😀 ... it's amazing!
Schestex
Schestex
AI & ML interests
None yet
Recent Activity
liked a model about 12 hours ago
WarlaxZ/Ternary-Bonsai-2-27B-NInfer-3090 new activity 1 day ago
bottlecapai/ThinkingCap-Qwen3.8-27B:ThinkingCap-Qwen3.8-27B-NInfer updated a model 3 days ago
Schestex/ThinkingCap-Qwen3.8-27B-NInferOrganizations
None yet
ThinkingCap-Qwen3.8-27B-NInfer
🚀 2
5
#4 opened about 1 month ago
by
Schestex
Fewer tokens usage with a other kernel ...
#4 opened 18 days ago
by
Schestex
Great work! Any news on that "[vocab-trimmer] release soon tool"?
2
#2 opened 3 months ago
by
vimr
Could ThinkingCap be combined with Amazon LAR-27B for stronger agentic coding?
#8 opened 10 days ago
by
Schestex
BFCL Benchmark after QLoRA
3
#5 opened 17 days ago
by
Schestex
Reducing the vocabulary to free up more VRAM and potentially improve decoding
6
#3 opened about 1 month ago
by
Schestex
24 GB optimized NVFP4 NInfer artifact
4
#7 opened about 2 months ago
by
Schestex
D48 + O44 + I64 PERFORMANCE CEILING
1
#2 opened about 1 month ago
by
Schestex
Great!
3
#1 opened about 1 month ago
by
pushkinops
VERSION / QUANT: NVFP4 NInfer / Blackwell build of Cold-Fusion V1.1 / Blackwell Pro 4000
👍❤️ 2
5
#18 opened about 2 months ago
by
Schestex
New Model based on Qwen 3.8
➕🔥 27
18
#23 opened 3 months ago
by
jaskfsafdjlk
MTP Hybrid-A1 Controlled OLD W8-MTP vs Hybrid-A1 benchmarks
2
#1 opened about 2 months ago
by
Schestex
How does this compare to neroued/Qwen3.8-27B-nvfp4-NInfer?
1
#3 opened 3 months ago
by
jmander11
nInfer version
🚀 2
6
#11 opened 2 months ago
by
ayucg
Blackwell Pro 4000 SFF / non SFF Prefill and MTP 3-5 Results Text / Code
#1 opened 2 months ago
by
Schestex
Would you consider GSQ-RCO IQ3_S + MTP for a Qwen3.8-27B coding-focused fine-tune?
🚀 1
6
#8 opened 2 months ago
by
Schestex
Worth validating GSQ-RCO IQ3_S against UD-IQ3_S on Dirk?
2
#16 opened 2 months ago
by
Schestex
NVFP4-MTP build available for each model type?
3
#4 opened 3 months ago
by
Schestex
This model is COOKING HARD
❤️🔥 1
25
#10 opened 2 months ago
by
nmh2k3