24 days ago
-
emu 3.2.2: every writing system draws, illustrated user guide, UI fixes; signed Windows build
-
emu 3.2.1: typeset maths, tooltips, download fixes; signed Windows build
25 days ago
-
emu 3.2: Windows and Linux editions, Qwen3.5 9B/4B packs, docs for all three platforms
-
emu 3.1 for macOS: Qwen3.8 27B pack and the macOS app
-
emu 2.0.0: emu-2.0.0-linux-cpu.run
-
emu 2.0.0: emu-2.0.0-linux-cuda.run
-
emu 2.0.0: emu-2.0.0-windows-cpu-setup.exe
-
emu 2.0.0: emu-2.0.0-windows-cuda-setup.exe
-
Upload qwen3vl-2b-cpu.nbq with huggingface_hub
-
Upload qwen3vl-8b-cuda.nbq with huggingface_hub
-
Upload Qwen3.6-27B-int4-AutoRound.nbq
-
Add full self-contained unlimited-ocr.nbq (encoder+decoder+tokenizer)
-
Upload decoder.nbq
-
emu #34: multi-arch fatbin kernels (native SASS sm_75-120 + PTX floor) — GPU works on any CUDA-13 card, no JIT/toolkit; debug build
-
emu #34: NVRTC fallback self-contained (in-memory libc headers) so GPU JIT works on driver-only machines — debug build
-
emu #34 fix: NVRTC dl.load() in initWithHeaders (CUDA JIT fallback no longer corrupts heap) — debug build
-
emu 1.0.0 CUDA debug build (GitHub #34): crash log + fault handlers for field diagnosis
-
models: add ppocr-v6.nbq
-
models: add vieneu-tts.nbq
-
models: add pocket-tts-en.nbq
-
models: add gwen-tts-vn-0.6b.nbq
-
models: add qwen3-tts-0.6b.nbq
-
models: add qwen3-asr-0.6b.nbq
-
models: add qwen3.5-9b-q4_k_m.nbq
-
models: add qwen3.5-4b-q4_k_m.nbq
-
models: add gemma-4-e4b-it-q4_k_m.nbq
-
models: add gemma-4-12b-it-qat-q4_0.nbq
-
models: add gemma-4-12b-it-UD-Q2_K_XL.nbq
-
Add emu EULA in PDF format
-
Add Product Information (PDF)
-
Add User Guide (PDF)
-
Add Download & Installation Guide (PDF)
-
Add Windows CUDA installer v1.0.0 (CloudKites-signed)
-
Add Windows CPU installer v1.0.0 (CloudKites-signed)
-
linux apps
-
initial commit
