From b8b339abc832b33c60cfb025981a44075ee3767e Mon Sep 17 00:00:00 2001
From: Ludovic Moncla <moncla.ludovic@gmail.com>
Date: Mon, 28 Nov 2022 13:30:00 +0100
Subject: [PATCH] Update Predict_XAI.ipynb
---
notebooks/Predict_XAI.ipynb | 41 +++++++++++++++----------------------
1 file changed, 16 insertions(+), 25 deletions(-)
diff --git a/notebooks/Predict_XAI.ipynb b/notebooks/Predict_XAI.ipynb
index a75c4cc..1fdf013 100644
--- a/notebooks/Predict_XAI.ipynb
+++ b/notebooks/Predict_XAI.ipynb
@@ -26,8 +26,8 @@
"outputs": [],
"source": [
"!pip install transformers==4.10.3\n",
- "!pip install transformers_interpret\n",
- "!pip install sentencepiece"
+ "!pip install sentencepiece\n",
+ "!pip install transformers_interpret"
]
},
{
@@ -93,22 +93,24 @@
},
{
"cell_type": "code",
- "execution_count": 7,
+ "execution_count": 1,
"metadata": {
"id": "SkErnwgMMbRj"
},
"outputs": [],
"source": [
- "import os\n",
- "import pandas as pd \n",
- "import numpy as np\n",
"import pickle \n",
"import torch\n",
- "from tqdm import tqdm\n",
"\n",
- "from transformers import BertTokenizer, BertForSequenceClassification, CamembertTokenizer, CamembertForSequenceClassification\n",
+ "from transformers import BertTokenizer, BertForSequenceClassification\n",
"from transformers_interpret import SequenceClassificationExplainer\n",
- "from torch.utils.data import TensorDataset, DataLoader, SequentialSampler"
+ "\n",
+ "import numpy as np\n",
+ "import torch\n",
+ "from torch.utils.data import TensorDataset, DataLoader, SequentialSampler\n",
+ "from tqdm import tqdm\n",
+ "import os\n",
+ "import pandas as pd \n"
]
},
{
@@ -160,7 +162,7 @@
},
{
"cell_type": "code",
- "execution_count": 12,
+ "execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
@@ -177,16 +179,10 @@
},
{
"cell_type": "code",
- "execution_count": 3,
+ "execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
- "def padding(content, max_len):\n",
- " if len(content) > max_len:\n",
- " content[:max_len]\n",
- " else:\n",
- " content + [0] * (max_len - len(content))\n",
- " return \n",
"\n",
"def generate_dataloader(tokenizer, sentences, batch_size = 8, max_len = 512):\n",
"\n",
@@ -311,7 +307,7 @@
"id": "c5QKcXulhNJ-"
},
"source": [
- "## 3. Load Data\n",
+ "## 2. Load Data\n",
"\n",
"\n",
"!! A modifier: charger le corpus parallele : EDdA et LGE"
@@ -541,7 +537,6 @@
"#path = \"drive/MyDrive/Classification-EDdA/\"\n",
"path = \"../\"\n",
"model_name = \"bert-base-multilingual-cased\"\n",
- "#model_name = \"camembert-base\"\n",
"model_path = path + \"models/model_\" + model_name + \"_s10000.pt\""
]
},
@@ -559,12 +554,8 @@
}
],
"source": [
- "if model_name == 'bert-base-multilingual-cased' :\n",
- " print('Loading Bert Tokenizer...')\n",
- " tokenizer = BertTokenizer.from_pretrained(model_name)\n",
- "elif model_name == 'camembert-base':\n",
- " print('Loading Camembert Tokenizer...')\n",
- " tokenizer = CamembertTokenizer.from_pretrained(model_name)"
+ "print('Loading Bert Tokenizer...')\n",
+ "tokenizer = BertTokenizer.from_pretrained(model_name)"
]
},
{
--
GitLab