Quantization Request: tencent/Hy-MT2-30B-A3B (oQ4e / BF16)

#2
by 0xtimi2233 - opened

Hi splats,

Could you please process an oQe build for the following model when you have time?

  1. Model Link: tencent/Hy-MT2-30B-A3B
  2. Quantization Format: BF16 (Optimized for M3/M4 native support)
  3. Preferred Tiers: oQ4e

Thank you for your time and amazing work on these Apple Silicon optimized quants!

Sign up or log in to comment