Upload opus-mt-en-tr ONNX INT8

.gitattributes

@@ -34,3 +34,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 madlad400-3b-int8/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+opus-mt-en-tr/source.spm filter=lfs diff=lfs merge=lfs -text
+opus-mt-en-tr/target.spm filter=lfs diff=lfs merge=lfs -text

opus-mt-en-tr/LICENSE

@@ -0,0 +1,16 @@
+Original model: Helsinki-NLP/opus-tatoeba-en-tr
+License: Apache-2.0
+
+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.
+
+Model data licensed under CC-BY-4.0.

opus-mt-en-tr/README.md

@@ -0,0 +1,27 @@
+---
+license: apache-2.0
+base_model: Helsinki-NLP/opus-tatoeba-en-tr
+tags:
+  - translation
+  - onnx
+  - int8
+  - en-tr
+  - nobuds
+language:
+  - en
+  - tr
+---
+
+# OPUS-MT en→tr ONNX INT8
+
+INT8 quantized ONNX export of [Helsinki-NLP/opus-tatoeba-en-tr](https://huggingface.co/Helsinki-NLP/opus-tatoeba-en-tr) for on-device English to Turkish translation.
+
+- **Direction:** English → Turkish
+- **License:** Apache 2.0 (data: CC-BY-4.0)
+- **Quantization:** INT8 dynamic (optimum, avx2)
+- **Export tool:** optimum-cli
+
+## Usage
+
+Exported for the NoBuds on-device translation pipeline (Android/ONNX Runtime Mobile).
+For quality comparison against multilingual models (M2M-100, MADLAD-400).

opus-mt-en-tr/config.json

@@ -0,0 +1,40 @@
+{
+  "activation_dropout": 0.0,
+  "activation_function": "swish",
+  "architectures": [
+    "MarianMTModel"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "classifier_dropout": 0.0,
+  "d_model": 512,
+  "decoder_attention_heads": 8,
+  "decoder_ffn_dim": 2048,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 6,
+  "decoder_start_token_id": 59993,
+  "decoder_vocab_size": 59994,
+  "dropout": 0.1,
+  "dtype": "float32",
+  "encoder_attention_heads": 8,
+  "encoder_ffn_dim": 2048,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 6,
+  "eos_token_id": 0,
+  "forced_eos_token_id": 0,
+  "init_std": 0.02,
+  "is_encoder_decoder": true,
+  "max_length": null,
+  "max_position_embeddings": 512,
+  "model_type": "marian",
+  "normalize_embedding": false,
+  "num_beams": null,
+  "num_hidden_layers": 6,
+  "pad_token_id": 59993,
+  "scale_embedding": true,
+  "share_encoder_decoder_embeddings": true,
+  "static_position_embeddings": true,
+  "transformers_version": "4.57.6",
+  "use_cache": true,
+  "vocab_size": 59994
+}

opus-mt-en-tr/decoder_model_merged_quantized.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d059b778ef546ac536145008921b729253856d4a60344573d12215987a432b03
+size 57363154

opus-mt-en-tr/decoder_model_quantized.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:063c97138c22005a0f0a513d3d9f8c2b178f5180140b4a6721d93f7833bd7082
+size 57010861

opus-mt-en-tr/decoder_with_past_model_quantized.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b998e9b22842fb367c518fb5019ab935cfba7a9dc55adeca0ce9cdeca0f61092
+size 53799241

opus-mt-en-tr/encoder_model_quantized.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c4e6a035d2a4190ebbd4b9b1abcf8a7b7486649c6be1bebadac56626773d78c
+size 50172996

opus-mt-en-tr/generation_config.json

@@ -0,0 +1,16 @@
+{
+  "bad_words_ids": [
+    [
+      59993
+    ]
+  ],
+  "bos_token_id": 0,
+  "decoder_start_token_id": 59993,
+  "eos_token_id": 0,
+  "forced_eos_token_id": 0,
+  "max_length": 512,
+  "num_beams": 6,
+  "pad_token_id": 59993,
+  "renormalize_logits": true,
+  "transformers_version": "4.57.6"
+}

opus-mt-en-tr/ort_config.json

@@ -0,0 +1,33 @@
+{
+  "one_external_file": true,
+  "opset": null,
+  "optimization": {},
+  "quantization": {
+    "activations_dtype": "QUInt8",
+    "activations_symmetric": false,
+    "format": "QOperator",
+    "is_static": false,
+    "mode": "IntegerOps",
+    "nodes_to_exclude": [],
+    "nodes_to_quantize": [],
+    "operators_to_quantize": [
+      "Conv",
+      "MatMul",
+      "Attention",
+      "LSTM",
+      "Gather",
+      "Transpose",
+      "EmbedLayerNormalization"
+    ],
+    "per_channel": false,
+    "qdq_add_pair_to_weight": false,
+    "qdq_dedicated_pair": false,
+    "qdq_op_type_per_channel_support_to_axis": {
+      "MatMul": 1
+    },
+    "reduce_range": false,
+    "weights_dtype": "QUInt8",
+    "weights_symmetric": true
+  },
+  "use_external_data_format": false
+}

opus-mt-en-tr/source.spm

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:98eb24f0995a9d5f7cb0fb628c474628b1d2284615e881e857d062c0b651ce10
+size 793920

opus-mt-en-tr/special_tokens_map.json

@@ -0,0 +1,5 @@
+{
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
+}

opus-mt-en-tr/target.spm

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45cc6000ed513cdca8f80739087fbcbf9933dc50c9ae36c319c9670882f72e1b
+size 837876

opus-mt-en-tr/tokenizer_config.json

@@ -0,0 +1,39 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "59993": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "separate_vocabs": false,
+  "source_lang": "eng",
+  "sp_model_kwargs": {},
+  "target_lang": "tur",
+  "tokenizer_class": "MarianTokenizer",
+  "unk_token": "<unk>"
+}