src : add n_expert_used_max function - #28323
Merged
Merged
Conversation
With Commit c61b98b ("model: add NVIDIA Nemotron-3-Puzzle-75B-A9B (NemotronHPuzzle) support (ggml-org#25444)") it is now possible for each layer to have a specific number of experts but there are a few checks that need to be updated to handle this upon model loading. For example: ```console llama_model_load: error loading model: model has expert layers but no expert layers are used ``` And later: ```console /llama.cpp/src/llama-model-loader.cpp:955: GGML_ASSERT(n_ids_used > 0) failed ``` This commit adds the n_expert_used_max function so that these checks can use it. Refs: ggml-org#25444 (comment)
Contributor
|
Fixes the issue on my end! |
CISC
approved these changes
Sep 3, 2026
Contributor
|
Still works on my end, just tested and it fixes the nemotron issue |
bartowski1182
approved these changes
Sep 3, 2026
Contributor
|
any concerns with merging @CISC ? |
Member
Nope. 😆 |
MarkShark2
added a commit
to MarkShark2/llama.cpp
that referenced
this pull request
Sep 5, 2026
64 upstream commits. The bulk of the 57 conflicted files is one seam: upstream adopted the per-layer n_ff_exp_arr / n_expert_used_arr accessors and dropped the scalar _impl fallbacks (ggml-org#28323), so those hunks take upstream's side. Hand-resolved: the DSV4 media routing (upstream's exp_probs_b_vl path kept, the fork's routing_id hash gather kept for palette batches), build_attn_mha's new n_kv_max argument in the fork's sparse-attention else branch, the loader's parallel RPC population kept under upstream's non-host-first ordering, the Puzzle converter keeping its MTP export plus upstream's name normalization. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01PtrnaBuYDHGDvy73TRkFbG
x1250
pushed a commit
to x1250/llama.cpp
that referenced
this pull request
Sep 9, 2026
* src : add n_expert_used_max function With Commit c61b98b ("model: add NVIDIA Nemotron-3-Puzzle-75B-A9B (NemotronHPuzzle) support (ggml-org#25444)") it is now possible for each layer to have a specific number of experts but there are a few checks that need to be updated to handle this upon model loading. For example: ```console llama_model_load: error loading model: model has expert layers but no expert layers are used ``` And later: ```console /llama.cpp/src/llama-model-loader.cpp:955: GGML_ASSERT(n_ids_used > 0) failed ``` This commit adds the n_expert_used_max function so that these checks can use it. Refs: ggml-org#25444 (comment) * src : use hparams.n_expert_used_max in llama_model_base::load_hparams * src : use 0 as initial value for n_expert_used_max
zbrad
pushed a commit
to zbrad/llama.cpp
that referenced
this pull request
Sep 10, 2026
* src : add n_expert_used_max function With Commit 3615d1b ("model: add NVIDIA Nemotron-3-Puzzle-75B-A9B (NemotronHPuzzle) support (ggml-org#25444)") it is now possible for each layer to have a specific number of experts but there are a few checks that need to be updated to handle this upon model loading. For example: ```console llama_model_load: error loading model: model has expert layers but no expert layers are used ``` And later: ```console /llama.cpp/src/llama-model-loader.cpp:955: GGML_ASSERT(n_ids_used > 0) failed ``` This commit adds the n_expert_used_max function so that these checks can use it. Refs: ggml-org#25444 (comment) * src : use hparams.n_expert_used_max in llama_model_base::load_hparams * src : use 0 as initial value for n_expert_used_max
pl752
pushed a commit
to pl752/llama.cpp
that referenced
this pull request
Sep 15, 2026
* src : add n_expert_used_max function With Commit c61b98b ("model: add NVIDIA Nemotron-3-Puzzle-75B-A9B (NemotronHPuzzle) support (ggml-org#25444)") it is now possible for each layer to have a specific number of experts but there are a few checks that need to be updated to handle this upon model loading. For example: ```console llama_model_load: error loading model: model has expert layers but no expert layers are used ``` And later: ```console /llama.cpp/src/llama-model-loader.cpp:955: GGML_ASSERT(n_ids_used > 0) failed ``` This commit adds the n_expert_used_max function so that these checks can use it. Refs: ggml-org#25444 (comment) * src : use hparams.n_expert_used_max in llama_model_base::load_hparams * src : use 0 as initial value for n_expert_used_max
zsogitbe
pushed a commit
to zsogitbe/llama.cpp
that referenced
this pull request
Sep 17, 2026
* src : add n_expert_used_max function With Commit c61b98b ("model: add NVIDIA Nemotron-3-Puzzle-75B-A9B (NemotronHPuzzle) support (ggml-org#25444)") it is now possible for each layer to have a specific number of experts but there are a few checks that need to be updated to handle this upon model loading. For example: ```console llama_model_load: error loading model: model has expert layers but no expert layers are used ``` And later: ```console /llama.cpp/src/llama-model-loader.cpp:955: GGML_ASSERT(n_ids_used > 0) failed ``` This commit adds the n_expert_used_max function so that these checks can use it. Refs: ggml-org#25444 (comment) * src : use hparams.n_expert_used_max in llama_model_base::load_hparams * src : use 0 as initial value for n_expert_used_max
Te-eMster
pushed a commit
to Te-eMster/mx-llama.cpp
that referenced
this pull request
Sep 18, 2026
* src : add n_expert_used_max function With Commit c61b98b ("model: add NVIDIA Nemotron-3-Puzzle-75B-A9B (NemotronHPuzzle) support (ggml-org#25444)") it is now possible for each layer to have a specific number of experts but there are a few checks that need to be updated to handle this upon model loading. For example: ```console llama_model_load: error loading model: model has expert layers but no expert layers are used ``` And later: ```console /llama.cpp/src/llama-model-loader.cpp:955: GGML_ASSERT(n_ids_used > 0) failed ``` This commit adds the n_expert_used_max function so that these checks can use it. Refs: ggml-org#25444 (comment) * src : use hparams.n_expert_used_max in llama_model_base::load_hparams * src : use 0 as initial value for n_expert_used_max
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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.
Overview
With Commit c61b98b ("model: add NVIDIA Nemotron-3-Puzzle-75B-A9B (NemotronHPuzzle) support (#25444)") it is now possible for each layer to have a specific number of experts but there are a few checks that need to be updated to handle this upon model loading. For example:
llama_model_load: error loading model: model has expert layers but no expert layers are usedAnd later:
/llama.cpp/src/llama-model-loader.cpp:955: GGML_ASSERT(n_ids_used > 0) failedThis commit adds the n_expert_used_max function so that these checks can use it.
Additional information
Refs: #25444 (comment)
Requirements