Issues

npu-pack: layer norms (and the final norm) are located only for hidden 896

1bit-MONSTER/engine#734 · opened

on Nov 1, 2026

1

hrx: gfxhub permission page faults (UTCL2 client CPF) on the rebuilt box with the TheRock nightly libhsa — the no-SVM env trio is the only fault-free arm

1bit-MONSTER/engine#715 · opened

on Nov 1, 2026

1

build: the engine should fetch and pin its own HRX toolchain instead of requiring /opt/rocm-therock on the box

1bit-MONSTER/engine#713 · opened

on Oct 31, 2026

1

1bit OS v3: the installed appliance must be a full operating system, not the live RAM image

1bit-MONSTER/engine#712 · opened

on Oct 31, 2026

3

os: booted from the stick with an installed NVMe present, init mounts the NVMe's data partition and runs as the installed system

1bit-MONSTER/engine#711 · opened

on Oct 31, 2026

os: 1bit-install must show the whole install on screen - progress for every step, nothing silent

1bit-MONSTER/engine#710 · opened

on Oct 31, 2026

os: 1bit-install registers no firmware boot entry and the live kernel has no efivarfs - the installed disk only boots by firmware fallback

1bit-MONSTER/engine#709 · opened

on Oct 31, 2026

1

os: boot log shows 'Loading of unsigned module is rejected' and snd_usbmidi_lib 'Unknown symbol' errors

1bit-MONSTER/engine#708 · opened

on Oct 31, 2026

os: the appliance menu is unreadably small on the board's display and its highlight bar covers only the first 12 columns

1bit-MONSTER/engine#707 · opened

on Oct 31, 2026

os: the live stick's /data is 2 GB (351 MB free after boot) - no catalogue model can be pulled on the live appliance

1bit-MONSTER/engine#706 · opened

on Oct 31, 2026

1

os: two menus at boot (12 s chooser, then the appliance screen) and no visible network wait - one flow

1bit-MONSTER/engine#705 · opened

on Oct 31, 2026

2

os: boot chooser banner hard-codes 'v2026.43' while the shipped Lemonade is 2026.46.0

1bit-MONSTER/engine#704 · opened

on Oct 31, 2026

os: the live boot is not clean - kernel log spews on tty0 (console=tty0 ignore_loglevel loglevel=7)

1bit-MONSTER/engine#703 · opened

on Oct 31, 2026

tracking: zero-copy on the serving paths (hybrid KV handoff, NPU activations in BOs, ONNX EP IO binding, audit)

1bit-MONSTER/engine#698 · opened

on Oct 31, 2026

3

hrx: hrx_synchronous_h2d/d2h are not ordered against the compute stream

1bit-MONSTER/engine#696 · opened

on Oct 31, 2026

hw-gate: every GPU step takes flock ~/npu.lock, so the gate cannot run while an NPU serve is up (c16 gate cut 3x by its own 20-min hold)

1bit-MONSTER/engine#682 · opened

on Oct 31, 2026

hrx: a separate MTP draft head (-md) is rejected for qwen4exp on our route, while a cut-vocab head gives +55% decode on the fork

1bit-MONSTER/engine#681 · opened

on Oct 31, 2026

hrx: gfxhub page fault + MODE2 GPU reset when GTT nears the 120 GiB ceiling (Flash-Next UD-IQ3_XXS, 18.8k ctx, 106.5 GiB GTT)

1bit-MONSTER/engine#680 · opened

on Oct 31, 2026

hrx: long-context decode 9.70 tok/s at 18.8k ctx (prefill 74-93) vs 31.15 / 459 on a gfx1151 fork — same GGUF, same hour, short-prompt parity 28.6 vs 28.9

1bit-MONSTER/engine#679 · opened

on Oct 31, 2026

2

hrx: any graph whose weight ops all run on the CPU computes garbage (1 s reproducer)

1bit-MONSTER/engine#674 · opened

on Oct 31, 2026

1

npu: load Ryzen AI OGA NPU model packs on Linux through the engine (pack import -> Q4NX -> our kernels)

1bit-MONSTER/engine#671 · opened

on Oct 30, 2026

2

HRX route pins ~107 GiB outside process RSS on a 217 GiB mmap (DeepSeek-V4.1 UD-IQ1_S), independent of -ngl/ubatch/prefetch

1bit-MONSTER/engine#669 · opened

on Oct 30, 2026

5

test-backend-ops (fork pin 40c4ccb): a run that executes zero cases exits 0 and prints OK

1bit-MONSTER/engine#657 · opened

on Oct 30, 2026

2

npu: /v1/systemone decision endpoint (Decision 3.0 / d3 packs) — the gap that keeps d3 out of the catalog

1bit-MONSTER/engine#653 · opened

on Oct 30, 2026

1

npu: constrained decoding past GBNF — tool_choice, parallel_tool_calls, json_schema, and one answer parser for every family (FLM v1.0.10)

1bit-MONSTER/engine#652 · opened

on Oct 30, 2026

1

Navigation Menu

Repository navigation

Issues

npu-pack: layer norms (and the final norm) are located only for hidden 896

hrx: gfxhub permission page faults (UTCL2 client CPF) on the rebuilt box with the TheRock nightly libhsa — the no-SVM env trio is the only fault-free arm

build: the engine should fetch and pin its own HRX toolchain instead of requiring /opt/rocm-therock on the box

1bit OS v3: the installed appliance must be a full operating system, not the live RAM image

os: booted from the stick with an installed NVMe present, init mounts the NVMe's data partition and runs as the installed system

os: 1bit-install must show the whole install on screen - progress for every step, nothing silent

os: 1bit-install registers no firmware boot entry and the live kernel has no efivarfs - the installed disk only boots by firmware fallback

os: boot log shows 'Loading of unsigned module is rejected' and snd_usbmidi_lib 'Unknown symbol' errors

os: the appliance menu is unreadably small on the board's display and its highlight bar covers only the first 12 columns

os: the live stick's /data is 2 GB (351 MB free after boot) - no catalogue model can be pulled on the live appliance

os: two menus at boot (12 s chooser, then the appliance screen) and no visible network wait - one flow

os: boot chooser banner hard-codes 'v2026.43' while the shipped Lemonade is 2026.46.0

os: the live boot is not clean - kernel log spews on tty0 (console=tty0 ignore_loglevel loglevel=7)

tracking: zero-copy on the serving paths (hybrid KV handoff, NPU activations in BOs, ONNX EP IO binding, audit)

hrx: hrx_synchronous_h2d/d2h are not ordered against the compute stream

hw-gate: every GPU step takes flock ~/npu.lock, so the gate cannot run while an NPU serve is up (c16 gate cut 3x by its own 20-min hold)

hrx: a separate MTP draft head (-md) is rejected for qwen4exp on our route, while a cut-vocab head gives +55% decode on the fork

hrx: gfxhub page fault + MODE2 GPU reset when GTT nears the 120 GiB ceiling (Flash-Next UD-IQ3_XXS, 18.8k ctx, 106.5 GiB GTT)

hrx: long-context decode 9.70 tok/s at 18.8k ctx (prefill 74-93) vs 31.15 / 459 on a gfx1151 fork — same GGUF, same hour, short-prompt parity 28.6 vs 28.9

hrx: any graph whose weight ops all run on the CPU computes garbage (1 s reproducer)

npu: load Ryzen AI OGA NPU model packs on Linux through the engine (pack import -> Q4NX -> our kernels)

HRX route pins ~107 GiB outside process RSS on a 217 GiB mmap (DeepSeek-V4.1 UD-IQ1_S), independent of -ngl/ubatch/prefetch

test-backend-ops (fork pin 40c4ccb): a run that executes zero cases exits 0 and prints OK

npu: /v1/systemone decision endpoint (Decision 3.0 / d3 packs) — the gap that keeps d3 out of the catalog

npu: constrained decoding past GBNF — tool_choice, parallel_tool_calls, json_schema, and one answer parser for every family (FLM v1.0.10)