btibor91

X · 19 atoms · 2026-06-28 → 2026-06-29 · 信用档先验 cred3

🟦 fact 10 · 🟥 take 9 · stance ▲0/▼0/◆7

覆盖实体 (top 10)

• OpenAI — 4 atoms
• Qwen3.6-27B-FP8-MTP — 2 atoms
• Anthropic — 2 atoms
• DeepSeek-V4-Flash-IQ2XXS-w2Q2K-AProjQ8-SExpQ8-OutQ8-chat-v2-imatrix — 1 atoms
• Qwen3.6-35B-A3B — 1 atoms
• Qwen3.6-35B-A3B-GGUF:UD-Q3_K_M — 1 atoms
• OpenCode Go — 1 atoms
• DeepSeek V4-Flash — 1 atoms
• Qwen3.6-35B-A3B-UD-Q3_K_M — 1 atoms
• Ideogram-4-FP8-Scaled — 1 atoms

话题分布 (top 8)

硬件要求 (5) · 模型部署 (4) · 硬件需求 (3) · 推荐 (2) · 硬件合作 (2) · 产品开发 (2) · 硬件推荐 (1) · 模型协作 (1)

样例 atoms

• 2026-06-28 · DeepSeek-V4-Flash-IQ2XXS-w2Q2K · antirez/ds4 - DeepSeek-V4-Flash-IQ2XXS-w2Q2K-AProjQ8-SExpQ8-OutQ8-chat-v2-imatrix (128GB Mac machines, 250k context)
• 2026-06-28 · Qwen3.6-27B-FP8-MTP · vLLM - Qwen3.6-27B-FP8-MTP (DGX Spark, 250k context)
• 2026-06-28 · Qwen3.6-35B-A3B · llama.cpp - Qwen3.6-35B-A3B (NVIDIA GPUs with ~20GB VRAM, 65k context)
• 2026-06-28 · Qwen3.6-35B-A3B-GGUF:UD-Q3_K_M · Try unsloth/Qwen3.6-35B-A3B-GGUF:UD-Q3_K_M, via llama.cpp (64K context)
• 2026-06-28 · OpenCode Go · All three are in a group chat using Maestro via OpenCode Go

← Author 索引 · 实体目录