SwinIR-L real-world x4 GAN on Core AI (float16, 128-px tiles), Apache-2.0

.gitattributes

@@ -33,3 +33,4 @@ 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
+swinir-l-real-x4-128-float16.aimodel/main.mlirb filter=lfs diff=lfs merge=lfs -text

LICENSE

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [2021] [SwinIR Authors]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

NOTICE

@@ -0,0 +1,9 @@
+SwinIR-L real-world super-resolution x4 (GAN), converted to Core AI.
+
+Original model and weights: SwinIR, Jingyun Liang et al. (https://github.com/JingyunLiang/SwinIR),
+Apache License 2.0 (see LICENSE). Checkpoint: 003_realSR_BSRGAN_DFOWMFC_s64w8_SwinIR-L_x4_GAN.pth
+from the v0.0 release (SHA-256 99adfa91350a84c99e946c1eb3d8fce34bc28f57d807b09dc8fe40a316328c0a).
+
+Modifications: exported to Core AI (float16, fixed 128x128 -> 512x512 tile) with coreai-torch 0.4.3 /
+coreai-core 1.0.0b3; attention masks precomputed for the 128-px tile; output clamped to [0, 1].
+The export script is export_swinir.py in this repository.

README.md

@@ -0,0 +1,47 @@
+---
+license: apache-2.0
+library_name: coreai
+pipeline_tag: image-to-image
+base_model: JingyunLiang/SwinIR
+tags:
+- super-resolution
+- swinir
+- coreai
+- apple-silicon
+---
+
+# SwinIR-L real-world ×4 for Core AI
+
+The default upscaler of [`upscale`](https://github.com/arraypress/swift-upscale) and
+`imagine --upscale`: **faithful** 4× super-resolution — it recovers detail without inventing it.
+
+- `swinir-l-real-x4-128-float16.aimodel`: input `x` `[1,3,128,128]` float16 in `[0,1]` (RGB), output
+  `output` `[1,3,512,512]` in `[0,1]`. Tile any picture with ~16 px overlap and feather the seams
+  ([swift-upscale](https://github.com/arraypress/swift-upscale) does).
+
+## Why this one
+
+Chosen in a bake-off of seven upscalers (×4 shrink-and-back on 5 generated and 5 stock photos, LPIPS
+lower is better): SwinIR-L scored **0.185 / 0.208** (generated / real) and 0.213 / 0.282 with blur,
+noise and JPEG damage, beating Real-ESRGAN x4plus (0.195 / 0.236) in every cell, with the best PSNR
+(28.05 / 24.45 dB). It keeps hair and foliage natural where Real-ESRGAN smooths them.
+
+## Measured (M3 Max, macOS 27, GPU)
+
+- **Parity** with the PyTorch float32 model, per 128-px tile: **66–68 dB PSNR** (float16; float32
+  export 124–125 dB). Float16 is shipped: half the size, slightly faster, no visible difference.
+- **Tile size:** 128 px had the lowest cost per new pixel — 24.9 µs, against 26.9 at 256 px and 29.0
+  at 192 px.
+- **Seams:** a 256-px picture upscaled as overlapping tiles against PyTorch on the whole picture in one
+  run: 41.6 dB overall, and the seam bands match as well as everywhere else (41.8 dB) — no seams.
+- **Speed:** 1024×1024 → 4096×4096 in **25–29 s** (three runs on a quiet machine; first load 0.2 s
+  once compiled) — about 0.25 s per 128-px tile, so time grows with the number of tiles.
+
+## Provenance
+
+- Weights: SwinIR `v0.0` release, `003_realSR_BSRGAN_DFOWMFC_s64w8_SwinIR-L_x4_GAN.pth`, byte-identical
+  to the release asset (SHA-256 `99adfa91…328c0a`).
+- Export: `export_swinir.py` (this repository) — spandrel 0.4.1 loads the checkpoint; the attention
+  masks are precomputed for the 128-px tile (SwinIR rebuilds them per call at other sizes, which
+  crashed the converter); `coreai-torch 0.4.3`, `coreai-core 1.0.0b3`, float16 under autocast.
+- Licence: Apache-2.0 (`LICENSE`), with `NOTICE`.

export_swinir.py

@@ -0,0 +1,116 @@
+# /// script
+# requires-python = ">=3.11"
+# dependencies = [
+#     "coreai-core==1.0.0b3",
+#     "coreai-torch==0.4.3",
+#     "spandrel==0.4.1",
+#     "numpy",
+#     "pillow",
+# ]
+#
+# [tool.uv]
+# index-url       = "https://pypi.org/simple"
+# prerelease      = "allow"
+# index-strategy  = "unsafe-best-match"
+# ///
+"""Export SwinIR-L (real-world SR, x4, GAN) as a fixed-tile Core AI asset and check it
+against PyTorch float32, tile by tile.
+
+    uv run export_swinir.py <ckpt.pth> <out-dir> <tile> <dtype> [<image.png> ...]
+"""
+import asyncio, sys, time
+from pathlib import Path
+
+import numpy as np
+import torch
+from PIL import Image
+from spandrel import ModelLoader
+from coreai.runtime import AIModelAssetMetadata
+import coreai.runtime as rt
+from coreai_torch import TorchConverter, get_decomp_table
+
+ckpt, out_dir, tile, dtype_name = sys.argv[1], Path(sys.argv[2]), int(sys.argv[3]), sys.argv[4]
+images = sys.argv[5:]
+dtype = {"float16": torch.float16, "float32": torch.float32}[dtype_name]
+
+# Two separate loads: the reference must stay float32 even when the export is float16.
+model32 = ModelLoader().load_from_file(ckpt).model.eval().float()
+desc = ModelLoader().load_from_file(ckpt)
+print("arch", type(model32).__name__, "scale", desc.scale, flush=True)
+
+def fix_masks(net, size):
+    """SwinIR blocks use their precomputed attention mask only when the input matches
+    the resolution they were built for (64 px); otherwise every call rebuilds it with
+    in-place indexed writes, which crashed the Core AI converter (SIGSEGV). Rebuild
+    every block for the tile size so the masks are constants in the graph."""
+    n = 0
+    for m in net.modules():
+        if type(m).__name__ == "SwinTransformerBlock":
+            m.input_resolution = (size, size)
+            if m.shift_size > 0:
+                m.attn_mask = m.calculate_mask((size, size))
+            n += 1
+    for m in net.modules():
+        if hasattr(m, "input_resolution") and type(m).__name__ != "SwinTransformerBlock":
+            m.input_resolution = (size, size)
+    return n
+
+print("blocks re-planned:", fix_masks(desc.model, tile), flush=True)
+
+class Wrapped(torch.nn.Module):
+    """[0,1] RGB in, [0,1] RGB out, clamped — what the Swift side feeds and reads."""
+    def __init__(self, m): super().__init__(); self.m = m
+    def forward(self, x): return self.m(x).clamp(0, 1)
+
+model = Wrapped(desc.model.eval()).to(dtype)
+example = {"x": torch.rand(1, 3, tile, tile).to(dtype)}
+t0 = time.time()
+with torch.no_grad(), torch.autocast(device_type="cpu", dtype=dtype, enabled=dtype != torch.float32):
+    exported = torch.export.export(model, args=(), kwargs=example)
+exported = exported.run_decompositions(get_decomp_table())
+program = TorchConverter(mode=TorchConverter.Mode.RELEASE).add_exported_program(
+    exported_program=exported, input_names=["x"], output_names=["output"]).to_coreai()
+md = AIModelAssetMetadata()
+md.author = "J. Liang et al. (SwinIR)"
+md.license = "Apache-2.0"
+md.model_description = (f"SwinIR-L real-world super-resolution x4 (GAN), "
+                        f"{tile}x{tile} -> {tile*4}x{tile*4} tile, {dtype_name}. "
+                        "Source: https://github.com/JingyunLiang/SwinIR")
+md.creation_date = int(time.time())
+out_dir.mkdir(parents=True, exist_ok=True)
+path = out_dir / f"swinir-l-real-x4-{tile}-{dtype_name}.aimodel"
+if path.exists():
+    import shutil; shutil.rmtree(path)
+program.save_asset(path, md)
+print(f"saved {path} in {time.time()-t0:.0f} s", flush=True)
+
+def psnr(a, b):
+    mse = float(np.mean((a.astype(np.float64) - b.astype(np.float64)) ** 2))
+    return 99.0 if mse == 0 else 10 * np.log10(1.0 / mse)
+
+async def parity():
+    tiles = []
+    rng = np.random.default_rng(0)
+    for p in images:
+        im = np.asarray(Image.open(p).convert("RGB")).astype(np.float32) / 255
+        h, w, _ = im.shape
+        for _ in range(3):
+            y = int(rng.integers(0, max(1, h - tile))); x = int(rng.integers(0, max(1, w - tile)))
+            crop = im[y:y+tile, x:x+tile]
+            if crop.shape[:2] == (tile, tile): tiles.append(crop)
+    if not tiles: tiles = [np.random.default_rng(1).random((tile, tile, 3), dtype=np.float32)]
+    model_rt = await rt.AIModel.load(str(path))
+    fn = model_rt.load_function("main")
+    scores, times = [], []
+    for crop in tiles:
+        x = torch.from_numpy(crop).permute(2, 0, 1)[None].contiguous()
+        with torch.no_grad(): ref = model32(x).clamp(0, 1)[0].permute(1, 2, 0).numpy()
+        t = time.time()
+        out = await fn({"x": rt.NDArray(data=x.to(dtype).contiguous())})
+        times.append(time.time() - t)
+        got = out["output"].numpy().astype(np.float32)[0].transpose(1, 2, 0)
+        scores.append(psnr(ref, got))
+    print(f"parity vs PyTorch fp32 over {len(scores)} tiles: PSNR min {min(scores):.1f} dB, "
+          f"mean {np.mean(scores):.1f} dB; Python runtime {np.median(times)*1000:.0f} ms/tile (median)", flush=True)
+
+asyncio.run(parity())

swinir-l-real-x4-128-float16.aimodel/main.hash

@@ -0,0 +1,2 @@
+4H-
+�N���%�5��I��jbX �M��\C

swinir-l-real-x4-128-float16.aimodel/main.mlirb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34480c2d0ab94eccd5e585259e35cbc59e86490fbef16a625820ce4da0c65c43
+size 61774569

swinir-l-real-x4-128-float16.aimodel/metadata.json

@@ -0,0 +1,8 @@
+{
+  "creationDate" : "20261011T193957Z",
+  "description" : "SwinIR-L real-world super-resolution x4 (GAN), 128x128 -> 512x512 tile, float16. Source: https:\/\/github.com\/JingyunLiang\/SwinIR",
+  "producer" : "coreai-core 1.0.0b3",
+  "license" : "Apache-2.0",
+  "assetVersion" : "2.0",
+  "author" : "J. Liang et al. (SwinIR)"
+}