Feat: Add new model to model hub

[hahuyhoang411]

Describe Your Changes

Add 1-3B models with good coherence.

Self Checklist

• [x] Stable Zephyr 3B
• [x] Delete Yarn Mistral (issue: https://github.com/janhq/jan/issues/1549)
• [x] Dolphin Phi-2 (Thank you @tomasmcm https://github.com/janhq/jan/pull/1603)
• [ ] Nous-Hermes mixtral (https://huggingface.co/NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO#prompt-format)
• [ ] LlamaCorn-1.1B (https://huggingface.co/jan-hq/LlamaCorn-1.1B)

[tomasmcm]

Not many people have 30GB of RAM available to run the Mixtral models 🙈 Can we merge this without that? Having stable-zephyr and dolphin-phi would already be a great improvement.

[hahuyhoang411]

Exclude:

• Nous-Hermes mixtral: too big
• Stable Code 3B: not working

[hahuyhoang411]

Prepare to merge with https://github.com/janhq/jan/pull/1659