Search results

Misc. bug: WebUI, first application of default settings (JSON)

bugSomething isn't workingSomething isn't working

Status: Open.

#18185 In ggml-org/llama.cpp;

· ServeurpersoCom opened on Dec 18, 2025

Eval bug: Hermes 2 Pro parser accepts unwrapped JSON as tool call

bugSomething isn't workingSomething isn't working

chat parserIssues related to the chat parser and chat templatesIssues related to the chat parser and chat templates

Status: Open.

#18183 In ggml-org/llama.cpp;

· wadealexc opened on Dec 18, 2025

Feature Request: convert_hf_to_gguf.py to default to the original precision

breaking changeChanges that break ABIs, APIs, file formats, or other forms of backwards compatibility.Changes that break ABIs, APIs, file formats, or other forms of backwards compatibility.

enhancementNew feature or requestNew feature or request

good first issueGood for newcomersGood for newcomers

help wantedNeeds help from the communityNeeds help from the community

Status: Open.

#18182 In ggml-org/llama.cpp;

· mramendi opened on Dec 18, 2025

Misc. bug: fit params does not take vision encoders into account

bugSomething isn't workingSomething isn't working

Status: Open.

#18181 In ggml-org/llama.cpp;

· verygreen opened on Dec 18, 2025

Feature Request: add tab-autocomplete to llama-cli

cliIssues related to the command-line clientIssues related to the command-line client

enhancementNew feature or requestNew feature or request

Status: Open.

#18177 In ggml-org/llama.cpp;

· am17an opened on Dec 18, 2025

Feature Request: Tensor size prevents quantization of models like Nemotron to below 4 bit - auto-padding ?

enhancementNew feature or requestNew feature or request

needs evalNeeds maintainer evaluation for inclusion viabilityNeeds maintainer evaluation for inclusion viability

Status: Open.

#18176 In ggml-org/llama.cpp;

· cmp-nct opened on Dec 18, 2025

Eval bug: FIT not calculating properly

bug-unconfirmed

Status: Open.

#18175 In ggml-org/llama.cpp;

· yggdrasil75 opened on Dec 18, 2025

Misc. bug: Sampler/grammar exception: Unexpected empty grammar stack after accepting piece: {" with json.gbnf

bug-unconfirmed

need feedbackTesting and feedback with results are neededTesting and feedback with results are needed

regressionA regression introduced in a new build (something that was previously working correctly)A regression introduced in a new build (something that was previously working correctly)

Status: Open.

#18173 In ggml-org/llama.cpp;

· koh-osug opened on Dec 18, 2025

Misc. bug: Llama-server.exe in router mode errors and bails on processing image task, no error when in non-router mode

bugSomething isn't workingSomething isn't working

mtmdRelated to multimodal functionality (video/image/audio)Related to multimodal functionality (video/image/audio)

server

server/api

Status: Open.

#18170 In ggml-org/llama.cpp;

· alan-l opened on Dec 18, 2025

Eval bug: GLM4.6V Flash incoherent on Vulkan

bugSomething isn't workingSomething isn't working

merge readyindicates that this may be ready to merge soon and is just holding out in case of objectionsindicates that this may be ready to merge soon and is just holding out in case of objections

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

#18164 In ggml-org/llama.cpp;

· LostRuins opened on Dec 18, 2025

presets: add preset-only options

Status: Open.

#18163 In ggml-org/llama.cpp;

· ngxson opened on Dec 18, 2025

Misc. bug: Requested buffer size exceeds device buffer size limit for previously working configuraitons.

bugSomething isn't workingSomething isn't working

regressionA regression introduced in a new build (something that was previously working correctly)A regression introduced in a new build (something that was previously working correctly)

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open.

#18160 In ggml-org/llama.cpp;

· engrtipusultan opened on Dec 18, 2025