Merge pull request #54 from web-arena-x/new_eval

Massive PR on many parts

.github/workflows/tests.yml

@@ -34,7 +34,7 @@ jobs:
         mypy --version
         # Run this mypy instance against our main package.
         mypy --install-types --non-interactive .
-        mypy --strict .
+        mypy --strict . --exclude scripts
     - name: Enviroment prepare
       run: |
         bash prepare.sh

.gitignore

@@ -141,18 +141,19 @@ run.sh
 
 # trajectory visualization
 render_cache/*
+cache/*
 
 # TMP IGNORE
 agent/prompts/jsons/*
 log_files/
-config_files/*0.json
-config_files/*1.json
-config_files/*2.json
-config_files/*3.json
-config_files/*4.json
-config_files/*5.json
-config_files/*6.json
-config_files/*7.json
-config_files/*8.json
-config_files/*9.json
-config_files/test.json
+config_files*/*0.json
+config_files*/*1.json
+config_files*/*2.json
+config_files*/*3.json
+config_files*/*4.json
+config_files*/*5.json
+config_files*/*6.json
+config_files*/*7.json
+config_files*/*8.json
+config_files*/*9.json
+config_files*/test.json

agent/agent.py

@@ -15,11 +15,14 @@
     create_playwright_action,
 )
 from browser_env.utils import Observation, StateInfo
-from llms import lm_config
-from llms.providers.openai_utils import (
+from llms import (
+    call_llm,
+    generate_from_huggingface_completion,
     generate_from_openai_chat_completion,
     generate_from_openai_completion,
+    lm_config,
 )
+from llms.tokenizers import Tokenizer
 
 
 class Agent:
@@ -120,82 +123,50 @@ def next_action(
             trajectory, intent, meta_data
         )
         lm_config = self.lm_config
-        if lm_config.provider == "openai":
-            if lm_config.mode == "chat":
-                response = generate_from_openai_chat_completion(
-                    messages=prompt,
-                    model=lm_config.model,
-                    temperature=lm_config.gen_config["temperature"],
-                    top_p=lm_config.gen_config["top_p"],
-                    context_length=lm_config.gen_config["context_length"],
-                    max_tokens=lm_config.gen_config["max_tokens"],
-                    stop_token=None,
-                )
-            elif lm_config.mode == "completion":
-                response = generate_from_openai_completion(
-                    prompt=prompt,
-                    engine=lm_config.model,
-                    temperature=lm_config.gen_config["temperature"],
-                    max_tokens=lm_config.gen_config["max_tokens"],
-                    top_p=lm_config.gen_config["top_p"],
-                    stop_token=lm_config.gen_config["stop_token"],
-                )
-            else:
-                raise ValueError(
-                    f"OpenAI models do not support mode {lm_config.mode}"
+        n = 0
+        while True:
+            response = call_llm(lm_config, prompt)
+            force_prefix = self.prompt_constructor.instruction[
+                "meta_data"
+            ].get("force_prefix", "")
+            response = f"{force_prefix}{response}"
+            n += 1
+            try:
+                parsed_response = self.prompt_constructor.extract_action(
+                    response
                 )
-        else:
-            raise NotImplementedError(
-                f"Provider {lm_config.provider} not implemented"
-            )
-
-        try:
-            parsed_response = self.prompt_constructor.extract_action(response)
-            if self.action_set_tag == "id_accessibility_tree":
-                action = create_id_based_action(parsed_response)
-            elif self.action_set_tag == "playwright":
-                action = create_playwright_action(parsed_response)
-            else:
-                raise ValueError(f"Unknown action type {self.action_set_tag}")
-
-            action["raw_prediction"] = response
-
-        except ActionParsingError as e:
-            action = create_none_action()
-            action["raw_prediction"] = response
+                if self.action_set_tag == "id_accessibility_tree":
+                    action = create_id_based_action(parsed_response)
+                elif self.action_set_tag == "playwright":
+                    action = create_playwright_action(parsed_response)
+                else:
+                    raise ValueError(
+                        f"Unknown action type {self.action_set_tag}"
+                    )
+                action["raw_prediction"] = response
+                break
+            except ActionParsingError as e:
+                if n >= lm_config.gen_config["max_retry"]:
+                    action = create_none_action()
+                    action["raw_prediction"] = response
+                    break
 
         return action
 
     def reset(self, test_config_file: str) -> None:
         pass
 
 
-def construct_llm_config(args: argparse.Namespace) -> lm_config.LMConfig:
-    llm_config = lm_config.LMConfig(
-        provider=args.provider, model=args.model, mode=args.mode
-    )
-    if args.provider == "openai":
-        llm_config.gen_config["temperature"] = args.temperature
-        llm_config.gen_config["top_p"] = args.top_p
-        llm_config.gen_config["context_length"] = args.context_length
-        llm_config.gen_config["max_tokens"] = args.max_tokens
-        llm_config.gen_config["stop_token"] = args.stop_token
-        llm_config.gen_config["max_obs_length"] = args.max_obs_length
-    else:
-        raise NotImplementedError(f"provider {args.provider} not implemented")
-    return llm_config
-
-
 def construct_agent(args: argparse.Namespace) -> Agent:
-    llm_config = construct_llm_config(args)
+    llm_config = lm_config.construct_llm_config(args)
 
     agent: Agent
     if args.agent_type == "teacher_forcing":
         agent = TeacherForcingAgent()
     elif args.agent_type == "prompt":
         with open(args.instruction_path) as f:
             constructor_type = json.load(f)["meta_data"]["prompt_constructor"]
-        tokenizer = tiktoken.encoding_for_model(llm_config.model)
+        tokenizer = Tokenizer(args.provider, args.model)
         prompt_constructor = eval(constructor_type)(
             args.instruction_path, lm_config=llm_config, tokenizer=tokenizer
         )

agent/prompts/prompt_constructor.py

@@ -3,14 +3,12 @@
 from pathlib import Path
 from typing import Any, TypedDict
 
-import tiktoken
-
 from browser_env import Action, ActionParsingError, Trajectory
 from browser_env.env_config import URL_MAPPINGS
 from browser_env.utils import StateInfo
 from llms import lm_config
-
-APIInput = str | list[Any] | dict[str, Any]
+from llms.tokenizers import Tokenizer
+from llms.utils import APIInput
 
 
 class Instruction(TypedDict):
@@ -27,12 +25,12 @@ def __init__(
         self,
         instruction_path: str | Path,
         lm_config: lm_config.LMConfig,
-        tokenizer: tiktoken.core.Encoding,
+        tokenizer: Tokenizer,
     ):
-        self.instrction_path = Path(instruction_path)
+        self.instruction_path = Path(instruction_path)
         self.obs_modality = "text"
         self.lm_config = lm_config
-        instruction = json.load(open(self.instrction_path))
+        instruction = json.load(open(self.instruction_path))
         instruction["examples"] = [tuple(e) for e in instruction["examples"]]
         self.instruction: Instruction = instruction
         self.tokenizer = tokenizer
@@ -77,6 +75,37 @@ def get_lm_api_input(
                 raise ValueError(
                     f"OpenAI models do not support mode {self.lm_config.mode}"
                 )
+        elif "huggingface" in self.lm_config.provider:
+            # https://huggingface.co/blog/llama2#how-to-prompt-llama-2
+            # https://github.com/facebookresearch/llama/blob/main/llama/generation.py#L320
+            if "Llama-2" in self.lm_config.model:
+                if self.lm_config.mode == "chat":
+                    B_INST, E_INST = "[INST]", "[/INST]"
+                    B_SYS, E_SYS = "<<SYS>>\n", "\n<</SYS>>\n\n"
+                    BOS, EOS = "<s>", "</s>"
+                    # adding the system message to be the starting of the first example
+                    examples = [
+                        (
+                            B_SYS + intro + E_SYS + examples[0][0],
+                            examples[0][1],
+                        )
+                    ] + examples[1:]
+                    message = "".join(
+                        [
+                            f"{BOS}{B_INST} {x.strip()} {E_INST} {y.strip()} {EOS}"
+                            for (x, y) in examples
+                        ]
+                    )
+                    # add the current observation
+                    message += f"{BOS}{B_INST} {current.strip()} {E_INST} {self.instruction['meta_data'].get('force_prefix', '')}"
+
+                    return message
+                else:
+                    raise ValueError("Only chat mode is supported for Llama-2")
+            else:
+                raise ValueError(
+                    f"Huggingface models do not support model_tag {self.lm_config.gen_config['model_tag']}"
+                )
         else:
             raise NotImplementedError(
                 f"Provider {self.lm_config.provider} not implemented"
@@ -102,6 +131,9 @@ def map_url_to_local(self, url: str) -> str:
         for i, j in URL_MAPPINGS.items():
             if j in url:
                 url = url.replace(j, i)
+            # https
+            if j.replace("http", "https") in url:
+                url = url.replace(j.replace("http", "https"), i)
         return url
 
     def _extract_action(self, response: str) -> str:
@@ -120,7 +152,7 @@ def __init__(
         self,
         instruction_path: str | Path,
         lm_config: lm_config.LMConfig,
-        tokenizer: tiktoken.core.Encoding,
+        tokenizer: Tokenizer,
     ):
         super().__init__(instruction_path, lm_config, tokenizer)
 
@@ -161,10 +193,10 @@ def construct(
 
     def _extract_action(self, response: str) -> str:
         action_splitter = self.instruction["meta_data"]["action_splitter"]
-        pattern = rf"{action_splitter}(.*?){action_splitter}"
+        pattern = rf"{action_splitter}((.|\n)*?){action_splitter}"
         match = re.search(pattern, response)
         if match:
-            return match.group(1)
+            return match.group(1).strip()
         else:
             raise ActionParsingError(
                 f"Cannot parse action from response {response}"
@@ -178,7 +210,7 @@ def __init__(
         self,
         instruction_path: str | Path,
         lm_config: lm_config.LMConfig,
-        tokenizer: tiktoken.core.Encoding,
+        tokenizer: Tokenizer,
     ):
         super().__init__(instruction_path, lm_config, tokenizer)
         self.answer_phrase = self.instruction["meta_data"]["answer_phrase"]
@@ -218,10 +250,10 @@ def construct(
     def _extract_action(self, response: str) -> str:
         # find the first occurence of action
         action_splitter = self.instruction["meta_data"]["action_splitter"]
-        pattern = rf"{action_splitter}(.*?){action_splitter}"
+        pattern = rf"{action_splitter}((.|\n)*?){action_splitter}"
         match = re.search(pattern, response)
         if match:
-            return match.group(1)
+            return match.group(1).strip()
         else:
             raise ActionParsingError(
                 f'Cannot find the answer phrase "{self.answer_phrase}" in "{response}"'

agent/prompts/raw/p_cot_id_actree_2s_no_na.py

@@ -0,0 +1,82 @@
+prompt = {
+	"intro": """You are an autonomous intelligent agent tasked with navigating a web browser. You will be given web-based tasks. These tasks will be accomplished through the use of specific actions you can issue.
+
+Here's the information you'll have:
+The user's objective: This is the task you're trying to complete.
+The current web page's accessibility tree: This is a simplified representation of the webpage, providing key information.
+The current web page's URL: This is the page you're currently navigating.
+The open tabs: These are the tabs you have open.
+The previous action: This is the action you just performed. It may be helpful to track your progress.
+
+The actions you can perform fall into several categories:
+
+Page Operation Actions:
+`click [id]`: This action clicks on an element with a specific id on the webpage.
+`type [id] [content] [press_enter_after=0|1]`: Use this to type the content into the field with id. By default, the "Enter" key is pressed after typing unless press_enter_after is set to 0.
+`hover [id]`: Hover over an element with id.
+`press [key_comb]`:  Simulates the pressing of a key combination on the keyboard (e.g., Ctrl+v).
+`scroll [direction=down|up]`: Scroll the page up or down.
+
+Tab Management Actions:
+`new_tab`: Open a new, empty browser tab.
+`tab_focus [tab_index]`: Switch the browser's focus to a specific tab using its index.
+`close_tab`: Close the currently active tab.
+
+URL Navigation Actions:
+`goto [url]`: Navigate to a specific URL.
+`go_back`: Navigate to the previously viewed page.
+`go_forward`: Navigate to the next page (if a previous 'go_back' action was performed).
+
+Completion Action:
+`stop [answer]`: Issue this action when you believe the task is complete. If the objective is to find a text-based answer, provide the answer in the bracket.
+
+Homepage:
+If you want to visit other websites, check out the homepage at http://homepage.com. It has a list of websites you can visit.
+http://homepage.com/password.html lists all the account name and password for the websites. You can use them to log in to the websites.
+
+To be successful, it is very important to follow the following rules:
+1. You should only issue an action that is valid given the current observation
+2. You should only issue one action at a time.
+3. You should follow the examples to reason step by step and then issue the next action.
+4. Generate the action in the correct format. Start with a "In summary, the next action I will perform is" phrase, followed by action inside ``````. For example, "In summary, the next action I will perform is ```click [1234]```".
+5. Issue stop action when you think you have achieved the objective. Don't generate anything after stop.""",
+	"examples": [
+		(
+			"""OBSERVATION:
+[1744] link 'HP CB782A#ABA 640 Inkjet Fax Machine (Renewed)'
+		[1749] StaticText '$279.49'
+		[1757] button 'Add to Cart'
+		[1760] button 'Add to Wish List'
+		[1761] button 'Add to Compare'
+URL: http://onestopmarket.com/office-products/office-electronics.html
+OBJECTIVE: What is the price of HP Inkjet Fax Machine
+PREVIOUS ACTION: None""",
+			"Let's think step-by-step. This page list the information of HP Inkjet Fax Machine, which is the product identified in the objective. Its price is $279.49. I think I have achieved the objective. I will issue the stop action with the answer. In summary, the next action I will perform is ```stop [$279.49]```",
+		),
+		(
+			"""OBSERVATION:
+[164] textbox 'Search' focused: True required: False
+[171] button 'Go'
+[174] link 'Find directions between two points'
+[212] heading 'Search Results'
+[216] button 'Close'
+URL: http://openstreetmap.org
+OBJECTIVE: Show me the restaurants near CMU
+PREVIOUS ACTION: None""",
+			"Let's think step-by-step. This page has a search box whose ID is [164]. According to the nominatim rule of openstreetmap, I can search for the restaurants near a location by \"restaurants near\". I can submit my typing by pressing the Enter afterwards. In summary, the next action I will perform is ```type [164] [restaurants near CMU] [1]```",
+		),
+	],
+	"template": """OBSERVATION:
+{observation}
+URL: {url}
+OBJECTIVE: {objective}
+PREVIOUS ACTION: {previous_action}""",
+	"meta_data": {
+		"observation": "accessibility_tree",
+		"action_type": "id_accessibility_tree",
+		"keywords": ["url", "objective", "observation", "previous_action"],
+		"prompt_constructor": "CoTPromptConstructor",
+		"answer_phrase": "In summary, the next action I will perform is",
+		"action_splitter": "```"
+	},
+}