Alexander Korolev
|
4e6b20c673
dflash2 lane: drafter kv_cache_dtype auto (fp8 KV invalid on SM121 outside B12X)
|
2 месяцев назад |
Alexander Korolev
|
a8ed91b8b5
dflash2 lane: pin drafter attention_backend FLASH_ATTN (auto-select poisoned by global B12X)
|
2 месяцев назад |
Alexander Korolev
|
3976ebe72f
dflash2 lane: drop drafter B12X (kv_cache_dtype constraint; main keeps B12X)
|
2 месяцев назад |
Alexander Korolev
|
bddf8d552e
add DFlash2 k=7 drafter lane (r18 img, A/B vs MTP)
|
2 месяцев назад |