chore: bump llama.cpp · gpustack/llama-box@dbd7284

Commit

chore: bump llama.cpp

Signed-off-by: thxCode <thxcode0824@gmail.com>

thxCode committed Jul 21, 2024

1 parent 50cdc8b commit dbd7284

llama.cpp

Submodule llama.cpp updated 25 files

+1 −1		.github/workflows/build.yml
+32 −26		convert_hf_to_gguf.py
+1 −0		convert_hf_to_gguf_update.py
+18 −8		convert_lora_to_gguf.py
+5 −0		examples/gguf/gguf.cpp
+3 −1		examples/llama.swiftui/llama.cpp.swift/LibLlama.swift
+1 −1		examples/llama.swiftui/llama.swiftui/Models/LlamaState.swift
+222 −137		examples/pydantic_models_to_grammar_examples.py
+24 −0		ggml/src/ggml-cuda/mmq.cu
+731 −646		ggml/src/ggml-cuda/mmq.cuh
+2 −1		ggml/src/ggml-cuda/template-instances/generate_cu_files.py
+5 −0		ggml/src/ggml-cuda/template-instances/mmq-instance-iq1_s.cu
+5 −0		ggml/src/ggml-cuda/template-instances/mmq-instance-iq2_s.cu
+5 −0		ggml/src/ggml-cuda/template-instances/mmq-instance-iq2_xs.cu
+5 −0		ggml/src/ggml-cuda/template-instances/mmq-instance-iq2_xxs.cu
+5 −0		ggml/src/ggml-cuda/template-instances/mmq-instance-iq3_s.cu
+5 −0		ggml/src/ggml-cuda/template-instances/mmq-instance-iq3_xxs.cu
+21 −0		ggml/src/ggml-cuda/vecdotq.cuh
+1 −1		ggml/src/ggml.c
+26 −9		gguf-py/gguf/metadata.py
+3 −3		gguf-py/gguf/utility.py
+38 −5		gguf-py/scripts/gguf_dump.py
+48 −3		gguf-py/tests/test_metadata.py
+1 −0		include/llama.h
+13 −0		src/llama.cpp

Please sign in to comment.