Georgi Gerganov e298d2fbd0 kv-cache : add SWA support (#13194) ...

* kv-cache : prepare for SWA

ggml-ci

* kv-cache : initial iSWA implementation

ggml-ci

* kv-cache : rework error recovery logic

ggml-ci

* models : fix Phi-3 SWA parameters

ggml-ci

* model : adjust Granite to rope factor changes

ggml-ci

* server : check if context can do shifts

ggml-ci

* iswa : for now, always enable shifts (experiment)

ggml-ci

* kv-cache : simplify SWA logic

ggml-ci

* kv-cache : apply defrag when we fail to find slots for the batch

ggml-ci

* llama : update docs about llama_decode

ggml-ci

* kv-cache : update warning logs when no space for the batch is available

ggml-ci

* llama : add llama_kv_self_seq_pos_min()

* kv-cache : keep track of partial SWA computes and print warnings

* server : disallow use cases involving partial SWA context

ggml-ci

* llama : add param to control SWA cache size

ggml-ci

* minor : clean-up

ggml-ci

2025-05-20 08:05:46 +03:00

..

batched-bench

batched-bench : fix pp batch contents (#13492)

2025-05-13 18:01:53 +03:00

cvector-generator

llama : move end-user examples to tools directory (#13249)

2025-05-02 20:27:13 +02:00

export-lora

llama : move end-user examples to tools directory (#13249)

2025-05-02 20:27:13 +02:00

gguf-split

llama : move end-user examples to tools directory (#13249)

2025-05-02 20:27:13 +02:00

imatrix

imatrix : Add --parse-special for enabling parsing of special tokens in imatrix calculation (#13389)

2025-05-09 11:53:58 +02:00

llama-bench

kv-cache : add SWA support (#13194)

2025-05-20 08:05:46 +03:00

main

llama : do not crash if there is no CPU backend (#13395)

2025-05-09 13:02:07 +02:00

mtmd

mtmd : add vision support for llama 4 (#13282)

2025-05-19 13:04:14 +02:00

perplexity

context : remove logits_all flag (#13284)

2025-05-08 14:26:50 +03:00

quantize

quantize : improve tensor-type pattern matching (#13033)

2025-05-13 19:12:31 +02:00

rpc

llama : do not crash if there is no CPU backend (#13395)

2025-05-09 13:02:07 +02:00

run

llama-run: add support for downloading models from ModelScope (#13370)

2025-05-09 10:25:50 +01:00

server

kv-cache : add SWA support (#13194)

2025-05-20 08:05:46 +03:00

tokenize

llama : move end-user examples to tools directory (#13249)

2025-05-02 20:27:13 +02:00

tts

llama : move end-user examples to tools directory (#13249)

2025-05-02 20:27:13 +02:00

CMakeLists.txt

mtmd : rename llava directory to mtmd (#13311)

2025-05-05 16:02:55 +02:00