Publish current 5M and 209M GGUF artifacts with validation limits

.gitattributes

@@ -33,3 +33,12 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+209M/timestral-209m-f16.gguf filter=lfs diff=lfs merge=lfs -text
+209M/timestral-209m-f32.gguf filter=lfs diff=lfs merge=lfs -text
+209M/timestral-209m-q4_k.gguf filter=lfs diff=lfs merge=lfs -text
+209M/timestral-209m-q5_k.gguf filter=lfs diff=lfs merge=lfs -text
+209M/timestral-209m-q6_k.gguf filter=lfs diff=lfs merge=lfs -text
+209M/timestral-209m-q8_0.gguf filter=lfs diff=lfs merge=lfs -text
+5M/timestral-5m-f16.gguf filter=lfs diff=lfs merge=lfs -text
+5M/timestral-5m-f32.gguf filter=lfs diff=lfs merge=lfs -text
+5M/timestral-5m-q8_0.gguf filter=lfs diff=lfs merge=lfs -text

209M/LICENSE

@@ -0,0 +1,10 @@
+Modified MIT License 
+
+Attribution notice: 2026 - Mistral AI
+
+Permission is hereby granted, free of charge, to any person obtaining a copy of the weights of this model and associated documentation files (the “Model”), to deal in the Model without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Model, and to permit persons to whom the Model is furnished to do so, subject to the following conditions: 
+
+1. The above attribution notice and this permission notice shall be included in all copies or substantial portions of the Model. 
+2. You are not authorized to exercise any rights under this license if the global consolidated monthly revenue of your company (or that of your employer) exceeds $20 million (or its equivalent in another currency) for the preceding month. This restriction in (b) applies to the Model and any derivatives, modifications, or combined works based on it, whether provided by Mistral AI or by a third party. You may contact Mistral AI (sales@mistral.ai) to request a commercial license, which Mistral AI may grant you at its sole discretion, or choose to use the Model on Mistral AI's hosted services available at https://mistral.ai/.
+
+THE MODEL IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL MISTRAL AI BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE MODEL OR THE USE OR OTHER DEALINGS IN THE MODEL.

209M/timestral-209m-f16.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24242f55b330a3794649700f6fadd7e59de320aabb42fe377d8df4ab811fb94f
+size 420310336

209M/timestral-209m-f32.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf7605545245a2dd4d32f7ac2f7bbe5c2a12d8f47370814a254862150255b8e8
+size 838347872

209M/timestral-209m-q4_k.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05b9a0c99c099ba23ecb0a8dabf1552a6f0fbd0b7e37df0163fe41c851d575a6
+size 119846112

209M/timestral-209m-q5_k.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e894e488baf17ba07a8dc20c685506a5a7bb89ca8ddb82e920cb2113401d1fef
+size 145973472

209M/timestral-209m-q6_k.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49f364c750906c2a0513a407af5c94cf16a6d711976ce803ef966b17a0f6fe4d
+size 173733792

209M/timestral-209m-q8_0.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bcf4b8a0b4042e06c3f26dd611b05e0a0bff7de5b7c7869aa0757f58383b479
+size 224355136

5M/LICENSE

@@ -0,0 +1,10 @@
+Modified MIT License 
+
+Attribution notice: 2026 - Mistral AI
+
+Permission is hereby granted, free of charge, to any person obtaining a copy of the weights of this model and associated documentation files (the “Model”), to deal in the Model without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Model, and to permit persons to whom the Model is furnished to do so, subject to the following conditions: 
+
+1. The above attribution notice and this permission notice shall be included in all copies or substantial portions of the Model. 
+2. You are not authorized to exercise any rights under this license if the global consolidated monthly revenue of your company (or that of your employer) exceeds $20 million (or its equivalent in another currency) for the preceding month. This restriction in (b) applies to the Model and any derivatives, modifications, or combined works based on it, whether provided by Mistral AI or by a third party. You may contact Mistral AI (sales@mistral.ai) to request a commercial license, which Mistral AI may grant you at its sole discretion, or choose to use the Model on Mistral AI's hosted services available at https://mistral.ai/.
+
+THE MODEL IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL MISTRAL AI BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE MODEL OR THE USE OR OTHER DEALINGS IN THE MODEL.

5M/timestral-5m-f16.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:933e2e518300e97e7799deb1920dd4e539da89a90c743b663cda772ee79d4a38
+size 11641824

5M/timestral-5m-f32.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c320276d8db95fb72fe5f4e043147e29142a1bf197c1b45daf6736780f6736ed
+size 21926656

5M/timestral-5m-q8_0.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5781a8c8e5c6d5ccadfba91c96dda92cea4dc84778eef0b8c3590535cd72a4c
+size 6820704

LICENSE

@@ -0,0 +1,10 @@
+Modified MIT License 
+
+Attribution notice: 2026 - Mistral AI
+
+Permission is hereby granted, free of charge, to any person obtaining a copy of the weights of this model and associated documentation files (the “Model”), to deal in the Model without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Model, and to permit persons to whom the Model is furnished to do so, subject to the following conditions: 
+
+1. The above attribution notice and this permission notice shall be included in all copies or substantial portions of the Model. 
+2. You are not authorized to exercise any rights under this license if the global consolidated monthly revenue of your company (or that of your employer) exceeds $20 million (or its equivalent in another currency) for the preceding month. This restriction in (b) applies to the Model and any derivatives, modifications, or combined works based on it, whether provided by Mistral AI or by a third party. You may contact Mistral AI (sales@mistral.ai) to request a commercial license, which Mistral AI may grant you at its sole discretion, or choose to use the Model on Mistral AI's hosted services available at https://mistral.ai/.
+
+THE MODEL IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL MISTRAL AI BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE MODEL OR THE USE OR OTHER DEALINGS IN THE MODEL.

README.md

@@ -0,0 +1,75 @@
+---
+license: other
+license_name: mistral-modified-mit
+license_link: LICENSE
+base_model:
+- mistralai/Timestral-5M
+- mistralai/Timestral-209M
+library_name: timestral.cpp
+tags:
+- gguf
+- time-series
+- forecasting
+- timestral
+---
+# Timestral GGUF models for timestral.cpp
+
+This interim release contains nine GGUF files for native CPU forecasting with [timestral.cpp](https://github.com/mudler/timestral.cpp).
+Inference requires no Python or libtorch. Conversion uses the pinned upstream checkpoints listed in `manifest.json`.
+
+## Available models
+
+| Directory | Available formats | Validation |
+| --- | --- | --- |
+| `5M/` | F32, F16, Q8_0 | All three pass storage, schema, lineage, and byte-encoding audits. F32 passes trained parity; reduced-format forecast quality is not yet evaluated. |
+| `209M/` | F32, F16, Q8_0, Q6_K, Q5_K, Q4_K | All six pass cohort/storage admission. Full quality evaluation completes, but the Q8 mean-deviation gate fails. |
+
+**28M and 123M GGUFs are not included.** Their upstream checkpoints are acquired; conversion and native qualification remain pending.
+The intended series includes all four models. This release does not claim that scope is complete.
+
+## Quality and performance limitations
+
+The qualified optional standalone MKL profile uses runtime `68a5a53` and the documented SDK version.
+Results do not establish parity or quality for every backend, SDK, or machine.
+No proprietary SDK libraries are included here.
+
+The 209M quality evaluation covers 79 items, six formats, 474 native calls, and 79 fresh F32 anchor pairs.
+All formats have complete coverage. Immutable closing checks pass.
+
+| 209M format | Aggregate mean normalized deviation from F32 | Absolute aggregate WQL change |
+| --- | ---: | ---: |
+| F32 | 0 | 0% |
+| F16 | 0.00015529176950670028 | 0.00274317535567626% |
+| Q8_0 | 0.005657737466196431 | 0.04116458747472307% |
+| Q6_K | 0.013512040093425381 | 0.06180385023158404% |
+| Q5_K | 0.02008887882387394 | 1.3580854256823647% |
+| Q4_K | 0.033048053192290705 | 2.770376555053301% |
+
+F16 passes its mean-deviation bound of 0.001 and WQL-change bound of 0.5%.
+**Q8_0 fails its mean-deviation bound of 0.005.** Its WQL change passes the 1% bound.
+WQL changes above are already percentages. Normalized deviation is not raw percentage forecast error.
+Strict mean-error ordering passes. K formats have no separate loss bound in this contract.
+MASE degeneracies remain reported, not excluded.
+
+**Formal latency acceptance fails at both tested thread counts.** Memory bounds pass.
+Review and optimization remain pending. This interim publication does not waive any acceptance gate.
+
+See the [complete quality report](https://github.com/mudler/timestral.cpp/blob/a4a1f95bb993a364c59d66edea17dafb2f17522a/docs/quantization-209m-result.md),
+[F32 replay](https://github.com/mudler/timestral.cpp/blob/a4a1f95bb993a364c59d66edea17dafb2f17522a/docs/coherent-f32-replay.md),
+and [runtime documentation](https://github.com/mudler/timestral.cpp/blob/a4a1f95bb993a364c59d66edea17dafb2f17522a/docs/standalone-mkl.md).
+Use the native CLI or C ABI documented in the code repository. GGUF storage does not imply compatibility with unrelated GGUF runtimes.
+
+## Licence
+
+These weights use Mistral AI's **Modified MIT License**, not unrestricted MIT.
+Read the complete [original licence](LICENSE), including its attribution, disclaimer, and revenue restriction.
+The licence restricts rights when your company or employer exceeds USD 20 million in consolidated monthly revenue for the preceding month.
+The restriction also covers derivatives. Contact Mistral AI for a commercial licence when required.
+Keep the original notices with copies and derivatives. This repository does not grant additional rights.
+
+## Integrity and provenance
+
+`manifest.json` records exact sizes, SHA-256 hashes, upstream revisions, and runtime/reference identities.
+`SHA256SUMS` contains checksums for all nine GGUF files.
+F32/F16/Q8_0 use the pinned converter; 209M K formats use the reviewed native quantizer.
+The 5M F32 file is byte-identical to the previously accepted trained-parity artifact.

SHA256SUMS

@@ -0,0 +1,9 @@
+c320276d8db95fb72fe5f4e043147e29142a1bf197c1b45daf6736780f6736ed  5M/timestral-5m-f32.gguf
+933e2e518300e97e7799deb1920dd4e539da89a90c743b663cda772ee79d4a38  5M/timestral-5m-f16.gguf
+a5781a8c8e5c6d5ccadfba91c96dda92cea4dc84778eef0b8c3590535cd72a4c  5M/timestral-5m-q8_0.gguf
+bf7605545245a2dd4d32f7ac2f7bbe5c2a12d8f47370814a254862150255b8e8  209M/timestral-209m-f32.gguf
+24242f55b330a3794649700f6fadd7e59de320aabb42fe377d8df4ab811fb94f  209M/timestral-209m-f16.gguf
+8bcf4b8a0b4042e06c3f26dd611b05e0a0bff7de5b7c7869aa0757f58383b479  209M/timestral-209m-q8_0.gguf
+49f364c750906c2a0513a407af5c94cf16a6d711976ce803ef966b17a0f6fe4d  209M/timestral-209m-q6_k.gguf
+e894e488baf17ba07a8dc20c685506a5a7bb89ca8ddb82e920cb2113401d1fef  209M/timestral-209m-q5_k.gguf
+05b9a0c99c099ba23ecb0a8dabf1552a6f0fbd0b7e37df0163fe41c851d575a6  209M/timestral-209m-q4_k.gguf

manifest.json

@@ -0,0 +1,72 @@
+{
+  "repository": "mudler/timestral-cpp-gguf",
+  "format": "GGUF",
+  "runtime_commit": "68a5a5394fda54bb852f3824e0a0465285edfe15",
+  "code_commit": "a4a1f95bb993a364c59d66edea17dafb2f17522a",
+  "reference_commit": "a534bdbd244d64f3d9d743f46c7bf6fc330ed4ff",
+  "files": [
+    {
+      "path": "5M/timestral-5m-f32.gguf",
+      "bytes": 21926656,
+      "sha256": "c320276d8db95fb72fe5f4e043147e29142a1bf197c1b45daf6736780f6736ed",
+      "upstream": "mistralai/Timestral-5M",
+      "revision": "bdd20cfc15a3fea0e4cb5ccef37f8d3ca7aa335f"
+    },
+    {
+      "path": "5M/timestral-5m-f16.gguf",
+      "bytes": 11641824,
+      "sha256": "933e2e518300e97e7799deb1920dd4e539da89a90c743b663cda772ee79d4a38",
+      "upstream": "mistralai/Timestral-5M",
+      "revision": "bdd20cfc15a3fea0e4cb5ccef37f8d3ca7aa335f"
+    },
+    {
+      "path": "5M/timestral-5m-q8_0.gguf",
+      "bytes": 6820704,
+      "sha256": "a5781a8c8e5c6d5ccadfba91c96dda92cea4dc84778eef0b8c3590535cd72a4c",
+      "upstream": "mistralai/Timestral-5M",
+      "revision": "bdd20cfc15a3fea0e4cb5ccef37f8d3ca7aa335f"
+    },
+    {
+      "path": "209M/timestral-209m-f32.gguf",
+      "bytes": 838347872,
+      "sha256": "bf7605545245a2dd4d32f7ac2f7bbe5c2a12d8f47370814a254862150255b8e8",
+      "upstream": "mistralai/Timestral-209M",
+      "revision": "9874d252ab0aae16f7f7c999f98f98d6ad7d8365"
+    },
+    {
+      "path": "209M/timestral-209m-f16.gguf",
+      "bytes": 420310336,
+      "sha256": "24242f55b330a3794649700f6fadd7e59de320aabb42fe377d8df4ab811fb94f",
+      "upstream": "mistralai/Timestral-209M",
+      "revision": "9874d252ab0aae16f7f7c999f98f98d6ad7d8365"
+    },
+    {
+      "path": "209M/timestral-209m-q8_0.gguf",
+      "bytes": 224355136,
+      "sha256": "8bcf4b8a0b4042e06c3f26dd611b05e0a0bff7de5b7c7869aa0757f58383b479",
+      "upstream": "mistralai/Timestral-209M",
+      "revision": "9874d252ab0aae16f7f7c999f98f98d6ad7d8365"
+    },
+    {
+      "path": "209M/timestral-209m-q6_k.gguf",
+      "bytes": 173733792,
+      "sha256": "49f364c750906c2a0513a407af5c94cf16a6d711976ce803ef966b17a0f6fe4d",
+      "upstream": "mistralai/Timestral-209M",
+      "revision": "9874d252ab0aae16f7f7c999f98f98d6ad7d8365"
+    },
+    {
+      "path": "209M/timestral-209m-q5_k.gguf",
+      "bytes": 145973472,
+      "sha256": "e894e488baf17ba07a8dc20c685506a5a7bb89ca8ddb82e920cb2113401d1fef",
+      "upstream": "mistralai/Timestral-209M",
+      "revision": "9874d252ab0aae16f7f7c999f98f98d6ad7d8365"
+    },
+    {
+      "path": "209M/timestral-209m-q4_k.gguf",
+      "bytes": 119846112,
+      "sha256": "05b9a0c99c099ba23ecb0a8dabf1552a6f0fbd0b7e37df0163fe41c851d575a6",
+      "upstream": "mistralai/Timestral-209M",
+      "revision": "9874d252ab0aae16f7f7c999f98f98d6ad7d8365"
+    }
+  ]
+}