{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"\n",
"
Modelowanie Języka
\n",
" 10. Model rekurencyjny z atencją [ćwiczenia]
\n",
" Jakub Pokrywka (2022)
\n",
"\n",
"\n",
""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"notebook na podstawie:\n",
"\n",
"# https://pytorch.org/tutorials/intermediate/seq2seq_translation_tutorial.html"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"from __future__ import unicode_literals, print_function, division\n",
"from io import open\n",
"import unicodedata\n",
"import string\n",
"import re\n",
"import random\n",
"\n",
"import torch\n",
"import torch.nn as nn\n",
"from torch import optim\n",
"import torch.nn.functional as F\n",
"\n",
"device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"SOS_token = 0\n",
"EOS_token = 1\n",
"\n",
"class Lang:\n",
" def __init__(self):\n",
" self.word2index = {}\n",
" self.word2count = {}\n",
" self.index2word = {0: \"SOS\", 1: \"EOS\"}\n",
" self.n_words = 2 # Count SOS and EOS\n",
"\n",
" def addSentence(self, sentence):\n",
" for word in sentence.split(' '):\n",
" self.addWord(word)\n",
"\n",
" def addWord(self, word):\n",
" if word not in self.word2index:\n",
" self.word2index[word] = self.n_words\n",
" self.word2count[word] = 1\n",
" self.index2word[self.n_words] = word\n",
" self.n_words += 1\n",
" else:\n",
" self.word2count[word] += 1"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"# def unicodeToAscii(s):\n",
"# return ''.join(\n",
"# c for c in unicodedata.normalize('NFD', s)\n",
"# if unicodedata.category(c) != 'Mn'\n",
"# )"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"pairs = []\n",
"with open('data/eng-pol.txt') as f:\n",
" for line in f:\n",
" eng_line, pol_line = line.lower().rstrip().split('\\t')\n",
"\n",
" eng_line = re.sub(r\"([.!?])\", r\" \\1\", eng_line)\n",
" eng_line = re.sub(r\"[^a-zA-Z.!?]+\", r\" \", eng_line)\n",
"\n",
" pol_line = re.sub(r\"([.!?])\", r\" \\1\", pol_line)\n",
" pol_line = re.sub(r\"[^a-zA-Z.!?ąćęłńóśźżĄĆĘŁŃÓŚŹŻ]+\", r\" \", pol_line)\n",
" \n",
"# eng_line = unicodeToAscii(eng_line)\n",
"# pol_line = unicodeToAscii(pol_line)\n",
"\n",
" pairs.append([eng_line, pol_line])\n",
"\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['hi .', 'cześć .']"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"pairs[1]"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
"MAX_LENGTH = 10\n",
"eng_prefixes = (\n",
" \"i am \", \"i m \",\n",
" \"he is\", \"he s \",\n",
" \"she is\", \"she s \",\n",
" \"you are\", \"you re \",\n",
" \"we are\", \"we re \",\n",
" \"they are\", \"they re \"\n",
")\n",
"\n",
"pairs = [p for p in pairs if len(p[0].split(' ')) < MAX_LENGTH and len(p[1].split(' ')) < MAX_LENGTH]\n",
"pairs = [p for p in pairs if p[0].startswith(eng_prefixes)]\n",
"\n",
"eng_lang = Lang()\n",
"pol_lang = Lang()\n",
"\n",
"for pair in pairs:\n",
" eng_lang.addSentence(pair[0])\n",
" pol_lang.addSentence(pair[1])"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['i m ok .', 'ze mną wszystko w porządku .']"
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"pairs[0]"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['i m up .', 'wstałem .']"
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"pairs[1]"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['i m tom .', 'jestem tom .']"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"pairs[2]"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [],
"source": [
"class EncoderRNN(nn.Module):\n",
" def __init__(self, input_size, hidden_size):\n",
" super(EncoderRNN, self).__init__()\n",
" self.hidden_size = hidden_size\n",
"\n",
" self.embedding = nn.Embedding(input_size, hidden_size)\n",
" self.gru = nn.GRU(hidden_size, hidden_size)\n",
"\n",
" def forward(self, input, hidden):\n",
" embedded = self.embedding(input).view(1, 1, -1)\n",
" output = embedded\n",
" output, hidden = self.gru(output, hidden)\n",
" return output, hidden\n",
"\n",
" def initHidden(self):\n",
" return torch.zeros(1, 1, self.hidden_size, device=device)"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [],
"source": [
"class DecoderRNN(nn.Module):\n",
" def __init__(self, hidden_size, output_size):\n",
" super(DecoderRNN, self).__init__()\n",
" self.hidden_size = hidden_size\n",
"\n",
" self.embedding = nn.Embedding(output_size, hidden_size)\n",
" self.gru = nn.GRU(hidden_size, hidden_size)\n",
" self.out = nn.Linear(hidden_size, output_size)\n",
" self.softmax = nn.LogSoftmax(dim=1)\n",
"\n",
" def forward(self, input, hidden):\n",
" output = self.embedding(input).view(1, 1, -1)\n",
" output = F.relu(output)\n",
" output, hidden = self.gru(output, hidden)\n",
" output = self.softmax(self.out(output[0]))\n",
" return output, hidden\n",
"\n",
" def initHidden(self):\n",
" return torch.zeros(1, 1, self.hidden_size, device=device)"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [],
"source": [
"class AttnDecoderRNN(nn.Module):\n",
" def __init__(self, hidden_size, output_size, dropout_p=0.1, max_length=MAX_LENGTH):\n",
" super(AttnDecoderRNN, self).__init__()\n",
" self.hidden_size = hidden_size\n",
" self.output_size = output_size\n",
" self.dropout_p = dropout_p\n",
" self.max_length = max_length\n",
"\n",
" self.embedding = nn.Embedding(self.output_size, self.hidden_size)\n",
" self.attn = nn.Linear(self.hidden_size * 2, self.max_length)\n",
" self.attn_combine = nn.Linear(self.hidden_size * 2, self.hidden_size)\n",
" self.dropout = nn.Dropout(self.dropout_p)\n",
" self.gru = nn.GRU(self.hidden_size, self.hidden_size)\n",
" self.out = nn.Linear(self.hidden_size, self.output_size)\n",
"\n",
" def forward(self, input, hidden, encoder_outputs):\n",
" embedded = self.embedding(input).view(1, 1, -1)\n",
" embedded = self.dropout(embedded)\n",
"\n",
" attn_weights = F.softmax(\n",
" self.attn(torch.cat((embedded[0], hidden[0]), 1)), dim=1)\n",
" attn_applied = torch.bmm(attn_weights.unsqueeze(0),\n",
" encoder_outputs.unsqueeze(0))\n",
"\n",
" output = torch.cat((embedded[0], attn_applied[0]), 1)\n",
" output = self.attn_combine(output).unsqueeze(0)\n",
"\n",
" output = F.relu(output)\n",
" output, hidden = self.gru(output, hidden)\n",
"\n",
" output = F.log_softmax(self.out(output[0]), dim=1)\n",
" return output, hidden, attn_weights\n",
"\n",
" def initHidden(self):\n",
" return torch.zeros(1, 1, self.hidden_size, device=device)"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [],
"source": [
"def tensorFromSentence(sentence, lang):\n",
" indexes = [lang.word2index[word] for word in sentence.split(' ')]\n",
" indexes.append(EOS_token)\n",
" return torch.tensor(indexes, dtype=torch.long, device=device).view(-1, 1)\n"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [],
"source": [
"teacher_forcing_ratio = 0.5\n",
"\n",
"def train_one_batch(input_tensor, target_tensor, encoder, decoder, encoder_optimizer, decoder_optimizer, criterion, max_length=MAX_LENGTH):\n",
" encoder_hidden = encoder.initHidden()\n",
"\n",
" encoder_optimizer.zero_grad()\n",
" decoder_optimizer.zero_grad()\n",
"\n",
" input_length = input_tensor.size(0)\n",
" target_length = target_tensor.size(0)\n",
"\n",
" encoder_outputs = torch.zeros(max_length, encoder.hidden_size, device=device)\n",
"\n",
" loss = 0\n",
"\n",
" for ei in range(input_length):\n",
" encoder_output, encoder_hidden = encoder(input_tensor[ei], encoder_hidden)\n",
" encoder_outputs[ei] = encoder_output[0, 0]\n",
"\n",
" decoder_input = torch.tensor([[SOS_token]], device=device)\n",
"\n",
" decoder_hidden = encoder_hidden\n",
"\n",
" use_teacher_forcing = True if random.random() < teacher_forcing_ratio else False\n",
"\n",
" if use_teacher_forcing:\n",
" for di in range(target_length):\n",
" decoder_output, decoder_hidden, decoder_attention = decoder(decoder_input, decoder_hidden, encoder_outputs)\n",
" loss += criterion(decoder_output, target_tensor[di])\n",
" decoder_input = target_tensor[di] # Teacher forcing\n",
"\n",
" else:\n",
" for di in range(target_length):\n",
" decoder_output, decoder_hidden, decoder_attention = decoder(decoder_input, decoder_hidden, encoder_outputs)\n",
" topv, topi = decoder_output.topk(1)\n",
" decoder_input = topi.squeeze().detach() # detach from history as input\n",
"\n",
" loss += criterion(decoder_output, target_tensor[di])\n",
" if decoder_input.item() == EOS_token:\n",
" break\n",
"\n",
" loss.backward()\n",
"\n",
" encoder_optimizer.step()\n",
" decoder_optimizer.step()\n",
"\n",
" return loss.item() / target_length"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [],
"source": [
"def trainIters(encoder, decoder, n_iters, print_every=1000, learning_rate=0.01):\n",
" print_loss_total = 0 # Reset every print_every\n",
"\n",
" encoder_optimizer = optim.SGD(encoder.parameters(), lr=learning_rate)\n",
" decoder_optimizer = optim.SGD(decoder.parameters(), lr=learning_rate)\n",
" \n",
" training_pairs = [random.choice(pairs) for _ in range(n_iters)]\n",
" training_pairs = [(tensorFromSentence(p[0], eng_lang), tensorFromSentence(p[1], pol_lang)) for p in training_pairs]\n",
" \n",
" criterion = nn.NLLLoss()\n",
"\n",
" for i in range(1, n_iters + 1):\n",
" training_pair = training_pairs[i - 1]\n",
" input_tensor = training_pair[0]\n",
" target_tensor = training_pair[1]\n",
"\n",
" loss = train_one_batch(input_tensor,\n",
" target_tensor,\n",
" encoder,\n",
" decoder,\n",
" encoder_optimizer,\n",
" decoder_optimizer,\n",
" criterion)\n",
" \n",
" print_loss_total += loss\n",
"\n",
" if i % print_every == 0:\n",
" print_loss_avg = print_loss_total / print_every\n",
" print_loss_total = 0\n",
" print(f'iter: {i}, loss: {print_loss_avg}')\n"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [],
"source": [
"def evaluate(encoder, decoder, sentence, max_length=MAX_LENGTH):\n",
" with torch.no_grad():\n",
" input_tensor = tensorFromSentence(sentence, eng_lang)\n",
" input_length = input_tensor.size()[0]\n",
" encoder_hidden = encoder.initHidden()\n",
"\n",
" encoder_outputs = torch.zeros(max_length, encoder.hidden_size, device=device)\n",
"\n",
" for ei in range(input_length):\n",
" encoder_output, encoder_hidden = encoder(input_tensor[ei],\n",
" encoder_hidden)\n",
" encoder_outputs[ei] += encoder_output[0, 0]\n",
"\n",
" decoder_input = torch.tensor([[SOS_token]], device=device) # SOS\n",
"\n",
" decoder_hidden = encoder_hidden\n",
"\n",
" decoded_words = []\n",
" decoder_attentions = torch.zeros(max_length, max_length)\n",
"\n",
" for di in range(max_length):\n",
" decoder_output, decoder_hidden, decoder_attention = decoder(\n",
" decoder_input, decoder_hidden, encoder_outputs)\n",
" decoder_attentions[di] = decoder_attention.data\n",
" topv, topi = decoder_output.data.topk(1)\n",
" if topi.item() == EOS_token:\n",
" decoded_words.append('')\n",
" break\n",
" else:\n",
" decoded_words.append(pol_lang.index2word[topi.item()])\n",
"\n",
" decoder_input = topi.squeeze().detach()\n",
"\n",
" return decoded_words, decoder_attentions[:di + 1]"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [],
"source": [
"def evaluateRandomly(encoder, decoder, n=10):\n",
" for i in range(n):\n",
" pair = random.choice(pairs)\n",
" print('>', pair[0])\n",
" print('=', pair[1])\n",
" output_words, attentions = evaluate(encoder, decoder, pair[0])\n",
" output_sentence = ' '.join(output_words)\n",
" print('<', output_sentence)\n",
" print('')"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"iter: 50, loss: 5.242557571774437\n",
"iter: 100, loss: 4.278488328888303\n",
"iter: 150, loss: 4.345584976514179\n",
"iter: 200, loss: 4.372515664100646\n",
"iter: 250, loss: 4.305285846165248\n",
"iter: 300, loss: 4.340697079749336\n",
"iter: 350, loss: 4.225462563787189\n",
"iter: 400, loss: 3.931193191331531\n",
"iter: 450, loss: 4.255609704971314\n",
"iter: 500, loss: 4.110652269242301\n",
"iter: 550, loss: 4.10534066640763\n",
"iter: 600, loss: 4.21553361088889\n",
"iter: 650, loss: 4.182825716926937\n",
"iter: 700, loss: 4.125172647006929\n",
"iter: 750, loss: 3.92194454044766\n",
"iter: 800, loss: 3.9588410959697904\n",
"iter: 850, loss: 4.045722855431693\n",
"iter: 900, loss: 3.989374102009668\n",
"iter: 950, loss: 3.8188858143791307\n",
"iter: 1000, loss: 4.083652021286979\n",
"iter: 1050, loss: 3.853677998118931\n",
"iter: 1100, loss: 4.240768341064452\n",
"iter: 1150, loss: 3.9389620546991857\n",
"iter: 1200, loss: 3.797674254826137\n",
"iter: 1250, loss: 3.9533765572441957\n",
"iter: 1300, loss: 4.227734768761528\n",
"iter: 1350, loss: 4.098812954554482\n",
"iter: 1400, loss: 3.970981749352953\n",
"iter: 1450, loss: 3.961285910742623\n",
"iter: 1500, loss: 3.7287075826554075\n",
"iter: 1550, loss: 4.011922280311584\n",
"iter: 1600, loss: 3.819334566025507\n",
"iter: 1650, loss: 4.058212718903073\n",
"iter: 1700, loss: 3.7410677611335874\n",
"iter: 1750, loss: 3.6925343862261077\n",
"iter: 1800, loss: 3.686991301317063\n",
"iter: 1850, loss: 4.038253510641673\n",
"iter: 1900, loss: 3.6877949990999133\n",
"iter: 1950, loss: 3.6264906252452302\n",
"iter: 2000, loss: 3.9525268210607845\n",
"iter: 2050, loss: 3.6234626099268596\n",
"iter: 2100, loss: 3.777446182023912\n",
"iter: 2150, loss: 3.6546845786609343\n",
"iter: 2200, loss: 3.5066249518167414\n",
"iter: 2250, loss: 3.4109464065082484\n",
"iter: 2300, loss: 3.3479860273467175\n",
"iter: 2350, loss: 3.4630342653289676\n",
"iter: 2400, loss: 3.565554211571103\n",
"iter: 2450, loss: 3.575563238794841\n",
"iter: 2500, loss: 4.010586709249588\n",
"iter: 2550, loss: 3.7003344478380105\n",
"iter: 2600, loss: 3.8373477258985003\n",
"iter: 2650, loss: 3.622782635825022\n",
"iter: 2700, loss: 3.539290331628588\n",
"iter: 2750, loss: 3.6115450941721594\n",
"iter: 2800, loss: 3.5639050323849633\n",
"iter: 2850, loss: 3.7455426768651083\n",
"iter: 2900, loss: 3.590438606050279\n",
"iter: 2950, loss: 3.6904614701346747\n",
"iter: 3000, loss: 3.8645651381356383\n",
"iter: 3050, loss: 3.704173568400126\n",
"iter: 3100, loss: 3.6185882792548534\n",
"iter: 3150, loss: 3.4422046549131\n",
"iter: 3200, loss: 3.4467696623802184\n",
"iter: 3250, loss: 3.5570392836237725\n",
"iter: 3300, loss: 3.4138823104585923\n",
"iter: 3350, loss: 3.1434556758668686\n",
"iter: 3400, loss: 3.683112149511065\n",
"iter: 3450, loss: 3.548991588297344\n",
"iter: 3500, loss: 3.8323247443380795\n",
"iter: 3550, loss: 3.51619815774191\n",
"iter: 3600, loss: 3.9732376934687297\n",
"iter: 3650, loss: 3.4628570978679356\n",
"iter: 3700, loss: 3.819421215375265\n",
"iter: 3750, loss: 3.7018858858819983\n",
"iter: 3800, loss: 3.332828684034802\n",
"iter: 3850, loss: 3.39565832292466\n",
"iter: 3900, loss: 3.6046563055099\n",
"iter: 3950, loss: 3.4032811139727404\n",
"iter: 4000, loss: 3.188702507541294\n",
"iter: 4050, loss: 3.246296736966995\n",
"iter: 4100, loss: 3.3872493017287484\n",
"iter: 4150, loss: 3.2912982750620166\n",
"iter: 4200, loss: 3.439030005250657\n",
"iter: 4250, loss: 3.6874865720536967\n",
"iter: 4300, loss: 3.2006266547081967\n",
"iter: 4350, loss: 3.3141552084968198\n",
"iter: 4400, loss: 3.1777613387107846\n",
"iter: 4450, loss: 3.306143865358262\n",
"iter: 4500, loss: 3.3490057452519726\n",
"iter: 4550, loss: 3.54855015988577\n",
"iter: 4600, loss: 3.1190093379020696\n",
"iter: 4650, loss: 3.1318349221547455\n",
"iter: 4700, loss: 3.3145397909482317\n",
"iter: 4750, loss: 3.6301960383823935\n",
"iter: 4800, loss: 3.497950396598331\n",
"iter: 4850, loss: 3.433724424384889\n",
"iter: 4900, loss: 3.099926131324163\n",
"iter: 4950, loss: 3.153078259695144\n",
"iter: 5000, loss: 3.4299044117473425\n",
"iter: 5050, loss: 3.2485543521245326\n",
"iter: 5100, loss: 3.288219501253158\n",
"iter: 5150, loss: 3.0275319642793566\n",
"iter: 5200, loss: 3.2333122690518694\n",
"iter: 5250, loss: 3.1438695950281055\n",
"iter: 5300, loss: 3.289688352705941\n",
"iter: 5350, loss: 3.4346048777671085\n",
"iter: 5400, loss: 3.3960607704435075\n",
"iter: 5450, loss: 3.134131056607716\n",
"iter: 5500, loss: 2.88015941856021\n",
"iter: 5550, loss: 3.223093853874812\n",
"iter: 5600, loss: 3.523275021235148\n",
"iter: 5650, loss: 3.2974130417430207\n",
"iter: 5700, loss: 3.291351405779521\n",
"iter: 5750, loss: 3.0594470017069857\n",
"iter: 5800, loss: 3.0294449334144593\n",
"iter: 5850, loss: 3.2555880333885314\n",
"iter: 5900, loss: 2.919657180456888\n",
"iter: 5950, loss: 3.0907614767362195\n",
"iter: 6000, loss: 2.961127914254628\n",
"iter: 6050, loss: 3.4255604133378896\n",
"iter: 6100, loss: 3.113428830744728\n",
"iter: 6150, loss: 3.2713393408457434\n",
"iter: 6200, loss: 2.808141718750909\n",
"iter: 6250, loss: 3.206718180179596\n",
"iter: 6300, loss: 2.961204339458829\n",
"iter: 6350, loss: 3.3583041914379788\n",
"iter: 6400, loss: 2.8745996781455148\n",
"iter: 6450, loss: 3.044813909867453\n",
"iter: 6500, loss: 3.0786628415698103\n",
"iter: 6550, loss: 3.1983558077206693\n",
"iter: 6600, loss: 3.2838380699536156\n",
"iter: 6650, loss: 3.299677872680482\n",
"iter: 6700, loss: 3.0458072693007336\n",
"iter: 6750, loss: 2.8759968113482937\n",
"iter: 6800, loss: 2.611457399186634\n",
"iter: 6850, loss: 3.1191990507443745\n",
"iter: 6900, loss: 2.8746687850649404\n",
"iter: 6950, loss: 3.266799050270565\n",
"iter: 7000, loss: 2.9557879123422834\n",
"iter: 7050, loss: 3.3536233253327623\n",
"iter: 7100, loss: 2.866518679376633\n",
"iter: 7150, loss: 3.0647721849698866\n",
"iter: 7200, loss: 3.0131801981396147\n",
"iter: 7250, loss: 3.3611434789687866\n",
"iter: 7300, loss: 2.896131462626987\n",
"iter: 7350, loss: 3.0051966722579224\n",
"iter: 7400, loss: 2.6453575278766577\n",
"iter: 7450, loss: 3.0411309962424027\n",
"iter: 7500, loss: 3.0933231606710523\n",
"iter: 7550, loss: 3.0312348983022908\n",
"iter: 7600, loss: 3.1757038073766797\n",
"iter: 7650, loss: 3.190331464472272\n",
"iter: 7700, loss: 2.518719242436545\n",
"iter: 7750, loss: 2.9345069105965758\n",
"iter: 7800, loss: 2.8456812357221337\n",
"iter: 7850, loss: 2.9130297107620837\n",
"iter: 7900, loss: 2.979178165594737\n",
"iter: 7950, loss: 2.901021231393965\n",
"iter: 8000, loss: 2.595174813210018\n",
"iter: 8050, loss: 2.7613930717271473\n",
"iter: 8100, loss: 2.746399310149844\n",
"iter: 8150, loss: 2.8843572297663913\n",
"iter: 8200, loss: 2.7994356728735426\n",
"iter: 8250, loss: 2.6970716561135784\n",
"iter: 8300, loss: 2.883459539050147\n",
"iter: 8350, loss: 2.7503165247099735\n",
"iter: 8400, loss: 2.9744199762647114\n",
"iter: 8450, loss: 3.0706924525518273\n",
"iter: 8500, loss: 2.888958851995922\n",
"iter: 8550, loss: 2.719320885154936\n",
"iter: 8600, loss: 2.8181346920444854\n",
"iter: 8650, loss: 2.8235925950890493\n",
"iter: 8700, loss: 3.051045098115527\n",
"iter: 8750, loss: 2.5698431457110815\n",
"iter: 8800, loss: 2.7776481211828807\n",
"iter: 8850, loss: 2.4384212581695075\n",
"iter: 8900, loss: 2.6480511212954445\n",
"iter: 8950, loss: 2.5756836236620706\n",
"iter: 9000, loss: 2.8125146527971534\n",
"iter: 9050, loss: 2.8097832722512504\n",
"iter: 9100, loss: 2.8278016069389533\n",
"iter: 9150, loss: 2.444784381949712\n",
"iter: 9200, loss: 2.8099934362154158\n",
"iter: 9250, loss: 2.984244331113876\n",
"iter: 9300, loss: 2.9806695501161005\n",
"iter: 9350, loss: 2.8827475949923187\n",
"iter: 9400, loss: 3.0439721408420137\n",
"iter: 9450, loss: 2.6807251452415706\n",
"iter: 9500, loss: 2.5094920273621875\n",
"iter: 9550, loss: 2.635116928410909\n",
"iter: 9600, loss: 2.587259496537466\n",
"iter: 9650, loss: 2.6364437070649767\n",
"iter: 9700, loss: 2.6659068493899842\n",
"iter: 9750, loss: 2.3925973146056365\n",
"iter: 9800, loss: 2.8345537455271157\n",
"iter: 9850, loss: 2.3069138811202277\n",
"iter: 9900, loss: 2.319064138798487\n",
"iter: 9950, loss: 2.4867696173228913\n",
"iter: 10000, loss: 2.614620875483468\n",
"iter: 10050, loss: 2.422453577261123\n",
"iter: 10100, loss: 2.643933411677678\n",
"iter: 10150, loss: 2.5282146744349645\n",
"iter: 10200, loss: 2.5393255345310486\n",
"iter: 10250, loss: 2.9825220655032565\n",
"iter: 10300, loss: 2.2635890780091277\n",
"iter: 10350, loss: 2.7769809711168683\n",
"iter: 10400, loss: 2.445600000275506\n",
"iter: 10450, loss: 2.453449849030328\n",
"iter: 10500, loss: 2.5520464683108854\n",
"iter: 10550, loss: 2.577900281663925\n",
"iter: 10600, loss: 2.4218848383086065\n",
"iter: 10650, loss: 2.5381085565317245\n",
"iter: 10700, loss: 2.196764139754431\n",
"iter: 10750, loss: 2.456448502972012\n",
"iter: 10800, loss: 2.560683703441468\n",
"iter: 10850, loss: 2.53125628255284\n",
"iter: 10900, loss: 2.7491349925086603\n",
"iter: 10950, loss: 2.6151628021588404\n",
"iter: 11000, loss: 2.507106682993117\n",
"iter: 11050, loss: 2.369231795661033\n",
"iter: 11100, loss: 2.5730169670676433\n",
"iter: 11150, loss: 2.3010029462254233\n",
"iter: 11200, loss: 2.633150562687526\n",
"iter: 11250, loss: 2.5919544999429163\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"iter: 11300, loss: 2.534775358722323\n",
"iter: 11350, loss: 2.3763021782466343\n",
"iter: 11400, loss: 2.477060817866099\n",
"iter: 11450, loss: 2.240788399741763\n",
"iter: 11500, loss: 2.4560615454931103\n",
"iter: 11550, loss: 2.4167706055300577\n",
"iter: 11600, loss: 2.5485691031482482\n",
"iter: 11650, loss: 2.4385872491881964\n",
"iter: 11700, loss: 2.262665515203325\n",
"iter: 11750, loss: 2.3140043601195015\n",
"iter: 11800, loss: 2.3840308377969834\n",
"iter: 11850, loss: 2.3109344417519044\n",
"iter: 11900, loss: 2.3575586484140825\n",
"iter: 11950, loss: 2.2054063754535855\n",
"iter: 12000, loss: 2.256502300773348\n",
"iter: 12050, loss: 2.4794748330608245\n",
"iter: 12100, loss: 2.337028050218309\n",
"iter: 12150, loss: 2.0973778800964356\n",
"iter: 12200, loss: 2.159631293109485\n",
"iter: 12250, loss: 2.3099975161770034\n",
"iter: 12300, loss: 2.421918697101729\n",
"iter: 12350, loss: 2.531752646151043\n",
"iter: 12400, loss: 2.320960243735995\n",
"iter: 12450, loss: 2.2293582723708387\n",
"iter: 12500, loss: 2.2750969548414623\n",
"iter: 12550, loss: 1.7618893385950534\n",
"iter: 12600, loss: 2.340753597418468\n",
"iter: 12650, loss: 2.297142340731999\n",
"iter: 12700, loss: 2.3628056962887443\n",
"iter: 12750, loss: 2.6683244729306956\n",
"iter: 12800, loss: 2.129389260549394\n",
"iter: 12850, loss: 2.200342266559601\n",
"iter: 12900, loss: 2.2035769688401903\n",
"iter: 12950, loss: 2.2374771443643264\n",
"iter: 13000, loss: 2.174828187268878\n",
"iter: 13050, loss: 2.4299814297888016\n",
"iter: 13100, loss: 2.2743770096472335\n",
"iter: 13150, loss: 2.2950440486839843\n",
"iter: 13200, loss: 2.4582167831375483\n",
"iter: 13250, loss: 2.540857286835474\n",
"iter: 13300, loss: 2.322540373416174\n",
"iter: 13350, loss: 2.319241341783887\n",
"iter: 13400, loss: 2.3551435836969854\n",
"iter: 13450, loss: 2.21491847473856\n",
"iter: 13500, loss: 2.0196543374175118\n",
"iter: 13550, loss: 2.294338379172105\n",
"iter: 13600, loss: 1.8327846462726596\n",
"iter: 13650, loss: 2.035525601328366\n",
"iter: 13700, loss: 2.2429525320794843\n",
"iter: 13750, loss: 2.034286926406243\n",
"iter: 13800, loss: 2.100517734408379\n",
"iter: 13850, loss: 2.0885622107670425\n",
"iter: 13900, loss: 1.90785291909793\n",
"iter: 13950, loss: 2.273535749908478\n",
"iter: 14000, loss: 1.838191468528339\n",
"iter: 14050, loss: 2.3195868289754507\n",
"iter: 14100, loss: 1.8965250493060974\n",
"iter: 14150, loss: 2.10934934569919\n",
"iter: 14200, loss: 2.151934117366397\n",
"iter: 14250, loss: 2.0066685717522152\n",
"iter: 14300, loss: 2.3296401413433134\n",
"iter: 14350, loss: 1.9384719442223746\n",
"iter: 14400, loss: 2.2025605153564425\n",
"iter: 14450, loss: 2.263807416308494\n",
"iter: 14500, loss: 1.9864815442051205\n",
"iter: 14550, loss: 1.7038374454577763\n",
"iter: 14600, loss: 2.274628053700167\n",
"iter: 14650, loss: 2.1628303778625675\n",
"iter: 14700, loss: 1.9897215003796989\n",
"iter: 14750, loss: 1.860605546917234\n",
"iter: 14800, loss: 1.9588362134335529\n",
"iter: 14850, loss: 1.8767746505396707\n",
"iter: 14900, loss: 1.834631380274182\n",
"iter: 14950, loss: 1.9499947649410792\n",
"iter: 15000, loss: 2.0015979091269624\n",
"iter: 15050, loss: 2.0649836547412574\n",
"iter: 15100, loss: 2.249369715940384\n",
"iter: 15150, loss: 1.5817453392441307\n",
"iter: 15200, loss: 2.1706447578157695\n",
"iter: 15250, loss: 1.9688029914564558\n",
"iter: 15300, loss: 2.046964565526871\n",
"iter: 15350, loss: 1.9338763165667892\n",
"iter: 15400, loss: 1.9137448829904438\n",
"iter: 15450, loss: 1.7699638532740727\n",
"iter: 15500, loss: 2.2515631875159245\n",
"iter: 15550, loss: 1.7620117027797395\n",
"iter: 15600, loss: 1.9152411586524003\n",
"iter: 15650, loss: 2.0947861353386017\n",
"iter: 15700, loss: 1.9149094790844687\n",
"iter: 15750, loss: 1.7210240173566909\n",
"iter: 15800, loss: 2.014472983038614\n",
"iter: 15850, loss: 2.1098430752697444\n",
"iter: 15900, loss: 2.023270213549099\n",
"iter: 15950, loss: 1.9570550824488917\n",
"iter: 16000, loss: 1.895675997123832\n",
"iter: 16050, loss: 1.837380247549642\n",
"iter: 16100, loss: 1.894489290089834\n",
"iter: 16150, loss: 2.075172846224573\n",
"iter: 16200, loss: 1.8212170035555248\n",
"iter: 16250, loss: 1.8570367700694095\n",
"iter: 16300, loss: 1.6184977439187818\n",
"iter: 16350, loss: 1.7351362812415\n",
"iter: 16400, loss: 1.872060403579758\n",
"iter: 16450, loss: 1.6218276036712858\n",
"iter: 16500, loss: 1.9870286158758497\n",
"iter: 16550, loss: 1.9007116212835387\n",
"iter: 16600, loss: 1.8743730505156142\n",
"iter: 16650, loss: 1.5293502329928537\n",
"iter: 16700, loss: 1.811881399162232\n",
"iter: 16750, loss: 1.5156562756375658\n",
"iter: 16800, loss: 1.6397469798794813\n",
"iter: 16850, loss: 2.2027597563172145\n",
"iter: 16900, loss: 1.8139538214131006\n",
"iter: 16950, loss: 2.1659815680677927\n",
"iter: 17000, loss: 1.947558910210927\n",
"iter: 17050, loss: 2.0774720856149993\n",
"iter: 17100, loss: 1.7940182881762112\n",
"iter: 17150, loss: 2.1425245618441746\n",
"iter: 17200, loss: 1.6630687274876097\n",
"iter: 17250, loss: 1.7448162170535044\n",
"iter: 17300, loss: 1.8790338722637718\n",
"iter: 17350, loss: 1.96936958753495\n",
"iter: 17400, loss: 1.8035021762762753\n",
"iter: 17450, loss: 1.784786748029883\n",
"iter: 17500, loss: 1.8431302896037933\n",
"iter: 17550, loss: 1.9356805955001288\n",
"iter: 17600, loss: 1.571500998784625\n",
"iter: 17650, loss: 1.849001149414551\n",
"iter: 17700, loss: 1.5969795638758038\n",
"iter: 17750, loss: 1.6012443284591038\n",
"iter: 17800, loss: 1.5525058465600012\n",
"iter: 17850, loss: 1.450256337930286\n",
"iter: 17900, loss: 1.7983906224483532\n",
"iter: 17950, loss: 1.7381368355050921\n",
"iter: 18000, loss: 1.6177345383224033\n",
"iter: 18050, loss: 1.835479336150582\n",
"iter: 18100, loss: 1.5402896869333964\n",
"iter: 18150, loss: 1.5447071926097078\n",
"iter: 18200, loss: 1.6833134629707485\n",
"iter: 18250, loss: 1.8886855756252532\n",
"iter: 18300, loss: 1.6310479882558186\n",
"iter: 18350, loss: 1.6417460731078708\n",
"iter: 18400, loss: 1.7383878009962657\n",
"iter: 18450, loss: 1.6342206524724057\n",
"iter: 18500, loss: 1.5872581603981197\n",
"iter: 18550, loss: 1.287150528927171\n",
"iter: 18600, loss: 1.6059650084300645\n",
"iter: 18650, loss: 1.28275570456045\n",
"iter: 18700, loss: 1.439326602407864\n",
"iter: 18750, loss: 1.7180297046511894\n",
"iter: 18800, loss: 1.6227167361766575\n",
"iter: 18850, loss: 1.437303775454324\n",
"iter: 18900, loss: 1.6929941054639364\n",
"iter: 18950, loss: 1.6776369486933662\n",
"iter: 19000, loss: 1.69069007818661\n",
"iter: 19050, loss: 1.8343193885277191\n",
"iter: 19100, loss: 1.3482130224931808\n",
"iter: 19150, loss: 1.4392069308530717\n",
"iter: 19200, loss: 1.4435342772607769\n",
"iter: 19250, loss: 1.4412190558891447\n",
"iter: 19300, loss: 1.7313999670062743\n",
"iter: 19350, loss: 1.6303069564179768\n",
"iter: 19400, loss: 1.8313010199240274\n",
"iter: 19450, loss: 1.476125830580318\n",
"iter: 19500, loss: 1.784752085032917\n",
"iter: 19550, loss: 1.900799496985617\n",
"iter: 19600, loss: 1.6683086817453778\n",
"iter: 19650, loss: 1.6018399291965693\n",
"iter: 19700, loss: 1.5080324055013201\n",
"iter: 19750, loss: 1.7074753486149838\n",
"iter: 19800, loss: 1.5588394280918059\n",
"iter: 19850, loss: 1.4063752451401856\n",
"iter: 19900, loss: 1.6571519161235722\n",
"iter: 19950, loss: 1.4880279254605846\n",
"iter: 20000, loss: 1.4425315815721234\n",
"iter: 20050, loss: 1.4204049231041045\n",
"iter: 20100, loss: 1.5411449456631194\n",
"iter: 20150, loss: 1.4098666115223417\n",
"iter: 20200, loss: 1.4514436369504011\n",
"iter: 20250, loss: 1.678218051835658\n",
"iter: 20300, loss: 1.3683213942356056\n",
"iter: 20350, loss: 1.4311776501555296\n",
"iter: 20400, loss: 1.44434953142537\n",
"iter: 20450, loss: 1.4809531437674215\n",
"iter: 20500, loss: 1.498182836138067\n",
"iter: 20550, loss: 1.6891843606990486\n",
"iter: 20600, loss: 1.307836448823176\n",
"iter: 20650, loss: 1.3191714194629873\n",
"iter: 20700, loss: 1.435782224451266\n",
"iter: 20750, loss: 1.4501241854064992\n",
"iter: 20800, loss: 1.570673788651587\n",
"iter: 20850, loss: 1.6726866277487031\n",
"iter: 20900, loss: 1.490093153404811\n",
"iter: 20950, loss: 1.3381259351434216\n",
"iter: 21000, loss: 1.4293887265811833\n",
"iter: 21050, loss: 1.5261030488553495\n",
"iter: 21100, loss: 1.4049972703861338\n",
"iter: 21150, loss: 1.3666674501952667\n",
"iter: 21200, loss: 1.544151809760502\n",
"iter: 21250, loss: 1.4767180123480546\n",
"iter: 21300, loss: 1.3458678885580055\n",
"iter: 21350, loss: 1.3158163404729633\n",
"iter: 21400, loss: 1.2743006317423922\n",
"iter: 21450, loss: 1.4159044456604926\n",
"iter: 21500, loss: 1.7186118897502385\n",
"iter: 21550, loss: 1.4735830772358276\n",
"iter: 21600, loss: 1.2575308752939818\n",
"iter: 21650, loss: 1.2709813627033006\n",
"iter: 21700, loss: 1.2383236987832047\n",
"iter: 21750, loss: 1.2756263920948618\n",
"iter: 21800, loss: 1.1783258064417612\n",
"iter: 21850, loss: 1.290928970362459\n",
"iter: 21900, loss: 1.2292051843586895\n",
"iter: 21950, loss: 1.4506985603798\n",
"iter: 22000, loss: 1.2761652381798578\n",
"iter: 22050, loss: 1.258709805733628\n",
"iter: 22100, loss: 1.5169502600658507\n",
"iter: 22150, loss: 1.384094950204804\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"iter: 22200, loss: 1.6116164426425141\n",
"iter: 22250, loss: 1.2793350757350999\n",
"iter: 22300, loss: 1.3017940769725374\n",
"iter: 22350, loss: 1.3736145087998537\n",
"iter: 22400, loss: 1.138627294466609\n",
"iter: 22450, loss: 1.310434480055457\n",
"iter: 22500, loss: 1.3961503054835491\n",
"iter: 22550, loss: 1.3724343404996964\n",
"iter: 22600, loss: 1.3563060793403596\n",
"iter: 22650, loss: 1.5050592094120523\n",
"iter: 22700, loss: 1.3752844895202485\n",
"iter: 22750, loss: 1.1653902740024384\n",
"iter: 22800, loss: 1.586769177135967\n",
"iter: 22850, loss: 1.2032956434232844\n",
"iter: 22900, loss: 1.3740250343538463\n",
"iter: 22950, loss: 1.1050300681430192\n",
"iter: 23000, loss: 1.2519222570203599\n",
"iter: 23050, loss: 1.42347088217688\n",
"iter: 23100, loss: 1.4116886373684991\n",
"iter: 23150, loss: 1.1194891034222785\n",
"iter: 23200, loss: 1.2006812089085581\n",
"iter: 23250, loss: 1.4476829275015806\n",
"iter: 23300, loss: 1.3017136716679447\n",
"iter: 23350, loss: 1.3050824448059475\n",
"iter: 23400, loss: 1.0946670590383667\n",
"iter: 23450, loss: 1.0267400648877734\n",
"iter: 23500, loss: 1.3339079014630544\n",
"iter: 23550, loss: 1.192698698204543\n",
"iter: 23600, loss: 1.2177529160323597\n",
"iter: 23650, loss: 1.0835593536988137\n",
"iter: 23700, loss: 1.0673790020138498\n",
"iter: 23750, loss: 1.3244361790158448\n",
"iter: 23800, loss: 1.1204376552663156\n",
"iter: 23850, loss: 1.2015556238284189\n",
"iter: 23900, loss: 1.4070206288629112\n",
"iter: 23950, loss: 1.3217124197293841\n",
"iter: 24000, loss: 1.0488405134649506\n",
"iter: 24050, loss: 1.1713006507924626\n",
"iter: 24100, loss: 1.2951658061903621\n",
"iter: 24150, loss: 1.2749495941726934\n",
"iter: 24200, loss: 1.2141112795103162\n",
"iter: 24250, loss: 1.3290269901402414\n",
"iter: 24300, loss: 1.094365536570549\n",
"iter: 24350, loss: 1.133138121331495\n",
"iter: 24400, loss: 1.3418169331314074\n",
"iter: 24450, loss: 0.9847527233404773\n",
"iter: 24500, loss: 1.1087985199188426\n",
"iter: 24550, loss: 1.4006639858985706\n",
"iter: 24600, loss: 1.1466205491246213\n",
"iter: 24650, loss: 1.1214664732799642\n",
"iter: 24700, loss: 1.1177163749280432\n",
"iter: 24750, loss: 1.05219458625714\n",
"iter: 24800, loss: 1.1949661584328566\n",
"iter: 24850, loss: 0.9802025896845353\n",
"iter: 24900, loss: 1.1272975780748655\n",
"iter: 24950, loss: 1.0976827581269404\n",
"iter: 25000, loss: 0.9013028181819688\n",
"iter: 25050, loss: 1.3180778384589484\n",
"iter: 25100, loss: 1.0977117643299557\n",
"iter: 25150, loss: 0.9444285869991021\n",
"iter: 25200, loss: 1.336973425586545\n",
"iter: 25250, loss: 1.2987125150627556\n",
"iter: 25300, loss: 1.0681130346740995\n",
"iter: 25350, loss: 0.9836248108498631\n",
"iter: 25400, loss: 1.1549646752675378\n",
"iter: 25450, loss: 1.0397938099617048\n",
"iter: 25500, loss: 1.4253321852816476\n",
"iter: 25550, loss: 1.397831358559548\n",
"iter: 25600, loss: 0.8681884022117372\n",
"iter: 25650, loss: 0.949661937920347\n",
"iter: 25700, loss: 1.018096680844114\n",
"iter: 25750, loss: 1.0033835446210135\n",
"iter: 25800, loss: 0.9399867170606815\n",
"iter: 25850, loss: 0.9365767020531115\n",
"iter: 25900, loss: 1.2080267537056453\n",
"iter: 25950, loss: 1.0215099297222634\n",
"iter: 26000, loss: 0.9733565044677448\n",
"iter: 26050, loss: 1.0712914910318834\n",
"iter: 26100, loss: 0.8407332850779805\n",
"iter: 26150, loss: 0.9271211279460363\n",
"iter: 26200, loss: 0.9953902960416108\n",
"iter: 26250, loss: 1.0131704654341178\n",
"iter: 26300, loss: 1.0885028305432156\n",
"iter: 26350, loss: 1.0190075791875521\n",
"iter: 26400, loss: 1.009052420553707\n",
"iter: 26450, loss: 1.0815212898623379\n",
"iter: 26500, loss: 0.9892340009240876\n",
"iter: 26550, loss: 1.0516380755560737\n",
"iter: 26600, loss: 0.9344196589528803\n",
"iter: 26650, loss: 0.8953249894132216\n",
"iter: 26700, loss: 0.9229552195980435\n",
"iter: 26750, loss: 0.7424087155598496\n",
"iter: 26800, loss: 0.911013327536129\n",
"iter: 26850, loss: 1.1781759474883002\n",
"iter: 26900, loss: 1.196274289493523\n",
"iter: 26950, loss: 1.0227981455389943\n",
"iter: 27000, loss: 0.9916679235928586\n",
"iter: 27050, loss: 0.9636169400480058\n",
"iter: 27100, loss: 0.8002338881918359\n",
"iter: 27150, loss: 0.800919440870247\n",
"iter: 27200, loss: 0.8211033871329966\n",
"iter: 27250, loss: 0.8155000005123162\n",
"iter: 27300, loss: 0.876837944473539\n",
"iter: 27350, loss: 1.1260614515467298\n",
"iter: 27400, loss: 1.058864346462583\n",
"iter: 27450, loss: 1.1114834662898192\n",
"iter: 27500, loss: 0.9796440882084387\n",
"iter: 27550, loss: 1.0277935135303036\n",
"iter: 27600, loss: 0.6979781560635284\n",
"iter: 27650, loss: 0.770827453770808\n",
"iter: 27700, loss: 1.1471699211550135\n",
"iter: 27750, loss: 0.8712478535033409\n",
"iter: 27800, loss: 0.7957819575319688\n",
"iter: 27850, loss: 1.0939111870155924\n",
"iter: 27900, loss: 0.9194521397224494\n",
"iter: 27950, loss: 0.8920607558945345\n",
"iter: 28000, loss: 0.8829188095186908\n",
"iter: 28050, loss: 0.9212011002366033\n",
"iter: 28100, loss: 0.7731392620366715\n",
"iter: 28150, loss: 1.056102939241699\n",
"iter: 28200, loss: 0.9831677025327132\n",
"iter: 28250, loss: 1.071929881365999\n",
"iter: 28300, loss: 0.9135961269267967\n",
"iter: 28350, loss: 0.8095226630355632\n",
"iter: 28400, loss: 0.9595384959959911\n",
"iter: 28450, loss: 0.7839641324215465\n",
"iter: 28500, loss: 0.9889460563829968\n",
"iter: 28550, loss: 1.0575634596305232\n",
"iter: 28600, loss: 1.05014324463218\n",
"iter: 28650, loss: 0.9521020337228501\n",
"iter: 28700, loss: 0.8122104218034515\n",
"iter: 28750, loss: 0.9600319336676408\n",
"iter: 28800, loss: 0.7290925218548092\n",
"iter: 28850, loss: 0.8589948218661168\n",
"iter: 28900, loss: 0.8876770969496832\n",
"iter: 28950, loss: 0.7668700665647076\n",
"iter: 29000, loss: 0.8810090623952094\n",
"iter: 29050, loss: 0.9807037507650397\n",
"iter: 29100, loss: 0.6704667443845952\n",
"iter: 29150, loss: 0.6698679181308975\n",
"iter: 29200, loss: 0.8776328837161972\n",
"iter: 29250, loss: 0.8806386950718503\n",
"iter: 29300, loss: 0.6410340730618862\n",
"iter: 29350, loss: 0.8755547849377472\n",
"iter: 29400, loss: 0.8818342795334163\n",
"iter: 29450, loss: 0.7442211986517623\n",
"iter: 29500, loss: 0.8927219600469348\n",
"iter: 29550, loss: 1.019919359203842\n",
"iter: 29600, loss: 0.8808109327583087\n",
"iter: 29650, loss: 0.8205070998280766\n",
"iter: 29700, loss: 1.019214930534363\n",
"iter: 29750, loss: 0.8730531409016206\n",
"iter: 29800, loss: 0.7633821407521053\n",
"iter: 29850, loss: 0.796077705860138\n",
"iter: 29900, loss: 0.7018148700419874\n",
"iter: 29950, loss: 1.1195493836871218\n",
"iter: 30000, loss: 0.8907366043790467\n",
"iter: 30050, loss: 0.9264667236958704\n",
"iter: 30100, loss: 1.0352731366356211\n",
"iter: 30150, loss: 0.7005343800724028\n",
"iter: 30200, loss: 0.9168639244217249\n",
"iter: 30250, loss: 0.8539114402177789\n"
]
},
{
"ename": "KeyboardInterrupt",
"evalue": "",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mKeyboardInterrupt\u001b[0m Traceback (most recent call last)",
"Input \u001b[0;32mIn [19]\u001b[0m, in \u001b[0;36m\u001b[0;34m()\u001b[0m\n\u001b[1;32m 2\u001b[0m encoder1 \u001b[38;5;241m=\u001b[39m EncoderRNN(eng_lang\u001b[38;5;241m.\u001b[39mn_words, hidden_size)\u001b[38;5;241m.\u001b[39mto(device)\n\u001b[1;32m 3\u001b[0m attn_decoder1 \u001b[38;5;241m=\u001b[39m AttnDecoderRNN(hidden_size, pol_lang\u001b[38;5;241m.\u001b[39mn_words, dropout_p\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m0.1\u001b[39m)\u001b[38;5;241m.\u001b[39mto(device)\n\u001b[0;32m----> 5\u001b[0m \u001b[43mtrainIters\u001b[49m\u001b[43m(\u001b[49m\u001b[43mencoder1\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mattn_decoder1\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m75000\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprint_every\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m50\u001b[39;49m\u001b[43m)\u001b[49m\n",
"Input \u001b[0;32mIn [16]\u001b[0m, in \u001b[0;36mtrainIters\u001b[0;34m(encoder, decoder, n_iters, print_every, learning_rate)\u001b[0m\n\u001b[1;32m 14\u001b[0m input_tensor \u001b[38;5;241m=\u001b[39m training_pair[\u001b[38;5;241m0\u001b[39m]\n\u001b[1;32m 15\u001b[0m target_tensor \u001b[38;5;241m=\u001b[39m training_pair[\u001b[38;5;241m1\u001b[39m]\n\u001b[0;32m---> 17\u001b[0m loss \u001b[38;5;241m=\u001b[39m \u001b[43mtrain_one_batch\u001b[49m\u001b[43m(\u001b[49m\u001b[43minput_tensor\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 18\u001b[0m \u001b[43m \u001b[49m\u001b[43mtarget_tensor\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 19\u001b[0m \u001b[43m \u001b[49m\u001b[43mencoder\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 20\u001b[0m \u001b[43m \u001b[49m\u001b[43mdecoder\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 21\u001b[0m \u001b[43m \u001b[49m\u001b[43mencoder_optimizer\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 22\u001b[0m \u001b[43m \u001b[49m\u001b[43mdecoder_optimizer\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 23\u001b[0m \u001b[43m \u001b[49m\u001b[43mcriterion\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 25\u001b[0m print_loss_total \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m loss\n\u001b[1;32m 27\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m i \u001b[38;5;241m%\u001b[39m print_every \u001b[38;5;241m==\u001b[39m \u001b[38;5;241m0\u001b[39m:\n",
"Input \u001b[0;32mIn [15]\u001b[0m, in \u001b[0;36mtrain_one_batch\u001b[0;34m(input_tensor, target_tensor, encoder, decoder, encoder_optimizer, decoder_optimizer, criterion, max_length)\u001b[0m\n\u001b[1;32m 39\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m decoder_input\u001b[38;5;241m.\u001b[39mitem() \u001b[38;5;241m==\u001b[39m EOS_token:\n\u001b[1;32m 40\u001b[0m \u001b[38;5;28;01mbreak\u001b[39;00m\n\u001b[0;32m---> 42\u001b[0m \u001b[43mloss\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbackward\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 44\u001b[0m encoder_optimizer\u001b[38;5;241m.\u001b[39mstep()\n\u001b[1;32m 45\u001b[0m decoder_optimizer\u001b[38;5;241m.\u001b[39mstep()\n",
"File \u001b[0;32m~/anaconda3/envs/zajeciaei/lib/python3.10/site-packages/torch/_tensor.py:363\u001b[0m, in \u001b[0;36mTensor.backward\u001b[0;34m(self, gradient, retain_graph, create_graph, inputs)\u001b[0m\n\u001b[1;32m 354\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m has_torch_function_unary(\u001b[38;5;28mself\u001b[39m):\n\u001b[1;32m 355\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m handle_torch_function(\n\u001b[1;32m 356\u001b[0m Tensor\u001b[38;5;241m.\u001b[39mbackward,\n\u001b[1;32m 357\u001b[0m (\u001b[38;5;28mself\u001b[39m,),\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 361\u001b[0m create_graph\u001b[38;5;241m=\u001b[39mcreate_graph,\n\u001b[1;32m 362\u001b[0m inputs\u001b[38;5;241m=\u001b[39minputs)\n\u001b[0;32m--> 363\u001b[0m \u001b[43mtorch\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mautograd\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbackward\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mgradient\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mretain_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcreate_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43minputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43minputs\u001b[49m\u001b[43m)\u001b[49m\n",
"File \u001b[0;32m~/anaconda3/envs/zajeciaei/lib/python3.10/site-packages/torch/autograd/__init__.py:173\u001b[0m, in \u001b[0;36mbackward\u001b[0;34m(tensors, grad_tensors, retain_graph, create_graph, grad_variables, inputs)\u001b[0m\n\u001b[1;32m 168\u001b[0m retain_graph \u001b[38;5;241m=\u001b[39m create_graph\n\u001b[1;32m 170\u001b[0m \u001b[38;5;66;03m# The reason we repeat same the comment below is that\u001b[39;00m\n\u001b[1;32m 171\u001b[0m \u001b[38;5;66;03m# some Python versions print out the first line of a multi-line function\u001b[39;00m\n\u001b[1;32m 172\u001b[0m \u001b[38;5;66;03m# calls in the traceback and some print out the last line\u001b[39;00m\n\u001b[0;32m--> 173\u001b[0m \u001b[43mVariable\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_execution_engine\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mrun_backward\u001b[49m\u001b[43m(\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;66;43;03m# Calls into the C++ engine to run the backward pass\u001b[39;49;00m\n\u001b[1;32m 174\u001b[0m \u001b[43m \u001b[49m\u001b[43mtensors\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mgrad_tensors_\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mretain_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcreate_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43minputs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 175\u001b[0m \u001b[43m \u001b[49m\u001b[43mallow_unreachable\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43maccumulate_grad\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m)\u001b[49m\n",
"\u001b[0;31mKeyboardInterrupt\u001b[0m: "
]
}
],
"source": [
"hidden_size = 256\n",
"encoder1 = EncoderRNN(eng_lang.n_words, hidden_size).to(device)\n",
"attn_decoder1 = AttnDecoderRNN(hidden_size, pol_lang.n_words, dropout_p=0.1).to(device)\n",
"\n",
"trainIters(encoder1, attn_decoder1, 75000, print_every=50)"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"> he s a very important person .\n",
"= on jest bardzo ważnym człowiekiem .\n",
"< on jest bardzo ważnym człowiekiem . \n",
"\n",
"> i m beautiful .\n",
"= jestem piękny .\n",
"< jestem piękna . \n",
"\n",
"> we re quite certain of that .\n",
"= jesteśmy tego całkiem pewni .\n",
"< jesteśmy tego całkiem pewni . \n",
"\n",
"> we are all looking forward to seeing you .\n",
"= miło nam będzie ponownie się z panem spotkać .\n",
"< miło nam nam ponownie się z tobą . . \n",
"\n",
"> i m inside .\n",
"= jestem w środku .\n",
"< jestem w środku . \n",
"\n",
"> i m giving up smoking .\n",
"= rzucam palenie .\n",
"< rzuciłem palenie . \n",
"\n",
"> we re not arguing .\n",
"= nie kłócimy się .\n",
"< nie wychodzimy . \n",
"\n",
"> i m not prepared to do that yet .\n",
"= nie jestem jeszcze przygotowany żeby to zrobić .\n",
"< nie jestem jeszcze przygotowany żeby to zrobić . \n",
"\n",
"> i m a free man .\n",
"= jestem wolnym człowiekiem .\n",
"< jestem wolnym człowiekiem . \n",
"\n",
"> i m still on the clock .\n",
"= jeszcze jestem w pracy .\n",
"< wciąż jestem w domu . \n",
"\n"
]
}
],
"source": [
"evaluateRandomly(encoder1, attn_decoder1)"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['i m ok .', 'ze mną wszystko w porządku .']"
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"pairs[0]"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"author": "Jakub Pokrywka",
"email": "kubapok@wmi.amu.edu.pl",
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"lang": "pl",
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.10.4"
},
"subtitle": "0.Informacje na temat przedmiotu[ćwiczenia]",
"title": "Ekstrakcja informacji",
"year": "2021"
},
"nbformat": 4,
"nbformat_minor": 4
}
|