RemoteProtocolError after streaming from Azure Deployement

[BrokenDuck]

SDK Version: 5.13.11

** Bug Summary **

When streaming from a Command R or Command R+ endpoint deployed on Azure, after all the content-delta chunks of the streaming response are received, httpx errors with RemoteProtocolError.

** Buggy code **

import cohere

co = cohere.ClientV2(api_key=os.getenv('COMMAND_R_API_KEY'), base_url=os.getenv('COMMAND_R_URL'))

res = co.chat_stream(
    model="command-r-plus-08-2024",
    messages=[{"role": "user", "content": "What is an LLM?"}],
)

for event in res:
    if event:
        if event.type == "content-delta":
            print(event.delta.message.content.text, end="")

** Detailed description **

During the code execution all the content chunks are returned properly and printed out. Afterwards, httpx return a RemoteProtocolError. I am using server deployments on Azure AI Foundry.

** Full stack trace **

---------------------------------------------------------------------------
RemoteProtocolError                       Traceback (most recent call last)
File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpx/_transports/default.py:101, in map_httpcore_exceptions()
    100 try:
--> 101     yield
    102 except Exception as exc:

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpx/_transports/default.py:127, in ResponseStream.__iter__(self)
    126 with map_httpcore_exceptions():
--> 127     for part in self._httpcore_stream:
    128         yield part

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpcore/_sync/connection_pool.py:407, in PoolByteStream.__iter__(self)
    406 self.close()
--> 407 raise exc from None

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpcore/_sync/connection_pool.py:403, in PoolByteStream.__iter__(self)
    402 try:
--> 403     for part in self._stream:
    404         yield part

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpcore/_sync/http11.py:342, in HTTP11ConnectionByteStream.__iter__(self)
    341     self.close()
--> 342 raise exc

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpcore/_sync/http11.py:334, in HTTP11ConnectionByteStream.__iter__(self)
    333 with Trace("receive_response_body", logger, self._request, kwargs):
--> 334     for chunk in self._connection._receive_response_body(**kwargs):
    335         yield chunk

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpcore/_sync/http11.py:203, in HTTP11Connection._receive_response_body(self, request)
    202 while True:
--> 203     event = self._receive_event(timeout=timeout)
    204     if isinstance(event, h11.Data):

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpcore/_sync/http11.py:213, in HTTP11Connection._receive_event(self, timeout)
    212 while True:
--> 213     with map_exceptions({h11.RemoteProtocolError: RemoteProtocolError}):
    214         event = self._h11_state.next_event()

File ~/.pyenv/versions/3.12.8/lib/python3.12/contextlib.py:158, in _GeneratorContextManager.__exit__(self, typ, value, traceback)
    157 try:
--> 158     self.gen.throw(value)
    159 except StopIteration as exc:
    160     # Suppress StopIteration *unless* it's the same exception that
    161     # was passed to throw().  This prevents a StopIteration
    162     # raised inside the "with" statement from being suppressed.

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpcore/_exceptions.py:14, in map_exceptions(map)
     13     if isinstance(exc, from_exc):
---> 14         raise to_exc(exc) from exc
     15 raise

RemoteProtocolError: peer closed connection without sending complete message body (incomplete chunked read)

The above exception was the direct cause of the following exception:

RemoteProtocolError                       Traceback (most recent call last)
Cell In[5], line 10
      3 co = cohere.ClientV2(api_key=os.getenv('COMMAND_R_API_KEY'), base_url=os.getenv('COMMAND_R_URL'))
      5 res = co.chat_stream(
      6     model="command-r-plus-08-2024",
      7     messages=[{"role": "user", "content": "What is an LLM?"}],
      8 )
---> 10 for event in res:
     11     if event:
     12         if event.type == "content-delta":

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/cohere/v2/client.py:252, in V2Client.chat_stream(self, model, messages, tools, strict_tools, documents, citation_options, response_format, safety_mode, max_tokens, stop_sequences, temperature, seed, frequency_penalty, presence_penalty, k, p, return_prompt, logprobs, tool_choice, request_options)
    250 if 200 <= _response.status_code < 300:
    251     _event_source = httpx_sse.EventSource(_response)
--> 252     for _sse in _event_source.iter_sse():
    253         try:
    254             yield typing.cast(
    255                 StreamedChatResponseV2,
    256                 construct_type(
   (...)
    259                 ),
    260             )

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpx_sse/_api.py:30, in EventSource.iter_sse(self)
     28 self._check_content_type()
     29 decoder = SSEDecoder()
---> 30 for line in self._response.iter_lines():
     31     line = line.rstrip("\n")
     32     sse = decoder.decode(line)

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpx/_models.py:929, in Response.iter_lines(self)
    927 decoder = LineDecoder()
    928 with request_context(request=self._request):
--> 929     for text in self.iter_text():
    930         for line in decoder.decode(text):
    931             yield line

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpx/_models.py:916, in Response.iter_text(self, chunk_size)
    914 chunker = TextChunker(chunk_size=chunk_size)
    915 with request_context(request=self._request):
--> 916     for byte_content in self.iter_bytes():
    917         text_content = decoder.decode(byte_content)
    918         for chunk in chunker.decode(text_content):

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpx/_models.py:897, in Response.iter_bytes(self, chunk_size)
    895 chunker = ByteChunker(chunk_size=chunk_size)
    896 with request_context(request=self._request):
--> 897     for raw_bytes in self.iter_raw():
    898         decoded = decoder.decode(raw_bytes)
    899         for chunk in chunker.decode(decoded):

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpx/_models.py:951, in Response.iter_raw(self, chunk_size)
    948 chunker = ByteChunker(chunk_size=chunk_size)
    950 with request_context(request=self._request):
--> 951     for raw_stream_bytes in self.stream:
    952         self._num_bytes_downloaded += len(raw_stream_bytes)
    953         for chunk in chunker.decode(raw_stream_bytes):

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpx/_client.py:153, in BoundSyncStream.__iter__(self)
    152 def __iter__(self) -> typing.Iterator[bytes]:
--> 153     for chunk in self._stream:
    154         yield chunk

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpx/_transports/default.py:126, in ResponseStream.__iter__(self)
    125 def __iter__(self) -> typing.Iterator[bytes]:
--> 126     with map_httpcore_exceptions():
    127         for part in self._httpcore_stream:
    128             yield part

File ~/.pyenv/versions/3.12.8/lib/python3.12/contextlib.py:158, in _GeneratorContextManager.__exit__(self, typ, value, traceback)
    156     value = typ()
    157 try:
--> 158     self.gen.throw(value)
    159 except StopIteration as exc:
    160     # Suppress StopIteration *unless* it's the same exception that
    161     # was passed to throw().  This prevents a StopIteration
    162     # raised inside the "with" statement from being suppressed.
    163     return exc is not value

File ~/.pyenv/versions/FullStack-RAG-Env/lib/python3.12/site-packages/httpx/_transports/default.py:118, in map_httpcore_exceptions()
    115     raise
    117 message = str(exc)
--> 118 raise mapped_exc(message) from exc

RemoteProtocolError: peer closed connection without sending complete message body (incomplete chunked read)