Agent groupchat stuck at one of the steps

[f-rower]

Agent groupchat gets stuck at a certain point and no progress is made anymore. In the example above, no output is being generated by BusinessObjectiveAgent after 40 minutes.

Next speaker: ConversationDelegationAgent

USING AUTO REPLY...
[autogen.oai.client: 07-21 16:52:43] {708} WARNING - Model llama3.2:latest is not found. The cost will be 0. In your config_list, add field {"price" : [prompt_price_per_1k, completion_token_price_per_1k]} for customized pricing.
[autogen.oai.client: 07-21 16:52:53] {708} WARNING - Model llama3.2:latest is not found. The cost will be 0. In your config_list, add field {"price" : [prompt_price_per_1k, completion_token_price_per_1k]} for customized pricing.
***** raw Memory *****
{"PROPOSAL_COUNT": 3, "DISCUSSION_ROUND": 2, "CONCENSUS_COUNT": 0, "final_output_obtained": false}

***** loaded Memory *****
illegal format
ConversationDelegationAgent (to chat_manager):

Consensus Step: Next speaker should reach an agreement with out speakers and finalise design.

Please proceed with reaching an agreement on the updated pipeline design, considering the evaluation results and recommendations.

---

Next speaker: BusinessObjectiveAgent

USING AUTO REPLY...
[autogen.oai.client: 07-21 17:03:42] {708} WARNING - Model llama3.2:latest is not found. The cost will be 0. In your config_list, add field {"price" : [prompt_price_per_1k, completion_token_price_per_1k]} for customized pricing.

[f-rower]

Could this be because we have reached the maximum number of rounds, which is currently set at 80?

[12yuens2]

Sounds like something to do with the illegal format which makes it never terminate. I'd guess output the memory of each agent and check its the right format each turn.