flake8 Lint

python-lint.yml

46 workflow runs

server: GET /lora-adapters answers from the set the main loop publishes, never queued behind decoding slots
queued:
flake8 Lint#46:Pull request#98openedbyjoelteply
d765a668/lora-report-published
Nov 7, 2026Queued
server: a failed prompt fill leaves the batch and its prompt clean (Windows bad_alloc abort)
waiting for another serialized run to finish:
flake8 Lint#45:Pull request#97synchronizebyjoelteply
fable/failed-fill-leaves-batch-clean
Nov 7, 2026Pending
server: a failed prompt fill leaves the batch and its prompt clean (Windows bad_alloc abort)
cancelled:
flake8 Lint#44:Pull request#97synchronizebyjoelteply
fable/failed-fill-leaves-batch-clean
Nov 7, 20268m 40s
server: a failed prompt fill leaves the batch and its prompt clean (Windows bad_alloc abort)
cancelled:
flake8 Lint#43:Pull request#97openedbyjoelteply
fable/failed-fill-leaves-batch-clean
Nov 7, 202613m 8s
server: a saved page's adapter identity is its active adapters; one carried at 0.0 never refuses a restore
cancelled:
flake8 Lint#42:Pull request#90openedbyjoelteply
cormac/page-identity-active
Nov 3, 20261d 0h 0m 2s
server: learn a past example: /learn's messages form serves its prompt through the ordinary completion path first, then steps
cancelled:
flake8 Lint#41:Pull request#89openedbyjoelteply
cormac/learn-past-example
Nov 2, 20261d 0h 0m 8s
server: the lost-reply binding: a learn names its execution, base step and engine; a repeat is answered from its receipt, a stale one is fenced
cancelled:
flake8 Lint#40:Pull request#88synchronizebyjoelteply
cormac/learn-receipts
Nov 1, 20261d 0h 0m 2s
server: the lost-reply binding: a learn names its execution, base step and engine; a repeat is answered from its receipt, a stale one is fenced
cancelled:
flake8 Lint#39:Pull request#88openedbyjoelteply
cormac/learn-receipts
Nov 1, 202618m 58s
server: one learning ledger (draft): every learn step debits the grant that admitted it
cancelled:
flake8 Lint#38:Pull request#87openedbyjoelteply
cormac/learn-ledger
Nov 1, 20261d 0h 0m 1s
server: POST /lora-adapters/snapshot writes a mind's shadow to an immutable GGUF between steps, and /learn reports the shadow's committed step
cancelled:
flake8 Lint#37:Pull request#86openedbyjoelteply
cormac/shadow-snapshot
Nov 1, 20261d 0h 0m 2s
server: one shadow per mind on the serving context: each shadow keeps its own optimizer, /learn trains any loaded shadow, and unload releases it
cancelled:
flake8 Lint#36:Pull request#85openedbyjoelteply
cormac/per-mind-shadow
Nov 1, 20261d 0h 0m 2s
ggml: CLAMP has a backward, and a training graph the gradient cannot pass through refuses by name instead of aborting the server
cancelled:
flake8 Lint#35:Pull request#84openedbyjoelteply
cormac/clamp-backward-refuse
Nov 1, 20261d 0h 0m 7s
metal: MUL_MAT_ID_BACK kernel, so a MoE LoRA's backward runs on the GPU (stacked on #82; continuum card 14b64aa9)
cancelled:
flake8 Lint#34:Pull request#83openedbyjoelteply
0a3a5106/metal-mul-mat-id-back-kernel-so-a-moe-lo
Oct 31, 20261d 0h 0m 2s
ggml: MUL_MAT_ID has a backward pass through frozen experts, so a LoRA trains on a MoE base (GGML_OP_MUL_MAT_ID_BACK; continuum card 14b64aa9)
cancelled:
flake8 Lint#33:Pull request#82openedbyjoelteply
cormac/mul-mat-id-backward
Oct 31, 20261d 0h 0m 1s
server: create dormant dream adapters and save immutable snapshots
cancelled:
flake8 Lint#32:Pull request#81synchronizebyjoelteply
codex/dream-adapters
Oct 28, 20261d 0h 0m 2s
server: create dormant dream adapters and save immutable snapshots
cancelled:
flake8 Lint#31:Pull request#81synchronizebyjoelteply
codex/dream-adapters
Oct 28, 202650m 56s
server: create dormant dream adapters and save immutable snapshots
cancelled:
flake8 Lint#30:Pull request#81synchronizebyjoelteply
codex/dream-adapters
Oct 28, 20269m 40s
server: create dormant dream adapters and save immutable snapshots
cancelled:
flake8 Lint#29:Pull request#81synchronizebyjoelteply
codex/dream-adapters
Oct 28, 202610m 24s
server: create dormant dream adapters and save immutable snapshots
cancelled:
flake8 Lint#28:Pull request#81openedbyjoelteply
codex/dream-adapters
Oct 28, 20266m 58s
server: a slot pages to named host RAM and back; checkpoints keep a byte budget; a page carries its draft KV
cancelled:
flake8 Lint#27:Pull request#70synchronizebyjoelteply
fable/ram-pages
Oct 28, 20261d 0h 0m 3s
server: a slot pages to named host RAM and back; checkpoints keep a byte budget; a page carries its draft KV
cancelled:
flake8 Lint#26:Pull request#70synchronizebyjoelteply
fable/ram-pages
Oct 28, 20264h 17m 17s
server: a slot pages to named host RAM and back; checkpoints keep a byte budget; a page carries its draft KV
cancelled:
flake8 Lint#25:Pull request#70synchronizebyjoelteply
fable/ram-pages
Oct 28, 20262s
server: the checkpoint budget's eviction gives up what a saved page gives up first, never oldest-first
cancelled:
flake8 Lint#24:Pull request#74openedbyjoelteply
164bbff9/server-a-saved-page-keeps-checkpoints-ju
Oct 25, 20261d 0h 0m 2s
server: a slot pages to named host RAM and back; checkpoints keep a byte budget; a page carries its draft KV
cancelled:
flake8 Lint#23:Pull request#70synchronizebyjoelteply
fable/ram-pages
Oct 24, 20261d 0h 0m 3s
server: a slot pages to named host RAM and back; checkpoints keep a byte budget; a page carries its draft KV
cancelled:
flake8 Lint#22:Pull request#70synchronizebyjoelteply
fable/ram-pages
Oct 24, 202610m 40s