biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

Adds method to read the pooling types from model's files (#9506)

Browse Source 
Signed-off-by: Flavia Beo <flavia.beo@ibm.com>
Signed-off-by: Max de Bayser <mbayser@br.ibm.com>
Co-authored-by: Max de Bayser <mbayser@br.ibm.com>
This commit is contained in:
Flávia Béo
2024-11-07 05:42:40 -03:00
committed by GitHub
parent e036e527a0
commit aa9078fa03
10 changed files with 342 additions and 25 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
tests/utils.py
View File

@@ -15,6 +15,7 @@ import openai
import pytest
import requests
import torch
import torch.nn.functional as F
from openai.types.completion import Completion
from typing_extensions import ParamSpec
@@ -515,13 +516,14 @@ def compare_all_settings(model: str,
ref_result = copy.deepcopy(ref_result)
compare_result = copy.deepcopy(compare_result)
if "embedding" in ref_result and method == "encode":
ref_embedding = torch.tensor(ref_result["embedding"])
compare_embedding = torch.tensor(
compare_result["embedding"])
mse = ((ref_embedding - compare_embedding)**2).mean()
assert mse < 1e-6, (
sim = F.cosine_similarity(
torch.tensor(ref_result["embedding"]),
torch.tensor(compare_result["embedding"]),
dim=0,
)
assert sim >= 0.999, (
f"Embedding for {model=} are not the same.\n"
f"mse={mse}\n")
f"cosine_similarity={sim}\n")
del ref_result["embedding"]
del compare_result["embedding"]
assert ref_result == compare_result, (