Downloads · 30 days
16
25% of all-time downloads
qwen3-8b-tool-recovery-FP8-dynamic
GurkanOz/qwen3-8b-tool-recovery-FP8-dynamic
qwen3-8b-tool-recovery-FP8-dynamic is a text generation model from GurkanOz. Use it when you need the model to write or continue text. The card lists the license as apache-2.0.
FP8-dynamic quantization of qwen3-8b-tool-recovery, from the QUENCH study (do trained agentic skills survive quantization?).
All-time downloads
64
Public
Parameters
8.2B
9.5 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors9.4 GB · 100%
Parameter types
How the weights are stored.
F8_E4M36.9B · 85%
Try a prompt
Base models
Model card
From the Hugging Face model README
qwen3-8b-tool-recovery-FP8-dynamic
FP8-dynamic quantization of qwen3-8b-tool-recovery, from the QUENCH study (do trained agentic skills survive quantization?).
| BF16 | this checkpoint (FP8) | |
|---|---|---|
| Recovery Rate (GLM-5.2 judge, 95% CI) | 65.6 (57.2–73.9) | 63.3 (55.0–71.7) — Δ -2.2pp, ns |
| Retry behavior | 87% | 88% |
| MMLU-400 guard | 71.9 | 71.4 (no loss) |
| Size | ~16 GB | ~9 GB |
Quantization: llm-compressor 0.12 FP8_DYNAMIC (per-token dynamic activations, no calibration), lm_head excluded — same single-path recipe as the study's base-model arm.
vllm serve GurkanOz/qwen3-8b-tool-recovery-FP8-dynamic \
--enable-auto-tool-choice --tool-call-parser hermes
See the BF16 card for training details, limitations, and the full QUENCH results grid; per-trial transcripts in the dataset.
- Task
- Text Generation
- Type
- qwen3
- License
- apache-2.0
- Languages
- en
- Created
- Jul 3, 2026
- Updated
- Jul 3, 2026