Skip to content

Pull requests: auroralabs-loci/llama.cpp

New pull request New
1,064 Open 154 Closed
1,064 Open 154 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

UPSTREAM PR #20019: gguf: add big-endian magic "FUGG" for explicit endianness detection
#1218 opened Mar 2, 2026 by loci-dev Loading…
UPSTREAM PR #20009: server: add Qwen3-Reranker instruction support
#1217 opened Mar 2, 2026 by loci-dev Loading…
1
UPSTREAM PR #20023: gguf-py: add type validation to GGUFWriter.add_key_value
#1216 opened Mar 2, 2026 by loci-dev Loading…
1
UPSTREAM PR #12727: Update llama-quant.cpp llama_tensor_get_type with DeepSeek friendly modifications
#1215 opened Mar 2, 2026 by loci-dev Loading…
UPSTREAM PR #19999: cuda: fix grid.y overflow in non-contiguous dequantize/convert kernels
#1214 opened Mar 1, 2026 by loci-dev Loading…
1
UPSTREAM PR #19976: vulkan: improve partial offloading performance on AMD
#1213 opened Mar 1, 2026 by loci-dev Loading…
1
UPSTREAM PR #19995: webui: add PWA support
#1212 opened Mar 1, 2026 by loci-dev Loading…
1
UPSTREAM PR #19966: cuda: fix ggml_cuda_cpy crash on partial GPU offload
#1211 opened Feb 28, 2026 by loci-dev Loading…
1
UPSTREAM PR #19938: scripts: ini_to_opencode.py
#1210 opened Feb 28, 2026 by loci-dev Loading…
1
UPSTREAM PR #19850: ggml-webgpu: Support non-contiguous src0 and overlapping src0/src1 in binary ops
#1209 opened Feb 28, 2026 by loci-dev Loading…
1
UPSTREAM PR #19770: quantize : fail-early on missing imatrix; refactor + optimize
#1208 opened Feb 28, 2026 by loci-dev Loading…
1
UPSTREAM PR #19841: server : add chat truncation to keep chat going
#1207 opened Feb 27, 2026 by loci-dev Loading…
5 of 6 tasks
1
UPSTREAM PR #19873: Mirroring /v1/responses to /responses
#1206 opened Feb 27, 2026 by loci-dev Loading…
1
UPSTREAM PR #19904: ggml : simplify amx tile config init
#1205 opened Feb 26, 2026 by loci-dev Loading…
1
UPSTREAM PR #19867: ggml : fix AMX and improve alignment checks
#1204 opened Feb 25, 2026 by loci-dev Loading…
1
UPSTREAM PR #19797: common : add more aliases for sampler CLI params
#1203 opened Feb 25, 2026 by loci-dev Loading…
1
UPSTREAM PR #19827: Kimi Linear block implementation
#1202 opened Feb 24, 2026 by loci-dev Loading…
3
UPSTREAM PR #19796: Add model metadata loading from huggingface for use with tests requiring real model data
#1201 opened Feb 23, 2026 by loci-dev Loading…
3
UPSTREAM PR #19819: hexagon refactor all Ops to use local context struct
#1200 opened Feb 23, 2026 by loci-dev Loading…
1
UPSTREAM PR #19791: tools : add learning-cache tool for persistent latent context
#1199 opened Feb 22, 2026 by loci-dev Loading…
7 of 10 tasks
1
UPSTREAM PR #19785: jinja: correct stats for tojson and string filters
#1198 opened Feb 22, 2026 by loci-dev Loading…
1
UPSTREAM PR #19784: ci : fix rocm release path
#1197 opened Feb 22, 2026 by loci-dev Loading…
1
UPSTREAM PR #19773: server : merge contiguous Responses input items into a single assistant message
#1196 opened Feb 21, 2026 by loci-dev Loading…
1
UPSTREAM PR #19750: feat: Ultra-Low-Bit Quantization Kernels (Q1_5_K, Q2_K_S)
#1195 opened Feb 21, 2026 by loci-dev Loading…
1
UPSTREAM PR #19769: WIP: ggml : add NVFP4 quantization type support
#1194 opened Feb 21, 2026 by loci-dev Loading…
2
ProTip! Follow long discussions with comments:>50.