fix streaming (#1944)

api/core/model_runtime/model_providers/openai_api_compatible/llm/llm.py

                 )
             )
 
-        for chunk in response.iter_content(chunk_size=2048):
+        for chunk in response.iter_lines(decode_unicode=True, delimiter='\n\n'):
             if chunk:
-                decoded_chunk = chunk.decode('utf-8').strip().lstrip('data: ').lstrip()
+                decoded_chunk = chunk.strip().lstrip('data: ').lstrip()
 
                 chunk_json = None
                 try:
                     continue
 
                 choice = chunk_json['choices'][0]
-                chunk_index = choice['index'] if 'index' in choice else chunk_index
+                chunk_index += 1
 
                 if 'delta' in choice:
                     delta = choice['delta']
                             message=assistant_prompt_message,
                         )
                     )
-            else:
-                yield create_final_llm_result_chunk(
-                    index=chunk_index + 1,
-                    message=AssistantPromptMessage(content=""),
-                    finish_reason="End of stream."
-                )
 
             chunk_index += 1
 

api/tests/integration_tests/model_runtime/openai_api_compatible/test_llm.py

             model='mistralai/Mixtral-8x7B-Instruct-v0.1',
             credentials={
                 'api_key': 'invalid_key',
-                'endpoint_url': 'https://api.together.xyz/v1/chat/completions',
+                'endpoint_url': 'https://api.together.xyz/v1/',
                 'mode': 'chat'
             }
         )
         model='mistralai/Mixtral-8x7B-Instruct-v0.1',
         credentials={
             'api_key': os.environ.get('TOGETHER_API_KEY'),
-            'endpoint_url': 'https://api.together.xyz/v1/chat/completions',
+            'endpoint_url': 'https://api.together.xyz/v1/',
             'mode': 'chat'
         }
     )
         model='mistralai/Mixtral-8x7B-Instruct-v0.1',
         credentials={
             'api_key': os.environ.get('TOGETHER_API_KEY'),
-            'endpoint_url': 'https://api.together.xyz/v1/completions',
+            'endpoint_url': 'https://api.together.xyz/v1/',
             'mode': 'completion'
         },
         prompt_messages=[
         model='mistralai/Mixtral-8x7B-Instruct-v0.1',
         credentials={
             'api_key': os.environ.get('TOGETHER_API_KEY'),
-            'endpoint_url': 'https://api.together.xyz/v1/chat/completions',
+            'endpoint_url': 'https://api.together.xyz/v1/',
             'mode': 'chat'
         },
         prompt_messages=[
         model='gpt-3.5-turbo',
         credentials={
             'api_key': os.environ.get('OPENAI_API_KEY'),
-            'endpoint_url': 'https://api.openai.com/v1/chat/completions',
+            'endpoint_url': 'https://api.openai.com/v1/',
             'mode': 'chat'
         },
         prompt_messages=[
         model='mistralai/Mixtral-8x7B-Instruct-v0.1',
         credentials={
             'api_key': os.environ.get('OPENAI_API_KEY'),
-            'endpoint_url': 'https://api.openai.com/v1/chat/completions'
+            'endpoint_url': 'https://api.openai.com/v1/'
         },
         prompt_messages=[
             SystemPromptMessage(

api/tests/integration_tests/model_runtime/openai_api_compatible/test_text_embedding.py

             model='text-embedding-ada-002',
             credentials={
                 'api_key': 'invalid_key',
-                'endpoint_url': 'https://api.openai.com/v1/embeddings',
-                'context_size': 8184,
-                'max_chunks': 32
+                'endpoint_url': 'https://api.openai.com/v1/',
+                'context_size': 8184
                 
             }
         )
         model='text-embedding-ada-002',
         credentials={
             'api_key': os.environ.get('OPENAI_API_KEY'),
-            'endpoint_url': 'https://api.openai.com/v1/embeddings',
-            'context_size': 8184,
-            'max_chunks': 32
+            'endpoint_url': 'https://api.openai.com/v1/',
+            'context_size': 8184
         }
     )
 
         model='text-embedding-ada-002',
         credentials={
             'api_key': os.environ.get('OPENAI_API_KEY'),
-            'endpoint_url': 'https://api.openai.com/v1/embeddings',
-            'context_size': 8184,
-            'max_chunks': 32
+            'endpoint_url': 'https://api.openai.com/v1/',
+            'context_size': 8184
         },
         texts=[
             "hello",
         credentials={
             'api_key': os.environ.get('OPENAI_API_KEY'),
             'endpoint_url': 'https://api.openai.com/v1/embeddings',
-            'context_size': 8184,
-            'max_chunks': 32
+            'context_size': 8184
         },
         texts=[
             "hello",