feat(llama.cpp): expose cache_type_k and cache_type_v for quant of kv cache

[mudler]

Description

This pull request introduces a new feature to handle cache types in the backend system. The changes span multiple files, adding new fields and parameters to support this functionality.

New cache type support:

• backend/backend.proto: Added CacheTypeKey and CacheTypeValue fields to the ModelOptions message.
• backend/cpp/llama/grpc-server.cpp: Updated the params_parse function to parse cache_type_k and cache_type_v from the ModelOptions request.
• core/backend/options.go: Modified the grpcModelOpts function to include CacheTypeKey and CacheTypeValue in the ModelOptions.
• core/config/backend_config.go: Added CacheTypeK and CacheTypeV fields to the LLMConfig struct.

To use it, enable flash_attention and specify a quant cache, like this:

cache_type_v: "q4_0"
flash_attention: true
cache_type_k: "q4_0"

Notes for Reviewers

Signed commits

• Yes, I signed my commits.

[netlify]

✅ Deploy Preview for localai ready!

Name	Link
🔨 Latest commit	da7a65d
🔍 Latest deploy log	https://app.netlify.com/sites/localai/deploys/67520ed2f8326e00084b2d27
😎 Deploy Preview	https://deploy-preview-4329--localai.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site configuration.

[netlify]

✅ Deploy Preview for localai ready!

Name	Link
🔨 Latest commit	c8bc017
🔍 Latest deploy log	https://app.netlify.com/sites/localai/deploys/6752b2b9d17be90008f9e53f
😎 Deploy Preview	https://deploy-preview-4329--localai.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site configuration.