Self-contained M3: weights, tokenizer, ONNX fp32/fp16/int8

.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
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

config.json

@@ -0,0 +1,30 @@
+{
+  "add_cross_attention": false,
+  "architectures": [
+    "XLMRobertaModel"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "dtype": "float32",
+  "eos_token_id": 2,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "is_decoder": false,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "xlm-roberta",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "output_past": true,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "tie_word_embeddings": true,
+  "transformers_version": "5.17.0",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 250002
+}

m3_config.json

@@ -2,7 +2,6 @@
   "model": "DibaOne-M3",
   "model_version": "m3",
   "architecture": "bi-encoder (shared encoder, cosine interaction)",
-  "backbone_init": "intfloat/multilingual-e5-base",
   "params_m": 278.0,
   "hidden_size": 768,
   "max_length": 128,
@@ -39,5 +38,13 @@
     "embeddings": "uint8",
     "calibration": "3,000 random rows of the v3 dev split"
   },
-  "tokenizer": "intfloat/multilingual-e5-base"
+  "files": {
+    "weights": "model.safetensors",
+    "tokenizer": "tokenizer.json",
+    "onnx": {
+      "fp32": "onnx/model.onnx",
+      "fp16": "onnx/model_fp16.onnx",
+      "int8": "onnx/model_int8.onnx"
+    }
+  }
 }

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b534f956ad8dc850e3a78f1ff53623d20c6a871a7bfade764ffec9f3f6ede63d
+size 1112197096

onnx/model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ca78b680ff8214e3c319a7422130f2a27c1cbb3006af6476016f26477146545
+size 1110043787

onnx/model_fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d250d5da45f596f2e51f5b2c75d141f755eb4d5c8890a6b0fa0d0cf6eebaf3e
+size 555205948

onnx/model_int8.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c02171af8833743e2370b2138edb0ff27cd8a08fed4204e5b85ee8664d0a84a5
+size 364972248

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cad551d5600a84242d0973327029452a1e3672ba6313c2a3c3d69c4310e12719
+size 17082987

tokenizer_config.json

@@ -0,0 +1,16 @@
+{
+  "add_prefix_space": true,
+  "backend": "tokenizers",
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "<mask>",
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "XLMRobertaTokenizer",
+  "unk_token": "<unk>"
+}