### Initialize and Run ChemAgent

Source: https://github.com/osu-nlp-group/chemtoolagent/blob/main/README.md

Demonstrates how to initialize the ChemAgent with a specified model and API keys, and then run a query to get the final answer, tool use chain, and conversation history.

```Python
from api_keys import api_keys
from chemagent import ChemAgent

agent = ChemAgent(model='gpt-4o-2024-08-06', api_keys=api_keys)

query = "What is the molecular weight the chemical compound Caffeine."
final_answer, tool_use_chain, conversation, conversation_with_icl = agent.run(query)
```

--------------------------------

### Set up Python Environment

Source: https://github.com/osu-nlp-group/chemtoolagent/blob/main/README.md

Installs project dependencies using pip and sets up a conda environment. It also mentions manual installation for Uni-Core.

```Bash
conda create -n chemagent python=3.9
pip install -r requirements.txt

# Manually install Uni-Core, following https://github.com/dptech-corp/Uni-Core
```

--------------------------------

### Start Jupyter Server

Source: https://github.com/osu-nlp-group/chemtoolagent/blob/main/README.md

Navigates to the python_server directory and starts a Jupyter server on port 8888.

```Bash
./start_jupyter_server.sh 8888
```

--------------------------------

### Import Libraries for ChemAgent

Source: https://github.com/osu-nlp-group/chemtoolagent/blob/main/playground.ipynb

Imports the logging module, API keys, and the ChemAgent class from the chemagent library. These are essential for setting up and running the agent.

```Python
import logging
from api_keys import api_keys
from chemagent import ChemAgent
```

--------------------------------

### Initialize ChemAgent

Source: https://github.com/osu-nlp-group/chemtoolagent/blob/main/playground.ipynb

Initializes an instance of the ChemAgent with a specified model ('gpt-4o-2024-08-06') and API keys. This sets up the agent for processing queries.

```Python
agent = ChemAgent(model='gpt-4o-2024-08-06', api_keys=api_keys)
```

--------------------------------

### Run ChemAgent with Rephrasing and Formatting Options

Source: https://github.com/osu-nlp-group/chemtoolagent/blob/main/playground.ipynb

Demonstrates running the ChemAgent with additional parameters: 'do_rephrasing=True' to enable query rephrasing and 'format' to control the output format, specifically requesting only the result number and unit.

```Python
# You could also use the 'do_rephrasing' and 'format' parameters to control the output
final_answer, tool_use_chain, conversation, conversation_with_icl = agent.run(query, do_rephrasing=True, format='Only return the result number and unit.')
```

--------------------------------

### Run ChemAgent with Default Parameters

Source: https://github.com/osu-nlp-group/chemtoolagent/blob/main/playground.ipynb

Executes the ChemAgent's run method with a predefined query. This call returns the final answer, the tool use chain, and the conversation history.

```Python
# Run the agent
final_answer, tool_use_chain, conversation, conversation_with_icl = agent.run(query)
```

--------------------------------

### Define Query for ChemAgent

Source: https://github.com/osu-nlp-group/chemtoolagent/blob/main/playground.ipynb

Defines a sample query string to be processed by the ChemAgent. This query asks for the molecular weight of Caffeine.

```Python
query = "What is the molecular weight the chemical compound Caffeine."
```

--------------------------------

### Configure Debug Logging for ChemAgent

Source: https://github.com/osu-nlp-group/chemtoolagent/blob/main/playground.ipynb

Provides commented-out code to enable debug logging for the ChemAgent. This allows for detailed monitoring of the agent's operations, including its internal processes and any potential issues.

```Python
# Uncomment the following line to enable debug logging

# logger = logging.getLogger('chemagent')
# logger.setLevel(logging.DEBUG)
# handler = logging.StreamHandler()
# formatter = logging.Formatter("[%(levelname)s] %(name)s: %(message)s")
# handler.setFormatter(formatter)
# logger.addHandler(handler)
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.