mirrors/hermes-agent
1
0
Fork 0
mirror of https://github.com/NousResearch/hermes-agent.git synced 2026-04-25 00:51:20 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

add a local vllm instance

Browse source
This commit is contained in:
dmahan93 2026-03-09 23:02:13 -05:00 • committed by teknium1
parent 13f5459670
commit 366de72a38
1 changed files with 5 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

9
environments/benchmarks/tblite/local_vllm.yaml
View file

@ -4,7 +4,7 @@
# #
# Start the vLLM server from the atropos directory: # Start the vLLM server from the atropos directory:
# python -m example_trainer.vllm_api_server \ # python -m example_trainer.vllm_api_server \
# --model Qwen/Qwen3-4B-Thinking-2507 \ # --model Qwen/Qwen3-4B-Instruct-2507 \
# --port 9001 \ # --port 9001 \
# --gpu-memory-utilization 0.8 \ # --gpu-memory-utilization 0.8 \
# --max-model-len=32000 # --max-model-len=32000
@ -26,14 +26,15 @@ env:
task_timeout: 1200 task_timeout: 1200
eval_concurrency: 8 eval_concurrency: 8
tool_call_parser: "hermes" tool_call_parser: "hermes"
tokenizer_name: "Qwen/Qwen3-4B-Thinking-2507" system_prompt: "You are an expert terminal agent. You MUST use the provided tools to complete tasks. Use the terminal tool to run shell commands, read_file to read files, write_file to write files, search_files to search, and patch to edit files. Do NOT write out solutions as text - execute them using the tools. Always start by exploring the environment with terminal commands."
tokenizer_name: "Qwen/Qwen3-4B-Instruct-2507"
use_wandb: false use_wandb: false
wandb_name: "tblite-qwen3-4b-thinking" wandb_name: "tblite-qwen3-4b-instruct"
ensure_scores_are_not_same: false ensure_scores_are_not_same: false
data_dir_to_save_evals: "environments/benchmarks/evals/tblite-qwen3-4b-local" data_dir_to_save_evals: "environments/benchmarks/evals/tblite-qwen3-4b-local"
openai: openai:
base_url: "http://localhost:9001" base_url: "http://localhost:9001"
model_name: "Qwen/Qwen3-4B-Thinking-2507" model_name: "Qwen/Qwen3-4B-Instruct-2507"
server_type: "vllm" server_type: "vllm"
health_check: false health_check: false