Merge pull request #28 from OpenManus/murphy/rollout-traj-gen

Initialize rollout trajectory Generation Env

Author: realtmxi

.gitignore

@@ -0,0 +1,8 @@
+__pycache__
+.DS_Store
+.cache
+.dockerfile
+.dockerfile-cache
+.idea
+.vscode
+.env

traj_generation/.env_example

@@ -0,0 +1,2 @@
+OAI_ENDPOINT=
+OAI_KEY=

traj_generation/configs/agents/api_agents.yaml

@@ -0,0 +1,15 @@
+gpt-4o:
+    import: "./openai-chat.yaml"
+    parameters:
+        name: "gpt-4o"
+        body:
+            model: "gpt-4o"
+            max_tokens: 512
+
+qwq-32b:
+    import: "./qwq.yaml"
+    parameters:
+        name: "qwq-32b"
+        body:
+            model: "qwq-32b"
+            max_tokens: 8192

traj_generation/configs/agents/openai-chat.yaml

@@ -0,0 +1,13 @@
+module: src.client.agents.HTTPAgent
+parameters:
+  url: https://api.openai.com/v1/chat/completions 
+  headers:
+    Content-Type: application/json
+    Authorization: Bearer <% PUT-YOUR-OPENAI-KEY-HERE %>
+  body:
+    temperature: 0
+  prompter:
+    name: role_content_dict
+    args:
+      agent_role: assistant
+  return_format: "{response[choices][0][message][content]}"

traj_generation/configs/agents/qwq.yaml

@@ -0,0 +1,18 @@
+import: definition.yaml
+
+concurrency:
+  task:
+    # dbbench-std: 5
+    os-std: 5
+
+  agent:
+    gpt-4o: 5
+
+assignments: # List[Assignment] | Assignment
+  - agent: # "task": List[str] | str ,  "agent": List[str] | str
+      - gpt-4o
+    task:
+      # - dbbench-std
+      - os-std
+
+output: "outputs/{TIMESTAMP}"

traj_generation/configs/assignments/default.yaml

@@ -0,0 +1,11 @@
+definition:
+  task:
+    overwrite:
+      module: src.client.TaskClient
+      parameters:
+        controller_address: "http://localhost:5000/api"
+    import: ../tasks/task_assembly.yaml
+  agent:
+    import:
+      - ../agents/api_agents.yaml
+      - ../agents/fs_agent.yaml

traj_generation/configs/assignments/definition.yaml

@@ -0,0 +1,20 @@
+import: definition.yaml
+
+concurrency:
+  task:
+    # dbbench-std: 5
+    os-std: 5
+
+  agent:
+    #gpt-4o: 5
+    qwq-32b: 5
+
+assignments: # List[Assignment] | Assignment
+  - agent: # "task": List[str] | str ,  "agent": List[str] | str
+      #- gpt-4o
+      - qwq-32b
+    task:
+      # - dbbench-std
+      - os-std
+
+output: "outputs/rollout_{TIMESTAMP}"

traj_generation/configs/assignments/rollout.yaml

@@ -0,0 +1,7 @@
+definition:
+  import: tasks/task_assembly.yaml
+
+start:
+    # dbbench-std: 5
+    os-std: 5
+    # webshop-std: 5

traj_generation/configs/start_task.yaml

@@ -0,0 +1,22 @@
+default:
+  module: src.server.tasks.alfworld.ALFWorld
+  docker:
+    image: longinyu/agentbench-alfworld
+    command: umask 0; [ -f /root/.setup.sh ] && bash /root/.setup.sh;
+  parameters:
+    name: alfworld-std
+    data_path: "/AgentBench/data/alfworld"
+    config_path: "src/server/tasks/alfworld/configs/base_config.yaml"
+    prompts_path: "src/server/tasks/alfworld/prompts/alfworld_multiturn_plan_first.json"
+    split: "standard"
+    max_step: 35
+
+alfworld-dev:
+  parameters:
+    name: alfworld-dev
+    split: "dev"
+
+alfworld-std:
+  parameters:
+    name: alfworld-std
+    split: "standard"