Hugging Face
Models
Datasets
Spaces
Buckets
Docs
Enterprise
Pricing
Website
Tasks
HuggingChat
Collections
Languages
Organizations
Community
Blog
Posts
Daily Papers
Hardware
Learn
Discord
Forum
GitHub
Solutions
Team & Enterprise
Hugging Face PRO
Enterprise Support
Inference Providers
Inference Endpoints
Storage Buckets
Log In
Sign Up
HelloSun
/
qwen35-4b-ssd-streaming
Like
0
GGUF
llama.cpp
qwen3
cpu-inference
ssd-streaming
low-memory
License:
apache-2.0
Model card
Files
Files and versions
xet
Community
Copy to bucket
main
qwen35-4b-ssd-streaming
/
validate
164 kB
Ctrl+K
Ctrl+K
2 contributors
History:
7 commits
Auto Upload Agent
上傳解除限制實驗的完整證據: validate/bench-logs-nolimit/ 六個 case 的 memwatch+chat 原始記錄 (swap 全 0, over_limit_samples 全 0) + bench-nolimit.jsonl; 另含 cmake 編譯記錄
364a89d
about 1 month ago
bench-logs-nolimit
上傳解除限制實驗的完整證據: validate/bench-logs-nolimit/ 六個 case 的 memwatch+chat 原始記錄 (swap 全 0, over_limit_samples 全 0) + bench-nolimit.jsonl; 另含 cmake 編譯記錄
about 1 month ago
bench-logs
實測確認 22 倍差距: pagecache-20ms 0.384GiB/2.67tok/s vs ssd-stream-20ms 0.226GiB/0.12tok/s, 兩者都合規; STATUS 記錄矩陣與三個量測 bug
about 1 month ago
bench-nolimit.jsonl
Safe
1.29 kB
解除記憶體限制的量測: 權重常駐 RAM 只快 6% (2.82 vs 2.67 tok/s) 但記憶體 7.2x (2.77 vs 0.384 GiB); 本機僅 2 vCPU, 8 threads 反而慢 9x
about 1 month ago
bench.jsonl
Safe
429 Bytes
實測確認 22 倍差距: pagecache-20ms 0.384GiB/2.67tok/s vs ssd-stream-20ms 0.226GiB/0.12tok/s, 兩者都合規; STATUS 記錄矩陣與三個量測 bug
about 1 month ago
chat-ctx1024.json
Safe
1 kB
驗證重跑: 全新編譯+模型在本機重現 v1 — ctx1024 峰值總 RSS 0.23GiB, swap 0, all_passed=true (patches 0001+0002 套用乾淨)
about 1 month ago
memwatch-ctx1024.json
Safe
40.4 kB
bench.sh: 修掉 bench.jsonl 被 mkdir 成目錄導致結果靜默遺失; 改 touch
about 1 month ago
validation.json
Safe
681 Bytes
驗證重跑: 全新編譯+模型在本機重現 v1 — ctx1024 峰值總 RSS 0.23GiB, swap 0, all_passed=true (patches 0001+0002 套用乾淨)
about 1 month ago