B2J/latent-state-tracking-minicpm5
Preview β’ Updated β’ 117
VID-IQ
B2J
AI & ML interests
None yet
Recent Activity
updated a dataset about 16 hours ago
B2J/latent-state-tracking-minicpm5 liked a Space about 18 hours ago
FINAL-Bench/Tetris-JEV-LAYA-ZTC liked a model 2 days ago
FINAL-Bench/POCKET-26B-GGUFOrganizations
updated a
dataset about 16 hours ago
liked a
Space about 18 hours ago
liked a
model 2 days ago
published a
dataset 7 days ago
liked a
Space 8 days ago
upvoted a collection 9 days ago
liked a
dataset about 1 month ago
liked 2
Spaces about 2 months ago
liked a
Space 2 months ago
reacted to SeaWolf-AI's post with π₯πβ€οΈ 2 months ago
Post
4112
π± POCKET β a 35-billion-parameter model that runs on your iPhone, and on your PC with no GPU
We're releasing POCKET, VIDRAFT's flagship Darwin-36B-Opus compressed for on-device use. No fork, no CUDA, no cloud β it runs on stock llama.cpp. It's a sparse Mixture-of-Experts model (256 experts, only 8 active per token), so the file can be large while the work per token stays small. That's what lets a 35B model run on a phone, and generate fast on a CPU with no graphics card.
Measured (POCKET-35B IQ1_M vs Bonsai-27B Q1_0):
β’ CPU generate (Xeon, 16 threads): 27.0 vs 10.1 tok/s β 2.69Γ faster
β’ GPU generate (H100): 197 vs 89 tok/s β 2.22Γ faster
β’ GPU prompt processing (H100): 753 vs 1816 β 0.41Γ (Bonsai wins this one β MoE prefill wakes every expert, so sparsity stops helping there. We say so.)
β’ Quality (HellaSwag, 400 q): 61.0% vs 60.0% β a tie (confidence intervals overlap)
On a real consumer laptop β MacBook M3 Pro (18 GB) β POCKET wins every axis, prompt processing included:
β’ Metal generate: 25.4 vs 12.8 β 1.99Γ
β’ CPU generate: 13.8 vs 4.4 β 3.13Γ
β’ Metal prompt: 240.7 vs 73.4 β 3.28Γ
One more quiet fact: the same-size, quality-oriented rival Ternary-Bonsai-27B (7.2 GB) fails to load in upstream llama.cpp at all β it needs the PrismML fork. POCKET runs on the tools you already have: LM Studio, Ollama, PocketPal, MLX.
π Full story (tech, measurements, recipes): https://huggingface.co/blog/FINAL-Bench/pocket
Models:
π¦ POCKET-35B-GGUF (PC / server, no GPU): FINAL-Bench/POCKET-35B-GGUF
π°π· POCKET-KR-GGUF (Android): FINAL-Bench/POCKET-KR-GGUF
π POCKET-KR-MLX (iPhone / Mac): FINAL-Bench/POCKET-KR-MLX
π POCKET-EN-GGUF (English phone / PC): FINAL-Bench/POCKET-EN-GGUF
π₯οΈ Live demo (answering on a CPU, no GPU): FINAL-Bench/POCKET-35B-CPU
π Collection: FINAL-Bench/pocket-models-6a618ee5d23eafb7e185a5c6
We're releasing POCKET, VIDRAFT's flagship Darwin-36B-Opus compressed for on-device use. No fork, no CUDA, no cloud β it runs on stock llama.cpp. It's a sparse Mixture-of-Experts model (256 experts, only 8 active per token), so the file can be large while the work per token stays small. That's what lets a 35B model run on a phone, and generate fast on a CPU with no graphics card.
Measured (POCKET-35B IQ1_M vs Bonsai-27B Q1_0):
β’ CPU generate (Xeon, 16 threads): 27.0 vs 10.1 tok/s β 2.69Γ faster
β’ GPU generate (H100): 197 vs 89 tok/s β 2.22Γ faster
β’ GPU prompt processing (H100): 753 vs 1816 β 0.41Γ (Bonsai wins this one β MoE prefill wakes every expert, so sparsity stops helping there. We say so.)
β’ Quality (HellaSwag, 400 q): 61.0% vs 60.0% β a tie (confidence intervals overlap)
On a real consumer laptop β MacBook M3 Pro (18 GB) β POCKET wins every axis, prompt processing included:
β’ Metal generate: 25.4 vs 12.8 β 1.99Γ
β’ CPU generate: 13.8 vs 4.4 β 3.13Γ
β’ Metal prompt: 240.7 vs 73.4 β 3.28Γ
One more quiet fact: the same-size, quality-oriented rival Ternary-Bonsai-27B (7.2 GB) fails to load in upstream llama.cpp at all β it needs the PrismML fork. POCKET runs on the tools you already have: LM Studio, Ollama, PocketPal, MLX.
π Full story (tech, measurements, recipes): https://huggingface.co/blog/FINAL-Bench/pocket
Models:
π¦ POCKET-35B-GGUF (PC / server, no GPU): FINAL-Bench/POCKET-35B-GGUF
π°π· POCKET-KR-GGUF (Android): FINAL-Bench/POCKET-KR-GGUF
π POCKET-KR-MLX (iPhone / Mac): FINAL-Bench/POCKET-KR-MLX
π POCKET-EN-GGUF (English phone / PC): FINAL-Bench/POCKET-EN-GGUF
π₯οΈ Live demo (answering on a CPU, no GPU): FINAL-Bench/POCKET-35B-CPU
π Collection: FINAL-Bench/pocket-models-6a618ee5d23eafb7e185a5c6
- 5 replies
liked 2
models 2 months ago
liked a
Space 2 months ago
liked 3
models 3 months ago
upvoted an article 3 months ago
Article
Chitos: From Detection to Proof β An Autonomous Security AI That Actually Exploits
FINAL-Bench
β’ β’ 19