config/eval.yaml

defaults:
  - envs
enable_response_mask: True
system:
  CUDA_VISIBLE_DEVICES: "0,1,2,3,4,5,6,7"

seed:
  train: 10000
  val: 123

model_path:  /mnt/general/wanghy/RAGEN_v2/saves_hf/qwen2.5_3B_it_rubikscube1_withthink_fulltraj_sa_rl/global_step_50/qwen2.5_3b_actor_hf
# /mnt/general/share/model/tyzhu/SPA-frozenlake-qwen2.5-1.5b-instruct

lora:
  rank: 0
  alpha: 64
  target_modules: all-linear

actor_rollout_ref:
  model:
    path: ${model_path}
    lora_rank: ${lora.rank}
    lora_alpha: ${lora.alpha}
    target_modules: ${lora.target_modules}
  rollout:
    name: vllm
    log_prob_micro_batch_size_per_gpu: ${micro_batch_size_per_gpu}
    tensor_model_parallel_size: 4
    dtype: bfloat16
    max_model_len: 7200
    prompt_length: 1
    response_length: 400
    gpu_memory_utilization: 0.7
    max_num_batched_tokens: 8192
    enforce_eager: True
    free_cache_engine: True
    enable_chunked_prefill: False
    disable_log_stats: False
    do_sample: True
    val_kwargs:
      do_sample: True
      temperature: 0.5
      top_p: 1.0
      top_k: -1
      logprobs: 20 # return top20 logprobs

agent_proxy:
  max_context_window: -1
  max_turn: 25
  action_sep: "||"
  max_actions_per_turn: 1
  use_turn_scores: False
  enable_think: True
  reward_normalization:
    grouping: "state"
    method: "identity"

es_manager:
  format_penalty: -0.1
  train:
    env_groups: 8
    group_size: 16
    env_configs:
      tags: ["rubikscube"]
      n_groups: [8]
  val:
    env_groups: 32
    group_size: 1
    env_configs:
      tags: ["rubikscube"]
      n_groups: [32]

ctx_manager:
  generation:
    gen_config:
      response_length: ${actor_rollout_ref.rollout.response_length}
      temperature: ${actor_rollout_ref.rollout.temperature}
      top_p: ${actor_rollout_ref.rollout.top_p}
      top_k: ${actor_rollout_ref.rollout.top_k}
      kwargs: null

output:
  dir: results/eval
  filename: val_rollouts.pkl
  append_timestamp: true
  keep_batch_keys: null
  keep_non_tensor_keys: null
  keep_meta_info: true