Robert Smith

tcclaviger

·

Rob-P-Smith

AI & ML interests

Local SWE supporting LLMs that can be used for private data. Currently expanding capabilities to enable use of larger models supported by a custom tool set environment providing context 7 like knowledge with added DSA, Software design, and web crawling capabilities to enable autonomous expansion of knowledge as the LLM deems necessary.

Recent Activity

new activity about 10 hours ago

tcclaviger/Qwen3.8-Flash-Next-MXFP4-FP8-GPTQ:Greedy decoding divergence on 28.04.10 / revision 2a052b4, including eager mode

new activity about 10 hours ago

tcclaviger/Qwen3.8-Flash-Next-MXFP4-FP8-GPTQ:General appreciation post

new activity about 10 hours ago

tcclaviger/Qwen3.8-Flash-Next-MXFP4-FP8-GPTQ:vLLM worker hang in `sample_tokens` → EngineDeadError (ROCm / RDNA4, MTP spec-decode)

View all activity

Organizations

None yet

New activity in tcclaviger/Qwen3.8-Flash-Next-MXFP4-FP8-GPTQ about 10 hours ago

Greedy decoding divergence on 28.04.10 / revision 2a052b4, including eager mode

#1 opened 11 days ago by

General appreciation post

#2 opened 7 days ago by

vLLM worker hang in `sample_tokens` → EngineDeadError (ROCm / RDNA4, MTP spec-decode)

#3 opened 7 days ago by

New activity in tcclaviger/Qwen3.8-Flash-Next-MXFP4-FP8 20 days ago

DockerHub tag is not public

#1 opened 29 days ago by

New activity in tcclaviger/Qwen3.8-27B-RFI8-AA about 1 month ago

Good quant, was fun to test

#1 opened about 1 month ago by

New activity in DavidAU/Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-NEO-MAX-MTP-GGUF about 1 month ago

Eval of 711, Tess, AEON, ThinkingCap, 27B base with no confounds

#35 opened 2 months ago by

New activity in meta-models/Muse-Glimmer-30B about 1 month ago

Even Big Lebowski doesn’t know where the money went: Meta’s "Superintelligence" is a sub-$5 basement-tier Frankenstein

#43 opened about 2 months ago by

A pelican test

#38 opened about 2 months ago by

vLLM DFlash spec decode: 6 fixes needed to get the muse-glimmer image working (with working Dockerfile)

#40 opened about 2 months ago by

New activity in DavidAU/Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-NEO-MAX-MTP-GGUF about 2 months ago

Absolute Beast : Claude (Opus) Level of performance confirmed

#41 opened 2 months ago by

Anyone have coding Benchmarks?

#45 opened about 2 months ago by

New activity in DavidAU/Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP 2 months ago

Let Great Quantization Run Begin

#1 opened 2 months ago by

New activity in DavidAU/Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-NEO-MAX-MTP-GGUF 2 months ago

THE TRUE GOD-TIER MASTERPIECE: UNMATCHED QUALITY UNLEASHED

#12 opened 2 months ago by

any plan to release the full precision model?

#3 opened 2 months ago by

New activity in nightmedia/Qwen3.6-27B-Architect-Polaris2-Fable-B-F451 2 months ago

Qwen3.6 27B family evaluation

#3 opened 2 months ago by

New activity in poolside/Laguna-S-2.1 2 months ago

Thinking is as broken as thinking gets

#17 opened 2 months ago by

Model is decent, its faster than Qwen and pretty close on real world tool calls

#22 opened 2 months ago by

New activity in DavidAU/Qwen3.6-40B-Claude-4.6-Opus-Deckard-Heretic-Uncensored-Thinking-NEO-CODE-Di-IMatrix-MAX-GGUF 2 months ago

Recommendation request for uncensored 40B-class model for physical synthetic companion project

#25 opened 3 months ago by

New activity in migtissera/Tess-4-27B 3 months ago

Recommended temperature and other settings?

#8 opened 3 months ago by

Complements To Your Work

#7 opened 3 months ago by