Merge pull request #69 from Not-Diamond/t7-update-pplx

deprecate old sonar, update to new

Author: r0ymanesco

notdiamond/llms/providers.py

@@ -76,7 +76,7 @@ class NDLLMProviders(Enum):
         REPLICATE_META_LLAMA_3_1_405B_INSTRUCT (NDLLMProvider): refers to "meta-llama-3.1-405b-instruct"
             model served via Replicate
 
-        LLAMA_3_1_SONAR_LARGE_128K_ONLINE (NDLLMProvider): refers to "llama-3.1-sonar-large-128k-online" model by Perplexity
+        SONAR (NDLLMProvider): refers to "sonar" model by Perplexity
     """
 
     GPT_3_5_TURBO = ("openai", "gpt-3.5-turbo")
@@ -157,9 +157,9 @@ class NDLLMProviders(Enum):
     )
     TOGETHER_DEEPSEEK_R1 = ("togetherai", "DeepSeek-R1")
 
-    LLAMA_3_1_SONAR_LARGE_128K_ONLINE = (
+    SONAR = (
         "perplexity",
-        "llama-3.1-sonar-large-128k-online",
+        "sonar",
     )
 
     REPLICATE_MISTRAL_7B_INSTRUCT_V0_2 = (

notdiamond/settings.py

@@ -341,14 +341,14 @@
     },
     "perplexity": {
         "models": [
-            "llama-3.1-sonar-large-128k-online",
+            "sonar",
         ],
         "api_key": PPLX_API_KEY,
         "openrouter_identifier": {
-            "llama-3.1-sonar-large-128k-online": "perplexity/llama-3.1-sonar-large-128k-online",
+            "sonar": "perplexity/sonar",
         },
         "price": {
-            "llama-3.1-sonar-large-128k-online": {"input": 1.0, "output": 1.0},
+            "sonar": {"input": 1.0, "output": 1.0},
         },
     },
     "replicate": {

notdiamond/toolkit/litellm/litellm_notdiamond.py

@@ -76,7 +76,7 @@
     "mistral/mistral-large-2407": "mistral/mistral-large-2407",
     "mistral/mistral-large-2402": "mistral/mistral-large-2402",
     # perplexity
-    "perplexity/llama-3.1-sonar-large-128k-online": "perplexity/llama-3.1-sonar-large-128k-online",
+    "perplexity/sonar": "perplexity/sonar",
     # cohere
     "cohere/command-r": "cohere_chat/command-r",
     "cohere/command-r-plus": "cohere_chat/command-r-plus",

tests/test_llm_calls/test_perplexity.py

@@ -8,8 +8,8 @@
 @pytest.mark.vcr
 @pytest.mark.longrun
 class Test_Perplexity_LLMs:
-    def test_llama_3_1_sonar_large_128k_online(self):
-        provider = NDLLMProviders.LLAMA_3_1_SONAR_LARGE_128K_ONLINE
+    def test_sonar(self):
+        provider = NDLLMProviders.SONAR
         provider.kwargs = {"max_tokens": 10}
         nd_llm = NotDiamond(
             llm_configs=[provider],

tests/test_toolkit/langchain/test_integration.py

@@ -165,7 +165,7 @@ def test_invokable(target_model: str, patch_class: str) -> None:
 
 
 def test_init_perplexity() -> None:
-    target_model = "perplexity/llama-3.1-sonar-large-128k-online"
+    target_model = "perplexity/sonar"
     nd_client = MagicMock(
         spec=NotDiamond,
         llm_configs=[target_model],

tests/test_toolkit/langchain/test_unit.py

@@ -276,7 +276,7 @@ def test_invokable_mock(self) -> None:
             ), f"{mock_client}"
 
     def test_init_perplexity(self) -> None:
-        target_model = "perplexity/llama-3.1-sonar-large-128k-online"
+        target_model = "perplexity/sonar"
         nd_client = MagicMock(
             spec=NotDiamond,
             llm_configs=[target_model],

tests/test_toolkit/test_litellm.py

@@ -40,7 +40,7 @@
     # {"provider": "mistral", "model": "open-mixtral-8x22b"},
     # {"provider": "mistral", "model": "mistral-large-2407"},
     # {"provider": "mistral", "model": "mistral-large-2402"},
-    {"provider": "perplexity", "model": "llama-3.1-sonar-large-128k-online"},
+    {"provider": "perplexity", "model": "sonar"},
     # {"provider": "cohere", "model": "command-r"},
     {"provider": "cohere", "model": "command-r-plus"},
     # {"provider": "google", "model": "gemini-pro"},