llama_cpp 0.9.3 → 0.9.5
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +4 -4
- data/CHANGELOG.md +8 -0
- data/ext/llama_cpp/src/ggml-alloc.c +1 -1
- data/ext/llama_cpp/src/ggml-cuda.cu +177 -98
- data/ext/llama_cpp/src/ggml-metal.m +29 -17
- data/ext/llama_cpp/src/ggml-metal.metal +93 -93
- data/ext/llama_cpp/src/ggml-opencl.cpp +5 -7
- data/ext/llama_cpp/src/ggml-quants.c +1 -1
- data/ext/llama_cpp/src/ggml.c +154 -30
- data/ext/llama_cpp/src/ggml.h +11 -3
- data/ext/llama_cpp/src/llama.cpp +316 -122
- data/ext/llama_cpp/src/llama.h +72 -4
- data/lib/llama_cpp/version.rb +2 -2
- metadata +3 -3
checksums.yaml
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
---
|
|
2
2
|
SHA256:
|
|
3
|
-
metadata.gz:
|
|
4
|
-
data.tar.gz:
|
|
3
|
+
metadata.gz: 4fd4e1a5e4d7e2442ab43255996da3ce92f898f9876f1bda343e2433c5050dd7
|
|
4
|
+
data.tar.gz: dece2da6c9befa15e6990d18fb58e2bf13d8da6c62033969b6b5104f82df736d
|
|
5
5
|
SHA512:
|
|
6
|
-
metadata.gz:
|
|
7
|
-
data.tar.gz:
|
|
6
|
+
metadata.gz: 51a383690b6e90e9493e1f318e916dfd94a909f4e554afd8ea822d047f05e96be3e2f371e83f0da5a37a9837d9ae5ecc6992bb9d9c0fd60a9de521bcd148e8f7
|
|
7
|
+
data.tar.gz: 15bbe94edb232d1979f2907c6c3ab7325a1089f9dcdd5d4262d7f0955fd6183e6b01cfee16593165f6e9901991e765ea30740bc1a83cca8fad60df4417551e3b
|
data/CHANGELOG.md
CHANGED
|
@@ -1,3 +1,11 @@
|
|
|
1
|
+
## [[0.9.5](https://github.com/yoshoku/llama_cpp.rb/compare/v0.9.4...v0.9.5)] - 2023-12-02
|
|
2
|
+
|
|
3
|
+
- Bump bundled llama.cpp from b1555 to b1593.
|
|
4
|
+
|
|
5
|
+
## [[0.9.4](https://github.com/yoshoku/llama_cpp.rb/compare/v0.9.3...v0.9.4)] - 2023-11-25
|
|
6
|
+
|
|
7
|
+
- Bump bundled llama.cpp from b1523 to b1555.
|
|
8
|
+
|
|
1
9
|
## [[0.9.3](https://github.com/yoshoku/llama_cpp.rb/compare/v0.9.2...v0.9.3)] - 2023-11-18
|
|
2
10
|
|
|
3
11
|
- Bump bundled llama.cpp from b1500 to b1523.
|
|
@@ -137,7 +137,7 @@ void ggml_tallocr_alloc(ggml_tallocr_t alloc, struct ggml_tensor * tensor) {
|
|
|
137
137
|
|
|
138
138
|
#ifdef GGML_ALLOCATOR_DEBUG
|
|
139
139
|
add_allocated_tensor(alloc, tensor);
|
|
140
|
-
size_t cur_max = (char*)addr - (char*)alloc->
|
|
140
|
+
size_t cur_max = (char*)addr - (char*)alloc->base + size;
|
|
141
141
|
if (cur_max > alloc->max_size) {
|
|
142
142
|
printf("max_size = %.2f MB: tensors: ", cur_max / 1024.0 / 1024.0);
|
|
143
143
|
for (int i = 0; i < 1024; i++) {
|