fairydreaming and Georgi Gerganov
b83d1671db
cuda : add support for f16->f16 GGML_OP_SET_ROWS (llama/25367)
2026-07-10 13:06:42 +03:00
fairydreaming and Georgi Gerganov
81c8bb90df
metal : add set_rows with src0 f16 (llama/25434)
...
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com >
2026-07-10 13:06:42 +03:00
72e3f210f3
ggml : add support for CPU f16->f16 GGML_OP_SET_ROWS (llama/25344)
...
* ggml : add support for CPU f16->f16 GGML_OP_SET_ROWS
* ggml : add missing type checks in f16 GGML_OP_SET_ROWS
* ggml : merge ggml_compute_forward_set_rows_f32() and ggml_compute_forward_set_rows_f16() into ggml_compute_forward_set_rows_impl()
* chore : replace assert() with GGML_ASSERT()
---------
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com >
2026-07-10 13:06:42 +03:00
bd5ce1ed48
vulkan : check src0 type in GGML_OP_SET_ROWS to avoid failures due to unimplemented f16 support (llama/25351)
...
* vulkan : check src0 type in GGML_OP_SET_ROWS to avoid failures due to unimplemented f16 support
* chore : get rid of else
---------
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com >
2026-07-10 13:06:42 +03:00
93cff36f15
cuda : concat implementation for quantized types (llama/25303)
...
* cuda : concat implementation for quantized types
* chore : apply am17an clever suggestion to shorten the code
---------
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com >
2026-07-10 13:06:42 +03:00
0a58b4b546
ggml : fix broken CPU concat implementation for quantized types (llama/25247)
...
* ggml : fix broken CPU concat implementation for quantized types
* tests : concat tests for quantized types
---------
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com >
2026-07-10 13:06:42 +03:00
52302312e6
cuda : prevent integer truncation and overflow errors when using KQ mask strides in flash_attn_mask_to_KV_max kernel (llama/24945)
...
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com >
2026-07-10 13:06:42 +03:00
82686b518f
ggml : address integer overflows in binary ops CUDA implementation (llama/24706)
...
* ggml : address integer overflows in binary ops CUDA implementation
* ggml : add size_t casts to avoid integer overflows
* ggml : add more asserts checking integer overflows in binary ops CUDA implementation
---------
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com >
2026-06-26 16:03:57 +03:00
c50e951afd
model : support for DeepseekV32ForCausalLM with generic DeepSeek Sparse Attention (DSA) implementation (llama/23346)
...
* llama : support DeepSeek V3.2 model family (with DSA lightning indexer)
* convert : handle DeepseekV32ForCausalLM architecture
* ggml : support for f16 GGML_OP_FILL
* memory : separate hparams argument in llama_kv_cache constructor
* memory : add llama_kv_cache_dsa memory (KV cache + lightning indexer cache)
* llama : support for LLM_ARCH_DEEPSEEK32
* model : llama_model_deepseek32 implementation
* model : merge two scale operations into one in DSA lightning indexer implementation
* chore : remove unused code
* model : support NVFP4 in DeepSeek V3.2
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com >
* memory : refactoring TODO
Co-authored-by: ggerganov <ggerganov@users.noreply.github.com >
---------
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com >
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com >
Co-authored-by: ggerganov <ggerganov@users.noreply.github.com >
2026-06-08 14:36:36 +03:00
60e420ff6a
cuda : fix KQ mask offset integer overflow in fattn MMA kernel (llama/23610)
...
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com >
2026-05-29 09:47:30 +03:00
c4c6e143a7
ggml : check return value of CUB calls used in argsort and top-k (they all return cudaError_t) (llama/21676)
...
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com >
2026-04-30 11:29:04 +03:00
3499dd83c0
SYCL : support non-contiguous tensors in binary ops (add, sub, etc) (llama/12399)
...
* sycl : support non-contiguous tensors in binary ops
* sycl : silence unused variable warning
---------
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com >
2025-03-27 11:06:03 +02:00
