github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 00:22:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #5663] Error: llama runner process has terminated: signal: abort trap error:vocab size mismatch. #50043

New Issue
Closed
opened 2026-04-28 13:55:55 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-04-28 13:55:55 -05:00
Owner
Copy Link

Originally created by @asap-blocky on GitHub (Jul 13, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/5663

What is the issue?

While attempting to run my fine tuned model using the Ollama library, I got this error message, "Error: llama runner process has terminated: signal: abort trap error:vocab size mismatch."

Model and Environment:

  • The model was fine-tuned using the FastLanguageModel from the unsloth library and saved in the GGUF format.
  • The tokenizer was applied using a chat template for formatting inputs.
  • The model and tokenizer were loaded correctly, and the inference process was initiated.

Error Occurrence:

  • The error occurs immediately after issuing the ollama run model-name command.
  • The detailed logs indicate a vocabulary size mismatch between the model and the tokenizer.

Model Metadata:

  • The model’s configuration (config.json) indicates a vocabulary size of 32064.
  • The tokenizer configuration (tokenizer.json) and metadata logs show different values, leading to the mismatch.

Steps Taken:

  1. Verified the consistency of vocabulary size across all relevant configuration files.
  2. Attempted to resize token embeddings to match the tokenizer’s vocabulary size during model loading.
  3. Checked for any missing or additional special tokens in the tokenizer configuration.

OS

macOS

GPU

Apple

CPU

Apple

Ollama version

ollama version is 0.2.2 Warning: client version is 0.2.1

Originally created by @asap-blocky on GitHub (Jul 13, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/5663 ### What is the issue? While attempting to run my fine tuned model using the Ollama library, I got this error message, "Error: llama runner process has terminated: signal: abort trap error:vocab size mismatch." ### Model and Environment: - The model was fine-tuned using the FastLanguageModel from the unsloth library and saved in the GGUF format. - The tokenizer was applied using a chat template for formatting inputs. - The model and tokenizer were loaded correctly, and the inference process was initiated. ### Error Occurrence: - The error occurs immediately after issuing the **ollama run model-name** command. - The detailed logs indicate a vocabulary size mismatch between the model and the tokenizer. ### Model Metadata: - The model’s configuration (config.json) indicates a vocabulary size of 32064. - The tokenizer configuration (tokenizer.json) and metadata logs show different values, leading to the mismatch. ### Steps Taken: 1. Verified the consistency of vocabulary size across all relevant configuration files. 2. Attempted to resize token embeddings to match the tokenizer’s vocabulary size during model loading. 3. Checked for any missing or additional special tokens in the tokenizer configuration. ### OS macOS ### GPU Apple ### CPU Apple ### Ollama version ollama version is 0.2.2 Warning: client version is 0.2.1
GiteaMirror added the bug label 2026-04-28 13:55:55 -05:00
GiteaMirror commented 2026-04-28 13:55:57 -05:00
Author
Owner
Copy Link

@ramenw4ve commented on GitHub (Aug 4, 2024):

so how did you fix it?

<!-- gh-comment-id:2267437570 --> @ramenw4ve commented on GitHub (Aug 4, 2024): so how did you fix it?
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 bug
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#50043
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?