github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-08 00:51:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #523] LLM falcon:text infinity loop #243

New Issue
Closed
opened 2026-04-12 09:45:52 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-04-12 09:45:52 -05:00
Owner
Copy Link

Originally created by @dcasota on GitHub (Sep 13, 2023).
Original GitHub issue: https://github.com/ollama/ollama/issues/523

Hi,

I was trying to run falcon, but it responds ... weired.

Setup recipe.

git clone https://github.com/jmorganca/ollama
cd .\ollama
mkdir ..\.ollama
go generate .\...
go build .

Then, start the server component of ollama.
start "Ollama server component" ollama.exe serve

Download the selected model.
ollama.exe pull falcon:text

Run the model.
ollama.exe run falcon:text

I've started a conversation with a simple "Hi".
The output started with listing date.getDay(); but it didn't stop, not at 7, not at 31, and not at 365...seems an infinity loop.

image

After ctrl-c, and restarting the 2nd and 3rd time a simple Hi, it doesn't list anymore getDay, but responds with extract from letters someone wrote.

What is the purpose of falcon:text?

Originally created by @dcasota on GitHub (Sep 13, 2023). Original GitHub issue: https://github.com/ollama/ollama/issues/523 Hi, I was trying to run falcon, but it responds ... weired. Setup recipe. ``` git clone https://github.com/jmorganca/ollama cd .\ollama mkdir ..\.ollama go generate .\... go build . ``` Then, start the server component of ollama. `start "Ollama server component" ollama.exe serve` Download the selected model. `ollama.exe pull falcon:text` Run the model. `ollama.exe run falcon:text` I've started a conversation with a simple "Hi". The output started with listing `date.getDay();` but it didn't stop, not at 7, not at 31, and not at 365...seems an infinity loop. <img src="https://github.com/jmorganca/ollama/assets/14890243/59f54411-f58d-4f9e-97ef-bce53cb6fedc" alt="image" width="200"> After ctrl-c, and restarting the 2nd and 3rd time a simple Hi, it doesn't list anymore getDay, but responds with extract from letters someone wrote. What is the purpose of falcon:text?
GiteaMirror commented 2026-04-12 09:45:53 -05:00
Author
Owner
Copy Link

@BruceMacD commented on GitHub (Sep 13, 2023):

Hi @dcasota, the falcon:text model is a raw model that is meant for fine tuning. Quoting from the original authors:

This is a raw, pretrained model, which should be further finetuned for most usecases. If you are looking for a version better suited to taking generic instructions in a chat format, we recommend taking a look at Falcon-7B-Instruct.

Try this instead, to run the instruct/chat Falcon model: ollama run falcon

<!-- gh-comment-id:1717687564 --> @BruceMacD commented on GitHub (Sep 13, 2023): Hi @dcasota, the falcon:text model is a raw model that is meant for fine tuning. Quoting from the original authors: > This is a raw, pretrained model, which should be further finetuned for most usecases. If you are looking for a version better suited to taking generic instructions in a chat format, we recommend taking a look at Falcon-7B-Instruct. Try this instead, to run the instruct/chat Falcon model: `ollama run falcon`
GiteaMirror commented 2026-04-12 09:45:54 -05:00
Author
Owner
Copy Link

@dcasota commented on GitHub (Sep 13, 2023):

@BruceMacD I switched to llama2:70b for langchain-document purposes. The response quality in English is very good.

Because of the wording "text" I though that the falcon LLM is more predestinated for answer/questions constellations than other LLMs. Imho it is not.

Thank you for the quick reply and clarification. As a beginner, it is difficult to distinct between "good" and "not good" LLMs for a specific purpose. You're right. It is necessary to read the LLM thesis and answers to choose wisely.

<!-- gh-comment-id:1717728077 --> @dcasota commented on GitHub (Sep 13, 2023): @BruceMacD I switched to llama2:70b for langchain-document purposes. The response quality in English is very good. Because of the wording "text" I though that the falcon LLM is more predestinated for answer/questions constellations than other LLMs. Imho it is not. Thank you for the quick reply and clarification. As a beginner, it is difficult to distinct between "good" and "not good" LLMs for a specific purpose. You're right. It is necessary to read the LLM thesis and answers to choose wisely.
GiteaMirror commented 2026-04-12 09:45:54 -05:00
Author
Owner
Copy Link

@BruceMacD commented on GitHub (Sep 13, 2023):

No worries, it is really confusing. Thanks for opening the issue!

<!-- gh-comment-id:1717754648 --> @BruceMacD commented on GitHub (Sep 13, 2023): No worries, it is really confusing. Thanks for opening the issue!
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#243
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?