chore(llm): univyed token size for agent usage · 1e3ad169c0 - LLM-Server - Gitea, hosted by Arthur

haylan/LLM-Server

0 Fork 0

haylan committed 2026-10-03 14:06:56 +02:00

1 parent b74c0e0ccc

commit 1e3ad169c0

1 file changed

+1

									
										docker-compose.yml
									
		+1
		
												View File
												
					@@ -23,6 +23,7 @@ services:

					      --n-gpu-layers ${LLAMA_GPU_LAYERS:-999}

					      --n-gpu-layers ${LLAMA_GPU_LAYERS:-999}

					      --ctx-size ${LLAMA_CTX_SIZE:-262144}

					      --ctx-size ${LLAMA_CTX_SIZE:-262144}

					      --parallel ${LLAMA_PARALLEL:-2}

					      --parallel ${LLAMA_PARALLEL:-2}

					      --kv-unified

					      --flash-attn on

					      --flash-attn on

					      --cache-type-k q8_0

					      --cache-type-k q8_0

					      --cache-type-v q8_0

					      --cache-type-v q8_0