[GH-ISSUE #561] Unexpected EOF with Falcon:40b #256

Closed
opened 2026-04-12 09:46:53 -05:00 by GiteaMirror · 4 comments
Owner

Originally created by @henry-prince-addepar on GitHub (Sep 20, 2023).
Original GitHub issue: https://github.com/ollama/ollama/issues/561

I'm getting an error from falcon:40b. Any help would be greatly appreciated. I'm currently running MacOS 13.5.2 (22G91) on a M1 Max with 32 GB of RAM. Thanks in advance!

➜  ~ ollama pull falcon:40b
pulling manifest
pulling a4a6e73500b0... 100% |██████████████████████████████████████████████████████████████████████████████████████████| (24/24 GB, 12 TB/s)
pulling d5311aab7c4c... 100% |██████████████████████████████████████████████████████████████████████████████████████████| (84/84 B, 103 kB/s)
pulling 0740207dce29... 100% |████████████████████████████████████████████████████████████████████████████████████████| (307/307 B, 3.9 MB/s)
verifying sha256 digest
writing manifest
removing any unused layers
success
➜  ~ ollama run falcon:40b
>>> Why is the sky blue?
Error: error reading llm response: unexpected EOF
➜  ~ ollama run falcon:40b
>>> Hi. This is a test.
Error: error reading llm response: unexpected EOF
Originally created by @henry-prince-addepar on GitHub (Sep 20, 2023). Original GitHub issue: https://github.com/ollama/ollama/issues/561 I'm getting an error from `falcon:40b`. Any help would be greatly appreciated. I'm currently running MacOS 13.5.2 (22G91) on a M1 Max with 32 GB of RAM. Thanks in advance! ``` ➜ ~ ollama pull falcon:40b pulling manifest pulling a4a6e73500b0... 100% |██████████████████████████████████████████████████████████████████████████████████████████| (24/24 GB, 12 TB/s) pulling d5311aab7c4c... 100% |██████████████████████████████████████████████████████████████████████████████████████████| (84/84 B, 103 kB/s) pulling 0740207dce29... 100% |████████████████████████████████████████████████████████████████████████████████████████| (307/307 B, 3.9 MB/s) verifying sha256 digest writing manifest removing any unused layers success ➜ ~ ollama run falcon:40b >>> Why is the sky blue? Error: error reading llm response: unexpected EOF ➜ ~ ollama run falcon:40b >>> Hi. This is a test. Error: error reading llm response: unexpected EOF ```
Author
Owner

@mxyng commented on GitHub (Sep 21, 2023):

Unfortunately it seems 32GB is not enough for falcon:40B. It doesn't fit into the allocated buffer for metal.

<!-- gh-comment-id:1730328581 --> @mxyng commented on GitHub (Sep 21, 2023): Unfortunately it seems 32GB is not enough for falcon:40B. It doesn't fit into the allocated buffer for metal.
Author
Owner

@henry-prince-addepar commented on GitHub (Sep 23, 2023):

Thanks for your insight on this, @mxyng. I'll go ahead and close this issue.

<!-- gh-comment-id:1732384277 --> @henry-prince-addepar commented on GitHub (Sep 23, 2023): Thanks for your insight on this, @mxyng. I'll go ahead and close this issue.
Author
Owner

@happy15 commented on GitHub (Sep 26, 2023):

same for vicuna:7b-16k on mbp 14:

(base) ➜  ~ ollama run vicuna:7b-16k
>>> hi
Error: error reading llm response: unexpected EOF
<!-- gh-comment-id:1734764394 --> @happy15 commented on GitHub (Sep 26, 2023): same for `vicuna:7b-16k` on mbp 14: ``` (base) ➜ ~ ollama run vicuna:7b-16k >>> hi Error: error reading llm response: unexpected EOF ```
Author
Owner

@yuis-ice commented on GitHub (Oct 22, 2023):

I get the Error: error reading llm response: unexpected EOF error message so often.

<!-- gh-comment-id:1774006966 --> @yuis-ice commented on GitHub (Oct 22, 2023): I get the `Error: error reading llm response: unexpected EOF` error message so often.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#256