docs: update mistral models in zoo

examples/4_engines_zoo.py

@@ -41,20 +41,13 @@
 from kani.engines.huggingface.llama2 import LlamaEngine
 engine = LlamaEngine(model_id="meta-llama/Llama-2-7b-chat-hf", use_auth_token=True)  # log in with huggingface-cli
 
-# ---- Mixtral-8x22B (Hugging Face) ----
+# ---- Mistral Small/Large (Hugging Face) ----
 from kani.engines.huggingface import HuggingEngine
-from kani.prompts.impl.mistral import MISTRAL_V3_PIPELINE, MixtralFunctionCallingAdapter
-model = HuggingEngine(
-    model_id="mistralai/Mixtral-8x22B-Instruct-v0.1",
-    prompt_pipeline=MISTRAL_V3_PIPELINE,
-    model_load_kwargs={"device_map": "auto", "torch_dtype": torch.bfloat16},
-)
-
-# to enable function calling:
-# NOTE: as of May 2024, the huggingface implementation of Mixtral-8x22B function calling is broken:
-# https://huggingface.co/mistralai/Mixtral-8x22B-Instruct-v0.1/discussions/27
-# this comment will be removed once it is fixed - until then, you should use another backend
-engine = MixtralFunctionCallingAdapter(model)
+from kani.prompts.impl.mistral import MISTRAL_V3_PIPELINE, MistralFunctionCallingAdapter
+# small (22B):  mistralai/Mistral-Small-Instruct-2409
+# large (123B): mistralai/Mistral-Large-Instruct-2407
+model = HuggingEngine(model_id="mistralai/Mistral-Small-Instruct-2409", prompt_pipeline=MISTRAL_V3_PIPELINE)
+engine = MistralFunctionCallingAdapter(model)
 
 # ---- Mistral-7B (Hugging Face) ----
 # v0.3 (supports function calling):
@@ -71,11 +64,8 @@
 # Also use the MISTRAL_V1_PIPELINE for Mixtral-8x7B (i.e. mistralai/Mixtral-8x7B-Instruct-v0.1).
 
 # ---- Command R (Hugging Face) ----
-import torch
 from kani.engines.huggingface.cohere import CommandREngine
-engine = CommandREngine(
-    model_id="CohereForAI/c4ai-command-r-v01", model_load_kwargs={"device_map": "auto", "torch_dtype": torch.float16}
-)
+engine = CommandREngine(model_id="CohereForAI/c4ai-command-r-v01")
 
 # ---- Gemma (Hugging Face) ----
 from kani.engines.huggingface import HuggingEngine