{"owner":"meta-llama","repo":"llama-cookbook","hasSkills":true,"totalSkillsCount":4,"totalTokensCount":1309,"categories":["subagent-persona","plugin-manifest"],"hasMcp":false,"mcpConfig":null,"found":["end-to-end-use-cases/agents/DeepLearningai_Course_Notebooks/README.md","src/llama_cookbook/tools/README.md","end-to-end-use-cases/agents/README.md","end-to-end-use-cases/agents/calendar_assistant/README.md"],"skills":{"end-to-end-use-cases/agents/DeepLearningai_Course_Notebooks/README.md":"# Quickstart Llama 3 Powered Agent Examples Ported from DeepLearning.ai Short Courses\n\nThe notebooks in this folder are ported from the 4 recent agent short courses on [Deeplearning.ai](https://www.deeplearning.ai) to use Llama 3 to build agent apps from scratch or with open source frameworks (LangChain, LlamaIndex, AutoGen).\n\n1. [Functions Tools and Agents with LangChain L1 Function Calling](Functions_Tools_and_Agents_with_LangChain_L1_Function_Calling.ipynb)\n\n2. [AI Agents in LangGraph L1 Build an Agent from Scratch](AI_Agents_in_LangGraph_L1_Build_an_Agent_from_Scratch.ipynb)\n\n3. [Building Agentic RAG with Llamaindex L1 Router Engine](Building_Agentic_RAG_with_Llamaindex_L1_Router_Engine.ipynb)\n\n4. [AI Agentic Design Patterns with AutoGen L4 Tool Use and Conversational Chess](AI_Agentic_Design_Patterns_with_AutoGen_L4_Tool_Use_and_Conversational_Chess.ipynb)\n","src/llama_cookbook/tools/README.md":"# Convert Hugging Face llama weights to official llama consolidated format\n\nThis is the reverse conversion for `convert_llama_weights_to_hf.py` script from the transformer package.\n\n## Step 0: Convert to consolidated format\n- Create an output directory for the converted weights, such as `test70B`.\n- Copy file params.json from the official llama download into that directory.\n- Run the conversion script. `model-path` can be a Hugging Face hub model or a local hf model directory.\n```\npython -m llama_cookbook.tools.convert_hf_weights_to_llama --model-path meta-llama/Meta-Llama-3.1-70B-Instruct --output-dir test70B --model-size 70B\n```\n\n## Step 1: Run inference\nCheckout the official llama 3 inference [repo](https://github.com/meta-llama/llama3). Test using chat or text completion.\n```\ntorchrun --nproc_per_node 8 example_chat_completion.py --ckpt_dir ./test70B --tokenizer_path ${llama_3_dir}/tokenizer.model\n```\n\nFor validation, please compare the converted weights with official llama 2 weights\n```\npython compare_llama_weights.py test70B ${Llama-3-70B-Instruct_dir}\n```\n","end-to-end-use-cases/agents/README.md":"## Agents and Tool Calling\n\nStructure:\n\n- Agents_Tutorial: Showcases 101 and 201 notebooks guidance for using tool calling with Llama models\n- DeepLearning_Course_Notebooks: Notebooks from the DL.ai course teaching Agents","end-to-end-use-cases/agents/calendar_assistant/README.md":"# Google Calendar Assistant with with Llama 3.2 3B Tool Calling\n\nThis notebook showcases how to go about building a digital assistant to schedule meetings with the Llama 3.2 3B model. The core concepts used to implement this are Prompt Engineering and Tool Calling. This demo shows how Llama can be used to interact with 3rd party apps like Google Contacts & Google Calendar and schedule a meeting requested by the user. Even though we are using prompt engineering to achieve this, the approach described doesn't degrade the model's ability to answer general queries. This approach can extended to perform other tasks in a similar manner without affecting the quality of other tasks\n\n\n## Approach\n\nInstead of using a complex system prompt with multiple conditions & expecting Llama to perform various tasks accurately out of the box, the approach here is to treat this as a 2 step process\n- Determine user intent - Task classification\n- Take action for the specific task using Tool Calling\n\n\n\nIn the diagram shown below,\n- system prompt 1 determines the classification of the query\n- In steps 2 & 3, we classify the task being requested.\n- system prompt 2 is chosen based on the classification result\n- Steps 4 & 5 implement the classified task.\n- For the sake of demo, we show 2 classes: General & Meeting\n\n![Tool Calling Flow Diagram](./assets/flow_diagram.png)\n\nBoth these tasks have a specific prompt. We use the same model with different system prompts depending on the classification result.\nAdditionally, this demo also showcases how Llama can be used to do double tool calling with 1 prompt. In the case of Meeting, Llama returns 2 function calls in Step 5\n```\n<function=google_contact>{{\"name\": \"John Constantine\"}}</function>\n<function=google_calendar>{{\"date\": \"Mar 31 \", \"time\": \"5:30 pm\", \"attendees\": \"John Constantine\"}}</function>\n```\n\n## Examples\n\n- General query\n\n  **User Prompt** : `Tell me about Paris`\n\n  **Prompt Classification**: `general`\n\n  **Answer**: \n```\nParis, the City of Light, is the capital of France. Known for its stunning architecture, art museums, and romantic atmosphere, Paris is a popular tourist destination. The city is famous for its iconic landmarks such as the Eiffel Tower, Notre-Dame Cathedral, and the Louvre Museum, which houses the Mona Lisa. Paris is also renowned for its fashion, cuisine, and wine. The city's charming streets and historic neighborhoods, like Montmartre and Le Marais, offer a glimpse\n```\n\n- Set up a Calendar Invite\n\n  **User Prompt** : `Schedule a meeting with John Constantine on Mar 31 at 5:30 pm`\n\n  **Prompt Classification**: `meeting`\n\n  **Answer**: \n```\n  <function=google_contact>{{\"name\": \"John Constantine\"}}</function>\n  <function=google_calendar>{{\"date\": \"Mar 31\", \"time\": \"5:30 pm\", \"attendees\": \"John Constantine\"}}</function>\n```\n\n### Actions based on tool calling output\n- The google_contact function call returned by the model is used to call [People API](https://developers.google.com/people) to look up the email address of the person of interest\n- The email address from the previous step is used to call [Calendar API](https://developers.google.com/calendar) along with the other information in the google_calendar tool calling output returned by the model \n\nThe end result is that a google meeting is scheduled with the person of interest at the date & time specified\n\n"},"files":{"end-to-end-use-cases/agents/DeepLearningai_Course_Notebooks/README.md":"# Quickstart Llama 3 Powered Agent Examples Ported from DeepLearning.ai Short Courses\n\nThe notebooks in this folder are ported from the 4 recent agent short courses on [Deeplearning.ai](https://www.deeplearning.ai) to use Llama 3 to build agent apps from scratch or with open source frameworks (LangChain, LlamaIndex, AutoGen).\n\n1. [Functions Tools and Agents with LangChain L1 Function Calling](Functions_Tools_and_Agents_with_LangChain_L1_Function_Calling.ipynb)\n\n2. [AI Agents in LangGraph L1 Build an Agent from Scratch](AI_Agents_in_LangGraph_L1_Build_an_Agent_from_Scratch.ipynb)\n\n3. [Building Agentic RAG with Llamaindex L1 Router Engine](Building_Agentic_RAG_with_Llamaindex_L1_Router_Engine.ipynb)\n\n4. [AI Agentic Design Patterns with AutoGen L4 Tool Use and Conversational Chess](AI_Agentic_Design_Patterns_with_AutoGen_L4_Tool_Use_and_Conversational_Chess.ipynb)\n","src/llama_cookbook/tools/README.md":"# Convert Hugging Face llama weights to official llama consolidated format\n\nThis is the reverse conversion for `convert_llama_weights_to_hf.py` script from the transformer package.\n\n## Step 0: Convert to consolidated format\n- Create an output directory for the converted weights, such as `test70B`.\n- Copy file params.json from the official llama download into that directory.\n- Run the conversion script. `model-path` can be a Hugging Face hub model or a local hf model directory.\n```\npython -m llama_cookbook.tools.convert_hf_weights_to_llama --model-path meta-llama/Meta-Llama-3.1-70B-Instruct --output-dir test70B --model-size 70B\n```\n\n## Step 1: Run inference\nCheckout the official llama 3 inference [repo](https://github.com/meta-llama/llama3). Test using chat or text completion.\n```\ntorchrun --nproc_per_node 8 example_chat_completion.py --ckpt_dir ./test70B --tokenizer_path ${llama_3_dir}/tokenizer.model\n```\n\nFor validation, please compare the converted weights with official llama 2 weights\n```\npython compare_llama_weights.py test70B ${Llama-3-70B-Instruct_dir}\n```\n","end-to-end-use-cases/agents/README.md":"## Agents and Tool Calling\n\nStructure:\n\n- Agents_Tutorial: Showcases 101 and 201 notebooks guidance for using tool calling with Llama models\n- DeepLearning_Course_Notebooks: Notebooks from the DL.ai course teaching Agents","end-to-end-use-cases/agents/calendar_assistant/README.md":"# Google Calendar Assistant with with Llama 3.2 3B Tool Calling\n\nThis notebook showcases how to go about building a digital assistant to schedule meetings with the Llama 3.2 3B model. The core concepts used to implement this are Prompt Engineering and Tool Calling. This demo shows how Llama can be used to interact with 3rd party apps like Google Contacts & Google Calendar and schedule a meeting requested by the user. Even though we are using prompt engineering to achieve this, the approach described doesn't degrade the model's ability to answer general queries. This approach can extended to perform other tasks in a similar manner without affecting the quality of other tasks\n\n\n## Approach\n\nInstead of using a complex system prompt with multiple conditions & expecting Llama to perform various tasks accurately out of the box, the approach here is to treat this as a 2 step process\n- Determine user intent - Task classification\n- Take action for the specific task using Tool Calling\n\n\n\nIn the diagram shown below,\n- system prompt 1 determines the classification of the query\n- In steps 2 & 3, we classify the task being requested.\n- system prompt 2 is chosen based on the classification result\n- Steps 4 & 5 implement the classified task.\n- For the sake of demo, we show 2 classes: General & Meeting\n\n![Tool Calling Flow Diagram](./assets/flow_diagram.png)\n\nBoth these tasks have a specific prompt. We use the same model with different system prompts depending on the classification result.\nAdditionally, this demo also showcases how Llama can be used to do double tool calling with 1 prompt. In the case of Meeting, Llama returns 2 function calls in Step 5\n```\n<function=google_contact>{{\"name\": \"John Constantine\"}}</function>\n<function=google_calendar>{{\"date\": \"Mar 31 \", \"time\": \"5:30 pm\", \"attendees\": \"John Constantine\"}}</function>\n```\n\n## Examples\n\n- General query\n\n  **User Prompt** : `Tell me about Paris`\n\n  **Prompt Classification**: `general`\n\n  **Answer**: \n```\nParis, the City of Light, is the capital of France. Known for its stunning architecture, art museums, and romantic atmosphere, Paris is a popular tourist destination. The city is famous for its iconic landmarks such as the Eiffel Tower, Notre-Dame Cathedral, and the Louvre Museum, which houses the Mona Lisa. Paris is also renowned for its fashion, cuisine, and wine. The city's charming streets and historic neighborhoods, like Montmartre and Le Marais, offer a glimpse\n```\n\n- Set up a Calendar Invite\n\n  **User Prompt** : `Schedule a meeting with John Constantine on Mar 31 at 5:30 pm`\n\n  **Prompt Classification**: `meeting`\n\n  **Answer**: \n```\n  <function=google_contact>{{\"name\": \"John Constantine\"}}</function>\n  <function=google_calendar>{{\"date\": \"Mar 31\", \"time\": \"5:30 pm\", \"attendees\": \"John Constantine\"}}</function>\n```\n\n### Actions based on tool calling output\n- The google_contact function call returned by the model is used to call [People API](https://developers.google.com/people) to look up the email address of the person of interest\n- The email address from the previous step is used to call [Calendar API](https://developers.google.com/calendar) along with the other information in the google_calendar tool calling output returned by the model \n\nThe end result is that a google meeting is scheduled with the person of interest at the date & time specified\n\n"},"items":[{"name":"README.md","path":"src/llama_cookbook/tools/README.md","rawUrl":"https://raw.githubusercontent.com/meta-llama/llama-cookbook/HEAD/src/llama_cookbook/tools/README.md","title":"tools Documentation","category":"plugin-manifest","format":"markdown","content":"# Convert Hugging Face llama weights to official llama consolidated format\n\nThis is the reverse conversion for `convert_llama_weights_to_hf.py` script from the transformer package.\n\n## Step 0: Convert to consolidated format\n- Create an output directory for the converted weights, such as `test70B`.\n- Copy file params.json from the official llama download into that directory.\n- Run the conversion script. `model-path` can be a Hugging Face hub model or a local hf model directory.\n```\npython -m llama_cookbook.tools.convert_hf_weights_to_llama --model-path meta-llama/Meta-Llama-3.1-70B-Instruct --output-dir test70B --model-size 70B\n```\n\n## Step 1: Run inference\nCheckout the official llama 3 inference [repo](https://github.com/meta-llama/llama3). Test using chat or text completion.\n```\ntorchrun --nproc_per_node 8 example_chat_completion.py --ckpt_dir ./test70B --tokenizer_path ${llama_3_dir}/tokenizer.model\n```\n\nFor validation, please compare the converted weights with official llama 2 weights\n```\npython compare_llama_weights.py test70B ${Llama-3-70B-Instruct_dir}\n```\n","isInternal":false,"tokens":268,"sizeBytes":1079},{"name":"README.md","path":"end-to-end-use-cases/agents/calendar_assistant/README.md","rawUrl":"https://raw.githubusercontent.com/meta-llama/llama-cookbook/HEAD/end-to-end-use-cases/agents/calendar_assistant/README.md","title":"Subagent: readme","category":"subagent-persona","format":"markdown","content":"# Google Calendar Assistant with with Llama 3.2 3B Tool Calling\n\nThis notebook showcases how to go about building a digital assistant to schedule meetings with the Llama 3.2 3B model. The core concepts used to implement this are Prompt Engineering and Tool Calling. This demo shows how Llama can be used to interact with 3rd party apps like Google Contacts & Google Calendar and schedule a meeting requested by the user. Even though we are using prompt engineering to achieve this, the approach described doesn't degrade the model's ability to answer general queries. This approach can extended to perform other tasks in a similar manner without affecting the quality of other tasks\n\n\n## Approach\n\nInstead of using a complex system prompt with multiple conditions & expecting Llama to perform various tasks accurately out of the box, the approach here is to treat this as a 2 step process\n- Determine user intent - Task classification\n- Take action for the specific task using Tool Calling\n\n\n\nIn the diagram shown below,\n- system prompt 1 determines the classification of the query\n- In steps 2 & 3, we classify the task being requested.\n- system prompt 2 is chosen based on the classification result\n- Steps 4 & 5 implement the classified task.\n- For the sake of demo, we show 2 classes: General & Meeting\n\n![Tool Calling Flow Diagram](./assets/flow_diagram.png)\n\nBoth these tasks have a specific prompt. We use the same model with different system prompts depending on the classification result.\nAdditionally, this demo also showcases how Llama can be used to do double tool calling with 1 prompt. In the case of Meeting, Llama returns 2 function calls in Step 5\n```\n<function=google_contact>{{\"name\": \"John Constantine\"}}</function>\n<function=google_calendar>{{\"date\": \"Mar 31 \", \"time\": \"5:30 pm\", \"attendees\": \"John Constantine\"}}</function>\n```\n\n## Examples\n\n- General query\n\n  **User Prompt** : `Tell me about Paris`\n\n  **Prompt Classification**: `general`\n\n  **Answer**: \n```\nParis, the City of Light, is the capital of France. Known for its stunning architecture, art museums, and romantic atmosphere, Paris is a popular tourist destination. The city is famous for its iconic landmarks such as the Eiffel Tower, Notre-Dame Cathedral, and the Louvre Museum, which houses the Mona Lisa. Paris is also renowned for its fashion, cuisine, and wine. The city's charming streets and historic neighborhoods, like Montmartre and Le Marais, offer a glimpse\n```\n\n- Set up a Calendar Invite\n\n  **User Prompt** : `Schedule a meeting with John Constantine on Mar 31 at 5:30 pm`\n\n  **Prompt Classification**: `meeting`\n\n  **Answer**: \n```\n  <function=google_contact>{{\"name\": \"John Constantine\"}}</function>\n  <function=google_calendar>{{\"date\": \"Mar 31\", \"time\": \"5:30 pm\", \"attendees\": \"John Constantine\"}}</function>\n```\n\n### Actions based on tool calling output\n- The google_contact function call returned by the model is used to call [People API](https://developers.google.com/people) to look up the email address of the person of interest\n- The email address from the previous step is used to call [Calendar API](https://developers.google.com/calendar) along with the other information in the google_calendar tool calling output returned by the model \n\nThe end result is that a google meeting is scheduled with the person of interest at the date & time specified\n\n","isInternal":false,"tokens":753,"sizeBytes":3364},{"name":"README.md","path":"end-to-end-use-cases/agents/DeepLearningai_Course_Notebooks/README.md","rawUrl":"https://raw.githubusercontent.com/meta-llama/llama-cookbook/HEAD/end-to-end-use-cases/agents/DeepLearningai_Course_Notebooks/README.md","title":"Subagent: readme","category":"subagent-persona","format":"markdown","content":"# Quickstart Llama 3 Powered Agent Examples Ported from DeepLearning.ai Short Courses\n\nThe notebooks in this folder are ported from the 4 recent agent short courses on [Deeplearning.ai](https://www.deeplearning.ai) to use Llama 3 to build agent apps from scratch or with open source frameworks (LangChain, LlamaIndex, AutoGen).\n\n1. [Functions Tools and Agents with LangChain L1 Function Calling](Functions_Tools_and_Agents_with_LangChain_L1_Function_Calling.ipynb)\n\n2. [AI Agents in LangGraph L1 Build an Agent from Scratch](AI_Agents_in_LangGraph_L1_Build_an_Agent_from_Scratch.ipynb)\n\n3. [Building Agentic RAG with Llamaindex L1 Router Engine](Building_Agentic_RAG_with_Llamaindex_L1_Router_Engine.ipynb)\n\n4. [AI Agentic Design Patterns with AutoGen L4 Tool Use and Conversational Chess](AI_Agentic_Design_Patterns_with_AutoGen_L4_Tool_Use_and_Conversational_Chess.ipynb)\n","isInternal":false,"tokens":239,"sizeBytes":874},{"name":"README.md","path":"end-to-end-use-cases/agents/README.md","rawUrl":"https://raw.githubusercontent.com/meta-llama/llama-cookbook/HEAD/end-to-end-use-cases/agents/README.md","title":"Subagent: readme","category":"subagent-persona","format":"markdown","content":"## Agents and Tool Calling\n\nStructure:\n\n- Agents_Tutorial: Showcases 101 and 201 notebooks guidance for using tool calling with Llama models\n- DeepLearning_Course_Notebooks: Notebooks from the DL.ai course teaching Agents","isInternal":false,"tokens":49,"sizeBytes":221}],"systemPromptSnippet":"<yakaai_skills repo=\"meta-llama/llama-cookbook\">\n<!-- File: src/llama_cookbook/tools/README.md (Tokens: ~268 | Category: plugin-manifest) -->\n# Convert Hugging Face llama weights to official llama consolidated format\n\nThis is the reverse conversion for `convert_llama_weights_to_hf.py` script from the transformer package.\n\n## Step 0: Convert to consolidated format\n- Create an output directory for the converted weights, such as `test70B`.\n- Copy file params.json from the official llama download into that directory.\n- Run the conversion script. `model-path` can be a Hugging Face hub model or a local hf model directory.\n```\npython -m llama_cookbook.tools.convert_hf_weights_to_llama --model-path meta-llama/Meta-Llama-3.1-70B-Instruct --output-dir test70B --model-size 70B\n```\n\n## Step 1: Run inference\nCheckout the official llama 3 inference [repo](https://github.com/meta-llama/llama3). Test using chat or text completion.\n```\ntorchrun --nproc_per_node 8 example_chat_completion.py --ckpt_dir ./test70B --tokenizer_path ${llama_3_dir}/tokenizer.model\n```\n\nFor validation, please compare the converted weights with official llama 2 weights\n```\npython compare_llama_weights.py test70B ${Llama-3-70B-Instruct_dir}\n```\n\n<!-- File: end-to-end-use-cases/agents/calendar_assistant/README.md (Tokens: ~753 | Category: subagent-persona) -->\n# Google Calendar Assistant with with Llama 3.2 3B Tool Calling\n\nThis notebook showcases how to go about building a digital assistant to schedule meetings with the Llama 3.2 3B model. The core concepts used to implement this are Prompt Engineering and Tool Calling. This demo shows how Llama can be used to interact with 3rd party apps like Google Contacts & Google Calendar and schedule a meeting requested by the user. Even though we are using prompt engineering to achieve this, the approach described doesn't degrade the model's ability to answer general queries. This approach can extended to perform other tasks in a similar manner without affecting the quality of other tasks\n\n\n## Approach\n\nInstead of using a complex system prompt with multiple conditions & expecting Llama to perform various tasks accurately out of the box, the approach here is to treat this as a 2 step process\n- Determine user intent - Task classification\n- Take action for the specific task using Tool Calling\n\n\n\nIn the diagram shown below,\n- system prompt 1 determines the classification of the query\n- In steps 2 & 3, we classify the task being requested.\n- system prompt 2 is chosen based on the classification result\n- Steps 4 & 5 implement the classified task.\n- For the sake of demo, we show 2 classes: General & Meeting\n\n![Tool Calling Flow Diagram](./assets/flow_diagram.png)\n\nBoth these tasks have a specific prompt. We use the same model with different system prompts depending on the classification result.\nAdditionally, this demo also showcases how Llama can be used to do double tool calling with 1 prompt. In the case of Meeting, Llama returns 2 function calls in Step 5\n```\n<function=google_contact>{{\"name\": \"John Constantine\"}}</function>\n<function=google_calendar>{{\"date\": \"Mar 31 \", \"time\": \"5:30 pm\", \"attendees\": \"John Constantine\"}}</function>\n```\n\n## Examples\n\n- General query\n\n  **User Prompt** : `Tell me about Paris`\n\n  **Prompt Classification**: `general`\n\n  **Answer**: \n```\nParis, the City of Light, is the capital of France. Known for its stunning architecture, art museums, and romantic atmosphere, Paris is a popular tourist destination. The city is famous for its iconic landmarks such as the Eiffel Tower, Notre-Dame Cathedral, and the Louvre Museum, which houses the Mona Lisa. Paris is also renowned for its fashion, cuisine, and wine. The city's charming streets and historic neighborhoods, like Montmartre and Le Marais, offer a glimpse\n```\n\n- Set up a Calendar Invite\n\n  **User Prompt** : `Schedule a meeting with John Constantine on Mar 31 at 5:30 pm`\n\n  **Prompt Classification**: `meeting`\n\n  **Answer**: \n```\n  <function=google_contact>{{\"name\": \"John Constantine\"}}</function>\n  <function=google_calendar>{{\"date\": \"Mar 31\", \"time\": \"5:30 pm\", \"attendees\": \"John Constantine\"}}</function>\n```\n\n### Actions based on tool calling output\n- The google_contact function call returned by the model is used to call [People API](https://developers.google.com/people) to look up the email address of the person of interest\n- The email address from the previous step is used to call [Calendar API](https://developers.google.com/calendar) along with the other information in the google_calendar tool calling output returned by the model \n\nThe end result is that a google meeting is scheduled with the person of interest at the date & time specified\n\n<!-- File: end-to-end-use-cases/agents/DeepLearningai_Course_Notebooks/README.md (Tokens: ~239 | Category: subagent-persona) -->\n# Quickstart Llama 3 Powered Agent Examples Ported from DeepLearning.ai Short Courses\n\nThe notebooks in this folder are ported from the 4 recent agent short courses on [Deeplearning.ai](https://www.deeplearning.ai) to use Llama 3 to build agent apps from scratch or with open source frameworks (LangChain, LlamaIndex, AutoGen).\n\n1. [Functions Tools and Agents with LangChain L1 Function Calling](Functions_Tools_and_Agents_with_LangChain_L1_Function_Calling.ipynb)\n\n2. [AI Agents in LangGraph L1 Build an Agent from Scratch](AI_Agents_in_LangGraph_L1_Build_an_Agent_from_Scratch.ipynb)\n\n3. [Building Agentic RAG with Llamaindex L1 Router Engine](Building_Agentic_RAG_with_Llamaindex_L1_Router_Engine.ipynb)\n\n4. [AI Agentic Design Patterns with AutoGen L4 Tool Use and Conversational Chess](AI_Agentic_Design_Patterns_with_AutoGen_L4_Tool_Use_and_Conversational_Chess.ipynb)\n\n<!-- File: end-to-end-use-cases/agents/README.md (Tokens: ~49 | Category: subagent-persona) -->\n## Agents and Tool Calling\n\nStructure:\n\n- Agents_Tutorial: Showcases 101 and 201 notebooks guidance for using tool calling with Llama models\n- DeepLearning_Course_Notebooks: Notebooks from the DL.ai course teaching Agents\n</yakaai_skills>"}