Skip to content

Dev #13

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
joaomonteirof merged 2 commits into from
Jun 9, 2023
Merged

Dev #13

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
92c07b5
Added notebook with code search eval.
joaomonteirof Jun 9, 2023
11720f3
Added notebook code_bert_score eval.
joaomonteirof Jun 9, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
273 changes: 273 additions & 0 deletions c2c_search_eval.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,273 @@
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"from typing import Dict\n",
"import torch\n",
"from src import datasets_loader\n",
"from src.utils import retrieval_eval, pool_and_normalize\n",
"from src.constants import GFG_DATA_PATH\n",
"from transformers import AutoModel, AutoTokenizer\n",
"from src.datasets_loader import prepare_tokenizer\n",
"from src.preprocessing_utils import truncate_sentences\n",
"from abc import ABC, abstractmethod\n"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"DEVICE = \"cuda:0\"\n",
"\n",
"EVAL_CONFIGS =[\n",
" {\"model_path\": \"starencoder\", \"maximum_raw_length\": 10000, \"maximum_input_length\": 1024, \"device\": DEVICE},\n",
" {\"model_path\": \"codebert\", \"maximum_raw_length\": 10000, \"maximum_input_length\": 512, \"device\": DEVICE}\n",
"]"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"def set_device(inputs: Dict[str, torch.Tensor], device: str) -> Dict[str, torch.Tensor]:\n",
" output_data = {}\n",
" for k, v in inputs.items():\n",
" output_data[k] = v.to(device)\n",
"\n",
" return output_data\n",
"\n",
"\n",
"def get_dataset(maximum_raw_length):\n",
" test_data = datasets_loader.get_dataset( # Geeks4Geeks data\n",
" dataset_name=\"gfg\",\n",
" path_to_cache=GFG_DATA_PATH,\n",
" split=\"test\",\n",
" maximum_raw_length=maximum_raw_length,\n",
" )\n",
"\n",
" return test_data\n",
"\n",
"\n",
"class BaseEncoder(torch.nn.Module, ABC):\n",
" def __init__(self, device, max_input_len, maximum_token_len, model_name):\n",
" super().__init__()\n",
"\n",
" self.model_name = model_name\n",
" self.tokenizer = prepare_tokenizer(model_name)\n",
" self.encoder = (\n",
" AutoModel.from_pretrained(model_name, use_auth_token=True).to(DEVICE).eval()\n",
" )\n",
" self.device = device\n",
" self.max_input_len = max_input_len\n",
" self.maximum_token_len = maximum_token_len\n",
"\n",
" @abstractmethod\n",
" def forward(\n",
" self,\n",
" ):\n",
" pass\n",
"\n",
" def encode(self, input_sentences, batch_size=32, **kwargs):\n",
" truncated_input_sentences = truncate_sentences(\n",
" input_sentences, self.max_input_len\n",
" )\n",
"\n",
" n_batches = len(truncated_input_sentences) // batch_size + int(\n",
" len(truncated_input_sentences) % batch_size > 0\n",
" )\n",
"\n",
" embedding_batch_list = []\n",
"\n",
" for i in range(n_batches):\n",
" start_idx = i * batch_size\n",
" end_idx = min((i + 1) * batch_size, len(truncated_input_sentences))\n",
"\n",
" with torch.no_grad():\n",
" embedding_batch_list.append(\n",
" self.forward(truncated_input_sentences[start_idx:end_idx])\n",
" .detach()\n",
" .cpu()\n",
" )\n",
"\n",
" return torch.cat(embedding_batch_list)\n",
"\n",
"\n",
"class StarEncoder(BaseEncoder):\n",
" def __init__(self, device, max_input_len, maximum_token_len):\n",
" super().__init__(\n",
" device,\n",
" max_input_len,\n",
" maximum_token_len,\n",
" model_name=\"bigcode/starencoder\",\n",
" )\n",
"\n",
" def forward(self, input_sentences):\n",
" inputs = self.tokenizer(\n",
" [\n",
" f\"{self.tokenizer.cls_token}{sentence}{self.tokenizer.sep_token}\"\n",
" for sentence in input_sentences\n",
" ],\n",
" padding=\"longest\",\n",
" max_length=self.maximum_token_len,\n",
" truncation=True,\n",
" return_tensors=\"pt\",\n",
" )\n",
"\n",
" outputs = self.encoder(**set_device(inputs, self.device))\n",
" embedding = pool_and_normalize(outputs.hidden_states[-1], inputs.attention_mask)\n",
"\n",
" return embedding\n",
"\n",
"\n",
"class CodeBERT(BaseEncoder):\n",
" def __init__(self, device, max_input_len, maximum_token_len):\n",
" super().__init__(\n",
" device,\n",
" max_input_len,\n",
" maximum_token_len,\n",
" model_name=\"microsoft/codebert-base\",\n",
" )\n",
"\n",
" self.tokenizer = AutoTokenizer.from_pretrained(\"microsoft/codebert-base\")\n",
"\n",
" def forward(self, input_sentences):\n",
" inputs = self.tokenizer(\n",
" [sentence for sentence in input_sentences],\n",
" padding=\"longest\",\n",
" max_length=self.maximum_token_len,\n",
" truncation=True,\n",
" return_tensors=\"pt\",\n",
" )\n",
"\n",
" inputs = set_device(inputs, self.device)\n",
"\n",
" outputs = self.encoder(inputs[\"input_ids\"], inputs[\"attention_mask\"])\n",
"\n",
" embedding = outputs[\"pooler_output\"]\n",
"\n",
" return torch.cat([torch.Tensor(el)[None, :] for el in embedding])\n",
"\n",
"\n",
"def evaluate(model_path, maximum_raw_length, maximum_input_length, device):\n",
" if \"starencoder\" in model_path.lower():\n",
" model = StarEncoder(\n",
" device, maximum_raw_length, maximum_input_length\n",
" )\n",
" elif \"codebert\" in model_path.lower():\n",
" model = CodeBERT(\n",
" device, maximum_raw_length, maximum_input_length\n",
" )\n",
" else:\n",
" raise ValueError(\n",
" \"Unsupported model type. We currently support starencoder and codebert.\"\n",
" )\n",
"\n",
" model = model.to(device)\n",
" model.eval()\n",
"\n",
" test_data = get_dataset(maximum_raw_length)\n",
"\n",
" source_entries, target_entries = [], []\n",
" for source, target in test_data:\n",
" source_entries.append(source)\n",
" target_entries.append(target)\n",
"\n",
" source_embeddings = model.encode(source_entries)\n",
" target_embeddings = model.encode(target_entries)\n",
"\n",
" recall_at_1, recall_at_5, mean_reciprocal_rank = retrieval_eval(\n",
" source_embeddings, target_embeddings\n",
" )\n",
"\n",
" print(\n",
" f\"\\n{model_path}: R@1: {recall_at_1.item()}, R@5: {recall_at_5.item()}, MRR: {mean_reciprocal_rank.item()}\"\n",
" )"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Using pad_token, but it is not set yet.\n",
"Using sep_token, but it is not set yet.\n",
"Using cls_token, but it is not set yet.\n",
"Using mask_token, but it is not set yet.\n",
"Some weights of the model checkpoint at bigcode/starencoder were not used when initializing BertModel: ['cls.predictions.transform.dense.weight', 'cls.seq_relationship.weight', 'cls.predictions.transform.LayerNorm.bias', 'cls.seq_relationship.bias', 'cls.predictions.bias', 'cls.predictions.decoder.weight', 'cls.predictions.decoder.bias', 'cls.predictions.transform.LayerNorm.weight', 'cls.predictions.transform.dense.bias']\n",
"- This IS expected if you are initializing BertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
"- This IS NOT expected if you are initializing BertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
"Loading cached shuffled indices for dataset at /mnt/home/research-BertBigCode/resources/data/transcoder_evaluation_gfg/cache-e9f62aa12abed28d.arrow\n",
"Loading cached processed dataset at /mnt/home/research-BertBigCode/resources/data/transcoder_evaluation_gfg/cache-62c8dbaa90db85ee_*_of_00096.arrow\n",
"Loading cached shuffled indices for dataset at /mnt/home/research-BertBigCode/resources/data/transcoder_evaluation_gfg/cache-f652c1e33d8c1a14.arrow\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"starencoder: R@1: 0.7222222089767456, R@5: 0.8767361044883728, MRR: 0.7930026054382324\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Loading cached shuffled indices for dataset at /mnt/home/research-BertBigCode/resources/data/transcoder_evaluation_gfg/cache-e9f62aa12abed28d.arrow\n",
"Loading cached processed dataset at /mnt/home/research-BertBigCode/resources/data/transcoder_evaluation_gfg/cache-62c8dbaa90db85ee_*_of_00096.arrow\n",
"Loading cached shuffled indices for dataset at /mnt/home/research-BertBigCode/resources/data/transcoder_evaluation_gfg/cache-f652c1e33d8c1a14.arrow\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"codebert: R@1: 0.0052083334885537624, R@5: 0.02777777798473835, MRR: 0.025095948949456215\n"
]
}
],
"source": [
"for eval_cfg in EVAL_CONFIGS:\n",
" evaluate(**eval_cfg)"
]
}
],
"metadata": {
"interpreter": {
"hash": "ae635839a86c404533bb974203baf1bd26d9dc49bfbf145b45e9350c30045fdd"
},
"kernelspec": {
"display_name": "Python 3.9.13 64-bit ('accelerate')",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.10.9"
},
"orig_nbformat": 4
},
"nbformat": 4,
"nbformat_minor": 2
}
Loading