Skip to content

Commit c77394b

Browse files
init27init27
committed
Fix llama-cli call
1 parent 386092c commit c77394b

File tree

1 file changed

+64
-117
lines changed

1 file changed

+64
-117
lines changed

‎recipes/quickstart/agents/Agents_101/Tool_Calling_101.ipynb

Lines changed: 64 additions & 117 deletions
Original file line numberDiff line numberDiff line change
@@ -299,7 +299,7 @@
299299
},
300300
{
301301
"cell_type": "code",
302-
"execution_count": 2,
302+
"execution_count": 12,
303303
"metadata": {},
304304
"outputs": [],
305305
"source": [
@@ -312,119 +312,98 @@
312312
"source": [
313313
"#### Now we can learn about the various prompt formats available \n",
314314
"\n",
315-
"When you run the cell below-you will see all the options for different template"
315+
"When you run the cell below-you will see models available and then we can check details for model specific prompts"
316316
]
317317
},
318318
{
319319
"cell_type": "code",
320-
"execution_count": 90,
320+
"execution_count": 20,
321321
"metadata": {},
322322
"outputs": [
323323
{
324324
"name": "stdout",
325325
"output_type": "stream",
326326
"text": [
327-
"+-----------+---------------------------------+\n",
328-
"\u001b[1m\u001b[97m| Role | Template Name |\u001b[0m\n",
329-
"+-----------+---------------------------------+\n",
330-
"| user | user-default |\n",
331-
"| assistant | assistant-builtin-tool-call |\n",
332-
"| assistant | assistant-custom-tool-call |\n",
333-
"| assistant | assistant-default |\n",
334-
"| system | system-builtin-and-custom-tools |\n",
335-
"| system | system-builtin-tools-only |\n",
336-
"| system | system-custom-tools-only |\n",
337-
"| system | system-default |\n",
338-
"| tool | tool-success |\n",
339-
"| tool | tool-failure |\n",
340-
"+-----------+---------------------------------+\n"
327+
"Traceback (most recent call last):\n",
328+
" File \"/opt/miniconda3/bin/llama\", line 8, in <module>\n",
329+
" sys.exit(main())\n",
330+
" ^^^^^^\n",
331+
" File \"/opt/miniconda3/lib/python3.12/site-packages/llama_toolchain/cli/llama.py\", line 44, in main\n",
332+
" parser.run(args)\n",
333+
" File \"/opt/miniconda3/lib/python3.12/site-packages/llama_toolchain/cli/llama.py\", line 38, in run\n",
334+
" args.func(args)\n",
335+
" File \"/opt/miniconda3/lib/python3.12/site-packages/llama_toolchain/cli/model/prompt_format.py\", line 59, in _run_model_template_cmd\n",
336+
" raise argparse.ArgumentTypeError(\n",
337+
"argparse.ArgumentTypeError: llama3_1 is not a valid Model. Choose one from --\n",
338+
"Llama3.1-8B\n",
339+
"Llama3.1-70B\n",
340+
"Llama3.1-405B\n",
341+
"Llama3.1-8B-Instruct\n",
342+
"Llama3.1-70B-Instruct\n",
343+
"Llama3.1-405B-Instruct\n",
344+
"Llama3.2-1B\n",
345+
"Llama3.2-3B\n",
346+
"Llama3.2-1B-Instruct\n",
347+
"Llama3.2-3B-Instruct\n",
348+
"Llama3.2-11B-Vision\n",
349+
"Llama3.2-90B-Vision\n",
350+
"Llama3.2-11B-Vision-Instruct\n",
351+
"Llama3.2-90B-Vision-Instruct\n"
341352
]
342353
}
343354
],
344355
"source": [
345-
"!llama model template"
346-
]
347-
},
348-
{
349-
"cell_type": "markdown",
350-
"metadata": {},
351-
"source": [
352-
"## Tool Calling: Using the correct Prompt Template\n",
353-
"\n",
354-
"With `llama-cli` we can learn the correct way of defining `System_prompt` and finally get the correct behaviour from the model"
356+
"!llama model prompt-format "
355357
]
356358
},
357359
{
358360
"cell_type": "code",
359-
"execution_count": 92,
361+
"execution_count": 21,
360362
"metadata": {},
361363
"outputs": [
362364
{
363365
"name": "stdout",
364366
"output_type": "stream",
365367
"text": [
366-
"+----------+--------------------------------------------------------------+\n",
367-
"| Name | \u001b[1m\u001b[97msystem-builtin-tools-only\u001b[0m |\n",
368-
"+----------+--------------------------------------------------------------+\n",
369-
"| Template | \u001b[1m\u001b[33m<|begin_of_text|>\u001b[0m\u001b[1m\u001b[33m<|start_header_id|>\u001b[0msystem\u001b[1m\u001b[33m<|end_header_id|>\u001b[0m↵ |\n",
370-
"| | ↵ |\n",
371-
"| | Environment: ipython↵ |\n",
372-
"| | Tools: brave_search, wolfram_alpha↵ |\n",
373-
"| | Cutting Knowledge Date: December 2023↵ |\n",
374-
"| | Today Date: 15 September 2024↵ |\n",
375-
"| | ↵ |\n",
376-
"| | You are a helpful assistant.↵ |\n",
377-
"| | \u001b[1m\u001b[33m<|eot_id|>\u001b[0m\u001b[1m\u001b[33m<|start_header_id|>\u001b[0massistant\u001b[1m\u001b[33m<|end_header_id|>\u001b[0m↵ |\n",
378-
"| | ↵ |\n",
379-
"| | |\n",
380-
"+----------+--------------------------------------------------------------+\n",
381-
"| Notes | ↵ represents newline |\n",
382-
"+----------+--------------------------------------------------------------+\n"
368+
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[m━━━━━━━━━━━━━━━━━━━┓\u001b[m\n",
369+
"┃ \u001b[1mLlama 3.1 - Prompt Formats\u001b[0m \u001b[m\u001b[1m\u001b[0m ┃\u001b[m\n",
370+
"┗━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[m━━━━━━━━━━━━━━━━━━━┛\u001b[m\n",
371+
"\u001b[m\n",
372+
"\u001b[m\n",
373+
" \u001b[1;4mTokens\u001b[0m \u001b[m\u001b[1;4m\u001b[0m \u001b[m\n",
374+
"\u001b[m\n",
375+
"Here is a list of special tokens that are supported by Llama 3.1: \u001b[m \u001b[m\n",
376+
"\u001b[m\n",
377+
"\u001b[1;33m • \u001b[0m\u001b[1;36;40m<|begin_of_text|>\u001b[0m: Specifies the start of the prompt \u001b[m\u001b[1;33m\u001b[0m\u001b[1;36;40m\u001b[0m \u001b[m\n",
378+
"\u001b[1;33m • \u001b[0m\u001b[1;36;40m<|end_of_text|>\u001b[0m: Model will cease to generate more tokens. This token is gene\u001b[m\u001b[1;33m\u001b[0m\u001b[1;36;40m\u001b[0mrated only by the \u001b[m\n",
379+
"\u001b[1;33m \u001b[0mbase models. \u001b[m\u001b[1;33m\u001b[0m \u001b[m\n",
380+
"\u001b[1;33m • \u001b[0m\u001b[1;36;40m<|finetune_right_pad_id|>\u001b[0m: This token is used for padding text sequences to t\u001b[m\u001b[1;33m\u001b[0m\u001b[1;36;40m\u001b[0mhe same length in a \u001b[m\n",
381+
"\u001b[1;33m \u001b[0mbatch. \u001b[m:\u001b[K"
383382
]
384383
}
385384
],
386385
"source": [
387-
"!llama model template --name system-builtin-tools-only"
386+
"!llama model prompt-format -m Llama3.1-8B"
388387
]
389388
},
390389
{
391390
"cell_type": "markdown",
392391
"metadata": {},
393392
"source": [
394-
"If everything is setup correctly-the model should now wrap function calls with the `|<python_tag>|` following the actualy function call. \n",
395-
"\n",
396-
"This can allow you to manage your function calling logic accordingly. \n",
393+
"## Tool Calling: Using the correct Prompt Template\n",
397394
"\n",
398-
"Time to test the theory"
395+
"With `llama-cli` we have already learned the right behaviour of the model"
399396
]
400397
},
401398
{
402-
"cell_type": "code",
403-
"execution_count": 94,
399+
"cell_type": "markdown",
404400
"metadata": {},
405-
"outputs": [
406-
{
407-
"name": "stdout",
408-
"output_type": "stream",
409-
"text": [
410-
"+----------+----------------------------------------------------------------------------------+\n",
411-
"| Name | \u001b[1m\u001b[97massistant-builtin-tool-call\u001b[0m |\n",
412-
"+----------+----------------------------------------------------------------------------------+\n",
413-
"| Template | \u001b[1m\u001b[33m<|begin_of_text|>\u001b[0m\u001b[1m\u001b[33m<|start_header_id|>\u001b[0massistant\u001b[1m\u001b[33m<|end_header_id|>\u001b[0m↵ |\n",
414-
"| | ↵ |\n",
415-
"| | \u001b[1m\u001b[33m<|python_tag|>\u001b[0mbrave_search.call(query=\"Who won NBA in |\n",
416-
"| | 2024?\")\u001b[1m\u001b[33m<|eom_id|>\u001b[0m\u001b[1m\u001b[33m<|start_header_id|>\u001b[0massistant\u001b[1m\u001b[33m<|end_header_id|>\u001b[0m↵ |\n",
417-
"| | ↵ |\n",
418-
"| | |\n",
419-
"+----------+----------------------------------------------------------------------------------+\n",
420-
"| Notes | ↵ represents newline |\n",
421-
"| | Notice <|python_tag|> |\n",
422-
"+----------+----------------------------------------------------------------------------------+\n"
423-
]
424-
}
425-
],
426401
"source": [
427-
"!llama model template --name assistant-builtin-tool-call"
402+
"If everything is setup correctly-the model should now wrap function calls with the `|<python_tag>|` following the actualy function call. \n",
403+
"\n",
404+
"This can allow you to manage your function calling logic accordingly. \n",
405+
"\n",
406+
"Time to test the theory"
428407
]
429408
},
430409
{
@@ -684,48 +663,7 @@
684663
"\n",
685664
"[Here](https://www.llama.com/docs/model-cards-and-prompt-formats/llama3_2#-tool-calling-(1b/3b)-) are the docs for your reference that we will be using. \n",
686665
"\n",
687-
"Let's verify the details from `llama-toolchain` again and then start the prompt engineering for the small Llamas."
688-
]
689-
},
690-
{
691-
"cell_type": "code",
692-
"execution_count": 18,
693-
"metadata": {},
694-
"outputs": [
695-
{
696-
"name": "stdout",
697-
"output_type": "stream",
698-
"text": [
699-
"Traceback (most recent call last):\n",
700-
" File \"/opt/miniconda3/bin/llama\", line 8, in <module>\n",
701-
" sys.exit(main())\n",
702-
" ^^^^^^\n",
703-
" File \"/opt/miniconda3/lib/python3.12/site-packages/llama_toolchain/cli/llama.py\", line 44, in main\n",
704-
" parser.run(args)\n",
705-
" File \"/opt/miniconda3/lib/python3.12/site-packages/llama_toolchain/cli/llama.py\", line 38, in run\n",
706-
" args.func(args)\n",
707-
" File \"/opt/miniconda3/lib/python3.12/site-packages/llama_toolchain/cli/model/prompt_format.py\", line 59, in _run_model_template_cmd\n",
708-
" raise argparse.ArgumentTypeError(\n",
709-
"argparse.ArgumentTypeError: llama3_1 is not a valid Model. Choose one from --\n",
710-
"Llama3.1-8B\n",
711-
"Llama3.1-70B\n",
712-
"Llama3.1-405B\n",
713-
"Llama3.1-8B-Instruct\n",
714-
"Llama3.1-70B-Instruct\n",
715-
"Llama3.1-405B-Instruct\n",
716-
"Llama3.2-1B\n",
717-
"Llama3.2-3B\n",
718-
"Llama3.2-1B-Instruct\n",
719-
"Llama3.2-3B-Instruct\n",
720-
"Llama3.2-11B-Vision\n",
721-
"Llama3.2-90B-Vision\n",
722-
"Llama3.2-11B-Vision-Instruct\n",
723-
"Llama3.2-90B-Vision-Instruct\n"
724-
]
725-
}
726-
],
727-
"source": [
728-
"!llama model prompt-format"
666+
"Excercise for viewer: Use `llama-toolchain` again to verify like we did earlier and then start the prompt engineering for the small Llamas."
729667
]
730668
},
731669
{
@@ -903,7 +841,16 @@
903841
"cell_type": "markdown",
904842
"metadata": {},
905843
"source": [
906-
"Hello Regex, my good old friend :) "
844+
"### Handling Tool-Calling logic for the model"
845+
]
846+
},
847+
{
848+
"cell_type": "markdown",
849+
"metadata": {},
850+
"source": [
851+
"Hello Regex, my good old friend :) \n",
852+
"\n",
853+
"With Regex, we can write a simple way to handle tool_calling and return either the model or tool call response"
907854
]
908855
},
909856
{

0 commit comments

Comments
 (0)