llama : adjust default context size + print warnings #10136
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
ngxson
approved these changes
Nov 2, 2024
There was a problem hiding this comment.
Thanks! This should prevent me from burning my swapfile whenever I forget to specify -c
Tested and it shows the log too:
> ./llama-cli -m ../models/Meta-Llama-3.1-8B-Instruct-Q4_K_M.gguf -cnv -p "You are a helpful assistant"
...
llama_new_context_with_model: n_ctx_per_seq (4096) < n_ctx_train (131072) -- the full capacity of the model will not be utilized
...
github-actions
bot
added
the
devops
slaren
approved these changes
Nov 2, 2024
|
Is 4096 a good value, or should we go lower? |
|
According to HF hub statistics, the most used model nowadays is the llama 3 (3.1, 3.2) 8B With a context size of 4096, the KV takes around 512MB which I think is a very reasonable amount. |
arthw
pushed a commit
to arthw/llama.cpp
that referenced
this pull request
Nov 15, 2024
* llama : adjust default context size + print warnings ggml-ci * ggml-ci : add missing gpu-layers + adjust context sizes
arthw
pushed a commit
to arthw/llama.cpp
that referenced
this pull request
Nov 18, 2024
* llama : adjust default context size + print warnings ggml-ci * ggml-ci : add missing gpu-layers + adjust context sizes
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
fix #8817, #9563 (comment)
By default, the examples will use a context size of 4096, instead of the training context of the model. In a lot of cases, the default training context can be very big - 32k to 128k tokens, which causes enormous KV cache allocation and failures for regular hardware.
Also, add warning logs when the specified context size per sequence does not match the training context.