LaBSE

.gitattributes

@@ -33,3 +33,15 @@ 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
+LaBSE-onnx/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+LABSE-Vit-L-14/textual/rknpu/rk3566/model.rknn filter=lfs diff=lfs merge=lfs -text
+LABSE-Vit-L-14/textual/rknpu/rk3568/model.rknn filter=lfs diff=lfs merge=lfs -text
+LABSE-Vit-L-14/textual/rknpu/rk3576/model.rknn filter=lfs diff=lfs merge=lfs -text
+LABSE-Vit-L-14/textual/rknpu/rk3588/model.rknn filter=lfs diff=lfs merge=lfs -text
+LABSE-Vit-L-14/textual/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+LABSE-Vit-L-14/visual/rknpu/rk3566/model.rknn filter=lfs diff=lfs merge=lfs -text
+LABSE-Vit-L-14/visual/rknpu/rk3568/model.rknn filter=lfs diff=lfs merge=lfs -text
+LABSE-Vit-L-14/visual/rknpu/rk3576/model.rknn filter=lfs diff=lfs merge=lfs -text
+LABSE-Vit-L-14/visual/rknpu/rk3588/model.rknn filter=lfs diff=lfs merge=lfs -text
+Language-agnostic[[:space:]]BERT[[:space:]]Sentence[[:space:]]Embedding.pdf filter=lfs diff=lfs merge=lfs -text
+sbert-LaBSE-onnx/tokenizer.json filter=lfs diff=lfs merge=lfs -text

LABSE-Vit-L-14/.gitattributes

@@ -0,0 +1,143 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz 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
+textual/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+visual/Constant_1562_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_1600_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_1632_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_1745_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_1764_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_1772_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_1801_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_1914_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_1933_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_1941_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_1970_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2083_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2102_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2110_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2139_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2252_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2271_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2279_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2308_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2421_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2440_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2448_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2477_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2590_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2609_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2617_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2646_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2759_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2778_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2786_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2815_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2928_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2947_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2955_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_2984_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3097_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3116_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3124_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3153_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3266_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3285_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3293_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3322_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3435_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3454_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3462_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3491_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3604_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3623_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3631_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3660_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3773_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3792_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3800_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3829_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3942_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3961_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3969_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_3998_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4111_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4130_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4138_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4167_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4280_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4299_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4307_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4336_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4449_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4468_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4476_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4505_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4618_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4637_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4645_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4674_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4787_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4806_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4814_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4843_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4956_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4975_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_4983_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5012_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5125_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5144_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5152_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5181_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5294_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5313_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5321_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5350_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5463_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5482_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5490_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5519_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5632_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5651_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5659_attr__value filter=lfs diff=lfs merge=lfs -text
+visual/Constant_5670_attr__value filter=lfs diff=lfs merge=lfs -text
+textual/rknpu/rk3566/model.rknn filter=lfs diff=lfs merge=lfs -text
+textual/rknpu/rk3576/model.rknn filter=lfs diff=lfs merge=lfs -text
+textual/rknpu/rk3588/model.rknn filter=lfs diff=lfs merge=lfs -text
+visual/rknpu/rk3566/model.rknn filter=lfs diff=lfs merge=lfs -text
+visual/rknpu/rk3576/model.rknn filter=lfs diff=lfs merge=lfs -text
+visual/rknpu/rk3588/model.rknn filter=lfs diff=lfs merge=lfs -text
+textual/rknpu/rk3568/model.rknn filter=lfs diff=lfs merge=lfs -text
+visual/rknpu/rk3568/model.rknn filter=lfs diff=lfs merge=lfs -text

LABSE-Vit-L-14/README.md

@@ -0,0 +1,11 @@
+---
+tags:
+ - immich
+ - clip
+ - multilingual
+---
+# Model Description
+
+This repo contains ONNX exports for the associated CLIP model by M-CLIP. See the [M-CLIP](https://huggingface.co/M-CLIP) repo for more info.
+
+This repo is specifically intended for use with [Immich](https://immich.app/), a self-hosted photo library.

LABSE-Vit-L-14/config.json

@@ -0,0 +1 @@
+{"embed_dim": 768, "vision_cfg": {"image_size": 224, "layers": 24, "width": 1024, "patch_size": 14}, "text_cfg": {"context_length": 77, "vocab_size": 49408, "width": 768, "heads": 12, "layers": 12}}

LABSE-Vit-L-14/source.txt

@@ -0,0 +1 @@
+https://huggingface.co/immich-app/LABSE-Vit-L-14

LABSE-Vit-L-14/textual/model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77333b8ed51cb501b0217c172c4c75bdd37ee15915335f3242dae063bf0f82b4
+size 1883874543

LABSE-Vit-L-14/textual/rknpu/rk3566/model.rknn

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f942de9726c58fea2b153a24f24f061e797af72a07ef366cdb6c479261b5a2b
+size 945356824

LABSE-Vit-L-14/textual/rknpu/rk3568/model.rknn

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5bad1a83b0d6ad86f66409365694b36d9a327bcf8a9fe2090e97f18004a28184
+size 945356824

LABSE-Vit-L-14/textual/rknpu/rk3576/model.rknn

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:036857011e24e80fd2df6398ac332ead044e6a96b0474d739b37321b30c1d926
+size 950513624

LABSE-Vit-L-14/textual/rknpu/rk3588/model.rknn

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3aed2ead635849e9a7ef90beb8f4f039dddef5bf8a66c0fb9f6ebdd2188bb9d3
+size 944952920

LABSE-Vit-L-14/textual/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
+  }
+}

LABSE-Vit-L-14/textual/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:09216b42d2697b7b4a26ac05ff09ba8bf52dc19b896c5ceee8bbff9f39055322
+size 13631919

LABSE-Vit-L-14/textual/tokenizer_config.json

@@ -0,0 +1,59 @@
+{
+  "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": false,
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": false,
+  "extra_special_tokens": {},
+  "full_tokenizer_file": null,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

LABSE-Vit-L-14/visual/model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b02d572f59c509f4b97b9c54a868453cca1a652cd5d60e1d51d0052f055cb8c
+size 1216297719

LABSE-Vit-L-14/visual/preprocess_cfg.json

@@ -0,0 +1 @@
+{"size": [224, 224], "mode": "RGB", "mean": [0.48145466, 0.4578275, 0.40821073], "std": [0.26862954, 0.26130258, 0.27577711], "interpolation": "bicubic", "resize_mode": "shortest", "fill_color": 0}

LABSE-Vit-L-14/visual/rknpu/rk3566/model.rknn

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec804bed50a21bd3774271613eb8479ad847cee6b4211a1e41c79c3ef5853f7e
+size 644018803

LABSE-Vit-L-14/visual/rknpu/rk3568/model.rknn

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfd435401c68bfe4942017528d7cd1ea096e87d9fa934e7aa1e648687b7e5a64
+size 644018803

LABSE-Vit-L-14/visual/rknpu/rk3576/model.rknn

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95675ecdd458ae23484e301cfae400f647d042eb77ccf604315ab879091390a9
+size 645072115

LABSE-Vit-L-14/visual/rknpu/rk3588/model.rknn

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef6067e14f92f4025eac000a7552309784288c1d3c75cccbed2c0d2334aff54b
+size 626315187

LaBSE-en-ru/.gitattributes

@@ -0,0 +1,17 @@
+*.bin.* filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tar.gz filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+model.safetensors filter=lfs diff=lfs merge=lfs -text

LaBSE-en-ru/1_Pooling/config.json

@@ -0,0 +1,7 @@
+{
+  "word_embedding_dimension": 768,
+  "pooling_mode_cls_token": true,
+  "pooling_mode_mean_tokens": false,
+  "pooling_mode_max_tokens": false,
+  "pooling_mode_mean_sqrt_len_tokens": false
+}

LaBSE-en-ru/2_Dense/config.json

@@ -0,0 +1 @@
+{"in_features": 768, "out_features": 768, "bias": true, "activation_function": "torch.nn.modules.activation.Tanh"}

LaBSE-en-ru/2_Dense/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06fb85120e40adf0ab188c4f0cc7684f702cb2023532947d1b85f325b0a3645c
+size 2363431

LaBSE-en-ru/README.md

@@ -0,0 +1,35 @@
+---
+language: ["ru", "en"]
+tags:
+- feature-extraction
+- embeddings
+- sentence-similarity
+---
+# LaBSE for English and Russian
+This is a truncated version of [sentence-transformers/LaBSE](https://huggingface.co/sentence-transformers/LaBSE), which is, in turn, a port of [LaBSE](https://tfhub.dev/google/LaBSE/1) by Google.
+
+The current model has only English and Russian tokens left in the vocabulary.
+Thus, the vocabulary is 10% of the original, and number of parameters in the whole model is 27% of the original, without any loss in the quality of English and Russian embeddings.
+ 
+To get the sentence embeddings, you can  use the following code:
+```python
+import torch
+from transformers import AutoTokenizer, AutoModel
+tokenizer = AutoTokenizer.from_pretrained("cointegrated/LaBSE-en-ru")
+model = AutoModel.from_pretrained("cointegrated/LaBSE-en-ru")
+sentences = ["Hello World", "Привет Мир"]
+encoded_input = tokenizer(sentences, padding=True, truncation=True, max_length=64, return_tensors='pt')
+with torch.no_grad():
+    model_output = model(**encoded_input)
+embeddings = model_output.pooler_output
+embeddings = torch.nn.functional.normalize(embeddings)
+print(embeddings)
+```
+
+The model has been truncated in [this notebook](https://colab.research.google.com/drive/1dnPRn0-ugj3vZgSpyCC9sgslM2SuSfHy?usp=sharing).
+You can adapt it for other languages (like [EIStakovskii/LaBSE-fr-de](https://huggingface.co/EIStakovskii/LaBSE-fr-de)), models or datasets.
+
+## Reference:
+Fangxiaoyu Feng, Yinfei Yang, Daniel Cer, Narveen Ari, Wei Wang. [Language-agnostic BERT Sentence Embedding](https://arxiv.org/abs/2007.01852). July 2020
+
+License: [https://tfhub.dev/google/LaBSE/1](https://tfhub.dev/google/LaBSE/1)

LaBSE-en-ru/config.json

@@ -0,0 +1,30 @@
+{
+  "_name_or_path": "cointegrated/LaBSE-en-ru",
+  "architectures": [
+    "BertForPreTraining"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "directionality": "bidi",
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "pooler_fc_size": 768,
+  "pooler_num_attention_heads": 12,
+  "pooler_num_fc_layers": 3,
+  "pooler_size_per_head": 128,
+  "pooler_type": "first_token_transform",
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.5.1",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 55083
+}

LaBSE-en-ru/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25678be7d335a05ceec55a820c49e9ec0d637c0d469ef09d5f7155ef42a41bd3
+size 516003632

LaBSE-en-ru/modules.json

@@ -0,0 +1,26 @@
+[
+  {
+    "idx": 0,
+    "name": "0",
+    "path": "",
+    "type": "sentence_transformers.models.Transformer"
+  },
+  {
+    "idx": 1,
+    "name": "1",
+    "path": "1_Pooling",
+    "type": "sentence_transformers.models.Pooling"
+  },
+  {
+    "idx": 2,
+    "name": "2",
+    "path": "2_Dense",
+    "type": "sentence_transformers.models.Dense"
+  },
+  {
+    "idx": 3,
+    "name": "3",
+    "path": "3_Normalize",
+    "type": "sentence_transformers.models.Normalize"
+  }
+]

LaBSE-en-ru/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d929e16c4cc9b40cdd96219e8ce3c1084129798435b3c67212efd68fa018673b
+size 516063655

LaBSE-en-ru/sentence_bert_config.json

@@ -0,0 +1,4 @@
+{
+  "max_seq_length": 512,
+  "do_lower_case": false
+}

LaBSE-en-ru/source.txt

@@ -0,0 +1 @@
+https://huggingface.co/cointegrated/LaBSE-en-ru

LaBSE-en-ru/special_tokens_map.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}

LaBSE-en-ru/tf_model.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78711a9ec0426e02e001f890ac237d7cb0022cf14941dc45495019a595effa7e
+size 687064460

LaBSE-en-ru/tokenizer_config.json

@@ -0,0 +1 @@
+{"do_lower_case": false, "model_max_length": 512}

LaBSE-fr-de/.gitattributes

@@ -0,0 +1,34 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz 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

LaBSE-fr-de/README.md

@@ -0,0 +1,31 @@
+---
+language: ["fr", "de"]
+tags:
+- feature-extraction
+- embeddings
+- sentence-similarity
+---
+# LaBSE for French and German
+This is a shortened version of [sentence-transformers/LaBSE](https://huggingface.co/sentence-transformers/LaBSE). The model was prepaired with the direct help of [cointegrated](https://huggingface.co/cointegrated), the author of the [LaBSE-en-ru model](https://huggingface.co/cointegrated/LaBSE-en-ru).
+
+The current model includes only French and German tokens, and the vocabulary is thus 10% of the original while number of parameters in the whole model is 27% of the original.
+ 
+To get the sentence embeddings, you can  use the following code:
+```python
+import torch
+from transformers import AutoTokenizer, AutoModel
+tokenizer = AutoTokenizer.from_pretrained("EIStakovskii/LaBSE-fr-de")
+model = AutoModel.from_pretrained("EIStakovskii/LaBSE-fr-de")
+sentences = ["Wie geht es dir?", "Comment vas-tu?"]
+encoded_input = tokenizer(sentences, padding=True, truncation=True, max_length=64, return_tensors='pt')
+with torch.no_grad():
+    model_output = model(**encoded_input)
+embeddings = model_output.pooler_output
+embeddings = torch.nn.functional.normalize(embeddings)
+print(embeddings)
+```
+
+## Reference:
+Fangxiaoyu Feng, Yinfei Yang, Daniel Cer, Narveen Ari, Wei Wang. [Language-agnostic BERT Sentence Embedding](https://arxiv.org/abs/2007.01852). July 2020
+
+License: [https://tfhub.dev/google/LaBSE/1](https://tfhub.dev/google/LaBSE/1)

LaBSE-fr-de/config.json

@@ -0,0 +1,32 @@
+{
+  "_name_or_path": "sentence-transformers/LaBSE",
+  "architectures": [
+    "BertForPreTraining"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "directionality": "bidi",
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "pooler_fc_size": 768,
+  "pooler_num_attention_heads": 12,
+  "pooler_num_fc_layers": 3,
+  "pooler_size_per_head": 128,
+  "pooler_type": "first_token_transform",
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.20.1",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 21754
+}

LaBSE-fr-de/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9d871e6e6336064e8260303b679b4da9fb530c3a853fe43a1443ccf5fbf5433
+size 413483584

LaBSE-fr-de/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2186e5f6fd3a8d517610234dcf8b7b4522767ad2a7ada93e48cad3d2da3dafbd
+size 413525283

LaBSE-fr-de/source.txt

@@ -0,0 +1 @@
+https://huggingface.co/EIStakovskii/LaBSE-fr-de

LaBSE-fr-de/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

LaBSE-fr-de/tokenizer_config.json

@@ -0,0 +1,17 @@
+{
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": false,
+  "full_tokenizer_file": null,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "name_or_path": "sentence-transformers/LaBSE",
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "special_tokens_map_file": "labse-pytorch/special_tokens_map.json",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

LaBSE-onnx/.gitattributes

@@ -0,0 +1,36 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz 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

LaBSE-onnx/1_Pooling/config.json

@@ -0,0 +1,7 @@
+{
+  "word_embedding_dimension": 768,
+  "pooling_mode_cls_token": true,
+  "pooling_mode_mean_tokens": false,
+  "pooling_mode_max_tokens": false,
+  "pooling_mode_mean_sqrt_len_tokens": false
+}

LaBSE-onnx/config.json

@@ -0,0 +1,31 @@
+{
+  "_name_or_path": "/Users/Shared/Libs/sentence_transformers/models--sentence-transformers--LaBSE/snapshots/e34fab64a3011d2176c99545a93d5cbddc9a91b7/",
+  "architectures": [
+    "BertModel"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "directionality": "bidi",
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "pooler_fc_size": 768,
+  "pooler_num_attention_heads": 12,
+  "pooler_num_fc_layers": 3,
+  "pooler_size_per_head": 128,
+  "pooler_type": "first_token_transform",
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.42.3",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 501153
+}

LaBSE-onnx/model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75801bb2ed33e25d24efe96f2bf7f7ef6d109778fc48ada8aa3b1fbd8969743c
+size 1881599307

LaBSE-onnx/source.txt

@@ -0,0 +1 @@
+https://huggingface.co/LightEmbed/LaBSE-onnx

LaBSE-onnx/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
+  }
+}