Merge pull request #138 from acon96/release/v0.2.16

Release v0.2.16

README.md

@@ -126,6 +126,7 @@ In order to facilitate running the project entirely on the system where Home Ass
 ## Version History
 | Version | Description                                                                                                                                                                                                          |
 |---------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
+| v0.2.16 | Fix for missing huggingface_hub package preventing startup                                                                                                                                                                   |
 | v0.2.15 | Fix startup error when using llama.cpp backend and add flash attention to llama.cpp backend                                                                                                                          |
 | v0.2.14 | Fix llama.cpp wheels + AVX detection                                                                                                                                                                                 |
 | v0.2.13 | Add support for Llama 3, build llama.cpp wheels that are compatible with non-AVX systems, fix an error with exposing script entities, fix multiple small Ollama backend issues, and add basic multi-language support |

custom_components/llama_conversation/agent.py

@@ -270,7 +270,7 @@ class LLaMAAgent(AbstractConversationAgent):
             intent_response = intent.IntentResponse(language=user_input.language)
             intent_response.async_set_error(
                 intent.IntentResponseErrorCode.UNKNOWN,
-                f"Sorry, there was a problem talking to the backend: {err}",
+                f"Sorry, there was a problem talking to the backend: {repr(err)}",
             )
             return ConversationResult(
                 response=intent_response, conversation_id=conversation_id

custom_components/llama_conversation/const.py

@@ -70,6 +70,7 @@ PROMPT_TEMPLATE_LLAMA3 = "llama3"
 PROMPT_TEMPLATE_NONE = "no_prompt_template"
 PROMPT_TEMPLATE_ZEPHYR = "zephyr"
 PROMPT_TEMPLATE_ZEPHYR2 = "zephyr2"
+PROMPT_TEMPLATE_ZEPHYR3 = "zephyr3"
 DEFAULT_PROMPT_TEMPLATE = PROMPT_TEMPLATE_CHATML
 PROMPT_TEMPLATE_DESCRIPTIONS = {
     PROMPT_TEMPLATE_CHATML: {
@@ -113,6 +114,12 @@ PROMPT_TEMPLATE_DESCRIPTIONS = {
         "assistant": { "prefix": "<|assistant|>\n", "suffix": "</s>" },
         "generation_prompt": "<|assistant|>\n"
     },
+    PROMPT_TEMPLATE_ZEPHYR3: {
+        "system": { "prefix": "<|system|>\n", "suffix": "<|end|>" },
+        "user": { "prefix": "<|user|>\n", "suffix": "<|end|>" },
+        "assistant": { "prefix": "<|assistant|>\n", "suffix": "<|end|>" },
+        "generation_prompt": "<|assistant|>\n"
+    },
     PROMPT_TEMPLATE_LLAMA3: {
         "system": { "prefix": "<|start_header_id|>system<|end_header_id|>\n\n", "suffix": "<|eot_id|>"},
         "user": { "prefix": "<|start_header_id|>user<|end_header_id|>\n\n", "suffix": "<|eot_id|>"},
@@ -271,8 +278,11 @@ OPTIONS_OVERRIDES = {
     "zephyr": {
         CONF_PROMPT: DEFAULT_PROMPT_BASE + ICL_EXTRAS,
         CONF_PROMPT_TEMPLATE: PROMPT_TEMPLATE_ZEPHYR,
+    },
+    "phi-3": {
+        CONF_PROMPT_TEMPLATE: PROMPT_TEMPLATE_ZEPHYR3
     }
 }
 
-INTEGRATION_VERSION = "0.2.15"
+INTEGRATION_VERSION = "0.2.16"
 EMBEDDED_LLAMA_CPP_PYTHON_VERSION = "0.2.69"

custom_components/llama_conversation/manifest.json

@@ -1,7 +1,7 @@
 {
   "domain": "llama_conversation",
   "name": "LLaMA Conversation",
-  "version": "0.2.15",
+  "version": "0.2.16",
   "codeowners": ["@acon96"],
   "config_flow": true,
   "dependencies": ["conversation"],
@@ -9,8 +9,8 @@
   "integration_type": "service",
   "iot_class": "local_polling",
   "requirements": [
-    "requests",
-    "huggingface-hub",
-    "webcolors"
+    "requests==2.31.0",
+    "huggingface-hub==0.23.0",
+    "webcolors==1.13"
   ]
 }

custom_components/llama_conversation/translations/en.json

@@ -164,6 +164,7 @@
                 "mistral": "Mistral",
                 "zephyr": "Zephyr (<|endoftext|>)",
                 "zephyr2": "Zephyr ('</s>')",
+                "zephyr3": "Zephyr (<|end|>)",
                 "llama3": "Llama 3",
                 "no_prompt_template": "None"
             }

custom_components/llama_conversation/utils.py

@@ -7,7 +7,6 @@ import multiprocessing
 import voluptuous as vol
 import webcolors
 from importlib.metadata import version
-from huggingface_hub import hf_hub_download, HfFileSystem
 
 from homeassistant.requirements import pip_kwargs
 from homeassistant.util.package import install_package, is_installed
@@ -48,6 +47,11 @@ def flatten_vol_schema(schema):
     return flattened
 
 def download_model_from_hf(model_name: str, quantization_type: str, storage_folder: str):
+    try:
+        from huggingface_hub import hf_hub_download, HfFileSystem
+    except Exception as ex:
+        raise Exception(f"Failed to import huggingface-hub library. Please re-install the integration.") from ex
+    
     fs = HfFileSystem()
     potential_files = [ f for f in fs.glob(f"{model_name}/*.gguf") ]
     wanted_file = [f for f in potential_files if (f".{quantization_type.lower()}." in f or f".{quantization_type.upper()}." in f)]