Grounding DINO tiny, fp16, with a text mask for any number of phrases

LICENSE

@@ -0,0 +1,202 @@
+
+                                 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 [yyyy] [name of copyright owner]
+
+   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.

README.md

@@ -0,0 +1,127 @@
+---
+license: apache-2.0
+library_name: transformers.js
+base_model: IDEA-Research/grounding-dino-tiny
+pipeline_tag: zero-shot-object-detection
+tags:
+  - onnx
+  - grounding-dino
+  - open-vocabulary-detection
+  - transformers.js
+  - onnxruntime-web
+  - webgpu
+---
+
+# Grounding DINO Tiny (ONNX, fp16), multi-phrase
+
+Grounding DINO Tiny as ONNX for [Transformers.js](https://github.com/huggingface/transformers.js),
+fp16. Give it an image and a prompt of one or more phrases (`goat. city. sky.`), and it returns
+boxes, each with the phrase it matched. [Card Anvil](https://github.com/Card-Anvil)'s Describe
+command uses it in the browser, on WebGPU or WebAssembly, to find what to select in card art.
+
+**Why another export.** [`onnx-community/grounding-dino-tiny-ONNX`](https://huggingface.co/onnx-community/grounding-dino-tiny-ONNX)
+scores only the **first** phrase of a prompt. Every phrase after it comes out near zero (0.02–0.05
+where PyTorch gives 0.5–0.8), so `goat. city. sky.` finds only goats. That export was made with
+transformers 4.48, which builds each phrase's text attention mask in a Python loop over the
+prompt's special tokens. Tracing unrolled the loop for a one-phrase example prompt.
+
+This export rebuilds that function from tensor comparisons and min/max. Before each export it is
+checked against Transformers' own on 140 prompts of 0–6 phrases, so the graph handles any number
+of phrases, up to the model's 256 tokens.
+
+It is a drop-in replacement: the same inputs, outputs and file name as onnx-community's, so
+Transformers.js loads it the same way.
+
+## Files
+
+| File                     | What                                                                                                    | Size   |
+| ------------------------ | ------------------------------------------------------------------------------------------------------- | ------ |
+| `onnx/model_fp16.onnx`   | The whole model, fp16 weights, float32 inputs and outputs                                               | 364 MB |
+| `config.json`, `preprocessor_config.json`, tokenizer files | Copied unchanged from onnx-community's repo at [`ff690b0a`](https://huggingface.co/onnx-community/grounding-dino-tiny-ONNX/tree/ff690b0a8050566c290287545bd059350f3e9096) | |
+| `export_grounding_dino.py` | The script that made it                                                                               |        |
+
+| Inputs                                                                  | Outputs                                                       |
+| ----------------------------------------------------------------------- | ------------------------------------------------------------- |
+| `pixel_values` [1, 3, 800, 800], `pixel_mask` [1, 800, 800]             | `logits` [1, 900, 256]: each query's score for each token     |
+| `input_ids`, `token_type_ids`, `attention_mask` [1, n], n from 4 to 256 | `pred_boxes` [1, 900, 4]: centre x, centre y, width, height, 0–1 |
+
+The processor stretches the image to 800 × 800, with no padding.
+
+## Usage
+
+```js
+import {
+  AutoProcessor,
+  GroundingDinoForObjectDetection,
+  RawImage,
+} from "@huggingface/transformers";
+
+const id = "CardAnvilModels/grounding-dino-tiny-ONNX";
+const model = await GroundingDinoForObjectDetection.from_pretrained(id, {
+  dtype: "fp16",
+  device: "webgpu", // or "wasm"
+});
+const processor = await AutoProcessor.from_pretrained(id);
+
+const image = await RawImage.fromURL("https://example.com/art.jpg");
+// Lower-case phrases, each ended with a full stop.
+const inputs = await processor(image, "goat. city. sky.");
+const outputs = await model(inputs);
+const [result] = processor.post_process_grounded_object_detection(
+  outputs,
+  inputs.input_ids,
+  {
+    box_threshold: 0.35,
+    text_threshold: 0.25,
+    target_sizes: [[image.height, image.width]],
+  },
+);
+// result.boxes: [x0, y0, x1, y1] in pixels; result.labels: the phrase each matched.
+```
+
+In one pass, the phrases compete for boxes, as they do in the original model. Something found on
+its own can lose to another phrase that claims the same region.
+
+## How it was checked
+
+With onnxruntime 1.31 on the CPU, against PyTorch (transformers 5.20, fp32) on the same inputs:
+
+- **fp32 export:** within 0.01 of PyTorch, on single phrases and on prompts of three and four
+  phrases.
+- **This fp16 file:** within 0.01 of PyTorch on single phrases, and close on joint prompts. The
+  largest gap seen was about 0.1, on one three-phrase prompt.
+
+In Microsoft Edge, with ONNX Runtime Web 1.31, on 24 images and 98 prompts:
+
+| Measure                                     | onnx-community's | This export                      |
+| ------------------------------------------- | ---------------- | -------------------------------- |
+| WebGPU, a pass (RTX 4070 Ti SUPER)          | 327 ms           | 347 ms                           |
+| WebAssembly, a pass (one thread)            | 13.4 s           | 14.7 s                           |
+| Single-phrase boxes                         | —                | the same: median IoU 0.995       |
+| Phrases found alone that a joint pass loses | 45 of 68         | 14 of 66, as in PyTorch          |
+
+All 48 `GridSample` nodes (the deformable attention) run on WebGPU.
+
+## How it was made
+
+`export_grounding_dino.py`, with torch 2.14.1, transformers 5.20.2, onnx 1.23.2, onnxscript 0.7.2
+and onnxconverter-common 1.16.0:
+
+1. **fp32 export** from [`IDEA-Research/grounding-dino-tiny`](https://huggingface.co/IDEA-Research/grounding-dino-tiny)
+   at `a2bb814d`, with the dynamo exporter, opset 18, and a dynamic sequence length. Two changes
+   make it export:
+   - the text-mask function rebuilt as above (Transformers' uses `torch.isin`, which exports as
+     `searchsorted`, which ONNX lacks);
+   - `TRANSFORMERS_DISABLE_TORCH_CHECK=1`, since its shape checks export as `_is_all_true`, which
+     ONNX also lacks.
+2. **fp16 conversion** with `onnxconverter_common`, inputs and outputs kept float32. The converter
+   leaves 23 `Cast` nodes aimed at float32; the script retypes them.
+
+## Credits and licence
+
+- Model: [Grounding DINO](https://github.com/IDEA-Research/GroundingDINO) by IDEA Research, from
+  _Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection_ (2023),
+  through [`IDEA-Research/grounding-dino-tiny`](https://huggingface.co/IDEA-Research/grounding-dino-tiny).
+- Config and tokenizer files: [ONNX Community](https://huggingface.co/onnx-community).
+
+Licensed under the Apache License 2.0, as the original model is. See [`LICENSE`](LICENSE).

config.json

@@ -0,0 +1,66 @@
+{
+  "_name_or_path": "IDEA-Research/grounding-dino-tiny",
+  "activation_dropout": 0.0,
+  "activation_function": "relu",
+  "architectures": [
+    "GroundingDinoForObjectDetection"
+  ],
+  "attention_dropout": 0.0,
+  "auxiliary_loss": false,
+  "backbone": null,
+  "backbone_config": {
+    "model_type": "swin",
+    "out_features": [
+      "stage2",
+      "stage3",
+      "stage4"
+    ],
+    "out_indices": [
+      2,
+      3,
+      4
+    ]
+  },
+  "backbone_kwargs": null,
+  "bbox_cost": 5.0,
+  "bbox_loss_coefficient": 5.0,
+  "class_cost": 1.0,
+  "d_model": 256,
+  "decoder_attention_heads": 8,
+  "decoder_bbox_embed_share": true,
+  "decoder_ffn_dim": 2048,
+  "decoder_layers": 6,
+  "decoder_n_points": 4,
+  "disable_custom_kernels": false,
+  "dropout": 0.1,
+  "embedding_init_target": true,
+  "encoder_attention_heads": 8,
+  "encoder_ffn_dim": 2048,
+  "encoder_layers": 6,
+  "encoder_n_points": 4,
+  "focal_alpha": 0.25,
+  "fusion_dropout": 0.0,
+  "fusion_droppath": 0.1,
+  "giou_cost": 2.0,
+  "giou_loss_coefficient": 2.0,
+  "init_std": 0.02,
+  "is_encoder_decoder": true,
+  "layer_norm_eps": 1e-05,
+  "max_text_len": 256,
+  "model_type": "grounding-dino",
+  "num_feature_levels": 4,
+  "num_queries": 900,
+  "position_embedding_type": "sine",
+  "positional_embedding_temperature": 20,
+  "query_dim": 4,
+  "text_config": {
+    "model_type": "bert"
+  },
+  "text_enhancer_dropout": 0.0,
+  "torch_dtype": "float32",
+  "transformers_version": "4.48.0.dev0",
+  "two_stage": true,
+  "two_stage_bbox_embed_share": false,
+  "use_pretrained_backbone": false,
+  "use_timm_backbone": false
+}

export_grounding_dino.py

@@ -0,0 +1,167 @@
+"""Grounding DINO tiny as ONNX, with a text mask that works for any number of phrases.
+
+onnx-community/grounding-dino-tiny-ONNX was exported with transformers 4.48, whose
+`generate_masks_with_special_tokens_and_transfer_map` loops in Python over the prompt's special
+tokens. Tracing unrolled that loop for its example prompt, so the export scores only a prompt's
+first phrase. Transformers 5 builds the mask from tensor operations, which export as they are.
+
+Same inputs and outputs as onnx-community's, so Transformers.js loads it unchanged:
+pixel_values, input_ids, token_type_ids, attention_mask, pixel_mask -> logits, pred_boxes.
+
+Usage, from this folder, with tools/sam3-export's .venv:
+  ../sam3-export/.venv/Scripts/python.exe -I export_grounding_dino.py --weights <dir> --out onnx
+"""
+
+import argparse
+import os
+from pathlib import Path
+
+import onnx
+import torch
+import transformers.models.grounding_dino.modeling_grounding_dino as grounding_dino
+from onnxconverter_common import float16
+from transformers import GroundingDinoForObjectDetection
+
+OPSET = 18
+IMAGE_SIZE = 800  # the export's processor stretches the art to 800 x 800
+INPUTS = ["pixel_values", "input_ids", "token_type_ids", "attention_mask", "pixel_mask"]
+OUTPUTS = ["logits", "pred_boxes"]
+
+
+def phrase_masks(input_ids: torch.LongTensor) -> tuple[torch.Tensor, torch.Tensor]:
+    """`generate_masks_with_special_tokens_and_transfer_map`, in operations ONNX has.
+
+    Transformers' version uses `torch.isin` (exported as `searchsorted`, which ONNX lacks) and
+    `cummax`/`cummin`. This finds each token's previous and next special token with a masked
+    min/max over a sequence x sequence grid instead: at most 256 x 256, so cheap. Same result;
+    `check_phrase_masks` compares the two.
+    """
+    batch_size, seq_len = input_ids.shape
+    device = input_ids.device
+    special = torch.zeros_like(input_ids, dtype=torch.bool)
+    for token in grounding_dino.SPECIAL_TOKENS:
+        special = special | (input_ids == token)
+    indices = torch.arange(seq_len, device=device)
+    row = indices.view(1, seq_len, 1)  # each token i
+    col = indices.view(1, 1, seq_len)  # each candidate special token j
+    is_special = special.unsqueeze(1)  # (B, 1, N)
+    prev_special = torch.where(is_special & (col <= row), col, -1).amax(dim=2)
+    next_special = torch.where(is_special & (col >= row), col, seq_len).amin(dim=2)
+    valid_block = (next_special != 0) & (next_special != seq_len - 1) & (next_special != seq_len)
+    attention_mask = (next_special.unsqueeze(2) == next_special.unsqueeze(1)) & valid_block.unsqueeze(1)
+    identity = (row == col).expand(batch_size, -1, -1)
+    attention_mask = identity | attention_mask
+    position_ids = indices.unsqueeze(0) - prev_special - 1
+    position_ids = torch.where(valid_block, position_ids, torch.zeros_like(position_ids))
+    position_ids = torch.clamp(position_ids, min=0).to(torch.long)
+    return attention_mask, position_ids
+
+
+def check_phrase_masks() -> None:
+    """`phrase_masks` against Transformers' own on prompts of 0–6 phrases, with padding."""
+    original = grounding_dino.generate_masks_with_special_tokens_and_transfer_map
+    generator = torch.Generator().manual_seed(0)
+    for phrases in range(7):
+        for _ in range(20):
+            ids = [101]
+            for _ in range(phrases):
+                words = int(torch.randint(1, 4, (1,), generator=generator))
+                ids += torch.randint(2000, 3000, (words,), generator=generator).tolist() + [1012]
+            ids += [102] + [0] * int(torch.randint(0, 3, (1,), generator=generator))
+            input_ids = torch.tensor([ids])
+            want, got = original(input_ids), phrase_masks(input_ids)
+            assert all(torch.equal(a, b) for a, b in zip(want, got)), ids
+    print("phrase_masks matches Transformers on 140 prompts")
+
+
+class Detector(torch.nn.Module):
+    def __init__(self, model):
+        super().__init__()
+        self.model = model
+
+    def forward(self, pixel_values, input_ids, token_type_ids, attention_mask, pixel_mask):
+        out = self.model(
+            pixel_values=pixel_values,
+            input_ids=input_ids,
+            token_type_ids=token_type_ids,
+            attention_mask=attention_mask,
+            pixel_mask=pixel_mask,
+        )
+        return out.logits, out.pred_boxes
+
+
+def example(tokens: int):
+    """Inputs for tracing: three phrases' worth of special tokens, so nothing assumes one."""
+    # [CLS] dragon . sword . fire . [SEP], padded with filler words to `tokens`.
+    ids = [101, 1000, 1012, 1001, 1012, 1002, 1012] + [1003] * (tokens - 8) + [102]
+    return (
+        torch.zeros(1, 3, IMAGE_SIZE, IMAGE_SIZE),
+        torch.tensor([ids], dtype=torch.long),
+        torch.zeros(1, tokens, dtype=torch.long),
+        torch.ones(1, tokens, dtype=torch.long),
+        torch.ones(1, IMAGE_SIZE, IMAGE_SIZE, dtype=torch.long),
+    )
+
+
+def export_fp32(weights: Path, out: Path) -> Path:
+    out.mkdir(parents=True, exist_ok=True)
+    check_phrase_masks()
+    # The model's forward calls the module-level function, so this swaps it for the export.
+    grounding_dino.generate_masks_with_special_tokens_and_transfer_map = phrase_masks
+    # Its shape asserts export as `_is_all_true`, which ONNX lacks; they're hints, not logic.
+    os.environ["TRANSFORMERS_DISABLE_TORCH_CHECK"] = "1"
+    model = Detector(GroundingDinoForObjectDetection.from_pretrained(weights).eval()).eval()
+    seq = torch.export.Dim("sequence", min=4, max=256)
+    path = out / "model.onnx"
+    with torch.no_grad():
+        torch.onnx.export(
+            model,
+            example(12),
+            str(path),
+            input_names=INPUTS,
+            output_names=OUTPUTS,
+            opset_version=OPSET,
+            dynamo=True,
+            dynamic_shapes=(None, {1: seq}, {1: seq}, {1: seq}, None),
+            external_data=True,
+            optimize=True,
+        )
+    return path
+
+
+def fix_casts(model: onnx.ModelProto) -> int:
+    """As in tools/sam3-export: retype each Cast the fp16 converter left aimed at float32."""
+    declared = {vi.name: vi.type.tensor_type.elem_type for vi in model.graph.value_info}
+    fixed = 0
+    for node in model.graph.node:
+        if node.op_type != "Cast" or declared.get(node.output[0]) != onnx.TensorProto.FLOAT16:
+            continue
+        for attribute in node.attribute:
+            if attribute.name == "to" and attribute.i == onnx.TensorProto.FLOAT:
+                attribute.i = onnx.TensorProto.FLOAT16
+                fixed += 1
+    return fixed
+
+
+def to_fp16(src: Path, dst: Path) -> None:
+    model = onnx.load(str(src))
+    del model.graph.value_info[:]
+    half = float16.convert_float_to_float16(model, keep_io_types=True, disable_shape_infer=False)
+    print(f"fp16: {fix_casts(half)} casts fixed")
+    onnx.save(half, str(dst))
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--weights", type=Path, required=True)
+    parser.add_argument("--out", type=Path, default=Path("onnx"))
+    parser.add_argument("--skip-fp32", action="store_true")
+    args = parser.parse_args()
+    fp32 = args.out / "fp32" / "model.onnx"
+    if not args.skip_fp32:
+        fp32 = export_fp32(args.weights, args.out / "fp32")
+    to_fp16(fp32, args.out / "model_fp16.onnx")
+
+
+if __name__ == "__main__":
+    main()

onnx/model_fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88a353167c3aa952f73ff7b0e380e155b25c7500e1c07c4d2898bb0ee42c2e89
+size 363683399

preprocessor_config.json

@@ -0,0 +1,27 @@
+{
+  "do_convert_annotations": true,
+  "do_normalize": true,
+  "do_pad": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "format": "coco_detection",
+  "image_mean": [
+    0.485,
+    0.456,
+    0.406
+  ],
+  "image_processor_type": "GroundingDinoImageProcessor",
+  "image_std": [
+    0.229,
+    0.224,
+    0.225
+  ],
+  "pad_size": null,
+  "processor_class": "GroundingDinoProcessor",
+  "resample": 2,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "width": 800,
+    "height": 800
+  }
+}

special_tokens_map.json

@@ -0,0 +1,37 @@
+{
+  "cls_token": {
+    "content": "[CLS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "[MASK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "sep_token": {
+    "content": "[SEP]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer_config.json

@@ -0,0 +1,57 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "extra_special_tokens": {},
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "processor_class": "GroundingDinoProcessor",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}