Duplicate from RisticksAI/Zaglyt2-transformer-test

Co-authored-by: - - - <ierhon@users.noreply.huggingface.co>

.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

README.md

@@ -0,0 +1,13 @@
+---
+title: Zaglyt2 Transformer Test
+emoji: 🚀
+colorFrom: pink
+colorTo: purple
+sdk: gradio
+sdk_version: 3.33.1
+app_file: app.py
+pinned: false
+duplicated_from: RisticksAI/Zaglyt2-transformer-test
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,14 @@
+import gradio as gr
+import net
+
+def generate(text):
+    o = text
+    r = []
+    for i in range(5):
+        t = net.gen(o)
+        o += " " + t
+        r.append(t)
+    return text + " *"+' '.join(r)+"*"
+
+iface = gr.Interface(fn=generate, inputs="text", outputs="text")
+iface.launch()

m_conf.py

@@ -0,0 +1,3 @@
+input_length = 20
+emb_dim = 128
+emb_o_dim = 256

net.py

@@ -0,0 +1,82 @@
+import word_emb
+from m_conf import *
+import numpy as np
+from gensim.models import Word2Vec
+from tensorflow.keras.models import Sequential
+from tensorflow.keras.layers import Dense, Dropout, Flatten, Embedding
+from keras_self_attention import SeqSelfAttention, SeqWeightedAttention
+from tensorflow.keras.optimizers import Adam
+from tensorflow.keras.preprocessing.text import Tokenizer
+from tensorflow.keras.losses import MeanSquaredError
+from tensorflow.keras.preprocessing.sequence import pad_sequences
+
+w2v = Word2Vec.load("w2v.model")
+
+# загрузка датасета
+with open('train.txt', 'r') as file:
+    text = file.readlines()
+
+# создание Tokenizerа
+tokenizer = Tokenizer()
+# обучение Tokenizer на основе текста из train.txt
+tokenizer.fit_on_texts(text)
+
+# преобразование текстовых данных в последовательности целых чисел с помощью объекта tokenizer
+tt = tokenizer.texts_to_sequences(text)
+
+t_sw = [[line[i:i+input_length] for i in range(len(line))] for line in tt]
+
+combined_list = []
+
+for line in t_sw:
+    combined_list.extend(line)
+
+y_t = [[w2v.wv[str(token)] for token in line] for line in tt]
+
+y = []
+for line in y_t:
+    y.extend(line)
+
+# задать длинну входа до переменной input_length, заполняя пустоту нулями
+X = pad_sequences(combined_list, maxlen=input_length, padding='pre')
+
+# получаем количество токенов в тексте
+vocab_size = len(tokenizer.word_index)
+
+# создание модели машинного обучения и задание её параметров
+model = Sequential()
+emb = Embedding(input_dim=vocab_size+1, output_dim=emb_dim, input_length=input_length)
+model.add(emb)
+model.add(SeqWeightedAttention())
+model.add(Flatten())
+model.add(Dense(512, activation="tanh"))
+model.add(Dropout(0.5))
+model.add(Dense(256, activation="tanh"))
+model.add(Dropout(0.5))
+model.add(Dense(128, activation="tanh"))
+model.add(Dense(emb_o_dim, activation="tanh"))
+
+# компилирование модели с функцией потерь mse и отображением accuracy
+model.compile(optimizer=Adam(learning_rate=0.001), loss="mse", metrics=["accuracy"])
+
+# обучение модели
+set_limit = 2000
+model.fit(np.array(X[:set_limit]), np.array(y[:set_limit]), epochs=10, batch_size=4)
+
+def find_closest_token(o, temperature=0.0, top_p=1):
+    token_distances = []
+    for token in w2v.wv.index_to_key:
+        vector = w2v.wv[token]
+        distance = np.sum((o - vector)**2)
+        token_distances.append((token, distance))
+
+    token_distances = sorted(token_distances, key=lambda x: x[1])
+    closest_token = token_distances[0][0]
+
+    return closest_token
+
+def gen(text):
+    # преобразовать текст в понимаемую нейросетью информацию
+    inp = pad_sequences(tokenizer.texts_to_sequences([text]), maxlen=input_length, padding='pre')
+    # сделать предположение и его возвратить
+    return str(tokenizer.index_word[int(find_closest_token(model.predict(inp)[0]))])

requirements.txt

@@ -0,0 +1,5 @@
+numpy
+gensim
+tensorflow
+keras
+keras_self_attention

word_emb.py

@@ -0,0 +1,15 @@
+from m_conf import *
+from keras.preprocessing.text import Tokenizer
+from gensim.models import Word2Vec
+
+with open('train.txt', 'r') as file:
+    lines = file.readlines()
+
+tokenizer = Tokenizer()
+tokenizer.fit_on_texts(lines)
+sequences = tokenizer.texts_to_sequences(lines)
+tokens = [[str(i) for i in seq] for seq in sequences]
+
+model = Word2Vec(tokens, window=3, min_count=1, vector_size=emb_o_dim)
+
+model.save("w2v.model")