making readme more human-friendly

README.md

@@ -16,122 +16,39 @@ widget:
 
 # SBERT_PQ
 
-This is a [sentence-transformers](https://www.SBERT.net) model: It maps texts & questions to a 312 dimensional dense vector space and can be used for tasks like clustering or semantic search.
+Это [sentence-transformers](https://www.SBERT.net) модель, предназначенная
+для определения релевантности короткого текста (преимущественно 1 предложение до 10-15 слов) и вопроса.
 
-<!--- Describe your model here -->
+Модель вычисляет для текста и вопроса векторы размерностью 312. Косинус угла между этими векторами
+дает оценку того, содержит ли текст ответ на заданный вопрос. В [проекте диалоговой системы](https://github.com/Koziev/chatbot) 
+она используется для семантического поиска записей в базе фактов по заданному собеседником вопросу.
 
-## Usage (Sentence-Transformers)
+Модель основана на [cointegrated/rubert-tiny2]. Она имеет очень небольшой размер и быстро выполняет инференс даже на CPU.
 
-Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
 
-```
-pip install -U sentence-transformers
-```
+## Использование с библиотекой (Sentence-Transformers)
 
-Then you can use the model like this:
-
-```python
-from sentence_transformers import SentenceTransformer
-sentences = ["Кошка ловит мышку.", "Чем занята кошка?"]
+Для удобства установите [sentence-transformers](https://www.SBERT.net):
 
-model = SentenceTransformer('inkoziev/sbert_pq')
-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
-
-
-#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('inkoziev/sbert_pq')
-model = AutoModel.from_pretrained('inkoziev/sbert_pq')
-
-# 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)
+pip install -U sentence-transformers
 ```
 
+Чтобы определить релевантность в одной паре "текст-вопрос", можно использовать такой код:
 
-
-## 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**:
-
-`torch.utils.data.dataloader.DataLoader` of length 2320 with parameters:
-```
-{'batch_size': 2048, 'sampler': 'torch.utils.data.sampler.RandomSampler', 'batch_sampler': 'torch.utils.data.sampler.BatchSampler'}
 ```
+import sentence_transformers
 
-**Loss**:
-
-`sentence_transformers.losses.ContrastiveLoss.ContrastiveLoss` with parameters:
-  ```
-  {'distance_metric': 'SiameseDistanceMetric.COSINE_DISTANCE', 'margin': 0.5, 'size_average': True}
-  ```
-
-Parameters of the fit()-Method:
-```
-{
-    "callback": null,
-    "epochs": 10,
-    "evaluation_steps": 200,
-    "evaluator": "sentence_transformers.evaluation.BinaryClassificationEvaluator.BinaryClassificationEvaluator",
-    "max_grad_norm": 1,
-    "optimizer_class": "<class 'transformers.optimization.AdamW'>",
-    "optimizer_params": {
-        "lr": 2e-05
-    },
-    "scheduler": "WarmupLinear",
-    "steps_per_epoch": null,
-    "warmup_steps": 2320,
-    "weight_decay": 1e-05
-}
-```
+sentences = ["Кошка ловит мышку.", "Чем занята кошка?"]
 
+model = sentence_transformers.SentenceTransformer('inkoziev/sbert_pq')
+embeddings = model.encode(sentences)
 
-## Full Model Architecture
-```
-SentenceTransformer(
-  (0): Transformer({'max_seq_length': 2048, 'do_lower_case': False}) with Transformer model: BertModel 
-  (1): Pooling({'word_embedding_dimension': 312, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
-)
+s = sentence_transformers.util.cos_sim(a=embeddings[0], b=embeddings[1])
+print('text={} qquestion={} cossim={}'.format(sentences[0], sentences[1], s))
 ```
 
-## Citing & Authors
+## Контакты и цитирование
 
 ```
 @MISC{rugpt_chitchat,
@@ -141,4 +58,3 @@ SentenceTransformer(
     year    = 2022
 }
 ```
-