Reduce bundle size after ONNX post-export layout optimization

.gitattributes

@@ -40,3 +40,4 @@ languages.png filter=lfs diff=lfs merge=lfs -text
 model.SRC filter=lfs diff=lfs merge=lfs -text
 model.TGT filter=lfs diff=lfs merge=lfs -text
 tokenizer_tgt.json filter=lfs diff=lfs merge=lfs -text
+decoder_shared.onnx.data filter=lfs diff=lfs merge=lfs -text

README.md

@@ -72,17 +72,16 @@ Compared against the FP32 ONNX oracle on the golden evaluation fixtures.
 
 | Format | Model Size | Exact Match (Token) | Exact Match (Text) | SacreBLEU (Raw) | Latency (Mean) | Speedup vs. FP32 |
 | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
-| FP32 | 6.64 GB | 100.00% | 100.00% | 100.00 | 244.4 ms | 1.000x |
-| FP16 | 3.32 GB | 99.73% | 99.73% | 100.00 | 259.8 ms | 0.941x |
-| INT8 | 1.66 GB | 89.64% | 89.73% | 95.69 | 112.6 ms | 2.228x |
-| Q4F16 | 850.5 MB | 82.27% | 82.36% | 92.55 | 143.3 ms | 1.673x |
+| FP32 | 6.64 GB | 100.00% | 100.00% | 100.00 | 69.5 ms | 1.000x |
+| FP16 | 3.32 GB | 99.73% | 99.73% | 100.00 | 74.3 ms | 0.935x |
+| INT8 | 1.66 GB | 89.55% | 89.55% | 96.27 | 31.4 ms | 2.125x |
+| Q4F16 | 850.5 MB | 82.45% | 82.55% | 91.99 | 58.4 ms | 1.186x |
 
 
 ## Files
 
 - `encoder_model.onnx` (and optional `.onnx.data` weights sidecar)
-- `decoder_model.onnx` (and optional `.onnx.data` weights sidecar)
-- `decoder_with_past_model.onnx` (and optional `.onnx.data` weights sidecar)
+- `decoder_model.onnx` and `decoder_with_past_model.onnx` (share `decoder_shared.onnx.data` when present)
 - `translate.py` — self-contained Python inference helper (see Usage below)
 - Fast tokenizer config files (`tokenizer_src.json`, `tokenizer_tgt.json`, `tokenizer_meta.json`)
 - Model configuration configs (`config.json`, `generation_config.json`)

decoder_model.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:27177ad88e0ace28bb0931400a75a74a1e7e7db2218fdb69d6433d503bc2f1a7
-size 1811875
+oid sha256:a3c69727dc4d5b4510bb4dda10d616ada571d43ee665232015e76877af36288f
+size 424244

decoder_with_past_model.onnx.data → decoder_shared.onnx.data

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:24ba50c8f3dfbac032cf0c5b5517e1ebd423108668f8b6fc6c34cbd52c3bd9fe
-size 2667429888
+oid sha256:3f525dc7de9ccb2e11608accc530dec4ac7332739f2fb3ba7e5f56a233bdba69
+size 2819629056

decoder_with_past_model.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5a60ae0313a87dcab5b3aaf37425bdfce49e2d3e5fc7866ba26c435e5bcb62e4
-size 1747287
+oid sha256:e68d1d33b71656f71a7a0bdc301459ac7f94f556d2c8461de69a2a9fe7a7b9b2
+size 392757

encoder_model.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5049976b746d0f49bb831fcf437bc1f461883bae26ef11c8d8d4154b044ec73c
-size 1452442
+oid sha256:84b706d878bd5fc2680be90e89ebe0e923b098d7a1ef2dc28dc0cab8ebe99932
+size 236890

encoder_model.onnx.data

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7356a14c26e92a2118a4b7ac78d245934a705f8802b6c91f29d665a408c2915c
-size 1643601920
+oid sha256:e7877cb26cd20178fb1288bc6d16d1bb245fc99c1437d00e626389e92971818e
+size 1644658688