Downloads · 30 days
16
1% of all-time downloads
Garrettkinsman/Qwen3.5-122B-A10B-HALO-GAHO
Qwen3.5-122B-A10B-HALO-GAHO is a machine learning model from Garrettkinsman. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Custom quantized Qwen3.5-122B-A10B MoE model, optimized for AMD Strix Halo (Framework1) running via llama.cpp.
Downloads · 30 days
16
1% of all-time downloads
All-time downloads
2K
Public
Repo size
155 GB
Likes
1
Public
Click a slice to open those files.
.gguf232 GB · 100%
From the Hugging Face model README
Custom quantized Qwen3.5-122B-A10B MoE model, optimized for AMD Strix Halo (Framework1) running via llama.cpp.
Target: huge token injection for OpenClaw / SybilClaw autonomous agent orchestration. Built by Agent: Gaho 🪨
| Tensor Type | Quant | Reason |
|---|---|---|
| Attention Q/K/V/O | Q8_0 | Precision critical for reasoning & tool use |
| FFN Sparse Experts | Q4_K | 32 experts, only 6 active/token |
| FFN Shared Expert | Q6_K | Always active, don't crush quality |
| MoE Router Gate | Q6_K | Wrong routing = wrong expert = garbage |
| Token Embeddings | Q8_0 | Input quality matters |
| Output Projection | Q8_0 | Output quality matters |
| RMS Norms | F32 | Tiny, numerically sensitive |
| SSM (alpha,beta,dt,a,ba,out,conv1d) | F32 | Recurrence — must stay F32 |
llama-quantize --tensor-type (merged Q8 base)