github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-22 21:51:37 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #7376] Is there a way to track tokens/context window in real-time? #51200

New Issue
Closed
opened 2026-04-28 18:54:32 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-28 18:54:32 -05:00
Owner
Copy Link

Originally created by @robotom on GitHub (Oct 26, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/7376

I'd like to implement a counter in a front end app to track the tokens used in order to see if I'm close to exceeding the context window.

This is useful to me because if I feed a large document into the model, I'd like to know when it's "too large" and perhaps to break it down or do something else.

Originally created by @robotom on GitHub (Oct 26, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/7376 I'd like to implement a counter in a front end app to track the tokens used in order to see if I'm close to exceeding the context window. This is useful to me because if I feed a large document into the model, I'd like to know when it's "too large" and perhaps to break it down or do something else.
GiteaMirror added the feature request label 2026-04-28 18:54:32 -05:00
GiteaMirror commented 2026-04-28 18:54:34 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Oct 26, 2024):

prompt_eval_count, returned in the response, counts the tokens in the request. There's no way to incrementally update a prompt. There's an open issue, https://github.com/ollama/ollama/issues/3582, to add a tokenizer endpoint that would allow a client to count tokens before sending a completion.

<!-- gh-comment-id:2439730096 --> @rick-github commented on GitHub (Oct 26, 2024): [prompt_eval_count](https://github.com/ollama/ollama/blob/main/docs/api.md#generate-a-completion:~:text=loading%20the%20model-,prompt_eval_count), returned in the response, counts the tokens in the request. There's no way to incrementally update a prompt. There's an open issue, https://github.com/ollama/ollama/issues/3582, to add a tokenizer endpoint that would allow a client to count tokens before sending a completion.
GiteaMirror commented 2026-04-28 18:54:34 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Oct 26, 2024):

Note that if your model was derived from one on HuggingFace, you can count the tokens in the client:

#!/usr/bin/env python3

from transformers import AutoTokenizer

model = "CohereForAI/aya-expanse-8b"
messages = [ { "role":"user", "content":"hello world" } ]

autotokenizer = AutoTokenizer.from_pretrained(model)
autotokens = autotokenizer.apply_chat_template(messages)

print(autotokens)
print(len(autotokens))
print(autotokenizer.decode(autotokens))

$ ./7376.py 
[5, 255000, 255006, 34313, 3845, 255001]
6
<BOS_TOKEN><|START_OF_TURN_TOKEN|><|USER_TOKEN|>hello world<|END_OF_TURN_TOKEN|>

Be aware that the template applied by ollama may be different to the template in the HuggingFace model config, so the token count between AutoTokenizer and ollama may differ slightly.

<!-- gh-comment-id:2439749060 --> @rick-github commented on GitHub (Oct 26, 2024): Note that if your model was derived from one on HuggingFace, you can count the tokens in the client: ```python #!/usr/bin/env python3 from transformers import AutoTokenizer model = "CohereForAI/aya-expanse-8b" messages = [ { "role":"user", "content":"hello world" } ] autotokenizer = AutoTokenizer.from_pretrained(model) autotokens = autotokenizer.apply_chat_template(messages) print(autotokens) print(len(autotokens)) print(autotokenizer.decode(autotokens)) ``` ```console $ ./7376.py [5, 255000, 255006, 34313, 3845, 255001] 6 <BOS_TOKEN><|START_OF_TURN_TOKEN|><|USER_TOKEN|>hello world<|END_OF_TURN_TOKEN|> ``` Be aware that the template applied by ollama may be different to the template in the HuggingFace model config, so the token count between AutoTokenizer and ollama may differ slightly.
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 feature request
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#51200
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?