Collection of Qwen3.8-27B quantizations, fine-tunes, and speculative decoding models.
Jake
jakeatx
AI & ML interests
distillation, MoE tuning + optimizing, offloading
Recent Activity
updated a collection 4 days ago
Qwen3.8-27B Models published a model 4 days ago
jakeatx/ATX-Swift-1.5-Qwen3.8-27B-Uncensored-IQ4_XS-M-GGUF updated a model 4 days ago
jakeatx/ATX-Swift-1.5-Qwen3.8-27B-Uncensored-IQ4_XS-M-GGUFOrganizations
None yet
Qwen3.8 Perian checkpoints: 48->32 layers, 384->288 experts, 50% PLE, and rank-32 QLoRA on 12,558 math, agentic, and coding traces.
-
jakeatx/ream-288-rank32-3m-adapter
Text Generation • Updated • 399 • 1 -
jakeatx/slimder-qwen38-ream288-ngram50-mixed-recovery-qlora-v2
Text Generation • Updated • 52 -
jakeatx/slimder-qwen38-ream288-depth32-agentic-ngram50-mixed-nvfp4-v2
Text Generation • 57B • Updated • 818 -
jakeatx/Qwopus3.8-27B-Flash-ATX-IQ4_XS-M-GGUF
Text Generation • 27B • Updated • 1.44k • 2
Collection of Qwen3.8-27B quantizations, fine-tunes, and speculative decoding models.
Qwen3.8 Perian checkpoints: 48->32 layers, 384->288 experts, 50% PLE, and rank-32 QLoRA on 12,558 math, agentic, and coding traces.
-
jakeatx/ream-288-rank32-3m-adapter
Text Generation • Updated • 399 • 1 -
jakeatx/slimder-qwen38-ream288-ngram50-mixed-recovery-qlora-v2
Text Generation • Updated • 52 -
jakeatx/slimder-qwen38-ream288-depth32-agentic-ngram50-mixed-nvfp4-v2
Text Generation • 57B • Updated • 818 -
jakeatx/Qwopus3.8-27B-Flash-ATX-IQ4_XS-M-GGUF
Text Generation • 27B • Updated • 1.44k • 2
models 26
jakeatx/ATX-Swift-1.5-Qwen3.8-27B-Uncensored-IQ4_XS-M-GGUF
Text Generation • 27B • Updated • 522
jakeatx/ATX-Swift-Qwen3.8-27B-Uncensored-IQ4_XS-M-GGUF
Text Generation • 27B • Updated • 1.98k • 4
jakeatx/Qwen3.8-27B-Ternary-PQ2_0-MTP-GGUF
27B • Updated • 987
jakeatx/Qwen3.8-27B-Ternary-PTQ1_0-MTP-GGUF
27B • Updated • 496
jakeatx/Qwen3.8-27B-GGUF
27B • Updated • 920
jakeatx/ream-288-rank32-3m-adapter
Text Generation • Updated • 399 • 1
jakeatx/Qwen3.8-Perian-GGUF
Text Generation • 78B • Updated • 3.32k • 3
jakeatx/Qwen3.8-Perian-BF16
Text Generation • 75B • Updated • 372
jakeatx/Agnes-3.0-Flash-GGUF
33B • Updated • 1.64k • 1
jakeatx/slimder-qwen38-reap384-s0
Image-Text-to-Text • 147B • Updated • 47
datasets 21
jakeatx/llamAmpere-v0.4-prefill-vs-beellama
Preview • Updated • 37
jakeatx/ATX-Swift-Qwen3.8-27B-Uncensored-IQ4_XS-M-LiveCodeBench-v6
Preview • Updated • 61
jakeatx/slimder-qwen38-ngram-activation-refine-20260901
Updated • 1.18k
jakeatx/slimder-qwen38-condensed-agentic100-20260901
Updated • 501
jakeatx/slimder-qwen38-agentic-ream-runs-20260901
Viewer • Updated • 1 • 184
jakeatx/slimder-qwen38-s3-results-20260831
Updated • 150
jakeatx/slimder-qwen38-s2-canary-20260831
Updated • 40
jakeatx/slimder-qwen38-s1-results-20260831
Updated • 94
jakeatx/turboquant-plus-pr197-eval
Updated • 73
jakeatx/nex-n2-mini-tq3-eval
Updated • 481