forked from ggml-org/llama.cpp
-
Notifications
You must be signed in to change notification settings - Fork 15
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
CUDA graphs: speculative decoding repeatedly resets warmup state
bugSomething isn't workingSomething isn't workingStatus: Open.#45 In am17an/llama.cpp;server router: queued model evicts the model serving its triggering request
bugSomething isn't workingSomething isn't workingStatus: Open.#44 In am17an/llama.cpp;server: malformed tool-call history returns HTTP 500
bugSomething isn't workingSomething isn't workingStatus: Open.#43 In am17an/llama.cpp;server: identical multimodal embeddings differ with default unified KV cache
bugSomething isn't workingSomething isn't workingStatus: Open.#42 In am17an/llama.cpp;server: LFM2.5 MoE with DSpark aborts during CPU context init
bugSomething isn't workingSomething isn't workingStatus: Open.#41 In am17an/llama.cpp;server: model-less preset spawns another router
bugSomething isn't workingSomething isn't workingStatus: Open.#40 In am17an/llama.cpp;server: PEG until() accepts a missing final delimiter
bugSomething isn't workingSomething isn't workingStatus: Open.#39 In am17an/llama.cpp;server: cached repeat changes deterministic Llama 3 output
bugSomething isn't workingSomething isn't workingStatus: Open.#38 In am17an/llama.cpp;server: streaming tools reject logprobs while n_probs works
bugSomething isn't workingSomething isn't workingStatus: Open.#35 In am17an/llama.cpp;server: id_slot pinned to an empty slot skips the prompt cache
bugSomething isn't workingSomething isn't workingStatus: Open.#34 In am17an/llama.cpp;server: tool_choice required does not require a tool call
bugSomething isn't workingSomething isn't workingStatus: Open.#33 In am17an/llama.cpp;server: Unicode tool properties collide in PEG grammar
bugSomething isn't workingSomething isn't workingStatus: Open.#30 In am17an/llama.cpp;