2026-10-27: PP-DocLayoutV3_fp32_batchable.onnx, PP-OCRv6_medium_det.onnx, PP-OCRv6_small_det.onnx, PP-OCRv6_tiny_det.onnx, PP-OCRv6_tiny_rec_ctc_fp16.onnx, PP-OCRv6_medium_rec_ctc_fp16.onnx, PP-OCRv6_small_rec_ctc_fp16.onnx, PP-FormulaNet_plus-M_encoder.onnx, PP-FormulaNet_plus-M_prep.onnx, PP-FormulaNet_plus-M_decoder_step.onnx, PP-FormulaNet_plus-M_tokenizer.json, PP-FormulaNet_plus-S_encoder.onnx, PP-FormulaNet_plus-S_prep.onnx, PP-FormulaNet_plus-S_decoder_step.onnx, PP-FormulaNet_plus-S_tokenizer.json, SLANet_plus_encoder.onnx, SLANet_plus_decoder.bin

MANIFEST.txt

@@ -2,9 +2,9 @@
 eb13b44b25bb36f89528b68720af8a61d9cf381176107f465db1757b65d086e1  PP-OCRv6_medium_det.onnx
 d73e0058b7a8086bbd57f3d10b8bcd4ff95363f67e06e2762b5e814fe9c9410e  PP-OCRv6_small_det.onnx
 193bab7a04fca699a6c82e6abb5b81bdb28177f0abd4062552b04908dafb19f8  PP-OCRv6_tiny_det.onnx
-b9f7b60ef5bb0cb51824d16cb5d9c33f2e10c3c825a060e11ddf5c836ae46527  PP-OCRv6_tiny_rec_ctc.onnx
-f81aebe4dcc25be084e46e4ccffaee5ce0fd92e9e4e2c5443c655639a70b3ad2  PP-OCRv6_medium_rec_ctc.onnx
-ed513aba775144c855d5864cc52137a767f186124e23db5945dd41ca518f69ac  PP-OCRv6_small_rec_ctc.onnx
+722a822d9c00f64d8a8206ede923c7651a7ce235b722e1c87c8d00736882f858  PP-OCRv6_tiny_rec_ctc_fp16.onnx
+590e788eb7ce9d7cd4ece16fc09bee70426c2adcd5c2dca8509c568783bcdae9  PP-OCRv6_medium_rec_ctc_fp16.onnx
+30490f7b582c233248ba39b21bbd3d99cd0b89c0c444409ce8c5f5aee178f866  PP-OCRv6_small_rec_ctc_fp16.onnx
 e15dc1f333c680ba2b9eef5a8be713391ad9b587c074ff5120254063ea7a1244  PP-FormulaNet_plus-M_encoder.onnx
 6fd37a4dc090b28c1a2a62516ec863786f8033dca14638400c0f7e3ce71d061f  PP-FormulaNet_plus-M_prep.onnx
 bb4a0d6f868bf60e2cf551203332db6c3f3ba1b838110ba03f669a5fc025787c  PP-FormulaNet_plus-M_decoder_step.onnx

PP-OCRv6_medium_rec_ctc.onnx → PP-OCRv6_medium_rec_ctc_fp16.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f81aebe4dcc25be084e46e4ccffaee5ce0fd92e9e4e2c5443c655639a70b3ad2
-size 76555275
+oid sha256:590e788eb7ce9d7cd4ece16fc09bee70426c2adcd5c2dca8509c568783bcdae9
+size 38307860

PP-OCRv6_small_rec_ctc.onnx → PP-OCRv6_small_rec_ctc_fp16.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ed513aba775144c855d5864cc52137a767f186124e23db5945dd41ca518f69ac
-size 21159674
+oid sha256:30490f7b582c233248ba39b21bbd3d99cd0b89c0c444409ce8c5f5aee178f866
+size 10608083

PP-OCRv6_tiny_rec_ctc.onnx → PP-OCRv6_tiny_rec_ctc_fp16.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b9f7b60ef5bb0cb51824d16cb5d9c33f2e10c3c825a060e11ddf5c836ae46527
-size 4463021
+oid sha256:722a822d9c00f64d8a8206ede923c7651a7ce235b722e1c87c8d00736882f858
+size 2246736

README.md

@@ -18,8 +18,8 @@ ONNX exports of PaddlePaddle document models, one flat file per model.
 | --- | --- | --- |
 | `PP-DocLayoutV3_fp32_batchable.onnx` | layout detection | [`PaddlePaddle/PP-DocLayoutV3_onnx`](https://huggingface.co/PaddlePaddle/PP-DocLayoutV3_onnx), output reshaped to `[B,300,7]` |
 | `PP-OCRv6_{tiny,small,medium}_det.onnx` | text detection | [`PaddlePaddle/PP-OCRv6_*_det_onnx`](https://huggingface.co/PaddlePaddle), unchanged |
-| `PP-OCRv6_{tiny,small,medium}_rec_ctc.onnx` | text recognition | [`PaddlePaddle/PP-OCRv6_*_rec_onnx`](https://huggingface.co/PaddlePaddle), plus `ctc_idx` / `ctc_prob` outputs (argmax and max over the class axis) |
-| `SLANet_plus_encoder.onnx`, `SLANet_plus_decoder.bin` | table structure | [`PaddlePaddle/SLANet_plus_onnx`](https://huggingface.co/PaddlePaddle/SLANet_plus_onnx): the encoder graph, and the GRU decoder's weights as raw float32 |
-| `PP-FormulaNet_plus-{M,S}_{encoder,prep,decoder_step}.onnx`, `_tokenizer.json` | formula recognition | [`PaddlePaddle/PP-FormulaNet_plus-M`](https://huggingface.co/PaddlePaddle/PP-FormulaNet_plus-M) / [`-S`](https://huggingface.co/PaddlePaddle/PP-FormulaNet_plus-S), split into encoder, cross-attention precompute and one greedy decoder step (FP16, new-slice) whose KV axis is dynamic: one TensorRT engine per tier with a profile per KV length (256 / 512 / 1536) |
+| `PP-OCRv6_{tiny,small,medium}_rec_ctc_fp16.onnx` | text recognition | [`PaddlePaddle/PP-OCRv6_*_rec_onnx`](https://huggingface.co/PaddlePaddle), plus `ctc_idx` / `ctc_prob` outputs (argmax and max over the class axis), FP16 weights with FP32 inputs and outputs |
+| `SLANet_plus_encoder.onnx`, `SLANet_plus_decoder.bin` | table structure | [`PaddlePaddle/SLANet_plus_onnx`](https://huggingface.co/PaddlePaddle/SLANet_plus_onnx): the encoder graph, and the GRU decoder's weights as raw float32 (run by a Python model, `table_decode`) |
+| `PP-FormulaNet_plus-{M,S}_{encoder,prep,decoder_step}.onnx`, `_tokenizer.json` | formula recognition | [`PaddlePaddle/PP-FormulaNet_plus-M`](https://huggingface.co/PaddlePaddle/PP-FormulaNet_plus-M) / [`-S`](https://huggingface.co/PaddlePaddle/PP-FormulaNet_plus-S), split into encoder, cross-attention precompute and one greedy decoder step (FP16, new-slice) whose KV axis is dynamic: one TensorRT engine per tier with a profile per KV length (plus-S 256 / 512 / 1024, plus-M 256 / 512 / 1536) |
 
 Pre- and post-processing follow each source model's `inference.yml`.