exo
exo copied to clipboard
Shard is not defined
I have started using Exo with 2 Mac Studios. It identifies both computers (they are connected using thunderbolt). However, when I send a request using the tinychat UI, the model does not return the response.
I have run it using:
DEBUG=9 exo
I have checked the logs and I found these errors:
[22afe857-180d-48a8-8d9c-72b8af4d0ad0] result size: 88064, is finished: False, buffered tokens: 0
Next partition index: 1
Task exception was never retrieved
future: <Task finished name='Task-152' coro=<StandardNode.forward_to_next_shard() done, defined at
/Users/llmserver/Documents/Projects/exo/exo/orchestration/standard_node.py:273> exception=NameError("name 'shard' is not
defined")>
Traceback (most recent call last):
File "/Users/llmserver/Documents/Projects/exo/exo/orchestration/standard_node.py", line 289, in forward_to_next_shard
if DEBUG >= 2: print(f"Computed next from: {shard}, {self.topology}. Next partition: {next_partition}")
NameError: name 'shard' is not defined. Did you mean: 'Shard'?
I have used llama3.1 - 1B for testing.
Thank you.