24 Releases published by 1 person

• b5142

  published Apr 15, 2025

• b5143

  published Apr 16, 2025

• b5144

  published Apr 16, 2025

• b5145

  published Apr 16, 2025

• b5146

  published Apr 17, 2025

• b5147

  published Apr 17, 2025

• b5148

  published Apr 17, 2025

• b5149

  published Apr 17, 2025

• b5150

  published Apr 18, 2025

• b5151

  published Apr 18, 2025

• b5152

  published Apr 18, 2025

• b5153

  published Apr 18, 2025

• b5155

  published Apr 18, 2025

• b5156

  published Apr 19, 2025

• b5158

  published Apr 19, 2025

• b5159

  published Apr 20, 2025

• b5160

  published Apr 20, 2025

• b5161

  published Apr 20, 2025

• b5162

  published Apr 20, 2025

• b5163

  published Apr 21, 2025

• b5164

  published Apr 21, 2025

• b5165

  published Apr 21, 2025

• b5166

  published Apr 22, 2025

• b5169

  published Apr 22, 2025

28 Pull requests merged by 17 people

• mtmd : support SmolVLM (version 1 and 2)

  #13050 merged Apr 22, 2025

• security : add note about RPC and server functionality

  #13061 merged Apr 22, 2025

• metal : add memory pool for temp allocs

  #12850 merged Apr 22, 2025

• llava : update documentations

  #13055 merged Apr 22, 2025

• ggml : add SSE 4.2 and x64 base variant for CPUs without AVX

  #12871 merged Apr 21, 2025

• SYCL: Add non-contiguous support in ROPE

  #12993 merged Apr 21, 2025

• mtmd : merge llava, gemma3 and minicpmv CLI into single llama-mtmd-cli

  #13012 merged Apr 21, 2025

• convert : experimental support for --mmproj flag

  #13023 merged Apr 20, 2025

• llava: fix errors in clip.h on certain pure C compilers

  #13030 merged Apr 20, 2025

• vulkan: support noncontiguous rms_norm

  #13031 merged Apr 20, 2025

• metal: add neg operator

  #13029 merged Apr 20, 2025

• Disable CI cross-compile builds

  #13022 merged Apr 19, 2025

• gguf-py : fix upload python package workflow

  #13020 merged Apr 19, 2025

• clip : refactor, add image_manipulation and llava_uhd classes

  #13011 merged Apr 19, 2025

• main : Fix Ctrl+D/newline handling

  #12951 merged Apr 18, 2025

• gguf-py : GGUF Editor GUI - Python + Qt

  #12930 merged Apr 18, 2025

• server : use std::move whenever possible

  #12936 merged Apr 18, 2025

• SYCL: Refactor and enable FP16 in binary broadcast OPs

  #12975 merged Apr 18, 2025

• mtmd : add methods to access mtmd_image_tokens

  #12906 merged Apr 18, 2025

• rpc : add RPC_CMD_HELLO

  #12955 merged Apr 18, 2025

• graph : make FA compatible with MLA + add initial Metal kernels

  #12953 merged Apr 17, 2025

• ggml: Re-enable CUDA graphs in presence of CONT and DUP nodes

  #12970 merged Apr 17, 2025

• CANN: Add support for async operator submission

  #12864 merged Apr 17, 2025

• Recognize IBM Granite 3.3 FIM tokens. Makes llama-server /infill usable.

  #12988 merged Apr 17, 2025

• opencl: fix incorrect local_size index in profiling log

  #12868 merged Apr 16, 2025

• vulkan: enable coopmat2 FA gqa and split_k optimizations more often

  #12931 merged Apr 16, 2025

• [CANN]310P OPT Support

  #12962 merged Apr 16, 2025

• opencl: split ggml-opencl.cl into multiple files and cleanup

  #12886 merged Apr 15, 2025

23 Pull requests opened by 22 people

• sycl: use DNN in the first part of ggml_sycl_mul_mat_batched_sycl

  #12972 opened Apr 16, 2025

• Fix convert script for non-hf GLM4 checkpoints

  #12992 opened Apr 17, 2025

• [CANN] Add the n_graph_splits performance metric to llama-bench.

  #12994 opened Apr 17, 2025

• threading: support for GGML_SCHED_PRIO_LOW, update thread info on Windows to avoid throttling

  #12995 opened Apr 17, 2025

• make memset range dynamic

  #13002 opened Apr 18, 2025

• [SYCL][OPT] Fix reorder optimization for Q4_0

  #13003 opened Apr 18, 2025

• Nix portability improvements

  #13005 opened Apr 18, 2025

• CUDA: noncont MMVQ + batched bs1 MUL_MAT_ID

  #13014 opened Apr 18, 2025

• vulkan: matmul gcn tuning

  #13016 opened Apr 18, 2025

• Append mult-eos,half-rope,bos to GLM4-0414 and Z

  #13021 opened Apr 19, 2025

• Bitnet: directly use scale instead of inverting it twice

  #13026 opened Apr 19, 2025

• quantize: improve pattern matching for allowed tensors

  #13033 opened Apr 20, 2025

• gguf-py : avoid requiring PySide6 for packaged scripts

  #13036 opened Apr 20, 2025

• quantize: Handle user-defined pruning of whole layers (blocks)

  #13037 opened Apr 20, 2025

• [CANN]Support OP MUL_MAT_ID

  #13042 opened Apr 21, 2025

• llama-gemma3-cli: Sigint rework in gemma3 vision example

  #13043 opened Apr 21, 2025

• ggml-cpu: Integrate fp32=bf16xbf16 SME KleidiAI kernel

  #13053 opened Apr 21, 2025

• Update README.md for tts example to use afplay on MacOS

  #13056 opened Apr 22, 2025

• Fix ChatGLMModel for glm-4-9b cannot find tokenizer merges in model file

  #13058 opened Apr 22, 2025

• rpc : add command line option for number of threads for the CPU backend

  #13060 opened Apr 22, 2025

• cmake : do not include ./src as public for libllama

  #13062 opened Apr 22, 2025

• mtmd : Support Pixtral 12B (help needed - 2D RoPE)

  #13065 opened Apr 22, 2025

• fix(rpc): Improve input validation and error handling

  #13069 opened Apr 22, 2025

30 Issues closed by 12 people

• Feature Request:

  #13070 closed Apr 22, 2025

• Big performance regression of llama-bench with Vulkan backend using forced integer dot product code path (at least on NV 4070 latest driver) (from initial support in b5010)..

  #13063 closed Apr 22, 2025

• Feature Request: Support for Deepseek Janus-Pro-7B & Janus-1.3B

  #11490 closed Apr 22, 2025

• Feature Request: Convert deepseek-v3's mtp module to gguf and quantize to q4km

  #12242 closed Apr 22, 2025

• Misc. bug: Misc. bug: cannot convert GLM-4v-9B- (glm-4v-9b) to GGUF format #11263

  #12266 closed Apr 22, 2025

• Misc. bug: llama fails to run on older x86 hardware.

  #12866 closed Apr 21, 2025

• Eval bug: Deepseek V2 Lite no longer working with Vulkan (assert fail during tg)

  #12956 closed Apr 20, 2025

• Compile bug: there is a build bug in examples/llama.android and it will brings build failure in CI

  #12638 closed Apr 20, 2025

• Eval bug: CANNOT LINK EXECUTABLE "./llama-cli": library "libomp.so" not found: needed by main executable

  #11979 closed Apr 20, 2025

• Feature Request: Proposing User-Customizable RAG Integration in llama.cpp: A Path to Enhanced Contextual Retrieval

  #12129 closed Apr 20, 2025

• Refactor: rename `n_swa` to `n_aux`

  #13019 closed Apr 19, 2025

• Misc. bug: HIP compilation together with -DGGML_CPU_ALL_VARIANTS=ON does not load the model or detects the GPU

  #12175 closed Apr 19, 2025

• Eval bug: Server returns 500 error on /api/generate and /api/chat requests

  #12176 closed Apr 19, 2025

• Misc. bug: llama-rpc crashes when deciding memory on CPU with CUDA_VISIBLE_DEVICES=""

  #12203 closed Apr 19, 2025

• Misc. bug: Ctrl+D no longer works properly

  #12949 closed Apr 18, 2025

• Does V100 support flash-attention?

  #13008 closed Apr 18, 2025

• Feature Request: Support for Apriel-5B-Instruct

  #12926 closed Apr 18, 2025

• Feature Request: Allow more than 16 combined devices to participate in RPC

  #12967 closed Apr 17, 2025

• Misc. bug: HIP when using llama.bench and kv cache quant cpu is doing the work instead of gpu

  #12624 closed Apr 17, 2025

• Eval bug: The answers have some problems with the example/llama.android

  #12158 closed Apr 17, 2025

• Compile bug: issue compiling in ubuntu (desktop and server version) using virtualbox

  #12164 closed Apr 17, 2025

• Misc. bug: I have a 20 core and 30 thread CPU, and anything above 3 or 4 CPU thread pool size doesn't give any improvements in tokens per second.

  #12966 closed Apr 16, 2025

• Misc. bug: "Unexpected empty grammar stack after accepting piece" tool crash

  #12597 closed Apr 16, 2025

• Misc. bug: RISCV output bug when using rvv with vlen > 256bit

  #11041 closed Apr 16, 2025

• Eval bug: Error running Phi4-mini gguf: unknown pre-tokenizer type: 'gpt-4o'

  #12122 closed Apr 16, 2025

• Eval bug: In RISC-V, output tokens are broken

  #12124 closed Apr 16, 2025

• Feature Request:

  #12128 closed Apr 16, 2025

• Misc. bug: vulkan on Adreno GPU

  #12139 closed Apr 16, 2025

• Misc. bug: gguf-dump 'newbyteorder' was removed

  #12146 closed Apr 16, 2025

• Replacement for deprecated codevct string conversion

  #12151 closed Apr 16, 2025

35 Issues opened by 30 people

• Compile bug: Vulkan Cross compile for arm64

  #13068 opened Apr 22, 2025

• Misc. bug: RPC server crash on `SET_TENSOR` with invalid `ggml_type`

  #13067 opened Apr 22, 2025

• Model Repeats Nonsensical Output

  #13066 opened Apr 22, 2025

• Doc. bug: docs/multimodal/gemma3.md need to be updated

  #13064 opened Apr 22, 2025

• Compile bug: NVIDIA A800-SXM4-40GB ggml_cuda_init failed

  #13059 opened Apr 22, 2025

• Misc. bug: in version 0.16.2, the gguf-dump CLI tool fails due to a missing PySide6 module, indicating an unintended GUI depende

  #13054 opened Apr 21, 2025

• Misc. bug: Intel container images keep getting `No space left on device` during CI Build

  #13052 opened Apr 21, 2025

• Misc. bug: CPU Usage low in rpc-server mode

  #13051 opened Apr 21, 2025

• Slow token generation speed of Gemma 3 QAT Models

  #13048 opened Apr 21, 2025

• Eval bug: multimodal llama-gemma3-cli gives nonsensical outputs when used with Vulkan

  #13046 opened Apr 21, 2025

• Misc. bug: llama-cli (vulkan backend) output gibberish with old vulkan sdk

  #13044 opened Apr 21, 2025

• Eval bug: Error when load `bge-reranker-v2-gemma` model

  #13041 opened Apr 21, 2025

• Feature Request: Ability to pack multiple GGUFs into single one

  #13028 opened Apr 19, 2025

• Feature Proposal: Server Model Switching at Runtime

  #13027 opened Apr 19, 2025

• Misc. bug: (clip.cpp) q8_0 mmproj is broken on gemma 3

  #13025 opened Apr 19, 2025

• Eval bug: Vulkan: "Requested buffer size exceeds device memory allocation limit" even with `-ngl 0` when trying to run very large models

  #13024 opened Apr 19, 2025

• Eval bug: RWKV inference issue with llama-server

  #13018 opened Apr 19, 2025

• [Build] Some Build Options/Definitions seems Missing in ggml-base

  #13017 opened Apr 19, 2025

• Perplexity script for non GGUF quantization

  #13015 opened Apr 18, 2025

• Eval bug: why Gemma 3 model has run into CPU inference

  #13004 opened Apr 18, 2025

• Eval bug: Segmentation fault when running gemma3-cli on Android

  #13000 opened Apr 18, 2025

• gmake[2]: *** [tests/CMakeFiles/test-tokenizer-0.dir/build.make:107: bin/test-tokenizer-0] Error 1

  #12998 opened Apr 18, 2025

• Eval bug: microsoft/bitnet-b1.58-2B-4T-gguf

  #12997 opened Apr 17, 2025

• Eval bug: HIP: llama.cpp server locks up when running multiple instances on the same gpu

  #12991 opened Apr 17, 2025

• Eval bug: Quad P40 unable to run 70B models on recent releases

  #12990 opened Apr 17, 2025

• Feature Request: Add kv-quant fa kernel variants for head sizes other than 128

  #12989 opened Apr 17, 2025

• Misc. bug: Potential memory leak in backend registry

  #12986 opened Apr 16, 2025

• Feature Request: llama-tts: read from text files and pipe audio signals to stdout for direct audio conversion using ffmpeg

  #12984 opened Apr 16, 2025

• Feature Reequest: Multi model cli tools: Add a possibility to specify a image in conversation mode plus tab auto completion for path

  #12983 opened Apr 16, 2025

• Feature Request: Make chat sessions possible with multi model cli tools

  #12982 opened Apr 16, 2025

• Feature Request: multi model cli tools: Convert submitted images to best size and format for model

  #12981 opened Apr 16, 2025

• Misc. bug: Only using 1 compute core on AMD

  #12978 opened Apr 16, 2025

• Misc. bug: Vulkan performance depends on thread priority

  #12976 opened Apr 16, 2025

• Eval bug: Gemma-3 Vision failed with CUDA

  #12973 opened Apr 16, 2025

• Misc. bug: llama-server speculative decoding not as performant as llama-speculative-simple

  #12968 opened Apr 16, 2025

89 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• Resolved half rope,multi-EOS issues in convert_hf_togguf.py for GLM4Z Model

  #12957 commented on Apr 22, 2025 • 17 new comments

• Add Qwen2.5VL support

  #12402 commented on Apr 22, 2025 • 11 new comments

• server : (experimental) vision support via libmtmd

  #12898 commented on Apr 22, 2025 • 10 new comments

• sycl : implementation of reordered Q4_0 MMVQ for Intel GPUs

  #12858 commented on Apr 22, 2025 • 8 new comments

• llama-bench: enhance benchmark with improved token throughput measurements

  #12874 commented on Apr 21, 2025 • 2 new comments

• tts : implement sesame CSM + Mimi decoder

  #12648 commented on Apr 22, 2025 • 1 new comment

• llama-bench : Add `--override-tensors` arg

  #12922 commented on Apr 21, 2025 • 1 new comment

• imatrix : use GGUF to store importance matrices

  #9400 commented on Apr 15, 2025 • 0 new comments

• Eval bug: Loading fail on Gemma 3:12b > llama_model_load: error loading model: error loading model hyperparameters: key not found in model: gemma3.attention.layer_norm_rms_epsilon

  #12367 commented on Apr 22, 2025 • 0 new comments

• Eval bug: LLaVa convert_image_encoder_to_gguf.py fails to byteswap v.head.ffn_up.bias tensor on Big-Endian system

  #12863 commented on Apr 22, 2025 • 0 new comments

• Compile bug: There was a errror while compiling support for the backend Vulkan

  #12619 commented on Apr 22, 2025 • 0 new comments

• Feature Request: Interleaved sliding window attention support for gemma 2 and 3

  #12637 commented on Apr 22, 2025 • 0 new comments

• Feature Request: Improve model load time when using the RPC backend

  #12954 commented on Apr 22, 2025 • 0 new comments

• Eval bug: input is too large to process. increase the physical batch size

  #12295 commented on Apr 22, 2025 • 0 new comments

• Compile bug: Prooted Debian in Droid Termux only

  #12452 commented on Apr 22, 2025 • 0 new comments

• Eval bug: llama.swiftui Unexpectedly found nil while unwrapping an Optional value

  #12510 commented on Apr 22, 2025 • 0 new comments

• Misc. bug: test-backend-ops grad crash by GGML_ASSERT error

  #12520 commented on Apr 22, 2025 • 0 new comments

• Feature request: Graphical GGUF viewer

  #6715 commented on Apr 21, 2025 • 0 new comments

• Error while converting peft finetuned merged model to gguf

  #12494 commented on Apr 21, 2025 • 0 new comments

• Misc. bug: Buffer offset is not aligned on macOS / Intel / Vulkan

  #10984 commented on Apr 21, 2025 • 0 new comments

• Feature Request: YuE (music gen)

  #11467 commented on Apr 21, 2025 • 0 new comments

• Misc. bug: --no-context-shift OR --context-shift ?

  #12038 commented on Apr 21, 2025 • 0 new comments

• Eval bug: Gemma-3 vision don't work multilingual

  #12351 commented on Apr 21, 2025 • 0 new comments

• Feature Request: New sampling method that boosts reasoning performance - looks too good?

  #12479 commented on Apr 21, 2025 • 0 new comments

• Feature Request: deep/ recurrent processing like "thinking", but script based.

  #12486 commented on Apr 21, 2025 • 0 new comments

• Compile bug: Error build llama cpp on CUDA

  #12491 commented on Apr 21, 2025 • 0 new comments

• Llama-3_1-Nemotron-Ultra-253B-v1 support

  #12843 commented on Apr 22, 2025 • 0 new comments

• `common`: add partial regex support

  #12808 commented on Apr 19, 2025 • 0 new comments

• ci: fix cross-compile sync issues

  #12804 commented on Apr 19, 2025 • 0 new comments

• `server`: inject date_string in llama 3.x template + fix date for firefunction v2

  #12802 commented on Apr 19, 2025 • 0 new comments

• kv-cache : separate recurrent vs non-recurrent impl

  #12799 commented on Apr 22, 2025 • 0 new comments

• Support for OuteTTS 1.0

  #12794 commented on Apr 22, 2025 • 0 new comments

• (wip) support ultravox audio input

  #12745 commented on Apr 22, 2025 • 0 new comments

• imatrix: add option to display importance score statistics for a given imatrix file

  #12718 commented on Apr 22, 2025 • 0 new comments

• vulkan: Add bfloat16 support

  #12554 commented on Apr 22, 2025 • 0 new comments

• (draft) tts: Orpheus support

  #12487 commented on Apr 22, 2025 • 0 new comments

• Nomic Embed Text V2 with Mixture-of-Experts (MoE) architecture

  #12466 commented on Apr 22, 2025 • 0 new comments

• `server`: streaming of tool calls and thoughts when `--jinja` is on

  #12379 commented on Apr 16, 2025 • 0 new comments

• PR: Refine ggml-hexagon backend(Qualcomm Hexagon NPU backend) for latest ggml,whisper.cpp,llama.cpp

  #12326 commented on Apr 21, 2025 • 0 new comments

• [WIP]backend: Integrating QNN (Qualcomm AI Engine Direct) as a dedicated backend for Qualcomm NPUs

  #12063 commented on Apr 21, 2025 • 0 new comments

• Supporting Velvet model

  #11716 commented on Apr 16, 2025 • 0 new comments

• Allow user to compile with any cuda version using github actions

  #10928 commented on Apr 20, 2025 • 0 new comments

• llama/ggml: add LLM training support

  #10544 commented on Apr 22, 2025 • 0 new comments

• add FP8 support to gguf/llama:

  #10055 commented on Apr 18, 2025 • 0 new comments

• [Draft] Tensor Parallel support to llama.cpp

  #9648 commented on Apr 19, 2025 • 0 new comments

• Feature Request: Ovis2 Support

  #12358 commented on Apr 17, 2025 • 0 new comments

• Misc. bug: Loading models result in Fatal signal 4 (SIGILL) on some Android devices

  #12393 commented on Apr 17, 2025 • 0 new comments

• Misc. bug: Program closes without providing an error.

  #12417 commented on Apr 17, 2025 • 0 new comments

• Misc. bug: CUDA graph update failed OR Failed to allocate graph error and cuda hang

  #12420 commented on Apr 17, 2025 • 0 new comments

• Qualcomm Adreno : Compute pipeline creation failed for mul_mat_vec_q4_k_f32_f32_1

  #12421 commented on Apr 17, 2025 • 0 new comments

• Feature Request: Support inference for OVIS2 models - (GGUF Conversion & quantization done with success!)

  #12429 commented on Apr 17, 2025 • 0 new comments

• Compile bug:

  #12431 commented on Apr 17, 2025 • 0 new comments

• Compile bug: fatal: not a git repository (nor any of the parent directories): .git

  #12438 commented on Apr 17, 2025 • 0 new comments

• Imatrix quantization bug: OLMo-2-0325-32B-Instruct found nan value

  #12439 commented on Apr 17, 2025 • 0 new comments

• Compile bug: FAILED: examples/llava/CMakeFiles/llava.dir/llava.cpp.obj

  #12899 commented on Apr 16, 2025 • 0 new comments

• Feature Request: Add support of convert.py for model Qwen2.5-Omni-7B

  #12641 commented on Apr 16, 2025 • 0 new comments

• Eval bug: convert_hf_to_gguf.py AttributeError:

  #12847 commented on Apr 16, 2025 • 0 new comments

• Regarding llama-bench and llama-parallel commands

  #12106 commented on Apr 16, 2025 • 0 new comments

• Compile bug: how to enable opencl in termux

  #12911 commented on Apr 16, 2025 • 0 new comments

• Eval bug: <think> tag with DeepSeek-R1-Distill-Qwen-1.5B-Q5_K_M.gguf

  #11325 commented on Apr 16, 2025 • 0 new comments

• Eval bug: Gemma 3 extremly slow prompt processing when using quantized kv cache.

  #12352 commented on Apr 16, 2025 • 0 new comments

• How do I know which operator the code is computing has an error?

  #12389 commented on Apr 16, 2025 • 0 new comments

• Eval bug: KV cache changes the inference results, even when context fits and no quantization

  #12396 commented on Apr 16, 2025 • 0 new comments

• Eval bug: llama-qwen2vl-cli gives too short or cut response

  #12408 commented on Apr 16, 2025 • 0 new comments

• Eval bug: main: failed to load image /home/data1/protected/hyperscope/9/4/3/1/6/2025-02-21/Raw Crystal of Morganite Gemstone.jpg. Terminating

  #12410 commented on Apr 16, 2025 • 0 new comments

• Eval bug: GLM-Z1-9B-0414

  #12946 commented on Apr 15, 2025 • 0 new comments

• Compile bug: gcc-11: error: unrecognized command-line option '-compress-mode=size'

  #12325 commented on Apr 15, 2025 • 0 new comments

• tts : add support for SparkTTS

  #12495 commented on Apr 21, 2025 • 0 new comments

• server : improvements and maintenance

  #4216 commented on Apr 20, 2025 • 0 new comments

• Misc. bug: The KV cache is sometimes truncated incorrectly when making v1/chat/completions API calls

  #11970 commented on Apr 20, 2025 • 0 new comments

• Feature Request: (webui) Implement a experimental features on webui

  #11662 commented on Apr 20, 2025 • 0 new comments

• Bug tracker: (webui/experimental) Python interpreter via pyodide

  #11762 commented on Apr 20, 2025 • 0 new comments

• Eval bug: does llama.cpp support Intel AMX instruction? how to enable it

  #12003 commented on Apr 20, 2025 • 0 new comments

• Eval bug: getting assertion error when trying to use a gguf quantized model at inference "GGML_ASSERT(n_outputs_enc > 0 && "call llama_encode() first") failed"

  #12080 commented on Apr 20, 2025 • 0 new comments

• Misc. bug: The model's reasoning performance has significantly decreased despite using different versions of the same model architecture, identical parameters, and the same set of questions.

  #12816 commented on Apr 19, 2025 • 0 new comments

• Eval bug: OpenAI incompatible image handling in server multimodal

  #12947 commented on Apr 19, 2025 • 0 new comments

• server: Bring back multimodal support

  #8010 commented on Apr 19, 2025 • 0 new comments

• Research: Performance differences between Metal (macOS) and Vulkan (Linux)

  #10982 commented on Apr 19, 2025 • 0 new comments

• Misc. bug: Using `-c -1` results in `n_ctx = 4294967295` or `n_ctx = 8`

  #12414 commented on Apr 19, 2025 • 0 new comments

• Eval bug: RK3588 Unexpected inf values cause garbled output(or core dump) in llama-cli

  #12458 commented on Apr 19, 2025 • 0 new comments

• Feature Request: Qwen2.5 0.5b OpenCL backend support

  #12463 commented on Apr 19, 2025 • 0 new comments

• Eval bug: MiniCPM-2B-128k convert_hf_to_gguf Missing the required key: rope_scaling

  #12468 commented on Apr 19, 2025 • 0 new comments

• Misc. bug: rpc - Flash Attention Failure in Metal/CUDA RPC Mixed Environment

  #12655 commented on Apr 18, 2025 • 0 new comments

• Eval bug: Excessive stack usage during tool calling

  #12234 commented on Apr 18, 2025 • 0 new comments

• Do you add LLaDA model support?

  #12360 commented on Apr 18, 2025 • 0 new comments

• Eval bug: How to isolate chat history

  #12440 commented on Apr 18, 2025 • 0 new comments

• Feature Request: Cache nix builds on a public cache server?

  #12453 commented on Apr 18, 2025 • 0 new comments

• Suport for Jamba JambaForCausalLM

  #6372 commented on Apr 17, 2025 • 0 new comments

• [Feature request] convert_hf_to_gguf.py supports for bloomz

  #12356 commented on Apr 17, 2025 • 0 new comments