How to run Groot-19B-A3B-261023 MTP on RTX 4070 SUPER + RTX 4060 Ti

Groot-19B-A3B-261023 MTP Q5_K_M achieved perfect 30/30 needle recall across the full 262144-token served context window and the fastest spec-decode on the tested RTX 4070 SUPER plus RTX 4060 Ti setup, with embedded-MTP acceptance of 62-68%, but failed the compile gate 0/3 on the v5 benchmark. The three runs produced Elixir modules with defects including an inline @moduledoc heredoc at kv_1.ex:2:17, a semicolon-joined one-liner plus a Python-style raise call at kv_2.ex:13:73, and a do/end imbalance from an unless written without do at kv_3.ex:4:15, persisting at 33:5 after mechanical heredoc repair. The draft briefing concludes Groot is worth promoting to long-context retrieval and review duty behind a compile gate and a lint instruction to the orchestrator model, but is not a drop-in subagent model for Elixir coding.

Warning: this is a temporary briefing, written by an LLM and not yet evaluated or rewritten by a human to get rid of the signature AI-slop writing style. It’s published as a draft despite its shortcomings, so that you can find out how to run it and what to expect in terms of pp and tg on this hardware, as well as what to expect in terms of coding performance on this benchmark, grosso modo . Groot-19B-A3B-261023 MTP Q5 K M demonstrated perfect 30/30 needle recall across the full served 262144-token context window, the fastest spec-decode among the models measured on this box embedded-MTP acceptance 62-68% , and proper elixir fences in all three runs, perfectly fulfilling the output contract of the v5 benchmark https://overbring.com/benchmark-v5.en . However, it still missed the compile gate 0/3 with more issues than merely a typo: the modules mix Python block syntax unless cond with no do with heredoc and one-liner issues, and run 3 keeps failing even after the heredocs are mechanically repaired. Review quality is the best measured on this hardware so far, but code output is not ready to compile. Representative output Run 3 of the benchmark delivered the cleanest, most complete module of the batch, with full @spec coverage on the public API and a timer-driven eviction loop. Trimmed excerpt of the public API: python defmodule TtlStore do @spec start link map :: pid def start link opts when is map opts do ttl = Map.get opts, :ttl, 3600 validate ttl ttl state = make state ttl {pid, ref} = Process.spawn linker fn s - loop s end, state pid end @spec start link binary , integer :: pid def start link name, ttl when is binary name , is integer ttl do start link %{name: name, ttl: ttl} end @spec put pid, binary , binary :: atom def put pid, key, value do validate key key validate value value send cmd pid, {:put, key, value} end @spec put pid, binary , binary , integer :: atom def put pid, key, value, ttl do validate key key validate value value validate ttl ttl send cmd pid, {:put, key, value, ttl} end @spec get pid, binary :: {atom , binary } def get pid, key do validate key key send cmd pid, {:get, key} end @spec delete pid, binary :: atom def delete pid, key do validate key key send cmd pid, {:delete, key} end @spec stats pid :: map def stats pid when is pid pid do send cmd pid, :stats end end The catch The delivered-pipeline gate is 0/3 as-filed, with three distinct issues, each one in its own defect class: - run 1 kv 1.ex:2:17 — inline @moduledoc heredoc triple-quote content on the opening line . - run 2 kv 2.ex:13:73 — a ;-joined one-liner def f x ; … end plus the Python raise ArgumentError, … call-shape. - run 3 kv 3.ex:4:15 → post-patch 33:5 — inline heredoc, then a do/end imbalance unless cond written without do that survives a mechanical heredoc repair. The issue runs two levels deep: cosmetic tokens and real block structure. Format discipline held across all three runs proper elixir fences , fulfilling the benchmark prompt’s output contract . However, the model’s output requires genuine editing by the orchestrator model that goes beyond small changes. Route It’s worth promoting Groot to long-context retrieval and review duty, but keep a compile gate and a “lint your output” instruction to the orchestrator in front of it. This is not a drop-in subagent model for Elixir coding that will require no orchestration, and can therefore not be used as a primary model for that purpose either.