Downloads · 30 days
167
23% of all-time downloads
spoomplesmaxx-magpie-35B-A3-i1-GGUF
aimeri/spoomplesmaxx-magpie-35B-A3-i1-GGUF
spoomplesmaxx-magpie-35B-A3-i1-GGUF is a text generation model from aimeri. Use it when you need the model to write or continue text. It is set up for transformers.
Weighted / imatrix quants of aimeri/spoomplesmaxx-magpie-35B-A3.
All-time downloads
726
Public
Repo size
40.1 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.gguf40.1 GB · 100%
At a glance
- Task
- Text Generation
- Library
- transformers
- Access
- Public
- Created
- Jul 27, 2026
- Updated
- Jul 27, 2026
- SHA
- 86e159cb
Try a prompt
Base models
Model card
From the Hugging Face model README
spoomplesmaxx-magpie-35B-A3-i1-GGUF
Weighted / imatrix quants of aimeri/spoomplesmaxx-magpie-35B-A3.
Static quants are not provided: at every size that matters the imatrix quant of the same name is at least as good, so this repo is the whole set.
Provenance
| llama.cpp | b9222 (pinned; Qwen3.5 MoE conversion needs >= b9180) |
| Source | bf16 GGUF converted from the SFT checkpoint with convert_hf_to_gguf.py |
| imatrix | llama-imatrix --parse-special, 192 MB |
| Calibration | 48 MB, 12.0M in-domain tokens (952 documents) rendered through this model's own chat template and truncated to exact 8192-token windows so imatrix chunks land on document boundaries; the generic term is unsloth's published imatrix for this architecture, merged after our own pass |
The in-domain half exists because this is a sparse MoE: a token only touches its routed experts, so a small generic corpus leaves rare experts with no importance data at all. Coverage was gated before quantizing — every quantizable tensor has importance data and every expert row was activated above a token floor.
The Q4_K_M quant was verified after quantization: greedy generations terminate
on <|im_end|> (not the token cap), think blocks close, no stray glyphs.
Measured
KL divergence of each quant against this model's own bf16, on a holdout that was excluded from calibration (evaluating on your own calibration domain flatters the numbers). Lower KLD is closer to the unquantized model.
| quant | size | KLD mean | KLD median | KLD p99 | ΔPPL | same top-1 |
|---|---|---|---|---|---|---|
i1-IQ4_XS | 18.7 GB | 0.0243 | 0.0125 | 0.146 | +1.04% | 100.0% |
i1-Q4_K_M | 21.4 GB | 0.0206 | 0.0102 | 0.130 | +1.04% | 100.0% |
same top-1 is the share of holdout positions where the quant's most likely
token is unchanged from bf16.
Which one do I want?
Roughly, higher in the table is smaller and worse. i1-Q4_K_M is the
recommended default; i1-IQ4_XS if you need to save a little; i1-Q6_K if you
have the memory and want to stop thinking about it.
Usage
llama-server -m spoomplesmaxx-magpie-35B-A3-i1-GGUF.i1-Q4_K_M.gguf -ngl 99 -c 32768 --jinja
--jinja matters: the chat template is embedded in the GGUF and pre-opens the
<think> block, which is the behaviour the model was trained for. Tool calls use
Qwen3.5's XML convention
(<tool_call><function=NAME><parameter=KEY>VALUE</parameter></function></tool_call>).
- Task
- Text Generation
- Library
- transformers
- Created
- Jul 27, 2026
- Updated
- Jul 27, 2026