whn09 cedd237772 neuron: fp8-activation linears take the NKI kernel (Engram projection); drop unused fp8 hidden copies before region splits; log weights without NKI scales ... An fp8 tuple input fell back to the deep_gemm shim, which dequantized the whole fp8 weight in the graph on every call (Engram fused_kv_projection: 9 ms of an Engram layer's prefill graph, GpSimd POW for the scales). Now decoded to bf16 and run through the NKI kernel. TTFT 1.71 -> 1.66 s (bucket 512), tokens unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>		2026-10-27 11:50:34 +08:00
..
tilelang_ops	neuron: drafter forward as one region, compiled dflash metadata fill; verification + drafts regions on by default	2026-10-27 00:14:32 +08:00
__init__.py	initial commit	2026-10-20 12:09:42 +08:00
attention_backend.py	initial commit	2026-10-20 12:09:42 +08:00
layernorm.py	initial commit	2026-10-20 12:09:42 +08:00
linear.py	neuron: fp8-activation linears take the NKI kernel (Engram projection); drop unused fp8 hidden copies before region splits; log weights without NKI scales	2026-10-27 11:50:34 +08:00
mtp_info_pool.py	initial commit	2026-10-20 12:09:42 +08:00
rotary_embedding.py	initial commit	2026-10-20 12:09:42 +08:00
sampler_utils.py	initial commit	2026-10-20 12:09:42 +08:00
sampler.py	neuron: compiled decode verification behind DEEPINFER_NEURON_COMPILE_VERIFY (default off)	2026-10-26 21:33:34 +08:00
vocab_parallel_embedding.py	Neuron: eager embedding/gather without dtype casts or host scalar operands	2026-10-25 06:55:43 +08:00