Skip to content

Releases: ggml-org/llama.cpp

b4847

07 Mar 08:22
5e2d57b
Compare
Choose a tag to compare
metal : simplify kernel arguments using a struct (#3229) (#12194)

* metal : refactor im2col parameters into a struct

* metal: Change im2col offset types from int32_t to uint64_t to support larger memory offsets

* metal : refactor sum_rows parameters into a struct

* metal : refactor soft_max parameters into a struct

* metal : refactor diag_mask_inf parameters into a struct

* metal : refactor ssm_conv parameters into a struct

* metal : refactor ssm_scan parameters into a struct

* metal : refactor get_rows parameters into a struct

* metal : refactor group_norm parameters into a struct

* metal : refactor conv_transpose_1d parameters into a struct

* metal : refactor upscale parameters into a struct

* metal : refactor pad parameters into a struct

* metal : refactor pad_reflect_1d parameters into a struct

* metal : refactor arange parameters into a struct

* metal : refactor timestep_embedding parameters into a struct

* metal : refactor argsort parameters into a struct

* metal : refactor leaky_relu parameters into a struct

* metal : refactor pool_2d parameters into a struct

* metal : fix trailing whitespace

---------

Co-authored-by: alexju <[email protected]>

b4846

07 Mar 07:51
f1648e9
Compare
Choose a tag to compare
HIP: fix rocWMMA build flags under Windows (#12230)

b4837

06 Mar 08:01
e721c05
Compare
Choose a tag to compare
HIP/CUDA: set the paramerter value in maintain_cuda_graph instead of …

b4836

06 Mar 07:03
57b6abf
Compare
Choose a tag to compare
android : fix KV cache log message condition (#12212)

b4835

06 Mar 03:09
94bb63e
Compare
Choose a tag to compare
opencl : fix buffer alignment (#12197)

Fix the following error:

```
ggml-alloc.c:99: not enough space in the buffer
ggml_tallocr_alloc: not enough space in the buffer to allocate blk.17.ffn_down.weight (needed 27525120, available 27521024)
```

which occurs when `ggml_backend_opencl_context::alignment` is larger
than `cl_ptr_base` (hard-coded to `0x1000`).

Also, fix `ggml_backend_opencl_context::alignment` was set to
`CL_DEVICE_MEM_BASE_ADDR_ALIGN` which was treated as bytes but the
value is reported in bits.

b4834

06 Mar 02:51
f792439
Compare
Choose a tag to compare
opencl : fix `ulong` kernel args were set from `int` variables (#12174)

... which left garbage bits in the upper half of the kernel args. This
caused segmentation faults when running PoCL.

b4833

06 Mar 02:43
ed4ce0d
Compare
Choose a tag to compare
opencl : fix profile-related errors (#12095)

Co-authored-by: ubuntu <[email protected]>

b4832

06 Mar 02:40
07d1572
Compare
Choose a tag to compare
ggml-cpu: Faster IQ1 mul_mat_vec on AVX2 using BMI2 instructions (#12…

b4831

05 Mar 16:40
5e43f10
Compare
Choose a tag to compare
SYCL: Disable f16 Unary OPs as not supported by the kernels (#12201)

b4830

05 Mar 16:02
16e4b22
Compare
Choose a tag to compare
ggml : fix GGMLMetalClass ODR (#12200)

-- it might happen if ggml is loaded from 2 separate libraries since each one of them will expose the class. This is more of a guard since we want to use only Metal as embedded library and don't care about the other case.