Datasets:

shibatch
/

random_models

@@ -61,3 +61,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 livekit-turn-detector/gguf/model-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
 minicpm5/gguf/model-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
 tinyllama-chat/gguf/model-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text

 livekit-turn-detector/gguf/model-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
 minicpm5/gguf/model-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
 tinyllama-chat/gguf/model-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
+gemma4-random-model/gguf-q4_0/gemma4-random-model-Q4_0.gguf filter=lfs diff=lfs merge=lfs -text
+qwen3-random-model/gguf-q4_0/qwen3-random-model-Q4_0.gguf filter=lfs diff=lfs merge=lfs -text

GEMMA4_QWEN3_REPORT.json ADDED Viewed

	@@ -0,0 +1,40 @@

+{
+  "cases": [
+    {
+      "architecture": "gemma4",
+      "case": "gemma4-random-model",
+      "eog_token_ids": [
+        1
+      ],
+      "gguf_quantization": "Q4_0",
+      "gguf_sha256": "2b631587994acd127d3e6fcc67552c707d6fdba22beada963597387b9e814881",
+      "hf_dtype": "bfloat16",
+      "hf_sha256": "18c8b09759c831fbc9d6a8caf7df8e9d395508d4aec09336167f675527b3a360",
+      "parameter_count": 1519168,
+      "tensor_type_histogram": {
+        "F32": 44,
+        "Q4_0": 42
+      },
+      "validation": "passed"
+    },
+    {
+      "architecture": "qwen3",
+      "case": "qwen3-random-model",
+      "eog_token_ids": [
+        1
+      ],
+      "gguf_quantization": "Q4_0",
+      "gguf_sha256": "6af2e0d230a7919c6b1a02b6462f0441ffbadbb8188b1bd9c0f8174fecd9308f",
+      "hf_dtype": "bfloat16",
+      "hf_sha256": "0960438e0541dfa529703f5e2b37565aec66b354f6e45bfdf121ecadfcd86dd4",
+      "parameter_count": 508800,
+      "tensor_type_histogram": {
+        "F32": 9,
+        "Q4_0": 15
+      },
+      "validation": "passed"
+    }
+  ],
+  "llama_cpp_commit": "40b740ad05c531b9d57aca6698c3ed553a9e784c",
+  "schema_version": 1
+}

README.md CHANGED Viewed

@@ -1,315 +1,130 @@
----
-pretty_name: Real-Config-Derived Tiny Llama Models
-license: other
-size_categories:
-- n<1K
-tags:
-- llama
-- transformers
-- safetensors
-- gguf
-- synthetic
-- conformance
-- compatibility-testing
----
-# Real-Config-Derived Tiny Llama Models
-This dataset contains small, deterministic Llama-compatible models derived from `config.json` files of real models published on the Hugging Face Hub.
-The models are intended for model-format, loader, inference, compatibility, and conformance testing. They are not intended for language-model quality evaluation.
-Each case preserves an observed real-world configuration as much as possible while reducing model geometry enough to make the resulting model inexpensive to store and execute.
-The current collection contains seven Llama-family cases in both:
-* unquantized F32 Safetensors format;
-* GGUF `Q4_K_M` format.
-All model weights in this dataset are synthetic and deterministically generated. No original model weights are included.
-## Why this dataset exists
-Small hand-written test models are useful, but they tend to exercise only a small number of clean, canonical configuration patterns.
-Real model repositories contain considerably more variation, including:
-* MHA and GQA configurations;
-* different query/KV head ratios;
-* explicit versus implicit `head_dim`;
-* tied and untied embeddings;
-* different special-token layouts;
-* multiple EOS token IDs;
-* older and newer RoPE scaling representations;
-* configuration files produced by different Transformers versions.
-This dataset retains such real configuration patterns while replacing the original large geometry and weights with small deterministic equivalents.
-The goal is to provide compact test inputs that remain representative of model configurations that actually exist in the ecosystem.
-## Cases
-| Case                    | Source model                               | Parameters | F32 Safetensors | Notable configuration features             |
-| ----------------------- | ------------------------------------------ | ---------: | --------------: | ------------------------------------------ |
-| `tinyllama-chat`        | `TinyLlama/TinyLlama-1.1B-Chat-v1.0`       |    303,744 |     1,217,160 B | GQA, query/KV ratio 8                      |
-| `smollm2-instruct`      | `HuggingFaceTB/SmolLM2-135M-Instruct`      |     46,320 |       187,328 B | GQA, query/KV ratio 3                      |
-| `mobilellama-chat`      | `mtgv/MobileLLaMA-1.4B-Chat`               |      9,296 |        39,288 B | MHA                                        |
-| `minicpm5`              | `openbmb/MiniCPM5-1B`                      |  1,409,664 |     5,640,864 B | explicit `head_dim`, GQA, multiple EOS IDs |
-| `deepseek-coder`        | `deepseek-ai/deepseek-coder-6.7b-instruct` |      9,296 |        39,288 B | MHA, linear RoPE scaling                   |
-| `hermes3-llama31`       | `NousResearch/Hermes-3-Llama-3.1-8B`       |     86,336 |       347,488 B | GQA, Llama-3 RoPE scaling                  |
-| `livekit-turn-detector` | `livekit/turn-detector`                    |    132,336 |       531,400 B | explicit `head_dim`, GQA                   |
-All seven source configurations passed the source-side validation performed with Transformers `4.55.0`, and all seven generated F32 models passed construction, prefill, cached decode, save, reload, and post-reload execution checks.
-See `REPORT.md` for the complete shrink results and source revisions.
-## Directory structure
-Each case is self-contained:
-```text
-<case>/
-├── case.json
-├── provenance.json
-├── source-config.json
-├── shrunk-config.json
-├── config-diff.json
-├── validation.json
-├── inputs.safetensors
-│
-├── package/
-│   ├── config.json
-│   ├── generation_config.json
-│   └── model.safetensors
-│
-├── tokenizer/
-│   ├── tokenizer.model
-│   ├── tokenizer_config.json
-│   ├── special_tokens_map.json
-│   ├── vocabulary.json
-│   └── metadata.json
-│
-├── reference/
-│   └── outputs.safetensors
-│
-└── gguf/
-    ├── model-Q4_K_M.gguf
-    ├── metadata.json
-    └── reference.json
-```
-The top-level `manifest.json` provides machine-readable paths, hashes, and sizes for the available model packages and auxiliary tokenizers.
-## Real configurations, synthetic models
-A case has four distinct origins:
-### Source configuration
-`source-config.json` is the configuration obtained from the source Hugging Face repository at the immutable revision recorded in `provenance.json`.
-Its SHA-256 is recorded for traceability.
-### Shrunk configuration
-`shrunk-config.json` is produced by a preservation-first Llama shrinker.
-Only geometry-dependent fields required to make the model small are changed. Unrelated configuration fields are retained rather than normalized into a newly generated configuration.
-The transformations used for this collection include reductions of fields such as:
-* `hidden_size`;
-* `intermediate_size`;
-* `num_hidden_layers`;
-* `num_attention_heads`;
-* `num_key_value_heads`;
-* `vocab_size`;
-* `max_position_embeddings`.
-Out-of-range special-token IDs are remapped only when required by vocabulary reduction.
-Observed fields such as explicit `head_dim`, bias settings, embedding-tying settings, and the original RoPE scaling object are preserved where possible.
-`config-diff.json` records the transformation for each case.
-### Synthetic weights
-The model weights are not derived from the original model checkpoint.
-They are deterministically generated using `tlfloat::LCG64`, with the seed and exact tlfloat revision recorded in `provenance.json`.
-This makes the generated model reproducible without distributing or downloading the original model weights.
-### Synthetic auxiliary tokenizer
-The reduced models use a vocabulary of 128 token IDs, so the original source-model tokenizers are generally not compatible with them.
-Selected cases therefore contain a deterministic synthetic auxiliary tokenizer.
-These tokenizers:
-* cover the reduced token-ID space;
-* preserve the special-token semantics required by the corresponding shrunk configuration;
-* are used consistently for the Hugging Face and GGUF representations;
-* are not derived from the original source tokenizer;
-* do not attempt to reproduce the linguistic tokenization behavior of the original model.
-The tokenizer metadata explicitly records this distinction.
-For numerical model testing, explicit token IDs remain the primary input. Text tokenization is secondary.
-## Safetensors packages
-The `package/` directory is a normal small Hugging Face model package containing the shrunk configuration and deterministic F32 weights.
-Reference inputs are stored separately in `inputs.safetensors`, and expected F32 outputs are stored in `reference/outputs.safetensors`.
-These reference artifacts were generated and validated with Transformers `4.55.0`.
-Exact software versions and hashes are available in each case's `provenance.json`.
-## GGUF Q4_K_M packages
-Every case also contains:
-```text
-gguf/model-Q4_K_M.gguf
-```
-The GGUF files are derived from the same deterministic model weights as the Safetensors packages.
-They were generated with upstream `ggml-org/llama.cpp` at commit:
-```text
-40b740ad05c531b9d57aca6698c3ed553a9e784c
-```
-The conversion pipeline is:
-```text
-F32 Safetensors model
-        ↓
-llama.cpp HF-to-GGUF converter
-        ↓
-temporary F16 GGUF
-        ↓
-llama-quantize Q4_K_M
-        ↓
-retained Q4_K_M GGUF
-```
-The temporary F16 GGUF is hashed for provenance but is not retained.
-`Q4_K_M` is a lossy mixed quantization profile. In very small models, some tensor dimensions are too small for particular K-quant types, so the final GGUF may contain fallback tensor types.
-For this reason, each `gguf/metadata.json` records the actual tensor-type histogram found in the resulting file.
-For example, depending on the case, a `Q4_K_M` artifact may contain a mixture of:
-* F16;
-* F32;
-* Q4_K;
-* Q5_0;
-* Q6_K;
-* Q8_0.
-The label `Q4_K_M` identifies the requested llama.cpp quantization profile; it does not imply that every tensor is stored as Q4_K.
-## GGUF validation
-GGUF numerical validation uses explicit token IDs rather than text tokenization.
-For each case, the final GGUF was loaded through the same pinned llama.cpp revision and tested with deterministic:
-* prefill;
-* autoregressive decode;
-* logit extraction;
-* finite-value checks;
-* repeated-execution checks.
-The effective EOG token set was also checked over the complete reduced token-ID range and required to match the EOS semantics of the shrunk configuration.
-`gguf/reference.json` contains the GGUF reference output, while `gguf/metadata.json` records:
-* llama.cpp revision;
-* source-package SHA-256;
-* tokenizer SHA-256;
-* converter and quantizer invocations;
-* intermediate and final GGUF hashes;
-* actual tensor-type histogram;
-* effective EOG token IDs;
-* direct-token validation results;
-* observed numerical difference from the F32 reference.
-Because `Q4_K_M` is lossy, GGUF logits are not expected to be numerically identical to the F32 Transformers reference.
-The quantization differences are observations rather than exact-equivalence requirements.
-## Reproducibility and provenance
-Each case records enough information to trace:
-```text
-real Hugging Face config + immutable revision
-                    ↓
-         preservation-first shrink
-                    ↓
-        deterministic small config
-                    ↓
-       deterministic synthetic weights
-             ↙              ↘
-   F32 Safetensors       auxiliary tokenizer
-             \              /
-              \            /
-               GGUF conversion
-                     ↓
-                Q4_K_M GGUF
-```
-Relevant identities are represented by SHA-256 hashes rather than filenames or timestamps.
-See:
-* `provenance.json` for the source and F32-generation chain;
-* `tokenizer/metadata.json` for tokenizer provenance;
-* `gguf/metadata.json` for GGUF provenance;
-* `manifest.json` for the dataset-wide package index.
-## What this dataset is not
-This is not:
-* a collection of pretrained language models;
-* a benchmark of model quality;
-* a reproduction of the source models' weights;
-* a reproduction of the source models' tokenizers;
-* a recommendation to use these geometries for training;
-* a comprehensive set of every possible Llama configuration.
-The value of the dataset is the combination of real observed configuration patterns with small, deterministic, executable model artifacts.
-## Source models and licensing
-The source configurations originate from the Hugging Face repositories identified in each case's `provenance.json`.
-The generated model weights and auxiliary tokenizer data are synthetic, but source configuration files and associated metadata may remain subject to terms applicable to their respective source repositories.
-This repository therefore does not claim that a single upstream model license applies uniformly to every source-derived file.
-Users should consult the corresponding source repository when redistributing or using source-derived metadata under conditions where its license is relevant.
-No original source-model checkpoint weights are included.
-## Reports
-* `REPORT.md` — Llama shrinker results and source provenance summary.
-* `GGUF_Q4_K_M_REPORT.json` — consolidated GGUF generation and validation report.
-* `manifest.json` — machine-readable package inventory.
-## Current scope
-The current release contains seven Llama-compatible cases.
-The collection is intentionally small: it favors configurations observed in real, relatively popular model repositories rather than generating large numbers of artificial configuration combinations.
-Future collections may add other model families or additional observed configuration patterns while retaining the same principles of small artifacts, deterministic generation, explicit provenance, and format-native validation.

+---
+pretty_name: Deterministic Random Models
+license: other
+size_categories:
+- n<1K
+tags:
+- llama
+- gemma4
+- qwen3
+- transformers
+- safetensors
+- gguf
+- synthetic
+- conformance
+- compatibility-testing
+---
+# Deterministic Random Models
+This dataset contains nine small, deterministic language-model fixtures for
+model-format, loader, inference, compatibility, and conformance testing. They
+are not trained models and must not be used for language-model quality
+evaluation.
+All weights are synthetic and deterministically generated. No original model
+checkpoint weights are included.
+## Cases
+| Case | Architecture | Parameters | Hugging Face | GGUF | Notable feature |
+|---|---|---:|---|---|---|
+| `tinyllama-chat` | Llama | 303,744 | F32 | Q4_K_M | GQA, query/KV ratio 8 |
+| `smollm2-instruct` | Llama | 46,320 | F32 | Q4_K_M | GQA, query/KV ratio 3 |
+| `mobilellama-chat` | Llama | 9,296 | F32 | Q4_K_M | MHA |
+| `minicpm5` | Llama | 1,409,664 | F32 | Q4_K_M | explicit head dimension, multiple EOS IDs |
+| `deepseek-coder` | Llama | 9,296 | F32 | Q4_K_M | linear RoPE scaling |
+| `hermes3-llama31` | Llama | 86,336 | F32 | Q4_K_M | Llama 3 RoPE scaling |
+| `livekit-turn-detector` | Llama | 132,336 | F32 | Q4_K_M | explicit head dimension, GQA |
+| `gemma4-random-model` | Gemma 4 | 1,519,168 | BF16 | Q4_0 | five-local/one-global attention schedule |
+| `qwen3-random-model` | Qwen 3 | 508,800 | BF16 | Q4_0 | wide Q projection and Q/K head norms |
+The seven Llama cases are derived from real Hugging Face configuration files by
+a preservation-first shrinker. Gemma 4 and Qwen 3 cannot be reduced by selecting
+each field independently, so their tiny geometries preserve architecture-specific
+ratios, tensor inventories, and layer schedules observed in locally downloaded
+upstream GGUF models. Published case names use `random-model` rather than
+`tiny-model` to avoid collision with a separately maintained TinyModel collection.
+## Formats and layout
+The Llama cases retain the original dataset layout:
+```text
+<llama-case>/
+|-- package/model.safetensors       # canonical F32 weights
+|-- gguf/model-Q4_K_M.gguf
+|-- tokenizer/
+|-- reference/outputs.safetensors
+|-- inputs.safetensors
+|-- case.json
+|-- provenance.json
+|-- source-config.json
+|-- shrunk-config.json
+|-- config-diff.json
+`-- validation.json
+```
+The architecture-specific cases use:
+```text
+<random-model-case>/
+|-- hf-bf16/
+|   |-- config.json
+|   |-- model.safetensors
+|   |-- tokenizer.json
+|   `-- tokenizer_config.json
+|-- gguf-q4_0/
+|   |-- <case>-Q4_0.gguf
+|   `-- quantize.log
+|-- reference/
+|   |-- inputs.json
+|   |-- hf-outputs.safetensors
+|   `-- gguf-native.json
+|-- CONFIG_DECISION.md
+`-- metadata.json
+```
+`manifest.json` is the machine-readable index of all nine model packages and
+their SHA-256 hashes and sizes.
+## Synthetic weights and tokenizers
+Weights use the `tlfloat::LCG64` recurrence with multiplier
+`6364136223846793005`, increment `1442695040888963407`, and ten warm-up steps.
+Each case records its seed and provenance.
+The reduced models use deterministic 128-token auxiliary vocabularies. These
+tokenizers cover token IDs `0..127` and preserve each case's special-token
+semantics, but they do not reproduce the linguistic behavior of the original
+tokenizer. Explicit token IDs are the primary numerical-test interface.
+## GGUF generation and validation
+GGUF files were generated with upstream `ggml-org/llama.cpp` commit
+`40b740ad05c531b9d57aca6698c3ed553a9e784c`.
+Every retained GGUF was loaded through that revision and exercised with direct
+token IDs for prefill, cached decode, logit extraction, finite-value checks, and
+repeated-execution checks. The effective EOG token set was checked against the
+model EOS semantics. Per-case metadata records the actual tensor-type histogram,
+hashes, commands, and informational comparison with the corresponding
+Transformers reference.
+Q4_K_M and Q4_0 are lossy formats. Their logits are not required to equal the
+F32 or BF16 reference exactly.
+## Reproducibility and scope
+The Hugging Face weights, configs, and GGUF outputs for Gemma 4 and Qwen 3 were
+independently regenerated and found byte-identical. The Llama cases retain their
+source revisions, source-config hashes, shrink decisions, and generation
+provenance in each case directory.
+This dataset is not a pretrained-model collection, a model-quality benchmark,
+or a reproduction of upstream weights or tokenizers. Source-derived configuration
+and metadata files may remain subject to terms of their respective upstream
+repositories; consult their recorded provenance before redistribution.
+See `REPORT.md`, `GGUF_Q4_K_M_REPORT.json`, and
+`GEMMA4_QWEN3_REPORT.json` for collection-level summaries.

SHA256SUMS ADDED Viewed

	@@ -0,0 +1,164 @@

+a9498453b9883cadce5862a6d8a862949d8a49b3690a5a88673cea1b7b866b42  GEMMA4_QWEN3_REPORT.json
+18ec0cdad98853d7f10381790713f0b3190bd4f24962594bf576e11fdb07f76f  GGUF_Q4_K_M_ADDED_FOUR_REPORT.json
+695036ee0e223d3d79170fb3a0f795eb88f6337fb2e29c18f0b549979c057208  GGUF_Q4_K_M_REPORT.json
+5e8d5eeee84a35177cba72fe8487d04295d2a9c52eb790b16ef5414eabbbcf65  README.md
+95d67a407dc96839250fe23e85e0a5bab9d355ee463b0c87bf9e33164c952b52  REPORT.md
+02130c5ef4fe4e00bf26cc6e0eb288b90916acf827031bf572ffc02a48d5f3f0  deepseek-coder/case.json
+616662945b2382e3c4ef36c05b40ebacea17e2e16fa3a723b7fd49eb5d0e9fdd  deepseek-coder/config-diff.json
+f47850275b039df486445bb70ce51ae0c8b9872793231e02cca07b0dedf6fb8b  deepseek-coder/gguf/metadata.json
+8557d99db9de6b588628f8fead576ff85b26edbdc8df176050b59a28acf720f3  deepseek-coder/gguf/model-Q4_K_M.gguf
+450bd48506eab19d3bc5d94e7bb847ce46d98d8dd54c14a519e4ff2c7f5bf86e  deepseek-coder/gguf/reference.json
+2b49b7cca71e662452eb97b184578af4f59e9a93bc541d621986836cd7e2c395  deepseek-coder/inputs.safetensors
+1385d0151ad8bfe96845e0301e85e01cdab09ba16c11348e838754a2fb9505c9  deepseek-coder/package/config.json
+a524bd3d991c192f2159aea11827c28393355c254a8530667e0fe8aeda3b8eda  deepseek-coder/package/generation_config.json
+cba3bb2298077f3bdeee9da78a4d0ae7586c3a3964e7a40eb2f9e5dc2631c4cc  deepseek-coder/package/model.safetensors
+a09ac290f288f81d96113e30767f088b32c623e26d02d15581d318287d072122  deepseek-coder/provenance.json
+4ceef3c5d1268f4fea0922628dab26f78f0dba84a6cebc1528152ffc75f3da64  deepseek-coder/reference/outputs.safetensors
+1385d0151ad8bfe96845e0301e85e01cdab09ba16c11348e838754a2fb9505c9  deepseek-coder/shrunk-config.json
+a3a2ce1a623b367d3799fb0c1e1ba12d5b8fa51a8e7abba14d10dd3739cd73e4  deepseek-coder/source-config.json
+13c3140dfda642fa7435cb045273d269e0a687626896f21d8cbb42559824ac77  deepseek-coder/tokenizer/metadata.json
+b80f4edf3204563fd78e3c95498aa4ceb9af1770f52869c638d33be9124a43a5  deepseek-coder/tokenizer/special_tokens_map.json
+0f88117a2f8c52a9eabe2d61f7301b6f9b58b7a56644888da4ec9970b1efbcee  deepseek-coder/tokenizer/tokenizer.model
+603a68ffd4df8d0b01492da00fcd7b7386f7a8bbe90e92b81b82516a9231fbfb  deepseek-coder/tokenizer/tokenizer_config.json
+7b1258e1304c5f09f3bc64cd4b569e7ddfa55a062c0581d0e5cb1a3d7a17fcc8  deepseek-coder/tokenizer/vocabulary.json
+aa20ff35795ac323bbdbc81c292f0fa74bb81f9026b88cd01c40c4d81e428d8c  deepseek-coder/validation.json
+399fff213eafc2d462587c9031564b5aafcb99436fc965d9f4e6c8677b6f1837  gemma4-random-model/CONFIG_DECISION.md
+2b631587994acd127d3e6fcc67552c707d6fdba22beada963597387b9e814881  gemma4-random-model/gguf-q4_0/gemma4-random-model-Q4_0.gguf
+5b733665980cfba8e31508035a6088b0e53f9a09c68bd0fcb0f53721bb140f1e  gemma4-random-model/gguf-q4_0/quantize.log
+7386dc7fbb3fc1e83a7db585d088c646183c8c700581bfc12b95390a515c6ba1  gemma4-random-model/hf-bf16/config.json
+f0ce61128449f564ff1dccf316e54c54429f5c5ae9a398803de39d452a671e85  gemma4-random-model/hf-bf16/generation_config.json
+18c8b09759c831fbc9d6a8caf7df8e9d395508d4aec09336167f675527b3a360  gemma4-random-model/hf-bf16/model.safetensors
+8538827f687d1fd46c3de2ba983d0af5e705c0b0661bab1e8f261a9372046d4d  gemma4-random-model/hf-bf16/tokenizer.json
+729e9569f612365498a64b12bf1c2fc8816b8b56cba7be3c5841ab1b0a346a7f  gemma4-random-model/hf-bf16/tokenizer_config.json
+d5f3c821f738e7cea4c8a9abb9d8283f168b23e67cec7528dfaa84f7be48094f  gemma4-random-model/metadata.json
+ceb65d119a54448e1d7ebbb5431284979f33ce309966b3149f667dc4caa24911  gemma4-random-model/reference/gguf-native.json
+ffa41b6c27bab211af7994632041ddedf47c460ad6b89482909d793604831d38  gemma4-random-model/reference/hf-outputs.safetensors
+719a2f5ea966e92e42ebcc43dd73f0714b5d740c9f426d22c942dbb1e7efb6ab  gemma4-random-model/reference/inputs.json
+94a8d82acf2fde73ae5095df8f17c0e2303df4510d0836f3c5c6d65ed51444e5  hermes3-llama31/case.json
+6c0670415f473f79fed118e88160012de23a21df169ed664c2f56be817aa755d  hermes3-llama31/config-diff.json
+dae067e6b71670df248f9d3cec59ad9c16754f470808135e24590ff3369280ab  hermes3-llama31/gguf/metadata.json
+6b53e66bd258178cb1ceb6f89ad36a3b909e912c430ec5ccac06a476d6a652e4  hermes3-llama31/gguf/model-Q4_K_M.gguf
+d1feb8950a9351725c8db93b3d52ef2bae231a3a90f832f6008da4bcc2fd4633  hermes3-llama31/gguf/reference.json
+2b49b7cca71e662452eb97b184578af4f59e9a93bc541d621986836cd7e2c395  hermes3-llama31/inputs.safetensors
+5ddca639398c8e5571a4d8670ea1b111a78c1cdb4342eafb22c68abb1325efcc  hermes3-llama31/package/config.json
+a524bd3d991c192f2159aea11827c28393355c254a8530667e0fe8aeda3b8eda  hermes3-llama31/package/generation_config.json
+22d5f7094e33c1e50eb689b20f56094239f0df30a23b17ef508ce1a6b49c81dc  hermes3-llama31/package/model.safetensors
+6353104f174f21f74e825fe2a2dbf3d409244c759ba4bcb9e41984102ed64224  hermes3-llama31/provenance.json
+97918f673a3659e9cff825ec69b974cb6d8ef2cd8e19da0c4f3da2db0363f69d  hermes3-llama31/reference/outputs.safetensors
+5ddca639398c8e5571a4d8670ea1b111a78c1cdb4342eafb22c68abb1325efcc  hermes3-llama31/shrunk-config.json
+7e189a0da074da6a878d5a276336caf3d62c57138f461bc6239402639f1bc6f5  hermes3-llama31/source-config.json
+13c3140dfda642fa7435cb045273d269e0a687626896f21d8cbb42559824ac77  hermes3-llama31/tokenizer/metadata.json
+b80f4edf3204563fd78e3c95498aa4ceb9af1770f52869c638d33be9124a43a5  hermes3-llama31/tokenizer/special_tokens_map.json
+0f88117a2f8c52a9eabe2d61f7301b6f9b58b7a56644888da4ec9970b1efbcee  hermes3-llama31/tokenizer/tokenizer.model
+603a68ffd4df8d0b01492da00fcd7b7386f7a8bbe90e92b81b82516a9231fbfb  hermes3-llama31/tokenizer/tokenizer_config.json
+7b1258e1304c5f09f3bc64cd4b569e7ddfa55a062c0581d0e5cb1a3d7a17fcc8  hermes3-llama31/tokenizer/vocabulary.json
+0867d4a07507a89dfded679b1950faaefa38b6a7b80140fe5d4862bd5d86c8e0  hermes3-llama31/validation.json
+a4aab6c894dcc48aaace2f6b6e72f164c1e1e405b7bcfc00f1d532490d36dc43  livekit-turn-detector/case.json
+6c480775de9a5f6a2ddb895a1b3654ad0e8c70ce3138da76b10f3f2af036325f  livekit-turn-detector/config-diff.json
+1e6df9143a664dc87031c5b8562cd3e0d31e540f7a85274ac008cd64c7a64294  livekit-turn-detector/gguf/metadata.json
+29643572ec796cace456931e906b8357827a78df07007a937632bc72c4b20d0e  livekit-turn-detector/gguf/model-Q4_K_M.gguf
+9f8ecec5f05d83347e346754a20b202272786b58fc5d8aeb97dac86697d197c7  livekit-turn-detector/gguf/reference.json
+2b49b7cca71e662452eb97b184578af4f59e9a93bc541d621986836cd7e2c395  livekit-turn-detector/inputs.safetensors
+af24bb263d016a531709c1dc49250646cd70795235ec1379466e5f0069530ab2  livekit-turn-detector/package/config.json
+ffcd36a5dfc0a7dc50fcc0ebcdc6ad4b9fb5e34ae88b97ef086464759bbf106f  livekit-turn-detector/package/generation_config.json
+2c696aa8fac754eebf2534453f80e67502dad635d939bf329f765b3d4f19ad2c  livekit-turn-detector/package/model.safetensors
+6d751e12e6cb29421fd316d6f4468335ace6c1207ec8e06886317392ceb72e9c  livekit-turn-detector/provenance.json
+0f1d6be628b1cada9df780c26ee204f723a717e3f2a323081db535b22c847caf  livekit-turn-detector/reference/outputs.safetensors
+af24bb263d016a531709c1dc49250646cd70795235ec1379466e5f0069530ab2  livekit-turn-detector/shrunk-config.json
+e8feb37fa6f974232742d4e223e40928ccc8dabe1906ce97090e14b5753dbcc1  livekit-turn-detector/source-config.json
+08ea5e05ba247548639ef808941a4d8dc73b92e4aed3bed7edf5bf604babe0ba  livekit-turn-detector/tokenizer/metadata.json
+f003d32e36ddc9d7c7a50c186da5c544b2828a0d04ca62aeac741d765727a635  livekit-turn-detector/tokenizer/special_tokens_map.json
+0f88117a2f8c52a9eabe2d61f7301b6f9b58b7a56644888da4ec9970b1efbcee  livekit-turn-detector/tokenizer/tokenizer.model
+564a5f0d3380d83d0ffcfa509a5e5369c0610f586cb6b3d0d45ccaee76610327  livekit-turn-detector/tokenizer/tokenizer_config.json
+7b1258e1304c5f09f3bc64cd4b569e7ddfa55a062c0581d0e5cb1a3d7a17fcc8  livekit-turn-detector/tokenizer/vocabulary.json
+391312ee129392e2b35b6a238593ca5ddff3dd51c6c324135f1bb0114be1e068  livekit-turn-detector/validation.json
+4ef64a4405f5f193022a8ecbc70513ac49278ac6b1ce2f101a98c8bf56f4c9a2  manifest.json
+047c4545232dc055c2c7e4a30a70c22e0e75e9e777e539732128f65c32b0c1cb  minicpm5/case.json
+282a55aae55b59573109fe37222197f298600ad691573bcd84ea5d4dc1afcc0a  minicpm5/config-diff.json
+d088266a01cd2f9a6107877932d6af9f65e6d00202292226eab76af8ea0e889f  minicpm5/gguf/metadata.json
+b9e0e01198ac04fc0239b602007cd4fec8fb93a0e334361c2af92227e4ff8981  minicpm5/gguf/model-Q4_K_M.gguf
+67b086485cab147a3c6c944c9c04099b25c65afa4944c45fcf08aba13581e344  minicpm5/gguf/reference.json
+2b49b7cca71e662452eb97b184578af4f59e9a93bc541d621986836cd7e2c395  minicpm5/inputs.safetensors
+ce27ca4c5e50fdb470c92f0481a63536be162b79241a7dc790008f20dd2698fa  minicpm5/package/config.json
+be5801ef67d4d476194b66a04c174f27c9e9d5a17e9e6ea56b9e3b1fe6b16c2b  minicpm5/package/generation_config.json
+d938e83c12df6357c39660fbc4572b454b87e196acafdcc54699e2f5c25bc5ab  minicpm5/package/model.safetensors
+885f3f3a4892d68a8edc102fcd4a1061809109223afe4818bb0f56f5dbb81d28  minicpm5/provenance.json
+a322c51b7d22e56e983cca2f27eef1b9a42b18709a5886f41777d555d4f8243a  minicpm5/reference/outputs.safetensors
+ce27ca4c5e50fdb470c92f0481a63536be162b79241a7dc790008f20dd2698fa  minicpm5/shrunk-config.json
+6a6509b646cb3169616c5ffc3196e7ccaf9d4d6bc17b266581d241a31c217714  minicpm5/source-config.json
+5522a2a41f78185d34b54729b2a4e679fea4626c84bf44753650936859d88016  minicpm5/tokenizer/metadata.json
+4f35d5d0dfac2e51bf23a35e174ef64df4c6239ecdde93304759c611a031b382  minicpm5/tokenizer/special_tokens_map.json
+517673e2ef9f9b9a667d03914009d1cec25514c1d8a1f8e0eded4d9f44eab94d  minicpm5/tokenizer/tokenizer.model
+476f5f7e5b53aaea02250f7b152a68ae795f561abdeb5db465a1c93d1ba294bf  minicpm5/tokenizer/tokenizer_config.json
+cf5fd07e0df593dc5a43ae647f1f3d1898588a5ec6b5af6ea967cf82be72fd99  minicpm5/tokenizer/vocabulary.json
+ab69d173daccd2f7d4e2634b39f725a1b3a1d3285e7bb1f3ad4976296bec2e05  minicpm5/validation.json
+e3f0830c9882cf8aaf552a20ead0b3a8fb1c5151352d734b61d03c0717c478da  mobilellama-chat/case.json
+f8b26026f91be516da2a729f530530a0f0f585e81e43ea5534c24d75af275e0c  mobilellama-chat/config-diff.json
+741c1e5033fe692b48c4bb2e27ad7fe9ff718561f4af26f836ba0db2aa2b387d  mobilellama-chat/gguf/metadata.json
+c5f51217248a246dccd38037b1f1e3d9f1c0fce977a369fd8a848630259719f9  mobilellama-chat/gguf/model-Q4_K_M.gguf
+55734e6f5e15245b9fa177afe96109bce99b2003dd2776ba185b184a9aedeac6  mobilellama-chat/gguf/reference.json
+2b49b7cca71e662452eb97b184578af4f59e9a93bc541d621986836cd7e2c395  mobilellama-chat/inputs.safetensors
+c4543eacdaf147a644c61f03a9bef6e99b4b3b04812f12e7d07fd5b19aeeb54b  mobilellama-chat/package/config.json
+0550b7d878120fe4cf6217115dcfabada36893981a6c10876de3f9d0c84b6ea5  mobilellama-chat/package/generation_config.json
+cba3bb2298077f3bdeee9da78a4d0ae7586c3a3964e7a40eb2f9e5dc2631c4cc  mobilellama-chat/package/model.safetensors
+cdc4a98af434ba95803504b77e3276d09a9190564ae91cb253741b71716aa817  mobilellama-chat/provenance.json
+404bb1828fe7282768f796774cd7d50bad4804e20b756ddd0d46dfd821e591ff  mobilellama-chat/reference/outputs.safetensors
+c4543eacdaf147a644c61f03a9bef6e99b4b3b04812f12e7d07fd5b19aeeb54b  mobilellama-chat/shrunk-config.json
+1e0f3e66ba7c3824ae2a2352df692c50d0e1b80df3b1fd1338670084de905aaf  mobilellama-chat/source-config.json
+ae54c625121024de84c19c900cb4b969f43aaa838696a7cbfa6467a97a89d65e  mobilellama-chat/tokenizer/metadata.json
+c7869b435de14d46342ee8a81758ecfb18e564629d2a94f9f2a542fee970e16e  mobilellama-chat/tokenizer/special_tokens_map.json
+517673e2ef9f9b9a667d03914009d1cec25514c1d8a1f8e0eded4d9f44eab94d  mobilellama-chat/tokenizer/tokenizer.model
+b9da2129fedf78450a1861f2668553934d96b647b3763c28df748ea44de59017  mobilellama-chat/tokenizer/tokenizer_config.json
+cf5fd07e0df593dc5a43ae647f1f3d1898588a5ec6b5af6ea967cf82be72fd99  mobilellama-chat/tokenizer/vocabulary.json
+25552c5838a9362d8327585337fbe554560f6a50268d20059a5cc6bf9da06e67  mobilellama-chat/validation.json
+d1b96dbe3547f672787456c155a61f7e5416a81faf5e2a8836833da36d8b1690  qwen3-random-model/CONFIG_DECISION.md
+f5bd831677b2b350a581f768a5753ef15a49f0b29b89728692f18e498181e7e5  qwen3-random-model/gguf-q4_0/quantize.log
+6af2e0d230a7919c6b1a02b6462f0441ffbadbb8188b1bd9c0f8174fecd9308f  qwen3-random-model/gguf-q4_0/qwen3-random-model-Q4_0.gguf
+6edb7aa5560d7554cb76e21e54cef67116eff94f0891fd8da905e550c12bef63  qwen3-random-model/hf-bf16/config.json
+2e4002dc0b41024f8f58872187a00d4f113433c5af597ce89114e38c7ac3ab38  qwen3-random-model/hf-bf16/generation_config.json
+0960438e0541dfa529703f5e2b37565aec66b354f6e45bfdf121ecadfcd86dd4  qwen3-random-model/hf-bf16/model.safetensors
+8538827f687d1fd46c3de2ba983d0af5e705c0b0661bab1e8f261a9372046d4d  qwen3-random-model/hf-bf16/tokenizer.json
+729e9569f612365498a64b12bf1c2fc8816b8b56cba7be3c5841ab1b0a346a7f  qwen3-random-model/hf-bf16/tokenizer_config.json
+d7f1008d2d3f6d775fe2af5be0c434bc9b4a4f0a84caa2053489b44da4b54304  qwen3-random-model/metadata.json
+b8a1b0bc02de0bccf136c1783290c59b5a39a9db64bac6528ca07717f7ef6f49  qwen3-random-model/reference/gguf-native.json
+bd0c82ac539035bd28f34e75c09950e377db649c7e6a5d76b99846542d7768e7  qwen3-random-model/reference/hf-outputs.safetensors
+719a2f5ea966e92e42ebcc43dd73f0714b5d740c9f426d22c942dbb1e7efb6ab  qwen3-random-model/reference/inputs.json
+97dd70dd17bbd3a12ab12a69ed003ea333395686b79fe78e490195e18e885428  report.json
+ce49a080a5152505160d2791142206a2937931952d280b74a312417ba279e857  smollm2-instruct/case.json
+979df510e67124e13a58c1247d20c1a68ec9218232cea9b50ba0b10ce5f87ed2  smollm2-instruct/config-diff.json
+cf70ac911c0a6d4b86e264fe33a78b74b04516ed1d5fa3c67d7c22f40901c861  smollm2-instruct/gguf/metadata.json
+8224570d8c01310dc10083e28da8b077503469fe06d701f44b047011065fda43  smollm2-instruct/gguf/model-Q4_K_M.gguf
+63a6c6cce842bb21ce3cb0825814b33ad839af6cd56f375db1d825dfdde219da  smollm2-instruct/gguf/reference.json
+2b49b7cca71e662452eb97b184578af4f59e9a93bc541d621986836cd7e2c395  smollm2-instruct/inputs.safetensors
+37d236b396e6e8912abf08854c58185772270900d402169dd284618e32087f8c  smollm2-instruct/package/config.json
+bb9a9e08898b17e76b4dc8079ff2dfcb93b449144483c1aca61e84294ca5fdbc  smollm2-instruct/package/generation_config.json
+b5fd132288d1a8e8a4c3dd57ff7bc863ba4e3aa7d0c58b8fd3a178f4b4c70f7f  smollm2-instruct/package/model.safetensors
+3dc317b8e1f932b8a2eb632e2bef706e5e096fd958bd7809465a1b98d17e6c22  smollm2-instruct/provenance.json
+d4a932badf1309a3bf52a5711d6b614276f54c7985fcf3510ae6c67f75a16967  smollm2-instruct/reference/outputs.safetensors
+37d236b396e6e8912abf08854c58185772270900d402169dd284618e32087f8c  smollm2-instruct/shrunk-config.json
+8eb740e8bbe4cff95ea7b4588d17a2432deb16e8075bc5828ff7ba9be94d982a  smollm2-instruct/source-config.json
+54703cd111da1a351e83abc6a7ab74ff007558c75092c6d0f856ef63dfe5b29a  smollm2-instruct/tokenizer/metadata.json
+833c7a70faf348f64bc7e62aa070b575205b5acc47201e109ec37a3516e0eba8  smollm2-instruct/tokenizer/special_tokens_map.json
+517673e2ef9f9b9a667d03914009d1cec25514c1d8a1f8e0eded4d9f44eab94d  smollm2-instruct/tokenizer/tokenizer.model
+2932a00a19894e082de22185578d4535c33f8893f4987cdeb913ec95d08f3401  smollm2-instruct/tokenizer/tokenizer_config.json
+cf5fd07e0df593dc5a43ae647f1f3d1898588a5ec6b5af6ea967cf82be72fd99  smollm2-instruct/tokenizer/vocabulary.json
+62f70e96ae4a7b114b522c412c131b2710f85bae174d2630bc4d8879028a7483  smollm2-instruct/validation.json
+8ffefc53e4e83fb5e5cba25e59f8f8246ac38243769e8d3b6266144ef6aa9529  tinyllama-chat/case.json
+30b65ef261fed6560ea3c0073aade016c4285ef1c4a372921d01e4b29bcb2ecf  tinyllama-chat/config-diff.json
+9ec815048e6120c0d82b1e073891801c57ab89ba04dadc835a99b31282169284  tinyllama-chat/gguf/metadata.json
+32bfddeebc736370580c83983cf01380a157c215600a5f1faef2bbb2ffdf9bed  tinyllama-chat/gguf/model-Q4_K_M.gguf
+748c939b4596839dc008051bb6c2de4a9c33162d222a9ca0331384e379a7912e  tinyllama-chat/gguf/reference.json
+2b49b7cca71e662452eb97b184578af4f59e9a93bc541d621986836cd7e2c395  tinyllama-chat/inputs.safetensors
+43282fa7d6b33c370f1a17f431cc744542ba92c647521c1948bebbb58ebbbbd7  tinyllama-chat/package/config.json
+8ec2927388066dace37b43f0a84cbddee5889008d415fdaf925c9cc6eb4cdd8b  tinyllama-chat/package/generation_config.json
+36002fc8be09f37f051cf7fad569658f8f166d0c95cf134f42ba8a59e85f1e95  tinyllama-chat/package/model.safetensors
+5098485919a96ead0aa3772390a5355aac9caef5107d69c490f8d6b6ff509274  tinyllama-chat/provenance.json
+24858401afb8494ae64b576d837e3e9399304e7c27d0f98e742bf6dd99f5a63e  tinyllama-chat/reference/outputs.safetensors
+43282fa7d6b33c370f1a17f431cc744542ba92c647521c1948bebbb58ebbbbd7  tinyllama-chat/shrunk-config.json
+486bedda3a6988332e60d9638a09ca4b260d34ebcf1b19e22cf3b140b63d8fe9  tinyllama-chat/source-config.json
+31557b9dc91fb0900eec0d1ed18676fe6a478829d3a58392f95bb75a4de79fde  tinyllama-chat/tokenizer/metadata.json
+96bdbb8504d9967606e5f661ccc7cbbac44a3661af863a7a58614670a0ccab33  tinyllama-chat/tokenizer/special_tokens_map.json
+517673e2ef9f9b9a667d03914009d1cec25514c1d8a1f8e0eded4d9f44eab94d  tinyllama-chat/tokenizer/tokenizer.model
+431228749bbf143836adf73cbcf31709ad61c448bfce901ae2c8891bc094ba01  tinyllama-chat/tokenizer/tokenizer_config.json
+cf5fd07e0df593dc5a43ae647f1f3d1898588a5ec6b5af6ea967cf82be72fd99  tinyllama-chat/tokenizer/vocabulary.json
+2cc1784cb4776a88fbeabf05327d1cc4607e99aadb446365685450b8b44fe7dc  tinyllama-chat/validation.json

gemma4-random-model/CONFIG_DECISION.md ADDED Viewed

	@@ -0,0 +1,87 @@

+# Gemma 4 tiny fixture configuration decision
+## Status
+The text-only tiny configuration is fixed and generated under
+`artifacts/gemma4-v0/gemma4-random-model` as a BF16 Hugging Face package and a
+Q4_0 GGUF fixture.
+The fixed Hugging Face implementation is Transformers 5.14.1
+`Gemma4ForCausalLM` with `Gemma4TextConfig`. Its constructed trainable parameter
+count is 1,519,168.
+The local source file identifies itself as `general.architecture = gemma4`.
+Repository and revision provenance are not inferable from the local directory and
+remain intentionally unset.
+## Source observations
+The 12B GGUF v3 header uses alignment 32 and contains 667 tensors. Its core
+geometry is:
+- 48 layers arranged as eight repetitions of five sliding-window layers followed
+  by one global layer.
+- Hidden width 3840 and FFN width 15360.
+- 16 query heads.
+- Local layers: 8 KV heads and head width 256.
+- Global layers: 1 KV head and head width 512.
+- Global layers omit a separate `attn_v.weight`; this is part of the schema and
+  must not be filled in by a generic Llama tensor template.
+- Sliding window 1024, context 262144.
+- Global/local RoPE bases 1000000/10000.
+- Global/local RoPE dimensions 512/256.
+- Final logit softcap 30 and RMS epsilon 1e-6.
+- Tied token/output embedding; the official tensor inventory has no independent
+  `output.weight`.
+The separate projector file identifies itself as `general.architecture = clip`
+and `general.type = mmproj`. It is not included in the first text-only fixture.
+## Tiny geometry
+The selected geometry is defined in `configs/gemma4-tiny-v0.json`:
+- 6 layers: one complete five-local/one-global schedule.
+- Hidden width 128 and FFN width 512.
+- 4 query heads.
+- Local layers: 2 KV heads, head width 32.
+- Global layer: 1 KV head, head width 64, no independent V projection.
+- Context 128 and sliding window 64.
+- Vocabulary 128 with PAD/EOS/BOS/UNK/MASK IDs 0/1/2/3/4.
+This is a constrained scale-down rather than independently selected fields. It
+preserves the official layer schedule, the 2:1 global/local head-width ratio, the
+2:1 query/local-KV head ratio, the single global KV head, FFN ratio 4, separate
+RoPE regimes, softcap, tied embeddings, and global shared-KV tensor inventory.
+All matrix dimensions are multiples of 32. This is required so that a direct F32
+GGUF can subsequently be quantized through the pinned Q4_0 path without changing
+model geometry merely to satisfy quantization blocks.
+## Expected GGUF tensor dimensions
+GGUF dimensions are listed in reader order.
+| Role | Local layers 0-4 | Global layer 5 |
+|---|---:|---:|
+| `attn_q.weight` | `[128, 128]` | `[128, 256]` |
+| `attn_k.weight` | `[128, 64]` | `[128, 64]` |
+| `attn_v.weight` | `[128, 64]` | absent |
+| `attn_output.weight` | `[128, 128]` | `[256, 128]` |
+| `attn_q_norm.weight` | `[32]` | `[64]` |
+| `attn_k_norm.weight` | `[32]` | `[64]` |
+Every layer also has FFN gate/up `[128, 512]`, FFN down `[512, 128]`,
+hidden-width norms, and one scalar layer-output scale.
+## Generation gates
+Fixture generation is accepted only when:
+- the direct writer emits `general.architecture = gemma4` and the exact tensor
+  role schedule above;
+- a second generation is byte-identical;
+- the F32 GGUF loads in the pinned llama.cpp revision;
+- direct-token prefill and cached decode are finite and reproducible;
+- the Q4_0 conversion retains the same geometry and loads successfully;
+- F32 outputs are compared with an independently generated matched reference.

gemma4-random-model/gguf-q4_0/gemma4-random-model-Q4_0.gguf ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:2b631587994acd127d3e6fcc67552c707d6fdba22beada963597387b9e814881
+size 877088

gemma4-random-model/gguf-q4_0/quantize.log ADDED Viewed

	@@ -0,0 +1,136 @@

+llama_print_build_info: build = 1 (40b740a)
+llama_print_build_info: built with Clang 21.1.8 for Linux x86_64
+llama_quantize: quantizing '/home/codex/tmp/gemma4-final2-work-20260809/gemma4-random-model-F32.gguf' to 'artifacts/gemma4-v0/gemma4-random-model/gguf-q4_0/gemma4-random-model-Q4_0.gguf' as Q4_0
+llama_model_loader: loaded meta data with 39 key-value pairs and 86 tensors from /home/codex/tmp/gemma4-final2-work-20260809/gemma4-random-model-F32.gguf (version GGUF V3 (latest))
+llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
+llama_model_loader: - kv   0:                       general.architecture str              = gemma4
+llama_model_loader: - kv   1:                               general.type str              = model
+llama_model_loader: - kv   2:                               general.name str              = Tiny Gemma 4 Random
+llama_model_loader: - kv   3:                           general.basename str              = tiny-gemma4-random
+llama_model_loader: - kv   4:                          general.alignment u32              = 32
+llama_model_loader: - kv   5:                         gemma4.block_count u32              = 6
+llama_model_loader: - kv   6:                      gemma4.context_length u32              = 128
+llama_model_loader: - kv   7:                    gemma4.embedding_length u32              = 128
+llama_model_loader: - kv   8:                 gemma4.feed_forward_length u32              = 512
+llama_model_loader: - kv   9:                gemma4.attention.head_count u32              = 4
+llama_model_loader: - kv  10:             gemma4.attention.head_count_kv arr[i32,6]       = [2, 2, 2, 2, 2, 1]
+llama_model_loader: - kv  11:                      gemma4.rope.freq_base f32              = 1000000.000000
+llama_model_loader: - kv  12:                  gemma4.rope.freq_base_swa f32              = 10000.000000
+llama_model_loader: - kv  13:    gemma4.attention.layer_norm_rms_epsilon f32              = 0.000001
+llama_model_loader: - kv  14:                gemma4.attention.key_length u32              = 64
+llama_model_loader: - kv  15:              gemma4.attention.value_length u32              = 64
+llama_model_loader: - kv  16:            gemma4.attention.key_length_swa u32              = 32
+llama_model_loader: - kv  17:          gemma4.attention.value_length_swa u32              = 32
+llama_model_loader: - kv  18:             gemma4.final_logit_softcapping f32              = 30.000000
+llama_model_loader: - kv  19:            gemma4.attention.sliding_window u32              = 64
+llama_model_loader: - kv  20:          gemma4.attention.shared_kv_layers u32              = 0
+llama_model_loader: - kv  21:    gemma4.embedding_length_per_layer_input u32              = 0
+llama_model_loader: - kv  22:    gemma4.attention.sliding_window_pattern arr[bool,6]      = [true, true, true, true, true, false]
+llama_model_loader: - kv  23:                gemma4.rope.dimension_count u32              = 64
+llama_model_loader: - kv  24:            gemma4.rope.dimension_count_swa u32              = 32
+llama_model_loader: - kv  25:                          general.file_type u32              = 0
+llama_model_loader: - kv  26:               general.quantization_version u32              = 2
+llama_model_loader: - kv  27:                       tokenizer.ggml.model str              = gemma4
+llama_model_loader: - kv  28:                      tokenizer.ggml.tokens arr[str,128]     = ["<pad>", "<eos>", "<bos>", "<unk>", ...
+llama_model_loader: - kv  29:                      tokenizer.ggml.scores arr[f32,128]     = [-1000.000000, -1000.000000, -1000.00...
+llama_model_loader: - kv  30:                  tokenizer.ggml.token_type arr[i32,128]     = [3, 3, 3, 3, 3, 1, 1, 1, 1, 1, 1, 1, ...
+llama_model_loader: - kv  31:                      tokenizer.ggml.merges arr[str,1]       = ["a b"]
+llama_model_loader: - kv  32:                tokenizer.ggml.bos_token_id u32              = 2
+llama_model_loader: - kv  33:                tokenizer.ggml.eos_token_id u32              = 1
+llama_model_loader: - kv  34:            tokenizer.ggml.unknown_token_id u32              = 3
+llama_model_loader: - kv  35:            tokenizer.ggml.padding_token_id u32              = 0
+llama_model_loader: - kv  36:               tokenizer.ggml.mask_token_id u32              = 4
+llama_model_loader: - kv  37:               tokenizer.ggml.add_bos_token bool             = true
+llama_model_loader: - kv  38:            tokenizer.ggml.add_space_prefix bool             = false
+llama_model_loader: - type  f32:   86 tensors
+[   1/  86] output_norm.weight                   - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[   2/  86] rope_freqs.weight                    - [    32,      1,      1,      1], type =    f32, size =    0.000 MiB
+[   3/  86] token_embd.weight                    - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[   4/  86] blk.0.attn_k.weight                  - [   128,     64,      1,      1], type =    f32, converting to q4_0 .. size =     0.03 MiB ->     0.00 MiB
+[   5/  86] blk.0.attn_k_norm.weight             - [    32,      1,      1,      1], type =    f32, size =    0.000 MiB
+[   6/  86] blk.0.attn_norm.weight               - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[   7/  86] blk.0.attn_output.weight             - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[   8/  86] blk.0.attn_q.weight                  - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[   9/  86] blk.0.attn_q_norm.weight             - [    32,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  10/  86] blk.0.attn_v.weight                  - [   128,     64,      1,      1], type =    f32, converting to q4_0 .. size =     0.03 MiB ->     0.00 MiB
+[  11/  86] blk.0.ffn_down.weight                - [   512,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  12/  86] blk.0.ffn_gate.weight                - [   128,    512,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  13/  86] blk.0.ffn_norm.weight                - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  14/  86] blk.0.ffn_up.weight                  - [   128,    512,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  15/  86] blk.0.layer_output_scale.weight      - [     1,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  16/  86] blk.0.post_attention_norm.weight     - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  17/  86] blk.0.post_ffw_norm.weight           - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  18/  86] blk.1.attn_k.weight                  - [   128,     64,      1,      1], type =    f32, converting to q4_0 .. size =     0.03 MiB ->     0.00 MiB
+[  19/  86] blk.1.attn_k_norm.weight             - [    32,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  20/  86] blk.1.attn_norm.weight               - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  21/  86] blk.1.attn_output.weight             - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[  22/  86] blk.1.attn_q.weight                  - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[  23/  86] blk.1.attn_q_norm.weight             - [    32,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  24/  86] blk.1.attn_v.weight                  - [   128,     64,      1,      1], type =    f32, converting to q4_0 .. size =     0.03 MiB ->     0.00 MiB
+[  25/  86] blk.1.ffn_down.weight                - [   512,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  26/  86] blk.1.ffn_gate.weight                - [   128,    512,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  27/  86] blk.1.ffn_norm.weight                - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  28/  86] blk.1.ffn_up.weight                  - [   128,    512,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  29/  86] blk.1.layer_output_scale.weight      - [     1,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  30/  86] blk.1.post_attention_norm.weight     - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  31/  86] blk.1.post_ffw_norm.weight           - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  32/  86] blk.2.attn_k.weight                  - [   128,     64,      1,      1], type =    f32, converting to q4_0 .. size =     0.03 MiB ->     0.00 MiB
+[  33/  86] blk.2.attn_k_norm.weight             - [    32,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  34/  86] blk.2.attn_norm.weight               - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  35/  86] blk.2.attn_output.weight             - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[  36/  86] blk.2.attn_q.weight                  - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[  37/  86] blk.2.attn_q_norm.weight             - [    32,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  38/  86] blk.2.attn_v.weight                  - [   128,     64,      1,      1], type =    f32, converting to q4_0 .. size =     0.03 MiB ->     0.00 MiB
+[  39/  86] blk.2.ffn_down.weight                - [   512,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  40/  86] blk.2.ffn_gate.weight                - [   128,    512,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  41/  86] blk.2.ffn_norm.weight                - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  42/  86] blk.2.ffn_up.weight                  - [   128,    512,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  43/  86] blk.2.layer_output_scale.weight      - [     1,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  44/  86] blk.2.post_attention_norm.weight     - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  45/  86] blk.2.post_ffw_norm.weight           - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  46/  86] blk.3.attn_k.weight                  - [   128,     64,      1,      1], type =    f32, converting to q4_0 .. size =     0.03 MiB ->     0.00 MiB
+[  47/  86] blk.3.attn_k_norm.weight             - [    32,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  48/  86] blk.3.attn_norm.weight               - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  49/  86] blk.3.attn_output.weight             - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[  50/  86] blk.3.attn_q.weight                  - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[  51/  86] blk.3.attn_q_norm.weight             - [    32,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  52/  86] blk.3.attn_v.weight                  - [   128,     64,      1,      1], type =    f32, converting to q4_0 .. size =     0.03 MiB ->     0.00 MiB
+[  53/  86] blk.3.ffn_down.weight                - [   512,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  54/  86] blk.3.ffn_gate.weight                - [   128,    512,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  55/  86] blk.3.ffn_norm.weight                - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  56/  86] blk.3.ffn_up.weight                  - [   128,    512,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  57/  86] blk.3.layer_output_scale.weight      - [     1,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  58/  86] blk.3.post_attention_norm.weight     - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  59/  86] blk.3.post_ffw_norm.weight           - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  60/  86] blk.4.attn_k.weight                  - [   128,     64,      1,      1], type =    f32, converting to q4_0 .. size =     0.03 MiB ->     0.00 MiB
+[  61/  86] blk.4.attn_k_norm.weight             - [    32,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  62/  86] blk.4.attn_norm.weight               - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  63/  86] blk.4.attn_output.weight             - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[  64/  86] blk.4.attn_q.weight                  - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[  65/  86] blk.4.attn_q_norm.weight             - [    32,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  66/  86] blk.4.attn_v.weight                  - [   128,     64,      1,      1], type =    f32, converting to q4_0 .. size =     0.03 MiB ->     0.00 MiB
+[  67/  86] blk.4.ffn_down.weight                - [   512,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  68/  86] blk.4.ffn_gate.weight                - [   128,    512,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  69/  86] blk.4.ffn_norm.weight                - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  70/  86] blk.4.ffn_up.weight                  - [   128,    512,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  71/  86] blk.4.layer_output_scale.weight      - [     1,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  72/  86] blk.4.post_attention_norm.weight     - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  73/  86] blk.4.post_ffw_norm.weight           - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  74/  86] blk.5.attn_k.weight                  - [   128,     64,      1,      1], type =    f32, converting to q4_0 .. size =     0.03 MiB ->     0.00 MiB
+[  75/  86] blk.5.attn_k_norm.weight             - [    64,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  76/  86] blk.5.attn_norm.weight               - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  77/  86] blk.5.attn_output.weight             - [   256,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.12 MiB ->     0.02 MiB
+[  78/  86] blk.5.attn_q.weight                  - [   128,    256,      1,      1], type =    f32, converting to q4_0 .. size =     0.12 MiB ->     0.02 MiB
+[  79/  86] blk.5.attn_q_norm.weight             - [    64,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  80/  86] blk.5.ffn_down.weight                - [   512,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  81/  86] blk.5.ffn_gate.weight                - [   128,    512,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  82/  86] blk.5.ffn_norm.weight                - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  83/  86] blk.5.ffn_up.weight                  - [   128,    512,      1,      1], type =    f32, converting to q4_0 .. size =     0.25 MiB ->     0.04 MiB
+[  84/  86] blk.5.layer_output_scale.weight      - [     1,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  85/  86] blk.5.post_attention_norm.weight     - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  86/  86] blk.5.post_ffw_norm.weight           - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+llama_model_quantize_impl: model size  =     5.80 MiB (32.00 BPW)
+llama_model_quantize_impl: quant size  =     0.83 MiB (4.57 BPW)
+llama_quantize: quantize time =    49.59 ms
+llama_quantize:    total time =    49.59 ms

gemma4-random-model/hf-bf16/config.json ADDED Viewed

	@@ -0,0 +1,59 @@

+{
+  "architectures": [
+    "Gemma4ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "attention_k_eq_v": true,
+  "bos_token_id": 2,
+  "dtype": "bfloat16",
+  "enable_moe_block": false,
+  "eos_token_id": 1,
+  "final_logit_softcapping": 30.0,
+  "global_head_dim": 64,
+  "head_dim": 32,
+  "hidden_activation": "gelu_pytorch_tanh",
+  "hidden_size": 128,
+  "hidden_size_per_layer_input": 0,
+  "initializer_range": 0.02,
+  "intermediate_size": 512,
+  "layer_types": [
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 128,
+  "model_type": "gemma4_text",
+  "moe_intermediate_size": null,
+  "num_attention_heads": 4,
+  "num_experts": null,
+  "num_global_key_value_heads": 1,
+  "num_hidden_layers": 6,
+  "num_key_value_heads": 2,
+  "num_kv_shared_layers": 0,
+  "pad_token_id": 0,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "full_attention": {
+      "partial_rotary_factor": 0.25,
+      "rope_theta": 1000000.0,
+      "rope_type": "proportional"
+    },
+    "sliding_attention": {
+      "rope_theta": 10000.0,
+      "rope_type": "default"
+    }
+  },
+  "sliding_window": 64,
+  "tie_word_embeddings": true,
+  "top_k_experts": null,
+  "transformers_version": "5.14.1",
+  "use_bidirectional_attention": null,
+  "use_cache": true,
+  "use_double_wide_mlp": false,
+  "vocab_size": 128,
+  "vocab_size_per_layer_input": 128
+}

gemma4-random-model/hf-bf16/generation_config.json ADDED Viewed

	@@ -0,0 +1,10 @@

+{
+  "_from_model_config": true,
+  "bos_token_id": 2,
+  "eos_token_id": 1,
+  "output_attentions": false,
+  "output_hidden_states": false,
+  "pad_token_id": 0,
+  "transformers_version": "5.14.1",
+  "use_cache": true
+}

gemma4-random-model/hf-bf16/model.safetensors ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:18c8b09759c831fbc9d6a8caf7df8e9d395508d4aec09336167f675527b3a360
+size 3047372

gemma4-random-model/hf-bf16/tokenizer.json ADDED Viewed

	@@ -0,0 +1,217 @@

+{
+  "version": "1.0",
+  "truncation": null,
+  "padding": null,
+  "added_tokens": [
+    {
+      "id": 0,
+      "content": "<pad>",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
+    },
+    {
+      "id": 1,
+      "content": "<eos>",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
+    },
+    {
+      "id": 2,
+      "content": "<bos>",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
+    },
+    {
+      "id": 3,
+      "content": "<unk>",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
+    },
+    {
+      "id": 4,
+      "content": "<mask>",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
+    }
+  ],
+  "normalizer": null,
+  "pre_tokenizer": {
+    "type": "Whitespace"
+  },
+  "post_processor": {
+    "type": "TemplateProcessing",
+    "single": [
+      {
+        "Sequence": {
+          "id": "A",
+          "type_id": 0
+        }
+      }
+    ],
+    "pair": [
+      {
+        "Sequence": {
+          "id": "A",
+          "type_id": 0
+        }
+      },
+      {
+        "Sequence": {
+          "id": "B",
+          "type_id": 1
+        }
+      }
+    ],
+    "special_tokens": {}
+  },
+  "decoder": null,
+  "model": {
+    "type": "WordLevel",
+    "vocab": {
+      "<pad>": 0,
+      "<eos>": 1,
+      "<bos>": 2,
+      "<unk>": 3,
+      "<mask>": 4,
+      "a": 5,
+      "b": 6,
+      "ab": 7,
+      "<t008>": 8,
+      "<t009>": 9,
+      "<t010>": 10,
+      "<t011>": 11,
+      "<t012>": 12,
+      "<t013>": 13,
+      "<t014>": 14,
+      "<t015>": 15,
+      "<t016>": 16,
+      "<t017>": 17,
+      "<t018>": 18,
+      "<t019>": 19,
+      "<t020>": 20,
+      "<t021>": 21,
+      "<t022>": 22,
+      "<t023>": 23,
+      "<t024>": 24,
+      "<t025>": 25,
+      "<t026>": 26,
+      "<t027>": 27,
+      "<t028>": 28,
+      "<t029>": 29,
+      "<t030>": 30,
+      "<t031>": 31,
+      "<t032>": 32,
+      "<t033>": 33,
+      "<t034>": 34,
+      "<t035>": 35,
+      "<t036>": 36,
+      "<t037>": 37,
+      "<t038>": 38,
+      "<t039>": 39,
+      "<t040>": 40,
+      "<t041>": 41,
+      "<t042>": 42,
+      "<t043>": 43,
+      "<t044>": 44,
+      "<t045>": 45,
+      "<t046>": 46,
+      "<t047>": 47,
+      "<t048>": 48,
+      "<t049>": 49,
+      "<t050>": 50,
+      "<t051>": 51,
+      "<t052>": 52,
+      "<t053>": 53,
+      "<t054>": 54,
+      "<t055>": 55,
+      "<t056>": 56,
+      "<t057>": 57,
+      "<t058>": 58,
+      "<t059>": 59,
+      "<t060>": 60,
+      "<t061>": 61,
+      "<t062>": 62,
+      "<t063>": 63,
+      "<t064>": 64,
+      "<t065>": 65,
+      "<t066>": 66,
+      "<t067>": 67,
+      "<t068>": 68,
+      "<t069>": 69,
+      "<t070>": 70,
+      "<t071>": 71,
+      "<t072>": 72,
+      "<t073>": 73,
+      "<t074>": 74,
+      "<t075>": 75,
+      "<t076>": 76,
+      "<t077>": 77,
+      "<t078>": 78,
+      "<t079>": 79,
+      "<t080>": 80,
+      "<t081>": 81,
+      "<t082>": 82,
+      "<t083>": 83,
+      "<t084>": 84,
+      "<t085>": 85,
+      "<t086>": 86,
+      "<t087>": 87,
+      "<t088>": 88,
+      "<t089>": 89,
+      "<t090>": 90,
+      "<t091>": 91,
+      "<t092>": 92,
+      "<t093>": 93,
+      "<t094>": 94,
+      "<t095>": 95,
+      "<t096>": 96,
+      "<t097>": 97,
+      "<t098>": 98,
+      "<t099>": 99,
+      "<t100>": 100,
+      "<t101>": 101,
+      "<t102>": 102,
+      "<t103>": 103,
+      "<t104>": 104,
+      "<t105>": 105,
+      "<t106>": 106,
+      "<t107>": 107,
+      "<t108>": 108,
+      "<t109>": 109,
+      "<t110>": 110,
+      "<t111>": 111,
+      "<t112>": 112,
+      "<t113>": 113,
+      "<t114>": 114,
+      "<t115>": 115,
+      "<t116>": 116,
+      "<t117>": 117,
+      "<t118>": 118,
+      "<t119>": 119,
+      "<t120>": 120,
+      "<t121>": 121,
+      "<t122>": 122,
+      "<t123>": 123,
+      "<t124>": 124,
+      "<t125>": 125,
+      "<t126>": 126,
+      "<t127>": 127
+    },
+    "unk_token": "<unk>"
+  }
+}

gemma4-random-model/hf-bf16/tokenizer_config.json ADDED Viewed

	@@ -0,0 +1,10 @@

+{
+  "backend": "tokenizers",
+  "bos_token": "<bos>",
+  "eos_token": "<eos>",
+  "mask_token": "<mask>",
+  "model_max_length": 128,
+  "pad_token": "<pad>",
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": "<unk>"
+}

gemma4-random-model/metadata.json ADDED Viewed

	@@ -0,0 +1,78 @@

+{
+  "architecture": "gemma4",
+  "gguf": {
+    "comparison_to_hf_bf16": {
+      "decode": {
+        "max_abs_error": 0.15816828608512878,
+        "mean_abs_error": 0.046988509595394135,
+        "top1_agreement": true,
+        "top5_overlap": 3
+      },
+      "policy": "informational because Q4_0 is lossy",
+      "prefill_last": {
+        "max_abs_error": 0.15831683576107025,
+        "mean_abs_error": 0.03859168291091919,
+        "top1_agreement": true,
+        "top5_overlap": 3
+      }
+    },
+    "direct_token_validation": {
+      "all_logits_finite": true,
+      "decode_token_id": 17,
+      "eog_token_ids": [
+        1
+      ],
+      "exactly_reproducible": true,
+      "prefill_token_ids": [
+        5,
+        7,
+        11,
+        13
+      ]
+    },
+    "f32_intermediate_retained": false,
+    "f32_intermediate_sha256": "634ea1838571d13516f8615d5be35cd4d2b35845e3e70445244971c4fa63ff24",
+    "model_sha256": "2b631587994acd127d3e6fcc67552c707d6fdba22beada963597387b9e814881",
+    "model_size_bytes": 877088,
+    "quantization": "Q4_0",
+    "quantizer_command": [
+      "LLAMA_QUANTIZE",
+      "--token-embedding-type",
+      "q4_0",
+      "F32_INPUT",
+      "OUTPUT",
+      "Q4_0"
+    ],
+    "tensor_type_histogram": {
+      "F32": 44,
+      "Q4_0": 42
+    }
+  },
+  "hf": {
+    "config_sha256": "7386dc7fbb3fc1e83a7db585d088c646183c8c700581bfc12b95390a515c6ba1",
+    "construct_prefill_decode_reload": "passed",
+    "dtype": "bfloat16",
+    "model_sha256": "18c8b09759c831fbc9d6a8caf7df8e9d395508d4aec09336167f675527b3a360",
+    "reload_tolerance": {
+      "atol": 0.02,
+      "rtol": 0.02
+    },
+    "tokenizer_json_sha256": "8538827f687d1fd46c3de2ba983d0af5e705c0b0661bab1e8f261a9372046d4d",
+    "transformers_version": "5.14.1"
+  },
+  "parameter_count": 1519168,
+  "reproducibility": {
+    "gguf_q4_0": true,
+    "hf_config": true,
+    "hf_model": true,
+    "second_generation_byte_identical": true
+  },
+  "rng": {
+    "increment": 1442695040888963407,
+    "multiplier": 6364136223846793005,
+    "name": "tlfloat LCG64 equation",
+    "warmup_steps": 10
+  },
+  "schema_version": 1,
+  "seed": 5135595944287666177
+}

gemma4-random-model/reference/gguf-native.json ADDED Viewed

	@@ -0,0 +1,10 @@

+{
+  "schema_version": 1,
+  "vocab_size": 128,
+  "prefill_token_ids": [5,7,11,13],
+  "decode_token_id": 17,
+  "eog_token_ids": [1],
+  "all_logits_finite": true,
+  "prefill_last_logits": [-0.243439302,0.0599161685,-0.0845743716,0.0522050597,0.0797347426,-0.03387703,-0.0461352468,-0.238350868,0.0519622341,0.0766301826,0.126724079,-0.0139343906,-0.0291910302,0.118146136,0.120272741,-0.0258229449,-0.159660712,-0.00222994783,0.00440907292,-0.0304125454,0.0980525315,0.0801480561,-0.00244864542,0.142669708,0.015994221,-0.0381400324,-0.0229879357,-0.0514277704,-0.0138719492,0.273642391,0.218537062,0.126278937,0.070124574,-0.0690698773,0.031898085,-0.0267212726,-0.0291228201,-0.0225784294,-0.239331439,0.167331189,-0.127066836,0.221266046,-0.126179755,0.206429392,-0.143929034,-0.0102543291,0.229470193,-0.0172765926,0.0780417323,-0.0969515741,0.0769872814,0.0613183305,0.0403772257,-0.168282136,0.231201753,0.0205171276,0.0230379496,-0.209012553,-0.0319907963,-0.0245357733,0.00959140435,0.0801036358,-0.0883303657,0.0970800146,-0.115716472,-0.0724429861,-0.124726042,0.124294072,-0.0134264585,0.0358970277,0.227650881,0.00189068564,0.155431747,-0.244728908,0.0816485435,0.0228621587,0.139080092,0.173921034,-0.100639582,-0.0593391173,-0.0474654809,0.0911222324,0.121439792,-0.149972886,-0.208900452,-0.188849151,0.218660593,0.0517382175,-0.197950065,-0.169730186,-0.163051397,0.113981888,0.172181055,-0.00614864333,0.115805477,-0.262914896,-0.132733464,0.0576047525,-0.108596526,-0.162751198,-0.0619868003,0.0415232368,-0.157174632,-0.00733210798,-0.12342795,-0.137874156,-0.101731159,0.0428587683,-0.10437981,0.0370034501,-0.0460586175,0.0158869345,-0.0467884727,0.101396501,-0.235209301,-0.0206604954,-0.0808326676,0.109388076,0.191693529,-0.0848258659,-0.00442003552,-0.0363867059,-0.154983073,-0.014337346,0.274131835,-0.263285935,0.0940570235,0.0207084212],
+  "decode_logits": [-0.00950608216,-0.0892616585,-0.0102732191,-0.0360522121,0.110710457,0.0217158236,0.0268691275,0.11116679,0.062429022,0.14466773,0.0287577808,0.120581679,-0.0712457374,-0.112511128,0.105950512,-0.0633386895,-0.0917795077,-0.0265769325,0.0225345306,-0.25535512,0.167570621,0.112365395,0.0230123922,0.0484296866,-0.228892609,-0.147982344,-0.116228707,-0.0728008151,-0.128445566,-0.0128700966,-0.0407301299,0.113672934,0.128408462,-0.0344361551,0.020843599,-0.107788987,-0.0234028604,0.213245809,-0.0919425413,0.0751381218,-0.253096074,0.262579083,-0.150305837,0.243521363,-0.145598575,-0.0906658098,0.0682541281,0.0127086695,0.0152440453,-0.0776171535,-0.0459475666,-0.00190381811,-0.148405954,-0.173953682,0.126784325,-0.213859275,0.116676956,-0.0121320002,-0.127416849,-0.0586256646,0.0533445515,0.010526645,-0.42138359,0.0762371942,-0.0866874307,0.181853861,-0.14425239,-0.0253368318,-0.152625233,0.0170733649,0.0313455909,0.0770332366,0.0380382352,-0.154644251,0.0201064609,0.0427017771,0.01267628,0.108827092,-0.219167039,-0.0583654344,-0.0495231189,0.0334515274,0.0151894297,0.0707492009,-0.0993220806,-0.126987964,0.0974186882,0.0565137677,-0.355304003,-0.134302884,-0.0124985855,0.104582824,0.145622939,-0.123666488,-0.0822865516,-0.383201808,-0.0574632362,-0.0576772653,-0.117835112,0.0597717986,-0.0567533672,0.122403659,-0.207931876,-0.062526606,-0.132351458,0.157630011,0.0877739787,-0.0715343878,-0.0917522386,-0.135613829,0.0618881881,-0.139158815,-0.0728474706,0.0418399647,0.147925481,0.122347273,-0.109589078,-0.0333428718,0.0310138203,-0.167800277,-0.00797175709,-0.00486898189,-0.0671501011,0.136077702,-0.0265318714,-0.110943288,-0.0150498049,0.147386327]
+}

gemma4-random-model/reference/hf-outputs.safetensors ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:ffa41b6c27bab211af7994632041ddedf47c460ad6b89482909d793604831d38
+size 2720

gemma4-random-model/reference/inputs.json ADDED Viewed

	@@ -0,0 +1,9 @@

+{
+  "decode_token_id": 17,
+  "prefill_token_ids": [
+    5,
+    7,
+    11,
+    13
+  ]
+}

manifest.json CHANGED Viewed

@@ -132,6 +132,46 @@
           "size_bytes": 1217160
         }
       }
     }
   ],
   "schema_version": 1

           "size_bytes": 1217160
         }
       }
+    },
+    {
+      "architecture": "gemma4",
+      "auxiliary_tokenizer": {
+        "path": "gemma4-random-model/hf-bf16/tokenizer.json",
+        "sha256": "8538827f687d1fd46c3de2ba983d0af5e705c0b0661bab1e8f261a9372046d4d"
+      },
+      "case": "gemma4-random-model",
+      "packages": {
+        "gguf_q4_0": {
+          "path": "gemma4-random-model/gguf-q4_0/gemma4-random-model-Q4_0.gguf",
+          "sha256": "2b631587994acd127d3e6fcc67552c707d6fdba22beada963597387b9e814881",
+          "size_bytes": 877088
+        },
+        "safetensors_bf16": {
+          "path": "gemma4-random-model/hf-bf16/model.safetensors",
+          "sha256": "18c8b09759c831fbc9d6a8caf7df8e9d395508d4aec09336167f675527b3a360",
+          "size_bytes": 3047372
+        }
+      }
+    },
+    {
+      "architecture": "qwen3",
+      "auxiliary_tokenizer": {
+        "path": "qwen3-random-model/hf-bf16/tokenizer.json",
+        "sha256": "8538827f687d1fd46c3de2ba983d0af5e705c0b0661bab1e8f261a9372046d4d"
+      },
+      "case": "qwen3-random-model",
+      "packages": {
+        "gguf_q4_0": {
+          "path": "qwen3-random-model/gguf-q4_0/qwen3-random-model-Q4_0.gguf",
+          "sha256": "6af2e0d230a7919c6b1a02b6462f0441ffbadbb8188b1bd9c0f8174fecd9308f",
+          "size_bytes": 294112
+        },
+        "safetensors_bf16": {
+          "path": "qwen3-random-model/hf-bf16/model.safetensors",
+          "sha256": "0960438e0541dfa529703f5e2b37565aec66b354f6e45bfdf121ecadfcd86dd4",
+          "size_bytes": 1020120
+        }
+      }
     }
   ],
   "schema_version": 1

qwen3-random-model/CONFIG_DECISION.md ADDED Viewed

	@@ -0,0 +1,16 @@

+# Qwen3 tiny fixture configuration decision
+The local 0.6B GGUF identifies itself as `general.architecture = qwen3` and
+contains 28 uniform decoder layers. Its characteristic geometry is query width
+twice the hidden width, KV width equal to hidden width, Q/K per-head norms, and an
+FFN ratio of three.
+The tiny fixture preserves those relationships with hidden width 128, four query
+heads, two KV heads, head dimension 64, FFN width 384, two layers, context 128,
+and RoPE base 1000000. All matrix quantization axes are multiples of 32.
+The fixed Hugging Face implementation is Transformers 5.14.1
+`Qwen3ForCausalLM`. The trainable parameter count is 508,800. The generated
+package uses BF16 and tied token/output embeddings. The GGUF is derived from the
+same BF16 values through a temporary F32 serialization and pinned Q4_0
+quantization; the F32 intermediate is not retained.

qwen3-random-model/gguf-q4_0/quantize.log ADDED Viewed

	@@ -0,0 +1,61 @@

+llama_print_build_info: build = 1 (40b740a)
+llama_print_build_info: built with Clang 21.1.8 for Linux x86_64
+llama_quantize: quantizing '/home/codex/tmp/qwen3-random-work-20260810/qwen3-random-model-F32.gguf' to 'artifacts/qwen3-v0/qwen3-random-model/gguf-q4_0/qwen3-random-model-Q4_0.gguf' as Q4_0
+llama_model_loader: loaded meta data with 26 key-value pairs and 24 tensors from /home/codex/tmp/qwen3-random-work-20260810/qwen3-random-model-F32.gguf (version GGUF V3 (latest))
+llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
+llama_model_loader: - kv   0:                       general.architecture str              = qwen3
+llama_model_loader: - kv   1:                               general.type str              = model
+llama_model_loader: - kv   2:                               general.name str              = Tiny Qwen3 Random
+llama_model_loader: - kv   3:                           general.basename str              = tiny-qwen3-random
+llama_model_loader: - kv   4:                          general.alignment u32              = 32
+llama_model_loader: - kv   5:                          qwen3.block_count u32              = 2
+llama_model_loader: - kv   6:                       qwen3.context_length u32              = 128
+llama_model_loader: - kv   7:                     qwen3.embedding_length u32              = 128
+llama_model_loader: - kv   8:                  qwen3.feed_forward_length u32              = 384
+llama_model_loader: - kv   9:                 qwen3.attention.head_count u32              = 4
+llama_model_loader: - kv  10:              qwen3.attention.head_count_kv u32              = 2
+llama_model_loader: - kv  11:                       qwen3.rope.freq_base f32              = 1000000.000000
+llama_model_loader: - kv  12:     qwen3.attention.layer_norm_rms_epsilon f32              = 0.000001
+llama_model_loader: - kv  13:                 qwen3.attention.key_length u32              = 64
+llama_model_loader: - kv  14:               qwen3.attention.value_length u32              = 64
+llama_model_loader: - kv  15:                          general.file_type u32              = 0
+llama_model_loader: - kv  16:               general.quantization_version u32              = 2
+llama_model_loader: - kv  17:                       tokenizer.ggml.model str              = gpt2
+llama_model_loader: - kv  18:                         tokenizer.ggml.pre str              = qwen2
+llama_model_loader: - kv  19:                      tokenizer.ggml.tokens arr[str,128]     = ["<pad>", "<eos>", "<bos>", "<unk>", ...
+llama_model_loader: - kv  20:                  tokenizer.ggml.token_type arr[i32,128]     = [3, 3, 3, 3, 3, 1, 1, 1, 1, 1, 1, 1, ...
+llama_model_loader: - kv  21:                      tokenizer.ggml.merges arr[str,1]       = ["a b"]
+llama_model_loader: - kv  22:                tokenizer.ggml.eos_token_id u32              = 1
+llama_model_loader: - kv  23:            tokenizer.ggml.padding_token_id u32              = 0
+llama_model_loader: - kv  24:            tokenizer.ggml.unknown_token_id u32              = 3
+llama_model_loader: - kv  25:               tokenizer.ggml.add_bos_token bool             = false
+llama_model_loader: - type  f32:   24 tensors
+[   1/  24] output_norm.weight                   - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[   2/  24] token_embd.weight                    - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[   3/  24] blk.0.attn_k.weight                  - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[   4/  24] blk.0.attn_k_norm.weight             - [    64,      1,      1,      1], type =    f32, size =    0.000 MiB
+[   5/  24] blk.0.attn_norm.weight               - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[   6/  24] blk.0.attn_output.weight             - [   256,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.12 MiB ->     0.02 MiB
+[   7/  24] blk.0.attn_q.weight                  - [   128,    256,      1,      1], type =    f32, converting to q4_0 .. size =     0.12 MiB ->     0.02 MiB
+[   8/  24] blk.0.attn_q_norm.weight             - [    64,      1,      1,      1], type =    f32, size =    0.000 MiB
+[   9/  24] blk.0.attn_v.weight                  - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[  10/  24] blk.0.ffn_down.weight                - [   384,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.19 MiB ->     0.03 MiB
+[  11/  24] blk.0.ffn_gate.weight                - [   128,    384,      1,      1], type =    f32, converting to q4_0 .. size =     0.19 MiB ->     0.03 MiB
+[  12/  24] blk.0.ffn_norm.weight                - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  13/  24] blk.0.ffn_up.weight                  - [   128,    384,      1,      1], type =    f32, converting to q4_0 .. size =     0.19 MiB ->     0.03 MiB
+[  14/  24] blk.1.attn_k.weight                  - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[  15/  24] blk.1.attn_k_norm.weight             - [    64,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  16/  24] blk.1.attn_norm.weight               - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  17/  24] blk.1.attn_output.weight             - [   256,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.12 MiB ->     0.02 MiB
+[  18/  24] blk.1.attn_q.weight                  - [   128,    256,      1,      1], type =    f32, converting to q4_0 .. size =     0.12 MiB ->     0.02 MiB
+[  19/  24] blk.1.attn_q_norm.weight             - [    64,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  20/  24] blk.1.attn_v.weight                  - [   128,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.06 MiB ->     0.01 MiB
+[  21/  24] blk.1.ffn_down.weight                - [   384,    128,      1,      1], type =    f32, converting to q4_0 .. size =     0.19 MiB ->     0.03 MiB
+[  22/  24] blk.1.ffn_gate.weight                - [   128,    384,      1,      1], type =    f32, converting to q4_0 .. size =     0.19 MiB ->     0.03 MiB
+[  23/  24] blk.1.ffn_norm.weight                - [   128,      1,      1,      1], type =    f32, size =    0.000 MiB
+[  24/  24] blk.1.ffn_up.weight                  - [   128,    384,      1,      1], type =    f32, converting to q4_0 .. size =     0.19 MiB ->     0.03 MiB
+llama_model_quantize_impl: model size  =     1.94 MiB (32.00 BPW)
+llama_model_quantize_impl: quant size  =     0.28 MiB (4.55 BPW)
+llama_quantize: quantize time =     3.12 ms
+llama_quantize:    total time =     3.12 ms

qwen3-random-model/gguf-q4_0/qwen3-random-model-Q4_0.gguf ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:6af2e0d230a7919c6b1a02b6462f0441ffbadbb8188b1bd9c0f8174fecd9308f
+size 294112

qwen3-random-model/hf-bf16/config.json ADDED Viewed

	@@ -0,0 +1,37 @@

+{
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": null,
+  "dtype": "bfloat16",
+  "eos_token_id": 1,
+  "head_dim": 64,
+  "hidden_act": "silu",
+  "hidden_size": 128,
+  "initializer_range": 0.02,
+  "intermediate_size": 384,
+  "layer_types": [
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 128,
+  "max_window_layers": 28,
+  "model_type": "qwen3",
+  "num_attention_heads": 4,
+  "num_hidden_layers": 2,
+  "num_key_value_heads": 2,
+  "pad_token_id": 0,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "rope_theta": 1000000.0,
+    "rope_type": "default"
+  },
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.14.1",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 128
+}

qwen3-random-model/hf-bf16/generation_config.json ADDED Viewed

	@@ -0,0 +1,9 @@

+{
+  "_from_model_config": true,
+  "eos_token_id": 1,
+  "output_attentions": false,
+  "output_hidden_states": false,
+  "pad_token_id": 0,
+  "transformers_version": "5.14.1",
+  "use_cache": true
+}

qwen3-random-model/hf-bf16/model.safetensors ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:0960438e0541dfa529703f5e2b37565aec66b354f6e45bfdf121ecadfcd86dd4
+size 1020120

qwen3-random-model/hf-bf16/tokenizer.json ADDED Viewed

	@@ -0,0 +1,217 @@

+{
+  "version": "1.0",
+  "truncation": null,
+  "padding": null,
+  "added_tokens": [
+    {
+      "id": 0,
+      "content": "<pad>",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
+    },
+    {
+      "id": 1,
+      "content": "<eos>",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
+    },
+    {
+      "id": 2,
+      "content": "<bos>",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
+    },
+    {
+      "id": 3,
+      "content": "<unk>",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
+    },
+    {
+      "id": 4,
+      "content": "<mask>",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
+    }
+  ],
+  "normalizer": null,
+  "pre_tokenizer": {
+    "type": "Whitespace"
+  },
+  "post_processor": {
+    "type": "TemplateProcessing",
+    "single": [
+      {
+        "Sequence": {
+          "id": "A",
+          "type_id": 0
+        }
+      }
+    ],
+    "pair": [
+      {
+        "Sequence": {
+          "id": "A",
+          "type_id": 0
+        }
+      },
+      {
+        "Sequence": {
+          "id": "B",
+          "type_id": 1
+        }
+      }
+    ],
+    "special_tokens": {}
+  },
+  "decoder": null,
+  "model": {
+    "type": "WordLevel",
+    "vocab": {
+      "<pad>": 0,
+      "<eos>": 1,
+      "<bos>": 2,
+      "<unk>": 3,
+      "<mask>": 4,
+      "a": 5,
+      "b": 6,
+      "ab": 7,
+      "<t008>": 8,
+      "<t009>": 9,
+      "<t010>": 10,
+      "<t011>": 11,
+      "<t012>": 12,
+      "<t013>": 13,
+      "<t014>": 14,
+      "<t015>": 15,
+      "<t016>": 16,
+      "<t017>": 17,
+      "<t018>": 18,
+      "<t019>": 19,
+      "<t020>": 20,
+      "<t021>": 21,
+      "<t022>": 22,
+      "<t023>": 23,
+      "<t024>": 24,
+      "<t025>": 25,
+      "<t026>": 26,
+      "<t027>": 27,
+      "<t028>": 28,
+      "<t029>": 29,
+      "<t030>": 30,
+      "<t031>": 31,
+      "<t032>": 32,
+      "<t033>": 33,
+      "<t034>": 34,
+      "<t035>": 35,
+      "<t036>": 36,
+      "<t037>": 37,
+      "<t038>": 38,
+      "<t039>": 39,
+      "<t040>": 40,
+      "<t041>": 41,
+      "<t042>": 42,
+      "<t043>": 43,
+      "<t044>": 44,
+      "<t045>": 45,
+      "<t046>": 46,
+      "<t047>": 47,
+      "<t048>": 48,
+      "<t049>": 49,
+      "<t050>": 50,
+      "<t051>": 51,
+      "<t052>": 52,
+      "<t053>": 53,
+      "<t054>": 54,
+      "<t055>": 55,
+      "<t056>": 56,
+      "<t057>": 57,
+      "<t058>": 58,
+      "<t059>": 59,
+      "<t060>": 60,
+      "<t061>": 61,
+      "<t062>": 62,
+      "<t063>": 63,
+      "<t064>": 64,
+      "<t065>": 65,
+      "<t066>": 66,
+      "<t067>": 67,
+      "<t068>": 68,
+      "<t069>": 69,
+      "<t070>": 70,
+      "<t071>": 71,
+      "<t072>": 72,
+      "<t073>": 73,
+      "<t074>": 74,
+      "<t075>": 75,
+      "<t076>": 76,
+      "<t077>": 77,
+      "<t078>": 78,
+      "<t079>": 79,
+      "<t080>": 80,
+      "<t081>": 81,
+      "<t082>": 82,
+      "<t083>": 83,
+      "<t084>": 84,
+      "<t085>": 85,
+      "<t086>": 86,
+      "<t087>": 87,
+      "<t088>": 88,
+      "<t089>": 89,
+      "<t090>": 90,
+      "<t091>": 91,
+      "<t092>": 92,
+      "<t093>": 93,
+      "<t094>": 94,
+      "<t095>": 95,
+      "<t096>": 96,
+      "<t097>": 97,
+      "<t098>": 98,
+      "<t099>": 99,
+      "<t100>": 100,
+      "<t101>": 101,
+      "<t102>": 102,
+      "<t103>": 103,
+      "<t104>": 104,
+      "<t105>": 105,
+      "<t106>": 106,
+      "<t107>": 107,
+      "<t108>": 108,
+      "<t109>": 109,
+      "<t110>": 110,
+      "<t111>": 111,
+      "<t112>": 112,
+      "<t113>": 113,
+      "<t114>": 114,
+      "<t115>": 115,
+      "<t116>": 116,
+      "<t117>": 117,
+      "<t118>": 118,
+      "<t119>": 119,
+      "<t120>": 120,
+      "<t121>": 121,
+      "<t122>": 122,
+      "<t123>": 123,
+      "<t124>": 124,
+      "<t125>": 125,
+      "<t126>": 126,
+      "<t127>": 127
+    },
+    "unk_token": "<unk>"
+  }
+}

qwen3-random-model/hf-bf16/tokenizer_config.json ADDED Viewed

	@@ -0,0 +1,10 @@

+{
+  "backend": "tokenizers",
+  "bos_token": "<bos>",
+  "eos_token": "<eos>",
+  "mask_token": "<mask>",
+  "model_max_length": 128,
+  "pad_token": "<pad>",
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": "<unk>"
+}

qwen3-random-model/metadata.json ADDED Viewed

	@@ -0,0 +1,79 @@

+{
+  "architecture": "qwen3",
+  "gguf": {
+    "comparison_to_hf_bf16": {
+      "decode_logits": {
+        "max_absolute_error": 0.04789973795413971,
+        "mean_absolute_error": 0.01485803909599781,
+        "top1_agreement": true,
+        "top5_overlap": 4
+      },
+      "policy": "informational; quantized and BF16 executions are not required to be byte-identical",
+      "prefill_last_logits": {
+        "max_absolute_error": 0.062290437519550323,
+        "mean_absolute_error": 0.017863117158412933,
+        "top1_agreement": true,
+        "top5_overlap": 5
+      }
+    },
+    "direct_token_validation": {
+      "all_logits_finite": true,
+      "decode_token_id": 17,
+      "eog_token_ids": [
+        1
+      ],
+      "prefill_token_ids": [
+        5,
+        7,
+        11,
+        13
+      ],
+      "repeated_output_byte_identical": true,
+      "status": "passed"
+    },
+    "f32_intermediate_retained": false,
+    "f32_intermediate_sha256": "2dec29125ee372738404f1fe272aee7b98015c28925be8466a5a3574ff4ea51c",
+    "model_sha256": "6af2e0d230a7919c6b1a02b6462f0441ffbadbb8188b1bd9c0f8174fecd9308f",
+    "model_size_bytes": 294112,
+    "quantization": "Q4_0",
+    "quantizer_command": [
+      "LLAMA_QUANTIZE",
+      "--token-embedding-type",
+      "q4_0",
+      "F32_INPUT",
+      "OUTPUT",
+      "Q4_0"
+    ],
+    "tensor_type_histogram": {
+      "F32": 9,
+      "Q4_0": 15
+    }
+  },
+  "hf": {
+    "config_sha256": "6edb7aa5560d7554cb76e21e54cef67116eff94f0891fd8da905e550c12bef63",
+    "construct_prefill_decode_reload": "passed",
+    "dtype": "bfloat16",
+    "model_sha256": "0960438e0541dfa529703f5e2b37565aec66b354f6e45bfdf121ecadfcd86dd4",
+    "reload_tolerance": {
+      "atol": 0.02,
+      "rtol": 0.02
+    },
+    "tokenizer_json_sha256": "8538827f687d1fd46c3de2ba983d0af5e705c0b0661bab1e8f261a9372046d4d",
+    "transformers_version": "5.14.1"
+  },
+  "parameter_count": 508800,
+  "reproducibility": {
+    "gguf_q4_0_byte_identical": true,
+    "hf_config_byte_identical": true,
+    "hf_model_byte_identical": true,
+    "native_repeated_output_byte_identical": true
+  },
+  "rng": {
+    "increment": 1442695040888963407,
+    "multiplier": 6364136223846793005,
+    "name": "tlfloat LCG64 equation",
+    "warmup_steps": 10
+  },
+  "schema_version": 1,
+  "seed": 5861229642211393537
+}

qwen3-random-model/reference/gguf-native.json ADDED Viewed

	@@ -0,0 +1,10 @@

+{
+  "schema_version": 1,
+  "vocab_size": 128,
+  "prefill_token_ids": [5,7,11,13],
+  "decode_token_id": 17,
+  "eog_token_ids": [1],
+  "all_logits_finite": true,
+  "prefill_last_logits": [-0.059569791,-0.00279867183,-0.129129365,-0.0497001708,-0.0163497981,0.0050463574,-0.118148349,-0.119329594,0.0471271574,0.165694386,0.121689022,0.0868684873,-0.285888314,0.464197576,0.0721694306,-0.0828261524,-0.00478410348,0.0489596315,-0.0185307804,0.131819114,0.150955066,-0.00900212489,-0.112120651,-0.0487837568,0.215351298,-0.0394815058,-0.246770024,0.123624079,0.440773666,-0.0608280487,-0.188993871,-0.0776931942,0.191938415,0.0849059895,0.161211357,0.209928721,0.19375211,0.0377947018,-0.0473469868,-0.0526904091,-0.342998058,0.00151638919,-0.281735033,-0.325757653,-0.0873417258,-0.0726597607,-0.0898307115,0.0961848348,0.0175519064,0.19467853,0.144317746,-0.0446694307,0.194096655,-0.0906346738,-0.00140182674,-0.248219967,-0.0332248844,-0.0761152059,-0.00496444013,0.0411773697,-0.100529052,0.184473619,0.12824671,-0.154807985,0.203380048,0.000879283529,0.0606221855,0.0596247315,-0.0945963487,0.101349384,-0.110620856,0.0342999995,0.260351092,0.122167736,-0.0360923372,0.173424855,-0.0171568785,0.01655237,0.0413330048,-0.0247170739,0.0867993608,-0.283120692,0.0356137976,0.131538391,-0.0686923563,-0.108378917,-0.234724298,0.0432123467,-0.0833863765,-0.136126012,-0.0844061077,0.0185204614,-0.106897838,-0.0382344723,0.0735053346,0.192892328,-0.072386384,0.0755672604,0.227094755,0.319558173,-0.0431868248,0.177255318,0.047903154,-0.0529437549,-0.0846508369,0.023263108,0.092685841,0.12609227,-0.0929150283,-0.0202170853,-0.0485494062,0.130493596,0.112307258,-0.0679365769,0.31350714,0.0275667477,0.412057728,-0.150501117,-0.0235658735,0.0479809642,0.120794863,-0.0657523423,-0.00684578717,0.128233552,-0.299440444,-0.0540933758,-0.0660768449,0.178839311],
+  "decode_logits": [-0.128728911,-0.0680600926,-0.0213499181,-0.0814930722,-0.0304954946,0.076275982,-0.0558523759,-0.193083286,-0.00124091003,0.180128112,-0.0235419907,-0.0499896929,-0.146302819,-0.0829787552,0.0822936594,-0.319333464,0.087422736,0.640825927,-0.0770442486,-0.00533498824,0.198489577,-0.0673802495,-0.044373773,-0.0899470448,0.22747986,-0.165194958,-0.206123427,0.184741259,0.430428028,0.201801389,-0.096214503,-0.182015613,0.113418564,-0.102462783,0.000823613489,0.201972425,0.286742687,0.00922726374,0.0207918584,0.139977247,-0.154677019,0.0959414467,-0.200118259,-0.286479235,-0.0979933813,-0.0308089741,-0.0729282647,0.0911978558,0.0227652844,0.0366883315,0.105987392,0.186607987,0.213508934,-0.0158322453,0.0286238045,-0.192842454,0.0747692138,-0.0121877929,-0.0764493123,0.0502483658,-0.00815661065,0.173016042,0.172006324,0.0342967473,0.21316421,0.00920947641,0.0562966801,-0.0439821929,-0.116201296,0.0822032988,-0.066759266,-0.0262424406,-0.0112875737,-0.043456845,-0.0415013656,0.271588266,-0.0526413396,-0.0800668597,-0.0901914164,0.0734161511,0.134245142,-0.195974514,-0.138575435,0.0581642762,-0.0191159453,-0.194816291,-0.104702979,0.0591645055,-0.14727436,-0.000515548047,0.0253087468,-0.0757197812,-0.0797108263,-0.217172638,-0.0814730152,0.19474645,-0.118938178,0.216702476,0.301452458,0.150767744,-0.178662762,0.103499614,0.164770931,0.0262989514,-0.0365286097,-0.0491542518,-0.0496671125,-0.0759799853,0.000831546262,-0.0431131274,-0.106817111,0.0936398432,0.0876262784,-0.0671788529,0.196318924,0.0209422112,0.214178532,-0.168416709,-0.0207815226,-0.00951737165,0.339859873,0.0654355064,-0.0690690279,0.105128914,-0.37723881,0.104125872,-0.151014224,0.0857063606]
+}

qwen3-random-model/reference/hf-outputs.safetensors ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:bd0c82ac539035bd28f34e75c09950e377db649c7e6a5d76b99846542d7768e7
+size 2720

qwen3-random-model/reference/inputs.json ADDED Viewed

	@@ -0,0 +1,9 @@

+{
+  "decode_token_id": 17,
+  "prefill_token_ids": [
+    5,
+    7,
+    11,
+    13
+  ]
+}