turboderp 8877b99855 QCache: Skip dequant when possible outside of SWA window

2026-04-13 00:52:46 +02:00

..

generator_stresstest.py

Bump to v0.0.29

2026-04-12 02:27:23 +02:00

smoke_gemma4_arch.py

feat(gemma4): add multimodal generation support

2026-04-03 12:52:51 +09:00

smoke_qwen3_5_arch.py

Qwen3.5: Smoke test

2026-03-02 15:49:29 +01:00

test_bighead_attn.py

Add bighead-attn and bighead-attn-paged kernels

2026-04-12 01:34:54 +02:00

test_cache_rotate.py

Generator: Periodically defragment paged cache

2025-05-28 22:23:08 +02:00

test_ext_norm.py

Initial commit

2025-04-06 14:42:49 +02:00

test_kv_quant.py

QCache: Skip dequant when possible outside of SWA window

2026-04-13 00:52:46 +02:00

test_qgemm.py

Initial commit

2025-04-06 14:42:49 +02:00

test_quant_fn.py

GEMM: Lock MCG multiplier to 0xCBAC1FED and MUL1 to 0x83DCD12D. Make MCG the default codebook for new models.

2025-10-12 22:09:01 +02:00

test_rope.py

New RoPE kernel with fused head norm

2025-05-25 13:33:39 +02:00

test_sampler.py

Add repetition, presence and frequency penalties

2025-04-27 01:09:33 +02:00

util.py

Initial commit

2025-04-06 14:42:49 +02:00