Fails when testing Haiku on AWS Bedrock

[NidPlays]

When testing AWS Bedrock, for Haiku, get this error and fails, testing for higher concurrency values.

Error Details

[2024-11-28 14:55:36,221] p11314 {clientwrap.py:98} WARNING - ERROR:fmbench.scripts.stream_responses:Error occurred while generating and computing metrics associated with the streaming response: litellm.APIConnectionError: Bad response code, expected 200: {'status_code': 400, 'headers': {':exception-type': 'serviceUnavailableException', ':content-type': 'application/json', ':message-type': 'exception'}, 'body': b'{"message":"Bedrock is unable to process your request."}'} Traceback (most recent call last): File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/litellm/utils.py", line 7919, in __next__ chunk = next(self.completion_stream) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/litellm/llms/bedrock/chat/invoke_handler.py", line 1233, in iter_bytes message = self._parse_message_from_event(event) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/litellm/llms/bedrock/chat/invoke_handler.py", line 1259, in _parse_message_from_event raise ValueError(f"Bad response code, expected 200: {response_dict}") ValueError: Bad response code, expected 200: {'status_code': 400, 'headers': {':exception-type': 'serviceUnavailableException', ':content-type': 'application/json', ':message-type': 'exception'}, 'body': b'{"message":"Bedrock is unable to process your request."}'} Traceback (most recent call last): File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/litellm/utils.py", line 7919, in __next__ chunk = next(self.completion_stream) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/litellm/llms/bedrock/chat/invoke_handler.py", line 1233, in iter_bytes message = self._parse_message_from_event(event) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/litellm/llms/bedrock/chat/invoke_handler.py", line 1259, in _parse_message_from_event raise ValueError(f"Bad response code, expected 200: {response_dict}") ValueError: Bad response code, expected 200: {'status_code': 400, 'headers': {':exception-type': 'serviceUnavailableException', ':content-type': 'application/json', ':message-type': 'exception'}, 'body': b'{"message":"Bedrock is unable to process your request."}'}

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/fmbench/scripts/stream_responses.py", line 100, in get_response_stream
for event in event_iterator:
File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/litellm/utils.py", line 8012, in next
raise exception_type(
^^^^^^^^^^^^^^^
File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/litellm/litellm_core_utils/exception_mapping_utils.py", line 2116, in exception_type
raise e
File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/litellm/litellm_core_utils/exception_mapping_utils.py", line 2092, in exception_type
raise APIConnectionError(
litellm.exceptions.APIConnectionError: litellm.APIConnectionError: Bad response code, expected 200: {'status_code': 400, 'headers': {':exception-type': 'serviceUnavailableException', ':content-type': 'application/json', ':message-type': 'exception'}, 'body': b'{"message":"Bedrock is unable to process your request."}'}
Traceback (most recent call last):
File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/litellm/utils.py", line 7919, in next
chunk = next(self.completion_stream)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/litellm/llms/bedrock/chat/invoke_handler.py", line 1233, in iter_bytes
message = self._parse_message_from_event(event)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/ec2-user/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages/litellm/llms/bedrock/chat/invoke_handler.py", line 1259, in _parse_message_from_event
raise ValueError(f"Bad response code, expected 200: {response_dict}")
ValueError: Bad response code, expected 200: {'status_code': 400, 'headers': {':exception-type': 'serviceUnavailableException', ':content-type': 'application/json', ':message-type': 'exception'}, 'body': b'{"message":"Bedrock is unable to process your request."}'}

[2024-11-28 14:55:36,222] p11314 {clientwrap.py:98} WARNING - INFO:fmbench.scripts.stream_responses:Final result: None

[2024-11-28 14:55:36,225] p11314 {clientwrap.py:98} WARNING - ERROR:custom_bedrock_predictor:Unexpected error during prediction, endpoint_name=eu.anthropic.claude-3-haiku-20240307-v1:0, exception='NoneType' object has no attribute 'get'

[2024-11-28 14:55:36,552] p11314 {clientwrap.py:91} INFO -
Give Feedback / Get Help: https://github.com/BerriAI/litellm/issues/new

also want to know if it is possible to fail the request when we reach the rate limit instead of retrying