wx44wx/pipecat
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Receiving the answer

Browse Source
This commit is contained in:
Filipi Fuchter
2025-11-28 15:16:24 -03:00
parent e1def85747
commit 73b9bbaad2
2 changed files with 13 additions and 73 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
examples/aws-agentcore/agents/pipecat/pipecat-agent.py
View File

@@ -60,27 +60,7 @@ async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
# Returning the answer
if isinstance(transport, SmallWebRTCTransport):
yield {"status": "ANSWER:START"}
answer_dict = transport._client._webrtc_connection.get_answer()
answer_json_str = json.dumps(answer_dict)
answer_bytes = answer_json_str.encode("utf-8")
encoded_answer = base64.b64encode(answer_bytes).decode("ascii")
# Break encoded_answer into multiple messages
chunk_size = 1000 # Adjust this size based on your needs
total_chunks = (
len(encoded_answer) + chunk_size - 1
) // chunk_size # Calculate total chunks
for i in range(0, len(encoded_answer), chunk_size):
chunk = encoded_answer[i : i + chunk_size]
chunk_index = i // chunk_size
yield {
"answer_chunk": chunk,
"chunk_index": chunk_index,
"total_chunks": total_chunks,
"is_last_chunk": chunk_index == total_chunks - 1,
}
yield {"answer": transport._client._webrtc_connection.get_answer()}
yield {"status": "ANSWER:END"}
stt = DeepgramSTTService(api_key=os.getenv("DEEPGRAM_API_KEY"))