Set FASTGREEDY=0 default in env; add deployment notes

This commit is contained in:
Frank
2026-08-22 12:54:28 -07:00
parent e0329caef8
commit f347d5336f
+9 -4
View File
@@ -82,7 +82,12 @@ SPEC_DRAFT_P_SPLIT=0.10
# --reasoning-budget -1 (unrestricted) and chat_template_kwargs
# preserve_thinking=true (preserve thinking tokens across turns).
#
# For fast-greedy bench: comment out the --reasoning on line in serve script
# and set FASTGREEDY=1 in env (see serve script FASTGREEDY section).
# FASTGREEDY=1 flips to: reasoning off, argmax fastpath=1, sampling defaults.
# FASTGREEDY=0 (default) uses the production values above.
# FASTGREEDY mode: set to 1 for fast-greedy benchmark (card's default profile)
# FASTGREEDY=0 (default): production chat config (reasoning on, argmax off)
# FASTGREEDY=1: reasoning off, LLAMA_TARGET_GREEDY_ARGMAX_FASTPATH=1, deterministic
# sampler. Mirrors the card's 'fast greedy (default)' benchmark config.
FASTGREEDY=0
# For fast-greedy bench: set FASTGREEDY=1 above. This flips to: reasoning off,
# argmax fastpath=1, sampling defaults. Deploy with
# FASTGREEDY=1 ./deploy.sh dankclaw to test the fast-greedy profile.