convert: allow using quantized Mistral weight #17889
Merged
+24
−4
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
Collaborator
Traceback (most recent call last):
File "/devel/tools/llama.cpp/convert_hf_to_gguf.py", line 10635, in <module>
main()
~~~~^^
File "/devel/tools/llama.cpp/convert_hf_to_gguf.py", line 10629, in main
model_instance.write()
~~~~~~~~~~~~~~~~~~~~^^
File "/devel/tools/llama.cpp/convert_hf_to_gguf.py", line 660, in write
self.prepare_tensors()
~~~~~~~~~~~~~~~~~~~~^^
File "/devel/tools/llama.cpp/convert_hf_to_gguf.py", line 2540, in prepare_tensors
super().prepare_tensors()
~~~~~~~~~~~~~~~~~~~~~~~^^
File "/devel/tools/llama.cpp/convert_hf_to_gguf.py", line 531, in prepare_tensors
for new_name, data_torch in (self.modify_tensors(data_torch, name, bid)):
~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^
File "/devel/tools/llama.cpp/convert_hf_to_gguf.py", line 2862, in modify_tensors
assert data_torch.nelement() == 0 # unused by the model
^^^^^^^^^^^^^^^^^^^^^^^^^^
AssertionError |
This comment was marked as resolved.
This comment was marked as resolved.
Collaborator
Author
|
should work now with |
Co-authored-by: compilade <compilade@users.noreply.github.com>
Contributor
|
confirming this works :) https://huggingface.co/bartowski/mistralai_Devstral-Small-2-24B-Instruct-2512-GGUF LGTM! |
CISC
approved these changes
Dec 9, 2025
Ethan-a2
pushed a commit
to Ethan-a2/llama.cpp
that referenced
this pull request
Dec 12, 2025
* convert: allow using quantized Mistral weight * data_torch.ndim * update dequant fn Co-authored-by: compilade <compilade@users.noreply.github.com> --------- Co-authored-by: compilade <compilade@users.noreply.github.com>
0Marble
pushed a commit
to 0Marble/llama.cpp
that referenced
this pull request
Dec 18, 2025
* convert: allow using quantized Mistral weight * data_torch.ndim * update dequant fn Co-authored-by: compilade <compilade@users.noreply.github.com> --------- Co-authored-by: compilade <compilade@users.noreply.github.com>
Anico2
added a commit
to Anico2/llama.cpp
that referenced
this pull request
Jan 15, 2026
* convert: allow using quantized Mistral weight * data_torch.ndim * update dequant fn Co-authored-by: compilade <compilade@users.noreply.github.com> --------- Co-authored-by: compilade <compilade@users.noreply.github.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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.
target model: https://huggingface.co/mistralai/Devstral-Small-2-24B-Instruct-2512
needif you have problem with tokenizer conversion, install--mistral-format, otherwise it has problem with tokenizertransformersversion 5.0.0