Your model already knows it's wrong. Asking costs 0.06 seconds and zero tokens.
โข 12
AI & ML interests
Contact: arxivgpt@gmail.com
Recent Activity
Papers
View all Papers Articles
SeaWolf-AIย
posted an update about 1 hour ago
Post
34
๐ผ๏ธ NO GPU, Only CPU : Z-Image model
Zero graphics cards. 46 seconds. Photoreal.
That laptop you're reading this on. No graphics card, right? It generates images.
No CUDA install. No Python environment. No driver changes. One binary, three model files. Done.
๐ Measured โ GPU count used: zero
512ร512 : 46.4 s
Korean prompt : 45.3 s (faster than English)
1024ร1024 : 192.7 s
Peak RAM : 6.42 GB
GPUs used : 0
(Intel Xeon Gold 6526Y ร2, 48 threads, Q4_0, 3 steps)
โก From 244 seconds to 46 โ 5.3ร
Run it on defaults and it takes 244 s. Switch to 3 steps and it's 48.6 s. Add VAE tiling and it's 46.4 s.
The biggest culprit was the default. Z-Image Turbo is distilled to paint in few strokes, but the tool's default is 20. We were throwing away 5ร for no reason. So were we, at first.
3 is the floor. Put 4 and 3 side by side and you cannot tell them apart. At 2 it collapses โ water droplets and wood grain vanish, and the surface turns cloth-like.
๐ Links
Model
FINAL-Bench/POCKET-Zimage-CPU
Live demo Space (runs on CPU)
FINAL-Bench/POCKET-Zimage-CPU
POCKET collection
FINAL-Bench/pocket-models-6a618ee5d23eafb7e185a5c6
Zero graphics cards. 46 seconds. Photoreal.
That laptop you're reading this on. No graphics card, right? It generates images.
No CUDA install. No Python environment. No driver changes. One binary, three model files. Done.
๐ Measured โ GPU count used: zero
512ร512 : 46.4 s
Korean prompt : 45.3 s (faster than English)
1024ร1024 : 192.7 s
Peak RAM : 6.42 GB
GPUs used : 0
(Intel Xeon Gold 6526Y ร2, 48 threads, Q4_0, 3 steps)
โก From 244 seconds to 46 โ 5.3ร
Run it on defaults and it takes 244 s. Switch to 3 steps and it's 48.6 s. Add VAE tiling and it's 46.4 s.
The biggest culprit was the default. Z-Image Turbo is distilled to paint in few strokes, but the tool's default is 20. We were throwing away 5ร for no reason. So were we, at first.
3 is the floor. Put 4 and 3 side by side and you cannot tell them apart. At 2 it collapses โ water droplets and wood grain vanish, and the surface turns cloth-like.
๐ Links
Model
FINAL-Bench/POCKET-Zimage-CPU
Live demo Space (runs on CPU)
FINAL-Bench/POCKET-Zimage-CPU
POCKET collection
FINAL-Bench/pocket-models-6a618ee5d23eafb7e185a5c6
SeaWolf-AIย
updated a
collection about 1 hour ago
SeaWolf-AIย
published a
Space about 2 hours ago
SeaWolf-AIย
published a
model about 2 hours ago
SeaWolf-AIย
updated 6
models about 2 hours ago
FINAL-Bench/POCKET-Image-Zimage
Text-to-Image โข 3B โข Updated โข 43 โข 42
FINAL-Bench/POCKET-KR-MLX
Text Generation โข 15B โข Updated โข 170 โข 34
FINAL-Bench/POCKET-EN-GGUF
Text Generation โข 23B โข Updated โข 313 โข 43
FINAL-Bench/POCKET-KR-GGUF
Text Generation โข 23B โข Updated โข 531 โข 36
FINAL-Bench/POCKET-26B-GGUF
Text Generation โข 25B โข Updated โข 334k โข 51
FINAL-Bench/POCKET-35B-GGUF
Text Generation โข 35B โข Updated โข 748k โข 73
SeaWolf-AIย
updated a
Space about 2 hours ago
SeaWolf-AIย
updated a
model about 2 hours ago
SeaWolf-AIย
updated a
Space about 23 hours ago
SeaWolf-AIย
updated 2
Spaces 1 day ago
SeaWolf-AIย
posted an update 2 days ago
Post
3350
The cost of a judging gate is usually quoted as a number. This puts it on a Tetris board.
Three boards get the same piece order, and on every move the same proposal and the same noise โ a paired comparison. The gate decides one thing: keep this move, or draw again. Each board gets the same 60 seconds of gate time.
The text-writing gates get through 15โ22 moves. The generation-free gate gets through 40โ50. The boards that stop simply run out of clock.
It does not win on accuracy: on the same 2,018-question LODO set, JEV scores AUC 0.7350 against ZTC-Judge-27B's 0.7289. The separation is elsewhere. Clock โ 2.1 s vs 0.0615 s per call, and on a 200-candidate agent screen one judging call measured 3.206 s generative vs 0.033 s readout, same server. Calibration โ a gate is a threshold, and at ECE 0.4985 (vs ZTC 0.0245) a threshold stops carrying information. Mechanism โ a text judge can name option 42 when there is no option 42; a scoring readout cannot. Not a lower error rate. No path.
The curve in the ZTC panel is real online fitting, scored prequentially โ predict first, learn after โ with base weights untouched. Not recursive self-improvement.
Limits, also stated on the page: Laya's AUC and latency are not our measurements and are set equal to JEV's, so calibration is the only measured axis it differs on. The page is a simulation driven by measured constants.
KO / EN / ZH.
FINAL-Bench/Tetris-JEV-LAYA-ZTC
FINAL-Bench/ZTC-Judge-27B
Three boards get the same piece order, and on every move the same proposal and the same noise โ a paired comparison. The gate decides one thing: keep this move, or draw again. Each board gets the same 60 seconds of gate time.
The text-writing gates get through 15โ22 moves. The generation-free gate gets through 40โ50. The boards that stop simply run out of clock.
It does not win on accuracy: on the same 2,018-question LODO set, JEV scores AUC 0.7350 against ZTC-Judge-27B's 0.7289. The separation is elsewhere. Clock โ 2.1 s vs 0.0615 s per call, and on a 200-candidate agent screen one judging call measured 3.206 s generative vs 0.033 s readout, same server. Calibration โ a gate is a threshold, and at ECE 0.4985 (vs ZTC 0.0245) a threshold stops carrying information. Mechanism โ a text judge can name option 42 when there is no option 42; a scoring readout cannot. Not a lower error rate. No path.
The curve in the ZTC panel is real online fitting, scored prequentially โ predict first, learn after โ with base weights untouched. Not recursive self-improvement.
Limits, also stated on the page: Laya's AUC and latency are not our measurements and are set equal to JEV's, so calibration is the only measured axis it differs on. The page is a simulation driven by measured constants.
KO / EN / ZH.
FINAL-Bench/Tetris-JEV-LAYA-ZTC
FINAL-Bench/ZTC-Judge-27B
- 5 replies
SeaWolf-AIย
published a
Space 2 days ago
SeaWolf-AIย
updated 2
models 3 days ago