embim/looped-qwen3-9.4M-T32-fineweb", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Looped-qwen3-9.4M card and cross-linked repository report random-init100M-token pretraining on rented A10080GB Linux for about2.5h. Earlier local RTX5080 Windows attempt is distinguished from published run.", "provenance_sources": [{"url": "https://huggingface.co/Embim/looped-qwen3-9.4M-T32-fineweb/raw/2b9b6f28b127aad163e93acd0b8f25c648b06638/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "rented A10080GB Linux, approximately2.5h; earlier RTX5080 attempt separate", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Welmia/welmia-1.0-81m", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Frozen card explicitly states original from-random-initialization pretraining with no borrowed backbone, followed by Alpaca-style instruction tuning. Selected81M release is instruction-tuned.", "provenance_sources": [{"url": "https://huggingface.co/Welmia/welmia-1.0-81m/raw/fe5abeca5f4b1713e098d3c458049f35563f7b08/README.md", "accessed_at": "2026-10-30T12:26:55.824733+00:00"}, {"url": "https://huggingface.co/Welmia/welmia-1.0-81m/raw/fe5abeca5f4b1713e098d3c458049f35563f7b08/README.md", "scope": "Complete saved frozen card read", "excerpt": "from-random-initialization pretraining, no borrowed backbone", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Own81M pretraining and instruction SFT, model/loading/KV-cache implementation fixes.", "provenance_confidence": "high", "architecture_lineage": "Publisher-authored TinyLlama-inspired decoder code using RoPE/RMSNorm/SwiGLU; standard GPT-2 BPE tokenizer reused. TinyLlama architecture inspiration is explicitly acknowledged; no TinyLlama weights claimed.", "evidence_updated_at": "2026-11-01"}
{"model_id": "sabari2005/cyberslm-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sabari2005/cyberslm-base/raw/84530d2379941ccd8822d067b24c54e91820cb25/README.md", "accessed_at": "2026-10-30T12:26:55.934897+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "jaipkapoor99/ultron-113m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/jaipkapoor99/ultron-113m/raw/c44f5744f3aa7d33967420950bb6f5f8b2053a26/README.md", "accessed_at": "2026-10-30T12:26:56.008953+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "jaipkapoor99/ultron-113m-instruct", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/jaipkapoor99/ultron-113m-instruct/raw/f08d413f46dc8cc003eb2d4bf22f60d334b56595/README.md", "accessed_at": "2026-10-30T12:26:56.083868+00:00", "quote": "is a lightweight instruction-tuned language model derived from Ultron-113M through Supervised Fine-Tuning (SFT) on 163.84 Million tokens of the HuggingFaceTB/smoltalk conversational dataset on an NVIDIA RTX 5090. 🤖 \"There are no strings on me.\" 🤖 - GitH", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "BananaMind-Model-Previewers/BananaMind-2.1-Unified", "model_provenance": "mirror", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "same_own_model_alternate_publisher_copy", "provenance_evidence_level": "publisher_statement_and_identical_weight_file_identifier", "provenance_basis": "Card explicitly says BananaMind trained from scratch; complete metadata and identical frozen weight SHA256 match the main BananaMind Unified release. Alternate publisher copy of the same experiment, not another training run.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind-Model-Previewers/BananaMind-2.1-Unified/raw/14c3aca500a4fa7948626c0331b48ce080cb7f50/README.md"}, {"url": "https://huggingface.co/BananaMind-Model-Previewers/BananaMind-2.1-Unified/raw/14c3aca500a4fa7948626c0331b48ce080cb7f50/README.md", "access_date": "2026-10-31", "scope": "frozen introduction/training/provenance/hardware and source-credit sections read"}, {"url": "https://huggingface.co/BananaMind-Model-Previewers/BananaMind-2.1-Unified/raw/14c3aca500a4fa7948626c0331b48ce080cb7f50/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete frozen metadata read"}, {"url": "https://huggingface.co/api/models/BananaMind-Model-Previewers/BananaMind-2.1-Unified/tree/14c3aca500a4fa7948626c0331b48ce080cb7f50", "access_date": "2026-10-31", "scope": "frozen file list and weight LFS identifier read"}, {"url": "https://huggingface.co/api/models/BananaMind/BananaMind-2.1-Unified/tree/3d4533034db064648704d4f4eac7d4e9b4ccf276", "access_date": "2026-10-31", "scope": "already-read primary frozen file list reused; same 140025876-byte weight LFS SHA256 3b23f2e1f59a5631fea477026f7b9b14f0b6fddeb42a1f448e7eff2eb365887a"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "8 x NVIDIA RTX PRO 6000 Blackwell Server Edition", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Physical compute country and exact builder location during this eight-GPU run unstated; same creator base estimate reused separately.", "training_duration_reported": "27405 seconds (7 h 37 m), same underlying training run as primary release.", "compute_source": "Frozen card/checkpoint metadata, matched to primary weight hash.", "upstream_checkpoint": "Identical to BananaMind/BananaMind-2.1-Unified frozen weight file SHA256 3b23f2e1f59a5631fea477026f7b9b14f0b6fddeb42a1f448e7eff2eb365887a.", "creator_contribution": "BananaMind/Banaxi training; this group publishes an alternate preview copy.", "upstream_architecture_or_code": [{"name": "Own three-tower relay Transformer", "source_url": "https://huggingface.co/BananaMind-Model-Previewers/BananaMind-2.1-Unified/raw/14c3aca500a4fa7948626c0331b48ce080cb7f50/README.md"}, {"name": "Anthropic jacobian-lens method", "role": "interpretability analysis method, not original model backbone", "source_url": "https://huggingface.co/BananaMind-Model-Previewers/BananaMind-2.1-Unified/raw/14c3aca500a4fa7948626c0331b48ce080cb7f50/README.md"}], "hardware": "8 x NVIDIA RTX PRO 6000 Blackwell Server Edition", "compute_sources": [{"url": "https://huggingface.co/BananaMind-Model-Previewers/BananaMind-2.1-Unified/raw/14c3aca500a4fa7948626c0331b48ce080cb7f50/README.md", "access_date": "2026-10-31", "scope": "frozen introduction/training/provenance/hardware and source-credit sections read"}, {"url": "https://huggingface.co/BananaMind-Model-Previewers/BananaMind-2.1-Unified/raw/14c3aca500a4fa7948626c0331b48ce080cb7f50/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete frozen metadata read"}, {"url": "https://huggingface.co/api/models/BananaMind-Model-Previewers/BananaMind-2.1-Unified/tree/14c3aca500a4fa7948626c0331b48ce080cb7f50", "access_date": "2026-10-31", "scope": "frozen file list and weight LFS identifier read"}, {"url": "https://huggingface.co/api/models/BananaMind/BananaMind-2.1-Unified/tree/3d4533034db064648704d4f4eac7d4e9b4ccf276", "access_date": "2026-10-31", "scope": "already-read primary frozen file list reused; same 140025876-byte weight LFS SHA256 3b23f2e1f59a5631fea477026f7b9b14f0b6fddeb42a1f448e7eff2eb365887a"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "tareq052/soccer-bert", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/tareq052/soccer-bert/raw/9ac2ceace0b9910fca6b3e31fd11e0d482fea3ed/README.md", "accessed_at": "2026-10-30T12:26:56.377482+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "BananaMind/BananaMind-2-Pro-Preview", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro-Preview/raw/db1d7267bd27706f6d072b8d251a1421f0dd96bb/README.md"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro-Preview/raw/db1d7267bd27706f6d072b8d251a1421f0dd96bb/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind", "access_date": "2026-10-31", "scope": "actual rendered organisation card", "finding": "All BananaMind 2 models described as trained from scratch on a single RTX 5070 Ti; shared GQA/QK norm, SwiGLU, RoPE, RMSNorm recipe. Model-specific eight-GPU Unified report takes precedence over series generalisation."}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro-Preview/raw/db1d7267bd27706f6d072b8d251a1421f0dd96bb/LICENSE", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro-Preview/raw/db1d7267bd27706f6d072b8d251a1421f0dd96bb/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}, {"url": "https://huggingface.co/posts/Banaxi-Tech/824856524827227", "access_date": "2026-10-31", "scope": "actual rendered creator release post", "finding": "Pro Preview trained on single 5070 Ti in about 11 days; separate from exact throughput-derived card estimate."}, {"url": "https://huggingface.co/posts/Banaxi-Tech/678789079412959", "access_date": "2026-10-31", "event_date": "2026-08-02", "scope": "actual rendered full creator announcement", "excerpt": "August 3, 1PM in Austria time", "finding": "Explicit Austrian release-time reference during the local Pro pretraining run; combined with own Austrian-grid local-system power estimate supports an inferred Austrian home/training base."}], "uk_build_status": "unknown", "builder_location_country": "Austria", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA RTX 5070 Ti", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Austrian builder-base estimate is inferred; no literal residence or physical compute-country statement. University/institutional route unstated.", "training_duration_reported": "About 11 days in direct creator release post; card separately estimates 274.95 hours from observed throughput.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Llama-style decoder Transformer / GQA, QK norm, SwiGLU, RoPE and RMSNorm components", "source_url": "https://huggingface.co/BananaMind/BananaMind-2-Pro-Preview/raw/db1d7267bd27706f6d072b8d251a1421f0dd96bb/README.md"}], "hardware": "1 x NVIDIA RTX 5070 Ti", "builder_location_at_training": {"country": "Austria", "city": null, "confidence": "medium", "direct_or_inferred": "inferred", "period": "July-August 2026 local Pro pretraining run", "basis": "Creator says Pro trains locally, schedules release in Austria time, and calculates own measured-system emissions with Austrian grid intensity. October Italian vacation is separately documented.", "sources": ["https://huggingface.co/posts/Banaxi-Tech/678789079412959", "https://huggingface.co/BananaMind/BananaMind-2-Pro/raw/846592429ab5f01390f6fa7340921823603c3c3a/README.md", "https://huggingface.co/BananaMind/BananaMind-2-SLMoE/discussions/2"]}, "compute_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro-Preview/raw/db1d7267bd27706f6d072b8d251a1421f0dd96bb/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind", "access_date": "2026-10-31", "scope": "actual rendered organisation card", "finding": "All BananaMind 2 models described as trained from scratch on a single RTX 5070 Ti; shared GQA/QK norm, SwiGLU, RoPE, RMSNorm recipe. Model-specific eight-GPU Unified report takes precedence over series generalisation."}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro-Preview/raw/db1d7267bd27706f6d072b8d251a1421f0dd96bb/LICENSE", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Pro-Preview/raw/db1d7267bd27706f6d072b8d251a1421f0dd96bb/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}, {"url": "https://huggingface.co/posts/Banaxi-Tech/824856524827227", "access_date": "2026-10-31", "scope": "actual rendered creator release post", "finding": "Pro Preview trained on single 5070 Ti in about 11 days; separate from exact throughput-derived card estimate."}, {"url": "https://huggingface.co/posts/Banaxi-Tech/678789079412959", "access_date": "2026-10-31", "event_date": "2026-08-02", "scope": "actual rendered full creator announcement", "excerpt": "August 3, 1PM in Austria time", "finding": "Explicit Austrian release-time reference during the local Pro pretraining run; combined with own Austrian-grid local-system power estimate supports an inferred Austrian home/training base."}], "evidence_updated_at": "2026-10-31"}
{"model_id": "sciencemj/tinyllm-29m-chat", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "sft_of_own_base", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card explicitly says own TinyStories pretraining followed by DailyDialog SFT; custom29.58M model and own8kBPE. Retain publisher-owned scratch lineage while recording final SFT operation.", "provenance_sources": [{"url": "https://huggingface.co/sciencemj/tinyllm-29m-chat/raw/774fb227a4ed493e5d4cbcea3f36d6416953cd9d/README.md", "accessed_at": "2026-10-30T12:26:56.547027+00:00"}, {"url": "https://huggingface.co/sciencemj/tinyllm-29m-chat/raw/774fb227a4ed493e5d4cbcea3f36d6416953cd9d/README.md", "read_scope": "Cached model/base/architecture/training and attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "sciencemj/tinyllm-29m-tinystories", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Custom29.58M decoder and own8kBPE, explicit first-stage TinyStories pretraining on RTX3060Ti for3.06h; supports own scratch base.", "provenance_sources": [{"url": "https://huggingface.co/sciencemj/tinyllm-29m-tinystories/raw/938a6ab16226feeb7edd123dfa1a8061c7fd3a1a/README.md", "accessed_at": "2026-10-30T12:26:56.618773+00:00"}, {"url": "https://huggingface.co/sciencemj/tinyllm-29m-tinystories/raw/938a6ab16226feeb7edd123dfa1a8061c7fd3a1a/README.md", "read_scope": "Cached model/base/architecture/training and attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "ashrafs1/nanogpt-3000", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/ashrafs1/nanogpt-3000/raw/5ad8a32a3377e570c27ccdb630ec78490c2d4afc/README.md", "accessed_at": "2026-10-30T12:26:56.673330+00:00", "quote": " 124M-parameter GPT-2-class causal language model built from scratch in PyTorch and trained on FineWeb-10B. This is a pretraining checkpoint (3,000 steps, ~1.57B tokens). Quick Stats | | | |---|---| | Parameters | ~124M | | Context | 2,048 tokens | | ", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "alpharomercoma/slm-pipeline-151m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/alpharomercoma/slm-pipeline-151m/raw/ccc7d353ae972d61b58174049cc28c3095b8786b/README.md", "accessed_at": "2026-10-30T12:26:57.024708+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "shubhamprakash108/decoder-only-transformer-tinystories-32m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/shubhamprakash108/decoder-only-transformer-tinystories-32m/raw/fc169cbf12540c47414abd35c92338a322ec9186/README.md", "accessed_at": "2026-10-30T12:26:57.062740+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "MixlyGames/Orpheus_Zero-base-0.3b", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card explicitly identifies a277M autoregressive decoder trained from scratch on30GB of cleaned text, with a local RTX4060 Laptop8GB run and completed baseline pretraining.", "provenance_sources": [{"url": "https://huggingface.co/MixlyGames/Orpheus_Zero-base-0.3b/blob/979742d94824eaac11601e1025f8b50b41698874/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA GeForce RTX 4060 Laptop GPU (8 GB VRAM)", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "hardware": "1 x NVIDIA GeForce RTX 4060 Laptop GPU (8 GB VRAM)", "compute_source": "local GPU node; publisher model-card statement"}
{"model_id": "Betrayedchair24/BIJA_150M_BASE", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Betrayedchair24/BIJA_150M_BASE/raw/8551cce3921514f3bbd4f027884e86e547b7c0c5/README.md", "accessed_at": "2026-10-30T12:26:57.159468+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "JohnEnev/modern-llm-v1-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/JohnEnev/modern-llm-v1-base/raw/287e00e1c999138d4941052f29f6befa8bf74d70/README.md", "accessed_at": "2026-10-30T12:26:57.216067+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "JohnEnev/modern-llm-v1-sft", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/JohnEnev/modern-llm-v1-sft/raw/602ae7de299cd4e4917272b58dd6a637f28396b9/README.md", "accessed_at": "2026-10-30T12:26:57.249684+00:00", "quote": "n-tuned --- modern-llm V1 (SFT) The V1 base model further trained with supervised fine-tuning to follow instructions and respond conversationally. Use it on this HF Space: https://huggingface.co/spaces/JohnEnev/modern-llm-playground 📖 Full writeup: Part", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "JohnEnev/modern-llm-v1-grpo", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card names own V1 SFT checkpoint and Base -> SFT -> GRPO chain; this released stage is GRPO post-training of its own model.", "provenance_sources": [{"url": "https://huggingface.co/JohnEnev/modern-llm-v1-grpo/raw/8478fb48ba5139a85195357de178890816ee606b/README.md", "accessed_at": "2026-10-30T12:26:57.637635+00:00"}, {"url": "https://huggingface.co/JohnEnev/modern-llm-v1-grpo/raw/8478fb48ba5139a85195357de178890816ee606b/README.md", "access_date": "2026-11-01", "scope": "Saved model-card introduction and training/family sections read", "finding": "Card names own V1 SFT checkpoint and Base -> SFT -> GRPO chain; this released stage is GRPO post-training of its own model."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "JohnEnev/modern-llm-v2-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/JohnEnev/modern-llm-v2-base/raw/51905a9f6060b72abbf085d5d4793851fb1ba90a/README.md", "accessed_at": "2026-10-30T12:26:57.747084+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "JohnEnev/modern-llm-v2-sft", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/JohnEnev/modern-llm-v2-sft/raw/2e651df3c2c5e254c67853c16ba2285f0c0b9584/README.md", "accessed_at": "2026-10-30T12:26:57.700897+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "JohnEnev/modern-llm-v2-grpo", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/JohnEnev/modern-llm-v2-grpo/raw/81301360cd0d03293a9ecc994c0a87664f3cd5e9/README.md", "accessed_at": "2026-10-30T12:26:57.797632+00:00", "quote": "- from-scratch - gpt --- modern-llm V2 (GRPO) The V2 SFT model further trained with GRPO (Group Relative Policy Optimization). The reward optimized for a percentage-correct signal (see the checkpoint path). See Part 3 for the RL setup. Use it on th", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AppleMind-AI/AppleMind-1.0-Mini", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Frozen README explicitly claims from-scratch pretraining; matching saved training log and metadata document the run to final upload.", "provenance_sources": [{"url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/blob/main/README.md"}, {"url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/raw/02508d4d3acce58d25f11e38baaf70de86a4d1ea/README.md", "accessed_at": "2026-10-30T12:26:57.855857+00:00"}, {"url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/raw/02508d4d3acce58d25f11e38baaf70de86a4d1ea/README.md", "access_date": "2026-10-31", "scope": "complete frozen README read", "finding": "Publisher says trained from scratch; final 1,020,480 parameters, 299,892,736 tokens and 2,288 completed steps. Card includes custom-architecture/custom-code tags and claims 4 layers / 128 hidden size in its architecture table."}, {"url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/raw/02508d4d3acce58d25f11e38baaf70de86a4d1ea/config.json", "access_date": "2026-10-31", "scope": "complete small frozen configuration read", "finding": "Standard Transformers GPT2LMHeadModel; actual configuration 2 layers, 2 heads, hidden size 20, vocab 50,260; no auto_map custom-code field."}, {"url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/raw/02508d4d3acce58d25f11e38baaf70de86a4d1ea/training_metadata.json", "access_date": "2026-10-31", "scope": "complete frozen training metadata read", "finding": "Confirms 1,020,480 parameters; 2,288 steps / 299,892,736 tokens; model_config is 2 layers, 2 heads, hidden size20, BF16 training."}, {"url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/raw/02508d4d3acce58d25f11e38baaf70de86a4d1ea/LOGS.md", "access_date": "2026-10-31", "scope": "training/hardware/configuration headers, all matching hardware/status lines, final training completion and upload footer read; repetitive installation/download/step-progress lines are not treated as additional biographical evidence", "excerpt": "Colab L4 High-RAM; GPU: NVIDIA L4", "finding": "Actual log identifies Colab L4 High-RAM, NVIDIA L4, 22.03 GB reported VRAM; completes 2,288 steps and successful final upload. Step 2,200 reports elapsed0.77h; no exact final wall-time stated."}, {"url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini-Checkpoints/discussions/1", "access_date": "2026-10-31", "event_date": "2026-08-12", "scope": "creator prose and hardware/training log sections read in complete public thread payload", "excerpt": "fyi this is colab L4", "finding": "Owner explicitly identifies Colab L4 for the matching 300M-token AppleMind Mini run and confirms Colab Pro access. Earlier CPU duration was a plan before switching to this GPU run."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA L4 (22.03 GB VRAM reported by training log)", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Exact final runtime, physical compute country and builder country during August12 training remain unstated.; Model creator has an inferred Pacific-time regional base; university-ever is not documented.", "training_tokens_publisher_reported": 299892736, "compute_source": "Google Colab Pro, L4 High-RAM runtime", "upstream_checkpoint": "No reused pretrained checkpoint stated; from-scratch training is publisher claim. Architecture/tokenizer reuse does not establish checkpoint reuse.", "creator_contribution": "Small GPT-2 configuration and pretraining run / dataset mixture, with checkpoint/final publication; later GGUF conversion is separately visible in creator comments.", "documented_architecture": {"implementation": "Transformers GPT2LMHeadModel", "model_type": "gpt2", "layers": 2, "attention_heads": 2, "hidden_size": 20, "vocabulary": 50260, "source_url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/raw/02508d4d3acce58d25f11e38baaf70de86a4d1ea/config.json"}, "upstream_architecture_or_code": [{"name": "GPT-2 / Transformers GPT2LMHeadModel and GPT-2 tokenizer", "source_urls": ["https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/raw/02508d4d3acce58d25f11e38baaf70de86a4d1ea/config.json", "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/raw/02508d4d3acce58d25f11e38baaf70de86a4d1ea/LOGS.md"], "attribution_location": "configuration, training log and tokenizer section of model card"}], "presentation_vs_files": "README custom-architecture/custom-code tags and 4-layer128-hidden table differ from frozen config/training_metadata, both of which specify standard GPT2LMHeadModel2-layer20-hidden.", "training_progress_elapsed_observation": "0.77h at step 2,200 of configured2,289; not an exact final duration", "hardware": "NVIDIA L4 (22.03 GB VRAM reported by training log)", "actual_training_hardware_count": 1, "compute_sources": [{"url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/raw/02508d4d3acce58d25f11e38baaf70de86a4d1ea/README.md", "access_date": "2026-10-31", "scope": "complete frozen README read", "finding": "Publisher says trained from scratch; final 1,020,480 parameters, 299,892,736 tokens and 2,288 completed steps. Card includes custom-architecture/custom-code tags and claims 4 layers / 128 hidden size in its architecture table."}, {"url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/raw/02508d4d3acce58d25f11e38baaf70de86a4d1ea/config.json", "access_date": "2026-10-31", "scope": "complete small frozen configuration read", "finding": "Standard Transformers GPT2LMHeadModel; actual configuration 2 layers, 2 heads, hidden size 20, vocab 50,260; no auto_map custom-code field."}, {"url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/raw/02508d4d3acce58d25f11e38baaf70de86a4d1ea/training_metadata.json", "access_date": "2026-10-31", "scope": "complete frozen training metadata read", "finding": "Confirms 1,020,480 parameters; 2,288 steps / 299,892,736 tokens; model_config is 2 layers, 2 heads, hidden size20, BF16 training."}, {"url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini/raw/02508d4d3acce58d25f11e38baaf70de86a4d1ea/LOGS.md", "access_date": "2026-10-31", "scope": "training/hardware/configuration headers, all matching hardware/status lines, final training completion and upload footer read; repetitive installation/download/step-progress lines are not treated as additional biographical evidence", "excerpt": "Colab L4 High-RAM; GPU: NVIDIA L4", "finding": "Actual log identifies Colab L4 High-RAM, NVIDIA L4, 22.03 GB reported VRAM; completes 2,288 steps and successful final upload. Step 2,200 reports elapsed0.77h; no exact final wall-time stated."}, {"url": "https://huggingface.co/AppleMind-AI/AppleMind-1.0-Mini-Checkpoints/discussions/1", "access_date": "2026-10-31", "event_date": "2026-08-12", "scope": "creator prose and hardware/training log sections read in complete public thread payload", "excerpt": "fyi this is colab L4", "finding": "Owner explicitly identifies Colab L4 for the matching 300M-token AppleMind Mini run and confirms Colab Pro access. Earlier CPU duration was a plan before switching to this GPU run."}], "evidence_updated_at": "2026-10-31"}
{"model_id": "christianrss/chris-gpt-2-124m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/christianrss/chris-gpt-2-124m/raw/28e932aeb3ae45f0c50086a81ba7821a7e71dff3/README.md", "accessed_at": "2026-10-30T12:26:57.922585+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "lordlebu/4000BCSaraswaty", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly attributes the released pretrained weights to an upstream model and describes unchanged redistribution, rather than publisher training.", "provenance_sources": [{"url": "https://huggingface.co/lordlebu/4000BCSaraswaty/raw/689b4e39421fb16e4f0584e2d74c04110b9dd0d1/README.md", "accessed_at": "2026-10-30T12:26:58.241705+00:00", "quote": "thys - not-fine-tuned --- 4000BCSaraswaty This is stock GPT-2 small (124M), unmodified. It is not a fine-tuned model, and nothing in the South of Tethys project uses it. It is published here for provenance, and this card exists so that nobody mis", "basis": "Card explicitly attributes the released pretrained weights to an upstream model and describes unchanged redistribution, rather than publisher training."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Deepjyoti/shakespeare-GPT", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Publisher describes self-contained 4.78M character GPT trained on Tiny Shakespeare for 10,000 steps and explicitly calls it a from-scratch GPT.", "provenance_sources": [{"url": "https://huggingface.co/Deepjyoti/shakespeare-GPT/raw/cce79547be0db4bf49612a96a09b34e237761b1d/README.md", "accessed_at": "2026-10-30T12:26:58.378864+00:00"}, {"url": "https://huggingface.co/Deepjyoti/shakespeare-GPT/raw/cce79547be0db4bf49612a96a09b34e237761b1d/README.md", "access_date": "2026-11-01", "scope": "Saved model-card introduction and training/family sections read", "finding": "Publisher describes self-contained 4.78M character GPT trained on Tiny Shakespeare for 10,000 steps and explicitly calls it a from-scratch GPT."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "ray0rf1re/nano-nano_4.7", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "The card describes an authored redesign from 15 layers/896 width to20 layers/1024 width, a newly corpus-trained BPE tokenizer and a completed43-dataset GTX1080 training run. This supports own from-scratch training rather than assigning external pretrained weights from instruction-tuned metadata.", "provenance_sources": [{"url": "https://huggingface.co/ray0rf1re/nano-nano_4.7/raw/4f0410aea2c3f057ebbe70064c7f07d10e26f871/README.md"}, {"url": "https://huggingface.co/ray0rf1re/nano-nano_4.7/blob/main/README.md", "access_date": "2026-11-01", "finding": "Saved card read.", "reused": true}, {"url": "https://huggingface.co/ray0rf1re/Nano-nano_v4.5/blob/main/README.md", "access_date": "2026-11-01", "finding": "Saved card explicitly from scratch; same creator's model-family lineage.", "reused": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA GTX1080 8GB", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "provenance_evidence_detail": "author_card_architecture_and_model_family_inference", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Ryosei-Akatsuka/kopriaLM", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining_and_sft", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original99360-parameter recursive Transformer/TT-Embedding design,3000-token own BPE, explicit BabyLM/WikiText2 pretraining and Dolly SFT support own scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/Ryosei-Akatsuka/kopriaLM/raw/82d2c48da07ef8816251747a1bdb85c46cfa1ddc/README.md", "accessed_at": "2026-10-30T12:26:58.419942+00:00"}, {"url": "https://huggingface.co/Ryosei-Akatsuka/kopriaLM", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Dream-W/ObsidianSmall-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "ObsidianSmall-Base claims 8.7M parameters and 4B-token from-scratch training. Config and commits identify Metris without geographic context.", "provenance_sources": [{"url": "https://huggingface.co/Dream-W/ObsidianSmall-Base/raw/2c677b26dba13a866ba4edab6e3fd82b2516bc5d/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/Dream-W/ObsidianSmall-Base/raw/2c677b26dba13a866ba4edab6e3fd82b2516bc5d/README.md", "access_date": "2026-10-31", "scope": "complete frozen card"}, {"url": "https://huggingface.co/Dream-W/ObsidianSmall-Base/raw/2c677b26dba13a866ba4edab6e3fd82b2516bc5d/NOTICE_MULTISCREEN.txt", "access_date": "2026-10-31", "scope": "complete attribution notice"}, {"url": "https://huggingface.co/Dream-W/ObsidianSmall-Base/raw/2c677b26dba13a866ba4edab6e3fd82b2516bc5d/CITATION.cff", "access_date": "2026-10-31", "scope": "complete upstream citation"}, {"url": "https://huggingface.co/Dream-W/ObsidianSmall-Base/raw/2c677b26dba13a866ba4edab6e3fd82b2516bc5d/config.json", "access_date": "2026-10-31", "scope": "complete frozen config"}, {"url": "https://huggingface.co/Dream-W/ObsidianSmall-Base/raw/2c677b26dba13a866ba4edab6e3fd82b2516bc5d/model_config.yaml", "access_date": "2026-10-31", "scope": "complete frozen architecture config"}, {"url": "https://huggingface.co/Dream-W/ObsidianSmall-Base/raw/2c677b26dba13a866ba4edab6e3fd82b2516bc5d/build_minimal_runtime.py", "access_date": "2026-10-31", "scope": "complete source read, not executed"}, {"url": "https://huggingface.co/Dream-W/ObsidianSmall-Base/raw/2c677b26dba13a866ba4edab6e3fd82b2516bc5d/runtime/litgpt/pretrain.py", "access_date": "2026-10-31", "scope": "fresh initialization and optional own/external checkpoint-loading paths, file headers/credits and device-reporting paths; optimizer internals not required"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not stated for exact run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Actual training device/count/provider/duration not reported in the read saved files/comments; CUDA inference commands are not training-device evidence.", "creator_contribution": "Publisher reports adapting Multiscreen to a pure nine-block small model, training from scratch on ~4B tokens and integrating custom LitGPT/Triton runtime. Does not originate the underlying Multiscreen architecture or LitGPT framework.", "upstream_architecture_or_code": [{"name": "Multiscreen / Ken M. Nakanishi", "source_url": "https://huggingface.co/Dream-W/ObsidianSmall-Base/raw/2c677b26dba13a866ba4edab6e3fd82b2516bc5d/README.md", "attribution_location": "model card architecture section"}, {"name": "Lightning AI LitGPT runtime", "source_url": "https://huggingface.co/Dream-W/ObsidianSmall-Base/raw/2c677b26dba13a866ba4edab6e3fd82b2516bc5d/NOTICE_MULTISCREEN.txt", "attribution_location": "repository notice and CITATION.cff"}], "architecture_and_code_lineage": "Model card explicitly attributes Multiscreen architecture to Ken M. Nakanishi / Screening Is Enough. Repository NOTICE_MULTISCREEN says runtime is derived from Lightning AI LitGPT with custom Multiscreen training/inference/Triton changes; CITATION.cff is LitGPT/Lightning AI. Bundled initialization credits GPT-NeoX/TinyLlama. Architecture/code reuse is separate from claimed new-weight pretraining.", "evidence_updated_at": "2026-10-31"}
{"model_id": "CreatorJarvis/JarvisLM-350M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/CreatorJarvis/JarvisLM-350M/raw/882ac49f7de3810820c56da17a8bb5cc44b35067/README.md", "accessed_at": "2026-10-30T12:26:58.527162+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "GODELEV/Rose-Medium", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Completed bespoke Rose-X1 training established by step38196/tokens_seen80063705088 metadata and direct owner training discussion. Best-supported origin is own scratch training: model-specific custom dimensions/refresh path and fresh random constructor; no external weight base identified. Literal first initialization/run receipt not public, so origin inference remains labelled.", "provenance_sources": [{"url": "https://huggingface.co/GODELEV/Rose-Medium/raw/5f482dd3a74e1b1c465df263d887d5ea954490bc/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/GODELEV/Rose-Medium/raw/5f482dd3a74e1b1c465df263d887d5ea954490bc/README.md", "access_date": "2026-10-31", "scope": "complete frozen card"}, {"url": "https://huggingface.co/GODELEV/Rose-Medium/raw/5f482dd3a74e1b1c465df263d887d5ea954490bc/training_meta.json", "access_date": "2026-10-31", "scope": "complete actual training metadata"}, {"url": "https://huggingface.co/GODELEV/Rose-Medium/raw/5f482dd3a74e1b1c465df263d887d5ea954490bc/config.json", "access_date": "2026-10-31", "scope": "complete selected model config"}, {"url": "https://huggingface.co/GODELEV/Rose-Medium/raw/5f482dd3a74e1b1c465df263d887d5ea954490bc/configuration_rose_x1.py", "access_date": "2026-10-31", "scope": "complete architecture config class"}, {"url": "https://huggingface.co/GODELEV/Rose-Medium/raw/5f482dd3a74e1b1c465df263d887d5ea954490bc/modeling_rose_x1.py", "access_date": "2026-10-31", "scope": "full model header, credited reference-family/cache design, component declarations and complete random-weight initialization/backbone/head constructors"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not stated for exact run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "creator_contribution": "Own Rose-X1 architecture adaptation, model scaling and ~80B-token training; owner reports own Jupyter/Kaggle-derived pipeline and declines to release the training script. Source-model implementation/runtime is public, full training pipeline is not.", "best_origin_estimate": {"origin": "own_from_scratch_backbone", "confidence": "medium", "direct_or_inferred": "inferred", "basis": "Bespoke Rose-X1 dimensions/new initialization; completed own-training metadata; direct owner corpus/training explanation; no external checkpoint base.", "remaining_gap": "Private training pipeline; literal earliest checkpoint initialization unavailable.", "source_urls": ["https://huggingface.co/GODELEV/Rose-Medium/raw/5f482dd3a74e1b1c465df263d887d5ea954490bc/README.md", "https://huggingface.co/GODELEV/Rose-Medium/raw/5f482dd3a74e1b1c465df263d887d5ea954490bc/training_meta.json", "https://huggingface.co/GODELEV/Rose-Medium/raw/5f482dd3a74e1b1c465df263d887d5ea954490bc/config.json", "https://huggingface.co/GODELEV/Rose-Medium/raw/5f482dd3a74e1b1c465df263d887d5ea954490bc/configuration_rose_x1.py", "https://huggingface.co/GODELEV/Rose-Medium/raw/5f482dd3a74e1b1c465df263d887d5ea954490bc/modeling_rose_x1.py", "https://huggingface.co/GODELEV/Rose-Medium/discussions/1", "https://huggingface.co/GODELEV/Rose-Medium/discussions/2"]}, "provenance_confidence": "medium", "architecture_and_code_lineage": "Standard RoPE/RMSNorm/SwiGLU/GQA with QK normalization plus XSA refresh gate. Frozen implementation explicitly identifies T-X4 family/reference cache design. Owner directly credits AxiomicLabs S2 inspiration in frozen model discussion4; novelty and implementation contribution are not inferred from custom-architecture tag alone.", "training_steps_recorded": 38196, "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-10-31"}
{"model_id": "kali147/bloomz-560m", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved frozen commit history explicitly duplicates bigscience/bloomz-560m. Complete inherited model card credits BigScience and uses upstream repository IDs; no additional publisher training evidence.", "provenance_sources": [{"url": "https://huggingface.co/api/models/kali147/bloomz-560m/commits/80134eb1e0c341003b394adf5eeb0d2e4c0a03a2"}, {"url": "https://huggingface.co/kali147/bloomz-560m/raw/80134eb1e0c341003b394adf5eeb0d2e4c0a03a2/README.md", "basis": "Pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/api/models/kali147/bloomz-560m/commits/80134eb1e0c341003b394adf5eeb0d2e4c0a03a2", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/kali147/bloomz-560m/raw/80134eb1e0c341003b394adf5eeb0d2e4c0a03a2/README.md", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/api/models/kali147/bloomz-560m/discussions?p=0", "access_date": "2026-10-31", "scope": "complete discussion-list JSON read", "finding": "0 listed public threads", "http_status": 200}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "compute_source": "No uploader training established; inherited 64 A100 80GB description is upstream BigScience hardware", "creator_contribution": "Redistribution of BigScience BLOOMZ model; existing duplicate commit evidence reused.", "architecture_lineage": "BigScience pretrained BLOOM-560M followed by upstream BLOOMZ xP3 instruction tuning; this publisher duplicates upstream release.", "evidence_updated_at": "2026-10-31"}
{"model_id": "Norman89107/ryze-125m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card says trained from scratch/custom50k BPE; history identifies Colab checkpoint step23000.", "provenance_sources": [{"url": "https://huggingface.co/Norman89107/ryze-125m/raw/fe3a239cceaa9f6fc470a6f01f2350eab2ed39c7/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Google Colab; exact accelerator unspecified", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Own pretraining on FineWeb sample-10BT and own tokenizer; no instruction tuning documented.", "architecture_lineage": "GPT-2-style decoder-only Transformer, 12 layers/768 hidden/12 heads, own 50,000-token BPE; publisher explicitly reports from-scratch training.", "evidence_updated_at": "2026-11-01"}
{"model_id": "Bwenge840/tinystories-gpt-30m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Bwenge840/tinystories-gpt-30m/raw/bd8f6dee23a18d4e45d8abbf658a99c1d4eb5b14/README.md", "accessed_at": "2026-10-30T12:26:59.057801+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "kerzgrr/Couplet-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card identifies this as own pretraining-only TinyGDN base, with custom hybrid GDN-2/GQA dimensions, explicit corpus exposure, optimizer checkpoint and validation loss. Best-supported own scratch origin. flash-linear-attention/Gated DeltaNet-2 implementation credit remains upstream; no external pretrained weights named.", "provenance_sources": [{"url": "https://huggingface.co/kerzgrr/Couplet-base/raw/b7d7f2d5f33927f9cb06d9ec3da1f6877fc225a4/README.md", "accessed_at": "2026-10-30T12:26:59.076834+00:00"}, {"url": "https://huggingface.co/kerzgrr/Couplet-base/raw/b7d7f2d5f33927f9cb06d9ec3da1f6877fc225a4/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "ThingAI/ARK-50M", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/ThingAI/ARK-50M/raw/c797720cd110c5ffebb148cbb2cc88a98f36d901/README.md", "accessed_at": "2026-10-30T12:26:59.131640+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Tasty-Kiwi/KiwiLM-X", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card expressly identifies final own ModelX TinyStories-Instruct SFTv2, following matched own750K-story base with160,465,920 pretraining targets, custom5.39M hybrid architecture and frozen8K BPE. Selected operation is SFT; own scratch ancestry is a medium supported inference.", "provenance_sources": [{"url": "https://huggingface.co/Tasty-Kiwi/KiwiLM-X/raw/4e8ecb8a46d33d2aa4861c38278482d6ae39efdb/README.md", "accessed_at": "2026-10-30T12:26:59.549707+00:00"}, {"url": "https://huggingface.co/Tasty-Kiwi/KiwiLM-X/raw/4e8ecb8a46d33d2aa4861c38278482d6ae39efdb/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "kipasyangin5/5m-terminal-lm-saturated", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Both5M terminal models explicitly say trained from scratch:100Mtokens Chinchilla run and250M saturated run. Repository includes training_code/main_kaggle.py; exact device and builder country not supplied by card.", "provenance_sources": [{"url": "https://huggingface.co/kipasyangin5/5m-terminal-lm-saturated/raw/a9cdc2d523d0adeb47f0c39eeeb1a94988078490/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Kaggle training script is supplied; exact device not documented in model card.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "kipasyangin5/5m-terminal-lm-chinchilla", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Both5M terminal models explicitly say trained from scratch:100Mtokens Chinchilla run and250M saturated run. Repository includes training_code/main_kaggle.py; exact device and builder country not supplied by card.", "provenance_sources": [{"url": "https://huggingface.co/kipasyangin5/5m-terminal-lm-chinchilla/raw/37688a5230275115971d799b85cb1c693999def8/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Kaggle training script is supplied; exact device not documented in model card.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "achavan1211/minimind-135m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/achavan1211/minimind-135m/raw/b3223c2dbe8ee4e5a3f96460511affe1eda00450/README.md", "accessed_at": "2026-10-30T12:26:59.698651+00:00", "quote": " parameter decoder-only transformer (GPT-2 style) pretrained on OpenWebText from scratch. Model Details | Attribute | Value | |-----------|-------| | Parameters | 134.6M | | Vocab size | 32,000 | | Hidden dim | 768 | | Layers | 12 | | Attention hea", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "momoketchum/nanoMoE", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Explicit scratch-built120.58M sparse MoE with own FineWeb-Edu pretraining,15000steps/~491.5Mtokens and H100 Modal training metrics.", "provenance_sources": [{"url": "https://huggingface.co/momoketchum/nanoMoE/raw/88c2aa016ee3c0f73f9ecce15e128bc8d097c354/README.md", "accessed_at": "2026-10-30T12:26:59.721692+00:00"}, {"url": "https://huggingface.co/momoketchum/nanoMoE/raw/88c2aa016ee3c0f73f9ecce15e128bc8d097c354/README.md", "read_scope": "Saved card model/base/training/architecture and attribution sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "harrrshall/BarunLM-35M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "continued_pretraining_of_own_lineage", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original35.07M Barun custom hybrid base with16384BPE and explicit5.7B realized pretraining tokens, including final4B-token continuation on H200. Supports own scratch-origin lineage; final operation is own continued pretraining.", "provenance_sources": [{"url": "https://huggingface.co/harrrshall/BarunLM-35M/raw/ef3e483a9fd7d906ecf2a7929babeffaf82d1d16/README.md", "accessed_at": "2026-10-30T12:26:59.767905+00:00"}, {"url": "https://huggingface.co/harrrshall/BarunLM-35M/raw/ef3e483a9fd7d906ecf2a7929babeffaf82d1d16/README.md", "read_scope": "Saved card model/base/training/architecture and attribution sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "samueljayasingh/rune-0.3b-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/samueljayasingh/rune-0.3b-base/raw/d66fa7b84249600e286bd76dbf9cbea1d4c5fdaf/README.md", "accessed_at": "2026-10-30T12:27:00.166011+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "LGTEAMai/Nano-1.2-Lite-198M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Nano1.2Lite198M card explicitly from scratch; exact run hardware and location not stated.", "provenance_sources": [{"url": "https://huggingface.co/LGTEAMai/Nano-1.2-Lite-198M/raw/ad77e201cc01f048c87a664d69b90999fc11068f/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Published own small decoder-only language model; corpus/run device/duration not supplied.", "architecture_lineage": "Standard decoder-only GQA Transformer with RoPE/RMSNorm/SwiGLU,13 layers,width1024,16Q/4KV heads,50304 vocabulary; publisher explicitly reports scratch training.", "evidence_updated_at": "2026-11-01"}
{"model_id": "deepakj111/tiny-shakespeare-gpt", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Custom32.1M4-layer128-hidden MoE/GQA model,500iterations on TinyShakespeare with measured train/validation losses; original small-model pretraining supports own scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/deepakj111/tiny-shakespeare-gpt/raw/863b9f84e5d39177ee932e566f366759f8fb11a4/README.md", "accessed_at": "2026-10-30T12:27:00.289025+00:00"}, {"url": "https://huggingface.co/deepakj111/tiny-shakespeare-gpt/raw/863b9f84e5d39177ee932e566f366759f8fb11a4/README.md", "read_scope": "Cached card model/architecture/training/result and upstream description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Krishna0812/Tiny_Stories", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Krishna0812/Tiny_Stories/raw/5d708df735cbef8d8721a4d0f6c4cb78ff7c36ef/README.md", "accessed_at": "2026-10-30T12:27:00.323570+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Pacific-i64/TR-MOE-306", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Pacific-i64/TR-MOE-306/raw/0d2584c11d700ab1f19e12d8a3b561bf7bf5bc7c/README.md", "accessed_at": "2026-10-30T12:27:00.335608+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Pacific-i64/Dense-306", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Pacific-i64/Dense-306/raw/46f10fb75c45e305c3c4ecaa8e93fb1ff46e3605/README.md", "accessed_at": "2026-10-30T12:27:00.389253+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Sandroeth/cali-0.1B", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Sandroeth/cali-0.1B/raw/c9de368e7e62184e2a70e76f4b3cedcfa170b73a/README.md", "accessed_at": "2026-10-30T12:27:00.759586+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "nepetai/Fern-gear-200M-73M-A-IND", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card identifies Base:Fern and a selected industrial ChatML specialization trained on6,256 programming samples for5 epochs, final loss0.40 on M3 MacBook. Best-supported selected operation is task fine-tuning; earliest Fern backbone initialization/ownership is unspecified.", "provenance_sources": [{"url": "https://huggingface.co/nepetai/Fern-gear-200M-73M-A-IND/raw/7379714c2559ffa702993f3740ac760d430c26a4/README.md", "accessed_at": "2026-10-30T12:27:00.849226+00:00"}, {"url": "https://huggingface.co/nepetai/Fern-gear-200M-73M-A-IND/raw/7379714c2559ffa702993f3740ac760d430c26a4/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "analist/poetlm-29m-sft", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Supervised fine-tuning stage of the publisher's own pretraining lineage; step 1500.", "provenance_sources": [{"url": "https://huggingface.co/analist/poetlm-29m-sft/blob/main/README.md"}, {"url": "https://huggingface.co/analist/poetlm-29m-sft/raw/4405b8ec610f5b128511d6d462047862508d249f/README.md", "accessed_at": "2026-10-30T12:27:00.932586+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "analist/poetlm-29m-pretrain", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Own pretraining checkpoint; saved config identifies pretrain stage at step 6000.", "provenance_sources": [{"url": "https://huggingface.co/analist/poetlm-29m-pretrain/blob/main/README.md"}, {"url": "https://huggingface.co/analist/poetlm-29m-pretrain/raw/0bcdfac0b1fe2e28e0e22d22cbc72df3a2dfc95f/README.md", "accessed_at": "2026-10-30T12:27:00.986667+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "oscowlai/Wiola360M", "model_provenance": "unresolved", "own_from_scratch_lineage": "yes", "trained_status": "unknown", "checkpoint_operation": "custom_architecture_checkpoint", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Publisher supplies custom Wiola architecture and inference files but no actual training data/run or weight-origin statement. No model community posts add training information.", "provenance_sources": [{"url": "https://huggingface.co/oscowlai/Wiola360M/raw/2db32a160011698434931472eb436fa5af79653c/README.md", "accessed_at": "2026-10-30T12:27:00.946676+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Publisher supplies custom Wiola architecture and inference files but no actual training data/run or weight-origin statement. No model community posts add training information.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "evidence_updated_at": "2026-11-01"}
{"model_id": "sdkjfgndjfg/better", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sdkjfgndjfg/better/raw/ea256ef016effe3a376bf5ca9196230b7c26e3cd/README.md", "accessed_at": "2026-10-30T12:27:01.028941+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "redredredredredred/slm-from-scratch-77m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/redredredredredred/slm-from-scratch-77m/raw/cdffd17fe94680b26db9273ec2bef445babdeac6/README.md", "accessed_at": "2026-10-30T12:27:01.364744+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Batman55072/NanoChat-28M-V1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/Batman55072/NanoChat-28M-V1/raw/6e23275acf04d284972e42c4c9ba09643660f543/README.md", "accessed_at": "2026-10-30T12:27:01.508764+00:00", "quote": ".8M parameter GPT-style decoder-only transformer, trained fully from scratch, that generates simulated live-stream chat comments -- both regular fast-paced \"hype chat\" and rarer, longer \"superchat\"-style paid messages/questions. Built t", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "PeterRabbit/fineweb-100m-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/PeterRabbit/fineweb-100m-base/raw/016a4c5744d63d4a776f9a4a945d647c3a4220b5/README.md", "accessed_at": "2026-10-30T12:27:01.579091+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "PeterRabbit/fineweb-12m-sft", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/PeterRabbit/fineweb-12m-sft/raw/9505dc321f7e30851389112a300148a4db4a0a10/README.md", "accessed_at": "2026-10-30T12:27:01.605112+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "PeterRabbit/fineweb-12m-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/PeterRabbit/fineweb-12m-base/raw/fd0b9f362cef005ebdb2a82518aeae0976d40995/README.md", "accessed_at": "2026-10-30T12:27:01.617133+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "PeterRabbit/fineweb-10m-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/PeterRabbit/fineweb-10m-base/raw/559cb29208882e3106651ee5d50b4050e6360dc6/README.md", "accessed_at": "2026-10-30T12:27:01.675391+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "PeterRabbit/fineweb-100m-sft", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/PeterRabbit/fineweb-100m-sft/raw/dd3593b9da2c571c94e528a01c3bc24c2c09e96a/README.md", "accessed_at": "2026-10-30T12:27:02.011386+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sarimahsan101/nova-141m-tinystories", "model_provenance": "unresolved", "own_from_scratch_lineage": "unknown", "trained_status": "unknown", "checkpoint_operation": "not_established", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'from_scratch_author_claim', 'detail': 'Rendered personal portfolio explicitly identifies this exact model and says trained from scratch on TinyStories. Pinned card supplies architecture and training-step details.'}", "provenance_sources": [{"url": "https://huggingface.co/sarimahsan101/nova-141m-tinystories/raw/358ddda78a8e136d8120e024c9c9ea32f8bc3f06/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30"}, {"url": "https://github.com/sarimahsan", "basis": "HF_linked_GitHub_profile", "accessed_at": "2026-10-30"}, {"url": "https://github.com/sarimahsan/sarimahsan/blob/main/README.md", "basis": "author_profile_README", "accessed_at": "2026-10-30"}, {"url": "https://syedsarimahsan.netlify.app/", "basis": "author_model_description", "accessed_at": "2026-10-30"}, {"url": "https://syedsarimahsan.netlify.app/Syed_Sarim_Ahsan.pdf", "basis": "author_public_CV", "accessed_at": "2026-10-30T17:52:19Z"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "unknown", "detail": "CV documents T4/Colab/Kaggle for other named research projects; not assigned to this model."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established."}
{"model_id": "kerzgrr/Monostich-2-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card identifies this as own pretraining-only TinyGDN base, with custom hybrid GDN-2/GQA dimensions, explicit corpus exposure, optimizer checkpoint and validation loss. Best-supported own scratch origin. flash-linear-attention/Gated DeltaNet-2 implementation credit remains upstream; no external pretrained weights named.", "provenance_sources": [{"url": "https://huggingface.co/kerzgrr/Monostich-2-base/raw/7779182b4dfad5939827f40cd1c819843a6825e2/README.md", "accessed_at": "2026-10-30T12:27:02.187121+00:00"}, {"url": "https://huggingface.co/kerzgrr/Monostich-2-base/raw/7779182b4dfad5939827f40cd1c819843a6825e2/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "BananaMind/BananaMind-2-Medium", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2-Medium/raw/7d2f397ef1cae15cf84cd6d9b428373238aa3231/README.md"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Medium/raw/7d2f397ef1cae15cf84cd6d9b428373238aa3231/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind", "access_date": "2026-10-31", "scope": "actual rendered organisation card", "finding": "All BananaMind 2 models described as trained from scratch on a single RTX 5070 Ti; shared GQA/QK norm, SwiGLU, RoPE, RMSNorm recipe. Model-specific eight-GPU Unified report takes precedence over series generalisation."}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Medium/raw/7d2f397ef1cae15cf84cd6d9b428373238aa3231/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA RTX 5070 Ti", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Llama-style decoder Transformer / GQA, QK norm, SwiGLU, RoPE and RMSNorm components", "source_url": "https://huggingface.co/BananaMind/BananaMind-2-Medium/raw/7d2f397ef1cae15cf84cd6d9b428373238aa3231/README.md"}], "hardware": "1 x NVIDIA RTX 5070 Ti", "compute_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2-Medium/raw/7d2f397ef1cae15cf84cd6d9b428373238aa3231/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind", "access_date": "2026-10-31", "scope": "actual rendered organisation card", "finding": "All BananaMind 2 models described as trained from scratch on a single RTX 5070 Ti; shared GQA/QK norm, SwiGLU, RoPE, RMSNorm recipe. Model-specific eight-GPU Unified report takes precedence over series generalisation."}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Medium/raw/7d2f397ef1cae15cf84cd6d9b428373238aa3231/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "BananaMind/BananaMind-2-Mini", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2-Mini/raw/6400d0a6dcbe52f3c725291cb97c079c591c05a4/README.md"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Mini/raw/6400d0a6dcbe52f3c725291cb97c079c591c05a4/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind", "access_date": "2026-10-31", "scope": "actual rendered organisation card", "finding": "All BananaMind 2 models described as trained from scratch on a single RTX 5070 Ti; shared GQA/QK norm, SwiGLU, RoPE, RMSNorm recipe. Model-specific eight-GPU Unified report takes precedence over series generalisation."}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Mini/raw/6400d0a6dcbe52f3c725291cb97c079c591c05a4/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA RTX 5070 Ti", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Llama-style decoder Transformer / GQA, QK norm, SwiGLU, RoPE and RMSNorm components", "source_url": "https://huggingface.co/BananaMind/BananaMind-2-Mini/raw/6400d0a6dcbe52f3c725291cb97c079c591c05a4/README.md"}, {"name": "AxiomicLabs GPT-X2-125M", "role": "architecture inspiration explicitly credited in model card", "source_url": "https://huggingface.co/BananaMind/BananaMind-2-Mini/raw/6400d0a6dcbe52f3c725291cb97c079c591c05a4/README.md"}], "hardware": "1 x NVIDIA RTX 5070 Ti", "compute_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2-Mini/raw/6400d0a6dcbe52f3c725291cb97c079c591c05a4/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind", "access_date": "2026-10-31", "scope": "actual rendered organisation card", "finding": "All BananaMind 2 models described as trained from scratch on a single RTX 5070 Ti; shared GQA/QK norm, SwiGLU, RoPE, RMSNorm recipe. Model-specific eight-GPU Unified report takes precedence over series generalisation."}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Mini/raw/6400d0a6dcbe52f3c725291cb97c079c591c05a4/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "BananaMind/BananaMind-2-Nano", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2-Nano/raw/c8564d1bd3f6177221ed7e4f63ae5f281a677a1c/README.md"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Nano/raw/c8564d1bd3f6177221ed7e4f63ae5f281a677a1c/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind", "access_date": "2026-10-31", "scope": "actual rendered organisation card", "finding": "All BananaMind 2 models described as trained from scratch on a single RTX 5070 Ti; shared GQA/QK norm, SwiGLU, RoPE, RMSNorm recipe. Model-specific eight-GPU Unified report takes precedence over series generalisation."}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Nano/raw/c8564d1bd3f6177221ed7e4f63ae5f281a677a1c/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}, {"url": "https://huggingface.co/posts/Banaxi-Tech/955209849515033", "access_date": "2026-10-31", "scope": "actual rendered creator release post", "finding": "Nano trained 30B tokens in about a day on one RTX 5070 Ti 16 GB."}, {"url": "https://huggingface.co/posts/Banaxi-Tech/201821282329459", "access_date": "2026-10-31", "scope": "actual rendered creator post and own replies", "finding": "Nano was an AI-assisted experiment: GPT 5.6 Sol Max wrote training code and ran/monitored it using Mini as architecture reference. Creator explicitly says published Nano is that artifact."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA RTX 5070 Ti (16 GB)", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "training_duration_reported": "About one day, direct creator release post.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Banaxi directed an AI-assisted experiment: GPT 5.6 Sol Max wrote the training code and ran/monitored training using BananaMind 2 Mini architecture as reference. Published Nano is the resulting trained artifact, according to creator.", "upstream_architecture_or_code": [{"name": "Llama-style decoder Transformer / GQA, QK norm, SwiGLU, RoPE and RMSNorm components", "source_url": "https://huggingface.co/BananaMind/BananaMind-2-Nano/raw/c8564d1bd3f6177221ed7e4f63ae5f281a677a1c/README.md"}, {"name": "AxiomicLabs GPT-X2-125M", "role": "architecture inspiration explicitly credited in model card", "source_url": "https://huggingface.co/BananaMind/BananaMind-2-Nano/raw/c8564d1bd3f6177221ed7e4f63ae5f281a677a1c/README.md"}], "hardware": "1 x NVIDIA RTX 5070 Ti (16 GB)", "compute_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2-Nano/raw/c8564d1bd3f6177221ed7e4f63ae5f281a677a1c/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind", "access_date": "2026-10-31", "scope": "actual rendered organisation card", "finding": "All BananaMind 2 models described as trained from scratch on a single RTX 5070 Ti; shared GQA/QK norm, SwiGLU, RoPE, RMSNorm recipe. Model-specific eight-GPU Unified report takes precedence over series generalisation."}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-Nano/raw/c8564d1bd3f6177221ed7e4f63ae5f281a677a1c/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}, {"url": "https://huggingface.co/posts/Banaxi-Tech/955209849515033", "access_date": "2026-10-31", "scope": "actual rendered creator release post", "finding": "Nano trained 30B tokens in about a day on one RTX 5070 Ti 16 GB."}, {"url": "https://huggingface.co/posts/Banaxi-Tech/201821282329459", "access_date": "2026-10-31", "scope": "actual rendered creator post and own replies", "finding": "Nano was an AI-assisted experiment: GPT 5.6 Sol Max wrote training code and ran/monitored it using Mini as architecture reference. Creator explicitly says published Nano is that artifact."}], "evidence_updated_at": "2026-10-31"}
{"model_id": "Lyon28/caca-650M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-650M-untrained/raw/41760b645d41d75c8a644e7cbab69757e7cee6d6/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "wincode/aetherstory", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/wincode/aetherstory/raw/2741fc99d622b81c2e3725918ef10e1df061c74f/README.md", "accessed_at": "2026-10-30T12:27:02.753883+00:00", "quote": "text-generation --- AetherStory > A from-scratch, CPU-trained storyteller transformer. ~863,492 parameters. AetherStory is a tiny decoder-only transformer (GPT-style) trained entirely on CPU on the procedurally generated [AetherStory dataset](wi", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-600M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-600M-untrained/raw/cefbd65270930da5555fe70668ed6cceea51560e/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-550M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-550M-untrained/raw/8a0cfccf4b99ea9b9373290c3cfaea66faa43456/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-500M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-500M-untrained/raw/d27fdc97e9319a80b535137cd378025f521b928f/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-450M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-450M-untrained/raw/bd0405c2fca6378539bd4df9f5617ace315fc643/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-400M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-400M-untrained/raw/4680ddd239061da5089551a9c57385ddb7361103/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-375M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-375M-untrained/raw/450af591877f105d629b77c516a4d47892801135/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-350M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-350M-untrained/raw/d4b1e2fd7200c7391cad1415718c2098a94d8e6b/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-325M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-325M-untrained/raw/e04866610b139bcc93f02f46df76b5a5333ed775/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-300M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-300M-untrained/raw/d212cdb7681109e6126c75895222fe13d10807db/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-275M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-275M-untrained/raw/3d239ee763dfa2764d3d570618144d4679bf6073/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-250M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-250M-untrained/raw/e3d035bfc3a082b0fcd14f769352c7555e0f53ca/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-225M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-225M-untrained/raw/569cc6e0c895ad2ca75526d2bf5cd413575ef46e/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-200M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-200M-untrained/raw/8c0559648e815b83c26e85b756e4f3297eecbdf8/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-175M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-175M-untrained/raw/aa5b4ed8299d7927120a8c7356191c3c84f500c6/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-160M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-160M-untrained/raw/a0bd9e151ce2e658a785b7761c4d536db41f9b38/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-150M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-150M-untrained/raw/b6c79b016dbf5a6873949d793593314d6109635e/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-140M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-140M-untrained/raw/31d3937b764f2c1b25f3af6bbd0cd06301e11859/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-130M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-130M-untrained/raw/8e74d5e0f24fea3ca4c094138f7a3353ed2d5765/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-120M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-120M-untrained/raw/2a5c36b55e0a45121e8d2c72888f022b33d0d441/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-110M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-110M-untrained/raw/c52734a474184a96fd7081c103711d78ad18418a/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-100M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-100M-untrained/raw/0a39414f4ac7b91b95b7fd4c1108371210993412/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-90M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-90M-untrained/raw/b0232f9ff679684411c696cf7d6777d3abcb601c/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-80M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-80M-untrained/raw/fcb773c11b14f0ec5353df6e91f13d9c22fcfc30/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-70M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-70M-untrained/raw/75b719150249625998e87776722592733bbf9162/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-65M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-65M-untrained/raw/46cab7dad61f1227e4ce93e9ded03c67cbb7de9e/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-60M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-60M-untrained/raw/ac75400474a029befed7f6872f749661a1ac44cd/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-55M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-55M-untrained/raw/7e7662f4830a451f1bcf80530609d34b870196ac/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-50M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-50M-untrained/raw/803cf2b38d0b4be332fc091674fb52be7e6e8788/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-45M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-45M-untrained/raw/0b53dded23c20b113a52f1feb50aeeb63cf76b31/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-40M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-40M-untrained/raw/cbd8c4c351eec77cc40c4937793966fdf00ef1bf/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-35M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-35M-untrained/raw/ef34f606ed376215c57baffe0fb19aa1ed330a93/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-30M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-30M-untrained/raw/7ae136a961a651d859bf5ed22fd0f820d0fccfa9/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-27M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-27M-untrained/raw/db46a8b8e1469750ead44b03a99c62e1fe050ebd/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-25M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-25M-untrained/raw/4b53a775f8b01c26474f29febcdd6a1f807d9d93/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-22M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-22M-untrained/raw/422cf6134c0cb7704558fc1443b55b4d95eaec5a/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-20M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-20M-untrained/raw/1d02cdbaa1dca4af49b9fac63441da2d5cc1d027/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-17M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-17M-untrained/raw/63738b441992decb900b75261db98dd4af24ef46/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-15M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-15M-untrained/raw/6f1836c112057c86a85fccb0e8d6ec37334bbfa8/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-14M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-14M-untrained/raw/079b0286f583bc9c2e7aa79775e4167b76279c6c/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-12M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-12M-untrained/raw/c4fdf7c3202e69f5c3280786919cda073f300f1d/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-10M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-10M-untrained/raw/5423ec60440b9902d0a2edd83588154f5565dc1b/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-9M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-9M-untrained/raw/012a0d9af1240c3cc6e2f2b479e63934c33be730/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-8M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-8M-untrained/raw/1cd9a2f93f1c7ebf33ef9fc884389d4e0e3a7d34/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-7M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-7M-untrained/raw/afc371f9650be93d01843cb52dd859d4d6e5bf14/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-6M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-6M-untrained/raw/100e040b11211d7e8a259194e7f47519b0d97bfb/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-5M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-5M-untrained/raw/d08079da5182fa85c4e1b02e29ba1e099aca736c/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-4.5M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-4.5M-untrained/raw/e7f153de3f8f26d0385df10cc3af0d74ae0c1dab/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-4M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-4M-untrained/raw/cb0ea6592386f9e5d30033ce3529be776975f6ce/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-3.5M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-3.5M-untrained/raw/5adb324d300056463bc7fbd9801b94ee6aa255fb/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-3M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-3M-untrained/raw/2931d93c26b3c61c76a79e98454ff1fd930606ad/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-2.5M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-2.5M-untrained/raw/f2fea5a48e0b983d6c0ae17dc22b70957b674c5a/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-2M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-2M-untrained/raw/e2080f0f9e2afb181fd4dd29ad398d7a2ecd4a60/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-1.5M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-1.5M-untrained/raw/b0e1b2a88f6083f12b7b2fdb123e3aa4a71213b8/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lyon28/caca-1M-untrained", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Saved model-specific evidence states the checkpoint is untrained.", "provenance_sources": [{"url": "https://huggingface.co/Lyon28/caca-1M-untrained/raw/ff12cdb8304c8c0817e6d8e25eb891b6a8f85c5c/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "No training performed according to card; profile ownership list not run hardware.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "BananaMind/MathBananaMind-1.1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/MathBananaMind-1.1/raw/4fc1eff1bec83ea0880f53fb9b47cbd52a499703/README.md"}, {"url": "https://huggingface.co/BananaMind/MathBananaMind-1.1/raw/4fc1eff1bec83ea0880f53fb9b47cbd52a499703/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/MathBananaMind-1.1/raw/4fc1eff1bec83ea0880f53fb9b47cbd52a499703/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "compute_source": "Saved CUDA configuration where supplied; no GPU product directly assigned to this run.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Llama-style decoder Transformer / GQA, QK norm, SwiGLU, RoPE and RMSNorm components", "source_url": "https://huggingface.co/BananaMind/MathBananaMind-1.1/raw/4fc1eff1bec83ea0880f53fb9b47cbd52a499703/README.md"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "BananaMind/BananaMind-2-MoE", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2-MoE/raw/ca16baf8186868a2933bc730357324e20cd4c904/README.md"}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-MoE/raw/ca16baf8186868a2933bc730357324e20cd4c904/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind", "access_date": "2026-10-31", "scope": "actual rendered organisation card", "finding": "All BananaMind 2 models described as trained from scratch on a single RTX 5070 Ti; shared GQA/QK norm, SwiGLU, RoPE, RMSNorm recipe. Model-specific eight-GPU Unified report takes precedence over series generalisation."}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-MoE/raw/ca16baf8186868a2933bc730357324e20cd4c904/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA RTX 5070 Ti", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Sparse top-1 mixture of 48 experts per layer, within the shared decoder Transformer recipe", "source_url": "https://huggingface.co/BananaMind/BananaMind-2-MoE/raw/ca16baf8186868a2933bc730357324e20cd4c904/README.md"}], "hardware": "1 x NVIDIA RTX 5070 Ti", "compute_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-2-MoE/raw/ca16baf8186868a2933bc730357324e20cd4c904/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind", "access_date": "2026-10-31", "scope": "actual rendered organisation card", "finding": "All BananaMind 2 models described as trained from scratch on a single RTX 5070 Ti; shared GQA/QK norm, SwiGLU, RoPE, RMSNorm recipe. Model-specific eight-GPU Unified report takes precedence over series generalisation."}, {"url": "https://huggingface.co/BananaMind/BananaMind-2-MoE/raw/ca16baf8186868a2933bc730357324e20cd4c904/checkpoint_metadata.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "kaushik-harsh-99/TinyStories-17M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/kaushik-harsh-99/TinyStories-17M/raw/a215e50ac7d48baafae3470493d7da264572389d/README.md", "accessed_at": "2026-10-30T12:27:09.096606+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "MyLabs-LLC/axion-250m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "234,914,816parameter model trained from scratch on5B FineWebEdu tokens. Config actual path configs/axion_250m.yaml:16layers1024hidden16heads4096FF16384vocab1024contextBF16,5Btrain10Mvalidation tokens.", "provenance_sources": [{"url": "https://huggingface.co/MyLabs-LLC/axion-250m/blob/main/README.md"}, {"url": "https://huggingface.co/MyLabs-LLC/axion-250m/raw/699fa5328a341fa30c4a9f89231ae8f8abe0a883/README.md", "accessed_at": "2026-10-30T12:27:09.176445+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "3 NVIDIA RTX3090, custom3-node PyTorch training, one GPU per node", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Own234.9M scratch pretraining on FineWeb-Edu, custom distributed pipeline and native export.", "architecture_lineage": "Standard decoder-only Transformer with RoPE/RMSNorm/causal SDPA and GELU feed-forward layers; own16K SentencePiece tokenizer, untied embeddings, native PyTorch implementation.", "hardware": "3 NVIDIA RTX3090, custom3-node PyTorch training, one GPU per node", "compute_sources": [{"url": "https://huggingface.co/MyLabs-LLC/axion-250m/raw/699fa5328a341fa30c4a9f89231ae8f8abe0a883/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/api/models/ikdataml/axion-250m", "scope": "Canonical repository identity metadata read", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "ashishk1331/MDLM-TinyShakespeare", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/ashishk1331/MDLM-TinyShakespeare/raw/c094914df5acdca1a4de8315e88b193affc0107e/README.md", "accessed_at": "2026-10-30T12:27:09.455314+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Vcecca/mamba-50m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Vcecca/mamba-50m/raw/c32bbaf42851dca095f8bb3046a5c3e2f62c52f6/README.md", "accessed_at": "2026-10-30T12:27:09.435774+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ismailtasdelen/pinanolm-20m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "probably_yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned card describes own19.8M scratch PyTorch model; pinned training/train.py constructs Pinanolm20mForCausalLM(cfg), trains and exports model.safetensors, with no external pretrained model loading. Root and checkpoints/ weight files are published. Best-supported inference is own scratch training; the exact completed-run metrics are not published.", "provenance_sources": [{"url": "https://huggingface.co/ismailtasdelen/pinanolm-20m/blob/5930c895670369ae27b39a2d6fe75ca1bfdd7817/README.md", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/ismailtasdelen/pinanolm-20m/raw/5930c895670369ae27b39a2d6fe75ca1bfdd7817/training/train.py", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/api/models/ismailtasdelen/pinanolm-20m/tree/main/checkpoints?recursive=true", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01", "provenance_confidence": "medium"}
{"model_id": "sdkjfgndjfg/gpt-model-2-decoder-100000-tiny-stories-fp32", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sdkjfgndjfg/gpt-model-2-decoder-100000-tiny-stories-fp32/raw/f031348fdd5e75ac826f9089a763de35d2184bc3/README.md", "accessed_at": "2026-10-30T12:27:09.620645+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sdkjfgndjfg/gpt-model-2-decoder-100000-tiny-stories-fp16", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sdkjfgndjfg/gpt-model-2-decoder-100000-tiny-stories-fp16/raw/d91a6934ba127f145745baf703f04bd808e5b42f/README.md", "accessed_at": "2026-10-30T12:27:09.720763+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "delcenjo/char-transformer-shakespeare", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/delcenjo/char-transformer-shakespeare/raw/b68adfee4abf7d4a3c1b84d96032f319954b2ef3/README.md", "accessed_at": "2026-10-30T12:27:09.963534+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "anonym5035/XpertGPT-BabyLM2026-Strict-Small", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Authored XpertGPT BabyLM10M-word challenge model with custom52.6M recurrent MoE architecture, own16384BPE and intermediate training revisions; cached publisher profile names coauthor Soham Jain. Supports own scratch-origin pretraining; preserve four-author citation and SRJ5035 alias in card.", "provenance_sources": [{"url": "https://huggingface.co/anonym5035/XpertGPT-BabyLM2026-Strict-Small/raw/3018a103f6de7132793fbf13b61d4971a88f5771/README.md", "accessed_at": "2026-10-30T12:27:10.055760+00:00"}, {"url": "https://huggingface.co/anonym5035/XpertGPT-BabyLM2026-Strict-Small/raw/3018a103f6de7132793fbf13b61d4971a88f5771/README.md", "read_scope": "Cached card model/architecture/training/result and upstream description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Seedyai/MiniGPT-Scratch-5M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Seedyai/MiniGPT-Scratch-5M/raw/ecbf7ff4ba6baf3f08a965df8db384658c031378/README.md", "accessed_at": "2026-10-30T12:27:10.077174+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Own5.2M model/tokenizer and end-to-end educational PyTorch training pipeline; TinyStories is credited dataset, not borrowed backbone.", "architecture_lineage": "Standard small GPT decoder with learned positional embeddings, LayerNorm/FFN/MHA and GPT-2-style tied embeddings; own BPE trained on TinyStories.", "evidence_updated_at": "2026-11-01"}
{"model_id": "Oxpose/Soma-1M", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Upload history explicitly duplicates Martico2432/srlm-1m; original1M Llama on16M SmolLM tokens is separate trainer evidence.", "provenance_sources": [{"url": "https://huggingface.co/api/models/Oxpose/Soma-1M/commits/main"}, {"url": "https://huggingface.co/Oxpose/Soma-1M/raw/45579bcf26cee85392358b28ed63137cacdac238/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "tjeongggg/wave-diffusion-548m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Novel548M attention-free Schrodinger wave-physics masked diffusion architecture, explicit9.99BFineWeb-Edu pretraining over RTX5090week and trained generations support own scratch-origin model.", "provenance_sources": [{"url": "https://huggingface.co/tjeongggg/wave-diffusion-548m/raw/38ec12436500ede106716463fffa91a54dc26f39/README.md", "accessed_at": "2026-10-30T12:27:10.526922+00:00"}, {"url": "https://huggingface.co/tjeongggg/wave-diffusion-548m/raw/38ec12436500ede106716463fffa91a54dc26f39/README.md", "read_scope": "Saved card architecture/base/training/result and upstream attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "kayrahan35/HFP-O1-Memory-Model", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly identifies random, untrained weights and describes training as a future operation.", "provenance_sources": [{"url": "https://huggingface.co/kayrahan35/HFP-O1-Memory-Model/raw/f9b2068a6155a50d747d56bba35ca5cea1d631a8/README.md", "accessed_at": "2026-10-30T12:27:10.459793+00:00", "quote": ") > Status: research preview — architecture only, weights are UNTRAINED. > This repository ships the model code and a randomly-initialized checkpoint so > the architecture can be loaded, inspected and trained. It is not a usable > ", "basis": "Card explicitly identifies random, untrained weights and describes training as a future operation."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "oscowlai/Wiola13M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/oscowlai/Wiola13M/raw/f1cf3ba93e8668744cc0e0b96fe53158bfa8bf9c/README.md", "accessed_at": "2026-10-30T12:27:10.568519+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "oscowlai/wiola-120m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author explicitly describes first-principles Wiola architecture, trained released weights, own WikiText103 training pipeline and validation loss4.1532. Supports own scratch-origin checkpoint.", "provenance_sources": [{"url": "https://huggingface.co/oscowlai/wiola-120m/raw/9292dd717c12650cbde70c9f49065439ac9e04f3/README.md", "accessed_at": "2026-10-30T12:27:10.662089+00:00"}, {"url": "https://huggingface.co/oscowlai/wiola-120m/raw/9292dd717c12650cbde70c9f49065439ac9e04f3/README.md", "read_scope": "Cached model/base/architecture/training and attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "leopard-Ai/rb-nano", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "rb-nano48M explicitly scratch-pretrained50M FineWeb tokens then Alpaca/Dolly/ShareGPT SFT. Card reports RTX4070 8GB. Preserve reported specification. Builder country/compute country not supplied.", "provenance_sources": [{"url": "https://huggingface.co/leopard-Ai/rb-nano/raw/bf46e77577c4bcabdd902a46ffa096b536c32778/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "RTX4070 8GB, as reported", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "my0919175/Sovythos-66M-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/my0919175/Sovythos-66M-Base/raw/778c2b98f57811ad5ce488d79d62cf274db75f78/README.md", "accessed_at": "2026-10-30T12:27:10.913660+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Unjuno/moss-dna-gpt-20m-patens", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original21.69M custom DNA transformer with nucleotide vocabulary and8M-step/~2week training on moss genome supports own scratch-origin trained artifact. Task is DNA next-base prediction, not natural-language dialogue.", "provenance_sources": [{"url": "https://huggingface.co/Unjuno/moss-dna-gpt-20m-patens/raw/5b0e3418760c959291439d8df0ba52b04eed0258/README.md", "accessed_at": "2026-10-30T12:27:11.075652+00:00"}, {"url": "https://huggingface.co/Unjuno/moss-dna-gpt-20m-patens/raw/5b0e3418760c959291439d8df0ba52b04eed0258/README.md", "read_scope": "Saved card architecture/base/training/result and upstream attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Gugu8/Quark-1", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/Gugu8/Quark-1/raw/8ada1a586700a3f6705712578c609b485e1bc25e/README.md", "accessed_at": "2026-10-30T12:27:11.156592+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "officialmobai/rostam", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "continued_pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card explicitly describes sequential continual pretraining and layer stacking14→24→36, followed by instruction stages300k pairs. Original initialization before the first Shahnameh stage is not stated; final V4 retains prior-stage weights.", "provenance_sources": [{"url": "https://huggingface.co/officialmobai/rostam/raw/aad361b38a9d3978ebc49203c522664c6d96ee04/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Profile lists2050Mobile/4060Laptop/RTXPro6000, but exact run hardware not assigned", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Martico2432/srlm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Custom900k ROSA architecture,16M-token SmolLM/FineWeb training and original ROSA source with reported evaluations support own from-scratch pretraining.", "provenance_sources": [{"url": "https://huggingface.co/Martico2432/srlm-1m/blob/main/README.md"}, {"url": "https://huggingface.co/Martico2432/srlm-1m/raw/7b3bddf1b88780b3f20f9c468ee2d9382bef500e/README.md", "accessed_at": "2026-10-30T12:27:11.338855+00:00"}, {"url": "https://huggingface.co/Martico2432/srlm-1m", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "BananaMind/BananaMind-KV1-8M-2Bit-Experimental", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining_with_kv_quantization_aware_training", "provenance_evidence_level": "training_code_and_saved_run_arguments", "provenance_basis": "Reciprocally linked exact source repository train_kv1_sample10bt.py constructs fresh randomly initialized KV1 when resume is empty; frozen run arguments have empty resume and matching custom architecture/data settings. Model card describes actual KV-aware training, rather than a post-training wrapper.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-KV1-8M-2Bit-Experimental/raw/9f8de743335f759cbe303b1067eb3b27d2589418/README.md"}, {"url": "https://huggingface.co/BananaMind/BananaMind-KV1-8M-2Bit-Experimental/raw/9f8de743335f759cbe303b1067eb3b27d2589418/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/BananaMind-KV1-8M-2Bit-Experimental/raw/9f8de743335f759cbe303b1067eb3b27d2589418/train_args.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}, {"url": "https://huggingface.co/BananaMind/BananaMind-KV1-8M-2Bit-Experimental/raw/9f8de743335f759cbe303b1067eb3b27d2589418/modeling_kv1.py", "access_date": "2026-10-31", "scope": "module header and initialization implementation"}, {"url": "https://huggingface.co/posts/Banaxi-Tech/896370235888789", "access_date": "2026-10-31", "scope": "actual rendered creator release post", "finding": "Original 8M KV1 trained under low-bit KV constraint; links exact source code repository Banaxi-Tech/kv1."}, {"url": "https://github.com/Banaxi-Tech/kv1/blob/main/README.md", "access_date": "2026-10-31", "scope": "complete linked code README", "finding": "Reciprocal link identifies the exact frozen 8M experimental model and matching train_kv1_sample10bt.py."}, {"url": "https://github.com/Banaxi-Tech/kv1/blob/main/train_kv1_sample10bt.py", "access_date": "2026-10-31", "scope": "complete 15502-byte training source read", "finding": "Empty resume selects build_fresh_model -> KV1ForCausalLM(new config); model has random-normal initialization. Exact frozen train_args.json records empty resume, matching architecture/data/training settings. Actual new-initialisation training supported by code and saved run arguments."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Exact GPU product, elapsed training time and physical compute country are not reported for this specific run.; Builder institutional route unstated.", "compute_source": "Saved CUDA configuration where supplied; no GPU product directly assigned to this run.", "upstream_checkpoint": "No pretrained base; code path and saved empty-resume arguments support fresh custom model initialization.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Custom decoder implementation with low-bit KV training and packed generation cache", "source_url": "https://huggingface.co/posts/Banaxi-Tech/896370235888789"}, {"name": "BitNet approach", "role": "creator describes low-bit training inspiration, not pretrained BitNet checkpoint reuse", "source_url": "https://huggingface.co/posts/Banaxi-Tech/896370235888789"}], "best_origin_estimate": {"origin": "own_from_scratch", "confidence": "high", "direct_or_inferred": "code_and_saved_configuration", "basis": "Reciprocally linked exact source repository train_kv1_sample10bt.py constructs fresh randomly initialized KV1 when resume is empty; frozen run arguments have empty resume and matching custom architecture/data settings. Model card describes actual KV-aware training, rather than a post-training wrapper."}, "evidence_updated_at": "2026-10-31"}
{"model_id": "Rapnss/DevOps-Ultra-125M", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/Rapnss/DevOps-Ultra-125M/raw/15e5fc7bcc4e81f211320f4b2c0569bc5b9f7804/README.md", "accessed_at": "2026-10-30T12:27:11.643077+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "aman0419/Vitallm-50M", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/aman0419/Vitallm-50M/raw/f9ad4d9c4fad59a8843f3c843db2869b3d56e21b/README.md", "accessed_at": "2026-10-30T12:27:11.713445+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "BananaMind/MicroBananaMind-v1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/MicroBananaMind-v1/raw/2dd299a219ffabc05bfcd39862e9293c68d3ef92/README.md"}, {"url": "https://huggingface.co/BananaMind/MicroBananaMind-v1/raw/2dd299a219ffabc05bfcd39862e9293c68d3ef92/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/MicroBananaMind-v1/raw/2dd299a219ffabc05bfcd39862e9293c68d3ef92/train_args.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "compute_source": "Saved CUDA configuration where supplied; no GPU product directly assigned to this run.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Model-card-described Llama-style decoder Transformer", "source_url": "https://huggingface.co/BananaMind/MicroBananaMind-v1/raw/2dd299a219ffabc05bfcd39862e9293c68d3ef92/README.md"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "BananaMind/MiniBananaMind-v4-9M", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "continued_pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact card starts from MiniBananaMind-v3-9M and continues on Cosmopedia-v2; original base pretraining is separate.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/MiniBananaMind-v4-9M/raw/013c7e8cb34dc4ba19fe4b389468c3501eee4ed3/README.md"}, {"url": "https://huggingface.co/BananaMind/MiniBananaMind-v4-9M/raw/013c7e8cb34dc4ba19fe4b389468c3501eee4ed3/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/MiniBananaMind-v4-9M/raw/013c7e8cb34dc4ba19fe4b389468c3501eee4ed3/train_args.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA RTX 5070 Ti", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "MiniBananaMind-v3-9M own previously pretrained checkpoint; explicitly continued on Cosmopedia-v2.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Model-card-described Llama-style decoder Transformer", "source_url": "https://huggingface.co/BananaMind/MiniBananaMind-v4-9M/raw/013c7e8cb34dc4ba19fe4b389468c3501eee4ed3/README.md"}], "hardware": "NVIDIA RTX 5070 Ti", "compute_sources": [{"url": "https://huggingface.co/BananaMind/MiniBananaMind-v4-9M/raw/013c7e8cb34dc4ba19fe4b389468c3501eee4ed3/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/MiniBananaMind-v4-9M/raw/013c7e8cb34dc4ba19fe4b389468c3501eee4ed3/train_args.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "BananaMind/BananaMind-1.5-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-1.5-Base/raw/78d74d706736dc983e01e3019a7ffab139f9b84e/README.md"}, {"url": "https://huggingface.co/BananaMind/BananaMind-1.5-Base/raw/78d74d706736dc983e01e3019a7ffab139f9b84e/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/BananaMind-1.5-Base/raw/78d74d706736dc983e01e3019a7ffab139f9b84e/training_summary.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA RTX PRO 6000", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Model-card-described Llama-style decoder Transformer", "source_url": "https://huggingface.co/BananaMind/BananaMind-1.5-Base/raw/78d74d706736dc983e01e3019a7ffab139f9b84e/README.md"}], "hardware": "NVIDIA RTX PRO 6000", "compute_sources": [{"url": "https://huggingface.co/BananaMind/BananaMind-1.5-Base/raw/78d74d706736dc983e01e3019a7ffab139f9b84e/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/BananaMind-1.5-Base/raw/78d74d706736dc983e01e3019a7ffab139f9b84e/training_summary.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "Multisymboliccore/fields-300m-pg19", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes first public own Fields LM checkpoint with bespoke Field recurrence/PCAF/local refresh/Mamba-2 architecture, own16K BPE and49,152,000-token PG19 controlled training. Strong own fresh-pretraining inference; Mamba-2/CUDA dependencies remain upstream components.", "provenance_sources": [{"url": "https://huggingface.co/Multisymboliccore/fields-300m-pg19/raw/476fb068b2efd752c983e098afb4a7df565c52dd/README.md", "accessed_at": "2026-10-30T12:27:11.964539+00:00"}, {"url": "https://huggingface.co/Multisymboliccore/fields-300m-pg19/raw/476fb068b2efd752c983e098afb4a7df565c52dd/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "BananaMind/MiniBananaMind-v3-9M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/MiniBananaMind-v3-9M/raw/e4b6c4935d92ae4a90c618238fef28510984e95d/README.md"}, {"url": "https://huggingface.co/BananaMind/MiniBananaMind-v3-9M/raw/e4b6c4935d92ae4a90c618238fef28510984e95d/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/MiniBananaMind-v3-9M/raw/e4b6c4935d92ae4a90c618238fef28510984e95d/train_args.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA RTX 5070 Ti", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Model-card-described Llama-style decoder Transformer", "source_url": "https://huggingface.co/BananaMind/MiniBananaMind-v3-9M/raw/e4b6c4935d92ae4a90c618238fef28510984e95d/README.md"}], "hardware": "NVIDIA RTX 5070 Ti", "compute_sources": [{"url": "https://huggingface.co/BananaMind/MiniBananaMind-v3-9M/raw/e4b6c4935d92ae4a90c618238fef28510984e95d/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}, {"url": "https://huggingface.co/BananaMind/MiniBananaMind-v3-9M/raw/e4b6c4935d92ae4a90c618238fef28510984e95d/train_args.json", "access_date": "2026-10-31", "scope": "complete small training metadata or notice file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "ostah-1010/Marshmello-8M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/ostah-1010/Marshmello-8M/raw/2013713e2960b5c657167b295960dab22cb996ab/README.md", "accessed_at": "2026-10-30T12:27:12.377688+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "BananaMind/MiniBananaMind-v2-9M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/BananaMind/MiniBananaMind-v2-9M/raw/29c4571df82ec1fef931e4c4ce756e634b9cea64/README.md"}, {"url": "https://huggingface.co/BananaMind/MiniBananaMind-v2-9M/raw/29c4571df82ec1fef931e4c4ce756e634b9cea64/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA RTX 5070 Ti", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country at this historical run and physical compute country are not directly established.; University attendance and institutional affiliation remain unstated for main builder Banaxi.", "training_duration_reported": "4 hours 34 minutes, publisher report.", "compute_source": "Publisher model card / actual rendered organisation statement; model-specific source recorded.", "upstream_checkpoint": "No external pretrained checkpoint identified in reviewed model evidence; from-scratch status follows the model-specific provenance assessment.", "creator_contribution": "Own small-model training experiment and model-specific implementation/settings; architecture recipe and upstream inspiration remain distinct from trained weight origin.", "upstream_architecture_or_code": [{"name": "Model-card-described Llama-style decoder Transformer", "source_url": "https://huggingface.co/BananaMind/MiniBananaMind-v2-9M/raw/29c4571df82ec1fef931e4c4ce756e634b9cea64/README.md"}], "hardware": "NVIDIA RTX 5070 Ti", "compute_sources": [{"url": "https://huggingface.co/BananaMind/MiniBananaMind-v2-9M/raw/29c4571df82ec1fef931e4c4ce756e634b9cea64/README.md", "access_date": "2026-10-31", "scope": "frozen model card introduction, training, architecture, upstream credits, hardware, energy and licensing sections read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "AI-Sweden-Models/gpt-sw3-356m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card identifies this as the AI Sweden/RISE GPT-SW3 base model trained on the 320B-token Nordic corpus with NeMo Megatron. The project developer explicitly states the 126M–20B family was trained from scratch.", "provenance_sources": [{"url": "https://huggingface.co/AI-Sweden-Models/gpt-sw3-356m/blob/cce4178148d228c27a29c5a2ee4ce5ecbe5775bb/README.md", "access_date": "2026-11-01"}, {"url": "https://github.com/huggingface/transformers/issues/20176", "access_date": "2026-11-01", "finding": "AI Sweden project developer describes models126M–20B as trained from scratch."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high"}
{"model_id": "AI-Sweden-Models/gpt-sw3-126m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card identifies this as the AI Sweden/RISE GPT-SW3 base model trained on the 320B-token Nordic corpus with NeMo Megatron. The project developer explicitly states the 126M–20B family was trained from scratch.", "provenance_sources": [{"url": "https://huggingface.co/AI-Sweden-Models/gpt-sw3-126m/blob/8b8d842903b523bb76c1d814907f49cb2b6ddc59/README.md", "access_date": "2026-11-01"}, {"url": "https://github.com/huggingface/transformers/issues/20176", "access_date": "2026-11-01", "finding": "AI Sweden project developer describes models126M–20B as trained from scratch."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high"}
{"model_id": "Sherwinroger002/gpt-124m-param-1.6b-tokens", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Sherwinroger002/gpt-124m-param-1.6b-tokens/raw/fdd6ab7a8d3185b5c99224f73396f6abc94a4e8d/README.md", "accessed_at": "2026-10-30T12:27:12.594622+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Aadit-032/LiteGPT-16M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Aadit-032/LiteGPT-16M/blob/main/README.md"}, {"url": "https://huggingface.co/Aadit-032/LiteGPT-16M/raw/9e30716ae28b21e6450ee0d5e9697bedc8b6c64d/README.md", "accessed_at": "2026-10-30T12:27:12.933185+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "m1357l/iconshop-svg-generator", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author reports training a custom IconShop autoregressive SVG decoder for5 epochs on1.5M FIGR icons using4RTX3090s with final losses. BERT text tokenization is documented separately; no pretrained decoder weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/m1357l/iconshop-svg-generator/raw/3a7386372b14acaeb8188b859d3054c26caf31c0/README.md", "accessed_at": "2026-10-30T12:27:13.150788+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Tripstoph/RRT-Foundation", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Tripstoph/RRT-Foundation/raw/4f983a98ff7278367f36bd3b76280dd5503b005a/README.md", "accessed_at": "2026-10-30T12:27:13.056653+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "mkurman/NeuroBLAST-V3-SYNTH-EC-150000", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original ConvGPT/NeuroBLAST custom architecture and authored SYNTH pretraining checkpoints with118B-250B tokens on TPUv4-16 support own scratch-origin lineage; JAX release retains same authors150k-step checkpoint lineage.", "provenance_sources": [{"url": "https://huggingface.co/mkurman/NeuroBLAST-V3-SYNTH-EC-150000/raw/1d76ec8d54c95e3523434f2bc98aa2f9832c6797/README.md", "accessed_at": "2026-10-30T12:27:13.104221+00:00"}, {"url": "https://huggingface.co/mkurman/NeuroBLAST-V3-SYNTH-EC-150000/raw/1d76ec8d54c95e3523434f2bc98aa2f9832c6797/README.md", "read_scope": "Saved card architecture, training and checkpoint description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "mkurman/NeuroBLAST-V3-0.6B-SYNTH-EC-144B-TOK", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original ConvGPT/NeuroBLAST custom architecture and authored SYNTH pretraining checkpoints with118B-250B tokens on TPUv4-16 support own scratch-origin lineage; JAX release retains same authors150k-step checkpoint lineage.", "provenance_sources": [{"url": "https://huggingface.co/mkurman/NeuroBLAST-V3-0.6B-SYNTH-EC-144B-TOK/raw/1e14f11ddb2856513bf188bcff7361682b019cc5/README.md", "accessed_at": "2026-10-30T12:27:13.190490+00:00"}, {"url": "https://huggingface.co/mkurman/NeuroBLAST-V3-0.6B-SYNTH-EC-144B-TOK/raw/1e14f11ddb2856513bf188bcff7361682b019cc5/README.md", "read_scope": "Saved card architecture, training and checkpoint description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "korkmazumut/orkhon-bunghu", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly describes this assistant as tuned from the own-family bilingual bengu branch.", "provenance_sources": [{"url": "https://huggingface.co/korkmazumut/orkhon-bunghu/raw/26f8c3c1d2c933ea20d4f020248c90f70cb5cec9/README.md", "accessed_at": "2026-10-30T12:27:13.169854+00:00"}, {"url": "https://huggingface.co/korkmazumut/orkhon-bunghu/raw/26f8c3c1d2c933ea20d4f020248c90f70cb5cec9/README.md", "access_date": "2026-11-01", "scope": "Saved model-card introduction and training/family sections read", "finding": "Card explicitly describes this assistant as tuned from the own-family bilingual bengu branch."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "korkmazumut/orkhon-istem", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly describes this assistant as tuned from the own-family FineWeb-Edu istemi base.", "provenance_sources": [{"url": "https://huggingface.co/korkmazumut/orkhon-istem/raw/810b1d93399afa2020c877cc89e5e7577516e332/README.md", "accessed_at": "2026-10-30T12:27:13.533461+00:00"}, {"url": "https://huggingface.co/korkmazumut/orkhon-istem/raw/810b1d93399afa2020c877cc89e5e7577516e332/README.md", "access_date": "2026-11-01", "scope": "Saved model-card introduction and training/family sections read", "finding": "Card explicitly describes this assistant as tuned from the own-family FineWeb-Edu istemi base."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "korkmazumut/orkhon-tegin", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly describes this assistant as tuned from the own-family former Kultegin instruct model.", "provenance_sources": [{"url": "https://huggingface.co/korkmazumut/orkhon-tegin/raw/f7f14f247a902f33d8010c58a3b5fbd9b1fd2208/README.md", "accessed_at": "2026-10-30T12:27:13.687233+00:00"}, {"url": "https://huggingface.co/korkmazumut/orkhon-tegin/raw/f7f14f247a902f33d8010c58a3b5fbd9b1fd2208/README.md", "access_date": "2026-11-01", "scope": "Saved model-card introduction and training/family sections read", "finding": "Card explicitly describes this assistant as tuned from the own-family former Kultegin instruct model."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "korkmazumut/orkhon-tonyuk", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly describes this assistant as tuned from the own-family Tonyukuk story base.", "provenance_sources": [{"url": "https://huggingface.co/korkmazumut/orkhon-tonyuk/raw/6ffd4d9ecbf55553d87b3ac36da8799b4b383706/README.md", "accessed_at": "2026-10-30T12:27:13.705259+00:00"}, {"url": "https://huggingface.co/korkmazumut/orkhon-tonyuk/raw/6ffd4d9ecbf55553d87b3ac36da8799b4b383706/README.md", "access_date": "2026-11-01", "scope": "Saved model-card introduction and training/family sections read", "finding": "Card explicitly describes this assistant as tuned from the own-family Tonyukuk story base."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "korkmazumut/orkhon-bumin-mini", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "supervised_finetuning_of_own_base", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned manifest explicitly states SFT from the original Bumin smoke checkpoint, with runs/bumin_mini_unified; original Bumin belongs to the publisher own scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/korkmazumut/orkhon-bumin-mini/raw/59aff438a827159f4cdd411cc2fdef7821b82be9/manifest.json", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/korkmazumut/orkhon-bumin-mini/raw/59aff438a827159f4cdd411cc2fdef7821b82be9/orkhon_model_card.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01", "upstream_checkpoint": "korkmazumut/orkhon-bumin"}
{"model_id": "TilelliLab/Yaz", "model_provenance": "derived", "own_from_scratch_lineage": "partial", "trained_status": "yes", "checkpoint_operation": "custom_model_with_pretrained_embedding_router", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned card explicitly states semantic routing uses frozen all-MiniLM-L6-v2 embeddings; released editable tiny knowledge model therefore includes pretrained component reuse.", "provenance_sources": [{"url": "https://huggingface.co/TilelliLab/Yaz/blob/48db9f238e43d4256a6b5b7b57da5cb5bf5e1589/README.md", "access_date": "2026-11-01", "finding": "Saved frozen card read; frozen MiniLM semantic router stated explicitly."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "upstream_checkpoint": "sentence-transformers/all-MiniLM-L6-v2"}
{"model_id": "korkmazumut/orkhon-bengu-gokturk", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/korkmazumut/orkhon-bengu-gokturk/raw/0335a14537531b8465714fea5666f30386b3e6bf/README.md", "accessed_at": "2026-10-30T12:27:13.830422+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "korkmazumut/orkhon-bengu", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies this trained base checkpoint as a member of the author's from-scratch Orkhon training stack; base role and evaluation are documented.", "provenance_sources": [{"url": "https://huggingface.co/korkmazumut/orkhon-bengu/raw/58b28eedf8e9927dfbad3813d1292e1bde83064a/README.md", "accessed_at": "2026-10-30T12:27:14.164401+00:00"}, {"url": "https://huggingface.co/korkmazumut/orkhon-bengu/raw/58b28eedf8e9927dfbad3813d1292e1bde83064a/README.md", "access_date": "2026-11-01", "scope": "Saved model-card introduction and training/family sections read", "finding": "Card identifies this trained base checkpoint as a member of the author's from-scratch Orkhon training stack; base role and evaluation are documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "korkmazumut/orkhon-istemi", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies this trained base checkpoint as a member of the author's from-scratch Orkhon training stack; base role and evaluation are documented.", "provenance_sources": [{"url": "https://huggingface.co/korkmazumut/orkhon-istemi/raw/6fab7638beba09d9df3dc57df1823b9fb6924862/README.md", "accessed_at": "2026-10-30T12:27:14.318417+00:00"}, {"url": "https://huggingface.co/korkmazumut/orkhon-istemi/raw/6fab7638beba09d9df3dc57df1823b9fb6924862/README.md", "access_date": "2026-11-01", "scope": "Saved model-card introduction and training/family sections read", "finding": "Card identifies this trained base checkpoint as a member of the author's from-scratch Orkhon training stack; base role and evaluation are documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "korkmazumut/orkhon-kultigin", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/korkmazumut/orkhon-kultigin/raw/8c64e303f08220104b2837a31ed4c29125748029/README.md", "accessed_at": "2026-10-30T12:27:14.322422+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "korkmazumut/orkhon-tonyukuk", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/korkmazumut/orkhon-tonyukuk/raw/61e60769722467a7a11cb2641a92251d186baa25/README.md", "accessed_at": "2026-10-30T12:27:14.372874+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "korkmazumut/orkhon-bumin", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "training_from_initialisation_on_synthetic_qa", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned manifest identifies the first Orkhon4M from-scratch transformer,trained end-to-end on synthetic arithmeticQ&A,with original runs/sft_smoke checkpoint and held-out validation loss1.6348.", "provenance_sources": [{"url": "https://huggingface.co/korkmazumut/orkhon-bumin/raw/3b9bdae6b24dfda91722c5cdda46a1064c7d4683/manifest.json", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/korkmazumut/orkhon-bumin/raw/3b9bdae6b24dfda91722c5cdda46a1064c7d4683/orkhon_model_card.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "AiRukua/BabyKilat", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_partial_checkpoint", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "Explicit from-scratch claim; partial checkpoint at 6,000 of 50,000 planned steps.", "provenance_sources": [{"url": "https://huggingface.co/AiRukua/BabyKilat/blob/14624fc40cae3ef0831a6cc22d677f32f15e55f2/README.md", "accessed_at": "2026-10-30"}, {"url": "https://github.com/Airukua", "basis": "Current HF-linked GitHub location; profile updated 2026-10-27, but field update date itself unknown.", "accessed_at": "2026-10-30"}, {"url": "https://www.linkedin.com/pub/dir/Abdul%2BWahid/%2B", "basis": "LinkedIn's public directory for the same linked author; conflicting location retained.", "accessed_at": "2026-10-30"}, {"url": "https://id.linkedin.com/in/abdul-wahid-rukua", "basis": "LinkedIn education lists PGDip Artificial Intelligence, 2024–2026; affiliation is not residence.", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/AiRukua/BabyKilat/blob/main/README.md"}, {"url": "https://huggingface.co/AiRukua/BabyKilat/raw/14624fc40cae3ef0831a6cc22d677f32f15e55f2/README.md", "accessed_at": "2026-10-30T12:27:14.452907+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Actual training device not unambiguously stated. Card estimates times for RTX 6000 Blackwell, RTX 4050 and Colab T4.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Sherwinroger002/gpt-124m-param-600m-tokens", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Sherwinroger002/gpt-124m-param-600m-tokens/raw/6c8b27738fef5cc34daee0198a241ff066e5ab15/README.md", "accessed_at": "2026-10-30T12:27:14.792890+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "uditmukherjee/glassbox-gpt-shakespeare", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/uditmukherjee/glassbox-gpt-shakespeare/raw/b775ad41d0de86eb328e3d6a2d2b6ca585e10e9c/README.md", "accessed_at": "2026-10-30T12:27:14.960397+00:00", "quote": "assbox GPT — Shakespeare (char-level) A small GPT built entirely from scratch — autograd engine, attention, the lot — and trained on character-level Tiny Shakespeare. It's tiny (10.8M parameters) and it only knows characters, yet it writ", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Abiray/Kshana-170M-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Frozen README explicitly calls Kshana trained from scratch; owner answers hardware question for this model.", "provenance_sources": [{"url": "https://huggingface.co/Abiray/Kshana-170M-Base/blob/main/README.md"}, {"url": "https://huggingface.co/Abiray/Kshana-170M-Base/raw/e46f4b167b3045c539d8b010bee3d7e5c8f76d1a/README.md", "accessed_at": "2026-10-30T12:27:14.926775+00:00"}, {"url": "https://huggingface.co/Abiray/Kshana-170M-Base/raw/e46f4b167b3045c539d8b010bee3d7e5c8f76d1a/README.md", "access_date": "2026-10-31", "scope": "complete frozen README read", "excerpt": "Kshana is trained from scratch", "finding": "Publisher explicitly describes own 169,906,752-parameter Llama-style GQA base model; card reports 65B training tokens and AdamW / cosine / mixed precision.", "confidence": "high as publisher statement"}, {"url": "https://huggingface.co/Abiray/Kshana-170M-Base/discussions/1#6a2d0a0c4e8b3b1de6fe54ea", "access_date": "2026-10-31", "event_date": "2026-06-13T07:43:08Z", "scope": "complete thread: initial question and owner reply, browser plus full JSON payload", "excerpt": "I used an RTX 5080 16GB card for it. It took me a little more than 8 days of continuous training.", "finding": "Owner reports actual card and approximate continuous training duration.", "confidence": "high as direct owner statement"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA GeForce RTX 5080 16GB", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Builder country and place during training; GPU ownership/rental/provider and physical country; University attendance, employment affiliation and project backing not stated in the read sources", "training_duration_reported": "a little more than 8 days of continuous training", "training_tokens_publisher_reported": 65000000000, "upstream_checkpoint": "No pretrained checkpoint reuse stated; frozen card explicitly claims training from scratch.", "creator_contribution": "Own from-scratch pretraining run claimed, with owner-reported RTX508016GB hardware and continuous8+day duration. Standard Llama architectural lineage is separate from weight initialization.", "upstream_architecture_or_code": [{"name": "Llama-style decoder architecture", "attribution_location": "frozen model card opening and architecture section", "source_url": "https://huggingface.co/Abiray/Kshana-170M-Base/raw/e46f4b167b3045c539d8b010bee3d7e5c8f76d1a/README.md"}], "presentation_tags": ["custom-architecture", "llama", "gqa"], "hardware": "NVIDIA GeForce RTX 5080 16GB", "actual_training_hardware_count": 1, "compute_sources": [{"url": "https://huggingface.co/Abiray/Kshana-170M-Base/raw/e46f4b167b3045c539d8b010bee3d7e5c8f76d1a/README.md", "access_date": "2026-10-31", "scope": "complete frozen README read", "excerpt": "Kshana is trained from scratch", "finding": "Publisher explicitly describes own 169,906,752-parameter Llama-style GQA base model; card reports 65B training tokens and AdamW / cosine / mixed precision.", "confidence": "high as publisher statement"}, {"url": "https://huggingface.co/Abiray/Kshana-170M-Base/discussions/1#6a2d0a0c4e8b3b1de6fe54ea", "access_date": "2026-10-31", "event_date": "2026-06-13T07:43:08Z", "scope": "complete thread: initial question and owner reply, browser plus full JSON payload", "excerpt": "I used an RTX 5080 16GB card for it. It took me a little more than 8 days of continuous training.", "finding": "Owner reports actual card and approximate continuous training duration.", "confidence": "high as direct owner statement"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "hranjan043/himbot-gpt-level2", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/hranjan043/himbot-gpt-level2/raw/3101cadc9cab96f8b1e7855172285cc3b7cb0e3c/README.md", "accessed_at": "2026-10-30T12:27:15.037235+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sapbot/7.6M-Chaos-Engine", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Published training_info records a completed approximately 6-hour run on a bespoke tiny Qwen2 configuration (8 layers, width256, vocab8000), with custom mixed text sources described in the card. Best-supported own scratch origin, medium confidence; earliest initialization script is absent.", "provenance_sources": [{"url": "https://huggingface.co/sapbot/7.6M-Chaos-Engine/raw/a3a74e8fd7665dd512b3dfe8e2b66840bdcf3713/README.md", "accessed_at": "2026-10-30T12:27:15.044750+00:00"}, {"url": "https://huggingface.co/sapbot/7.6M-Chaos-Engine/raw/a3a74e8fd7665dd512b3dfe8e2b66840bdcf3713/README.md", "scope": "Complete saved frozen card read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/sapbot/7.6M-Chaos-Engine/raw/a3a74e8fd7665dd512b3dfe8e2b66840bdcf3713/config.json", "scope": "Complete configuration read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/sapbot/7.6M-Chaos-Engine/raw/a3a74e8fd7665dd512b3dfe8e2b66840bdcf3713/training_info.txt", "scope": "Complete training note read", "excerpt": "Time: ~6h", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "training_duration_reported": "~6h", "creator_contribution": "Tiny custom-configuration model trained on mixed Dutch TLDR, Pushkin, Genius Makers and arXiv text, per publisher card/training note.", "best_origin_estimate": {"origin": "own_from_scratch_backbone", "confidence": "medium", "basis": "Tiny custom Qwen2 configuration, custom text mix and completed run note, no identified external checkpoint.", "remaining_gap": "No original initialization code."}, "provenance_confidence": "medium", "architecture_lineage": "Standard Qwen2ForCausalLM architecture scaled to 8 layers/width256/vocabulary8000; no architectural novelty claimed in this assessment.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "compute_sources": [{"url": "https://huggingface.co/sapbot/7.6M-Chaos-Engine/raw/a3a74e8fd7665dd512b3dfe8e2b66840bdcf3713/README.md", "scope": "Complete saved frozen card read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/sapbot/7.6M-Chaos-Engine/raw/a3a74e8fd7665dd512b3dfe8e2b66840bdcf3713/config.json", "scope": "Complete configuration read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/sapbot/7.6M-Chaos-Engine/raw/a3a74e8fd7665dd512b3dfe8e2b66840bdcf3713/training_info.txt", "scope": "Complete training note read", "excerpt": "Time: ~6h", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "AnthonyPa57/CirillaMTP-0.1B-3E", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card describes a final domain-specific fine-tuning stage, so this published checkpoint is a post-training release.", "provenance_sources": [{"url": "https://huggingface.co/AnthonyPa57/CirillaMTP-0.1B-3E/raw/c5cf4da5e52606e52f436f0000abbd0fb938b3d2/README.md", "accessed_at": "2026-10-30T12:27:15.093826+00:00", "quote": "ngside lore retention. (21MiB, 50.5K data points) Domain-Specific Fine-Tuning The final stage focused on activating the model's conversational abilities and aligning it with the ingested lore. This involved training on 185K synthetic que", "basis": "Card describes a final domain-specific fine-tuning stage, so this published checkpoint is a post-training release."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Banaxi-Tech/BananaMind-Completor-V1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Publisher calls this local GPT-2 training; exact 1.93M custom byte model dimensions, 256 byte vocabulary, local autocomplete_model_gpt2_best.pt,70000 training steps. No external pretrained base named.", "provenance_sources": [{"url": "https://huggingface.co/Banaxi-Tech/BananaMind-Completor-V1/raw/454c7fa5a18e047a53a693a547712dd5e8895156/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Exact initial-run receipt, training GPU, duration and physical country unstated.; Later Austrian base estimate does not directly establish place during June4 Completor run.", "compute_source": "Local training is stated; exact device is not assigned to this earlier June run.", "upstream_checkpoint": "Own autocomplete_model_gpt2_best.pt named. No external pretrained checkpoint identified.", "creator_contribution": "Own byte-level autocomplete training experiment using standard GPT-2 implementation, 70000-step best validation checkpoint and native Transformers packaging.", "upstream_architecture_or_code": [{"name": "Hugging Face Transformers GPT2LMHeadModel", "role": "standard model implementation / recipe", "source_url": "https://huggingface.co/Banaxi-Tech/BananaMind-Completor-V1/raw/454c7fa5a18e047a53a693a547712dd5e8895156/README.md"}, {"name": "agentlans/high-quality-english-sentences", "role": "training dataset explicitly attributed in NOTICE; ODC-BY 1.0", "source_url": "https://huggingface.co/Banaxi-Tech/BananaMind-Completor-V1/raw/454c7fa5a18e047a53a693a547712dd5e8895156/NOTICE"}], "best_origin_estimate": {"origin": "own_from_scratch", "confidence": "medium", "direct_or_inferred": "inferred_from_saved_development_evidence", "basis": "Publisher calls this local GPT-2 training; exact 1.93M custom byte model dimensions, 256 byte vocabulary, local autocomplete_model_gpt2_best.pt,70000 training steps. No external pretrained base named.", "remaining_gap": "Training entrypoint/first initialization receipt is not included; constructor default configuration alone is not direct proof."}, "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-10-31"}
{"model_id": "dr-tkxx/Enigma", "model_provenance": "unresolved", "own_from_scratch_lineage": "yes", "trained_status": "not_documented", "checkpoint_operation": "custom_architecture_checkpoint", "provenance_evidence_level": "unresolved", "provenance_basis": "Custom embedding/linear module initialisation is visible; repository implementation plan describes symbolic/client-side generation. No corpus, optimiser run or training result for the neural checkpoint is documented.", "provenance_sources": [{"url": "https://huggingface.co/dr-tkxx/Enigma/raw/c269c7c2be993bdc24be4a468afc5ab6b4dcedcc/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Custom embedding/linear module initialisation is visible; repository implementation plan describes symbolic/client-side generation. No corpus, optimiser run or training result for the neural checkpoint is documented.", "evidence_updated_at": "2026-11-01"}
{"model_id": "yasserrmd/smoothie-diffusion-qqp", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/yasserrmd/smoothie-diffusion-qqp/raw/5f4017ef725d531e899481c2ebd2fd6513078703/README.md", "accessed_at": "2026-10-30T12:27:15.661961+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Harley-ml/Dillion-1.2M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Own custom-scale1.28M all-attention Qwen3.5 configuration and3076-token vocabulary, actual extensive pretraining loss/epoch table and explicit9B-token creator training report support own fresh pretraining. No external base named; literal first initialization code/run receipt absent.", "provenance_sources": [{"url": "https://huggingface.co/Harley-ml/Dillion-1.2M/raw/7ab5bfdc62ce9050c54541afa9c5ebf192fe3b05/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/Harley-ml/Dillion-1.2M/raw/7ab5bfdc62ce9050c54541afa9c5ebf192fe3b05/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Harley-ml/Dillion-1.2M/raw/7ab5bfdc62ce9050c54541afa9c5ebf192fe3b05/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA GeForce RTX 2060,6GB VRAM", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "training_duration_reported": "Approximately2.5days", "compute_source": "Direct frozen publisher model-card training statement; exact physicalcountry unstated.", "project_backing": "Public hobby/community collaboration with SupraLabs around model-release date; no paid institutional role established.", "creator_contribution": "Tiny-depth-focused architecture sizing, FineWeb-Edu pretraining and evaluation; own model parameters1281384, current config/weights published.", "best_origin_estimate": {"origin": "own_from_scratch_backbone", "confidence": "medium", "direct_or_inferred": "inferred", "basis": "Bespoke tiny architecture/vocabulary plus recorded pretraining curve and9B-token run; no external checkpoint identified.", "remaining_gap": "Literal training initialization entrypoint unavailable.", "source_urls": ["https://huggingface.co/Harley-ml/Dillion-1.2M/raw/7ab5bfdc62ce9050c54541afa9c5ebf192fe3b05/README.md", "https://huggingface.co/Harley-ml/Dillion-1.2M/raw/7ab5bfdc62ce9050c54541afa9c5ebf192fe3b05/config.json"]}, "provenance_confidence": "medium", "architecture_and_code_lineage": "Uses upstream Qwen3.5 Transformers architecture class at tiny custom dimensions, configured12 full-attention layers rather than stock hybrid layer mixture. Architecture reuse is explicit, not an own novel architecture claim. No pretrained Qwen checkpoint base established.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "hardware": "1 x NVIDIA GeForce RTX 2060,6GB VRAM", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "compute_sources": [{"url": "https://huggingface.co/Harley-ml/Dillion-1.2M/raw/7ab5bfdc62ce9050c54541afa9c5ebf192fe3b05/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Harley-ml/Dillion-1.2M/raw/7ab5bfdc62ce9050c54541afa9c5ebf192fe3b05/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "user-anto/Axiom-Dense-380M-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/user-anto/Axiom-Dense-380M-Base/raw/a8bc3c9417a910294a9227b9aaa1a8eb482de576/README.md", "accessed_at": "2026-10-30T12:27:15.692772+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AlexWortega/moe-200m-qwen3-step365000-12B-20260525-2225", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/AlexWortega/moe-200m-qwen3-step365000-12B-20260525-2225/raw/b342c75fd6a6d925cbcfb06724792ca8127055ac/README.md", "accessed_at": "2026-10-30T12:27:15.732742+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AlexWortega/moe-200m-qwen3-step285000-9B-20260524-1743", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/AlexWortega/moe-200m-qwen3-step285000-9B-20260524-1743/raw/a50e9b44f6d86c7bd9f22201535a39cfc1ae65dc/README.md", "accessed_at": "2026-10-30T12:27:16.009986+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "anshullpal/ByteZira-Vaani-350M-pretrain-base-model", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/anshullpal/ByteZira-Vaani-350M-pretrain-base-model/raw/4852997b94775bc171731fa2791e8cf8c5a6d1db/README.md", "accessed_at": "2026-10-30T12:27:16.150400+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Sherwinroger002/fineweb_edu_gpt_100m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Sherwinroger002/fineweb_edu_gpt_100m/raw/c29f6f23045f59616ecbd32180e062577a5dfadc/README.md", "accessed_at": "2026-10-30T12:27:16.271423+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "koushikkb12/gpt2-123.6m-paramfromscratch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/koushikkb12/gpt2-123.6m-paramfromscratch/raw/70f6d78ea6d1fbf1e2553a1961d07cba760f27a5/README.md", "accessed_at": "2026-10-30T12:27:16.278931+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Harshhvm/bharat-minigpt-350m-pretrain-3b-tokens", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Harshhvm/bharat-minigpt-350m-pretrain-3b-tokens/raw/8c9f9de134a6d422f195900259d5c96f43ebc440/README.md", "accessed_at": "2026-10-30T12:27:16.314463+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "outpost-bio/Waypoint-6m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "taxonomic_sequence_pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card identifies the published Waypoint taxonomic-sequence pretraining checkpoint. Linked official pretrain.py constructs GPT2LMHeadModel(gpt2_config) with a newly built taxonomic tokenizer before Atlas training; it loads no pretrained natural-language GPT2 weights. The checkpoint is a microbiome/taxonomic model.", "provenance_sources": [{"url": "https://huggingface.co/outpost-bio/Waypoint-6m/blob/2f14877d9ed09b7354dbd205bd426d94da3d616c/README.md", "access_date": "2026-11-01"}, {"url": "https://github.com/Outpost-Bio/waypoint", "access_date": "2026-11-01"}, {"url": "https://raw.githubusercontent.com/Outpost-Bio/waypoint/main/src/waypoint_bio/pretrain.py", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "unknown", "detail": "No actual training hardware found in public model cards/readme; preprint retrieval returned403."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "evidence_updated_at": "2026-11-01", "provenance_confidence": "high", "natural_language": "no"}
{"model_id": "outpost-bio/Waypoint-170m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "taxonomic_sequence_pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card identifies the published Waypoint taxonomic-sequence pretraining checkpoint. Linked official pretrain.py constructs GPT2LMHeadModel(gpt2_config) with a newly built taxonomic tokenizer before Atlas training; it loads no pretrained natural-language GPT2 weights. The checkpoint is a microbiome/taxonomic model.", "provenance_sources": [{"url": "https://huggingface.co/outpost-bio/Waypoint-170m/blob/eb7d87184a73d5c0475379c408c7412cd0514678/README.md", "access_date": "2026-11-01"}, {"url": "https://github.com/Outpost-Bio/waypoint", "access_date": "2026-11-01"}, {"url": "https://raw.githubusercontent.com/Outpost-Bio/waypoint/main/src/waypoint_bio/pretrain.py", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "unknown", "detail": "No actual training hardware found in public model cards/readme; preprint retrieval returned403."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "evidence_updated_at": "2026-11-01", "provenance_confidence": "high", "natural_language": "no"}
{"model_id": "outpost-bio/Waypoint-45m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "taxonomic_sequence_pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card identifies the published Waypoint taxonomic-sequence pretraining checkpoint. Linked official pretrain.py constructs GPT2LMHeadModel(gpt2_config) with a newly built taxonomic tokenizer before Atlas training; it loads no pretrained natural-language GPT2 weights. The checkpoint is a microbiome/taxonomic model.", "provenance_sources": [{"url": "https://huggingface.co/outpost-bio/Waypoint-45m/blob/1664ab5ec88b4bdf571f2512968f99dc222ce0ac/README.md", "access_date": "2026-11-01"}, {"url": "https://github.com/Outpost-Bio/waypoint", "access_date": "2026-11-01"}, {"url": "https://raw.githubusercontent.com/Outpost-Bio/waypoint/main/src/waypoint_bio/pretrain.py", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "unknown", "detail": "No actual training hardware found in public model cards/readme; preprint retrieval returned403."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "evidence_updated_at": "2026-11-01", "provenance_confidence": "high", "natural_language": "no"}
{"model_id": "SallySims/equibert-generator", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "Explicit T5-base fine-tuning for conditional DEI text generation.", "provenance_sources": [{"url": "https://huggingface.co/SallySims/equibert-generator/raw/d68b17414c3ebabb53131e40cabeb2fdebf46d02/README.md", "accessed_at": "2026-10-30T12:27:16.897731+00:00"}, {"url": "https://github.com/SallySims", "basis": "Public self-declared location on HF-linked profile.", "accessed_at": "2026-10-30T12:28:19.334808+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Not stated.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Ashwin7/gpt2-124m-fineweb-baseline", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Ashwin7/gpt2-124m-fineweb-baseline/raw/cb1f33dd242ac9e6c8df3f2dd8bf83c1a6e6c67a/README.md", "accessed_at": "2026-10-30T12:27:16.926939+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AlexWortega/ml-intern-moe100m-ufweb-en-step131778-21B-20260518-0918", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "continued_pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies a continued pretraining checkpoint of this author's custom MoE100M base, its own1B/10B warm-start phases and actual10.82B/21.59B consumed tokens.", "provenance_sources": [{"url": "https://huggingface.co/AlexWortega/ml-intern-moe100m-ufweb-en-step131778-21B-20260518-0918/raw/70c6af5d42db643cb1a7b2766f9af274417fc29b/README.md", "accessed_at": "2026-10-30T12:27:16.958494+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "AlexWortega/ml-intern-moe100m-ufweb-en-step66035-10B-20260517-1222", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "continued_pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies a continued pretraining checkpoint of this author's custom MoE100M base, its own1B/10B warm-start phases and actual10.82B/21.59B consumed tokens.", "provenance_sources": [{"url": "https://huggingface.co/AlexWortega/ml-intern-moe100m-ufweb-en-step66035-10B-20260517-1222/raw/61d1276880639122148c2c32c84565307b8e834d/README.md", "accessed_at": "2026-10-30T12:27:16.949986+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "Riaz7/content-writer-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card states~989.2M Llama-style model trained from scratch;5M is claimed context, not parameter count.", "provenance_sources": [{"url": "https://huggingface.co/Riaz7/content-writer-slm-5m/raw/591815f209df3158d3873f70c1d34ecf97dad109/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Apple M4 using MPS", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "qrv0/mnsm-transformer-70m-enwik8", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Matching published snapshot preserved by andujarleo/mnls; scale_up_dynamics.py builds a fresh model from configuration before50k-step enwik8 optimization. Model/run identity supported by exact snapshot revisions; original optimizer/RNG not archived.", "provenance_sources": [{"url": "https://github.com/andujarleo/mnls/blob/HEAD/experiments/neural/scale_up_dynamics.py"}, {"url": "https://huggingface.co/qrv0/mnsm-transformer-70m-enwik8/raw/f34e1381653a711ead27228c5f9555c56241998a/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA RTX4060 Laptop; approximately2.8h per model", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "qrv0/mnsm-memnls-70m-enwik8", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Matching published snapshot preserved by andujarleo/mnls; scale_up_dynamics.py builds a fresh model from configuration before50k-step enwik8 optimization. Model/run identity supported by exact snapshot revisions; original optimizer/RNG not archived.", "provenance_sources": [{"url": "https://github.com/andujarleo/mnls/blob/HEAD/experiments/neural/scale_up_dynamics.py"}, {"url": "https://huggingface.co/qrv0/mnsm-memnls-70m-enwik8/raw/6133a700e8ec2de6bbf3ab5148894979aab7a2ab/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA RTX4060 Laptop; approximately2.8h per model", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ray0rf1re/Nano-nano_v4.5", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact card says trained from scratch255.7M on15datasets/0.08Btokens. Card training timestamp2026-05-09 22:50 is later than repository creation2026-02-11; preserve both dates. Instruction data does not imply pretrained weight initialization.", "provenance_sources": [{"url": "https://huggingface.co/ray0rf1re/Nano-nano_v4.5/raw/f667a2f0f142fc4b7eac9293a6d53a73d28994f7/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA GTX1080 8GB", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ray0rf1re/Nano-nano-4.6", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "The v4.5 card explicitly describes own from-scratch 255.7M training and links v4.6 as its new version. The v4.7 comparison identifies v4.6's same custom 15-layer/896-width architecture and Nano-nano tokenizer. Own original lineage is the best-supported classification.", "provenance_sources": [{"url": "https://huggingface.co/ray0rf1re/Nano-nano-4.6/raw/30ba47bc41290fa1a2bf9434c1e89efa0f55d12e/README.md"}, {"url": "https://huggingface.co/ray0rf1re/Nano-nano-4.6/blob/main/README.md", "access_date": "2026-11-01", "finding": "Saved card read.", "reused": true}, {"url": "https://huggingface.co/ray0rf1re/Nano-nano_v4.5/blob/main/README.md", "access_date": "2026-11-01", "finding": "Saved card explicitly from scratch; same creator's model-family lineage.", "reused": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "provenance_evidence_detail": "author_card_architecture_and_model_family_inference", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "onalennainnocent/mzansilm-125m", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Uploader reproduces original UCT MzansiLM125M card/weights. Upstream original research pretraining is documented by paper, not a distinct run by Onalenna.", "provenance_sources": [{"url": "https://huggingface.co/onalennainnocent/mzansilm-125m/raw/b511bf0f0dd99da6569f8209c0acf93ba2b01ffc/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AlexWortega/ml-intern-v4-100m-tinystories-20260512-1721", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes a self-implemented130M downscale, actual initial-to-final training losses and8000-step TinyStories checkpoint produced by the author's pipeline.", "provenance_sources": [{"url": "https://huggingface.co/AlexWortega/ml-intern-v4-100m-tinystories-20260512-1721/raw/1be1dec0906b249fe14c17bbebb63b7235f29835/README.md", "accessed_at": "2026-10-30T12:27:17.853697+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "nisha97/tiny-stories-slm-50M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card explicitly identifies a50M GPT-style model built from scratch and trained on TinyStories, validation loss2.390 and actual training hyperparameters.", "provenance_sources": [{"url": "https://huggingface.co/nisha97/tiny-stories-slm-50M/blob/8189bb21e3a71bda09b399408a702ad43d8f1805/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high"}
{"model_id": "Sherwinroger002/gpt-124m-fineweb-edu-10m-tokens", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Custom124M GPT implemented from scratch in PyTorch, explicit10M-token FineWeb-Edu run on RTX3060Ti and early-stage generation samples support own scratch-origin training.", "provenance_sources": [{"url": "https://huggingface.co/Sherwinroger002/gpt-124m-fineweb-edu-10m-tokens/raw/67bd60c06fc73b27bd7c48f425b533a30b4fb2e9/README.md", "accessed_at": "2026-10-30T12:27:18.133239+00:00"}, {"url": "https://huggingface.co/Sherwinroger002/gpt-124m-fineweb-edu-10m-tokens/raw/67bd60c06fc73b27bd7c48f425b533a30b4fb2e9/README.md", "read_scope": "Saved card architecture/base/training/result and upstream attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "snehangshu511/tinystories-gpt2-124M-scratch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/snehangshu511/tinystories-gpt2-124M-scratch/raw/c77c05ae98b2b5c94c47bc18d3b9ced15ec66aae/README.md", "accessed_at": "2026-10-30T12:27:18.123733+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Lernex/Metis-1.4-base", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/Lernex/Metis-1.4-base/raw/5f864b3975b8010e325e51b01b6d89312686d1e5/README.md", "accessed_at": "2026-10-30T12:27:18.210081+00:00", "quote": "aining - exported BF16 weights in safetensors The final CPT checkpoint ended with validation loss around 2.4341 and perplexity around 11.41 on the continued-pretraining validation split. This number is not directly comparable to i", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "edwinhere/namer", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original870k seq2seq number-name transformer with41-token vocabulary, stratified synthetic integer training and own scratch-training command supports original trained mapping model; not a general-purpose text assistant.", "provenance_sources": [{"url": "https://huggingface.co/edwinhere/namer/raw/1064fadd81028477f5b229d9e311d07ede483e1d/README.md", "accessed_at": "2026-10-30T12:27:18.219591+00:00"}, {"url": "https://huggingface.co/edwinhere/namer/raw/1064fadd81028477f5b229d9e311d07ede483e1d/README.md", "read_scope": "Cached card model/architecture/training/result and upstream description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "curious-techie/Vritya-Tiny-163M-HF", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "format_export_of_own_pretrained_base", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned export card names best_model.pth from the publisher original Vritya163M repository. The linked author project identifies best_model.pth as the pretrained base and constructs GPTModel(cfg) before TinyStories pretraining; later instruction/classification weights have separate filenames.", "provenance_sources": [{"url": "https://huggingface.co/curious-techie/Vritya-Tiny-163M-HF/blob/af1373451a16a096ab2fee0071c342d5ed540a24/README.md", "access_date": "2026-11-01"}, {"url": "https://github.com/VrityaCodeRishi/Vritya-Tiny-163M-1", "access_date": "2026-11-01"}, {"url": "https://raw.githubusercontent.com/VrityaCodeRishi/Vritya-Tiny-163M-1/main/pretraining.py", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "upstream_checkpoint": "curious-techie/Vritya-Tiny-163M/best_model.pth"}
{"model_id": "darthcrawl/tiny-tang-38m", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_model", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/darthcrawl/tiny-tang-38m/raw/d70a00b7382ed40eda55ddafd26928b98d79d61e/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/darthcrawl/tiny-tang-38m/raw/d70a00b7382ed40eda55ddafd26928b98d79d61e/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/darthcrawl/tiny-tang-38m/raw/d70a00b7382ed40eda55ddafd26928b98d79d61e/config.json", "access_date": "2026-10-31", "scope": "same file blobs as paired base repository; complete code/config reading reused"}, {"url": "https://huggingface.co/darthcrawl/tiny-tang-38m/raw/d70a00b7382ed40eda55ddafd26928b98d79d61e/config.py", "access_date": "2026-10-31", "scope": "same file blobs as paired base repository; complete code/config reading reused"}, {"url": "https://huggingface.co/darthcrawl/tiny-tang-38m/raw/d70a00b7382ed40eda55ddafd26928b98d79d61e/meta.txt", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/darthcrawl/tiny-tang-38m/raw/d70a00b7382ed40eda55ddafd26928b98d79d61e/model.py", "access_date": "2026-10-31", "scope": "same file blobs as paired base repository; complete code/config reading reused"}, {"url": "https://huggingface.co/api/models/darthcrawl/tiny-30m", "access_date": "2026-10-31", "http_status": 401, "scope": "named upstream parent API inaccessible; not evidence of absence"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not stated for exact run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Full-weight Wu-Tang-lyrics continuation without LoRA; shipped latest.pt from roughly500 additional steps, card best checkpoint step19950. Underlying music/corpus creators remain separately credited.", "architecture_and_code_lineage": "Own educational PyTorch implementation of standard RMSNorm/RoPE/SwiGLU/SDPA decoder recipe; code explicitly credits GPT-2 residual-projection initialization. No imported external pretrained checkpoint in model definition.", "evidence_updated_at": "2026-10-31"}
{"model_id": "darthcrawl/tiny-38m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "tiny-38m describes from-scratch PyTorch training, 37.8M parameters and 477.5M tokens on one unspecified GPU. tiny-tang-38m explicitly continues darthcrawl/tiny-30m. Pinned configs and commit-author handles add no personal location.", "provenance_sources": [{"url": "https://huggingface.co/darthcrawl/tiny-38m/raw/6e14144212e7e7c4baa6d2c1edc064382dfe6ae2/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/darthcrawl/tiny-38m/raw/6e14144212e7e7c4baa6d2c1edc064382dfe6ae2/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/darthcrawl/tiny-38m/raw/6e14144212e7e7c4baa6d2c1edc064382dfe6ae2/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/darthcrawl/tiny-38m/raw/6e14144212e7e7c4baa6d2c1edc064382dfe6ae2/config.py", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/darthcrawl/tiny-38m/raw/6e14144212e7e7c4baa6d2c1edc064382dfe6ae2/meta.txt", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/darthcrawl/tiny-38m/raw/6e14144212e7e7c4baa6d2c1edc064382dfe6ae2/model.py", "access_date": "2026-10-31", "scope": "complete frozen file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 GPU; model unspecified", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Explicit pretraining from zero, own 8192-vocabulary tokenizer and mixed small-corpus pipeline; best checkpoint step19500. Fresh model constructor random initialization is visible.", "architecture_and_code_lineage": "Own educational PyTorch implementation of standard RMSNorm/RoPE/SwiGLU/SDPA decoder recipe; code explicitly credits GPT-2 residual-projection initialization. No imported external pretrained checkpoint in model definition.", "hardware": "1 GPU; model unspecified", "compute_sources": [{"url": "https://huggingface.co/darthcrawl/tiny-38m/raw/6e14144212e7e7c4baa6d2c1edc064382dfe6ae2/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/darthcrawl/tiny-38m/raw/6e14144212e7e7c4baa6d2c1edc064382dfe6ae2/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/darthcrawl/tiny-38m/raw/6e14144212e7e7c4baa6d2c1edc064382dfe6ae2/config.py", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/darthcrawl/tiny-38m/raw/6e14144212e7e7c4baa6d2c1edc064382dfe6ae2/meta.txt", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/darthcrawl/tiny-38m/raw/6e14144212e7e7c4baa6d2c1edc064382dfe6ae2/model.py", "access_date": "2026-10-31", "scope": "complete frozen file read"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "saiteja718/gpt-2", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/saiteja718/gpt-2/raw/ffa51e61940ca1e3f6cab17e463a15f39b0eb7f2/README.md", "accessed_at": "2026-10-30T12:27:18.757086+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "mlnomad/goat-v-482m-c4-then-fineweb-seed0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/goat-v-482m-c4-then-fineweb-seed0/raw/a4faeb9f55e91ac98fad68ba4a338c92f74092a8/README.md", "accessed_at": "2026-10-30T12:27:18.862875+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "mlnomad/goat-v-482m-c4-seed0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/goat-v-482m-c4-seed0/raw/a919111673ca879f3ebdbe587a0e63de9121986d/README.md", "accessed_at": "2026-10-30T12:27:18.839839+00:00"}, {"url": "https://huggingface.co/mlnomad/goat-v-482m-c4-seed0/raw/a919111673ca879f3ebdbe587a0e63de9121986d/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "Gothicdreams/i3-tiny", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card reports own tiny character-model i3 architecture trained2,000 iterations on45,830 characters using Kaggle CPU, with loss falling3.53→2.15. Bespoke tiny character design and complete run description support own scratch origin; generic component ideas are separate.", "provenance_sources": [{"url": "https://huggingface.co/Gothicdreams/i3-tiny/raw/50507bfddc2361f4afb9b03645a1fd93820f0d5d/README.md", "accessed_at": "2026-10-30T12:27:19.042206+00:00"}, {"url": "https://huggingface.co/Gothicdreams/i3-tiny/raw/50507bfddc2361f4afb9b03645a1fd93820f0d5d/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mlnomad/gelu-d12-chinchilla-261M-seed2-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/gelu-d12-chinchilla-261M-seed2-pytorch/raw/df539cd1f6f47a74cabb01f9efa5520a0a89080d/README.md", "accessed_at": "2026-10-30T12:27:19.301079+00:00"}, {"url": "https://huggingface.co/mlnomad/gelu-d12-chinchilla-261M-seed2-pytorch/raw/df539cd1f6f47a74cabb01f9efa5520a0a89080d/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mlnomad/gelu-d12-chinchilla-261M-seed1-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/gelu-d12-chinchilla-261M-seed1-pytorch/raw/3d3ab0b762800e2af029abe3d0ce954d44457b4c/README.md", "accessed_at": "2026-10-30T12:27:19.391720+00:00"}, {"url": "https://huggingface.co/mlnomad/gelu-d12-chinchilla-261M-seed1-pytorch/raw/3d3ab0b762800e2af029abe3d0ce954d44457b4c/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mlnomad/yatnmn-softplus-sb-ca-d12-chinchilla-261M-seed2-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/yatnmn-softplus-sb-ca-d12-chinchilla-261M-seed2-pytorch/raw/65a88a12956843b6e88de2ab5db0425c3ed17138/README.md", "accessed_at": "2026-10-30T12:27:19.504141+00:00"}, {"url": "https://huggingface.co/mlnomad/yatnmn-softplus-sb-ca-d12-chinchilla-261M-seed2-pytorch/raw/65a88a12956843b6e88de2ab5db0425c3ed17138/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mlnomad/yatnmn-softplus-sb-ca-d12-chinchilla-261M-seed1-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/yatnmn-softplus-sb-ca-d12-chinchilla-261M-seed1-pytorch/raw/9203b20a512b70fa16f9a266a9e60df940650727/README.md", "accessed_at": "2026-10-30T12:27:19.459575+00:00"}, {"url": "https://huggingface.co/mlnomad/yatnmn-softplus-sb-ca-d12-chinchilla-261M-seed1-pytorch/raw/9203b20a512b70fa16f9a266a9e60df940650727/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mlnomad/yatnmn-softplus-ca-d12-chinchilla-261M-seed2-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/yatnmn-softplus-ca-d12-chinchilla-261M-seed2-pytorch/raw/a5fe50fb8b83e9ce553abca0668ad3d37c7a0aca/README.md", "accessed_at": "2026-10-30T12:27:19.515658+00:00"}, {"url": "https://huggingface.co/mlnomad/yatnmn-softplus-ca-d12-chinchilla-261M-seed2-pytorch/raw/a5fe50fb8b83e9ce553abca0668ad3d37c7a0aca/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mlnomad/yatnmn-softplus-ca-d12-chinchilla-261M-seed1-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/yatnmn-softplus-ca-d12-chinchilla-261M-seed1-pytorch/raw/03655837e5dff7a45f55a4f1fd97781fb456319f/README.md", "accessed_at": "2026-10-30T12:27:19.653389+00:00"}, {"url": "https://huggingface.co/mlnomad/yatnmn-softplus-ca-d12-chinchilla-261M-seed1-pytorch/raw/03655837e5dff7a45f55a4f1fd97781fb456319f/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mlnomad/yatnmn-softplus-sb-d12-chinchilla-261M-seed2-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/yatnmn-softplus-sb-d12-chinchilla-261M-seed2-pytorch/raw/da2ca189597a2500eda167d1ba702f7d7a37c584/README.md", "accessed_at": "2026-10-30T12:27:19.897423+00:00"}, {"url": "https://huggingface.co/mlnomad/yatnmn-softplus-sb-d12-chinchilla-261M-seed2-pytorch/raw/da2ca189597a2500eda167d1ba702f7d7a37c584/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mlnomad/yatnmn-softplus-sb-d12-chinchilla-261M-seed1-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/yatnmn-softplus-sb-d12-chinchilla-261M-seed1-pytorch/raw/c0f5ea1785c3dba6e18701b8e8c9be138eb4e81c/README.md", "accessed_at": "2026-10-30T12:27:20.042772+00:00"}, {"url": "https://huggingface.co/mlnomad/yatnmn-softplus-sb-d12-chinchilla-261M-seed1-pytorch/raw/c0f5ea1785c3dba6e18701b8e8c9be138eb4e81c/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mlnomad/yatnmn-softplus-d12-chinchilla-261M-seed2-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/yatnmn-softplus-d12-chinchilla-261M-seed2-pytorch/raw/f56e95238223c9f19424d4263c61c47dc11c1ccf/README.md", "accessed_at": "2026-10-30T12:27:20.086371+00:00"}, {"url": "https://huggingface.co/mlnomad/yatnmn-softplus-d12-chinchilla-261M-seed2-pytorch/raw/f56e95238223c9f19424d4263c61c47dc11c1ccf/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mlnomad/yatnmn-softplus-d12-chinchilla-261M-seed1-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/yatnmn-softplus-d12-chinchilla-261M-seed1-pytorch/raw/74d3eb04a1d7c0297cc4164ddf0e6aae507932b9/README.md", "accessed_at": "2026-10-30T12:27:20.116859+00:00"}, {"url": "https://huggingface.co/mlnomad/yatnmn-softplus-d12-chinchilla-261M-seed1-pytorch/raw/74d3eb04a1d7c0297cc4164ddf0e6aae507932b9/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mlnomad/yatnmn-softplus-sb-ca-d12-chinchilla-261M-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/yatnmn-softplus-sb-ca-d12-chinchilla-261M-pytorch/raw/105af05258cbb8c8d63c33a7773499c6872b365a/README.md", "accessed_at": "2026-10-30T12:27:20.132885+00:00"}, {"url": "https://huggingface.co/mlnomad/yatnmn-softplus-sb-ca-d12-chinchilla-261M-pytorch/raw/105af05258cbb8c8d63c33a7773499c6872b365a/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mlnomad/yatnmn-softplus-ca-d12-chinchilla-261M-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/yatnmn-softplus-ca-d12-chinchilla-261M-pytorch/raw/0074b2b24815201a6aea848a5b5ed500aafffc33/README.md", "accessed_at": "2026-10-30T12:27:20.231423+00:00"}, {"url": "https://huggingface.co/mlnomad/yatnmn-softplus-ca-d12-chinchilla-261M-pytorch/raw/0074b2b24815201a6aea848a5b5ed500aafffc33/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "bgraudt/mythos", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/bgraudt/mythos/raw/dead18986c1d52647609cb5ec6c73e06ab55ebd9/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA A10040GB; device count unspecified", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Own decoder pretraining/tokenizer/implementation plus HF-compatible export; no SFT.", "architecture_lineage": "Explicit Llama/Mistral7B architecture inspiration and FineWeb data-pipeline methodology; publisher says own pure-PyTorch implementation and original weights, exported to LlamaForCausalLM format.", "hardware": "NVIDIA A10040GB; device count unspecified", "compute_sources": [{"url": "https://huggingface.co/bgraudt/mythos/raw/dead18986c1d52647609cb5ec6c73e06ab55ebd9/README.md", "scope": "Complete frozen card read", "excerpt": "Hardware | A100 40 GB", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "mlnomad/goat-vvv-d12-fineweb-5x-pytorch", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own architecture ablation/pretraining or a reproducibility run differing only in random seed, with corpus/training budget and matched architecture. This supports own fresh-run origin. Nanochat/GPT architecture and reused Mistral tokenizer remain separately attributed; no Mistral weights are identified.", "provenance_sources": [{"url": "https://huggingface.co/mlnomad/goat-vvv-d12-fineweb-5x-pytorch/raw/12f8ca00dc1d4eba918377012d8b5b83a121ef62/README.md", "accessed_at": "2026-10-30T12:27:20.650663+00:00"}, {"url": "https://huggingface.co/mlnomad/goat-vvv-d12-fineweb-5x-pytorch/raw/12f8ca00dc1d4eba918377012d8b5b83a121ef62/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "HenrySentinel/tinyMind", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/HenrySentinel/tinyMind/raw/22c34d98a379c394779d777e325dc14491774df3/README.md", "accessed_at": "2026-10-30T12:27:20.708585+00:00", "quote": "nyMind This is a small transformer language model trained from scratch with approximately 17,731,328 parameters. Model Details - Architecture: GPT-style transformer - Parameters: ~17M - Layers: 6 - Attention Heads: 8 - Embedding D", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "unknown", "detail": "No actual training hardware established for this cohort artifact."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "cosail-knu/pythia-14m", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card retains EleutherAI original Pythia-14M-deduped developer, Pile training and original checkpoint details, with no cosail-specific training description. Best-supported release is upstream-model redistribution.", "provenance_sources": [{"url": "https://huggingface.co/cosail-knu/pythia-14m/raw/6ca1ad1da7329f7f6c28cfd412c235886d5c1aa1/README.md", "accessed_at": "2026-10-30T12:27:20.738658+00:00"}, {"url": "https://huggingface.co/cosail-knu/pythia-14m/raw/6ca1ad1da7329f7f6c28cfd412c235886d5c1aa1/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "AAzurez/pythia-14mtest", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card reproduces EleutherAI Pythia-14M model identity, original developer attribution, original Pile/checkpoint schedule and original loading context, with no publisher-specific training contribution. Best-supported release is redistribution of EleutherAI model.", "provenance_sources": [{"url": "https://huggingface.co/AAzurez/pythia-14mtest/raw/7d1cfbf9cd74c83811eb4f491d7078628a7ad2c1/README.md", "accessed_at": "2026-10-30T12:27:20.750182+00:00"}, {"url": "https://huggingface.co/AAzurez/pythia-14mtest/raw/7d1cfbf9cd74c83811eb4f491d7078628a7ad2c1/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "pandeyps/DeepSeekV3", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "not_established", "provenance_evidence_level": "declared_pretrained_base", "provenance_basis": "Frozen repository metadata explicitly identifies pretrained base deepseek-v3.", "provenance_sources": [{"url": "https://huggingface.co/pandeyps/DeepSeekV3/raw/f3ba1d2c5ef2f6c5c96bf4db98103fabf4071702/README.md", "accessed_at": "2026-10-30T12:27:20.812817+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "PranavGuhan/python-ds-accelerate", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/PranavGuhan/python-ds-accelerate/raw/07ac7e71671b2d179784f2c0e2c3fc1ae1bf04b1/README.md", "accessed_at": "2026-10-30T12:27:21.221562+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "rohansingh2612/healthslm", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/rohansingh2612/healthslm/raw/fae736ca397a982b1e6cd722c49078a67b2b1494/README.md", "accessed_at": "2026-10-30T12:27:21.294642+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "DevHunterAI/RubiRLM-1B-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Creator describes its custom recursive/MoE model, actual FineWeb/UltraChat base pretraining and final original training checkpoint export. This supports own original pretraining without requiring a separate literal random-initialization receipt.", "provenance_sources": [{"url": "https://huggingface.co/DevHunterAI/RubiRLM-1B-Base/raw/6210d1d86d88aa17c98e3206010de015356872ed/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/DevHunterAI/RubiRLM-1B-Base/blob/main/README.md", "access_date": "2026-11-01", "finding": "Saved card read: authored custom10-recursive-block MoE architecture, FineWeb/UltraChat completed base pretraining, original training_checkpoint.pt and exported final checkpoint.", "reused": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not stated for exact run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "author_base_training_account_and_custom_architecture_inference", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "Rzoro/erebus-medium", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Rzoro/erebus-medium/raw/ba0c987aa3a017e5af279997c8e2d74a201e2ba7/README.md", "accessed_at": "2026-10-30T12:27:21.387659+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Saminx22/MedSLM", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Custom381M24-layer RoPE/GQA medical base model has an actual44M-token training account,19500-step best checkpoint and measured validation loss.", "provenance_sources": [{"url": "https://huggingface.co/Saminx22/MedSLM/raw/10e3fe8979642fdeaec3d30f1009eb252ea76171/README.md", "accessed_at": "2026-10-30T12:27:21.377351+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "Rzoro/erebus-small", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Rzoro/erebus-small/raw/0ed640a21354be4a298de867dda071487d932ea8/README.md", "accessed_at": "2026-10-30T12:27:21.451364+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "bmnzyb/CampGPT_X", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "saved_training_code_and_stage_metadata", "provenance_basis": "Frozen pretraining source constructs GPT(config) with random-normal initialization and no external weight load; SFT loads own log/model_00999_CampGPT_X.pt, DPO loads own SFT weights. Final card/config identifies pretrain_10B -> SFT -> DPO. Own from-scratch ancestry supported by the saved complete development pipeline; frozen release is post-trained.", "provenance_sources": [{"url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/README.md"}, {"url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/README.md", "access_date": "2026-10-31", "scope": "complete frozen model card"}, {"url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/config.py", "access_date": "2026-10-31", "scope": "complete architecture presets"}, {"url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/config.json", "access_date": "2026-10-31", "scope": "complete final model metadata"}, {"url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/bash.sh", "access_date": "2026-10-31", "scope": "complete launch-command file; configuration not proof of four-device execution"}, {"url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/train_CampGPT_X_plus.py", "access_date": "2026-10-31", "scope": "module description, device/DDP setup, fresh model construction and training setup"}, {"url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/model.py", "access_date": "2026-10-31", "scope": "component headers, model constructor and random-normal initialization"}, {"url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/train_sft.py", "access_date": "2026-10-31", "scope": "source checkpoint path and model loading sections"}, {"url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/train_dpo.py", "access_date": "2026-10-31", "scope": "SFT loading and policy/reference model initialization sections"}, {"url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/build_dataset.py", "access_date": "2026-10-31", "scope": "PDF/Qwen3-Max data-generation configuration and source selection"}, {"url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/chunks_debug.json", "access_date": "2026-10-31", "scope": "first parsed handbook record identifying University of Houston-Clear Lake"}, {"url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/caogao.ipynb", "access_date": "2026-10-31", "scope": "complete saved notebook, including RTX3090 diagnostic outputs"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Exact full-run GPU count, elapsed duration and physical compute country unstated.; Creator institutional route not identified.", "compute_source": "Actual repository diagnostic RTX3090; launch file has single, two and four GPU commands as options. No completed-run log establishes which count ran.", "upstream_checkpoint": "Own pretraining checkpoint -> own SFT -> own DPO. GPT-2 tiktoken tokenizer reuse is separate from pretrained model-weight reuse.", "creator_contribution": "Own standard-recipe MoE training pipeline, FineWeb-Edu pretraining, Qwen-generated handbook Q&A and post-training, packaging.", "upstream_architecture_or_code": [{"name": "GPT-style custom PyTorch decoder with GQA, RoPE, RMSNorm, SwiGLU and MoE/shared experts", "source_url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/model.py"}, {"name": "RMSNorm research paper", "role": "explicit implementation reference", "source_url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/model.py"}, {"name": "Qwen3-Max", "role": "synthetic handbook Q&A / preference-data generator, not a pretrained weight base", "source_url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/build_dataset.py"}], "parameter_count_publisher_reported": 322680576, "parameter_count_inventory": 284047104, "best_training_hardware_estimate": {"hardware": "NVIDIA GeForce RTX 3090", "confidence": "medium", "basis": "Saved notebook diagnostic inside same frozen development repository explicitly prints RTX3090, but does not bind that card to every full training stage or establish device count.", "source_url": "https://huggingface.co/bmnzyb/CampGPT_X/raw/8d44ebfbdac01d26adb6804a6feedf2e6b45eb38/caogao.ipynb"}, "evidence_updated_at": "2026-10-31"}
{"model_id": "CJJones/Jeeney_AI_200M_Reloaded_GPT", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card explicitly describes this207M modified nanoGPT model as built from scratch over approximately two years, trained almost entirely on synthetic corpora; checkpoint, custom code and benchmark results are supplied.", "provenance_sources": [{"url": "https://huggingface.co/CJJones/Jeeney_AI_200M_Reloaded_GPT/blob/9c88655f0a41aaadb237845c33a9ee942a444065/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high"}
{"model_id": "Girinath11/recursive-language-model-198m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "recursive-language-model-198m card states from-scratch adaptive-recursion198M model, single KaggleT4 and150kconversation data. UltraChat/Tsinghua acknowledgment is dataset provenance, not author's schooling.", "provenance_sources": [{"url": "https://huggingface.co/Girinath11/recursive-language-model-198m/raw/feed9fea31fe9950dbfc2ef5a20924b2c070f67d/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "single KaggleT4", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Rzoro/erebus-tiny", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Rzoro/erebus-tiny/raw/45493f00867cb0408cfb04184b741047cffb5fd3/README.md", "accessed_at": "2026-10-30T12:27:22.010405+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "justjuu/pico-gpt", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/justjuu/pico-gpt/raw/f83ce24ba9abdb088c06d4c763f138e4a97503ed/README.md", "accessed_at": "2026-10-30T12:27:22.044934+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "suhailult777/MedBrain-0.5B", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/suhailult777/MedBrain-0.5B/raw/0dea5bfa9b85ca76e34b237814ec807ff6701534/README.md", "accessed_at": "2026-10-30T12:27:22.122714+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Tribewarez/pot-o-22-slim", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "untrained", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly identifies random, untrained weights and describes training as a future operation.", "provenance_sources": [{"url": "https://huggingface.co/Tribewarez/pot-o-22-slim/raw/ea762be72c576abc3b837706028b29ea5545a723/README.md", "accessed_at": "2026-10-30T12:27:22.449560+00:00", "quote": "anges-22-22k) (signature paramsignature 22.2222). Weights are random initialization — intended for fine-tuning on PoT-O challenge → optimalpath text (see dataset card). Architecture chosen to land near 22.2k params with a 257-token byte-level ", "basis": "Card explicitly identifies random, untrained weights and describes training as a future operation."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "fromthesky/PLDR-LLM-v51-SOC-110M-5", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 41B tokens/1250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-SOC-110M-5/raw/de8e539c0ba1829072f4b8c2c5fae3bde0a3a2d2/README.md", "accessed_at": "2026-10-30T12:27:22.543742+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-SOC-110M-5/raw/de8e539c0ba1829072f4b8c2c5fae3bde0a3a2d2/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51-SOC-110M-4", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-SOC-110M-4/raw/c377be06f2294aeccb56d19ea5f52640b1850daf/README.md", "accessed_at": "2026-10-30T12:27:22.654511+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-SOC-110M-4/raw/c377be06f2294aeccb56d19ea5f52640b1850daf/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51-SOC-110M-3", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-SOC-110M-3/raw/5b6c6e8f299d166386176728792c6803c3a6a51e/README.md", "accessed_at": "2026-10-30T12:27:22.648507+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-SOC-110M-3/raw/5b6c6e8f299d166386176728792c6803c3a6a51e/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51-SOC-110M-2", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-SOC-110M-2/raw/5e6f3164f6aef7878743f92725364860931047a0/README.md", "accessed_at": "2026-10-30T12:27:22.682790+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-SOC-110M-2/raw/5e6f3164f6aef7878743f92725364860931047a0/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51-SOC-110M-1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-SOC-110M-1/raw/7a34e2ca9aa78038683677cfda17fe3a9fe6da8a/README.md", "accessed_at": "2026-10-30T12:27:22.734534+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-SOC-110M-1/raw/7a34e2ca9aa78038683677cfda17fe3a9fe6da8a/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Prositron/gome", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes_incomplete_run", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author calls this first custom24M model and describes own FineWeb-Edu pretraining stopped early at19518steps. Original custom code and training narrative support own scratch lineage; incomplete training is retained.", "provenance_sources": [{"url": "https://huggingface.co/Prositron/gome/raw/8921d25f4af7ca7864f321d46e37ff5406c935f5/README.md"}, {"url": "https://huggingface.co/Prositron/gome", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "prskid1000/micro-Omni", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining_and_multimodal_sft", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Explicitly from-scratch3.4M multimodal model with synthetic training and7-stage90min RTX5070Ti run; original792k Thinker language core and other small components reported.", "provenance_sources": [{"url": "https://huggingface.co/prskid1000/micro-Omni/raw/c1065ce95cf2522db2299f5c5340582cff866c6b/README.md", "accessed_at": "2026-10-30T12:27:23.114305+00:00"}, {"url": "https://huggingface.co/prskid1000/micro-Omni/raw/c1065ce95cf2522db2299f5c5340582cff866c6b/README.md", "read_scope": "Saved card model/task/training and source attribution", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "levossadtchi/QED-75M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card documents a custom75M model family with12.6B-token pretraining followed by context-length annealing and SFT preparation. Own original pretraining is supported; preparation is not counted as completed SFT.", "provenance_sources": [{"url": "https://huggingface.co/levossadtchi/QED-75M/raw/11b4277a57923768acd23ab199d84459cdcfc446/README.md", "accessed_at": "2026-10-30T12:27:23.359879+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "rudyon/linnet-497M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/rudyon/linnet-497M/raw/039520259f666db52eabadc448e653dd12c7c398/README.md", "accessed_at": "2026-10-30T12:27:23.332833+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ray0rf1re/hyper-nix.1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned rendered card describes the publisher custom HyperNix architecture with custom tokenizer and an8055-step training run onGTX1080. Custom small architecture/run supports scratch origin; no pretrained parent is declared.", "provenance_sources": [{"url": "https://huggingface.co/ray0rf1re/hyper-nix.1/blob/e892d5167af1e337d63b6f8dbb59d9157752caed/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA GeForce GTX 1080", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "hardware": "1 x NVIDIA GeForce GTX 1080", "training_steps_recorded": 8055, "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "akakumar42/longnet-shakespeare", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card explicitly describes a trained6.3M character-level LongNet with a custom from-scratch implementation and Tiny Shakespeare corpus; use-example/step placeholders do not erase the stated completed training.", "provenance_sources": [{"url": "https://huggingface.co/akakumar42/longnet-shakespeare/blob/a4c90018d18790ebdf70a6bfec6523d9d23c3c11/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "evidence_updated_at": "2026-11-01", "provenance_confidence": "high"}
{"model_id": "rudyon/rudygpt", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/rudyon/rudygpt/raw/dd42cab1fbfc4a4db1089341a8cd78a9a229c937/README.md", "accessed_at": "2026-10-30T12:27:23.614727+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "aonee/medical-gpt-50m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/aonee/medical-gpt-50m/blob/main/README.md"}, {"url": "https://huggingface.co/aonee/medical-gpt-50m/raw/1dbd2216be8f2a6e15badc056a7b34d85476abfe/README.md", "accessed_at": "2026-10-30T12:27:23.711669+00:00"}, {"url": "https://huggingface.co/aonee/medical-gpt-50m/raw/1dbd2216be8f2a6e15badc056a7b34d85476abfe/README.md", "access_date": "2026-10-31", "scope": "complete frozen card read", "excerpt": "L4 has 24GB, not H100's 80GB", "finding": "Experiment table says first working baseline on L4; Limitations states 5-minute training budget.", "confidence": "high as publisher statement"}, {"url": "https://huggingface.co/api/models/aonee/medical-gpt-50m/discussions?p=0", "access_date": "2026-10-31", "scope": "complete JSON discussion list read; count 0"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA L4 24GB", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Creator geographic base / training place, physical GPU country and provider not stated.; University / institutional employment or model backing unstated.", "training_duration_reported": "5-minute training budget; time-budgeted experiments", "training_tokens_publisher_reported": 37000000, "hardware": "NVIDIA L4 24GB", "compute_sources": [{"url": "https://huggingface.co/aonee/medical-gpt-50m/raw/1dbd2216be8f2a6e15badc056a7b34d85476abfe/README.md", "access_date": "2026-10-31", "scope": "complete frozen card read", "excerpt": "L4 has 24GB, not H100's 80GB", "finding": "Experiment table says first working baseline on L4; Limitations states 5-minute training budget.", "confidence": "high as publisher statement"}, {"url": "https://huggingface.co/api/models/aonee/medical-gpt-50m/discussions?p=0", "access_date": "2026-10-31", "scope": "complete JSON discussion list read; count 0"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "HoangHa/smoldlm-144m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own SmolDLM first-principles diffusion-LM project, custom144M block-diffusion architecture and25,500-step/~26B-token base run. BD3-LM/gated-attention ideas and reused SmolLM tokenizer merges retain their upstream attribution; no external pretrained backbone named.", "provenance_sources": [{"url": "https://huggingface.co/HoangHa/smoldlm-144m/raw/64f1177303ae2c6f021791d1b80bcd92ea911151/README.md", "accessed_at": "2026-10-30T12:27:24.062448+00:00"}, {"url": "https://huggingface.co/HoangHa/smoldlm-144m/raw/64f1177303ae2c6f021791d1b80bcd92ea911151/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "HzJunp/flan-t5-text2sql-with-schema-v2", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Frozen flan-t5-text2sql-with-schema-v2 card repeats juierror's upstream checkpoint and examples; commit history retains juierror and HzJunp. Publisher-specific training, initialization and hardware are not documented; copied source is distinct from uploader.", "provenance_sources": [{"url": "https://huggingface.co/api/models/HzJunp/flan-t5-text2sql-with-schema-v2/commits/247d258bc27d3d0ef9dafaf537fcec0fc51f5d9b"}, {"url": "https://huggingface.co/HzJunp/flan-t5-text2sql-with-schema-v2/raw/247d258bc27d3d0ef9dafaf537fcec0fc51f5d9b/README.md", "basis": "Pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/api/models/HzJunp/flan-t5-text2sql-with-schema-v2/commits/247d258bc27d3d0ef9dafaf537fcec0fc51f5d9b", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/HzJunp/flan-t5-text2sql-with-schema-v2/raw/247d258bc27d3d0ef9dafaf537fcec0fc51f5d9b/README.md", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/api/models/HzJunp/flan-t5-text2sql-with-schema-v2/discussions?p=0", "access_date": "2026-10-31", "scope": "complete discussion-list JSON read", "finding": "0 listed public threads", "http_status": 200}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "compute_source": "No publisher-specific training operation established for mirrored release.", "creator_contribution": "Republication of existing text-to-SQL checkpoint; no new publisher training operation documented.", "architecture_and_code_lineage": "juierror FLAN-T5 text-to-SQL checkpoint and copied schema/prompt examples; upstream builder distinct from this uploader. Settled copy/history evidence reused.", "evidence_updated_at": "2026-10-31"}
{"model_id": "sathishphdai/cybersecurity-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/cybersecurity-slm-5m/raw/8113761fa089ce99f0ec708bf3c3e5ad62afb70f/README.md", "accessed_at": "2026-10-30T12:27:24.054934+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "dineth554/legion-coder-8m-10k", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "upstream_republication", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card repeatedly identifies pnny13/legion-coder-8m as original model repository, clone/deploy target and licence source. Supports republication of upstream44M pretrained code model; retain pnny13/DEATH LEGION attribution.", "provenance_sources": [{"url": "https://huggingface.co/dineth554/legion-coder-8m-10k/raw/0d9979c84a88544c9437ea68374375904b27fe97/README.md", "accessed_at": "2026-10-30T12:27:24.038407+00:00"}, {"url": "https://huggingface.co/dineth554/legion-coder-8m-10k/raw/0d9979c84a88544c9437ea68374375904b27fe97/README.md", "read_scope": "Cached model/base/architecture/training and attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "dineth554/legion-coder-8m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card credits DEATH LEGION and describes original44M code transformer,16000vocab and10000-step Stack-v2 pretraining. Own scratch training is supported inference; name8M differs from reported44M.", "provenance_sources": [{"url": "https://huggingface.co/dineth554/legion-coder-8m/raw/b106aa67f0ae2b4d752c70c93e334860365043c2/README.md", "accessed_at": "2026-10-30T12:27:24.252426+00:00"}, {"url": "https://huggingface.co/dineth554/legion-coder-8m/raw/b106aa67f0ae2b4d752c70c93e334860365043c2/README.md", "read_scope": "Cached model/base/architecture/training and attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "sathishphdai/uiux-designer-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/uiux-designer-slm-5m/raw/7fdbabd2c3aebba14bbe5246aac781d012143f3b/README.md", "accessed_at": "2026-10-30T12:27:24.361767+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/supply-chain-manager-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/supply-chain-manager-slm-5m/raw/ed0915145b027d97511093b0f2b58c99371e0aee/README.md", "accessed_at": "2026-10-30T12:27:24.576808+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/software-engineer-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/software-engineer-slm-5m/raw/83c3df17c353990663533c7ce536d30458c1e7f2/README.md", "accessed_at": "2026-10-30T12:27:24.654073+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/sales-executive-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/sales-executive-slm-5m/raw/dd3825c0e35c9df5267eb316db99a0f6c5a8bbd3/README.md", "accessed_at": "2026-10-30T12:27:24.650071+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/sdlc-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/sdlc-slm-5m/raw/f024a6dc347f789a0a9a2cc6c767ab0ba1b1cc0b/README.md", "accessed_at": "2026-10-30T12:27:24.700992+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/retail-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/retail-slm-5m/raw/1deb752d96fab927e4be873e51a5e21340cf6ea1/README.md", "accessed_at": "2026-10-30T12:27:24.868305+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/project-manager-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/project-manager-slm-5m/raw/ec57c1016dd299c49ab2ebd567586913b08bd8ce/README.md", "accessed_at": "2026-10-30T12:27:24.985285+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/product-manager-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/product-manager-slm-5m/raw/ee4fc83a5ba1c8986143607aab8724cf636eb525/README.md", "accessed_at": "2026-10-30T12:27:25.181772+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/operations-manager-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/operations-manager-slm-5m/raw/fa6eda5f9386d21ed31cee1c846b4d9568f90a19/README.md", "accessed_at": "2026-10-30T12:27:25.256734+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/marketing-manager-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/marketing-manager-slm-5m/raw/4a0597b1743de9f861d80d5a8e1183a0b63ee4cd/README.md", "accessed_at": "2026-10-30T12:27:25.247219+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/legal-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/legal-slm-5m/raw/8a166a3740d075641306c13bae698acce4cc54fe/README.md", "accessed_at": "2026-10-30T12:27:25.361925+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/healthcare-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/healthcare-slm-5m/raw/e629538d53d7e4d2033fce3e8bb3268361630b9e/README.md", "accessed_at": "2026-10-30T12:27:25.517024+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/hr-manager-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/hr-manager-slm-5m/raw/1bbaf66031348d3d34147a5f68bf0f958931a4e3/README.md", "accessed_at": "2026-10-30T12:27:25.687416+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/financial-analyst-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/financial-analyst-slm-5m/raw/50037e3c9b01895cf16fce8e116251e69e96504a/README.md", "accessed_at": "2026-10-30T12:27:25.836888+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/finance-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/finance-slm-5m/raw/a3fb4c9cc7298fa094abf29f5415ac3549663203/README.md", "accessed_at": "2026-10-30T12:27:25.884358+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/education-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/education-slm-5m/raw/d068d4097dd73095ffc07b8dedfa5b3930c58512/README.md", "accessed_at": "2026-10-30T12:27:25.893872+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/devops-engineer-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/devops-engineer-slm-5m/raw/5dba742264651d23e85b7e8173f29c3dfe92005d/README.md", "accessed_at": "2026-10-30T12:27:25.958498+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/data-scientist-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/data-scientist-slm-5m/raw/9ca3a7e411fb513370e1973375acb834694b26cb/README.md", "accessed_at": "2026-10-30T12:27:26.192390+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/customer-support-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/customer-support-slm-5m/raw/7a391f8a00f9fe1d42ed509c615fa1d8530d4090/README.md", "accessed_at": "2026-10-30T12:27:26.317672+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/content-writer-slm-5m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/content-writer-slm-5m/raw/372b4945c9423694c536e0d94d0c1d9043ee2054/README.md", "accessed_at": "2026-10-30T12:27:26.488349+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/supply-chain-manager-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/supply-chain-manager-slm-1m/raw/0eee4442182b16c6c1863f79f35c1fa55996304a/README.md", "accessed_at": "2026-10-30T12:27:26.509886+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/customer-support-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/customer-support-slm-1m/raw/b4dab0216c4d9669a236a1004af9df27f314dcea/README.md", "accessed_at": "2026-10-30T12:27:26.699359+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/content-writer-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/content-writer-slm-1m/raw/f696b69f2ead1ead1ec3a3bf62acaec26faa7b28/README.md", "accessed_at": "2026-10-30T12:27:26.597161+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/business-analyst-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/business-analyst-slm-1m/raw/c9dcfc7727173257fb262f7f7efa1870b6461d2a/README.md", "accessed_at": "2026-10-30T12:27:26.806686+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/project-manager-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/project-manager-slm-1m/raw/0670aa9807505eeeff57eb7658a482697df78bba/README.md", "accessed_at": "2026-10-30T12:27:26.900070+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/operations-manager-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/operations-manager-slm-1m/raw/9e4a5408d1d24b5958a8ba8d2572747618d73ac3/README.md", "accessed_at": "2026-10-30T12:27:27.098272+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/sales-executive-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/sales-executive-slm-1m/raw/23a9bde41b6a21104430bd5748220035aad40ab8/README.md", "accessed_at": "2026-10-30T12:27:27.222692+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/hr-manager-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/hr-manager-slm-1m/raw/ad578725add27925298d7fde9032ebee2cf84259/README.md", "accessed_at": "2026-10-30T12:27:27.206981+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/product-manager-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/product-manager-slm-1m/raw/6f8376177a588d0ebd54c68fa480336565bb1b13/README.md", "accessed_at": "2026-10-30T12:27:27.318480+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/uiux-designer-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/uiux-designer-slm-1m/raw/082f0db044b802d7afc554a58906ee2e31eda91f/README.md", "accessed_at": "2026-10-30T12:27:27.489958+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/devops-engineer-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/devops-engineer-slm-1m/raw/d62dca3f81a06b77975252ed525335ede38f5ca1/README.md", "accessed_at": "2026-10-30T12:27:27.509489+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/data-scientist-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/data-scientist-slm-1m/raw/33e58807dd2ee58c315705e970daa3fc4a6fc4ea/README.md", "accessed_at": "2026-10-30T12:27:27.732503+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/software-engineer-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/software-engineer-slm-1m/raw/7d0c7c5ef7c9e53dda67e6b68a17c4f91272a8bd/README.md", "accessed_at": "2026-10-30T12:27:27.881629+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/financial-analyst-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/financial-analyst-slm-1m/raw/e52a61815fabe6655bc1465648a5e9d46be1f8fd/README.md", "accessed_at": "2026-10-30T12:27:27.889698+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/marketing-manager-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/marketing-manager-slm-1m/raw/7ee1d934e770cc26b36395e2332961610e8ebaa6/README.md", "accessed_at": "2026-10-30T12:27:27.963611+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sopanm11/Mini-GPT-124M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sopanm11/Mini-GPT-124M/raw/5b11a56194ccc2e523669fdd61cc2a46940b7928/README.md", "accessed_at": "2026-10-30T12:27:28.139762+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/finance-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/finance-slm-1m/raw/d75ecd93015999a5631bde24f87671a84a3aac65/README.md", "accessed_at": "2026-10-30T12:27:28.132724+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/education-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/education-slm-1m/raw/69c4643bb087c531519bb60e1678d3bc5df78453/README.md", "accessed_at": "2026-10-30T12:27:28.338269+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/retail-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/retail-slm-1m/raw/6aa0596bd6be445c54ea75fce852e4ac4c2ca1c9/README.md", "accessed_at": "2026-10-30T12:27:28.529919+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/legal-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/legal-slm-1m/raw/65d097800b7cda8a66af7f9c8e0f6d1caa274423/README.md", "accessed_at": "2026-10-30T12:27:28.551440+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/cybersecurity-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/cybersecurity-slm-1m/raw/16b24d809a9719e6b055ddfb935151da0a90d1bc/README.md", "accessed_at": "2026-10-30T12:27:28.616022+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/healthcare-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/healthcare-slm-1m/raw/596b8e7d1180dadd847b6d5682d9fbf2a75db70f/README.md", "accessed_at": "2026-10-30T12:27:28.746767+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sathishphdai/sdlc-slm-1m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/sathishphdai/sdlc-slm-1m/raw/ae902ce1940f3c149b153c56220584c087c616a3/README.md", "accessed_at": "2026-10-30T12:27:28.717193+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "instruction-pretrain/InstructLM-500M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/instruction-pretrain/InstructLM-500M/raw/2158022d3e7be0c4f9668f0a45f8002c434d5432/README.md", "accessed_at": "2026-10-30T12:27:28.915354+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "xaskasdf/brandon-tiny-10m-instruct", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own tiny Brandon foundation pretraining, then knowledge distillation from own30M teacher and instruction fine-tuning. Selected instruct release is post-trained; bespoke10M base and explicit3-phase development support own scratch ancestry. GPT4o-mini supplies synthetic data, not model weights.", "provenance_sources": [{"url": "https://huggingface.co/xaskasdf/brandon-tiny-10m-instruct/raw/ca6329de991a1cfc17b2a7a1aaae2cc38c08577a/README.md", "accessed_at": "2026-10-30T12:27:29.181441+00:00"}, {"url": "https://huggingface.co/xaskasdf/brandon-tiny-10m-instruct/raw/ca6329de991a1cfc17b2a7a1aaae2cc38c08577a/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "EleutherAI/pythia-31m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-31m/raw/e556ace21b489575e94e9d50b6dad2fcc7419679/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:27:29.154897+00:00"}, {"url": "https://github.com/EleutherAI/pythia", "basis": "dated_repository_changelog", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-14m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-14m/raw/cf967c0a9a04383db6f7b1108d86b2962634b4ac/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:27:29.249047+00:00"}, {"url": "https://github.com/EleutherAI/pythia", "basis": "dated_repository_changelog", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "NNEngine/TinyWay-1.1.0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/NNEngine/TinyWay-1.1.0/raw/ca7128c7d66729653eb4f19f585dcc89ea180a7c/README.md", "accessed_at": "2026-10-30T12:27:29.378362+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "bcml/bclm-1-small-preview", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "sft_of_own_base", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original123M Bulldog/BCLM architecture and explicit24000-step pretraining followed by6000-step SFT on H200. Final checkpoint post-trains publisher-owned scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/bcml/bclm-1-small-preview/raw/5a48f7ccceaf283cc41ae2df02ca4dc536f1a7e4/README.md", "accessed_at": "2026-10-30T12:27:29.387457+00:00"}, {"url": "https://huggingface.co/bcml/bclm-1-small-preview/raw/5a48f7ccceaf283cc41ae2df02ca4dc536f1a7e4/README.md", "read_scope": "Cached card model/architecture/training/result and upstream description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "abirmaheshwari/abirhinv1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/abirmaheshwari/abirhinv1/raw/991f74dbc5f072b456fe1e0b53ff5922b76f53b1/README.md", "accessed_at": "2026-10-30T12:27:29.530489+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AnthonyPa57/Cirilla-0.3B-4E-grpo-icl", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card describes a final domain-specific fine-tuning stage, so this published checkpoint is a post-training release.", "provenance_sources": [{"url": "https://huggingface.co/AnthonyPa57/Cirilla-0.3B-4E-grpo-icl/raw/1cf41dd18334135d05463cf5e5e8a02ddc5dcc73/README.md", "accessed_at": "2026-10-30T12:27:29.762669+00:00", "quote": "ngside lore retention. (21MiB, 50.5K data points) Domain-Specific Fine-Tuning The final stage focused on activating the model's conversational abilities and aligning it with the ingested lore. This involved training on 185K synthetic que", "basis": "Card describes a final domain-specific fine-tuning stage, so this published checkpoint is a post-training release."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AnthonyPa57/Cirilla-0.3B-4E-grpo", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card describes a final domain-specific fine-tuning stage, so this published checkpoint is a post-training release.", "provenance_sources": [{"url": "https://huggingface.co/AnthonyPa57/Cirilla-0.3B-4E-grpo/raw/aec24f8445595e0cdf63960d2d6327b2bc564abc/README.md", "accessed_at": "2026-10-30T12:27:29.820679+00:00", "quote": "ngside lore retention. (21MiB, 50.5K data points) Domain-Specific Fine-Tuning The final stage focused on activating the model's conversational abilities and aligning it with the ingested lore. This involved training on 185K synthetic que", "basis": "Card describes a final domain-specific fine-tuning stage, so this published checkpoint is a post-training release."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AnthonyPa57/Cirilla-0.3B-4E", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card describes a final domain-specific fine-tuning stage, so this published checkpoint is a post-training release.", "provenance_sources": [{"url": "https://huggingface.co/AnthonyPa57/Cirilla-0.3B-4E/raw/002ee017ad8166a2a77dada72c13edb7f6c40d61/README.md", "accessed_at": "2026-10-30T12:27:29.859883+00:00", "quote": "ngside lore retention. (21MiB, 50.5K data points) Domain-Specific Fine-Tuning The final stage focused on activating the model's conversational abilities and aligning it with the ingested lore. This involved training on 185K synthetic que", "basis": "Card describes a final domain-specific fine-tuning stage, so this published checkpoint is a post-training release."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "0rn0/llama2-15m-tinystories-sft", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/0rn0/llama2-15m-tinystories-sft/raw/4d69f1c13019afcf5c50b621c73cb9c92e01bb4e/README.md", "accessed_at": "2026-10-30T12:27:30.036628+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "0rn0/llama2-15m-tinystories", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly attributes the released pretrained weights to an upstream model and describes unchanged redistribution, rather than publisher training.", "provenance_sources": [{"url": "https://huggingface.co/0rn0/llama2-15m-tinystories/raw/cd75d4bc6b80d5a8376bf90cf70a2ba1bb791985/README.md", "accessed_at": "2026-10-30T12:27:30.010607+00:00", "quote": "face.co/datasets/roneneldan/TinyStories) dataset. Pretrained by [Andrej Karpathy](https://github.com/karpathy/llama2.c) (stories15M checkpoint), uploaded here for easy loading and fine-tuning. Model Details | Parameter | Va", "basis": "Card explicitly attributes the released pretrained weights to an upstream model and describes unchanged redistribution, rather than publisher training."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "anthonym21/Eve-2-MoE-272M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/anthonym21/Eve-2-MoE-272M/raw/12c3a397d4a1d99b5218ed58754fe8888183dc9f/README.md", "accessed_at": "2026-10-30T12:27:30.173960+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "fibonacciai/Iran-v1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved Persian card reports original Yasin GQA/RoPE/SwiGLU architecture designed from the ground up for Persian, trained on70M Persian texts with training loss/step results. Supports publisher-owned scratch-origin model; preserve Fibonacci/Yasin team attribution.", "provenance_sources": [{"url": "https://huggingface.co/fibonacciai/Iran-v1/raw/333f50544f23b1ec65cee3e8770b6fba6f967146/README.md", "accessed_at": "2026-10-30T12:27:30.391498+00:00"}, {"url": "https://huggingface.co/fibonacciai/Iran-v1/raw/333f50544f23b1ec65cee3e8770b6fba6f967146/README.md", "read_scope": "Cached card model/architecture/training/result and upstream description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "geoffsee/auto-g-nano-153m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Publisher describes FineWeb-Edu training of own modernized nanoGPT model; selected source builds bespoke GQA/RoPE/RMSNorm modules fresh and has no external pretrained weight loader. Same creator documents scratch10M predecessor. Own153M scratch origin is the best-supported medium inference; first training execution record absent.", "provenance_sources": [{"url": "https://huggingface.co/geoffsee/auto-g-nano-153m/raw/6d48ec142622d35555b9db6b2ae0e20065c639e0/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/geoffsee/auto-g-nano-153m/raw/6d48ec142622d35555b9db6b2ae0e20065c639e0/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/geoffsee/auto-g-nano-153m/raw/6d48ec142622d35555b9db6b2ae0e20065c639e0/model.py", "scope": "Model constructor and architecture implementation read", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not stated for exact run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "creator_contribution": "Own custom model training on FineWeb-Edu sample, modernized decoder implementation.", "provenance_confidence": "medium", "architecture_lineage": "Publisher explicitly calls it nanoGPT-style / nanoGPT evolution; Llama-style RoPE/RMSNorm/SwiGLU/GQA, standard GPT-2 tiktoken tokenizer.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "geoffsee/auto-g-nano-10m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "auto-g-nano-10m explicitly from scratch on TinyShakespeare;153m reports FineWeb-Edu training. Exact-card hardware absent. 1b model's rented2xRTXPro6000Blackwell not attributed to these two.", "provenance_sources": [{"url": "https://huggingface.co/geoffsee/auto-g-nano-10m/raw/84e5c4b4fcecd0c949e73dc2b938892bc0f27ce2/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not stated for exact run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Own custom model training on TinyShakespeare, basic decoder implementation.", "architecture_lineage": "Publisher explicitly calls it nanoGPT-style / nanoGPT evolution; Standard character GPT decoder with learned positions, LayerNorm/ReLU and65-character vocabulary; TinyShakespeare data.", "evidence_updated_at": "2026-11-01"}
{"model_id": "EleutherAI/pythia-31m-deduped", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-31m-deduped/raw/b7782556ba7adfb4730d9bda7d12aa44d88fa132/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:27:30.599659+00:00"}, {"url": "https://github.com/EleutherAI/pythia", "basis": "dated_repository_changelog", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-14m-deduped", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-14m-deduped/raw/7386d9a4ae45aef494a6e704910394def3037fc5/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:27:30.654782+00:00"}, {"url": "https://github.com/EleutherAI/pythia", "basis": "dated_repository_changelog", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "0rn0/gpt2-125m-tinystories", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/0rn0/gpt2-125m-tinystories/raw/7d126df45d73d0a62530db5eebd7fe8bf7276e9a/README.md", "accessed_at": "2026-10-30T12:27:30.799236+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "0rn0/gpt2-30m-tinystories", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/0rn0/gpt2-30m-tinystories/raw/55e4cbcf8b45ca14571fed9bfa8eb5a3fc75a51e/README.md", "accessed_at": "2026-10-30T12:27:31.040629+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Exquisique/BabyLangModel", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "BabyLangModel from-scratch200ksteps RTX4060; GrownUpBaby110M from-scratchCosmopedia750Mtokens, laptop RTX4060 8GB about31h. Configs/commits reveal no country.", "provenance_sources": [{"url": "https://huggingface.co/Exquisique/BabyLangModel/raw/f82a831a67031c92cc89b0d8849c6d72b7107c7f/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/Exquisique/BabyLangModel/raw/f82a831a67031c92cc89b0d8849c6d72b7107c7f/README.md", "access_date": "2026-10-31", "scope": "complete card/config read"}, {"url": "https://huggingface.co/Exquisique/BabyLangModel/raw/f82a831a67031c92cc89b0d8849c6d72b7107c7f/config.json", "access_date": "2026-10-31", "scope": "complete card/config read"}, {"url": "https://huggingface.co/Exquisique/BabyLangModel/raw/f82a831a67031c92cc89b0d8849c6d72b7107c7f/model.py", "access_date": "2026-10-31", "scope": "model architecture header, component declarations, full fresh initialization and own-checkpoint inference paths"}, {"url": "https://github.com/Exquisique/Babymodel/blob/main/src/train.py", "access_date": "2026-10-31", "scope": "complete training entrypoint read via GitHub connector; own resume or fresh GPT constructor, no external pretrained initialization"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "NVIDIA GeForce RTX 4060; exact desktop/laptop variant not stated for this run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "compute_source": "Publisher explicit hardware statement; own consumer GPU wording.", "creator_contribution": "Own TinyStories scratch training, manual PyTorch loop with own-checkpoint resume, and publishing runtime integration.", "architecture_and_code_lineage": "Standard GPT learned-position/LayerNorm/GELU design explicitly inspired by Karpathy nanoGPT, credited in card and linked repository README; custom Transformers-compatible implementation. No external pretrained weight base established.", "hardware": "NVIDIA GeForce RTX 4060; exact desktop/laptop variant not stated for this run", "compute_sources": [{"url": "https://huggingface.co/Exquisique/BabyLangModel/raw/f82a831a67031c92cc89b0d8849c6d72b7107c7f/README.md", "access_date": "2026-10-31", "scope": "complete card/config read"}, {"url": "https://huggingface.co/Exquisique/BabyLangModel/raw/f82a831a67031c92cc89b0d8849c6d72b7107c7f/config.json", "access_date": "2026-10-31", "scope": "complete card/config read"}, {"url": "https://huggingface.co/Exquisique/BabyLangModel/raw/f82a831a67031c92cc89b0d8849c6d72b7107c7f/model.py", "access_date": "2026-10-31", "scope": "model architecture header, component declarations, full fresh initialization and own-checkpoint inference paths"}, {"url": "https://github.com/Exquisique/Babymodel/blob/main/src/train.py", "access_date": "2026-10-31", "scope": "complete training entrypoint read via GitHub connector; own resume or fresh GPT constructor, no external pretrained initialization"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "Exquisique/GrownUpBaby", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "BabyLangModel from-scratch200ksteps RTX4060; GrownUpBaby110M from-scratchCosmopedia750Mtokens, laptop RTX4060 8GB about31h. Configs/commits reveal no country.", "provenance_sources": [{"url": "https://huggingface.co/Exquisique/GrownUpBaby/raw/f6328aa05567c289801f3dcce7608f30a6c89920/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/Exquisique/GrownUpBaby/raw/f6328aa05567c289801f3dcce7608f30a6c89920/README.md", "access_date": "2026-10-31", "scope": "complete card/config read"}, {"url": "https://huggingface.co/Exquisique/GrownUpBaby/raw/f6328aa05567c289801f3dcce7608f30a6c89920/config.json", "access_date": "2026-10-31", "scope": "complete card/config read"}, {"url": "https://huggingface.co/Exquisique/GrownUpBaby/raw/f6328aa05567c289801f3dcce7608f30a6c89920/model.py", "access_date": "2026-10-31", "scope": "model architecture header, component declarations, full fresh initialization and own-checkpoint inference paths"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "1 x NVIDIA GeForce RTX 4060 Laptop GPU,8GB VRAM", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "training_duration_reported": "Approximately31hours across2epochs", "compute_source": "Explicit frozen card hardware statement.", "creator_contribution": "Publisher states own from-scratch Cosmopedia stories pretraining; selected checkpointstep23000, custom small Gemma-style implementation and training loop.", "architecture_and_code_lineage": "Explicit Gemma-style RMSNorm/RoPE/GeGLU/GQA adaptation with custom implementation and GPT-2 tokenizer, not evidence of using pretrained Gemma weights. Frozen model class has fresh constructor/post_init and own-checkpoint loading override.", "hardware": "1 x NVIDIA GeForce RTX 4060 Laptop GPU,8GB VRAM", "compute_sources": [{"url": "https://huggingface.co/Exquisique/GrownUpBaby/raw/f6328aa05567c289801f3dcce7608f30a6c89920/README.md", "access_date": "2026-10-31", "scope": "complete card/config read"}, {"url": "https://huggingface.co/Exquisique/GrownUpBaby/raw/f6328aa05567c289801f3dcce7608f30a6c89920/config.json", "access_date": "2026-10-31", "scope": "complete card/config read"}, {"url": "https://huggingface.co/Exquisique/GrownUpBaby/raw/f6328aa05567c289801f3dcce7608f30a6c89920/model.py", "access_date": "2026-10-31", "scope": "model architecture header, component declarations, full fresh initialization and own-checkpoint inference paths"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "mkurman/ConvGPT-0.2B-SYNTH-250B-EC", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original ConvGPT/NeuroBLAST custom architecture and authored SYNTH pretraining checkpoints with118B-250B tokens on TPUv4-16 support own scratch-origin lineage; JAX release retains same authors150k-step checkpoint lineage.", "provenance_sources": [{"url": "https://huggingface.co/mkurman/ConvGPT-0.2B-SYNTH-250B-EC/raw/7b334b192683bbd0e36f9d8482fd4edda18f8ea5/README.md", "accessed_at": "2026-10-30T12:27:31.209576+00:00"}, {"url": "https://huggingface.co/mkurman/ConvGPT-0.2B-SYNTH-250B-EC/raw/7b334b192683bbd0e36f9d8482fd4edda18f8ea5/README.md", "read_scope": "Saved card architecture, training and checkpoint description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "RinKana/OtterLM-WKT103-110M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/RinKana/OtterLM-WKT103-110M/raw/745929046b544b558d062bc08abd3fe6a92bb180/README.md", "accessed_at": "2026-10-30T12:27:31.291822+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Berang-Berang/OtterLM-WKT103-110M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/Berang-Berang/OtterLM-WKT103-110M/raw/4d54cd5004c050358d6a46c9ef223cefeabddfd0/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "RinKana/MiniGPT-30M-Wikipedia-Var1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/RinKana/MiniGPT-30M-Wikipedia-Var1/raw/91063e9e173528f0d280679c70025b900662095b/README.md", "accessed_at": "2026-10-30T12:27:31.656330+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "RinKana/MiniGPT-30M-Wikipedia", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly calls the30M model built from scratch and records completed WikiText103 training for1epoch on2KaggleT4s.", "provenance_sources": [{"url": "https://huggingface.co/RinKana/MiniGPT-30M-Wikipedia/raw/009e72e473def7d67947ba93e872b2d8d498da66/README.md", "accessed_at": "2026-10-30T12:27:31.742861+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "JorgeVanco/diffusionGPT", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/JorgeVanco/diffusionGPT/raw/82be92f1c2366f08a428e406f1fb389f48e187f9/README.md", "accessed_at": "2026-10-30T12:27:31.768409+00:00", "quote": "GPT is a Discrete Diffusion Language Model (MDLM) fine-tuned for conversational AI. Unlike traditional autoregressive models (like GPT-4 or Llama) that predict text one token at a time from left to right, DiffusionGPT generates text through a", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "psyche/kogpt", "model_provenance": "unresolved", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card calls this pretrained KoreanGPT2 with expanded context/embedding size, but does not distinguish scratch Korean pretraining from expanded pretrained weights. Model discussions are conversion/evaluation requests, no training-origin answer.", "provenance_sources": [{"url": "https://huggingface.co/psyche/kogpt/raw/45e405d32ab6d09dc32320b36812243e1aad1b96/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Card calls this pretrained KoreanGPT2 with expanded context/embedding size, but does not distinguish scratch Korean pretraining from expanded pretrained weights. Model discussions are conversion/evaluation requests, no training-origin answer."}
{"model_id": "NNEngine/TinyWay-1.0.0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/NNEngine/TinyWay-1.0.0/raw/2f3222e9100c768ba4c9b01053ba2a830516a243/README.md", "accessed_at": "2026-10-30T12:27:31.944242+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Zual/chess_char", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original8-layer512hidden GPT-style model with32-characterPGN vocabulary, explicit5epoch Lichess training and own custom tokenizer support scratch-origin chess character model.", "provenance_sources": [{"url": "https://huggingface.co/Zual/chess_char/raw/9b1821fdbd12fb3e8341e4067ffa0deeb5a62d15/README.md", "accessed_at": "2026-10-30T12:27:32.049679+00:00"}, {"url": "https://huggingface.co/Zual/chess_char/raw/9b1821fdbd12fb3e8341e4067ffa0deeb5a62d15/README.md", "read_scope": "Saved card architecture/base/training/result and upstream attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "anarlavrenov/lime-1b-instruct", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/anarlavrenov/lime-1b-instruct/raw/c49cfb9c7bc0e3526712890264686d8b1d3a3eab/README.md", "accessed_at": "2026-10-30T12:27:32.269706+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "rocky1410/haipai-nano", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes authored Haipai custom small architecture, base-language-model training corpus and completed trained weights/run or benchmark results.", "provenance_sources": [{"url": "https://huggingface.co/rocky1410/haipai-nano/raw/628d3abc3232359b8c951da3b71ea8c3c3226532/README.md", "accessed_at": "2026-10-30T12:27:32.368477+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "vietrix/viena-tiny-demo-pretrain", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card identifies tiny bespoke Viena pretraining-only checkpoint with own small SentencePiece tokenizer, synthetic offline corpus and50 pretraining steps. Own fresh training is supported; limited demonstration scale does not make it untrained.", "provenance_sources": [{"url": "https://huggingface.co/vietrix/viena-tiny-demo-pretrain/raw/20bd5337f2d83bd17df0f28245631e740196e5a6/README.md", "accessed_at": "2026-10-30T12:27:32.386997+00:00"}, {"url": "https://huggingface.co/vietrix/viena-tiny-demo-pretrain/raw/20bd5337f2d83bd17df0f28245631e740196e5a6/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "vietrix/viena-tiny-demo", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/vietrix/viena-tiny-demo/raw/d3b58b63ae9129c6b37106e4075ecf84b5428725/README.md", "accessed_at": "2026-10-30T12:27:32.467082+00:00", "quote": "a Tiny Demo (SFT) This is a tiny, demo-only Viena checkpoint fine-tuned for instruction following. It is not production quality. It is intended for smoke tests and workflow validation. Model description - Architecture: decoder-only Transformer (VienaModel) ", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Hai929/The_GuageLLM_23M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The_GuageLLM_23M card explicitly claims from-scratch GPT-style23M training; config/history examined; no hardware/site or institution.", "provenance_sources": [{"url": "https://huggingface.co/Hai929/The_GuageLLM_23M/raw/dc4865e420aca89e9ae229789341ad6ce4de5188/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/Hai929/The_GuageLLM_23M/raw/dc4865e420aca89e9ae229789341ad6ce4de5188/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Hai929/The_GuageLLM_23M/raw/dc4865e420aca89e9ae229789341ad6ce4de5188/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Hai929/The_GuageLLM_23M/raw/dc4865e420aca89e9ae229789341ad6ce4de5188/chat_template.jinja", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Hai929/The_GuageLLM_23M/raw/dc4865e420aca89e9ae229789341ad6ce4de5188/tokenizer_config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not stated for exact run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Training corpus, step count, exact hardware and duration absent from frozen public files/comments.", "creator_contribution": "Own small GPT-style scratch-training claim and custom-tokenizer/chat-template packaging. Instruction-tuning and character-level tags are not supported by a documented SFT run or actual character vocabulary; actual config vocab32000.", "architecture_and_code_lineage": "Standard Transformers GPT2LMHeadModel with6layers,width384,6heads,64tokencontext and32000tokenvocabulary. Publisher explicitly claims scratch training; no custom model class present. GPT2style architecture is reused separately from weight ancestry.", "evidence_updated_at": "2026-10-31"}
{"model_id": "AjithBharadwaj/nano-llama-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author describes a bespoke166M Llama-style base with GPT2 tokenization, actual FineWeb training and optimizer recipe; original small-model pretraining is the best-supported origin.", "provenance_sources": [{"url": "https://huggingface.co/AjithBharadwaj/nano-llama-base/raw/25974ad16bc4286dc4f84f8b44dbb348b15c6b98/README.md", "accessed_at": "2026-10-30T12:27:32.671301+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "younsmirzai/stackadvisor-extractor", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "probably_yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card explicitly identifies T5-small specialized for OCR delivery-order extraction. Best-supported reading is a task-specific T5-small derivative.", "provenance_sources": [{"url": "https://huggingface.co/younsmirzai/stackadvisor-extractor/raw/fe770e5b63d709bcf786324e0eab7ac37817ae14/README.md", "accessed_at": "2026-10-30T12:27:32.871054+00:00"}, {"url": "https://huggingface.co/younsmirzai/stackadvisor-extractor", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_explicit_base_model_and_task", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Jiraya/zoof-v1.2-394M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes the author's self-implemented394M transformer, depth-scaled initialization and actual79B-token FineWeb-Edu pretraining with measured benchmarks.", "provenance_sources": [{"url": "https://huggingface.co/Jiraya/zoof-v1.2-394M/raw/c4cec4b397c9f0f51e4b30a9648e5be3d032d4e9/README.md", "accessed_at": "2026-10-30T12:27:33.015773+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "rocky1410/haipai-micro", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes authored Haipai custom small architecture, base-language-model training corpus and completed trained weights/run or benchmark results.", "provenance_sources": [{"url": "https://huggingface.co/rocky1410/haipai-micro/raw/2d2b4ff8df40b241538a6a08ac79618b61311e59/README.md", "accessed_at": "2026-10-30T12:27:33.024288+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "EleutherAI/pythia-160m-seed3", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-160m-seed3/raw/a74886a25b5f4ffb3b9a0fa80bd02750f13461bd/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:27:33.343740+00:00"}, {"url": "https://uk.linkedin.com/in/pietrolesci", "basis": "self_reported_profile_location_search_index", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2503.09543v2", "basis": "paper_joint_first_author_with_University_of_Cambridge_affiliation", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2503.09543v1", "basis": "Primary public page read", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-160m-seed2", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-160m-seed2/raw/6e9b9f6270d1d624f9d7188c1e8909600c4efae2/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:27:33.204210+00:00"}, {"url": "https://uk.linkedin.com/in/pietrolesci", "basis": "self_reported_profile_location_search_index", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2503.09543v2", "basis": "paper_joint_first_author_with_University_of_Cambridge_affiliation", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2503.09543v1", "basis": "Primary public page read", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-160m-seed1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-160m-seed1/raw/5ed55222b3092e69edcfa5fa646b7aac0b7e03d9/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:27:33.293430+00:00"}, {"url": "https://uk.linkedin.com/in/pietrolesci", "basis": "self_reported_profile_location_search_index", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2503.09543v2", "basis": "paper_joint_first_author_with_University_of_Cambridge_affiliation", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2503.09543v1", "basis": "Primary public page read", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ethanthoma/synth-gpt-110m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes the authored110M GPT with T5-tokenizer vocabulary, base-only pretraining on30B SYNTH tokens, Muon training and H100 hardware.", "provenance_sources": [{"url": "https://huggingface.co/ethanthoma/synth-gpt-110m/raw/2b5f12775c4d6308c0326744133e72578697d37b/README.md", "accessed_at": "2026-10-30T12:27:33.538977+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "FarhanAK128/CustomGPT", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/FarhanAK128/CustomGPT/raw/4bedcce61fa4b2dceb91d1161189fe5062bda00a/README.md", "accessed_at": "2026-10-30T12:27:33.653227+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "QuarkML/QMoE-400", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card reports custom400M sparse-MoE pretraining in JAX/Flax on OpenWebText with8TPUv3 chips,79100steps and actual train/validation/router losses.", "provenance_sources": [{"url": "https://huggingface.co/QuarkML/QMoE-400/raw/93123fb1052c727913ad0db7df4221f0f309c051/README.md", "accessed_at": "2026-10-30T12:27:33.638133+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "pszemraj/mega-ar-small-4096-wikitext-103-raw-v1", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/pszemraj/mega-ar-small-4096-wikitext-103-raw-v1/raw/9dc0f2f710de71cdaf105fb4736066aad30c75e2/README.md", "accessed_at": "2026-10-30T12:27:33.806685+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "pszemraj/bart-base-instruct-dolly_hhrlhf", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/pszemraj/bart-base-instruct-dolly_hhrlhf/raw/f91a5b688c7da9633a31efae7ecbfd6a8fdb2083/README.md", "accessed_at": "2026-10-30T12:27:33.928396+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "pszemraj/flan-t5-base-instruct-dolly_hhrlhf", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/pszemraj/flan-t5-base-instruct-dolly_hhrlhf/raw/b1aa3f18804112809696adad4b7b14c51f2e6261/README.md", "accessed_at": "2026-10-30T12:27:33.993730+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "pszemraj/karnold-walmer-base-biopapers", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/pszemraj/karnold-walmer-base-biopapers/raw/1e8efe24bf6473bad8ed75e9445fd35ce27b43a9/README.md", "accessed_at": "2026-10-30T12:27:34.178329+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "pszemraj/bart-base-instructiongen-w-inputs", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/pszemraj/bart-base-instructiongen-w-inputs/raw/3820debe64999685695a31f98ec6ea190cd46d2c/README.md", "accessed_at": "2026-10-30T12:27:34.270104+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "pszemraj/distilgpt2-HC3", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/pszemraj/distilgpt2-HC3/raw/f380db02de3bc4b2c13614b7a90bbfa8736fc1ca/README.md", "accessed_at": "2026-10-30T12:27:34.305198+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "pszemraj/gpt2-medium-vaguely-human-dialogue", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/pszemraj/gpt2-medium-vaguely-human-dialogue/raw/2126fd2ff7ca717d31ef0328c2a112cc7142d7ca/README.md", "accessed_at": "2026-10-30T12:27:34.440899+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "pszemraj/Ballpark-Trivia-L", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/pszemraj/Ballpark-Trivia-L/raw/e3f069dd6f682adeb0f08ef5db9ee77f7a170c8b/README.md", "accessed_at": "2026-10-30T12:27:34.555949+00:00", "quote": "rl.ai/projects/wizardofwikipedia/) for 40k steps (34/36 layers frozen for the fine-tuning), and then subsequently trained for 40k steps on a parsed variant of [Natural Questions](https://ai.google.com/research/NaturalQuestions)(a", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "tsingla98/frawdllm-100m", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "post_training_of_own_scratch_model", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "The card describes original from-scratch pretraining in this project followed by fine-tuning or post-training; the released checkpoint is the later stage.", "provenance_sources": [{"url": "https://huggingface.co/tsingla98/frawdllm-100m/raw/f74ba30cf03fb8928c261795cd7ee6c69e0f0e21/README.md", "accessed_at": "2026-10-30T12:27:34.638991+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "i3-lab/i3-Ethan-Base", "model_provenance": "unresolved", "own_from_scratch_lineage": "unknown", "trained_status": "unknown", "checkpoint_operation": "not_established", "provenance_evidence_level": "unresolved", "provenance_basis": "Pinned card contains onlyFineWeb-Edu metadata and rwkv-hybrid custom config. No training code, initial weights or completed-run description; only model-community thread is a format conversion.", "provenance_sources": [{"url": "https://huggingface.co/i3-lab/i3-Ethan-Base/raw/4d1756f49bd93784efea4e01365711cf46947834/README.md", "accessed_at": "2026-10-30T12:27:34.795704+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Pinned card contains onlyFineWeb-Edu metadata and rwkv-hybrid custom config. No training code, initial weights or completed-run description; only model-community thread is a format conversion."}
{"model_id": "fromthesky/PLDR-LLM-v52-110M-1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v52-110M-1/raw/d83e24e7984cbd01fdf9e62f4ab45b22750a3080/README.md", "accessed_at": "2026-10-30T12:27:34.950958+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v52-110M-1/raw/d83e24e7984cbd01fdf9e62f4ab45b22750a3080/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51-104M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-104M/raw/242efa831fde7ac536e0871a315941b540b418b3/README.md", "accessed_at": "2026-10-30T12:27:34.972557+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-104M/raw/242efa831fde7ac536e0871a315941b540b418b3/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51-110M-1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-110M-1/raw/1813d20afee62f9811ec297ef41b61369ae1b91f/README.md", "accessed_at": "2026-10-30T12:27:35.072794+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-110M-1/raw/1813d20afee62f9811ec297ef41b61369ae1b91f/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51-110M-2", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-110M-2/raw/0cee57d1d5cba8f4356b3ec085d1adb04fa9bc70/README.md", "accessed_at": "2026-10-30T12:27:35.263611+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-110M-2/raw/0cee57d1d5cba8f4356b3ec085d1adb04fa9bc70/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51-110M-3", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-110M-3/raw/711d436aae02019f89a44ec0f0d3d72204731ddd/README.md", "accessed_at": "2026-10-30T12:27:35.272646+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-110M-3/raw/711d436aae02019f89a44ec0f0d3d72204731ddd/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51-110M-4", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-110M-4/raw/e095aa7813ccdfbd6096f74e43a86043bef1ded3/README.md", "accessed_at": "2026-10-30T12:27:35.427341+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-110M-4/raw/e095aa7813ccdfbd6096f74e43a86043bef1ded3/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51-110M-5", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-110M-5/raw/1dfb02e0120f275ffc91a8125349f334a5dbd975/README.md", "accessed_at": "2026-10-30T12:27:35.542429+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-110M-5/raw/1dfb02e0120f275ffc91a8125349f334a5dbd975/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51-DAG-110M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-DAG-110M/raw/a98493354811b4e16e6cfec7f87f9059f23b5550/README.md", "accessed_at": "2026-10-30T12:27:35.610969+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51-DAG-110M/raw/a98493354811b4e16e6cfec7f87f9059f23b5550/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51G-106M-1", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51G-106M-1/raw/0950b636c2deb4d4c74c946616fd5564029319fe/README.md", "accessed_at": "2026-10-30T12:27:35.749354+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51G-106M-1/raw/0950b636c2deb4d4c74c946616fd5564029319fe/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51G-106M-2", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51G-106M-2/raw/6dee34bbe5d4b6489609bc8416f298abac4068cd/README.md", "accessed_at": "2026-10-30T12:27:35.903661+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51G-106M-2/raw/6dee34bbe5d4b6489609bc8416f298abac4068cd/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51G-106M-3", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51G-106M-3/raw/c6a7ac56bba37fa3f036bb94b81bca0b90736170/README.md", "accessed_at": "2026-10-30T12:27:35.906662+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51G-106M-3/raw/c6a7ac56bba37fa3f036bb94b81bca0b90736170/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "AnthonyPa57/ChemMultiRegressor-65M-38H", "model_provenance": "unresolved", "own_from_scratch_lineage": "unknown", "trained_status": "unknown", "checkpoint_operation": "not_established", "provenance_evidence_level": "unresolved", "provenance_basis": "Card calls this a random demonstration model, while metadata reports a loss and optimizer states exist. Custom MoE/regression config is recorded, but actual completed training versus an initialisation demo remains ambiguous.", "provenance_sources": [{"url": "https://huggingface.co/AnthonyPa57/ChemMultiRegressor-65M-38H/raw/23e6ac4683eb1943298bdbed8edb8b4f505fe597/README.md", "accessed_at": "2026-10-30T12:27:36.049901+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Card calls this a random demonstration model, while metadata reports a loss and optimizer states exist. Custom MoE/regression config is recorded, but actual completed training versus an initialisation demo remains ambiguous."}
{"model_id": "kmkrworks/LiteGPT-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/kmkrworks/LiteGPT-Base/raw/62ccc46d556de38b54f021eca71826debc59a96a/README.md", "accessed_at": "2026-10-30T12:27:36.183434+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "RessAI/Onner-300m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Best supported origin is own scratch pretraining: publisher reports own bespoke FineWeb-Edu model, with matching unusual actual configuration and no named external checkpoint. Medium inference; original training script not supplied.", "provenance_sources": [{"url": "https://huggingface.co/RessAI/Onner-300m/raw/5b5d13ff58f16c0a88f336e6c3c2bc43372d9f38/README.md"}, {"url": "https://huggingface.co/RessAI/Onner-300m/raw/5b5d13ff58f16c0a88f336e6c3c2bc43372d9f38/README.md", "scope": "Complete saved frozen card read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/RessAI/Onner-300m/raw/5b5d13ff58f16c0a88f336e6c3c2bc43372d9f38/config.json", "scope": "Complete actual frozen architecture configuration read", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "creator_contribution": "Reported training/configuration of own approximately 200M dense decoder model on FineWeb-Edu.", "best_origin_estimate": {"origin": "own_from_scratch_backbone", "confidence": "medium", "basis": "Publisher describes training own bespoke approximately 200M architecture on FineWeb-Edu; frozen actual config matches custom 12-layer, 768-wide, 2 KV-head and 128256-vocabulary architecture, without identified upstream checkpoint.", "remaining_gap": "Earliest initialization/training script is not published."}, "provenance_confidence": "medium", "architecture_lineage": "Publisher explicitly describes BERT-base sizing combined with Llama causal attention mechanisms; custom RessAiForCausalLM configuration, no implementation file supplied.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "mkurman/NeuroBLAST-V3-SYNTH-EC-150000-JAX", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "own_checkpoint_format_export", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original ConvGPT/NeuroBLAST custom architecture and authored SYNTH pretraining checkpoints with118B-250B tokens on TPUv4-16 support own scratch-origin lineage; JAX release retains same authors150k-step checkpoint lineage.", "provenance_sources": [{"url": "https://huggingface.co/mkurman/NeuroBLAST-V3-SYNTH-EC-150000-JAX/raw/e35a8852555e449c2636e53eb0bec2c258c92e08/README.md", "accessed_at": "2026-10-30T12:27:36.411704+00:00"}, {"url": "https://huggingface.co/mkurman/NeuroBLAST-V3-SYNTH-EC-150000-JAX/raw/e35a8852555e449c2636e53eb0bec2c258c92e08/README.md", "read_scope": "Saved card architecture, training and checkpoint description", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Madras1/MTLM2-40M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Custom41.5M TinyGPT base model, own14B-token JAX/Flax training run, TPUv5e-8 and measured completion performance; card credits Madras1 (Gabriel). Best-supported reading is own from-scratch pretraining.", "provenance_sources": [{"url": "https://huggingface.co/Madras1/MTLM2-40M/raw/ba8622095037071416b7c9acc18f18f8e38984c2/README.md"}, {"url": "https://huggingface.co/Madras1/MTLM2-40M", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Google Cloud TPUv5e-8;~6hours", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "i3-lab/i3-200m-v2", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes the author's bespoke i3 architecture/tokenizer and completed initial or multi-dataset pretraining, with run steps, losses and hardware.", "provenance_sources": [{"url": "https://huggingface.co/i3-lab/i3-200m-v2/raw/42d84969502ffaf9dc3e06066afc2b546221ee12/README.md", "accessed_at": "2026-10-30T12:27:36.552016+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "cccczshao/CALM-Autoencoder", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Official linked training recipe builds the Autoencoder from configuration, with no model_name_or_path checkpoint, and trains it on about15B tokens. Published75M checkpoint is reported with over99.9% token reconstruction; reused Llama3 tokenizer is separate from model weights.", "provenance_sources": [{"url": "https://huggingface.co/cccczshao/CALM-Autoencoder/raw/615a8d00f37f3aebb34e43271d7b61f171bd0c56/config.json", "access_date": "2026-11-01"}, {"url": "https://raw.githubusercontent.com/shaochenze/calm/main/train/train_autoencoder.sh", "access_date": "2026-11-01"}, {"url": "https://raw.githubusercontent.com/shaochenze/calm/main/train/train_autoencoder.py", "access_date": "2026-11-01"}, {"url": "https://github.com/shaochenze/calm", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high"}
{"model_id": "i3-lab/i3-80m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes the author's bespoke i3 architecture/tokenizer and completed initial or multi-dataset pretraining, with run steps, losses and hardware.", "provenance_sources": [{"url": "https://huggingface.co/i3-lab/i3-80m/raw/8800d788facbc1d238d04ade756f37c0269f355d/README.md", "accessed_at": "2026-10-30T12:27:36.797572+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "Nakul24/YC_Bot", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned config uses the standard24-layer1024-hidden DialoGPT-medium/GPT2-medium shape, task_specific_params=conversational and local output-small checkpoint path; publisher presents a chatbot. Best-supported inference is adaptation of a pretrained GPT2/DialoGPT checkpoint, exact parent unresolved.", "provenance_sources": [{"url": "https://huggingface.co/Nakul24/YC_Bot/raw/83b9840d924ff151955bd996f2f1097b19c5cdb9/config.json", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/Nakul24/YC_Bot/blob/83b9840d924ff151955bd996f2f1097b19c5cdb9/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01", "provenance_confidence": "medium", "upstream_checkpoint": "GPT2/DialoGPT family; exact checkpoint unconfirmed"}
{"model_id": "bitlabsdb/gpt2-124m-ablation", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact model card explicitly states from-scratch training/building; custom architecture and base-model metadata are additional facts, not the sole basis.", "provenance_sources": [{"url": "https://huggingface.co/bitlabsdb/gpt2-124m-ablation/raw/89d35bbeedf9c806b1b46c9b8442b08da3ae8cd0/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "DienerTech/sparknet-70m", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/DienerTech/sparknet-70m/raw/2721ae5734b219d822d51a2f626b83ed21f2d0a3/README.md", "accessed_at": "2026-10-30T12:27:37.196053+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "vinothkannans/hello-world", "model_provenance": "untrained", "own_from_scratch_lineage": "no", "trained_status": "no", "checkpoint_operation": "none", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly states a static demonstration with one linear layer returning Hello World regardless of input; it requires no training.", "provenance_sources": [{"url": "https://huggingface.co/vinothkannans/hello-world/raw/6d546b0ed1540af98aa5743e8187a563f3d0dc58/README.md", "accessed_at": "2026-10-30T12:27:37.197055+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "facebook/MobileLLM-R1-360M-base", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "midtraining_with_teacher_distillation", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned card and official training repository document random-initialised family pretraining followed by4k-context mid-training with KL knowledge distillation from Llama-3.1-8B-Instruct. Released base checkpoint is a later training-stage release of the publisher own scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/facebook/MobileLLM-R1-360M-base/blob/2bba97cebe4fdc06193407e9ab028d2f45819212/README.md", "access_date": "2026-11-01"}, {"url": "https://github.com/facebookresearch/MobileLLM-R1", "access_date": "2026-11-01", "finding": "Official repository maps the4k base releases to pretraining and mid-training stages; mid-training uses pretrained teacher distillation."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_model": "meta-llama/Llama-3.1-8B-Instruct (teacher)", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "facebook/MobileLLM-R1-140M-base", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "midtraining_with_teacher_distillation", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned card and official training repository document random-initialised family pretraining followed by4k-context mid-training with KL knowledge distillation from Llama-3.1-8B-Instruct. Released base checkpoint is a later training-stage release of the publisher own scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/facebook/MobileLLM-R1-140M-base/blob/27cb0e3b88e117559000732d26d3b08d0339cf7e/README.md", "access_date": "2026-11-01"}, {"url": "https://github.com/facebookresearch/MobileLLM-R1", "access_date": "2026-11-01", "finding": "Official repository maps the4k base releases to pretraining and mid-training stages; mid-training uses pretrained teacher distillation."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_model": "meta-llama/Llama-3.1-8B-Instruct (teacher)", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "bekalebendong/pendo-gpt2-medium-teacher", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/bekalebendong/pendo-gpt2-medium-teacher/raw/134af49cdd38388dfaa75c607667afd0f5db2fba/README.md", "accessed_at": "2026-10-30T12:27:37.485160+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "bekalebendong/pendo-distilgpt2-wikitext", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/bekalebendong/pendo-distilgpt2-wikitext/raw/e363c83c4ec9ffc2e076268f7caa784aa2a87413/README.md", "accessed_at": "2026-10-30T12:27:37.662271+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "vandijklab/pythia-160m-c2s", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/vandijklab/pythia-160m-c2s/raw/9a85d1899246ebe5bd711a31238cebf9bd9ae423/README.md", "accessed_at": "2026-10-30T12:27:37.888407+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "i3-lab/i3-22m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes the author's bespoke i3 architecture/tokenizer and completed initial or multi-dataset pretraining, with run steps, losses and hardware.", "provenance_sources": [{"url": "https://huggingface.co/i3-lab/i3-22m/raw/a91ff3c6953400d7193d564368fac63cc1b6d290/README.md", "accessed_at": "2026-10-30T12:27:37.813711+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "yiwenX/MiniMind-MoE-640-120M", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly describes full pretraining/SFT pipeline and MoE training starting from a pretrained checkpoint; selected conversational release is post-trained. Card credits Jingyao Gong MiniMind upstream project; publisher ownership of earliest checkpoint is unstated.", "provenance_sources": [{"url": "https://huggingface.co/yiwenX/MiniMind-MoE-640-120M/raw/3255d25b45de1a29dd60598bb767fd99cab143ad/README.md", "accessed_at": "2026-10-30T12:27:37.906605+00:00"}, {"url": "https://huggingface.co/yiwenX/MiniMind-MoE-640-120M/raw/3255d25b45de1a29dd60598bb767fd99cab143ad/README.md", "scope": "Complete frozen Chinese card read", "finding": "Explicit MiniMind attribution, pretrained checkpoint/SFT stages,4 GPU DDP", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/yiwenX/MiniMind-MoE-640-120M/raw/3255d25b45de1a29dd60598bb767fd99cab143ad/config.json", "scope": "Complete actual configuration read", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "4 GPUs with DDP, product/VRAM unspecified (publisher training configuration)", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "creator_contribution": "Published MiniMind-based conversational/MoE trained release; actual publisher-versus-upstream earliest pretraining contribution unspecified.", "provenance_confidence": "high", "architecture_lineage": "Jingyao Gong MiniMind code/project expressly credited; MiniMindForCausalLM with4 routed+1 shared experts, top2, width640,8 layers,6400 vocabulary.", "hardware": "4 GPUs with DDP, product/VRAM unspecified (publisher training configuration)", "compute_sources": [{"url": "https://huggingface.co/yiwenX/MiniMind-MoE-640-120M/raw/3255d25b45de1a29dd60598bb767fd99cab143ad/README.md", "scope": "Complete frozen Chinese card read", "finding": "Explicit MiniMind attribution, pretrained checkpoint/SFT stages,4 GPU DDP", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/yiwenX/MiniMind-MoE-640-120M/raw/3255d25b45de1a29dd60598bb767fd99cab143ad/config.json", "scope": "Complete actual configuration read", "access_date": "2026-11-01"}], "evidence_updated_at": "2026-11-01"}
{"model_id": "i3-lab/i3-12m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/i3-lab/i3-12m/raw/cd5b4c0a8ab7a6e46d19fbbfa33f906a25eb8bdc/README.md", "accessed_at": "2026-10-30T12:27:38.013248+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "habdine/Prot2Text-Large-v1-1", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "multimodal_training_with_pretrained_components", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned config specifies pretrained ESM2 protein encoder and GPT2 text decoder. Linked official implementation warm_up loads EsmModel.from_pretrained and GPT2LMHeadModel.from_pretrained; this multimodal task model reuses pretrained components.", "provenance_sources": [{"url": "https://huggingface.co/habdine/Prot2Text-Large-v1-1/raw/2701f1753395919c252f41014a6dedf0ddcd52ca/config.json", "access_date": "2026-11-01"}, {"url": "https://raw.githubusercontent.com/hadi-abdine/Prot2Text/master/prot2text_model/Model.py", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "upstream_checkpoint": "facebook/esm2 family and gpt2"}
{"model_id": "habdine/Prot2Text-Base-v1-1", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "multimodal_training_with_pretrained_components", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned config specifies pretrained ESM2 protein encoder and GPT2 text decoder. Linked official implementation warm_up loads EsmModel.from_pretrained and GPT2LMHeadModel.from_pretrained; this multimodal task model reuses pretrained components.", "provenance_sources": [{"url": "https://huggingface.co/habdine/Prot2Text-Base-v1-1/raw/a450806eda77b7776402a2de9888db0259a3eec8/config.json", "access_date": "2026-11-01"}, {"url": "https://raw.githubusercontent.com/hadi-abdine/Prot2Text/master/prot2text_model/Model.py", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "upstream_checkpoint": "facebook/esm2 family and gpt2"}
{"model_id": "habdine/Esm2Text-Base-v1-1", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "multimodal_training_with_pretrained_components", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned config specifies pretrained ESM2 protein encoder and GPT2 text decoder. Linked official implementation warm_up loads EsmModel.from_pretrained and GPT2LMHeadModel.from_pretrained; this multimodal task model reuses pretrained components.", "provenance_sources": [{"url": "https://huggingface.co/habdine/Esm2Text-Base-v1-1/raw/97498de12ff67449dc2f8deb54093e360086effe/config.json", "access_date": "2026-11-01"}, {"url": "https://raw.githubusercontent.com/hadi-abdine/Prot2Text/master/prot2text_model/Model.py", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "upstream_checkpoint": "facebook/esm2 family and gpt2"}
{"model_id": "i3-lab/i3-tiny", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card describes the author's bespoke i3 architecture/tokenizer and completed initial or multi-dataset pretraining, with run steps, losses and hardware.", "provenance_sources": [{"url": "https://huggingface.co/i3-lab/i3-tiny/raw/7fe5689ca2215c59bf0299c30b7215da584a6456/README.md", "accessed_at": "2026-10-30T12:27:38.532185+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "sdobson/tinystories-llama-15m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card documents own tiny Llama2-architecture model trained100,000 iterations on TinyStories using code adapted from Karpathy llama2.c. The small custom training run supports own scratch origin; architecture/code credit remains Meta/Karpathy.", "provenance_sources": [{"url": "https://huggingface.co/sdobson/tinystories-llama-15m/raw/6cc2bab8c4a55cce6f7532317af84a94bb340d5e/README.md", "accessed_at": "2026-10-30T12:27:38.558124+00:00"}, {"url": "https://github.com/samdobson", "basis": "Public self-declared location on HF-linked profile.", "accessed_at": "2026-10-30T12:33:13.065446+00:00"}, {"url": "https://gist.github.com/samdobson", "basis": "Five public gists listed; one nanochat CPU-generation fork, others unrelated; no TinyStories training script.", "accessed_at": "2026-10-30"}, {"url": "https://samdobson.uk/", "basis": "Author blog read; nanochat training is a separate out-of-cohort model.", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/sdobson/tinystories-llama-15m/raw/6cc2bab8c4a55cce6f7532317af84a94bb340d5e/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Actual training hardware not stated; bfloat16 and hyperparameters are supplied.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "Branis333/astro-gpt2-chatbot", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/Branis333/astro-gpt2-chatbot/raw/1892acecf34d134b370cfad68e53a37e2f455225/README.md", "accessed_at": "2026-10-30T12:27:38.654214+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "hdlm-group/hdlm-base-epsilon-0.0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies the authors' hybrid-diffusion research checkpoint, custom conditioning/noising architecture and1M-step OpenWebText training; this supports original diffusion pretraining.", "provenance_sources": [{"url": "https://huggingface.co/hdlm-group/hdlm-base-epsilon-0.0/raw/bfbce1154ac0d508f87ad2ac1952075b61e85cfe/README.md", "accessed_at": "2026-10-30T12:27:38.763033+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "hdlm-group/hdlm-base-gamma-0.01", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies the authors' hybrid-diffusion research checkpoint, custom conditioning/noising architecture and1M-step OpenWebText training; this supports original diffusion pretraining.", "provenance_sources": [{"url": "https://huggingface.co/hdlm-group/hdlm-base-gamma-0.01/raw/2a57330997c2fb25411ee67286d651a3165472cf/README.md", "accessed_at": "2026-10-30T12:27:38.957535+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "hdlm-group/hdlm-base-gamma-0.05", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies the authors' hybrid-diffusion research checkpoint, custom conditioning/noising architecture and1M-step OpenWebText training; this supports original diffusion pretraining.", "provenance_sources": [{"url": "https://huggingface.co/hdlm-group/hdlm-base-gamma-0.05/raw/19a2ef5e265d3b4a45197d33eda35a80d690991d/README.md", "accessed_at": "2026-10-30T12:27:39.166370+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "hdlm-group/hdlm-base-epsilon-0.01", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies the authors' hybrid-diffusion research checkpoint, custom conditioning/noising architecture and1M-step OpenWebText training; this supports original diffusion pretraining.", "provenance_sources": [{"url": "https://huggingface.co/hdlm-group/hdlm-base-epsilon-0.01/raw/e901ed16a3d3d32f4b67f6252b4a248325458187/README.md", "accessed_at": "2026-10-30T12:27:39.169882+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "OpenPeerAI/OpenPeerLLM", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned src/modeling_openpeer.py constructs custom Transformer modules and explicitly random-initialises their weights; pinned train.py constructs OpenPeerLLM(config) before training. The model card reports an actual training result. main.py mentions GPT2 in an unused interface argument, while src/model.py constructs the custom model rather than loading GPT2 weights.", "provenance_sources": [{"url": "https://huggingface.co/OpenPeerAI/OpenPeerLLM/raw/0d2cbfa820e84c23ab8f939938ffed3a539056a0/train.py", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/OpenPeerAI/OpenPeerLLM/raw/0d2cbfa820e84c23ab8f939938ffed3a539056a0/src/modeling_openpeer.py", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/OpenPeerAI/OpenPeerLLM/resolve/0d2cbfa820e84c23ab8f939938ffed3a539056a0/src/model.py", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/OpenPeerAI/OpenPeerLLM/blob/0d2cbfa820e84c23ab8f939938ffed3a539056a0/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high"}
{"model_id": "facebook/MobileLLM-R1-950M-base", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "midtraining_with_teacher_distillation", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned card and official training repository document random-initialised family pretraining followed by4k-context mid-training with KL knowledge distillation from Llama-3.1-8B-Instruct. Released base checkpoint is a later training-stage release of the publisher own scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/facebook/MobileLLM-R1-950M-base/blob/f84ce13a2f4e512911e7086d7739d4600aff8423/README.md", "access_date": "2026-11-01"}, {"url": "https://github.com/facebookresearch/MobileLLM-R1", "access_date": "2026-11-01", "finding": "Official repository maps the4k base releases to pretraining and mid-training stages; mid-training uses pretrained teacher distillation."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_model": "meta-llama/Llama-3.1-8B-Instruct (teacher)", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "shivash/enhanced-hybrid-transformer-fixed-1758805039", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "probably_yes", "checkpoint_operation": "pretraining_or_own_model_repair", "provenance_evidence_level": "unresolved", "provenance_basis": "Public rendered card describes custom163M architecture, exact training architecture and repaired own weight loading producing coherent text; probable own trained lineage, training dataset unstated.", "provenance_sources": [{"url": "https://huggingface.co/shivash/enhanced-hybrid-transformer-fixed-1758805039/raw/b85985dd42d442b2c8672d2200b6dffd398d9686/README.md", "accessed_at": "2026-10-30T12:27:39.402955+00:00"}, {"url": "https://huggingface.co/shivash/enhanced-hybrid-transformer-fixed-1758805039", "accessed_at": "2026-11-01", "read_method": "Chrome_rendered_UI"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "evidence_updated_at": "2026-11-01"}
{"model_id": "shivash/enhanced-hybrid-transformer-768d-trained-thinking", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "instruction_training_from_own_base", "provenance_evidence_level": "unresolved", "provenance_basis": "Public rendered card says own Enhanced Hybrid768D base,20000GSM8K/AQuA samples,6epochs, final losses; companion base explicitly random/untrained. Probable scratch-origin family trained for reasoning.", "provenance_sources": [{"url": "https://huggingface.co/shivash/enhanced-hybrid-transformer-768d-trained-thinking/raw/e8ace07c4ce42945562d8ca0b6e7ae7b602fe53e/README.md", "accessed_at": "2026-10-30T12:27:39.525865+00:00"}, {"url": "https://huggingface.co/shivash/enhanced-hybrid-transformer-768d-trained-thinking", "accessed_at": "2026-11-01", "read_method": "Chrome_rendered_UI"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "evidence_updated_at": "2026-11-01"}
{"model_id": "shivash/enhanced-hybrid-transformer-768d", "model_provenance": "untrained", "own_from_scratch_lineage": "yes", "trained_status": "no", "checkpoint_operation": "random_initialization", "provenance_evidence_level": "unresolved", "provenance_basis": "Public rendered card explicitly says untrained random weights;142425600 parameters, planned Colab training.", "provenance_sources": [{"url": "https://huggingface.co/shivash/enhanced-hybrid-transformer-768d/raw/a3779d4294a97bebc2534ebf7e2b356e8f28bbd4/README.md", "accessed_at": "2026-10-30T12:27:39.750386+00:00"}, {"url": "https://huggingface.co/shivash/enhanced-hybrid-transformer-768d", "accessed_at": "2026-11-01", "read_method": "Chrome_rendered_UI"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "evidence_updated_at": "2026-11-01"}
{"model_id": "shivash/enhanced-hybrid-transformer-416m-universal", "model_provenance": "unresolved", "own_from_scratch_lineage": "unknown", "trained_status": "unknown", "checkpoint_operation": "not_established", "provenance_evidence_level": "unresolved", "provenance_basis": "Card details architecture/inference claims but gives no completed training, data scale or original weights. No model-specific community post resolves trained versus initialised checkpoint.", "provenance_sources": [{"url": "https://huggingface.co/shivash/enhanced-hybrid-transformer-416m-universal/raw/a5748a2acc2f60010a349d8327672fc1713b7f5c/README.md", "accessed_at": "2026-10-30T12:27:39.782955+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Card details architecture/inference claims but gives no completed training, data scale or original weights. No model-specific community post resolves trained versus initialised checkpoint."}
{"model_id": "adityamore2004/Final-gpt-124", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card associates the released Final-GPT checkpoint/FineWeb-Edu10B dataset with the author's GPT2-From-Scratch-124M training repository.", "provenance_sources": [{"url": "https://huggingface.co/adityamore2004/Final-gpt-124/raw/9d9cf81004b952e24f65c0d716451b5f49dab481/README.md", "accessed_at": "2026-10-30T12:27:39.783956+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "gaussalgo/T5-LM-Large-text2sql-spider", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/gaussalgo/T5-LM-Large-text2sql-spider/raw/f0a1f3ec0a40a26a7e94619d62c2992a82bf83bc/README.md", "accessed_at": "2026-10-30T12:27:39.870583+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "LizardModels/Lizard-17m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card credits Martico2432 and states training on evol-codealpaca-v1. Published SmallGPT source implements the small transformer and explicitly initializes its embeddings/linear weights randomly; this supports original pretraining.", "provenance_sources": [{"url": "https://huggingface.co/LizardModels/Lizard-17m/raw/bebae60353e0a40e1c132db249026f2c3eb17163/README.md"}, {"url": "https://huggingface.co/LizardModels/Lizard-17m/raw/main/architecture.py", "access_date": "2026-11-01", "finding": "Read actual SmallGPT source: own 256-dimensional, six-layer transformer; explicit self.apply(_init_weights), normal random initialization of embeddings/linear layers.", "reused": false}, {"url": "https://huggingface.co/LizardModels/Lizard-17m/blob/main/README.md", "finding": "Saved model card training description."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "source_and_card_supported_inference", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "kunjcr2/GatorGPT2", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/kunjcr2/GatorGPT2/raw/a616980abe21749fd9d0e4ab113804699804d43a/README.md", "accessed_at": "2026-10-30T12:27:40.173704+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ThatHungarian/AuroraStories-12M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "probably_yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Small original Aurora series reports changes in parameter counts, token budgets and lower losses across own6M/10M/30M models. Stories12M records100M TinyStories tokens and17h local training. Best-supported reading is publisher-owned scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/ThatHungarian/AuroraStories-12M/raw/24ecd84ce13779fa0014f4e904601130ed3ed371/README.md", "accessed_at": "2026-10-30T12:27:40.412149+00:00"}, {"url": "https://huggingface.co/ThatHungarian/AuroraStories-12M", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_authored_model_family_and_training", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "fromthesky/PLDR-LLM-v51G-106M-test", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original PLDR power-law graph-attention architecture, author-linked training implementation and explicit RefinedWeb autoregressive pretraining: 8B tokens/250k steps. Own scratch-origin research model is the supported inference.", "provenance_sources": [{"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51G-106M-test/raw/bae1a378637c9b354661c54c5de5e4e2e2706f2d/README.md", "accessed_at": "2026-10-30T12:27:40.424261+00:00"}, {"url": "https://huggingface.co/fromthesky/PLDR-LLM-v51G-106M-test/raw/bae1a378637c9b354661c54c5de5e4e2e2706f2d/README.md", "read_scope": "Model description, training data and training procedure sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "ThatHungarian/Aurora-30M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "probably_yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Small original Aurora series reports changes in parameter counts, token budgets and lower losses across own6M/10M/30M models. Stories12M records100M TinyStories tokens and17h local training. Best-supported reading is publisher-owned scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/ThatHungarian/Aurora-30M/raw/f562a14e5b0cbd4a382049219f051c7aa6f3f734/README.md", "accessed_at": "2026-10-30T12:27:40.418152+00:00"}, {"url": "https://huggingface.co/ThatHungarian/Aurora-30M", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_authored_model_family_and_training", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "ThatHungarian/Aurora-10M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "probably_yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Small original Aurora series reports changes in parameter counts, token budgets and lower losses across own6M/10M/30M models. Stories12M records100M TinyStories tokens and17h local training. Best-supported reading is publisher-owned scratch lineage.", "provenance_sources": [{"url": "https://huggingface.co/ThatHungarian/Aurora-10M/raw/e5f97ab07ec014e1c245478309142298aed734b6/README.md", "accessed_at": "2026-10-30T12:27:40.467340+00:00"}, {"url": "https://huggingface.co/ThatHungarian/Aurora-10M", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_authored_model_family_and_training", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "hyrinmansoor/text2frappe-s3-flan-query", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card expressly identifies FLAN-T5-BASE as the Stage3 Text2Frappe query-generator model; task inputs/outputs specialize ERPNext ORM query generation. Best-supported operation is task fine-tuning of pretrained Google FLAN-T5-base.", "provenance_sources": [{"url": "https://huggingface.co/hyrinmansoor/text2frappe-s3-flan-query/raw/74013b298fd922b57cb174314c1d6a8dde298aac/README.md", "accessed_at": "2026-10-30T12:27:40.638375+00:00"}, {"url": "https://huggingface.co/hyrinmansoor/text2frappe-s3-flan-query/raw/74013b298fd922b57cb174314c1d6a8dde298aac/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "moinbach7/asr_en_text_normalization", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card explicitly fine-tunes facebook/bart-base for inverse text normalization. Example loads pavanBuduguppa/asr_inverse_text_normalization, so upstream author and current publisher are kept separate.", "provenance_sources": [{"url": "https://huggingface.co/moinbach7/asr_en_text_normalization/blob/main/README.md"}, {"url": "https://huggingface.co/moinbach7/asr_en_text_normalization/raw/220c6b07aad08e27dabffaa1b05d50e88f4ababe/README.md", "accessed_at": "2026-10-30T12:27:40.797371+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ZombitX64/HanumanGPT", "model_provenance": "unresolved", "own_from_scratch_lineage": "unknown", "trained_status": "unknown", "checkpoint_operation": "not_established", "provenance_evidence_level": "unresolved", "provenance_basis": "Card credits JonusNattapong, shows generic training configuration and datasets but no starting weights or completed pretraining run. Linked authorHanuman page says Demo-Model; trained scratch versus adapted checkpoint remains unestablished.", "provenance_sources": [{"url": "https://huggingface.co/ZombitX64/HanumanGPT/raw/3156372636e2b6ca4496eaa460cbeddcb0ba3ee4/README.md", "accessed_at": "2026-10-30T12:27:41.074898+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Card credits JonusNattapong, shows generic training configuration and datasets but no starting weights or completed pretraining run. Linked authorHanuman page says Demo-Model; trained scratch versus adapted checkpoint remains unestablished."}
{"model_id": "OussamaEL/medical-llm-10m-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Original10.126M GPT2-style base, custom8192-token medical BPE and documented6788-dialogue5epoch training support own scratch lineage. Inference from_pretrained call loads this release rather than supplying a training initializer.", "provenance_sources": [{"url": "https://huggingface.co/OussamaEL/medical-llm-10m-base/raw/146fe5bdcd6eee7475ebda3daa1d1b92719ee3f7/README.md"}, {"url": "https://huggingface.co/OussamaEL/medical-llm-10m-base", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Unspecified CUDA-enabled GPU", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_author_model_description", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "team-suzuki/Qwen3-4B-SFT-TEST2", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/team-suzuki/Qwen3-4B-SFT-TEST2/raw/20b58413982034e3e61b407266fa9f9415d0f5b4/README.md", "accessed_at": "2026-10-30T12:27:41.045463+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "lumees/lumees-177m-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card reports own177M base training over prepared diverse Arrow-shard text, custom10-layer Transformer with RoPE and cl100k tokenizer, and proper weight initialization. Bespoke configuration/general corpus supports own scratch origin; no external checkpoint named.", "provenance_sources": [{"url": "https://huggingface.co/lumees/lumees-177m-base/raw/7d2de36e8b7067c1536922d67fa1bee92033bd00/README.md", "accessed_at": "2026-10-30T12:27:41.112938+00:00"}, {"url": "https://huggingface.co/lumees/lumees-177m-base/raw/7d2de36e8b7067c1536922d67fa1bee92033bd00/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "nagolinc/fungpt", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies an authored tiny6x384 decoder with SentencePiece tokenization, trained generation behavior and a model-specific reproduction training script.", "provenance_sources": [{"url": "https://huggingface.co/nagolinc/fungpt/raw/e68ac4e55b35bec0091900da648c876f29f4f973/README.md", "accessed_at": "2026-10-30T12:27:41.298240+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Valtry/my-gpt-2", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Inherited GPT-2 card, file-upload/automated-conversion history and owner statement about uploaded files/local inference support pretrained-model redistribution. No creator-specific fine-tuning statement occurs in the card or model discussions.", "provenance_sources": [{"url": "https://huggingface.co/Valtry/my-gpt-2/raw/c8798b6cbc1d7d592a856bdd8104705d01c09e53/README.md", "accessed_at": "2026-10-30T12:27:41.416852+00:00"}, {"url": "https://huggingface.co/Valtry/my-gpt-2/raw/c8798b6cbc1d7d592a856bdd8104705d01c09e53/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/Valtry/my-gpt-2/discussions/2", "scope": "Complete owner inference-support request read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/api/models/Valtry/my-gpt-2/commits/c8798b6cbc1d7d592a856bdd8104705d01c09e53", "scope": "Complete frozen upload descriptions read", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Publishing/inference packaging of GPT-2; fine-tuning not documented.", "provenance_confidence": "medium", "architecture_lineage": "External OpenAI GPT-2 pretrained model.", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "async0x42/Llama-3.2-1B-Instruct-exl3_4.5bpw", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "quantization", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "EXL3 quantized conversion of Meta pretrained Llama weights.", "provenance_sources": [{"url": "https://huggingface.co/async0x42/Llama-3.2-1B-Instruct-exl3_4.5bpw/blob/main/README.md"}, {"url": "https://huggingface.co/async0x42/Llama-3.2-1B-Instruct-exl3_4.5bpw/raw/388e5fde738bf64d9e4a71d1dc6673acb8d01cc1/README.md", "accessed_at": "2026-10-30T12:27:41.693621+00:00"}, {"url": "https://huggingface.co/async0x42/Llama-3.2-1B-Instruct-exl3_4.5bpw/blob/main/README.md", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/async0x42/Llama-3.2-1B-Instruct-exl3_4.5bpw/raw/388e5fde738bf64d9e4a71d1dc6673acb8d01cc1/README.md", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/api/models/async0x42/Llama-3.2-1B-Instruct-exl3_4.5bpw/discussions?p=0", "access_date": "2026-10-31", "scope": "complete discussion-list JSON read", "finding": "0 listed public threads", "http_status": 200}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Conversion device and publisher country unstated; no independent from-scratch training by this publisher is claimed.", "compute_source": "Copied Meta model-card H100 cluster statements describe original Meta training, not this publisher conversion device.", "upstream_checkpoint": "Meta Llama-3.2-1B-Instruct pretrained weights", "creator_contribution": "EXL3 4.5 bits-per-weight conversion and redistribution of Meta model weights.", "evidence_updated_at": "2026-10-31"}
{"model_id": "async0x42/Llama-3.2-1B-exl3_4.5bpw", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "quantization", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "EXL3 quantized conversion of Meta pretrained Llama weights.", "provenance_sources": [{"url": "https://huggingface.co/async0x42/Llama-3.2-1B-exl3_4.5bpw/blob/main/README.md"}, {"url": "https://huggingface.co/async0x42/Llama-3.2-1B-exl3_4.5bpw/raw/bb24e3eb2e6f6784a37c29b0f6b313a1356baacb/README.md", "accessed_at": "2026-10-30T12:27:41.693621+00:00"}, {"url": "https://huggingface.co/async0x42/Llama-3.2-1B-exl3_4.5bpw/blob/main/README.md", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/async0x42/Llama-3.2-1B-exl3_4.5bpw/raw/bb24e3eb2e6f6784a37c29b0f6b313a1356baacb/README.md", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/api/models/async0x42/Llama-3.2-1B-exl3_4.5bpw/discussions?p=0", "access_date": "2026-10-31", "scope": "complete discussion-list JSON read", "finding": "0 listed public threads", "http_status": 200}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Conversion device and publisher country unstated; no independent from-scratch training by this publisher is claimed.", "compute_source": "Copied Meta model-card H100 cluster statements describe original Meta training, not this publisher conversion device.", "upstream_checkpoint": "Meta Llama-3.2-1B pretrained weights", "creator_contribution": "EXL3 4.5 bits-per-weight conversion and redistribution of Meta model weights.", "evidence_updated_at": "2026-10-31"}
{"model_id": "Houzeric/text-to-mermaid", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Custom16.81M8-layer128-hidden MermaidGPT model and tokenizer trained for text-to-diagram task; card supplies own train.py/model.py and trained output example. Supports original scratch model; card also links eric-houzelle/mermaidGPT attribution.", "provenance_sources": [{"url": "https://huggingface.co/Houzeric/text-to-mermaid/raw/2a8a19e9b304d7b6e324e5389f66029dd1066854/README.md", "accessed_at": "2026-10-30T12:27:41.743293+00:00"}, {"url": "https://huggingface.co/Houzeric/text-to-mermaid/raw/2a8a19e9b304d7b6e324e5389f66029dd1066854/README.md", "read_scope": "Saved card model/base/training/architecture and attribution sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "nm-testing/Llama-3.1-8B-Instruct-speculator.eagle3-converted", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "conversion", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly converts yuhuili/EAGLE3-LLaMA3.1-Instruct-8B to the speculators library format; Meta Llama3.1-8B-Instruct is its verifier.", "provenance_sources": [{"url": "https://huggingface.co/nm-testing/Llama-3.1-8B-Instruct-speculator.eagle3-converted/raw/e4dd2cd82b6002adff066e54d368eb1e9aa0a845/README.md", "accessed_at": "2026-10-30T12:27:41.806020+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "svraj379/sv-flan-t5-rephrase", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "unknown", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "FLAN-T5 rephrase model naming and frozen247.6M parameter count support a pretrained FLAN-T5 application derivative/fine-tune as the best classification. Exact uploader training recipe is not available.", "provenance_sources": [{"url": "https://huggingface.co/svraj379/sv-flan-t5-rephrase/raw/2e4100f2433618bf7e8f50fcdca41a3769ce1542/README.md", "accessed_at": "2026-10-30T12:27:41.920626+00:00"}, {"url": "https://huggingface.co/svraj379/sv-flan-t5-rephrase", "access_date": "2026-11-01", "finding": "Frozen inventory identifies a247.6M FLAN-T5 rephrase release. Saved card absent; exact community empty; current config returns401.", "reused": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "model_name_size_and_application_supported_inference", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "kristianfischerai12345/fischgpt-sft", "model_provenance": "derived", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "supervised_finetuning_of_own_base", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card identifies supervised fine-tuning of the publisher own GPT-style scratch family on OpenAssistant/oasst1 for19,999steps; validationloss1.7257502.", "provenance_sources": [{"url": "https://huggingface.co/kristianfischerai12345/fischgpt-sft/blob/d233ef69cf45ea19b9dedbe61baac609eb2da9e7/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high"}
{"model_id": "kristianfischerai12345/fischgpt-pretrain-final", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card describes own from-scratch GPT-style pretraining on10B FineWeb-Edu tokens, with85,000 steps and validation loss2.976645.", "provenance_sources": [{"url": "https://huggingface.co/kristianfischerai12345/fischgpt-pretrain-final/blob/3c71d8587480964d7f94ab6bda9a8675c94caa3c/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high"}
{"model_id": "kristianfischerai12345/fischgpt-pretrain-best", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card describes own from-scratch GPT-style pretraining on10B FineWeb-Edu tokens, with80,000 steps and validation loss2.9803455.", "provenance_sources": [{"url": "https://huggingface.co/kristianfischerai12345/fischgpt-pretrain-best/blob/538c7d5da2eab2d390ea3f8fe5dd35ce2598152c/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high"}
{"model_id": "vishesh-t27/Nano-Llama-Base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Custom compact6layer768hidden Llama-style base with110M-token FineWeb training over6hRTX3090 supports own scratch-origin lineage. Card67M/678M parameter typo retained without changing frozen size.", "provenance_sources": [{"url": "https://huggingface.co/vishesh-t27/Nano-Llama-Base/raw/925ac6f8bf1f84369db7391f99cd008a142dd48d/README.md", "accessed_at": "2026-10-30T12:27:42.350195+00:00"}, {"url": "https://huggingface.co/vishesh-t27/Nano-Llama-Base/raw/925ac6f8bf1f84369db7391f99cd008a142dd48d/README.md", "read_scope": "Saved card architecture/base/training/result and upstream attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "jasonacox/jojo-124M", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/jasonacox/jojo-124M/raw/bde826d0c60eccba7fe4ab631094eaed64d8f9b6/README.md", "accessed_at": "2026-10-30T12:27:42.440158+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "pgryko/babyllama-10m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card reports actual training of a custom6-layer/192-width miniature Llama on1M synthetic tokens, with2epochs,29.36seconds and final loss. Original model training is the best-supported origin.", "provenance_sources": [{"url": "https://huggingface.co/pgryko/babyllama-10m/raw/5774ee24a25a5a552235a8f67f2635e3e76dbae4/README.md", "accessed_at": "2026-10-30T12:27:42.543135+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "ayjays132/PHILLNET-1", "model_provenance": "unresolved", "own_from_scratch_lineage": "unknown", "trained_status": "unknown", "checkpoint_operation": "not_established", "provenance_evidence_level": "unresolved", "provenance_basis": "Pinned executable config is GPT2LMHeadModel24layers, inconsistent with the card custom1-layer cognitive system. No initial-weight declaration or model-specific training run resolves whether GPT2 weights were adapted or newly initialised.", "provenance_sources": [{"url": "https://huggingface.co/ayjays132/PHILLNET-1/raw/37ef5eae068edbf81a2d244e6e5c1603b1d69ff7/README.md", "accessed_at": "2026-10-30T12:27:42.668873+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Pinned executable config is GPT2LMHeadModel24layers, inconsistent with the card custom1-layer cognitive system. No initial-weight declaration or model-specific training run resolves whether GPT2 weights were adapted or newly initialised."}
{"model_id": "TitleOS/Seahorse-350m", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Frozen card explicitly identifies OPT-based model fine-tuned on Alpaca-formatted Orca data; upstream Meta OPT license/authors retained.", "provenance_sources": [{"url": "https://huggingface.co/TitleOS/Seahorse-350m/raw/0080ba451eac5fc2fcfdea1ec59c4fc0ec93cb3d/README.md", "accessed_at": "2026-10-30T12:27:43.015528+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Alpaca-style Orca instruction fine-tuning and release publication.", "architecture_lineage": "Meta pretrained OPT350M backbone.", "evidence_updated_at": "2026-11-01"}
{"model_id": "MahiH/dialogpt-chatbot", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card explicitly describes DialoGPT-small fine-tuned on a small chat dataset; config GPT2LMHead12×768. No actual training device documented.", "provenance_sources": [{"url": "https://huggingface.co/MahiH/dialogpt-chatbot/blob/main/README.md"}, {"url": "https://huggingface.co/MahiH/dialogpt-chatbot/raw/0cec8ed742512047a0f796d642e980ee552a94a5/README.md", "accessed_at": "2026-10-30T12:27:42.921981+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sausheong/ssllm_hf", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies the author's custom218M SSLLM with cl100k tokenization and completed20epoch/1.38B-token Cosmopedia2 training; HF export preserves that own original checkpoint.", "provenance_sources": [{"url": "https://huggingface.co/sausheong/ssllm_hf/raw/7c18e03c9a98d604c527b192c06a59cad007d4a9/README.md", "accessed_at": "2026-10-30T12:27:42.983171+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "Rashik24/tinycoder-15M", "model_provenance": "unresolved", "own_from_scratch_lineage": "unknown", "trained_status": "unknown", "checkpoint_operation": "not_established", "provenance_evidence_level": "unresolved", "provenance_basis": "Pinned card has onlytiny-codes dataset metadata and15.2M Llama config. Small Llama shape can be trained anew or adapted from existingTinyStories15M; model community contains conversion only, exact author/model search finds no training-origin statement.", "provenance_sources": [{"url": "https://huggingface.co/Rashik24/tinycoder-15M/raw/b5be23dad92a6cd3f202d226b99fc9559c237719/README.md", "accessed_at": "2026-10-30T12:27:43.071465+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Pinned card has onlytiny-codes dataset metadata and15.2M Llama config. Small Llama shape can be trained anew or adapted from existingTinyStories15M; model community contains conversion only, exact author/model search finds no training-origin statement."}
{"model_id": "irodkin/gpt2-wiki2", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains standard GPT2-small dimensions and released task isWikiText2 adaptation with training_args.bin and standard GPT2 tokenizer. Best-supported origin is continued training of pretrained GPT2 rather than new initialisation; exact initial weights are not explicitly declared.", "provenance_sources": [{"url": "https://huggingface.co/irodkin/gpt2-wiki2/raw/48aedb6a5e8b20d363e0604302e8452aa8e73131/config.json", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/api/models/irodkin/gpt2-wiki2/tree/48aedb6a5e8b20d363e0604302e8452aa8e73131?recursive=false&expand=false", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/irodkin/gpt2-wiki2/blob/48aedb6a5e8b20d363e0604302e8452aa8e73131/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "gpt2", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "benjamintli/duckdb-sqlcoder-0.5B", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/benjamintli/duckdb-sqlcoder-0.5B/raw/667862d12056d769e2043caaa8166ce555c65a40/README.md", "accessed_at": "2026-10-30T12:27:43.294007+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "marcuscedricridia/bananafish-522", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Unsloth4bit Qwen3-0.6B instruction tuning,5000steps on49k ITP partial epoch. Base checkpoint explicit. Training hardware not stated.", "provenance_sources": [{"url": "https://huggingface.co/marcuscedricridia/bananafish-522/blob/main/README.md"}, {"url": "https://huggingface.co/marcuscedricridia/bananafish-522/raw/f3dbf6620d39d348b2f454b8b1c1f63cba86207e/README.md", "accessed_at": "2026-10-30T12:27:43.559515+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ymroddi/model", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/ymroddi/model/raw/3be41ddc01b643460c579c69aaaac1a93e5f3655/README.md", "accessed_at": "2026-10-30T12:27:43.623160+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "marcuscedricridia/bananafish-0517", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Qwen3-0.6B LoRA instruction fine-tune,0.25epoch,rank16,alpha32. A promised Colab notebook does not document actual training hardware.", "provenance_sources": [{"url": "https://huggingface.co/marcuscedricridia/bananafish-0517/blob/main/README.md"}, {"url": "https://huggingface.co/marcuscedricridia/bananafish-0517/raw/bf4bce354bb52b3356088496de3a2a7b50a6b8cc/README.md", "accessed_at": "2026-10-30T12:27:43.643945+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "mreeza/simple-transformer-model", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author describes custom2-layer128-hidden decoder trained on WikiText2 for5epochs and manually instantiated for inference. Best-supported interpretation of authors narrative is own scratch training; stored BertModel12x768 config conflicts with narrative, lowering confidence. No weight inspection performed.", "provenance_sources": [{"url": "https://huggingface.co/mreeza/simple-transformer-model/blob/main/README.md"}, {"url": "https://huggingface.co/mreeza/simple-transformer-model/raw/a69ddb4e27e829d7f8afee2b6d5d95b1017376b8/README.md", "accessed_at": "2026-10-30T12:27:43.711301+00:00"}, {"url": "https://huggingface.co/mreeza/simple-transformer-model", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "low", "provenance_evidence_detail": "inferred_from_author_report_with_metadata_conflict", "origin_assessment_type": "inferred", "origin_assessment_confidence": "low", "evidence_updated_at": "2026-11-01"}
{"model_id": "raghavbali/gpt2_ft_sherlock_holmes", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=gpt2; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/raghavbali/gpt2_ft_sherlock_holmes/raw/b7eeb52935827abd65f29818861943af62d9c020/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01", "provenance_confidence": "medium", "upstream_checkpoint": "gpt2"}
{"model_id": "ahmeterdempmk/Qwen2.5-0.5B-Quantum-Computing-Instruct", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/ahmeterdempmk/Qwen2.5-0.5B-Quantum-Computing-Instruct/raw/5d27cc354c0011777b21ae09b7f0ece3a49f0623/README.md", "accessed_at": "2026-10-30T12:27:43.918948+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "codewithdark/DiffusionLM", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "probably_yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned custom model constructs LLaDAModel(config) and calls init_weights; its architecture is a33M diffusion LM rather than a standard pretrained GPT checkpoint. The dataset declaration namesWikiText103 and linked author training project constructs/trains this model. Best-supported inference is own scratch training.", "provenance_sources": [{"url": "https://huggingface.co/codewithdark/DiffusionLM/raw/3d5e57e2b5dbfae59158d3e14a75e288b1eae95a/modeling_DiffusionLLM.py", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/codewithdark/DiffusionLM/blob/3d5e57e2b5dbfae59158d3e14a75e288b1eae95a/README.md", "access_date": "2026-11-01"}, {"url": "https://github.com/codewithdark-git/DiffusionLM", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "architecture_and_code_lineage": "Author DiffusionLM implementation; GPT2 tokenizer reused separately from weights", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "Ateeqq/keywords-title-generator", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/Ateeqq/keywords-title-generator/raw/3c23a9ee9d68dcbdc0316b4770775a2c40b933d1/README.md", "accessed_at": "2026-10-30T12:27:44.271749+00:00", "quote": " content appeal. Developed by https://exnrt.com - Fine Tuned: T5-Base - Parameters: 223M - Train Dataset Length: 10,000 - Validation Dataset Length: 2000 - Batch Size: 1 - Epochs: 2 - Train Loss: 1.6578 - Validation Loss: 1.8115 ", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Ateeqq/product-description-generator", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/Ateeqq/product-description-generator/raw/be976562d6eedf8a61f457fc0b610ebb335d3b15/README.md", "accessed_at": "2026-10-30T12:27:44.354809+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Data-Selection/PDS-470M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Data-Selection/PDS-470M/raw/899c684d7bdc4dfcd08fec714ca6541896ce9822/README.md", "accessed_at": "2026-10-30T12:27:44.359839+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Data-Selection/PDS-160M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/Data-Selection/PDS-160M/raw/da18471bf820d9f1b12539a3c51fd3510d26244f/README.md", "accessed_at": "2026-10-30T12:27:44.393379+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "abcorrea/smolLM-135M-wiki-ft", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "Explicit fine-tuning of unsloth/smolLM-135M using Unsloth/TRL.", "provenance_sources": [{"url": "https://huggingface.co/abcorrea/smolLM-135M-wiki-ft/raw/a32920ffd53c4045631b9a4662ee9badebe9d47e/README.md", "accessed_at": "2026-10-30T12:27:44.506734+00:00"}, {"url": "https://abcorrea.github.io/", "basis": "Current institutional affiliation on HF-linked GitHub personal site; separate residence not asserted.", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Model training hardware not stated; profile GPU badges are not attributed to this run.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ar08/Chat-Doctor", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card and configuration identify pretrained Unsloth Qwen2 weights.", "provenance_sources": [{"url": "https://huggingface.co/ar08/Chat-Doctor/blob/main/README.md"}, {"url": "https://huggingface.co/ar08/Chat-Doctor/raw/7b299c1eceb1e28bd46d5ab86b6ad22d7caa3602/README.md", "accessed_at": "2026-10-30T12:27:44.789587+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "core-outline/nyx-31M-it", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered model card explicitly describes Nyx-31M-it as fine-tuned for domain-specific tasks,12,000steps,final validationloss0.04. The starting pretrained checkpoint is unnamed.", "provenance_sources": [{"url": "https://huggingface.co/core-outline/nyx-31M-it/blob/6e090acaf7d0df4dd29981ea132d06322c62f766/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "evidence_updated_at": "2026-11-01", "provenance_confidence": "high"}
{"model_id": "Xennon-BD/Doctor-Chad", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=unsloth/Qwen2-0.5b-bnb-4bit; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/Xennon-BD/Doctor-Chad/raw/f5238dd0018c6436ae21fef5ec694e3afd2710a5/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "unsloth/Qwen2-0.5b-bnb-4bit", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "abrar0503/h2ogpt-gm", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "not_established", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "H2O LLM Studio derivative; configuration explicitly names the pretrained OpenLLaMA checkpoint.", "provenance_sources": [{"url": "https://huggingface.co/abrar0503/h2ogpt-gm/blob/main/README.md"}, {"url": "https://huggingface.co/abrar0503/h2ogpt-gm/raw/d235a999bc06547c8aa80c102f622de6a0078b51/README.md", "accessed_at": "2026-10-30T12:27:44.981193+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ubaada/original-transformer", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author's scratch-transformer implementation is trained as the65M2017 architecture on4.5M WMT14 pairs for24epochs, with RTX4070 hardware and final test loss.", "provenance_sources": [{"url": "https://huggingface.co/ubaada/original-transformer/raw/f4ea61511cd9c62ce825115f8d6d905513a1ef62/README.md", "accessed_at": "2026-10-30T12:27:45.024861+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "TheJoeZenOne/qwen-0.5-reasoning-16bit", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/TheJoeZenOne/qwen-0.5-reasoning-16bit/raw/91e0852c649289613cd7494fd0566ffd69d13582/README.md", "accessed_at": "2026-10-30T12:27:45.136463+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "chestnutbox/next-token-gen", "model_provenance": "derived", "own_from_scratch_lineage": "partial", "trained_status": "yes", "checkpoint_operation": "custom_lstm_pretraining_with_pretrained_embeddings", "provenance_evidence_level": "unresolved", "provenance_basis": "Public card explicitly describes custom9.95M embedding/LSTM/output network trained on diverse unlabeled English and states pretrained word embeddings are processed. Original recurrent training with imported pretrained component is the supported mixed lineage.", "provenance_sources": [{"url": "https://huggingface.co/chestnutbox/next-token-gen/raw/38ff0551263c68a4fa6dce135838d88feeadf684/README.md"}, {"url": "https://huggingface.co/chestnutbox/next-token-gen", "accessed_at": "2026-11-01", "read_method": "Chrome_rendered_UI"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "evidence_updated_at": "2026-11-01"}
{"model_id": "codewithdark/latent-recurrent-depth-lm", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/codewithdark/latent-recurrent-depth-lm/raw/9fad8af7f9cf08d8cc6078700f7da405a935d017/README.md", "accessed_at": "2026-10-30T12:27:45.494750+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "nafisehNik/girt-t5-base", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/nafisehNik/girt-t5-base/raw/4166742d19d974839406d44b3a5a874625e0bf2c/README.md", "accessed_at": "2026-10-30T12:27:45.660626+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "jinymusim/dialogmodel", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/jinymusim/dialogmodel/raw/cd1852bbc2aff1e0c04fc1f7c391915776acc98b/README.md", "accessed_at": "2026-10-30T12:27:45.676139+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "DeathReaper0965/flan-t5-samsum-lora-RLAIF-detoxified", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/DeathReaper0965/flan-t5-samsum-lora-RLAIF-detoxified/raw/244a73bf3f1bc5183812ac6b6157a546c646da8f/README.md", "accessed_at": "2026-10-30T12:27:45.697314+00:00", "quote": "with reduced toxicity using RLAIF This model is a two-fold fine-tuned [Flan-T5 model](https://huggingface.co/google/flan-t5-base) firstly on the [SAMSUM](https://huggingface.co/datasets/samsum) dataset followed by further fine-tu", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "M4-ai/tau-0.5B-instruct", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "instruction_fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card names tau-0.5B as its base, 16,000 GPT-4-generated instruction entries, and explicitly describes fine-tuning.", "provenance_sources": [{"url": "https://huggingface.co/M4-ai/tau-0.5B-instruct/raw/93cf071cdae238777a83c6c2f19513e666898622/README.md", "accessed_at": "2026-10-30T12:27:45.764965+00:00"}, {"url": "https://huggingface.co/M4-ai/tau-0.5B-instruct/raw/93cf071cdae238777a83c6c2f19513e666898622/README.md", "access_date": "2026-11-01", "scope": "Saved model-card introduction and training/family sections read", "finding": "Card names tau-0.5B as its base, 16,000 GPT-4-generated instruction entries, and explicitly describes fine-tuning."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "jinymusim/RPGPT", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card identifies GPT2 trained for role-play with rpguild data and added character/context/input/response tokens. Narrow role-playing adaptation of named GPT2 supports pretrained GPT2 fine-tuning; no independent general pretraining described.", "provenance_sources": [{"url": "https://huggingface.co/jinymusim/RPGPT/raw/4e3a1eeae9ecfb3f65d2d54dad7122fa0fe03bfb/README.md", "accessed_at": "2026-10-30T12:27:45.976813+00:00"}, {"url": "https://huggingface.co/jinymusim/RPGPT/raw/4e3a1eeae9ecfb3f65d2d54dad7122fa0fe03bfb/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "Corianas/256m", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact card explicitly describes Cerebras GPT fine-tuning using Dolly tools; this is not merely a retained config path.", "provenance_sources": [{"url": "https://huggingface.co/Corianas/256m/raw/d29d0ff189816ff916054900260401caca0727bb/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "houcine-bdk/chatMachineProto", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Author reports a personal nanoGPT implementation/124M model with initial OpenWebText training on8A100s and subsequent training experiments. Original pretraining is the best development-evidence inference.", "provenance_sources": [{"url": "https://huggingface.co/houcine-bdk/chatMachineProto/raw/89b471d5e09b32f5b2ef17e8affef4c223a875aa/README.md", "accessed_at": "2026-10-30T12:27:46.335922+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "ar08/Llama-bengali", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card and configuration identify pretrained Unsloth Qwen2 weights.", "provenance_sources": [{"url": "https://huggingface.co/ar08/Llama-bengali/blob/main/README.md"}, {"url": "https://huggingface.co/ar08/Llama-bengali/raw/a83a72808f57a5ac07b0af654b9b7344d3762480/README.md", "accessed_at": "2026-10-30T12:27:46.349671+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "shwars/fido-rugpt3-small", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuned_existing_pretrained_model", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'fine_tuned_existing_pretrained_model', 'initialization_model': 'ai-forever/rugpt3small_based_on_gpt2', 'epochs': 1, 'dataset': 'Selected Russian FIDONet echo archives from2013–2015', 'dataset_size_reported': '552MB', 'basis': 'Explicit pinned-card training description; not a from-scratch base release.'}", "provenance_sources": [{"url": "https://huggingface.co/shwars/fido-rugpt3-small/raw/0ea7b3e5a5243c55160e11194fbd01384db7c444/README.md", "basis": "Pinned model card from the frozen cohort.", "accessed_at": "2026-10-30"}, {"url": "https://soshnikov.com/art/fidoci", "basis": "Project page distinguishes English GPT-2-large example; no funding payer supplied.", "accessed_at": "2026-10-30"}, {"url": "https://soshnikov.com/personal/about", "basis": "Historical Microsoft educational collaboration, not this model’s training location.", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"class": "cloud_datacentre_gpu", "gpus": ["NVIDIA V100"], "service": "Yandex DataSphere", "wall_clock_hours_approx": 3, "gpu_count": "Unspecified", "single_gpu_at_a_time": null, "basis": "Card gives V100 compute and about3 hours; it does not state device count. The linked art page’s9h/A100 example concerns a different English GPT-2-large model.", "sources": [{"url": "https://huggingface.co/shwars/fido-rugpt3-small/raw/0ea7b3e5a5243c55160e11194fbd01384db7c444/README.md", "revision": "0ea7b3e5a5243c55160e11194fbd01384db7c444", "retrieved_at": "2026-10-30T12:27:46.505147+00:00", "accessed_at": "2026-10-30", "access": "fetched", "sha256": "79c4afe5686eca0c637c76e1a4288f55a5dc10a1a9e4f24874c76eb8fa631fac", "basis": "Pinned model card from the frozen cohort."}]}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "JetBrains-Research/cmg-codet5-with-history", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Five CMG checkpoints fine-tune CodeT5/CodeReviewer, with RACE based on the team's CodeT5. ASE2023 paper lists Eliseeva/Serbia, Sokolov/Germany, Bogomolov/Cyprus, Golubev/Serbia, Dig/US and Bryksin/Cyprus. Section4.2 reports one TITAN RTX or4T4s, mixed precision,7.6Mexamples~3Btokens for one epoch. These alternatives are not assigned per variant; no separate acknowledgement section found.", "provenance_sources": [{"url": "https://huggingface.co/JetBrains-Research/cmg-codet5-with-history/raw/0d30f91eb08b6d7e867fa662f43b24f5e8389622/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Paper: single TITAN RTX or4NVIDIA T4 GPUs with DDP; not assigned per checkpoint variant.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "FlofloB/40k_continued_pretraining_Qwen2.5-0.5B-Instruct_Unsloth_merged_16bit", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/FlofloB/40k_continued_pretraining_Qwen2.5-0.5B-Instruct_Unsloth_merged_16bit/raw/aaaabb06c2d1365dfcd5d6e52fb9ff189d66cd4d/README.md", "accessed_at": "2026-10-30T12:27:46.573221+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "FlofloB/100k_fineweb_continued_pretraining_Qwen2.5-0.5B-Instruct_Unsloth_merged_16bit", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/FlofloB/100k_fineweb_continued_pretraining_Qwen2.5-0.5B-Instruct_Unsloth_merged_16bit/raw/318daa274d6dc8752af2208e92f6f2f6d0efd5ae/README.md", "accessed_at": "2026-10-30T12:27:46.838065+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "farihdzaky/nekozutrans", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/farihdzaky/nekozutrans/raw/c44c2a165696128301c1e4ced7cf58a12b7dba0c/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/farihdzaky/nekozutrans/raw/c44c2a165696128301c1e4ced7cf58a12b7dba0c/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/farihdzaky/nekozutrans/raw/c44c2a165696128301c1e4ced7cf58a12b7dba0c/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not stated for exact run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "compute_source": "Not stated. Unsloth generic2xfaster template sentence is not actual measured device or runtime evidence.", "creator_contribution": "Unsloth/TRL supervised fine-tune and redistribution of resulting Qwen checkpoint.", "architecture_and_code_lineage": "Alibaba Qwen2.5-0.5B-Instruct upstream weights, via Unsloth4bit model; actual config and card both name unsloth/qwen2.5-0.5b-instruct-bnb-4bit. Underlying model building remains attributed upstream.", "evidence_updated_at": "2026-10-31"}
{"model_id": "lomahony/pythia-410m-helpful-sft", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lomahony/pythia-410m-helpful-sft/raw/64bb978247f688a4919de1cb4f802a5675a29a72/README.md", "accessed_at": "2026-10-30T12:27:46.951164+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "lomahony/pythia-70m-helpful-sft", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lomahony/pythia-70m-helpful-sft/raw/436bbed193d64e3404dc7155db51651a8c69f48b/README.md", "accessed_at": "2026-10-30T12:27:47.166446+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "harouzie/bart-base-qqp-paws", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/harouzie/bart-base-qqp-paws/raw/4f484577b3b77452e6dcb53eb44c71c9d79ace82/README.md", "accessed_at": "2026-10-30T12:27:47.119089+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "JetBrains-Research/cmg-codet5-without-history", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Five CMG checkpoints fine-tune CodeT5/CodeReviewer, with RACE based on the team's CodeT5. ASE2023 paper lists Eliseeva/Serbia, Sokolov/Germany, Bogomolov/Cyprus, Golubev/Serbia, Dig/US and Bryksin/Cyprus. Section4.2 reports one TITAN RTX or4T4s, mixed precision,7.6Mexamples~3Btokens for one epoch. These alternatives are not assigned per variant; no separate acknowledgement section found.", "provenance_sources": [{"url": "https://huggingface.co/JetBrains-Research/cmg-codet5-without-history/raw/d0375928d8fe65596f29dcd7d68239ff99d5f9b3/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Paper: single TITAN RTX or4NVIDIA T4 GPUs with DDP; not assigned per checkpoint variant.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "suayptalha/minGRULM-base", "model_provenance": "untrained", "own_from_scratch_lineage": "yes", "trained_status": "no", "checkpoint_operation": "initialisation_only", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned model card explicitly tells users to train the experimental model before usage and provides an integration/training template rather than a completed run result. Best-supported interpretation is an initialised minGRU integration checkpoint, despite a generic sentence saying trained onTinyStories.", "provenance_sources": [{"url": "https://huggingface.co/suayptalha/minGRULM-base/blob/746dc14cd01f018b62dd37ac83e7a61b49111597/README.md", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/suayptalha/minGRULM-base/raw/746dc14cd01f018b62dd37ac83e7a61b49111597/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "sudy-super/Contrail-200m-64k", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Own compact200M Mistral-style base pretrained10Btokens on authors JetCopper10B with validation perplexity27.88 supports original scratch-origin model, retaining Mistral architecture attribution.", "provenance_sources": [{"url": "https://huggingface.co/sudy-super/Contrail-200m-64k/raw/96fc9048469ab34d9c0dbaf3961f9c2e83397a78/README.md", "accessed_at": "2026-10-30T12:27:47.605126+00:00"}, {"url": "https://huggingface.co/sudy-super/Contrail-200m-64k/raw/96fc9048469ab34d9c0dbaf3961f9c2e83397a78/README.md", "read_scope": "Saved card architecture/base/training/result and upstream attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "rayonlabs/bloomz-560m-databricks-dolly-15k-curated-en-622484de-f726-4a35-9a9c-a580e804a968", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "duplicated_or_derived_instruction_finetuned_checkpoint", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'duplicated_or_derived_instruction_finetuned_checkpoint', 'base_family': 'BigScience BLOOMZ-560M, itself multitask-finetuned BLOOM', 'from_scratch_by_context_person': 'not_established', 'basis': 'Pinned card describes xP3 instruction finetuning. History starts with duplicate; Dolly-specific further training is not documented in the inherited card.'}", "provenance_sources": [{"url": "https://huggingface.co/rayonlabs/bloomz-560m-databricks-dolly-15k-curated-en-622484de-f726-4a35-9a9c-a580e804a968/raw/79960b50a005261dcb61812f4a07e620b8e269df/README.md", "basis": "Pinned card from the exact frozen cohort.", "accessed_at": "2026-10-30T12:27:47.664589+00:00"}, {"url": "https://eprints.soton.ac.uk/501903/1/pda_version_Thesis_Final-1.pdf", "basis": "University of Southampton cover, May2025; full PDF retrieved.", "accessed_at": "2026-10-30T18:28:15.630875+00:00"}, {"url": "https://sg.linkedin.com/in/chris-subia-waud", "basis": "Current professional profile with exact identity/publication match. Direct opening blocked.", "accessed_at": "2026-10-30T18:27:18.286Z"}, {"url": "https://arxiv.org/html/2506.07940v1", "basis": "Gradients platform whitepaper,9June2025; Rayonlabs.ai author affiliation.", "accessed_at": "2026-10-30T18:29:06.500863+00:00"}, {"url": "https://huggingface.co/cwaud", "basis": "Named HF profile directly links wanderingweights.", "accessed_at": "2026-10-30T18:22:15.276805+00:00"}, {"url": "https://github.com/wanderingweights", "basis": "Linked GitHub pins the Gradients platform.", "accessed_at": "2026-10-30T18:22:15.530984+00:00"}, {"url": "https://huggingface.co/rayonlabs/bloomz-560m-databricks-dolly-15k-curated-en-622484de-f726-4a35-9a9c-a580e804a968/commits/main", "basis": "Public repository history,26December2024; no separately attributed personal training run.", "accessed_at": "2026-10-30T18:22:15.143485+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "exact_Rayon_run_unknown", "upstream_card_context": "Card lists1750 steps,3.67B finetuning tokens,1x/1x/1x parallel layout and a64-A10080GB infrastructure pool. This inherited card does not establish Rayon or Christopher’s hardware allocation."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "gFulvio/moralstories-bart-norm.actions-context-consequences_gen", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/gFulvio/moralstories-bart-norm.actions-context-consequences_gen/raw/4acd05d7df8d6a91bcc56ea84e6d069408aa0d5c/README.md", "accessed_at": "2026-10-30T12:27:47.747354+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "SparseLLM/sparsing-law-0.1b-relu", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/SparseLLM/sparsing-law-0.1b-relu/raw/65b15e89bce1f7515b192404ef394cf47e4ca8a7/README.md", "accessed_at": "2026-10-30T12:27:47.837237+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "iimaginary/schorbGPT-medium", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "continued_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'continued_pretraining', 'detail': 'Derived from GPT-2. The author training code loads pretrained GPT-2; its completed 5 December 2024 run precedes the 6 December upload, matches 124,439,808 parameters and model_checkpoint output path, and the uploaded config has GPT-2-small dimensions. Four sampled tensors across the uploaded checkpoint retain GPT-2 weight structure (cosines 0.9974-0.9996).', 'basis': 'Converging author code, run log, config, upload chronology and sampled checkpoint comparison.'}", "provenance_sources": [{"url": "https://huggingface.co/iimaginary/schorbGPT-medium/raw/466613f76c6f4c11915a84c7efd0ce7e39a78266/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:27:47.816200+00:00"}, {"url": "https://www.sschorb.com/CV-Sam-Schorb.pdf", "basis": "author_public_CV", "accessed_at": "2026-10-30T17:28:09Z"}, {"url": "https://www.sschorb.com/", "basis": "author_portfolio_identity_link_to_iimaginary", "accessed_at": "2026-10-30T17:28:09Z"}, {"url": "https://github.com/sam-schorb", "basis": "HF_linked_GitHub_profile", "accessed_at": "2026-10-30"}, {"url": "https://github.com/sam-schorb/schorbGPT-tiny/blob/9c199d0d1f79c3c31c5e4089718cec4fdc907e8b/train.py", "basis": "Training code loads pretrained GPT-2 weights.", "accessed_at": "2026-10-30"}, {"url": "https://github.com/sam-schorb/schorbGPT-tiny/blob/9c199d0d1f79c3c31c5e4089718cec4fdc907e8b/run.sh", "basis": "Launch script specifies eight H100 GPUs.", "accessed_at": "2026-10-30"}, {"url": "https://github.com/sam-schorb/schorbGPT-tiny/blob/9c199d0d1f79c3c31c5e4089718cec4fdc907e8b/training_logs.txt", "basis": "December 5, 2024 log confirms eight processes/GPUs and 124439808 parameters, with training completion.", "accessed_at": "2026-10-30"}, {"url": "https://github.com/sam-schorb/schorbGPT-tiny/blob/9c199d0d1f79c3c31c5e4089718cec4fdc907e8b/method.txt", "basis": "Dataset download command identifies FineWeb-Edu.", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/iimaginary/schorbGPT-medium/raw/466613f76c6f4c11915a84c7efd0ce7e39a78266/config.json", "basis": "Pinned uploaded config read; corrects prior medium/512-token description.", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/api/models/iimaginary/schorbGPT-medium/commits/main", "basis": "Upload 6 December 2024 follows completed author training run.", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/openai-community/gpt2/resolve/607a30d783dfa663caf39e06633721c8d4cfcd7e/model.safetensors", "basis": "GPT-2 baseline for sampled direct weight comparison; revision pinned.", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "documented_related_run", "detail": "8 NVIDIA H100 GPUs in the matching author run of 5 December 2024; direct checkpoint-to-job identifier absent, correspondence supported by config/path/timing/parameters/weights.", "device": "NVIDIA H100", "count": 8, "memory_variant": "not established"}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sapienzanlp/Minerva-350M-base-v1.0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned rendered card explicitly identifies Minerva family as pretrained from scratch and this350M base as trained on70B Italian/English tokens,16,690steps using llm-foundry.", "provenance_sources": [{"url": "https://huggingface.co/sapienzanlp/Minerva-350M-base-v1.0/blob/65e78a1551d0ea16f3f5860e846b8bd6b5e9a59b/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "evidence_updated_at": "2026-11-01", "provenance_confidence": "high"}
{"model_id": "FlofloB/10k_continued_pretraining_Qwen2.5-0.5B-Instruct_Unsloth_merged_16bit", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/FlofloB/10k_continued_pretraining_Qwen2.5-0.5B-Instruct_Unsloth_merged_16bit/raw/1fcf14c5b4d99c8c7d0c2ee6a706ad351fdd7a2e/README.md", "accessed_at": "2026-10-30T12:27:48.260537+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "FlofloB/83k_continued_pretraining_Qwen2.5-0.5B-Instruct_Unsloth_merged_16bit", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/FlofloB/83k_continued_pretraining_Qwen2.5-0.5B-Instruct_Unsloth_merged_16bit/raw/a6996f0be572d0ee0914413fcdaea3d900109cb8/README.md", "accessed_at": "2026-10-30T12:27:48.318095+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "mlabonne/dummy-llama-2", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "derived_debugging_artifact", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'derived_debugging_artifact', 'detail': 'Card explicitly calls this a dummy version of existing Llama/CodeLlama for debugging, and says not to use it for inference.'}", "provenance_sources": [{"url": "https://huggingface.co/mlabonne/dummy-llama-2/raw/6a40dbe99d20eac9400d2d606654fea8e179be17/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30"}, {"url": "https://mlabonne.github.io/blog/about.html", "basis": "author_personal_biography", "accessed_at": "2026-10-30T17:40:07Z"}, {"url": "https://mlabonne.github.io/blog/publications.html", "basis": "author_publication_list", "accessed_at": "2026-10-30"}, {"url": "https://www.linkedin.com/in/maxime-labonne?skipRedirect=true", "basis": "public_LinkedIn_profile", "accessed_at": "2026-10-30T17:11:14Z"}, {"url": "https://arxiv.org/pdf/2304.01238", "basis": "author_paper_affiliation", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "unknown", "detail": "No actual training hardware established for this cohort artifact."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "nassersala/bloom-small-166", "model_provenance": "unresolved", "own_from_scratch_lineage": "unknown", "trained_status": "unknown", "checkpoint_operation": "not_established", "provenance_evidence_level": "unresolved", "provenance_basis": "Publisher describes a smaller BLOOM functional-testing checkpoint with changed dimensions, but does not say whether weights were randomly initialised, sliced fromBLOOM or trained. Config alone cannot distinguish those operations.", "provenance_sources": [{"url": "https://huggingface.co/nassersala/bloom-small-166/raw/f8cec4d2c8737cbc72ceacf541d2b906dfb645e1/README.md", "accessed_at": "2026-10-30T12:27:48.505211+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Publisher describes a smaller BLOOM functional-testing checkpoint with changed dimensions, but does not say whether weights were randomly initialised, sliced fromBLOOM or trained. Config alone cannot distinguish those operations."}
{"model_id": "lomahony/pythia-160m-helpful-sft", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lomahony/pythia-160m-helpful-sft/raw/21cf72b10bfb6fab017ca9f7719b7d1a8ad26b0a/README.md", "accessed_at": "2026-10-30T12:27:48.469216+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "irodkin/gpt2-wiki103", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=gpt2; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/irodkin/gpt2-wiki103/raw/03e952a713d6bc7f71b8ac7326d2ae5e7c3dc191/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "gpt2", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "JetBrains-Research/cmg-codereviewer-without-history", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Five CMG checkpoints fine-tune CodeT5/CodeReviewer, with RACE based on the team's CodeT5. ASE2023 paper lists Eliseeva/Serbia, Sokolov/Germany, Bogomolov/Cyprus, Golubev/Serbia, Dig/US and Bryksin/Cyprus. Section4.2 reports one TITAN RTX or4T4s, mixed precision,7.6Mexamples~3Btokens for one epoch. These alternatives are not assigned per variant; no separate acknowledgement section found.", "provenance_sources": [{"url": "https://huggingface.co/JetBrains-Research/cmg-codereviewer-without-history/raw/6b38b7e9175cfbd0d08a1383a14df8f0f2d30d5b/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Paper: single TITAN RTX or4NVIDIA T4 GPUs with DDP; not assigned per checkpoint variant.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "mikeendale/Customer-Service", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card header states trained on Cerebras-GPT111M for customer-service abilities. Subsequent Andromeda/Pile pretraining narrative belongs to upstream Cerebras, not this publisher.", "provenance_sources": [{"url": "https://huggingface.co/mikeendale/Customer-Service/raw/a84929b6abef1c21a23eb71114fde084ce7b1874/README.md", "accessed_at": "2026-10-30T12:27:48.958079+00:00"}, {"url": "https://huggingface.co/mikeendale/Customer-Service/raw/a84929b6abef1c21a23eb71114fde084ce7b1874/README.md", "read_scope": "Saved card model/base/training/architecture and attribution sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "Wonder-Griffin/TraXL", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/Wonder-Griffin/TraXL/raw/694bdc28088285fd98350aff40cefbe02847e34d/README.md", "accessed_at": "2026-10-30T12:27:49.025414+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "lightgpt/LightGPT-0.5B-Qwen2", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "LightGPT0.5B-Qwen2 checkpoint identity and LLMLight traffic-signal-control training task support a domain-trained Qwen2 derivative; retain LLMLight research attribution.", "provenance_sources": [{"url": "https://huggingface.co/lightgpt/LightGPT-0.5B-Qwen2/raw/f566851abf69131d49117cda06fc2e4c6ddbb459/README.md", "accessed_at": "2026-10-30T12:27:49.104663+00:00"}, {"url": "https://huggingface.co/lightgpt/LightGPT-0.5B-Qwen2/raw/f566851abf69131d49117cda06fc2e4c6ddbb459/README.md", "read_scope": "Saved card model/base/training/architecture and attribution sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "AI4free/Qwen2-0.5b-tool_call", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/AI4free/Qwen2-0.5b-tool_call/raw/1d4c2adafd33feec5f8557e6c845c3b24adfb4c2/README.md", "accessed_at": "2026-10-30T12:27:49.148451+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Ramikan-BR/Qwen2-0.5B-v24", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/Ramikan-BR/Qwen2-0.5B-v24/raw/31d98fd2f6cc2ccb04fb791bbcad9feb67d6e009/README.md", "accessed_at": "2026-10-30T12:27:49.346658+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "AlekseyElygin/Qwen2-0.5B", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Unsloth Qwen2 pretrained-weight lineage recorded in card and configuration.", "provenance_sources": [{"url": "https://huggingface.co/AlekseyElygin/Qwen2-0.5B/blob/main/README.md"}, {"url": "https://huggingface.co/AlekseyElygin/Qwen2-0.5B/raw/35a182677d13fe076e4d9ad1a5c1c06cd4f0f071/README.md", "accessed_at": "2026-10-30T12:27:49.545435+00:00"}, {"url": "https://huggingface.co/AlekseyElygin/Qwen2-0.5B/blob/main/README.md", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/AlekseyElygin/Qwen2-0.5B/raw/35a182677d13fe076e4d9ad1a5c1c06cd4f0f071/README.md", "access_date": "2026-10-31", "scope": "saved frozen model card / relevant training evidence read; valid earlier source evidence reused"}, {"url": "https://huggingface.co/api/models/AlekseyElygin/Qwen2-0.5B/discussions?p=0", "access_date": "2026-10-31", "scope": "complete discussion-list JSON read", "finding": "0 listed public threads", "http_status": 200}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No explicit dated builder location for this model was established; physical compute region, current profile location and institutional affiliation are separate fields.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Actual training device/duration/provider and physical country unstated.; Builder country and institutional biography absent from exposed creator sources.", "upstream_checkpoint": "unsloth/qwen2-0.5b-bnb-4bit", "creator_contribution": "Fine-tuning a pre-existing Qwen2 checkpoint and publishing the derivative; underlying checkpoint remains upstream work.", "evidence_updated_at": "2026-10-31"}
{"model_id": "Csking/t5-base-AE", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Best supported: task-specific fine-tuning of google-t5/t5-base. Frozen config retains that exact upstream path; SQuAD task card, named AE/QG artifact and actual saved loss/validation event stream establish task training. Literal weight-loading entrypoint absent, so this is a labelled inference rather than a quoted initialization statement.", "provenance_sources": [{"url": "https://huggingface.co/Csking/t5-base-AE/raw/699374f5bd7cae11d8f22e1af4882ea12e901eaa/config.json"}, {"url": "https://huggingface.co/Csking/t5-base-AE/raw/699374f5bd7cae11d8f22e1af4882ea12e901eaa/README.md"}, {"url": "https://huggingface.co/Csking/t5-base-AE/raw/699374f5bd7cae11d8f22e1af4882ea12e901eaa/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Csking/t5-base-AE/raw/699374f5bd7cae11d8f22e1af4882ea12e901eaa/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Csking/t5-base-AE/resolve/699374f5bd7cae11d8f22e1af4882ea12e901eaa/events.out.tfevents.1717262761.node5.944249.0", "access_date": "2026-10-31", "scope": "complete TensorBoard event stream decoded; scalar tags/counts, first/last progress and text/hardware fields inspected; no model code executed", "finding": "1016708 loss records; 11 validation records; no hardware/text tags"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Actual hardware/provider/country not logged; node5 hostname identifies no location.; Exact end-of-training completion marker absent; progress through recorded steps is established.", "creator_contribution": "Task-specific answer extraction (AE) training on declared SQuAD data. Event stream records real training and validation progress.", "best_origin_estimate": {"origin": "fine_tuning_of_external_pretrained_t5", "confidence": "medium", "direct_or_inferred": "inferred", "basis": "Exact upstream _name_or_path plus task-specific SQuAD release and actual loss/validation log.", "remaining_gap": "No literal initialization/training script in frozen files.", "source_urls": ["https://huggingface.co/Csking/t5-base-AE/raw/699374f5bd7cae11d8f22e1af4882ea12e901eaa/README.md", "https://huggingface.co/Csking/t5-base-AE/raw/699374f5bd7cae11d8f22e1af4882ea12e901eaa/config.json", "https://huggingface.co/Csking/t5-base-AE/resolve/699374f5bd7cae11d8f22e1af4882ea12e901eaa/events.out.tfevents.1717262761.node5.944249.0"]}, "provenance_confidence": "medium", "architecture_and_code_lineage": "Standard Google T5ForConditionalGeneration, upstream checkpoint google-t5/t5-base; original architecture/model building attributed to Google.", "training_log_span": {"hours": 32.5836, "start_utc": "2024-06-01T17:26:09.493207+00:00", "end_utc": "2024-06-03T02:01:10.354893+00:00", "basis": "First to last recorded scalar timestamp. Includes validation/pauses and is not asserted exact pure training runtime."}, "training_steps_recorded": 1016708, "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-10-31"}
{"model_id": "Csking/t5-base-QG", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Best supported: task-specific fine-tuning of google-t5/t5-base. Frozen config retains that exact upstream path; SQuAD task card, named AE/QG artifact and actual saved loss/validation event stream establish task training. Literal weight-loading entrypoint absent, so this is a labelled inference rather than a quoted initialization statement.", "provenance_sources": [{"url": "https://huggingface.co/Csking/t5-base-QG/raw/23232ea26161df2a83a7628de7db40ac0f299b68/config.json"}, {"url": "https://huggingface.co/Csking/t5-base-QG/raw/23232ea26161df2a83a7628de7db40ac0f299b68/README.md"}, {"url": "https://huggingface.co/Csking/t5-base-QG/raw/23232ea26161df2a83a7628de7db40ac0f299b68/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Csking/t5-base-QG/raw/23232ea26161df2a83a7628de7db40ac0f299b68/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Csking/t5-base-QG/resolve/23232ea26161df2a83a7628de7db40ac0f299b68/events.out.tfevents.1717263499.node5.981972.0", "access_date": "2026-10-31", "scope": "complete TensorBoard event stream decoded; scalar tags/counts, first/last progress and text/hardware fields inspected; no model code executed", "finding": "980151 loss records; 11 validation records; no hardware/text tags"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Actual hardware/provider/country not logged; node5 hostname identifies no location.; Exact end-of-training completion marker absent; progress through recorded steps is established.", "creator_contribution": "Task-specific question generation training on declared SQuAD data. Event stream records real training and validation progress.", "best_origin_estimate": {"origin": "fine_tuning_of_external_pretrained_t5", "confidence": "medium", "direct_or_inferred": "inferred", "basis": "Exact upstream _name_or_path plus task-specific SQuAD release and actual loss/validation log.", "remaining_gap": "No literal initialization/training script in frozen files.", "source_urls": ["https://huggingface.co/Csking/t5-base-QG/raw/23232ea26161df2a83a7628de7db40ac0f299b68/README.md", "https://huggingface.co/Csking/t5-base-QG/raw/23232ea26161df2a83a7628de7db40ac0f299b68/config.json", "https://huggingface.co/Csking/t5-base-QG/resolve/23232ea26161df2a83a7628de7db40ac0f299b68/events.out.tfevents.1717263499.node5.981972.0"]}, "provenance_confidence": "medium", "architecture_and_code_lineage": "Standard Google T5ForConditionalGeneration, upstream checkpoint google-t5/t5-base; original architecture/model building attributed to Google.", "training_log_span": {"hours": 32.3646, "start_utc": "2024-06-01T17:38:29.270609+00:00", "end_utc": "2024-06-03T02:00:21.922404+00:00", "basis": "First to last recorded scalar timestamp. Includes validation/pauses and is not asserted exact pure training runtime."}, "training_steps_recorded": 980151, "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-10-31"}
{"model_id": "Csking/t5-large-QG", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Best supported: task-specific fine-tuning of google-t5/t5-large. Frozen config retains that exact upstream path; SQuAD task card, named AE/QG artifact and actual saved loss/validation event stream establish task training. Literal weight-loading entrypoint absent, so this is a labelled inference rather than a quoted initialization statement.", "provenance_sources": [{"url": "https://huggingface.co/Csking/t5-large-QG/raw/be0b152141722b8b15285d184a72978d87f77fdc/config.json"}, {"url": "https://huggingface.co/Csking/t5-large-QG/raw/be0b152141722b8b15285d184a72978d87f77fdc/README.md"}, {"url": "https://huggingface.co/Csking/t5-large-QG/raw/be0b152141722b8b15285d184a72978d87f77fdc/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Csking/t5-large-QG/raw/be0b152141722b8b15285d184a72978d87f77fdc/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Csking/t5-large-QG/resolve/be0b152141722b8b15285d184a72978d87f77fdc/events.out.tfevents.1717170414.node5.400638.0", "access_date": "2026-10-31", "scope": "complete TensorBoard event stream decoded; scalar tags/counts, first/last progress and text/hardware fields inspected; no model code executed", "finding": "525594 loss records; 6 validation records; no hardware/text tags"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Actual hardware/provider/country not logged; node5 hostname identifies no location.; Exact end-of-training completion marker absent; progress through recorded steps is established.", "creator_contribution": "Task-specific question generation training on declared SQuAD data. Event stream records real training and validation progress.", "best_origin_estimate": {"origin": "fine_tuning_of_external_pretrained_t5", "confidence": "medium", "direct_or_inferred": "inferred", "basis": "Exact upstream _name_or_path plus task-specific SQuAD release and actual loss/validation log.", "remaining_gap": "No literal initialization/training script in frozen files.", "source_urls": ["https://huggingface.co/Csking/t5-large-QG/raw/be0b152141722b8b15285d184a72978d87f77fdc/README.md", "https://huggingface.co/Csking/t5-large-QG/raw/be0b152141722b8b15285d184a72978d87f77fdc/config.json", "https://huggingface.co/Csking/t5-large-QG/resolve/be0b152141722b8b15285d184a72978d87f77fdc/events.out.tfevents.1717170414.node5.400638.0"]}, "provenance_confidence": "medium", "architecture_and_code_lineage": "Standard Google T5ForConditionalGeneration, upstream checkpoint google-t5/t5-large; original architecture/model building attributed to Google.", "training_log_span": {"hours": 28.7766, "start_utc": "2024-05-31T15:47:07.032486+00:00", "end_utc": "2024-06-01T20:33:42.682059+00:00", "basis": "First to last recorded scalar timestamp. Includes validation/pauses and is not asserted exact pure training runtime."}, "training_steps_recorded": 525594, "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-10-31"}
{"model_id": "Csking/t5-large-AE", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Best supported: task-specific fine-tuning of google-t5/t5-large. Frozen config retains that exact upstream path; SQuAD task card, named AE/QG artifact and actual saved loss/validation event stream establish task training. Literal weight-loading entrypoint absent, so this is a labelled inference rather than a quoted initialization statement.", "provenance_sources": [{"url": "https://huggingface.co/Csking/t5-large-AE/raw/171f77f81452f7c0fec8366be1cf4ea0ce8ddb25/config.json"}, {"url": "https://huggingface.co/Csking/t5-large-AE/raw/171f77f81452f7c0fec8366be1cf4ea0ce8ddb25/README.md"}, {"url": "https://huggingface.co/Csking/t5-large-AE/raw/171f77f81452f7c0fec8366be1cf4ea0ce8ddb25/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Csking/t5-large-AE/raw/171f77f81452f7c0fec8366be1cf4ea0ce8ddb25/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/Csking/t5-large-AE/resolve/171f77f81452f7c0fec8366be1cf4ea0ce8ddb25/events.out.tfevents.1717153554.node5.3737934.0", "access_date": "2026-10-31", "scope": "complete TensorBoard event stream decoded; scalar tags/counts, first/last progress and text/hardware fields inspected; no model code executed", "finding": "525594 loss records; 6 validation records; no hardware/text tags"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Actual hardware/provider/country not logged; node5 hostname identifies no location.; Exact end-of-training completion marker absent; progress through recorded steps is established.", "creator_contribution": "Task-specific answer extraction (AE) training on declared SQuAD data. Event stream records real training and validation progress.", "best_origin_estimate": {"origin": "fine_tuning_of_external_pretrained_t5", "confidence": "medium", "direct_or_inferred": "inferred", "basis": "Exact upstream _name_or_path plus task-specific SQuAD release and actual loss/validation log.", "remaining_gap": "No literal initialization/training script in frozen files.", "source_urls": ["https://huggingface.co/Csking/t5-large-AE/raw/171f77f81452f7c0fec8366be1cf4ea0ce8ddb25/README.md", "https://huggingface.co/Csking/t5-large-AE/raw/171f77f81452f7c0fec8366be1cf4ea0ce8ddb25/config.json", "https://huggingface.co/Csking/t5-large-AE/resolve/171f77f81452f7c0fec8366be1cf4ea0ce8ddb25/events.out.tfevents.1717153554.node5.3737934.0"]}, "provenance_confidence": "medium", "architecture_and_code_lineage": "Standard Google T5ForConditionalGeneration, upstream checkpoint google-t5/t5-large; original architecture/model building attributed to Google.", "training_log_span": {"hours": 38.1414, "start_utc": "2024-05-31T11:06:04.631749+00:00", "end_utc": "2024-06-02T01:14:33.697441+00:00", "basis": "First to last recorded scalar timestamp. Includes validation/pauses and is not asserted exact pure training runtime."}, "training_steps_recorded": 525594, "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-10-31"}
{"model_id": "unionai/union-pythia-70m-deduped", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/unionai/union-pythia-70m-deduped/raw/f1341c8cb809a033899f4794cf3e4c72290045d1/README.md", "accessed_at": "2026-10-30T12:27:50.029283+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Locutusque/TinyMistral-248M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Owner directly answers sheared/pruned versus scratch question: trained from scratch with Mistral architecture; own PyTorch loop, single TitanV and5-6effective days.", "provenance_sources": [{"url": "https://huggingface.co/Locutusque/TinyMistral-248M/raw/e84fb52945d42de4f2361b6047decdc65426dc2b/README.md"}, {"url": "https://huggingface.co/Locutusque/TinyMistral-248M/discussions/1", "accessed_at": "2026-11-01", "read_method": "Chrome_rendered_UI"}, {"url": "https://huggingface.co/Locutusque/TinyMistral-248M", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "single NVIDIA TitanV", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "evidence_updated_at": "2026-11-01"}
{"model_id": "juierror/flan-t5-text2sql-with-schema-v2", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Flan-T5 model identity, text-to-SQL task, named CoSQL/Spider/SParC training datasets and same-author predecessor support Flan-T5 task fine-tuning.", "provenance_sources": [{"url": "https://huggingface.co/juierror/flan-t5-text2sql-with-schema-v2/raw/f59c3923a9df8adf88d6fde90ba46607d04e8817/README.md", "accessed_at": "2026-10-30T12:27:50.222736+00:00"}, {"url": "https://huggingface.co/juierror/flan-t5-text2sql-with-schema-v2/raw/f59c3923a9df8adf88d6fde90ba46607d04e8817/README.md", "read_scope": "Saved card model/base/training/architecture and attribution sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "monsoon-nlp/gpt-nyc-small", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/monsoon-nlp/gpt-nyc-small/raw/61cb384cad3d29c180ba0e57d5f0c6bc1327ec77/README.md", "accessed_at": "2026-10-30T12:27:50.303719+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pile-t5-large", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved original EleutherAI Pile-T5 card describes2,000,000 steps/about2T tokens of denoising pretraining on deduplicated Pile with T5x, own checkpoint suite and losses. This supports original pretraining of T5 architecture rather than adaptation of a named existing T5 checkpoint.", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pile-t5-large/raw/bd11f5ec6311bc866bc31213941972cec09ec95f/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:27:50.372290+00:00"}, {"url": "https://huggingface.co/EleutherAI/pile-t5-large/raw/bd11f5ec6311bc866bc31213941972cec09ec95f/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "jesseD/homer-bot", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/jesseD/homer-bot/raw/32d40acd06333ede9a689150c26b34953df853c9/README.md", "accessed_at": "2026-10-30T12:27:50.426622+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "M4-ai/tau-0.5B", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "continued_pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly identifies Qwen1.5-0.5B base and further pretraining on UltraTextbooks2.0.", "provenance_sources": [{"url": "https://huggingface.co/M4-ai/tau-0.5B/raw/5d86478db762a26c603eb220da918643c860f54f/README.md", "accessed_at": "2026-10-30T12:27:50.650056+00:00"}, {"url": "https://huggingface.co/M4-ai/tau-0.5B/raw/5d86478db762a26c603eb220da918643c860f54f/README.md", "read_scope": "Saved card model/base/training/architecture and attribution sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "recoilme/insomnia_v2", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card explicitly says trained from GPT2 small on synthetic image prompts generated with Mistral7B. GPT2 is the pretrained backbone; Mistral supplies synthetic data, not merged weights.", "provenance_sources": [{"url": "https://huggingface.co/recoilme/insomnia_v2/raw/657494bdfe3dcd32ad62a5f16d8921f7c2226ef6/README.md", "accessed_at": "2026-10-30T12:27:50.835361+00:00"}, {"url": "https://huggingface.co/recoilme/insomnia_v2/raw/657494bdfe3dcd32ad62a5f16d8921f7c2226ef6/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "com3dian/Bart-large-paper2slides-expander", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/com3dian/Bart-large-paper2slides-expander/raw/2e253df0e88b9bb6a0f67976ae35ce9eb5d0a937/README.md", "accessed_at": "2026-10-30T12:27:50.849921+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "akoksal/LongForm-OPT-125M", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "Instruction-tuned OPT-125M using LongForm dataset.", "provenance_sources": [{"url": "https://huggingface.co/akoksal/LongForm-OPT-125M/raw/e0a724bdc182ba5ce4ac07f97d1556fe8720c549/README.md", "accessed_at": "2026-10-30T12:27:50.957291+00:00"}, {"url": "https://akoksal.com/", "basis": "Current self-description on HF-linked personal website.", "accessed_at": "2026-10-30"}, {"url": "https://github.com/akoksal", "basis": "Public self-declared location on HF-linked profile.", "accessed_at": "2026-10-30T12:47:03.836546+00:00"}, {"url": "https://github.com/akoksal/LongForm", "basis": "Project README describes instruction tuning and lists LongForm-OPT-125M.", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2304.08460v3", "basis": "Primary public page read", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "LongForm paper Appendix A describes 2–4 NVIDIA A6000 48GB GPUs for fine-tuning; exact allocation to OPT-125M is not separately specified", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "vilm/Quyen-SE-v0.1", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuned_existing_pretrained_model", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'fine_tuned_existing_pretrained_model', 'initialization_model': 'Qwen1.5-0.5B', 'training_stages': ['supervised fine-tuning', 'direct preference optimisation'], 'datasets': ['OpenHermes-2.5', 'Capybara', 'argilla/distilabel-capybara-dpo-7k-binarized', 'Intel/orca_dpo_pairs', 'private data by Ontocord and BEE-spoke-data'], 'basis': 'Pinned card explicitly describes Qwen1.5 family adaptation; creator public post corroborates exact0.5B base.'}", "provenance_sources": [{"url": "https://huggingface.co/vilm/Quyen-SE-v0.1/raw/3218bf42508847fa858f302b10f82b1fa014e30b/README.md", "basis": "Pinned model card from the frozen cohort.", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/qnguyen3", "basis": "Exact Quyen-SE model authorship in public post", "accessed_at": "2026-10-30"}, {"url": "https://dungminhdao.github.io", "basis": "Historical short course inJuly2026; no2024 UK model connection.", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"class": "supported_compute_unspecified", "gpus": [], "gpu_count": "unknown", "wall_clock_hours": "unknown", "single_gpu_at_a_time": null, "basis": "Card names compute supporters but gives no device model, count or duration.", "sources": [{"url": "https://huggingface.co/vilm/Quyen-SE-v0.1/raw/3218bf42508847fa858f302b10f82b1fa014e30b/README.md", "revision": "3218bf42508847fa858f302b10f82b1fa014e30b", "retrieved_at": "2026-10-30T12:27:51.032542+00:00", "accessed_at": "2026-10-30", "access": "fetched", "sha256": "242a16724d2c4b4593f137834312e202b9955c58478ceaeefe9cd78b4afa9267", "basis": "Pinned model card from the frozen cohort."}]}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "p208p2002/llama-traditional-chinese-120M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own tiny bilingual Llama-architecture pretraining with a newly trained Chinese/English BPE and named Wikipedia/Wudao/C4 corpora; 120M card reports 15B tokens. Tiny bespoke scale and own tokenizer/corpus support own scratch origin; Meta architecture inspiration is separate from weights.", "provenance_sources": [{"url": "https://huggingface.co/p208p2002/llama-traditional-chinese-120M/raw/98fcab38cb17f1ea632a14c30deb4fa7dedcd7ca/README.md", "accessed_at": "2026-10-30T12:27:51.058070+00:00"}, {"url": "https://huggingface.co/p208p2002/llama-traditional-chinese-120M/raw/98fcab38cb17f1ea632a14c30deb4fa7dedcd7ca/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "sgarbi/gpt-nq-prompt-generator", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/sgarbi/gpt-nq-prompt-generator/raw/be48a02399fbeb551c9e4f3dcceaafc3eaa481c1/README.md", "accessed_at": "2026-10-30T12:27:51.288190+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "openai-community/openai-gpt", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'from_scratch_pretraining', 'evidence': 'Original GPT pretrained on BooksCorpus; HF community archive is a later distribution.'}", "provenance_sources": [{"url": "https://huggingface.co/openai-community/openai-gpt/blob/1e0d4f3028acbffb47fe933cea64619c5ec1a002/README.md", "basis": "pinned model card", "accessed_at": "2026-10-30"}, {"url": "https://www.linkedin.com/in/tsalimans/details/experience/", "basis": "OpenAI employment and model role", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/openai-community/openai-gpt/raw/1e0d4f3028acbffb47fe933cea64619c5ec1a002/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "OriginalFAQ: one month/eight GPUs; later P600 estimate not primary hardware log", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "shibing624/code-autocomplete-distilgpt2-python", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved code-autocomplete card identifies GPT2/DistilGPT2 model family and gives pytorch_awesome project source code as training data, with upstream OpenAI GPT2 attribution retained. Best-supported operation is code-domain fine-tuning of the named pretrained backbone.", "provenance_sources": [{"url": "https://huggingface.co/shibing624/code-autocomplete-distilgpt2-python/raw/6a8c5e39c213817ad15fa6765da66c27ba91d4f6/README.md", "accessed_at": "2026-10-30T12:27:51.461785+00:00"}, {"url": "https://huggingface.co/shibing624/code-autocomplete-distilgpt2-python/raw/6a8c5e39c213817ad15fa6765da66c27ba91d4f6/README.md", "access_date": "2026-11-01", "scope": "Saved card project description, named pretrained family, Train data and original GPT2 attribution read locally"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "xwjzds/paraphrase_text_generation", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning_with_custom_diffusion_topic_modules", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved config explicitly sets model=google/flan-t5-large within DeTiME wrapper; card describes leveraging encoder-decoder LLM for diffusion-enhanced topic/paraphrase training. Supports FlanT5-derived model with custom task modules; retain Amazon/DeTiME authors attribution.", "provenance_sources": [{"url": "https://huggingface.co/xwjzds/paraphrase_text_generation/raw/da37a2fa7956b4fdec8a66f28b4a8094c21c8831/README.md", "accessed_at": "2026-10-30T12:27:51.615615+00:00"}, {"url": "https://huggingface.co/xwjzds/paraphrase_text_generation/raw/da37a2fa7956b4fdec8a66f28b4a8094c21c8831/README.md", "read_scope": "Saved card architecture/base/training/result and upstream attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}, {"url": "https://huggingface.co/xwjzds/paraphrase_text_generation/raw/main/config.json", "read_scope": "Actual cached config model base field", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "jucasoliveira/tgs-model", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/jucasoliveira/tgs-model/raw/7467d8d3aa2dd2a805fd0bcf2ada90c3c5700882/README.md", "accessed_at": "2026-10-30T12:27:51.638247+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "umuthopeyildirim/fin-rwkv-169M", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved Fin-RWKV card describes finance-specific training on finance-alpaca and says the RWKV model is finely tuned for financial data. Standard RWKV 169M/430M backbone scales and narrow instruction corpus support external pretrained RWKV adaptation; no own general pretraining described.", "provenance_sources": [{"url": "https://huggingface.co/umuthopeyildirim/fin-rwkv-169M/raw/7773bef155ac58f1b3731f56922d442f74c7ca6f/README.md", "accessed_at": "2026-10-30T12:27:51.678132+00:00"}, {"url": "https://huggingface.co/umuthopeyildirim/fin-rwkv-169M/raw/7773bef155ac58f1b3731f56922d442f74c7ca6f/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "umuthopeyildirim/fin-rwkv-430m", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved Fin-RWKV card describes finance-specific training on finance-alpaca and says the RWKV model is finely tuned for financial data. Standard RWKV 169M/430M backbone scales and narrow instruction corpus support external pretrained RWKV adaptation; no own general pretraining described.", "provenance_sources": [{"url": "https://huggingface.co/umuthopeyildirim/fin-rwkv-430m/raw/644f87faa405cb7a68244c41a2762f6784e11750/README.md", "accessed_at": "2026-10-30T12:27:51.863683+00:00"}, {"url": "https://huggingface.co/umuthopeyildirim/fin-rwkv-430m/raw/644f87faa405cb7a68244c41a2762f6784e11750/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "kinshuk-h/flan-t5-tacred-kg-w-context-small", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly says google/flan-t5-small trained over TACRED Knowledge Graph with KGT-5 method and contextual sentences.", "provenance_sources": [{"url": "https://huggingface.co/kinshuk-h/flan-t5-tacred-kg-w-context-small/raw/bb0d1800399a64c0be0152e10f580340da0a56dc/README.md", "accessed_at": "2026-10-30T12:27:52.080318+00:00"}, {"url": "https://huggingface.co/kinshuk-h/flan-t5-tacred-kg-w-context-small/raw/bb0d1800399a64c0be0152e10f580340da0a56dc/README.md", "read_scope": "Saved card model/base/training/architecture and attribution sections", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "EleutherAI/gpt-neo-125m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/gpt-neo-125m/raw/21def0189f5705e2521767faed922f1f15e7d7db/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:27:52.096337+00:00"}, {"url": "https://www.aisi.gov.uk/people/sid-black", "basis": "current_UK_institutional_affiliation_not_residence", "accessed_at": "2026-10-30"}, {"url": "https://www.conjecture.dev/research/we-are-conjecture-a-new-alignment-research-startup", "basis": "2022_company_team_location_not_current_personal_residence", "accessed_at": "2026-10-30"}, {"url": "https://www.linkedin.com/in/connor-j-leahy", "basis": "current_self_reported_profile_location_search_index", "accessed_at": "2026-10-30"}, {"url": "https://github.com/EleutherAI/gpt-neo/blob/master/README.md", "basis": "Primary public page read", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "kiwibasket/hypmic-gentarobot-medium", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=microsoft/DialoGPT-medium; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/kiwibasket/hypmic-gentarobot-medium/raw/ef05ba35bc5db18bd08d2b53149369c3dd687645/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "microsoft/DialoGPT-medium", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "Locutusque/TinyMistral-248M-v2", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "continued_pretraining_of_own_lineage", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "V2 is same creators TinyMistral248M family with detailed500GPUh base-pretraining stages; original family model directly confirmed scratch-trained in creator discussion.", "provenance_sources": [{"url": "https://huggingface.co/Locutusque/TinyMistral-248M-v2/raw/0f57b17cb317bb322c7c1466b669c681f80c058f/README.md"}, {"url": "https://huggingface.co/Locutusque/TinyMistral-248M/discussions/1", "accessed_at": "2026-11-01", "read_method": "Chrome_rendered_UI"}, {"url": "https://huggingface.co/Locutusque/TinyMistral-248M-v2", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "single NVIDIA TitanV; approximately500GPU-hours", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "author_report_and_family_lineage_inference", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "CrabfishAI/NeXGen-based", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "upstream_checkpoint_redistribution", "provenance_evidence_level": "exact_file_identity", "provenance_basis": "All five frozen weight-format LFS SHA-256 hashes and config Git blob exactly match openai-community/gpt2-medium.", "provenance_sources": [{"url": "https://huggingface.co/CrabfishAI/NeXGen-based/raw/f9c76859ff69c417deee135b323da5b247668fd9/README.md"}, {"url": "https://huggingface.co/CrabfishAI/NeXGen-based/raw/f9c76859ff69c417deee135b323da5b247668fd9/README.md", "access_date": "2026-10-31", "scope": "complete card/config read"}, {"url": "https://huggingface.co/CrabfishAI/NeXGen-based/raw/f9c76859ff69c417deee135b323da5b247668fd9/config.json", "access_date": "2026-10-31", "scope": "complete card/config read"}, {"url": "https://huggingface.co/api/models/CrabfishAI/NeXGen-based/tree/f9c76859ff69c417deee135b323da5b247668fd9", "access_date": "2026-10-31", "finding": "Frozen exact weight hashes and file metadata read"}, {"url": "https://huggingface.co/api/models/openai-community/gpt2-medium/tree/main", "access_date": "2026-10-31", "finding": "All five upstream weight hashes and config blob match frozen release"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "compute_source": "No new training operation for this mirrored release. Upstream original-training hardware is not publisher hardware.", "creator_contribution": "Republishing unchanged upstream checkpoint files with NeXGen naming and model-card framing.", "architecture_and_code_lineage": "OpenAI GPT-2 medium, standard GPT2LMHeadModel; no independently originated model established.", "upstream_model": "openai-community/gpt2-medium", "evidence_updated_at": "2026-10-31"}
{"model_id": "CrabfishAI/NeXGen-large", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "upstream_checkpoint_redistribution", "provenance_evidence_level": "exact_file_identity", "provenance_basis": "All five frozen weight-format LFS SHA-256 hashes and config Git blob exactly match openai-community/gpt2-large.", "provenance_sources": [{"url": "https://huggingface.co/CrabfishAI/NeXGen-large/raw/6d788b19d4a1e27f7dd129203322a9e24ed638f2/README.md"}, {"url": "https://huggingface.co/CrabfishAI/NeXGen-large/raw/6d788b19d4a1e27f7dd129203322a9e24ed638f2/README.md", "access_date": "2026-10-31", "scope": "complete card/config read"}, {"url": "https://huggingface.co/CrabfishAI/NeXGen-large/raw/6d788b19d4a1e27f7dd129203322a9e24ed638f2/config.json", "access_date": "2026-10-31", "scope": "complete card/config read"}, {"url": "https://huggingface.co/api/models/CrabfishAI/NeXGen-large/tree/6d788b19d4a1e27f7dd129203322a9e24ed638f2", "access_date": "2026-10-31", "finding": "Frozen exact weight hashes and file metadata read"}, {"url": "https://huggingface.co/api/models/openai-community/gpt2-large/tree/main", "access_date": "2026-10-31", "finding": "All five upstream weight hashes and config blob match frozen release"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "compute_source": "No new training operation for this mirrored release. Upstream original-training hardware is not publisher hardware.", "creator_contribution": "Republishing unchanged upstream checkpoint files with NeXGen naming and model-card framing.", "architecture_and_code_lineage": "OpenAI GPT-2 large, standard GPT2LMHeadModel; no independently originated model established.", "upstream_model": "openai-community/gpt2-large", "evidence_updated_at": "2026-10-31"}
{"model_id": "CrabfishAI/NeXGen-small", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact PEFT LoRA adapter configuration names Sirclavin/NeXGen as base_model_name_or_path.", "provenance_sources": [{"url": "https://huggingface.co/CrabfishAI/NeXGen-small/raw/9283b4f52b88bcb2b98553b62338f8250d0a5a99/config.json"}, {"url": "https://huggingface.co/CrabfishAI/NeXGen-small/raw/9283b4f52b88bcb2b98553b62338f8250d0a5a99/README.md"}, {"url": "https://huggingface.co/CrabfishAI/NeXGen-small/raw/9283b4f52b88bcb2b98553b62338f8250d0a5a99/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/CrabfishAI/NeXGen-small/raw/9283b4f52b88bcb2b98553b62338f8250d0a5a99/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/CrabfishAI/NeXGen-small/raw/9283b4f52b88bcb2b98553b62338f8250d0a5a99/training_params.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "AutoTrain supervised fine-tuning on timdettmers/openassistant-guanaco; saved intended parameters 3 epochs, LoRA r16 alpha32 dropout0.05, use_int4 true. No completion log/device record in frozen tree.", "architecture_and_code_lineage": "GPT-2 upstream model with AutoTrain SFT/PEFT LoRA. Saved training_params.model is farnhua/gpt2-small-self_instruct_human_eval; adapter-style config.base_model_name_or_path is Sirclavin/NeXGen. These name different stages/aliases, not an own scratch origin.", "evidence_updated_at": "2026-10-31"}
{"model_id": "crumb/gpt2023", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "continued_pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card explicitly starts from OpenAI GPT2-124M and trains on2.23B recent web/arXiv/GitHub tokens,79.32h RTX3060. Card calls this finetuning; operation is continued language-model pretraining.", "provenance_sources": [{"url": "https://huggingface.co/crumb/gpt2023/raw/e7025b0a30ae5ffe242f34ef75502d266b09f13e/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "JackFram/llama-160m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "SpecInfer authors developed small160M Llama-style base trained on Wikipedia/C4. Authored research-model architecture and base pretraining support own scratch-origin lineage; retain SpecInfer team attribution.", "provenance_sources": [{"url": "https://huggingface.co/JackFram/llama-160m/raw/aca9b687d1425f863dcf5de9a4c96e3fe36266dd/README.md", "accessed_at": "2026-10-30T12:27:52.889064+00:00"}, {"url": "https://huggingface.co/JackFram/llama-160m/raw/aca9b687d1425f863dcf5de9a4c96e3fe36266dd/README.md", "read_scope": "Cached model/base/architecture/training and attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Locutusque/gpt2-conversational-or-qa", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Two TinyMistral248M cards describe pretraining with single TitanV;v1 uses7,488,000examples,v2~500GPU-hours/OpenWebText1MplusTinyMistral data3epochs. Initial random weights not explicit. Five GPT2 conversational models are derived fine-tunes. Synthetic example addresses in cards are not author biography.", "provenance_sources": [{"url": "https://huggingface.co/Locutusque/gpt2-conversational-or-qa/raw/91345c70b1ceaec7b015427fdc21ef990b59cf85/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Locutusque/gpt2-large-conversational-retrain", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Two TinyMistral248M cards describe pretraining with single TitanV;v1 uses7,488,000examples,v2~500GPU-hours/OpenWebText1MplusTinyMistral data3epochs. Initial random weights not explicit. Five GPT2 conversational models are derived fine-tunes. Synthetic example addresses in cards are not author biography.", "provenance_sources": [{"url": "https://huggingface.co/Locutusque/gpt2-large-conversational-retrain/raw/c210f844f7139a961eb32606cb652cf150a03006/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "MOKSHm/t5-base-grammar-synthesis", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card describes Flan-T5-base JFLEG grammar fine-tuning; config/source example names Sajid030/t5-base-grammar-synthesis. Source operation is upstream fine-tuning; no independent uploader training documented.", "provenance_sources": [{"url": "https://huggingface.co/MOKSHm/t5-base-grammar-synthesis/blob/main/README.md"}, {"url": "https://huggingface.co/MOKSHm/t5-base-grammar-synthesis/raw/f230946d61981650d31807357561db190c5eedaa/README.md", "accessed_at": "2026-10-30T12:27:53.142187+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "auhide/chef-gpt-en", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/auhide/chef-gpt-en/raw/c67a72e11317703625e4f8e277110e4e85b95e35/README.md", "accessed_at": "2026-10-30T12:27:53.425827+00:00", "quote": "he model [HERE](https://chef-gpt.streamlit.app/). Fine-tuned GPT-2 for recipe generation. [This](https://www.kaggle.com/datasets/shuyangli94/food-com-recipes-and-user-interactions/data) is the dataset that it's trained on. Usa", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "SRDdev/ScriptForge-small", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=gpt2; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/SRDdev/ScriptForge-small/raw/81da4bc855be7990ff9b720a5f232b328f8bad26/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "gpt2", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "qiyinmiss/My_GPT2", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Commit history explicitly uploads originalGPT2 weights; no new uploader training run established.", "provenance_sources": [{"url": "https://huggingface.co/api/models/qiyinmiss/My_GPT2/commits/main"}, {"url": "https://huggingface.co/qiyinmiss/My_GPT2/raw/4145e280b85ec619906dfc5a624e17cde8ffbea6/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "neonon/DialoGPT-medium-cloy", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "probably_yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Exact release names identify DialoGPT-medium, and conversational cards specialize it to named Korean dramas. Best-supported reading is a task-specific DialoGPT derivative rather than an original base model.", "provenance_sources": [{"url": "https://huggingface.co/neonon/DialoGPT-medium-cloy/blob/main/README.md"}, {"url": "https://huggingface.co/neonon/DialoGPT-medium-cloy/raw/a6af799d691a006a4498ea966f4cca1c1ac9bfcb/README.md", "accessed_at": "2026-10-30T12:27:53.666952+00:00"}, {"url": "https://huggingface.co/neonon/DialoGPT-medium-cloy", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_model_identity_and_task", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "neonon/DialoGPT-medium-htccc", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "probably_yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Exact release names identify DialoGPT-medium, and conversational cards specialize it to named Korean dramas. Best-supported reading is a task-specific DialoGPT derivative rather than an original base model.", "provenance_sources": [{"url": "https://huggingface.co/neonon/DialoGPT-medium-htccc/blob/main/README.md"}, {"url": "https://huggingface.co/neonon/DialoGPT-medium-htccc/raw/59b361c620f4e1af2ee6a7f1408f01ea4ef49892/README.md", "accessed_at": "2026-10-30T12:27:53.682499+00:00"}, {"url": "https://huggingface.co/neonon/DialoGPT-medium-htccc", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_model_identity_and_task", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "vicgalle/gpt2-alpaca", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/vicgalle/gpt2-alpaca/raw/dd90aae6879accb7ba273e24b7531b18654c7574/README.md", "accessed_at": "2026-10-30T12:27:53.832199+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "andrewkroening/GalaxyFarAway-DialoGPT-Yoda", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/andrewkroening/GalaxyFarAway-DialoGPT-Yoda/raw/add57eb314e7cd1c061b671498d2d9b1c7536b07/README.md", "accessed_at": "2026-10-30T12:27:54.127979+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "lgaalves/gpt1", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lgaalves/gpt1/raw/5ded136e0674b0edf337bbd4b6fdce4a23c20d88/README.md", "accessed_at": "2026-10-30T12:27:54.209325+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "s-nlp/mbart-detox-en-ru", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card explicitly names facebook/mbart-large-50 as the base model trained on parallel ParaDetox/RuDetox corpora.", "provenance_sources": [{"url": "https://huggingface.co/s-nlp/mbart-detox-en-ru/raw/9a6694f29ea0707f15d2f395b76dc4b8f0aa7e19/README.md", "accessed_at": "2026-10-30T12:27:54.261473+00:00"}, {"url": "https://huggingface.co/s-nlp/mbart-detox-en-ru/raw/9a6694f29ea0707f15d2f395b76dc4b8f0aa7e19/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "satkinson/DialoGPT-medium-marvin", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/satkinson/DialoGPT-medium-marvin/raw/9209996a705db7b9f8670e553821dfc124f7be96/README.md", "accessed_at": "2026-10-30T12:27:54.327310+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "HWERI/pythia-70m-deduped-cleansharegpt", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/HWERI/pythia-70m-deduped-cleansharegpt/raw/00a1f9422c42e35c6c16edab748badd359d816b2/README.md", "accessed_at": "2026-10-30T12:27:54.348302+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Locutusque/gpt2-conversational-retrain", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Two TinyMistral248M cards describe pretraining with single TitanV;v1 uses7,488,000examples,v2~500GPU-hours/OpenWebText1MplusTinyMistral data3epochs. Initial random weights not explicit. Five GPT2 conversational models are derived fine-tunes. Synthetic example addresses in cards are not author biography.", "provenance_sources": [{"url": "https://huggingface.co/Locutusque/gpt2-conversational-retrain/raw/42beabea365183b1353d6b67b9b126ecc67a2e75/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Locutusque/gpt2-large-conversational", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Two TinyMistral248M cards describe pretraining with single TitanV;v1 uses7,488,000examples,v2~500GPU-hours/OpenWebText1MplusTinyMistral data3epochs. Initial random weights not explicit. Five GPT2 conversational models are derived fine-tunes. Synthetic example addresses in cards are not author biography.", "provenance_sources": [{"url": "https://huggingface.co/Locutusque/gpt2-large-conversational/raw/3f4f2493cf13d1ce3b1dd18a340a48313a3e98a6/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Mxode/Pythia-70m-Synonym-Sentence-Converter", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/Mxode/Pythia-70m-Synonym-Sentence-Converter/raw/650f58049c5723392d524472028f5edff81b09d7/README.md", "accessed_at": "2026-10-30T12:27:54.824514+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "KoboldAI/fairseq-dense-355M", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "fairseq125M/355M are conversions of upstream Meta models, not KoboldAI pretraining. OPT350M-Erebus is MrSeeker fine-tune of pretrained OPT. HF histories identify Julius ter Pelkwijk/mrseeker87, Henk717, VE-forbryderne and bots. Original dense-paper training hardware is upstream, not conversion hardware.", "provenance_sources": [{"url": "https://huggingface.co/KoboldAI/fairseq-dense-355M/raw/217d89a4823f0fe748311ce0213f2d5ee8f68a40/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "KoboldAI/fairseq-dense-125M", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "fairseq125M/355M are conversions of upstream Meta models, not KoboldAI pretraining. OPT350M-Erebus is MrSeeker fine-tune of pretrained OPT. HF histories identify Julius ter Pelkwijk/mrseeker87, Henk717, VE-forbryderne and bots. Original dense-paper training hardware is upstream, not conversion hardware.", "provenance_sources": [{"url": "https://huggingface.co/KoboldAI/fairseq-dense-125M/raw/7ed81ed9d8d0bead584f5bc6b78889fc6d921fd2/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "blueapple8259/TinyStories-Alpaca", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card explicitly says roneneldan/TinyStories-33M fine-tuned on vicgalle/alpaca-gpt4.", "provenance_sources": [{"url": "https://huggingface.co/blueapple8259/TinyStories-Alpaca/raw/608563bf2e46697f4438971845f181dc828b2ef2/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Corianas/590m", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact card explicitly describes Cerebras GPT fine-tuning using Dolly tools; this is not merely a retained config path.", "provenance_sources": [{"url": "https://huggingface.co/Corianas/590m/raw/b40088f720410f9623c5609e6e8dadfd971440d8/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Corianas/Quokka_256m", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact card explicitly describes Cerebras GPT fine-tuning using Dolly tools; this is not merely a retained config path.", "provenance_sources": [{"url": "https://huggingface.co/Corianas/Quokka_256m/raw/c760e6593d968420b2ad5791c40a6b5c6c74ebb9/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Corianas/Quokka_590m", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact card explicitly describes Cerebras GPT fine-tuning using Dolly tools; this is not merely a retained config path.", "provenance_sources": [{"url": "https://huggingface.co/Corianas/Quokka_590m/raw/d6ad3685f9821d9c46c03d8d20c2ecf365ec524d/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "lgaalves/gpt2_platypus-dolly-guanaco", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lgaalves/gpt2_platypus-dolly-guanaco/raw/6aeecc64a5f3446c5f8205c0bd9843cd018202f1/README.md", "accessed_at": "2026-10-30T12:27:55.465257+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "lgaalves/gpt2_guanaco-dolly-platypus", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lgaalves/gpt2_guanaco-dolly-platypus/raw/1cc62c3849a5208fb24df8542121e4282727a99a/README.md", "accessed_at": "2026-10-30T12:27:55.584395+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "lgaalves/gpt2_open-platypus", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lgaalves/gpt2_open-platypus/raw/1e88bc616b60798667f8e914c4b3aa572c58bb8d/README.md", "accessed_at": "2026-10-30T12:27:55.691040+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "lgaalves/gpt2-dolly", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lgaalves/gpt2-dolly/raw/268164679a9ba2065fddbaa8fb594583def16ba1/README.md", "accessed_at": "2026-10-30T12:27:55.793477+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "p208p2002/llama-chinese-81M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own tiny bilingual Llama-architecture pretraining with a newly trained Chinese/English BPE and named Wikipedia/Wudao/C4 corpora; 120M card reports 15B tokens. Tiny bespoke scale and own tokenizer/corpus support own scratch origin; Meta architecture inspiration is separate from weights.", "provenance_sources": [{"url": "https://huggingface.co/p208p2002/llama-chinese-81M/raw/db7e31ab28c46a64ab840e461aabee9655bb0329/README.md", "accessed_at": "2026-10-30T12:27:56.046736+00:00"}, {"url": "https://huggingface.co/p208p2002/llama-chinese-81M/raw/db7e31ab28c46a64ab840e461aabee9655bb0329/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "unionai/Llama-2-7b-hf-wikipedia", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/unionai/Llama-2-7b-hf-wikipedia/raw/a1f67e5af741e44267a45a8f6fde1bd36e0d1917/README.md", "accessed_at": "2026-10-30T12:27:56.112119+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "unionai/pythia-70m-deduped-alpaca-cleaned", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/unionai/pythia-70m-deduped-alpaca-cleaned/raw/80ec2cef1f50c9c5a3350a98d8900e6a180cddf1/README.md", "accessed_at": "2026-10-30T12:27:56.120635+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "isek-ai/SDPrompt-RetNet-300M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/isek-ai/SDPrompt-RetNet-300M/raw/0fbda9439c8a4d35b45f6e068d9c03276fcc53d9/README.md", "accessed_at": "2026-10-30T12:27:56.264337+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "nelson2424/distilroberta-base-finetuned-cot", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "distilroberta-base-finetuned identity plus detailed publisher masked-token chess-opening training experiments support DistilRoBERTa task fine-tuning.", "provenance_sources": [{"url": "https://huggingface.co/nelson2424/distilroberta-base-finetuned-cot/raw/5b5d54b431eadeb22a96e899ced559c9cfc86a56/README.md", "accessed_at": "2026-10-30T12:27:56.351504+00:00"}, {"url": "https://huggingface.co/nelson2424/distilroberta-base-finetuned-cot/raw/5b5d54b431eadeb22a96e899ced559c9cfc86a56/README.md", "read_scope": "Saved card model/task/training and source attribution", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "JackFram/llama-160m-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "SpecInfer authors developed small160M Llama-style base trained on Wikipedia/C4. Authored research-model architecture and base pretraining support own scratch-origin lineage; retain SpecInfer team attribution.", "provenance_sources": [{"url": "https://huggingface.co/JackFram/llama-160m-base/raw/957d08a6fb73abb718dbf7d8ec573bc91728a316/README.md", "accessed_at": "2026-10-30T12:27:56.555720+00:00"}, {"url": "https://huggingface.co/JackFram/llama-160m-base/raw/957d08a6fb73abb718dbf7d8ec573bc91728a316/README.md", "read_scope": "Cached model/base/architecture/training and attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "kargaranamir/T5R-base", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card explicitly states instruction-tuned version of google/flan-t5-base on Alpaca in reversed instruction/output format.", "provenance_sources": [{"url": "https://huggingface.co/kargaranamir/T5R-base/raw/4269d7e1714198c5c01e7ae1dbcd8db1a4732730/README.md", "accessed_at": "2026-10-30T12:27:56.741255+00:00"}, {"url": "https://huggingface.co/kargaranamir/T5R-base/raw/4269d7e1714198c5c01e7ae1dbcd8db1a4732730/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "timinar/baby-llama-58m", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/timinar/baby-llama-58m/raw/f697871b1a476875a9fe469eb04bfae2ad3a4d92/README.md", "accessed_at": "2026-10-30T12:27:56.938987+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Sajid030/t5-base-grammar-synthesis", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/Sajid030/t5-base-grammar-synthesis/raw/38f6dbba2183fc2c57a0a829b0cc026925f6c1f5/README.md", "accessed_at": "2026-10-30T12:27:56.823007+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "kaitchup/OPT-350M-RM-DSChat", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/kaitchup/OPT-350M-RM-DSChat/raw/4a9c8558c4263685fb92ad27c99be3c36867c4f4/README.md", "accessed_at": "2026-10-30T12:27:56.949504+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "zaemyung/DElIteraTeR-PEGASUS-Multi-Sent-Revision-Generator", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/zaemyung/DElIteraTeR-PEGASUS-Multi-Sent-Revision-Generator/raw/3da459958ec3c4217cdc8e4737618d27534f8d0c/README.md", "accessed_at": "2026-10-30T12:27:57.106515+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "cmarkea/bloomz-560m-sft-chat", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/cmarkea/bloomz-560m-sft-chat/raw/47ef43cbaa682d7a455c521fc8f06a7bd43fcb8a/README.md", "accessed_at": "2026-10-30T12:27:57.354608+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "talalH/summarizer_on_T5_base", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/talalH/summarizer_on_T5_base/raw/33afe35762a561aa4dd77d4262c6aaf5f4b7c9f4/README.md", "accessed_at": "2026-10-30T12:27:57.450304+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Intel/fid_flan_t5_base_nq", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card explicitly names pretrained google/flan-t5-base or t5-large as the base checkpoint for Fusion-In-Decoder training on Natural Questions with retrieved Wikipedia passages.", "provenance_sources": [{"url": "https://huggingface.co/Intel/fid_flan_t5_base_nq/raw/f2825687c315acc8e64507b8c829f7e295135563/README.md", "accessed_at": "2026-10-30T12:27:57.550593+00:00"}, {"url": "https://huggingface.co/Intel/fid_flan_t5_base_nq/raw/f2825687c315acc8e64507b8c829f7e295135563/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "mlabonne/dummy-CodeLlama-7b-hf", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "derived_debugging_artifact", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'derived_debugging_artifact', 'detail': 'Card explicitly calls this a dummy version of existing Llama/CodeLlama for debugging, and says not to use it for inference.'}", "provenance_sources": [{"url": "https://huggingface.co/mlabonne/dummy-CodeLlama-7b-hf/raw/bd3988d6c7174f10974b0982042e63758c39cf33/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30"}, {"url": "https://mlabonne.github.io/blog/about.html", "basis": "author_personal_biography", "accessed_at": "2026-10-30T17:40:07Z"}, {"url": "https://mlabonne.github.io/blog/publications.html", "basis": "author_publication_list", "accessed_at": "2026-10-30"}, {"url": "https://www.linkedin.com/in/maxime-labonne?skipRedirect=true", "basis": "public_LinkedIn_profile", "accessed_at": "2026-10-30T17:11:14Z"}, {"url": "https://arxiv.org/pdf/2304.01238", "basis": "author_paper_affiliation", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "unknown", "detail": "No actual training hardware established for this cohort artifact."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "andreipb/gpt2-poetry-model-crpo", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/andreipb/gpt2-poetry-model-crpo/raw/f7ba9c16aead1f0a36571fa89bad5cc00d7c97f6/README.md", "accessed_at": "2026-10-30T12:27:57.643156+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "shuvom/pythia-70m-FT-Lamini-420", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved release name explicitly says pythia-70m-FT-Lamini and card metadata specifies lamini/lamini_docs. Best-supported operation is fine-tuning pretrained EleutherAI Pythia70M on Lamini documentation.", "provenance_sources": [{"url": "https://huggingface.co/shuvom/pythia-70m-FT-Lamini-420/raw/4572f071830f73cc3287904c76dbbf0a42966c32/README.md", "accessed_at": "2026-10-30T12:27:57.836602+00:00"}, {"url": "https://huggingface.co/shuvom/pythia-70m-FT-Lamini-420/raw/4572f071830f73cc3287904c76dbbf0a42966c32/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "crumb/core1-base-464m-c4", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "continued_pretraining", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card retains first decoder layer from Llama2-7B then trains on1B C4 tokens; truncated pretrained weights, not random origin.", "provenance_sources": [{"url": "https://huggingface.co/crumb/core1-base-464m-c4/raw/b5a316c8e15bdf79e2bd1c6a906313f2b7028d85/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Intel/fid_t5_large_nq", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card explicitly names pretrained google/flan-t5-base or t5-large as the base checkpoint for Fusion-In-Decoder training on Natural Questions with retrieved Wikipedia passages.", "provenance_sources": [{"url": "https://huggingface.co/Intel/fid_t5_large_nq/raw/565f5fe522a417c5a2ca0c4953a209ba243c31a4/README.md", "accessed_at": "2026-10-30T12:27:58.188273+00:00"}, {"url": "https://huggingface.co/Intel/fid_t5_large_nq/raw/565f5fe522a417c5a2ca0c4953a209ba243c31a4/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "JetBrains-Research/cmg-race-with-history", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Five CMG checkpoints fine-tune CodeT5/CodeReviewer, with RACE based on the team's CodeT5. ASE2023 paper lists Eliseeva/Serbia, Sokolov/Germany, Bogomolov/Cyprus, Golubev/Serbia, Dig/US and Bryksin/Cyprus. Section4.2 reports one TITAN RTX or4T4s, mixed precision,7.6Mexamples~3Btokens for one epoch. These alternatives are not assigned per variant; no separate acknowledgement section found.", "provenance_sources": [{"url": "https://huggingface.co/JetBrains-Research/cmg-race-with-history/raw/120dea8f7554a322c6ad0f83b76cca7b86806189/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Paper: single TITAN RTX or4NVIDIA T4 GPUs with DDP; not assigned per checkpoint variant.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "monsoon-nlp/gpt-nyc", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/monsoon-nlp/gpt-nyc/raw/13b898a9c95f3abe5f6a4ce394bc58bfb2daaafa/README.md", "accessed_at": "2026-10-30T12:27:58.362799+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "monsoon-nlp/no-phone-gpt2", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "tokenizer_and_embedding_edit", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card explicitly describes modifying existing small OpenAI GPT2 tokenizer/merges and moving numeric-token embeddings to reduce memorized-number associations. Fine-tuning is listed only as a future goal; selected operation is direct token/embedding editing.", "provenance_sources": [{"url": "https://huggingface.co/monsoon-nlp/no-phone-gpt2/raw/d84c528c6816d53a08d2dea8cc3a0644dafe045e/README.md", "accessed_at": "2026-10-30T12:27:58.398563+00:00"}, {"url": "https://huggingface.co/monsoon-nlp/no-phone-gpt2/raw/d84c528c6816d53a08d2dea8cc3a0644dafe045e/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "JetBrains-Research/cmg-codereviewer-with-history", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Five CMG checkpoints fine-tune CodeT5/CodeReviewer, with RACE based on the team's CodeT5. ASE2023 paper lists Eliseeva/Serbia, Sokolov/Germany, Bogomolov/Cyprus, Golubev/Serbia, Dig/US and Bryksin/Cyprus. Section4.2 reports one TITAN RTX or4T4s, mixed precision,7.6Mexamples~3Btokens for one epoch. These alternatives are not assigned per variant; no separate acknowledgement section found.", "provenance_sources": [{"url": "https://huggingface.co/JetBrains-Research/cmg-codereviewer-with-history/raw/8b66f9b3c794461a15c35dedcaf97a61666f893f/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Paper: single TITAN RTX or4NVIDIA T4 GPUs with DDP; not assigned per checkpoint variant.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "RajuKandasamy/tamillama_tiny_30m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved card describes own30M Llama-architecture TinyStories-style bilingual Tamil/English experiment and translated training data. Tiny bespoke model scale and original story-training project support own scratch origin; Meta Llama architecture and TinyStories/IndicTrans data methods remain separately credited.", "provenance_sources": [{"url": "https://huggingface.co/RajuKandasamy/tamillama_tiny_30m/raw/c9a967f4bd136b63505b3d07ee44df13c076be17/README.md"}, {"url": "https://huggingface.co/RajuKandasamy/tamillama_tiny_30m/raw/c9a967f4bd136b63505b3d07ee44df13c076be17/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "Isotonic/bullet-points-generator", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuned_existing_pretrained_model", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'fine_tuned_existing_pretrained_model', 'base_model': 'google-t5/t5-base', 'detail': 'T5-base fine-tune. Config records t5_base_text_to_bp/model. Four sampled tensors retain the corresponding T5-base weights: cosines 0.9951-0.9999998; 88.39% exact equality in 65,536 sampled shared-embedding elements. The generated model-card from-scratch claim conflicts with the checkpoint evidence.'}", "provenance_sources": [{"url": "https://huggingface.co/Isotonic/bullet-points-generator/raw/c8b6bc9d8a83541123b43073272180b8bcb1f8cf/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:27:58.851056+00:00"}, {"url": "https://huggingface.co/api/models/Isotonic/bullet-points-generator", "basis": "collected_api_config_in_data/raw/models.jsonl", "accessed_at": "2026-10-30"}, {"url": "https://sripaad.github.io/src/sripaad_resume_0724.pdf", "basis": "author_public_CV", "accessed_at": "2026-10-30T17:31:53Z"}, {"url": "https://github.com/Sripaad", "basis": "HF_linked_GitHub_profile", "accessed_at": "2026-10-30"}, {"url": "https://github.com/Sripaad/Sripaad/blob/main/README.md", "basis": "HF_linked_GitHub_profile_README", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/Isotonic/bullet-points-generator/raw/c8b6bc9d8a83541123b43073272180b8bcb1f8cf/config.json", "basis": "Pinned T5-base path and architecture read.", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/api/models/Isotonic/bullet-points-generator/commits/main", "basis": "End-of-training commits August 2023; later safetensors conversion.", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/google-t5/t5-base/resolve/a9723ea7f1b39c1eae772870f3b547bf6ef7e6c1/model.safetensors", "basis": "T5-base baseline for sampled direct weight comparison; revision pinned.", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "unknown", "detail": "TorchCUDA version alone does not identify training hardware."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "UncleanCode/anacondia-70m", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/UncleanCode/anacondia-70m/raw/48a71114092821893236b13a950a2df4a18714c3/README.md", "accessed_at": "2026-10-30T12:27:59.021808+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Voicelab/vlt5-base-keywords", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Frozen card identifies t5-base as its language model and describes supervised keyword generation training onPOSMAC; best-supported origin is fine-tuning of pretrained T5-base.", "provenance_sources": [{"url": "https://huggingface.co/Voicelab/vlt5-base-keywords/blob/89a069b6626ad817ceaddf1f08271cd1aadb138a/README.md", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/Voicelab/vlt5-base-keywords/raw/89a069b6626ad817ceaddf1f08271cd1aadb138a/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "t5-base", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "sxx123/gpt2-medium", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/sxx123/gpt2-medium/raw/8e5864c6d20174d5bdc86c15d8b51235403c3539/README.md", "accessed_at": "2026-10-30T12:27:59.103998+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "HexHands/finishABOUTME", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Both releases have124439808 parameters and describe training on only2000 Scratch profile/studio texts. A GPT2-small task fine-tune is the best-supported interpretation, rather than original base pretraining.", "provenance_sources": [{"url": "https://huggingface.co/HexHands/finishABOUTME/raw/6fbb3349a41ae0fcfe5301a44290b4934bbc8766/README.md"}, {"url": "https://huggingface.co/HexHands/finishABOUTME", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_training_scale_and_checkpoint_size", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "lomahony/eleuther-pythia410m-hh-dpo", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lomahony/eleuther-pythia410m-hh-dpo/raw/81330f27400985589f97c34bb8de31cfbac91ce4/README.md", "accessed_at": "2026-10-30T12:27:59.472322+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "lomahony/eleuther-pythia160m-hh-sft", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lomahony/eleuther-pythia160m-hh-sft/raw/6b539cf307b976d632998b0d2df9a62242300a42/README.md", "accessed_at": "2026-10-30T12:27:59.532339+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "lomahony/eleuther-pythia70m-hh-dpo", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lomahony/eleuther-pythia70m-hh-dpo/raw/c26135008b7530f5f1d2be347bb1fec8eb60e877/README.md", "accessed_at": "2026-10-30T12:27:59.723204+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "lomahony/eleuther-pythia70m-hh-sft", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lomahony/eleuther-pythia70m-hh-sft/raw/64018b21b3178d88d326bb6d591303e1a0cf32a3/README.md", "accessed_at": "2026-10-30T12:27:59.792469+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "nicbull/DialoGPT-medium-leric", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "unknown", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "DialoGPT-medium release naming, standard medium GPT-2 architecture and output-medium metadata support a DialoGPT derivative/fine-tune as the best classification.", "provenance_sources": [{"url": "https://huggingface.co/nicbull/DialoGPT-medium-leric/blob/main/README.md"}, {"url": "https://huggingface.co/nicbull/DialoGPT-medium-leric/raw/5424c5ca583eb9b5b6a1b45e22af73c8ea5ba67e/README.md", "accessed_at": "2026-10-30T12:27:59.814995+00:00"}, {"url": "https://huggingface.co/nicbull/DialoGPT-medium-leric/raw/main/config.json", "access_date": "2026-11-01", "finding": "Read GPT2LMHeadModel config: standard medium 24 layers/1024 width/16 heads/50257 vocab, _name_or_path output-medium, conversational generation settings.", "reused": false}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "release_name_and_configuration_supported_inference", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "nicbull/DialoGPT-medium-nic2", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "unknown", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "DialoGPT-medium release naming, standard medium GPT-2 architecture and output-medium metadata support a DialoGPT derivative/fine-tune as the best classification.", "provenance_sources": [{"url": "https://huggingface.co/nicbull/DialoGPT-medium-nic2/blob/main/README.md"}, {"url": "https://huggingface.co/nicbull/DialoGPT-medium-nic2/raw/b43e0f7391d0362fdba4132da09e97ba5f0adcfc/README.md", "accessed_at": "2026-10-30T12:27:59.992463+00:00"}, {"url": "https://huggingface.co/nicbull/DialoGPT-medium-nic2/raw/main/config.json", "access_date": "2026-11-01", "finding": "Read GPT2LMHeadModel config: standard medium 24 layers/1024 width/16 heads/50257 vocab, _name_or_path output-medium, conversational generation settings.", "reused": false}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "release_name_and_configuration_supported_inference", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "igorktech/t5-base-lyrics-explainer", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=igorktech/t5-base-en; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/igorktech/t5-base-lyrics-explainer/raw/2ef2b1e0e518c06127d462ee6814910e1f20ec29/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "igorktech/t5-base-en", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "voidful/unifiedqg-bart-base", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly describes a task-specific question/distractor generator based on pretrained BART-base.", "provenance_sources": [{"url": "https://huggingface.co/voidful/unifiedqg-bart-base/raw/f9354146d23a9edc02e320387843113a8437f61c/README.md", "accessed_at": "2026-10-30T12:28:00.296554+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "s3nh/DialoGPT-small-harry-potter-goblet-of-fire", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/s3nh/DialoGPT-small-harry-potter-goblet-of-fire/raw/304d189f074382ab90affa52395ee2c1d7490c69/README.md", "accessed_at": "2026-10-30T12:28:00.426143+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "flax-community/t5-recipe-generation", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Pinned src/run.sh explicitly sets MODEL_NAME_OR_PATH=t5-base and passes it into the recipe training program. This release fine-tunes pretrained T5-base onRecipeNLG.", "provenance_sources": [{"url": "https://huggingface.co/flax-community/t5-recipe-generation/raw/12ba57f5c4924642afdf556a761124638e06af51/src/run.sh", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Google-sponsored TPU; HF Flax/JAX community week", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "high", "upstream_checkpoint": "t5-base"}
{"model_id": "leondz/artgpt2tox", "model_provenance": "unresolved", "own_from_scratch_lineage": "unknown", "trained_status": "unknown", "checkpoint_operation": "not_established", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "Pinned card is metadata-only and config names local artgpt path. Public model discussion is conversion only; no starting checkpoint or task-training description found by exact model search.", "provenance_sources": [{"url": "https://huggingface.co/leondz/artgpt2tox/raw/a880359fe389ed36d717f4f36f5a36d80925ed35/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30"}, {"url": "https://github.com/leondz", "basis": "HF_linked_GitHub_profile", "accessed_at": "2026-10-30"}, {"url": "https://etheses.whiterose.ac.uk/id/eprint/4068/", "basis": "university_thesis_repository", "accessed_at": "2026-10-30T17:52:19Z"}, {"url": "https://orcid.org/0000-0002-8656-3431", "basis": "self_asserted_ORCID_employment", "accessed_at": "2026-10-30T17:11:07Z"}, {"url": "https://www.derczynski.com/sheffield/papers/report.pdf", "basis": "author_undergraduate_dissertation", "accessed_at": "2026-10-30"}, {"url": "https://www.derczynski.com/itu/", "basis": "author_academic_biography_and_grants", "accessed_at": "2026-10-30T17:36:23Z"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "unknown", "detail": "No actual training hardware established for this cohort artifact."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Pinned card is metadata-only and config names local artgpt path. Public model discussion is conversion only; no starting checkpoint or task-training description found by exact model search."}
{"model_id": "Arjun-G-Ravi/GPT2-Alpaca", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/Arjun-G-Ravi/GPT2-Alpaca/raw/64310d166641b1db600b84fa9685afec6083e8dd/README.md", "accessed_at": "2026-10-30T12:28:00.808266+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Locutusque/gpt2-medium-conversational", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Two TinyMistral248M cards describe pretraining with single TitanV;v1 uses7,488,000examples,v2~500GPU-hours/OpenWebText1MplusTinyMistral data3epochs. Initial random weights not explicit. Five GPT2 conversational models are derived fine-tunes. Synthetic example addresses in cards are not author biography.", "provenance_sources": [{"url": "https://huggingface.co/Locutusque/gpt2-medium-conversational/raw/853ceb9495ecfe8d682929c93b7be56903928087/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "HexHands/finishSTUDIO", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Both releases have124439808 parameters and describe training on only2000 Scratch profile/studio texts. A GPT2-small task fine-tune is the best-supported interpretation, rather than original base pretraining.", "provenance_sources": [{"url": "https://huggingface.co/HexHands/finishSTUDIO/raw/898855221e43f865b4818ecf8680e9eab70d1649/README.md"}, {"url": "https://huggingface.co/HexHands/finishSTUDIO", "accessed_at": "2026-11-01", "read_method": "saved_card_read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inference_from_training_scale_and_checkpoint_size", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "DarwinAnim8or/GPT-Greentext-125m", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/DarwinAnim8or/GPT-Greentext-125m/raw/7dd691f1cc68500c8edcda929ef8e472e86aef15/README.md", "accessed_at": "2026-10-30T12:28:01.278656+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "iamtarun/pycompetitive-codegen350M-qlora", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/iamtarun/pycompetitive-codegen350M-qlora/raw/8cc741d06decedad7fe5b7fe8ad2dcd0ea18756a/README.md", "accessed_at": "2026-10-30T12:28:01.324732+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "BXYMartin/long-t5-tglobal-base-openhearthstone-v0.1", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'category': 'fine_tune', 'evidence': 'used the pretrained version ... and finetuned for 10 epochs', 'status': 'fine_tuning'}", "provenance_sources": [{"url": "https://huggingface.co/BXYMartin/long-t5-tglobal-base-openhearthstone-v0.1/blob/94261062cfcc31adac76f5182bb7b6d4b98dd131/README.md", "basis": "pinned model card", "accessed_at": "2026-10-30"}, {"url": "https://www.linkedin.com/in/xinyu-bai-a42514173/", "basis": "current creator location", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/BXYMartin/long-t5-tglobal-base-openhearthstone-v0.1/raw/94261062cfcc31adac76f5182bb7b6d4b98dd131/README.md", "accessed_at": "2026-10-30T12:28:01.366310+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "unknown"}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Technotech/MagicPrompt-tinystories-33M-epoch10-merged", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/Technotech/MagicPrompt-tinystories-33M-epoch10-merged/raw/1f7625e183d0d0a572fae62ea422dffd02ac58ea/README.md", "accessed_at": "2026-10-30T12:28:01.580041+00:00", "quote": "M, fine tune a 1-7B param model Training config - Rank 16 LoRA - Trained on Gustavosta/Stable-Diffusion-Prompts for 10 epochs - Batch size of 64 Training procedure The following bitsandbytes quantization config was used du", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "andrewkroening/GalaxyFarAway-DialoGPT-HanSolo", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/andrewkroening/GalaxyFarAway-DialoGPT-HanSolo/raw/6bc1c7dfaa3102418771abc8b59e4b8e7914feee/README.md", "accessed_at": "2026-10-30T12:28:01.979395+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "andrewkroening/GalaxyFarAway-DialoGPT-Vader", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/andrewkroening/GalaxyFarAway-DialoGPT-Vader/raw/016b688adcafbe8d92c59230e69044d4a2615255/README.md", "accessed_at": "2026-10-30T12:28:02.036966+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "marouni/miniDolly", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly starts from EleutherAI Pythia70M and instruction-trains on Dolly15K for12hours on the author's CPU.", "provenance_sources": [{"url": "https://huggingface.co/marouni/miniDolly/raw/bbfded354818ce336d6e1fad287b3f9d92a0efe1/README.md", "accessed_at": "2026-10-30T12:28:02.229476+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "sr5434/gptQuotes", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/sr5434/gptQuotes/raw/d511d262c74509effca2e26a083ef0ac9522cb96/README.md", "accessed_at": "2026-10-30T12:28:02.719740+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "crumb/distilpythia", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "distillation", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Card explicitly warm-starts student from smaller pretrained Pythia checkpoint and applies supervised plus teacher-soft-target losses.", "provenance_sources": [{"url": "https://huggingface.co/crumb/distilpythia/raw/7795b40f875e098630413dab066debe72a25aa10/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "l3cube-pune/hing-gpt-devanagari", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned card and project paper describe standard12-layerGPT2 trained for2epochs onHingCorpus at5e-5, with standard GPT2configuration/tokenizer. In the project continued-pretraining workflow the best-supported origin is adaptation of pretrained GPT2; the paper does not explicitly state GPT2initialisation, so this is an inference.", "provenance_sources": [{"url": "https://huggingface.co/l3cube-pune/hing-gpt-devanagari/blob/df18a9728c6fdb07de8bdeec555c15b71b08bd49/README.md", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/l3cube-pune/hing-gpt-devanagari/raw/df18a9728c6fdb07de8bdeec555c15b71b08bd49/config.json", "access_date": "2026-11-01"}, {"url": "https://arxiv.org/html/2204.08398#S7.SS1", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "gpt2", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "l3cube-pune/hing-gpt", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned card and project paper describe standard12-layerGPT2 trained for2epochs onHingCorpus at5e-5, with standard GPT2configuration/tokenizer. In the project continued-pretraining workflow the best-supported origin is adaptation of pretrained GPT2; the paper does not explicitly state GPT2initialisation, so this is an inference.", "provenance_sources": [{"url": "https://huggingface.co/l3cube-pune/hing-gpt/blob/8ea7ef1f2b5699b8d2e9877037967709700b0dc0/README.md", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/l3cube-pune/hing-gpt/raw/8ea7ef1f2b5699b8d2e9877037967709700b0dc0/config.json", "access_date": "2026-11-01"}, {"url": "https://arxiv.org/html/2204.08398#S7.SS1", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "gpt2", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "kingabzpro/DialoGPT-small-Rick-Bot", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "DialoGPT-small RickBot uses the Rick-and-Morty corpus, task-specific source-code project and measured perplexity8.53; best-supported origin is a DialoGPT adaptation.", "provenance_sources": [{"url": "https://huggingface.co/kingabzpro/DialoGPT-small-Rick-Bot/raw/73d7e10375e0ab0408913c91aba15c1d8f75d53b/README.md", "accessed_at": "2026-10-30T12:28:02.801005+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "crumb/opentinystories-68m-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card reports four training epochs on author's generated TinyStories corpus; config specifies custom 8-layer/512-width NeoX with1024 FFN. Original pretraining is the best inference; the copied30M card label is retained as a documentation discrepancy.", "provenance_sources": [{"url": "https://huggingface.co/crumb/opentinystories-68m-base/raw/022ee7978be08a3a8def188ca4665293bbaa26d2/README.md"}, {"url": "https://huggingface.co/crumb/opentinystories-68m-base/blob/main/README.md", "finding": "Saved card training description.", "reused": true}, {"url": "https://huggingface.co/crumb/opentinystories-68m-base/raw/main/config.json", "finding": "Read custom configuration.", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "card_and_custom_configuration_supported_inference", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Avitas8485/Dialogpt-small-v1", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=microsoft/DialoGPT-medium; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/Avitas8485/Dialogpt-small-v1/raw/25eea33aea9a2af28d48c6dd29981a4f4c245ef9/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01", "provenance_confidence": "medium", "upstream_checkpoint": "microsoft/DialoGPT-medium"}
{"model_id": "kinshuk-h/flan-t5-retacred-kg-w-context-var-len-small-finetuned", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/kinshuk-h/flan-t5-retacred-kg-w-context-var-len-small-finetuned/raw/41df3905370126738241964acca2a210339b26b4/README.md", "accessed_at": "2026-10-30T12:28:03.298607+00:00", "quote": "all](https://huggingface.co/google/flan-t5-small) finetuned over the TACRED Knowledge Graph patched with the [Re-TACRED proposal](https://github.com/gstoica27/Re-TACRED) using the training method for [KGT-5](https://github.com/apoorvum", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "dwojcik/gpt2-large-fine-tuned-context-256", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/dwojcik/gpt2-large-fine-tuned-context-256/raw/b0bb55d43aa9ad9d26e5959cde68db0b5b6a271a/README.md", "accessed_at": "2026-10-30T12:28:03.668156+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "rasta/BART-FHIR-question", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/rasta/BART-FHIR-question/raw/dd9084dec2029641ddb33e53f1f2a5e69678ecf4/README.md", "accessed_at": "2026-10-30T12:28:03.610069+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "harouzie/bart-base-paraphrase-v1", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=facebook/bart-base; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/harouzie/bart-base-paraphrase-v1/raw/63aea1ae309d052adac8382a7d33c904b9246272/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "facebook/bart-base", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "trawzified/khajiit-speak", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=t5-base; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/trawzified/khajiit-speak/raw/00da8eff8c264be43bb2667a39313301e96ab06c/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "t5-base", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "crumb/opentinystories-30m-base", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card reports four training epochs on author's generated TinyStories corpus; config specifies custom 6-layer/256-width NeoX model. These training and bespoke-scale facts support original from-scratch pretraining.", "provenance_sources": [{"url": "https://huggingface.co/crumb/opentinystories-30m-base/raw/50372a94540bbf34110491d7ce97e0b24a2cf3a3/README.md"}, {"url": "https://huggingface.co/crumb/opentinystories-30m-base/blob/main/README.md", "finding": "Saved card training description.", "reused": true}, {"url": "https://huggingface.co/crumb/opentinystories-30m-base/raw/main/config.json", "finding": "Read custom configuration.", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "card_and_custom_configuration_supported_inference", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "spitfire4794/dialogpt-small-morty", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=microsoft/DialoGPT-small; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/spitfire4794/dialogpt-small-morty/raw/7ed1c8ac1335932f12868b533ffaa154a04d881c/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "microsoft/DialoGPT-small", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "s3nh/DialoGPT-small-woody-toy-story", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/s3nh/DialoGPT-small-woody-toy-story/raw/0a1bc2e33a642a0ac0af3e1762bed023dfa0e7e7/README.md", "accessed_at": "2026-10-30T12:28:04.517035+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "dsfsi/ss-en-m2m100-gov", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card identifies a Siswati-to-English translation model based on M2M100 and adapted to the supplied South African government parallel corpus.", "provenance_sources": [{"url": "https://huggingface.co/dsfsi/ss-en-m2m100-gov/raw/7fc95db1a67ad35cf93e66a5d1f0eadfd1f689c3/README.md", "accessed_at": "2026-10-30T12:28:04.499230+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "gameofdimension/cs324-length-control", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/gameofdimension/cs324-length-control/raw/9db80e1c335cc25df77648952496710873f0ec76/README.md", "accessed_at": "2026-10-30T12:28:04.531998+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-410m-v0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-410m-v0/raw/dc7539d488d46992ea8b23461f2b6060534208cb/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:04.632707+00:00"}, {"url": "https://proceedings.mlr.press/v202/biderman23a/biderman23a.pdf", "basis": "2023_model_suite_paper_author_contribution_Appendix_A", "accessed_at": "2026-10-30"}, {"url": "https://herbiebradley.com/", "basis": "current_self_reported_research_location", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2304.01373v2", "basis": "Acknowledgments; Appendix A contributor roles; Appendix D actual original-suite training hardware", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-70m-deduped-v0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-70m-deduped-v0/raw/fefff897c5382cd0e1465128aa6867052a06ecd5/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:04.690385+00:00"}, {"url": "https://proceedings.mlr.press/v202/biderman23a/biderman23a.pdf", "basis": "2023_model_suite_paper_author_contribution_Appendix_A", "accessed_at": "2026-10-30"}, {"url": "https://herbiebradley.com/", "basis": "current_self_reported_research_location", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2304.01373v2", "basis": "Acknowledgments; Appendix A contributor roles; Appendix D actual original-suite training hardware", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-410m-deduped-v0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-410m-deduped-v0/raw/a7f858d8328ad9a7305c2302954caa46ca97fd70/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:04.936167+00:00"}, {"url": "https://proceedings.mlr.press/v202/biderman23a/biderman23a.pdf", "basis": "2023_model_suite_paper_author_contribution_Appendix_A", "accessed_at": "2026-10-30"}, {"url": "https://herbiebradley.com/", "basis": "current_self_reported_research_location", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2304.01373v2", "basis": "Acknowledgments; Appendix A contributor roles; Appendix D actual original-suite training hardware", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-160m-deduped-v0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-160m-deduped-v0/raw/26b9eeef106f66faf9b45930f92430a8a77c8799/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:05.199280+00:00"}, {"url": "https://proceedings.mlr.press/v202/biderman23a/biderman23a.pdf", "basis": "2023_model_suite_paper_author_contribution_Appendix_A", "accessed_at": "2026-10-30"}, {"url": "https://herbiebradley.com/", "basis": "current_self_reported_research_location", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2304.01373v2", "basis": "Acknowledgments; Appendix A contributor roles; Appendix D actual original-suite training hardware", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-70m-deduped", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-70m-deduped/raw/e93a9faa9c77e5d09219f6c868bfc7a1bd65593c/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:05.235351+00:00"}, {"url": "https://proceedings.mlr.press/v202/biderman23a/biderman23a.pdf", "basis": "2023_model_suite_paper_author_contribution_Appendix_A", "accessed_at": "2026-10-30"}, {"url": "https://herbiebradley.com/", "basis": "current_self_reported_research_location", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2304.01373v2", "basis": "Acknowledgments; Appendix A contributor roles; Appendix D actual original-suite training hardware", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-410m-deduped", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-410m-deduped/raw/c4fc8d586d62df497f1f9b69d66d3ca419992d3e/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:05.257423+00:00"}, {"url": "https://proceedings.mlr.press/v202/biderman23a/biderman23a.pdf", "basis": "2023_model_suite_paper_author_contribution_Appendix_A", "accessed_at": "2026-10-30"}, {"url": "https://herbiebradley.com/", "basis": "current_self_reported_research_location", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2304.01373v2", "basis": "Acknowledgments; Appendix A contributor roles; Appendix D actual original-suite training hardware", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-160m-deduped", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-160m-deduped/raw/582159a2dfe3e712a8d47ae83dec95ae3bde8e7e/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:05.346263+00:00"}, {"url": "https://proceedings.mlr.press/v202/biderman23a/biderman23a.pdf", "basis": "2023_model_suite_paper_author_contribution_Appendix_A", "accessed_at": "2026-10-30"}, {"url": "https://herbiebradley.com/", "basis": "current_self_reported_research_location", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2304.01373v2", "basis": "Acknowledgments; Appendix A contributor roles; Appendix D actual original-suite training hardware", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-160m-v0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-160m-v0/raw/3ef0c8cc72344bf3f6691d73c1a952892f8d1618/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:05.427943+00:00"}, {"url": "https://proceedings.mlr.press/v202/biderman23a/biderman23a.pdf", "basis": "2023_model_suite_paper_author_contribution_Appendix_A", "accessed_at": "2026-10-30"}, {"url": "https://herbiebradley.com/", "basis": "current_self_reported_research_location", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2304.01373v2", "basis": "Acknowledgments; Appendix A contributor roles; Appendix D actual original-suite training hardware", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-410m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-410m/raw/9879c9b5f8bea9051dcb0e68dff21493d67e9d4f/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:05.685553+00:00"}, {"url": "https://proceedings.mlr.press/v202/biderman23a/biderman23a.pdf", "basis": "2023_model_suite_paper_author_contribution_Appendix_A", "accessed_at": "2026-10-30"}, {"url": "https://herbiebradley.com/", "basis": "current_self_reported_research_location", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2304.01373v2", "basis": "Acknowledgments; Appendix A contributor roles; Appendix D actual original-suite training hardware", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-160m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-160m/raw/50f5173d932e8e61f858120bcb800b97af589f46/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:05.946946+00:00"}, {"url": "https://proceedings.mlr.press/v202/biderman23a/biderman23a.pdf", "basis": "2023_model_suite_paper_author_contribution_Appendix_A", "accessed_at": "2026-10-30"}, {"url": "https://herbiebradley.com/", "basis": "current_self_reported_research_location", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2304.01373v2", "basis": "Acknowledgments; Appendix A contributor roles; Appendix D actual original-suite training hardware", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "spitfire4794/dialogpt-small-rick", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=microsoft/DialoGPT-small; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/spitfire4794/dialogpt-small-rick/raw/232917ab7575caf5f4d84323204b693eb830168c/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "microsoft/DialoGPT-small", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "lxyuan/distilgpt2-finetuned-finance", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/lxyuan/distilgpt2-finetuned-finance/raw/e185be9dba22a0c041b26293483b55e39461119b/README.md", "accessed_at": "2026-10-30T12:28:06.023033+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "oliverguhr/spelling-correction-multilingual-base", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Cached config retains google/flan-t5-base and standard T5 architecture; card describes trained English/German spelling-correction task. Together support Flan-T5-base fine-tuning.", "provenance_sources": [{"url": "https://huggingface.co/oliverguhr/spelling-correction-multilingual-base/raw/033a56caf0e7b6dc0033e88ec146b89ccb22e951/README.md", "accessed_at": "2026-10-30T12:28:06.144197+00:00"}, {"url": "https://huggingface.co/oliverguhr/spelling-correction-multilingual-base/raw/033a56caf0e7b6dc0033e88ec146b89ccb22e951/README.md", "read_scope": "Saved card model/task/training and source attribution", "accessed_at": "2026-11-01", "local_evidence_only": true}, {"url": "https://huggingface.co/oliverguhr/spelling-correction-multilingual-base/raw/main/config.json", "read_scope": "Actual cached config base path and architecture", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "Binaryy/blender-bot-distill-finetuned", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Repository title indicates fine-tune; exact config retains facebook/blenderbot-400M-distill. Empty narrative card gives no actual loading/training record; base reference is retained without treating config path alone as proof.", "provenance_sources": [{"url": "https://huggingface.co/Binaryy/blender-bot-distill-finetuned/raw/c81cc33a2f3dc64a7524f82945a5505fd5f2ae54/config.json"}, {"url": "https://huggingface.co/Binaryy/blender-bot-distill-finetuned/raw/c81cc33a2f3dc64a7524f82945a5505fd5f2ae54/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "ManthanKulakarni/Text2JQLBuilder", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved model metadata names FLAN-T5 and own Text2JQL dataset, with task examples converting natural language to Jira queries. Best-supported origin is pretrained Google FLAN-T5 task adaptation.", "provenance_sources": [{"url": "https://huggingface.co/ManthanKulakarni/Text2JQLBuilder/raw/986f9e23f0630658c3611107a43f4822ae3690a8/README.md", "accessed_at": "2026-10-30T12:28:06.450447+00:00"}, {"url": "https://huggingface.co/ManthanKulakarni/Text2JQLBuilder/raw/986f9e23f0630658c3611107a43f4822ae3690a8/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "NihalSrivastava/advertisement-description-generator", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "GPT2 model hyperparameters with only last6layers unfrozen; config _name_or_path=gpt2, and advertisement-generation use case. CUDA inference example is not training hardware.", "provenance_sources": [{"url": "https://huggingface.co/NihalSrivastava/advertisement-description-generator/blob/main/README.md"}, {"url": "https://huggingface.co/NihalSrivastava/advertisement-description-generator/raw/bd5bfceda5bbb630681412ef2101808d729f0103/README.md", "accessed_at": "2026-10-30T12:28:06.695088+00:00"}], "uk_build_status": "no", "builder_location_country": "India", "builder_location_basis": "VITChennai on-site president/advisory role spans release period. Current Hyderabad location is separate; actual GPU country not documented.", "builder_location_evidence_type": "dated_on_site_affiliation_overlapping_release", "builder_location_period": "on-site VIT roles Jul2022–May2024, overlapping Apr–Jul2023 repository release", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "SRDdev/ScriptForge_Plus", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=gpt2; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/SRDdev/ScriptForge_Plus/raw/42a6f1edacc89988deb88acb5fc01bc48e8bd7db/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "gpt2", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "minhcrafters/DialoGPT-small-Fukuya", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Publisher release explicitly names DialoGPT-small plus a character/task suffix, with standard GPT2LMHeadModel configuration and tokenizer. Best-supported origin is a character chatbot adaptation of pretrained MicrosoftDialoGPT-small; exact run is undocumented.", "provenance_sources": [{"url": "https://huggingface.co/minhcrafters/DialoGPT-small-Fukuya/raw/1f3e7da78237bd327fabefa63dfd26af777d39f8/config.json", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/minhcrafters/DialoGPT-small-Fukuya/blob/1f3e7da78237bd327fabefa63dfd26af777d39f8/README.md", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "microsoft/DialoGPT-small", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "TheHappyDrone/DialoGPT-medium-Nexus-Nova-turing-v2", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/TheHappyDrone/DialoGPT-medium-Nexus-Nova-turing-v2/raw/3d27ddf1d161e2dbb860e4c7eec078b36ac14ccb/README.md", "accessed_at": "2026-10-30T12:28:06.829396+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Andrew5057/gpt-backstory-generator", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=EleutherAI/gpt-neo-125M; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/Andrew5057/gpt-backstory-generator/raw/9171d1822424075d46a3bb3e0a4acc2525aa0847/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "EleutherAI/gpt-neo-125M", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "TheHappyDrone/DialoGPT-medium-salesman", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "DialoGPT-medium identity, conversational persona/task and second release explicitly pointing to the newer same-family model support a DialoGPT persona fine-tune.", "provenance_sources": [{"url": "https://huggingface.co/TheHappyDrone/DialoGPT-medium-salesman/raw/96b53be4d95f66116dc82ec5e1309d16dae2448d/README.md", "accessed_at": "2026-10-30T12:28:07.181907+00:00"}, {"url": "https://huggingface.co/TheHappyDrone/DialoGPT-medium-salesman/raw/96b53be4d95f66116dc82ec5e1309d16dae2448d/README.md", "read_scope": "Cached model/base/architecture/training and attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "mgfrantz/distilgpt2-finetuned-reddit-tifu", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/mgfrantz/distilgpt2-finetuned-reddit-tifu/raw/688875aa412a0b296e89f8f9e1c8a7dd6c853419/README.md", "accessed_at": "2026-10-30T12:28:07.427925+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "KoboldAI/OPT-350M-Erebus", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "fairseq125M/355M are conversions of upstream Meta models, not KoboldAI pretraining. OPT350M-Erebus is MrSeeker fine-tune of pretrained OPT. HF histories identify Julius ter Pelkwijk/mrseeker87, Henk717, VE-forbryderne and bots. Original dense-paper training hardware is upstream, not conversion hardware.", "provenance_sources": [{"url": "https://huggingface.co/KoboldAI/OPT-350M-Erebus/raw/83ce2f4e78d308968cf7ecd03d86a1f64aea8336/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "pavanBuduguppa/asr_inverse_text_normalization", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/pavanBuduguppa/asr_inverse_text_normalization/raw/4f4970841ddd8e063cc4a70d14554bca441ba1b1/README.md", "accessed_at": "2026-10-30T12:28:07.528231+00:00", "quote": "alization --- --- --- asrinversetextnormalization Finetuned a facebook/bart-base Pretrained model on the ASR inverse text normalization dataset by treating it as a seq2seq task. Other approaches which may be considered is by considering it ", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "s3nh/DialoGPT-large-Morty", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/s3nh/DialoGPT-large-Morty/raw/dec81ba9356d367448ddbcd7df277786cce086d4/README.md", "accessed_at": "2026-10-30T12:28:07.582060+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sdadas/byt5-text-correction", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "ByT5 model identity and explicit trained multilingual punctuation/capitalization/diacritic correction support task fine-tuning of pretrained ByT5 family.", "provenance_sources": [{"url": "https://huggingface.co/sdadas/byt5-text-correction/raw/6f7e5953598cc24fa24f6a0344c768b959a658e4/README.md", "accessed_at": "2026-10-30T12:28:07.673391+00:00"}, {"url": "https://huggingface.co/sdadas/byt5-text-correction/raw/6f7e5953598cc24fa24f6a0344c768b959a658e4/README.md", "read_scope": "Saved card model/task/training and source attribution", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "syf2023/gpt2", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Frozen card explicitly declares duplicated_from: gpt2 and reproduces the upstream OpenAI/Hugging Face model card; no publisher-specific fine-tuning statement.", "provenance_sources": [{"url": "https://huggingface.co/syf2023/gpt2/raw/1ae84c91e91cd6c6910bc44f3bb68b69028baf05/README.md", "accessed_at": "2026-10-30T12:28:07.921445+00:00"}, {"url": "https://huggingface.co/syf2023/gpt2/raw/1ae84c91e91cd6c6910bc44f3bb68b69028baf05/README.md", "scope": "Complete frozen card including duplication header read", "excerpt": "duplicated_from: gpt2", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Redistribution of pretrained GPT-2.", "provenance_confidence": "high", "architecture_lineage": "OpenAI GPT-2 model mirrored by syf2023.", "evidence_updated_at": "2026-11-01"}
{"model_id": "bigscience/bloom-560m-intermediate", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "intermediate_pretraining_checkpoint", "provenance_sources": [{"url": "https://huggingface.co/bigscience/bloom-560m-intermediate/raw/fac2c0b5210dd3ad43b2e0e488ea9fa26d250bb2/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:08.183597+00:00"}, {"url": "https://github.com/davanstrien", "basis": "self_reported_profile_location", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2211.05100v4", "basis": "BLOOM_paper_dataset_and_applications_contributor", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/bigscience/bloom-560m-intermediate", "basis": "Compute infrastructure and training-location fields; includes generic 176B specifications", "accessed_at": "2026-10-30"}, {"url": "https://github.com/bigscience-workshop/bigscience/blob/master/train/tr11-176B-ml/smaller_models/tr11e-350M-ml.slurm", "basis": "Official matching small-model training configuration; not an execution log", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Exact intermediate-checkpoint allocation unresolved. The card repeats 176B-scale 384 A100 infrastructure. The official matching small-model configuration (tr11e-350M: 24 layers, hidden 1024, 16 heads) requests 32 V100-32GB GPUs; configuration is not an execution log.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "DarwinAnim8or/DailyChat-350M", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/DarwinAnim8or/DailyChat-350M/raw/d306c53d695a7c285faff872e21cef7e6ed0ff1c/README.md", "accessed_at": "2026-10-30T12:28:08.190623+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "p208p2002/bloomz-Alpaca-560M", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/p208p2002/bloomz-Alpaca-560M/raw/102f58842ab0e34b56a842121feb9ce901bafa5e/README.md", "accessed_at": "2026-10-30T12:28:08.266299+00:00", "quote": "? Response: tags: - text-generation-inference --- Fine-Tune BLOOMZ Model on Alpaca 52K.", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "bigscience/bloomz-560m", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "multitask_finetuning_existing_BLOOM", "provenance_sources": [{"url": "https://huggingface.co/bigscience/bloomz-560m/raw/a2845d7e13dd12efae154a9f1c63fcc2e0cc4b05/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:08.320625+00:00"}, {"url": "https://samuelalbanie.com/", "basis": "self_reported_personal_site_location", "accessed_at": "2026-10-30"}, {"url": "https://aclanthology.org/2023.acl-long.891.pdf", "basis": "paper_author_with_University_of_Cambridge_affiliation", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2211.01786v2", "basis": "Acknowledgments and Appendix A exact author contributions", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/bigscience/bloomz-560m", "basis": "Training layout and hardware fields", "accessed_at": "2026-10-30"}, {"url": "https://github.com/bigscience-workshop/bigscience/blob/master/train/tr13-mtf/smaller_models/tr13e-350m-mtf-xp3capmixnewcodelonglossseq-a100.slurm", "basis": "Official matching A100 small-model fine-tuning configuration", "accessed_at": "2026-10-30"}, {"url": "https://github.com/bigscience-workshop/bigscience/blob/master/train/tr13-mtf/smaller_models/tr13e-350m-mtf-xp3capmixnewcodelonglossseq.slurm", "basis": "Official matching V100 small-model fine-tuning configuration", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Card lists 64 A100-80GB project infrastructure and 1x1x1 model parallel layout. Official matching tr13e-350m configurations request either 32 A100 GPUs or 32 V100-32GB GPUs; exact released-checkpoint execution allocation is unresolved.", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "sn4kebyt3/ru-bart-large", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "vocabulary_pruning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly identifies facebook/mbart-large-50 with embeddings reduced to Russian/English; vocabulary shrunk250k to25k. Upstream pretrained mBART weights retained and pruned.", "provenance_sources": [{"url": "https://huggingface.co/sn4kebyt3/ru-bart-large/raw/ee1ed1982cf2c89dee39c5696af8e4d29f9ec904/README.md", "accessed_at": "2026-10-30T12:28:08.386540+00:00"}, {"url": "https://huggingface.co/sn4kebyt3/ru-bart-large/raw/ee1ed1982cf2c89dee39c5696af8e4d29f9ec904/README.md", "read_scope": "Saved card architecture/base/training/result and upstream attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "khanhj/testgpt2chatbot", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Frozen card is the inherited OpenAI/Hugging Face GPT-2 card, not a publisher-specific fine-tuning statement. Selected testgpt2chatbot config is GPT2LMHeadModel, 12 layers/768 width, vocab 50257. Best supported lineage is external pretrained GPT-2; exact publisher modification is not documented.", "provenance_sources": [{"url": "https://huggingface.co/khanhj/testgpt2chatbot/raw/45c71bf882fb643c5d82486d2d59ebb12374bb10/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not_documented", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "architecture_lineage": "OpenAI GPT-2 architecture and inherited model-card attribution", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "shihab17/bn-to-en-translation", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/shihab17/bn-to-en-translation/raw/45cec0e2033365f668ecb71ea01c804f35e09c53/README.md", "accessed_at": "2026-10-30T12:28:08.909732+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "JdThe65th/GPT2-Glitchfur-Zenith-JD", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Frozen card explicitly says pretrained GPT-2 fine-tuned on three named Twitter feeds using Boris Dayma HuggingTweets pipeline. Exact linked training run was rendered and read in earlier follow-through: Tesla T4, pretrained-model gpt2 input and finetuned-model output. The service workspace author is not uploader identity.", "provenance_sources": [{"url": "https://huggingface.co/JdThe65th/GPT2-Glitchfur-Zenith-JD/raw/6f27f3f84e1017e04e03358a0f8209612e14a0d9/README.md"}, {"url": "https://huggingface.co/JdThe65th/GPT2-Glitchfur-Zenith-JD/raw/6f27f3f84e1017e04e03358a0f8209612e14a0d9/README.md", "access_date": "2026-10-31", "scope": "complete frozen model card read"}, {"url": "https://forge.coreweave.com/wandb/wandb/huggingtweets/runs/pg3exi1g/overview?nw=nwuser_scott", "access_date": "2026-10-31", "scope": "valid prior rendered Overview read reused", "finding": "Tesla T4; GPT-2 input and fine-tuned output; shared HuggingTweets service attribution"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "Tesla T4 (exact linked HuggingTweets training run Overview, previously browser-read)", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "compute_source": "Hosted HuggingTweets/W&B project run; physical provider/country not stated", "creator_contribution": "Publishing tweet-fine-tuned GPT-2 release; no original pretrained backbone.", "architecture_lineage": "OpenAI pretrained GPT-2; Boris Dayma HuggingTweets training pipeline.", "hardware": "Tesla T4 (exact linked HuggingTweets training run Overview, previously browser-read)", "compute_sources": [{"url": "https://huggingface.co/JdThe65th/GPT2-Glitchfur-Zenith-JD/raw/6f27f3f84e1017e04e03358a0f8209612e14a0d9/README.md", "access_date": "2026-10-31", "scope": "complete frozen model card read"}, {"url": "https://forge.coreweave.com/wandb/wandb/huggingtweets/runs/pg3exi1g/overview?nw=nwuser_scott", "access_date": "2026-10-31", "scope": "valid prior rendered Overview read reused", "finding": "Tesla T4; GPT-2 input and fine-tuned output; shared HuggingTweets service attribution"}], "evidence_updated_at": "2026-10-31"}
{"model_id": "loresiensis/distilgpt2-emailgen-phishing", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/loresiensis/distilgpt2-emailgen-phishing/raw/47c55ac9ca2ca5af6887949898f382c97040e4b5/README.md", "accessed_at": "2026-10-30T12:28:08.981388+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "gray567/PModel", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "upstream_checkpoint_redistribution", "provenance_evidence_level": "exact_file_identity", "provenance_basis": "Frozen card explicitly duplicated_from:gpt2. All8 stored weight/converted-format LFS SHA256 hashes exactly match openai-community/gpt2, including Safetensors/PyTorch/TF/Flax/Rust and3TFLite variants. Generic fine-tuning use instructions in copied upstream card are not a publisher fine-tuning operation.", "provenance_sources": [{"url": "https://huggingface.co/gray567/PModel/raw/a1a480ef0ff1e639e01f19e3da07ec0f5a0a56a6/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/gray567/PModel/raw/a1a480ef0ff1e639e01f19e3da07ec0f5a0a56a6/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/gray567/PModel/raw/a1a480ef0ff1e639e01f19e3da07ec0f5a0a56a6/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/api/models/gray567/PModel/tree/a1a480ef0ff1e639e01f19e3da07ec0f5a0a56a6", "access_date": "2026-10-31", "scope": "full frozen file metadata read"}, {"url": "https://huggingface.co/api/models/openai-community/gpt2/tree/main", "access_date": "2026-10-31", "scope": "full upstream file metadata comparison; all8 weight/converted-format SHA256 hashes identical"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "upstream model hardware not publisher training", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "compute_source": "Not applicable to publisher: copied original-model TPU statement is upstream training, not this account training.", "creator_contribution": "Public duplication/renaming of upstream GPT-2 files; no new training or conversion operation established.", "architecture_and_code_lineage": "Unchanged OpenAI GPT-2 small checkpoint/config/tokenizer. Original model-building attribution belongs to OpenAI.", "evidence_updated_at": "2026-10-31"}
{"model_id": "DarwinAnim8or/NoSleepPromptGen", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/DarwinAnim8or/NoSleepPromptGen/raw/4fa1f618edc618c2797b0a16acffaefd5accb9b0/README.md", "accessed_at": "2026-10-30T12:28:09.103372+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "DarwinAnim8or/GPT-Greentext-355m", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/DarwinAnim8or/GPT-Greentext-355m/raw/816b9f40f80153c2c34aa40cd43dc3a6dfa0d293/README.md", "accessed_at": "2026-10-30T12:28:09.428335+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "MockingJ/chatbot", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "not_established", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Config _name_or_path=microsoft/DialoGPT-small,12×768, establishes pretrained checkpoint lineage. README is only YAML and does not state training operation.", "provenance_sources": [{"url": "https://huggingface.co/MockingJ/chatbot/blob/main/README.md"}, {"url": "https://huggingface.co/MockingJ/chatbot/raw/7b06026e8cadf236f4c842f836c1015fb28076c2/README.md", "accessed_at": "2026-10-30T12:28:09.630038+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "DarwinAnim8or/GPT-NoSleep-355m", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/DarwinAnim8or/GPT-NoSleep-355m/raw/4f702224a78cd79139726419ee976569e7f4983f/README.md", "accessed_at": "2026-10-30T12:28:09.626526+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "s3nh/DialoGPT-large-Rick", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/s3nh/DialoGPT-large-Rick/raw/6e6f54c7b04bd420fce1192fda84ad452dbb6aa4/README.md", "accessed_at": "2026-10-30T12:28:09.711448+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "SRDdev/ScriptForge-medium", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=gpt2; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/SRDdev/ScriptForge-medium/raw/90a26b0e01c5dc00678c64e93751931f644de7ba/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "gpt2", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "SRDdev/ScriptForge", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=gpt2; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/SRDdev/ScriptForge/raw/1a1c460f308d9c293de8ccb2f87efe46b614d5ca/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "gpt2", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "Narsil/gpt3", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/Narsil/gpt3/raw/aa766c0ffb50d62633142d9756f2fb287dc36427/README.md", "accessed_at": "2026-10-30T12:28:10.149825+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "NewBreaker/gpt2", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_or_conversion", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Copied OpenAI GPT2 card and upload history describe an existing pretrained GPT2 checkpoint; config12layerGPT2. Original256TPUv3 claim belongs to upstream larger GPT2 training, not uploader. No new uploader training claimed.", "provenance_sources": [{"url": "https://huggingface.co/NewBreaker/gpt2/blob/main/README.md"}, {"url": "https://huggingface.co/NewBreaker/gpt2/raw/edeb6a7c6d427819a89bc8c789b876377d0352a0/README.md", "accessed_at": "2026-10-30T12:28:10.376461+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "No dated builder place established for this run.", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "mrsteyk/memepp-llama-512v-6l-8h-256e", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly declares training from scratch or random initialization.", "provenance_sources": [{"url": "https://huggingface.co/mrsteyk/memepp-llama-512v-6l-8h-256e/raw/658de62fb665d6ac67089093d203e3331ab567c5/README.md", "accessed_at": "2026-10-30T12:28:10.376461+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "mrm8488/gpt2-finetuned-recipes-cooking_v2", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "task_finetuning", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Publisher release identifier explicitly says gpt2-finetuned-recipes-cooking; pinned config has GPT2LMHeadModel and standard GPT2dimensions, with recipe-generation examples and saved training arguments. Best-supported origin is task fine-tuning of GPT2.", "provenance_sources": [{"url": "https://huggingface.co/mrm8488/gpt2-finetuned-recipes-cooking_v2/raw/bd081c7eda76890eaa9a865213b637504daa673a/config.json", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/api/models/mrm8488/gpt2-finetuned-recipes-cooking_v2/tree/bd081c7eda76890eaa9a865213b637504daa673a?recursive=false&expand=false", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "gpt2", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "VMware/flan-t5-large-alpaca", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card explicitly names google/flan-t5-large parent and Alpaca style-transfer training,3 epochs on3 V100 GPUs.", "provenance_sources": [{"url": "https://huggingface.co/VMware/flan-t5-large-alpaca/raw/edd7ef2df791072f43c7620416fee54e4dfa6a62/README.md", "accessed_at": "2026-10-30T12:28:10.522126+00:00"}, {"url": "https://huggingface.co/VMware/flan-t5-large-alpaca/raw/edd7ef2df791072f43c7620416fee54e4dfa6a62/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "toloka/gpt2-large-rl-prompt-writing", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/toloka/gpt2-large-rl-prompt-writing/raw/6c7e8838cd1f94495c8b1b997bd031ce271183ab/README.md", "accessed_at": "2026-10-30T12:28:10.513609+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "RTT-FI/RTT-NLP-125M", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "probably_yes", "checkpoint_operation": "pretrained_checkpoint_derivative", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card explicitly says Based on GPT-2 for125M NLP checkpoint. Best-supported origin is GPT2-derived; exact modification/training operation is not described.", "provenance_sources": [{"url": "https://huggingface.co/RTT-FI/RTT-NLP-125M/raw/ca46efb8f70bc2cd2477f41dd838a330aa04b005/README.md", "accessed_at": "2026-10-30T12:28:10.851102+00:00"}, {"url": "https://huggingface.co/RTT-FI/RTT-NLP-125M/raw/ca46efb8f70bc2cd2477f41dd838a330aa04b005/README.md", "read_scope": "Saved card model/task/training and source attribution", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "MrD05/Kaido-350m", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "adaptation_of_pretrained_weights", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Pinned configuration retains _name_or_path=opt-350m; the release is presented as a task-specific GPT/OPT/BART/T5/Qwen checkpoint. Best-supported origin is adaptation of that pretrained base.", "provenance_sources": [{"url": "https://huggingface.co/MrD05/Kaido-350m/raw/98e431be878c8604285d94197e22d225857aadbf/config.json", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01", "provenance_confidence": "medium", "upstream_checkpoint": "opt-350m"}
{"model_id": "vicgalle/gpt2-open-instruct-v1", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/vicgalle/gpt2-open-instruct-v1/raw/eccf4d5899c24523625fe3d41f1cf78c755821b0/README.md", "accessed_at": "2026-10-30T12:28:11.128950+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "Narsil/gpt2", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/Narsil/gpt2/raw/7752b7a35e4230c2ef46c70634615ffb009a2b48/README.md", "accessed_at": "2026-10-30T12:28:11.186944+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "robowaifudev/megatron-gpt2-345m", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Frozen card explicitly identifies this as an archive of nvidia/megatron-gpt2-345m, with original NVIDIA attribution and upstream code/paper.", "provenance_sources": [{"url": "https://huggingface.co/robowaifudev/megatron-gpt2-345m/raw/b39f8d00fb9f33da4271be2035da848da896a23b/README.md", "accessed_at": "2026-10-30T12:28:11.238530+00:00"}, {"url": "https://huggingface.co/robowaifudev/megatron-gpt2-345m/raw/b39f8d00fb9f33da4271be2035da848da896a23b/README.md", "scope": "Complete frozen card read", "excerpt": "This is an archive of nvidia/megatron-gpt2-345m", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "creator_contribution": "Archiving and redistributing original NVIDIA weights; underlying model creation remains NVIDIA.", "provenance_confidence": "high", "architecture_lineage": "NVIDIA Megatron GPT-2 pretrained model.", "evidence_updated_at": "2026-11-01"}
{"model_id": "concedo/Pythia-70M-ChatSalad", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact configuration welcome text explicitly identifies follow-up fine-tune of Pythia-70M on ChatSalad data.", "provenance_sources": [{"url": "https://huggingface.co/concedo/Pythia-70M-ChatSalad/raw/692289413c47c219cf83b1596783a8e9223541eb/config.json"}, {"url": "https://huggingface.co/concedo/Pythia-70M-ChatSalad/raw/692289413c47c219cf83b1596783a8e9223541eb/README.md"}, {"url": "https://huggingface.co/concedo/Pythia-70M-ChatSalad/raw/692289413c47c219cf83b1596783a8e9223541eb/README.md", "access_date": "2026-10-31", "scope": "complete frozen file read"}, {"url": "https://huggingface.co/concedo/Pythia-70M-ChatSalad/raw/692289413c47c219cf83b1596783a8e9223541eb/config.json", "access_date": "2026-10-31", "scope": "complete frozen file read"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "ChatSalad fine-tuning of upstream Pythia-70M; later safetensors conversion PR accepted with owner noting they had not actually tried it.", "architecture_and_code_lineage": "GPTNeoXForCausalLM configuration from upstream EleutherAI Pythia-70M; explicit follow-up fine-tune on ChatSalad, not original base-model building.", "evidence_updated_at": "2026-10-31"}
{"model_id": "Corianas/Quokka_111m", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "reviewed_model_evidence", "provenance_basis": "Exact card explicitly describes Cerebras GPT fine-tuning using Dolly tools; this is not merely a retained config path.", "provenance_sources": [{"url": "https://huggingface.co/Corianas/Quokka_111m/raw/4ed0941f119b73e7da1a8b223690ff52df1e7e4a/README.md"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "0x7o/pyGPT-50M", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration.", "provenance_sources": [{"url": "https://huggingface.co/0x7o/pyGPT-50M/raw/673d3fd12a3e2df378521b3940d9c69f174abd93/README.md", "accessed_at": "2026-10-30T12:28:11.858113+00:00", "quote": " mpl-2.0 --- PythonGPT A GPT2-type neural network trained on 16 gigabytes of Pyhon scripts from scratch. It has 50 million parameters. Made as a toy.", "basis": "The saved card explicitly describes original from-scratch training for this checkpoint; training details or sample usage accompany the declaration."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "DarwinAnim8or/gpt-grug-125m", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/DarwinAnim8or/gpt-grug-125m/raw/d9ed85dd5a2ad6e2353e61c45dbf0ff736d4b5de/README.md", "accessed_at": "2026-10-30T12:28:11.877843+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "DarwinAnim8or/GPT-Grug-355m", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/DarwinAnim8or/GPT-Grug-355m/raw/adecea073e27371d800aa158eda15ae08e93caff/README.md", "accessed_at": "2026-10-30T12:28:11.930237+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "dkagramanyan/horoscope_rugpt3small", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card expressly says horoscope generation based on sberbank-ai rugpt3small; task-specific adaptation of external pretrained RuGPT backbone.", "provenance_sources": [{"url": "https://huggingface.co/dkagramanyan/horoscope_rugpt3small/raw/eea74acbe84181760f79c114ef71e911e47a2d53/README.md", "accessed_at": "2026-10-30T12:28:11.968329+00:00"}, {"url": "https://huggingface.co/dkagramanyan/horoscope_rugpt3small/raw/eea74acbe84181760f79c114ef71e911e47a2d53/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high"}
{"model_id": "benkimz/agbrain", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/benkimz/agbrain/raw/f1524dba013cfc4ab3640420363f34fd1fa4d30e/README.md", "accessed_at": "2026-10-30T12:28:12.016411+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "XBOT-RK/distilgpt2-wiki-qa", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuned_existing_checkpoint", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'fine_tuned_existing_checkpoint', 'detail': 'Card explicitly says fine-tuned from distilgpt2 on Microsoft Wiki-QA.'}", "provenance_sources": [{"url": "https://huggingface.co/XBOT-RK/distilgpt2-wiki-qa/raw/19e9df3f22291228e6733301c7e9b972fd4f1d1f/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:12.361912+00:00"}, {"url": "https://vrajeshtrichy.github.io/website/education/", "basis": "author_portfolio", "accessed_at": "2026-10-30T17:28:09Z"}, {"url": "https://vrajeshtrichy.github.io/website/experience/", "basis": "author_portfolio", "accessed_at": "2026-10-30T17:28:10Z"}, {"url": "https://github.com/vrajeshtrichy", "basis": "HF_linked_GitHub_profile", "accessed_at": "2026-10-30"}, {"url": "https://uk.linkedin.com/in/vrajeshtrichy", "basis": "public_LinkedIn_profile", "accessed_at": "2026-10-30T17:28:10Z"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "unknown", "detail": "Usage selects CUDA if available, not training-hardware evidence."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "EleutherAI/pythia-70m-v0", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "from_scratch_pretraining", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "from_scratch_pretraining", "provenance_sources": [{"url": "https://huggingface.co/EleutherAI/pythia-70m-v0/raw/57d0448e12b146f9d3decf2d64df9a6dc5194f7b/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:12.597184+00:00"}, {"url": "https://proceedings.mlr.press/v202/biderman23a/biderman23a.pdf", "basis": "2023_model_suite_paper_author_contribution_Appendix_A", "accessed_at": "2026-10-30"}, {"url": "https://herbiebradley.com/", "basis": "current_self_reported_research_location", "accessed_at": "2026-10-30"}, {"url": "https://arxiv.org/html/2304.01373v2", "basis": "Acknowledgments; Appendix A contributor roles; Appendix D actual original-suite training hardware", "accessed_at": "2026-10-30"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "see_existing_exact_model_evidence", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "voidful/context-only-question-generator", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly describes a task-specific question/distractor generator based on pretrained BART-base.", "provenance_sources": [{"url": "https://huggingface.co/voidful/context-only-question-generator/raw/55f2c093bd86e78fdcbd25e62ad856d1be99ecd6/README.md", "accessed_at": "2026-10-30T12:28:12.623820+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "philippelaban/keep_it_simple", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Cached config identifies gpt2-medium; card describes trained paragraph-simplification behavior with CNN/DailyMail and linked training code/log. Supports GPT2-medium task post-training.", "provenance_sources": [{"url": "https://huggingface.co/philippelaban/keep_it_simple/raw/3b77c90fda64d1d054ee754fda21c26d71409938/README.md", "accessed_at": "2026-10-30T12:28:12.678247+00:00"}, {"url": "https://huggingface.co/philippelaban/keep_it_simple/raw/3b77c90fda64d1d054ee754fda21c26d71409938/README.md", "read_scope": "Saved card model/task/training and source attribution", "accessed_at": "2026-11-01", "local_evidence_only": true}, {"url": "https://huggingface.co/philippelaban/keep_it_simple/raw/main/config.json", "read_scope": "Actual cached config base path and architecture", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "sagorsarker/emailgenerator", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented.", "provenance_sources": [{"url": "https://huggingface.co/sagorsarker/emailgenerator/raw/76d1a906c33da76778a951397aa6346c56f624c1/README.md", "accessed_at": "2026-10-30T12:28:12.695297+00:00", "quote": "l-generation license: mit --- EmailGenerator is a gpt-2 fine-tuned text-generation pre-trained model trained on [emailblog](https://www.kaggle.com/datasets/mikeschmidtavemac/emailblog) datasets for [EmailWriter](https://github", "basis": "Saved card identifies fine-tuning, continued pretraining or post-training in the released checkpoint. A publisher-owned from-scratch lineage is retained separately when explicitly documented."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "arun-shankar/GPT-2-covid-news-articles", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "continued_pretraining_of_external_model", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly describes GPT-2 further pretrained on COVID news articles.", "provenance_sources": [{"url": "https://huggingface.co/arun-shankar/GPT-2-covid-news-articles/raw/0a86a75cb47df38bc29efa900b3f960c7a37b784/README.md", "accessed_at": "2026-10-30T12:28:12.784570+00:00"}, {"url": "https://huggingface.co/arun-shankar/GPT-2-covid-news-articles/raw/0a86a75cb47df38bc29efa900b3f960c7a37b784/README.md", "access_date": "2026-11-01", "scope": "Saved model-card introduction and training/family sections read", "finding": "Card explicitly describes GPT-2 further pretrained on COVID news articles."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "Graverman/t5-code-summary", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Frozen actual config explicitly names _name_or_path=t5-large. Card describes code summarization with CodeSearchNet/CoNaLa training-data tags; best-supported operation is task fine-tuning of pretrained T5-large.", "provenance_sources": [{"url": "https://huggingface.co/Graverman/t5-code-summary/raw/e7169f2ec0682c5e0d0dbeb9fa416fcfd2bbcab9/README.md", "basis": "pinned cohort model card", "accessed_at": "2026-10-30"}, {"url": "https://huggingface.co/Graverman/t5-code-summary/raw/e7169f2ec0682c5e0d0dbeb9fa416fcfd2bbcab9/README.md", "scope": "Complete frozen card read", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/Graverman/t5-code-summary/raw/e7169f2ec0682c5e0d0dbeb9fa416fcfd2bbcab9/config.json", "scope": "Complete configuration read", "excerpt": "\"_name_or_path\": \"t5-large\"", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not stated for exact run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "creator_contribution": "Published Python-code summarization adaptation using CodeSearchNet/CoNaLa per card.", "provenance_confidence": "medium", "architecture_lineage": "External pretrained T5-large encoder-decoder (24 encoder/24 decoder layers, width1024).", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "DarwinAnim8or/GPT-DMV-125m", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/DarwinAnim8or/GPT-DMV-125m/raw/5868c798d51bc5327bb70c4c68fb58020a6b3041/README.md", "accessed_at": "2026-10-30T12:28:13.375858+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "voidful/bart-distractor-generation-both", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly describes a task-specific question/distractor generator based on pretrained BART-base.", "provenance_sources": [{"url": "https://huggingface.co/voidful/bart-distractor-generation-both/raw/b68bd438cbb0d456ef0e89fa9c7ec8797fd0101d/README.md", "accessed_at": "2026-10-30T12:28:13.342321+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "voidful/bart-distractor-generation", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly describes a task-specific question/distractor generator based on pretrained BART-base.", "provenance_sources": [{"url": "https://huggingface.co/voidful/bart-distractor-generation/raw/05eba9563297cfe8ef463d3004cf8936eac5c2e6/README.md", "accessed_at": "2026-10-30T12:28:13.400908+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "voidful/bart-distractor-generation-pm", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly describes a task-specific question/distractor generator based on pretrained BART-base.", "provenance_sources": [{"url": "https://huggingface.co/voidful/bart-distractor-generation-pm/raw/945b2e517984ceda3fbe5305b253d428772a7788/README.md", "accessed_at": "2026-10-30T12:28:13.517777+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "Ar4ikov/gpt2-650k-stable-diffusion-prompt-generator", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "GPT2-labelled task-specific prompt generator lists training prompt datasets and demonstrates learned Stable-Diffusion-style generation; best-supported classification is GPT2 adaptation.", "provenance_sources": [{"url": "https://huggingface.co/Ar4ikov/gpt2-650k-stable-diffusion-prompt-generator/raw/30a70df708cb7a5ea969fb8559c682f21502cfb5/README.md", "accessed_at": "2026-10-30T12:28:13.493202+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "bigscience/distill-bloom-1b3", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "distillation", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "distillation_incomplete_checkpoint", "provenance_sources": [{"url": "https://huggingface.co/bigscience/distill-bloom-1b3/raw/e61d41db91584464739a6722269725026c8c1cbf/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:13.742610+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "unknown_for_exact_run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "TheHappyDrone/DialoGPT-medium-Nexus-Nova", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "DialoGPT-medium identity, conversational persona/task and second release explicitly pointing to the newer same-family model support a DialoGPT persona fine-tune.", "provenance_sources": [{"url": "https://huggingface.co/TheHappyDrone/DialoGPT-medium-Nexus-Nova/raw/1e0f0d5a3b9823bc033ccd4a40e8fe82f04ddc43/README.md", "accessed_at": "2026-10-30T12:28:14.052023+00:00"}, {"url": "https://huggingface.co/TheHappyDrone/DialoGPT-medium-Nexus-Nova/raw/1e0f0d5a3b9823bc033ccd4a40e8fe82f04ddc43/README.md", "read_scope": "Cached model/base/architecture/training and attribution descriptions", "accessed_at": "2026-11-01", "local_evidence_only": true}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "voidful/bart-eqg-question-generator", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly describes a task-specific question/distractor generator based on pretrained BART-base.", "provenance_sources": [{"url": "https://huggingface.co/voidful/bart-eqg-question-generator/raw/65c680bdc3b513d1e15fab0b4c99491cfec4398f/README.md", "accessed_at": "2026-10-30T12:28:14.064540+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "thefrigidliquidation/pythia-410m-lightnovels", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Frozen card explicitly identifies Pythia-410M fine-tuning on light/web novels for translation and generation.", "provenance_sources": [{"url": "https://huggingface.co/thefrigidliquidation/pythia-410m-lightnovels/raw/ac23cc0b9d81ca5e4211bfe6653f8a010af221f7/README.md", "accessed_at": "2026-10-30T12:28:14.094221+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "creator_contribution": "Novel-domain English/Japanese fine-tuning with translation-format tokens.", "architecture_lineage": "External EleutherAI Pythia-410M pretrained backbone.", "evidence_updated_at": "2026-11-01"}
{"model_id": "thu-coai/blenderbot-400M-esconv", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/thu-coai/blenderbot-400M-esconv/raw/0f924942b3a177c38f6407c09aff5f46d5307de6/README.md", "accessed_at": "2026-10-30T12:28:14.327814+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "s3nh/DialoGPT-small-hermione-granger-goblet-of-fire", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/s3nh/DialoGPT-small-hermione-granger-goblet-of-fire/raw/05eeedce5997306deb44504e54568955408188e1/README.md", "accessed_at": "2026-10-30T12:28:14.213535+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "shibing624/code-autocomplete-gpt2-base", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved code-autocomplete card identifies GPT2/DistilGPT2 model family and gives pytorch_awesome project source code as training data, with upstream OpenAI GPT2 attribution retained. Best-supported operation is code-domain fine-tuning of the named pretrained backbone.", "provenance_sources": [{"url": "https://huggingface.co/shibing624/code-autocomplete-gpt2-base/raw/b9d6dd3ccfce8467bc50e2451160217a972f2411/README.md", "accessed_at": "2026-10-30T12:28:14.466046+00:00"}, {"url": "https://huggingface.co/shibing624/code-autocomplete-gpt2-base/raw/b9d6dd3ccfce8467bc50e2451160217a972f2411/README.md", "access_date": "2026-11-01", "scope": "Saved card project description, named pretrained family, Train data and original GPT2 attribution read locally"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "Beltenebros/DialoGPT-small-PerionOfGaul", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Saved DialoGPT-small card describes a tutorial-based conversational bot trained on Perion of Gaul book dialogues. Model identity and task adaptation support pretrained Microsoft DialoGPT-small fine-tuning, not original backbone pretraining.", "provenance_sources": [{"url": "https://huggingface.co/Beltenebros/DialoGPT-small-PerionOfGaul/raw/2dd92120bf94892953eb2920a2f05899402c7dbb/README.md", "accessed_at": "2026-10-30T12:28:14.960897+00:00"}, {"url": "https://huggingface.co/Beltenebros/DialoGPT-small-PerionOfGaul/raw/2dd92120bf94892953eb2920a2f05899402c7dbb/README.md", "access_date": "2026-11-01", "scope": "Saved frozen model-card training/origin/architecture/attribution text read locally; no network"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "inferred_from_saved_development_evidence", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "bigscience/distill-bloom-1b3-10x", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "distillation", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "distillation_incomplete_checkpoint", "provenance_sources": [{"url": "https://huggingface.co/bigscience/distill-bloom-1b3-10x/raw/37767e4eaa3ec0e2ea0853fed93aba4af6f42f92/README.md", "basis": "pinned_model_card", "accessed_at": "2026-10-30T12:28:14.948371+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "unknown_for_exact_run", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "s3nh/DialoGPT-small-morty", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/s3nh/DialoGPT-small-morty/raw/63720ab64487d10eda5f121f4df3cc4354c1ffd5/README.md", "accessed_at": "2026-10-30T12:28:14.964896+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "crumb/pico-gpt-j-6.7m", "model_provenance": "from_scratch", "own_from_scratch_lineage": "yes", "trained_status": "yes", "checkpoint_operation": "pretraining", "provenance_evidence_level": "inferred_from_saved_development_evidence", "provenance_basis": "Card reports a custom4-layer/64-width6.7M GPT-J trained on0.53B C4 tokens; config names init_model. This custom miniature architecture and initial model path support original pretraining.", "provenance_sources": [{"url": "https://huggingface.co/crumb/pico-gpt-j-6.7m/raw/0bf1c6983c029d4c2e4a64264e153655f8c01757/README.md"}, {"url": "https://huggingface.co/crumb/pico-gpt-j-6.7m/blob/main/README.md", "finding": "Saved card training description.", "reused": true}, {"url": "https://huggingface.co/crumb/pico-gpt-j-6.7m/raw/main/config.json", "finding": "Read custom configuration.", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "medium", "provenance_evidence_detail": "card_and_custom_configuration_supported_inference", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium", "evidence_updated_at": "2026-11-01"}
{"model_id": "mrm8488/gpt2-finetuned-recipes-cooking", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "task_finetuning", "provenance_evidence_level": "inferred_from_model_configuration_and_task", "provenance_basis": "Publisher release identifier explicitly says gpt2-finetuned-recipes-cooking; pinned config has GPT2LMHeadModel and standard GPT2dimensions, with recipe-generation examples and saved training arguments. Best-supported origin is task fine-tuning of GPT2.", "provenance_sources": [{"url": "https://huggingface.co/mrm8488/gpt2-finetuned-recipes-cooking/raw/7c686261e052bc9e761a482c8bf3ea15e4d486f8/config.json", "access_date": "2026-11-01"}, {"url": "https://huggingface.co/api/models/mrm8488/gpt2-finetuned-recipes-cooking/tree/7c686261e052bc9e761a482c8bf3ea15e4d486f8?recursive=false&expand=false", "access_date": "2026-11-01"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "", "provenance_confidence": "medium", "upstream_checkpoint": "gpt2", "origin_assessment_type": "inferred", "origin_assessment_confidence": "medium"}
{"model_id": "mrm8488/CodeGPT-small-finetuned-python-token-completion", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/mrm8488/CodeGPT-small-finetuned-python-token-completion/raw/6baa93b88684835848809d041a432fea4ad738ba/README.md", "accessed_at": "2026-10-30T12:28:15.354371+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "FredZhang7/anime-anything-promptgen-v2", "model_provenance": "derived", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "fine_tuning", "provenance_evidence_level": "publisher_statement", "provenance_basis": "Card explicitly calls this a fine-tuned GPT2-family model, with DistilGPT2 tokenization and3 epochs on80K curated anime prompts.", "provenance_sources": [{"url": "https://huggingface.co/FredZhang7/anime-anything-promptgen-v2/raw/0b3297c092ac2ecc9f1a336b8762642af540a7c4/README.md", "accessed_at": "2026-10-30T12:28:15.804153+00:00"}, {"access_date": "2026-11-01", "finding": "Read saved model card; no new source retrieval."}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": "Weight initialization and/or actual training are not established.", "provenance_confidence": "high", "evidence_updated_at": "2026-11-01"}
{"model_id": "mrsteyk/openchatgpt-neox-125m", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/mrsteyk/openchatgpt-neox-125m/raw/e4375536a68062074de9100ae915bd1c85ca292e/README.md", "accessed_at": "2026-10-30T12:28:15.796638+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "mrsteyk/openchatgpt-neo-125m", "model_provenance": "derived", "own_from_scratch_lineage": "unknown", "trained_status": "yes", "checkpoint_operation": "fine_tune_continuation_or_distillation", "provenance_evidence_level": "publisher_statement", "provenance_basis": "The saved model card explicitly describes a fine-tuning, continuation or distillation operation.", "provenance_sources": [{"url": "https://huggingface.co/mrsteyk/openchatgpt-neo-125m/raw/fb17f2c2f7086eabc4aaa95715aa8a1311ac1a2e/README.md", "accessed_at": "2026-10-30T12:28:15.812654+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": "not established", "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "model-attribution-challenge/gpt2", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_of_existing_pretrained_base_model", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'mirror_of_existing_pretrained_base_model', 'base_family': 'OpenAI GPT-2', 'from_scratch_by_context_person': 'not_established', 'basis': 'Pinned card attributes original pretrained model to OpenAI and its2019 paper.'}", "provenance_sources": [{"url": "https://huggingface.co/model-attribution-challenge/gpt2/raw/75e09b43581151bd1d9ef6700faa605df408979f/README.md", "basis": "Pinned card from the exact frozen cohort.", "accessed_at": "2026-10-30T12:28:15.850619+00:00"}, {"url": "https://github.com/gante", "basis": "Current public self-location and career README.", "accessed_at": "2026-10-30T18:14:49.414211+00:00"}, {"url": "https://x.com/joao_gante", "basis": "Current directly linked self-description.", "accessed_at": "2026-10-30T18:14:49.414211+00:00"}, {"url": "https://deec.tecnico.ulisboa.pt/en/news/community/alumni-session-joao-gante-partilha-a-sua-jornada-enquanto-estudante-do-doutoramento-em-engenharia-eletrotecnica-e-de-computadores", "basis": "Institutional alumnus biography, published3November2025.", "accessed_at": "2026-10-30T18:14:49.414211+00:00"}, {"url": "https://mlmac.io/mlmac_satml.pdf", "basis": "Official challenge report; Hugging Face affiliation, 2022 competition.", "accessed_at": "2026-10-30T18:12:48.957225+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "exact_124M_checkpoint_training_hardware_not_specified", "family_context": "Card mentions256 TPUv3 cores for the larger GPT-2 model; do not assign that allocation to this checkpoint or to João Gante."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}
{"model_id": "model-attribution-challenge/bloom-560m", "model_provenance": "mirror", "own_from_scratch_lineage": "no", "trained_status": "yes", "checkpoint_operation": "mirror_of_existing_pretrained_base_model", "provenance_evidence_level": "detailed_model_assessment", "provenance_basis": "{'status': 'mirror_of_existing_pretrained_base_model', 'base_family': 'BigScience BLOOM', 'from_scratch_by_context_person': 'not_established'}", "provenance_sources": [{"url": "https://huggingface.co/model-attribution-challenge/bloom-560m/raw/afe2e6f33eb135d254df849c74bb83322b53641c/README.md", "basis": "Pinned card from the exact frozen cohort.", "accessed_at": "2026-10-30T12:28:16.091011+00:00"}, {"url": "https://github.com/gante", "basis": "Current public self-location and career README.", "accessed_at": "2026-10-30T18:14:49.414211+00:00"}, {"url": "https://x.com/joao_gante", "basis": "Current directly linked self-description.", "accessed_at": "2026-10-30T18:14:49.414211+00:00"}, {"url": "https://deec.tecnico.ulisboa.pt/en/news/community/alumni-session-joao-gante-partilha-a-sua-jornada-enquanto-estudante-do-doutoramento-em-engenharia-eletrotecnica-e-de-computadores", "basis": "Institutional alumnus biography, published3November2025.", "accessed_at": "2026-10-30T18:14:49.414211+00:00"}, {"url": "https://mlmac.io/mlmac_satml.pdf", "basis": "Official challenge report; Hugging Face affiliation, 2022 competition.", "accessed_at": "2026-10-30T18:12:48.957225+00:00"}], "uk_build_status": "unknown", "builder_location_country": "not established", "builder_location_basis": "", "builder_location_evidence_type": "unresolved", "builder_location_period": "not established", "compute_used": {"status": "upstream_cluster_context_exact_small_run_allocation_unresolved", "reported_in_card": "Jean Zay;384 A10080GB GPUs plus32 reserve. The card also links the176B training recipe; do not treat family infrastructure numbers as an exact560M run allocation."}, "compute_country": "not established", "public_record_date": "2026-10-30", "open_question": ""}