knn retrieval evaluation #770
Conversation
GentleZhu
added
the
draft
| @@ -0,0 +1,128 @@ | |||
| import torch as th | |||
There was a problem hiding this comment.
Please add a license head for each python code.
| import torch as th | ||
| import time | ||
| import graphstorm as gs | ||
| from graphstorm.utils import is_distributed |
There was a problem hiding this comment.
It is better to move import graphstorm related code together.
And for import, usually the order will be:
import system/builtin libraries like os, time, etc.
import pip packages
import local codes.
| pred = set(pred) | ||
|
|
||
| overlap = len(pred & ground_truth) | ||
| #if overlap > 0: |
| from graphstorm.utils import setup_device | ||
| from graphstorm.model.utils import load_gsgnn_embeddings | ||
|
|
||
| def calculate_recall(pred, ground_truth): |
There was a problem hiding this comment.
Can you give a description of how you compute recall in the function doc?
| index_dimension = embs[config.target_ntype].size(1) | ||
| # Number of clusters (higher values lead to better recall but slower search) | ||
| #nlist = 750 | ||
| #quantizer = faiss.IndexFlatL2(index_dimension) # Use Flat index for quantization |
There was a problem hiding this comment.
Remove the commented codes.
| index = faiss.IndexFlatIP(index_dimension) | ||
| index.add(embs[config.target_ntype]) | ||
|
|
||
| #print(scores.abs().mean()) |
| len_dataloader = max_num_batch = len(test_dataloader) | ||
| tensor = th.tensor([len_dataloader], device=device) | ||
| if is_distributed(): | ||
| th.distributed.all_reduce(tensor, op=th.distributed.ReduceOp.MAX) |
There was a problem hiding this comment.
Do we need to make it distributed?
| #print(blocks[0].edges(form='uv', etype='also_buy')) | ||
| #breakpoint() | ||
| # print(dgl.NID) | ||
| if 'also_buy' in blocks[0].etypes: |
There was a problem hiding this comment.
Is this implemented specifically for amazon review?
| @@ -1065,6 +1066,32 @@ def save_full_node_embeddings(g, save_embed_path, | |||
|
|
|||
| save_shuffled_node_embeddings(shuffled_embs, save_embed_path, save_embed_format) | |||
|
|
|||
| def load_gsgnn_embeddings(emb_path): | |||
There was a problem hiding this comment.
Can you check if load_pytorch_embedding is useful?
Issue #, if available:
Description of changes:
This is a draft PR because I found the recall@K is extremly low and it seems the remap get wrong from reading embeddings.
By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.