Migrate to V2 Pydantic interface (#3262)

Signed-off-by: Emmanuel Ferdman <[email protected]>

Author: emmanuel-ferdman

docs/source/basic_tutorials/using_guidance.md

@@ -138,10 +138,10 @@ client = InferenceClient("http://localhost:3000")
 
 user_input = "I saw a puppy a cat and a raccoon during my bike ride in the park"
 resp = client.text_generation(
-    f"convert to JSON: 'f{user_input}'. please use the following schema: {Animals.schema()}",
+    f"convert to JSON: 'f{user_input}'. please use the following schema: {Animals.model_json_schema()}",
     max_new_tokens=100,
     seed=42,
-    grammar={"type": "json", "value": Animals.schema()},
+    grammar={"type": "json", "value": Animals.model_json_schema()},
 )
 
 print(resp)

integration-tests/models/test_grammar_response_format_llama.py

@@ -34,7 +34,7 @@ class Weather(BaseModel):
         "messages": [
             {
                 "role": "system",
-                "content": f"Respond to the users questions and answer them in the following format: {Weather.schema()}",
+                "content": f"Respond to the users questions and answer them in the following format: {Weather.model_json_schema()}",
             },
             {
                 "role": "user",
@@ -43,7 +43,7 @@ class Weather(BaseModel):
         ],
         "seed": 42,
         "max_tokens": 500,
-        "response_format": {"type": "json_object", "value": Weather.schema()},
+        "response_format": {"type": "json_object", "value": Weather.model_json_schema()},
     }
     # send the request
     response = requests.post(
@@ -75,7 +75,7 @@ class Weather(BaseModel):
 
     json_payload["response_format"] = {
         "type": "json_schema",
-        "value": {"name": "weather", "strict": True, "schema": Weather.schema()},
+        "value": {"name": "weather", "strict": True, "schema": Weather.model_json_schema()},
     }
     response = requests.post(
         f"{llama_grammar.base_url}/v1/chat/completions",
@@ -109,7 +109,7 @@ class Weather(BaseModel):
             "messages": [
                 {
                     "role": "system",
-                    "content": f"Respond to the users questions and answer them in the following format: {Weather.schema()}",
+                    "content": f"Respond to the users questions and answer them in the following format: {Weather.model_json_schema()}",
                 },
                 {
                     "role": "user",
@@ -119,7 +119,7 @@ class Weather(BaseModel):
             "seed": 42,
             "max_tokens": 500,
             "tools": [],
-            "response_format": {"type": "json_object", "value": Weather.schema()},
+            "response_format": {"type": "json_object", "value": Weather.model_json_schema()},
         },
     )