FINAL-Bench/POCKET-Qwen3.8-Flash-Next-GGUF
Text Generation โข 177B โข Updated โข 36
openfree
openfree
AI & ML interests
Contact: arxivgpt@gmail.com
Recent Activity
liked a model 21 minutes ago
FINAL-Bench/POCKET-Qwen3.8-Flash-Next-GGUF liked a model about 3 hours ago
FINAL-Bench/POCKET-Zimage-CPU liked a Space about 3 hours ago
FINAL-Bench/POCKET-Zimage-CPUOrganizations
liked a
model 21 minutes ago
liked a
model about 3 hours ago
liked a
Space about 3 hours ago
reacted to SeaWolf-AI's post with โค๏ธ๐๐๐ฅ about 3 hours ago
Post
205
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
updated a
bucket about 13 hours ago
updated a
Space about 21 hours ago
liked a
Space 2 days ago
reacted to SeaWolf-AI's post with ๐๐ค 2 days ago
Post
3354
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
liked 2
models 5 days ago
reacted to SeaWolf-AI's post with ๐ฅ 5 days ago
Post
3145
Ask a language model how confident it is and you get an AUC of 0.5000. Exactly a coin flip. We measured it across 2,018 items.
FINAL-Bench/gate-tetris
https://huggingface.co/blog/FINAL-Bench/ztc
Collection: https://huggingface.co/collections/FINAL-Bench/ztc-models-jev-ecosystems
Zero-Token Confidence (ZTC) reads it. One forward pass over the model's hidden state returns a calibrated probability that the answer is correct. Zero generated tokens.
It sits at the top of the shared board. Same 2,018 items, same harness for every entry: ZTC on Darwin-397B 0.7394, JEV 0.7335, ZTC-Judge-27B 0.7255, a surface baseline that reads only answer length and formatting 0.7036, Lynx 8B 0.5157, the model's own self-reported confidence 0.5000, HHEM 0.4852. First and third place both emit nothing at all.
The number worth staring at is 0.7036. That is a baseline reading no content whatsoever, just how long the answer is and how it is formatted. Any verifier scoring below it is not reading content either.
On speed, one gate call costs 0.0615 seconds, measured on four B200s across 2,000 items. Generating a single candidate answer takes 1.631 seconds, so the gate is 26 times cheaper than the work it guards. A verifier that generates competes with your agent for the same budget. A verifier that only reads can be attached to every action instead of a sampled few.
We built it so you can watch it decide. Three lanes receive the same stream of proposed actions and the same time budget. One has no gate and must execute everything. One uses a text-reading verifier. One uses ZTC. Right action plus one, wrong action minus one, hold zero. Over 400 matches: no gate minus 3.9, text verifier plus 13.0, ZTC plus 29.1, with ZTC taking 98 percent of matches. Gating lifts executed accuracy from 49 percent to 65 percent.
FINAL-Bench/gate-tetris
https://huggingface.co/blog/FINAL-Bench/ztc
Collection: https://huggingface.co/collections/FINAL-Bench/ztc-models-jev-ecosystems
Zero-Token Confidence (ZTC) reads it. One forward pass over the model's hidden state returns a calibrated probability that the answer is correct. Zero generated tokens.
It sits at the top of the shared board. Same 2,018 items, same harness for every entry: ZTC on Darwin-397B 0.7394, JEV 0.7335, ZTC-Judge-27B 0.7255, a surface baseline that reads only answer length and formatting 0.7036, Lynx 8B 0.5157, the model's own self-reported confidence 0.5000, HHEM 0.4852. First and third place both emit nothing at all.
The number worth staring at is 0.7036. That is a baseline reading no content whatsoever, just how long the answer is and how it is formatted. Any verifier scoring below it is not reading content either.
On speed, one gate call costs 0.0615 seconds, measured on four B200s across 2,000 items. Generating a single candidate answer takes 1.631 seconds, so the gate is 26 times cheaper than the work it guards. A verifier that generates competes with your agent for the same budget. A verifier that only reads can be attached to every action instead of a sampled few.
We built it so you can watch it decide. Three lanes receive the same stream of proposed actions and the same time budget. One has no gate and must execute everything. One uses a text-reading verifier. One uses ZTC. Right action plus one, wrong action minus one, hold zero. Over 400 matches: no gate minus 3.9, text verifier plus 13.0, ZTC plus 29.1, with ZTC taking 98 percent of matches. Gating lifts executed accuracy from 49 percent to 65 percent.
liked a
Space 5 days ago
liked a
model 5 days ago
upvoted a collection 5 days ago
liked 2
Spaces 5 days ago