Quantized variants share exact source weights within each family. Measured QFS KL; RTN format tests, not optimizer-quality rankings.
Michel Belleau PRO
malaiwah
AI & ML interests
None yet
Recent Activity
liked a model about 12 hours ago
zai-org/GLM-5.3 new activity about 19 hours ago
malaiwah/Qwen3.8-27B-EXL3-K5K6-context:Getting garbled text updated a dataset 1 day ago
malaiwah/GLM-5.3-Flash-fidelity-suite-v1Organizations
None yet
Tiny random models, quantized format fixtures and verified CPU captures for QFS. Pipeline tests, not assistants or quality benchmarks.
-
malaiwah/glm-moe-dsa-tiny-random-bf16
Text Generation • 278k • Updated • 429 -
malaiwah/qwen3-5-tiny-random-bf16
Text Generation • 245k • Updated • 401 -
malaiwah/glm5-next-tiny-random-bf16
Image-Text-to-Text • 422k • Updated • 28 -
malaiwah/qwen4-exp-tiny-random-bf16
Text Generation • 342k • Updated • 380
Four runtime-specific Qwen3.8 EXL3 builds. v5 10M shared-head, body-only KL vs BF16; advisory, not a quality ranking. Context tested separately.
-
malaiwah/Qwen3.8-27B-EXL3-K5K6-hydrated
Image-Text-to-Text • 11B • Updated • 12.3k • 11 -
malaiwah/Qwen3.8-27B-EXL3-K5K6
Image-Text-to-Text • 15B • Updated • 1.18k • 1 -
malaiwah/Qwen3.8-27B-K4
Image-Text-to-Text • 14B • Updated • 1.12k -
malaiwah/Qwen3.8-27B-EXL3-K5K6-context
Image-Text-to-Text • 10B • Updated • 851 • 8
Quantized variants share exact source weights within each family. Measured QFS KL; RTN format tests, not optimizer-quality rankings.
Independent tiny random native architectures and qualified CPU roots. No trained weights or production-model fine-tuning lineage.
Tiny random models, quantized format fixtures and verified CPU captures for QFS. Pipeline tests, not assistants or quality benchmarks.
-
malaiwah/glm-moe-dsa-tiny-random-bf16
Text Generation • 278k • Updated • 429 -
malaiwah/qwen3-5-tiny-random-bf16
Text Generation • 245k • Updated • 401 -
malaiwah/glm5-next-tiny-random-bf16
Image-Text-to-Text • 422k • Updated • 28 -
malaiwah/qwen4-exp-tiny-random-bf16
Text Generation • 342k • Updated • 380
GLM-5.3-Flash receipt-backed fidelity: named panels, lanes and capture scopes. Descriptive evidence, not universal or native-serving rankings.
Four runtime-specific Qwen3.8 EXL3 builds. v5 10M shared-head, body-only KL vs BF16; advisory, not a quality ranking. Context tested separately.
-
malaiwah/Qwen3.8-27B-EXL3-K5K6-hydrated
Image-Text-to-Text • 11B • Updated • 12.3k • 11 -
malaiwah/Qwen3.8-27B-EXL3-K5K6
Image-Text-to-Text • 15B • Updated • 1.18k • 1 -
malaiwah/Qwen3.8-27B-K4
Image-Text-to-Text • 14B • Updated • 1.12k -
malaiwah/Qwen3.8-27B-EXL3-K5K6-context
Image-Text-to-Text • 10B • Updated • 851 • 8