Skip to content

Finetune model #44

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

Open
ivanleomk wants to merge 3 commits into
base: main
Choose a base branch
from
Open

Finetune model #44

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
147f211
Refactored Justin's script for fine-tuning and generated some initial…
ivanleomk Feb 10, 2024
4533fef
Current Status
ivanleomk Feb 12, 2024
3e60115
Modified finetune.py and updated the generted benchmarks with the values
ivanleomk Feb 13, 2024
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
178 changes: 178 additions & 0 deletions applications/finetune-quora-embeddings/finetune.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,178 @@
{
"accuracy": 0.8534110003360591,
"precision": 0.6358073009321853,
"recall": 0.8624517374517374,
"AUC": 0.856565037216022,
"model": "BAAI/bge-base-en-v1.5",
"dataset_size": 1000,
"epochs": 6,
"warmup_steps": 500,
"scheduler": "warmupcosine",
"batch_size": 64,
"loss": [
0.786621867396219,
0.7872971821573767,
0.7884130523672186,
0.7900412655915894,
0.7917944440127781,
0.7939522746116873
]
}
{
"accuracy": 0.8606698778985101,
"precision": 0.6471675314521288,
"recall": 0.8788610038610039,
"AUC": 0.8670162057962933,
"model": "BAAI/bge-base-en-v1.5",
"dataset_size": 2000,
"epochs": 6,
"warmup_steps": 500,
"scheduler": "warmupcosine",
"batch_size": 64,
"loss": [
0.7873146839028387,
0.7901949982729476,
0.7947561069240852,
0.7996619526215534,
0.8042983311953955,
0.8074228142497679
]
}
{
"accuracy": 0.8689817407863784,
"precision": 0.6643356643356644,
"recall": 0.8803088803088803,
"AUC": 0.8729334335898887,
"model": "BAAI/bge-base-en-v1.5",
"dataset_size": 4000,
"epochs": 6,
"warmup_steps": 500,
"scheduler": "warmupcosine",
"batch_size": 64,
"loss": [
0.7903197146907419,
0.7999941228386948,
0.8096067455629543,
0.8145869930614067,
0.8169794554574514,
0.8153368349249744
]
}
{
"accuracy": 0.8765990814383331,
"precision": 0.6825707405177603,
"recall": 0.8754826254826255,
"AUC": 0.8762095840761048,
"model": "BAAI/bge-base-en-v1.5",
"dataset_size": 8000,
"epochs": 6,
"warmup_steps": 500,
"scheduler": "warmupcosine",
"batch_size": 64,
"loss": [
0.7995496614305293,
0.8161136835788527,
0.8199591566340426,
0.8237846756261787,
0.8205980826591113,
0.8205005653419446
]
}
{
"accuracy": 0.8808334266830963,
"precision": 0.6961020773360305,
"recall": 0.8636100386100386,
"AUC": 0.8748247129592863,
"model": "BAAI/bge-base-en-v1.5",
"dataset_size": 16000,
"epochs": 6,
"warmup_steps": 500,
"scheduler": "warmupcosine",
"batch_size": 64,
"loss": [
0.8170246518823161,
0.8263173567661242,
0.8277094062084467,
0.8263304392574006,
0.8269694628304216,
0.8265691784767804
]
}
{
"accuracy": 0.8865240282289683,
"precision": 0.7131814155729125,
"recall": 0.8549227799227799,
"AUC": 0.8754993184807188,
"model": "BAAI/bge-base-en-v1.5",
"dataset_size": 32000,
"epochs": 6,
"warmup_steps": 500,
"scheduler": "warmupcosine",
"batch_size": 64,
"loss": [
0.8289369361547437,
0.8309696091067318,
0.8334979335164595,
0.8297416940865507,
0.8296396127399528,
0.8295118543727288
]
}
{
"accuracy": 0.8893917329449983,
"precision": 0.721835883171071,
"recall": 0.8516409266409266,
"AUC": 0.8762216303518273,
"model": "BAAI/bge-base-en-v1.5",
"dataset_size": 64000,
"epochs": 6,
"warmup_steps": 500,
"scheduler": "warmupcosine",
"batch_size": 64,
"loss": [
0.8343571459849803,
0.8343571459849803,
0.8387984050821868,
0.8387984050821868,
0.8352079736992297,
0.8352079736992297,
0.8334507561970751,
0.8334507561970751,
0.8332350622287716,
0.8332350622287716,
0.8327780129486679,
0.8327780129486679
]
}
{
"accuracy": 0.8937605018483253,
"precision": 0.7401265173533937,
"recall": 0.8357142857142857,
"AUC": 0.8735099510263624,
"model": "BAAI/bge-base-en-v1.5",
"dataset_size": 128000,
"epochs": 6,
"warmup_steps": 500,
"scheduler": "warmupcosine",
"batch_size": 64,
"loss": [
0.8336498402530864,
0.8407761364043034,
0.8407761364043034,
0.8387109362309478,
0.8408796564561053,
0.8408796564561053,
0.8391623132712105,
0.8404157875124303,
0.8404157875124303,
0.8392159143841129,
0.8396718885264413,
0.8396718885264413,
0.8371604374328758,
0.8377142666446907,
0.8377142666446907,
0.8373806355548399,
0.8373202179603153,
0.8373202179603153
]
}
Loading