Init commit.

1_Pooling/config.json

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

README.md

@@ -1,33 +1,129 @@
 ---
-language: 
-  - en
-datasets:
-- SNLI
-- MNLI
+pipeline_tag: sentence-similarity
 tags:
-- zero-shot-classification
+- sentence-transformers
+- feature-extraction
+- sentence-similarity
+- transformers
 ---
 
+# {MODEL_NAME}
 
-A cross attention NLI model trained for zero-shot and few-shot text classification.
+This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.
 
-The base model is [mpnet-base](https://huggingface.co/microsoft/mpnet-base), trained with the code from [here](https://github.com/facebookresearch/anli).
-on [SNLI](https://nlp.stanford.edu/projects/snli/) and [MNLI](https://cims.nyu.edu/~sbowman/multinli/).
+<!--- Describe your model here -->
 
-Usage:
+## Usage (Sentence-Transformers)
+
+Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
+
+```
+pip install -U sentence-transformers
+```
+
+Then you can use the model like this:
 
 ```python
-from transformers import AutoModelForSequenceClassification, AutoTokenizer
+from sentence_transformers import SentenceTransformer
+sentences = ["This is an example sentence", "Each sentence is converted"]
+
+model = SentenceTransformer('{MODEL_NAME}')
+embeddings = model.encode(sentences)
+print(embeddings)
+```
+
+
+
+## Usage (HuggingFace Transformers)
+Without [sentence-transformers](https://www.SBERT.net), you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.
+
+```python
+from transformers import AutoTokenizer, AutoModel
 import torch
-import numpy as np
 
-model = AutoModelForSequenceClassification.from_pretrained("symanto/mpnet-base-snli-mnli")
-tokenizer = AutoTokenizer.from_pretrained("symanto/mpnet-base-snli-mnli")
 
-input_pairs = [("I like this pizza.", "The sentence is positive."), ("I like this pizza.", "The sentence is negative.")]
-inputs = tokenizer(["</s></s>".join(input_pair) for input_pair in input_pairs], return_tensors="pt")
-logits = model(**inputs).logits
-probs =  torch.softmax(logits, dim=1).tolist()
-print("probs", probs)
-np.testing.assert_almost_equal(probs, [[0.86, 0.14, 0.00], [0.16, 0.15, 0.69]], decimal=2)
+#Mean Pooling - Take attention mask into account for correct averaging
+def mean_pooling(model_output, attention_mask):
+    token_embeddings = model_output[0] #First element of model_output contains all token embeddings
+    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
+    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
+
+
+# Sentences we want sentence embeddings for
+sentences = ['This is an example sentence', 'Each sentence is converted']
+
+# Load model from HuggingFace Hub
+tokenizer = AutoTokenizer.from_pretrained('{MODEL_NAME}')
+model = AutoModel.from_pretrained('{MODEL_NAME}')
+
+# Tokenize sentences
+encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
+
+# Compute token embeddings
+with torch.no_grad():
+    model_output = model(**encoded_input)
+
+# Perform pooling. In this case, max pooling.
+sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
+
+print("Sentence embeddings:")
+print(sentence_embeddings)
 ```
+
+
+
+## Evaluation Results
+
+<!--- Describe how your model was evaluated -->
+
+For an automated evaluation of this model, see the *Sentence Embeddings Benchmark*: [https://seb.sbert.net](https://seb.sbert.net?model_name={MODEL_NAME})
+
+
+## Training
+The model was trained with the parameters:
+
+**DataLoader**:
+
+`zsde.training.NoDuplicatesDataLoader` of length 75000 with parameters:
+```
+{'batch_size': 16}
+```
+
+**Loss**:
+
+`sentence_transformers.losses.MultipleNegativesRankingLoss.MultipleNegativesRankingLoss` with parameters:
+  ```
+  {'scale': 20.0, 'similarity_fct': 'cos_sim'}
+  ```
+
+Parameters of the fit()-Method:
+```
+{
+    "callback": null,
+    "epochs": 1,
+    "evaluation_steps": 7500,
+    "evaluator": "sentence_transformers.evaluation.EmbeddingSimilarityEvaluator.EmbeddingSimilarityEvaluator",
+    "max_grad_norm": 1,
+    "optimizer_class": "<class 'transformers.optimization.AdamW'>",
+    "optimizer_params": {
+        "lr": 2e-05
+    },
+    "scheduler": "WarmupLinear",
+    "steps_per_epoch": 75000,
+    "warmup_steps": 7500,
+    "weight_decay": 0.01
+}
+```
+
+
+## Full Model Architecture
+```
+SentenceTransformer(
+  (0): Transformer({'max_seq_length': 128, 'do_lower_case': False}) with Transformer model: MPNetModel 
+  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
+)
+```
+
+## Citing & Authors
+
+<!--- Describe where people can find more information -->

config.json

@@ -1,7 +1,7 @@
 {
-  "_name_or_path": "symanto/mpnet-base-snli-mnli",
+  "_name_or_path": "microsoft/mpnet-base",
   "architectures": [
-    "MPNetForSequenceClassification"
+    "MPNetModel"
   ],
   "attention_probs_dropout_prob": 0.1,
   "bos_token_id": 0,
@@ -9,18 +9,8 @@
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.1,
   "hidden_size": 768,
-  "id2label": {
-    "0": "ENTAILMENT",
-    "1": "NEUTRAL",
-    "2": "CONTRADICTION"
-  },
   "initializer_range": 0.02,
   "intermediate_size": 3072,
-  "label2id": {
-    "ENTAILMENT": 0,
-    "NEUTRAL": 1,
-    "CONTRADICTION": 2
-  },
   "layer_norm_eps": 1e-05,
   "max_position_embeddings": 514,
   "model_type": "mpnet",
@@ -28,7 +18,6 @@
   "num_hidden_layers": 12,
   "pad_token_id": 1,
   "relative_attention_num_buckets": 32,
-  "torch_dtype": "float32",
-  "transformers_version": "4.11.1",
+  "transformers_version": "4.6.0",
   "vocab_size": 30527
 }

config_sentence_transformers.json

@@ -0,0 +1,7 @@
+{
+  "__version__": {
+    "sentence_transformers": "2.0.0",
+    "transformers": "4.6.0",
+    "pytorch": "1.7.0"
+  }
+}

modules.json

@@ -0,0 +1,14 @@
+[
+  {
+    "idx": 0,
+    "name": "0",
+    "path": "",
+    "type": "sentence_transformers.models.Transformer"
+  },
+  {
+    "idx": 1,
+    "name": "1",
+    "path": "1_Pooling",
+    "type": "sentence_transformers.models.Pooling"
+  }
+]

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f624388b921afb0de61fe012a99b52069b9743f8f7266696d8657cfa9fa57d12
-size 438035181
+oid sha256:1c83f1bfd69577fa7210ba31a94b019b6a3f0112a2aa79a99bff92261a92cecf
+size 438028807

sentence_bert_config.json

@@ -0,0 +1,4 @@
+{
+  "max_seq_length": 128,
+  "do_lower_case": false
+}

tokenizer_config.json

@@ -1 +1 @@
-{"do_lower_case": true, "bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "[UNK]", "pad_token": "<pad>", "mask_token": "<mask>", "tokenize_chinese_chars": true, "strip_accents": null, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "microsoft/mpnet-base", "tokenizer_class": "MPNetTokenizer"}
+{"do_lower_case": true, "bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "[UNK]", "pad_token": "<pad>", "mask_token": "<mask>", "tokenize_chinese_chars": true, "strip_accents": null, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "microsoft/mpnet-base"}