NousResearch/atropos
1
0
Fork 0
mirror of https://github.com/NousResearch/atropos.git synced 2026-04-24 17:04:55 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

major refactor 2

Browse source
This commit is contained in:
Jai Suphavadeeprasit 2026-01-22 13:21:50 -05:00
parent 2e6c3b3e12
commit 1b1d8a5909
2 changed files with 45 additions and 27 deletions
Download patch file Download diff file Expand all files Collapse all files

30
example_trainer/scripts/run_comparison.sh
View file

@ -189,10 +189,12 @@ wait_for_vllm 9001 "legacy (internal)" || { echo "Legacy vLLM failed to start";
# NOW start environment server (after vLLM is ready)
echo " Starting environment server..."
python environments/gsm8k_server.py serve \
--slurm.num_gpus 0 \
--env.tokenizer_name $MODEL \
--openai.base_url http://localhost:9001/v1 \
python -u environments/gsm8k_server.py serve \
--env.tokenizer_name "$MODEL" \
--env.use_wandb=False \
--openai.model_name "$MODEL" \
--openai.base_url "http://localhost:9001/v1" \
--openai.server_type vllm \
--server.port 8001 \
> $LOGDIR/env_legacy.log 2>&1 &
LEGACY_ENV_PID=$!
@ -228,10 +230,12 @@ wait_for_vllm 9002 "shared" || { echo "Failed to start shared vLLM"; exit 1; }
# Start environment server for Shared
echo " Starting environment server..."
python environments/gsm8k_server.py serve \
--slurm.num_gpus 0 \
--env.tokenizer_name $MODEL \
--openai.base_url http://localhost:9002/v1 \
python -u environments/gsm8k_server.py serve \
--env.tokenizer_name "$MODEL" \
--env.use_wandb=False \
--openai.model_name "$MODEL" \
--openai.base_url "http://localhost:9002/v1" \
--openai.server_type vllm \
--server.port 8002 \
> $LOGDIR/env_shared.log 2>&1 &
SHARED_ENV_PID=$!
@ -286,10 +290,12 @@ wait_for_vllm 9003 "lora" || { echo "Failed to start lora vLLM"; exit 1; }
# Start environment server for LoRA
echo " Starting environment server..."
python environments/gsm8k_server.py serve \
--slurm.num_gpus 0 \
--env.tokenizer_name $MODEL \
--openai.base_url http://localhost:9003/v1 \
python -u environments/gsm8k_server.py serve \
--env.tokenizer_name "$MODEL" \
--env.use_wandb=False \
--openai.model_name "$MODEL" \
--openai.base_url "http://localhost:9003/v1" \
--openai.server_type vllm \
--server.port 8003 \
> $LOGDIR/env_lora.log 2>&1 &
LORA_ENV_PID=$!