Support for f16 vector metrics

[generall]

Is your feature request related to a problem? Please describe.

Currently Qdrant supports storage of raw vectors with 2 datatypes: f32 and uint8.
In some applications it is also desirable to have f16 datatype support.

Qdrant VectorStorage itself is generic to the datatype and technically can store vectors of any type with no additional changes.

There is, however, a component which depends on datatype and requires specialized implementation - that is metrics.

Describe the solution you'd like

Provide SIMD-optimized implementations for the following:

impl Metric<f16> for DotProductMetric {
impl Metric<f16> for CosineMetric {
impl Metric<f16> for EuclidMetric {
impl Metric<f16> for ManhattanMetric {

for avx2, sse and neon architectures in a similar way as it is implemented for f32.

Please also include unit-tests to ensure correctness of the implementation.

Describe alternatives you've considered

• add a compile-time option to switch VectorElementType between f32 and f16, but this approach is less flexible and requires more changes

Additional context

[generall]

/bounty $300

[algora-pbc]

💎 $300 bounty • Qdrant

Steps to solve:

1. Start working: Comment /attempt #4110 with your implementation plan
2. Submit work: Create a pull request including /claim #4110 in the PR body to claim the bounty
3. Receive payment: 100% of the bounty is received 2-5 days post-reward. Make sure you are eligible for payouts

Additional opportunities:

• 🔴 Livestream on Algora TV while solving this bounty & earn $200 upon merge! Comment /livestream once live

Thank you for contributing to qdrant/qdrant!

Add a bounty • Share on socials

Attempt	Started (GMT+0)	Solution
🟢 @TheQuantumFractal	Apr 24, 2024, 6:56:51 PM	#4122

[TheQuantumFractal]

Happy to help with this. You just want SIMD implementations and tests for f16? (ie. code doesn't have to be refactored to account for f16 datatype correct?)
/attempt #4110

Options

• Cancel my attempt

[generall]

Happy to help with this.

Thanks 😃

code doesn't have to be refactored to account for f16 datatype correct?

yeah, integration of f16 into API and storage would be a separate issue.

[abhishek818]

Found a crate based on simd for f16.

[algora-pbc]

💡 @TheQuantumFractal submitted a pull request that claims the bounty. You can visit your bounty board to reward.

[abhishek818]

Great work @TheQuantumFractal !

[algora-pbc]

@TheQuantumFractal: You've been awarded a $360 bounty by Qdrant! 👉 Complete your Algora onboarding to collect the bounty.

[algora-pbc]

🎉🎈 @TheQuantumFractal has been awarded $360! 🎈🎊

• Share on socials
• Give feedback