ggerganov/llama.cpp

Pull Requests Commits

llama : fix simple splits when the batch contains embeddings

compilade committed 280 days ago

80626503

Merge branch 'master' into compilade/batch-splits

compilade committed 281 days ago

80d9d2a5

llama : support for `falcon-mamba` architecture (#9074)

younesbelkada committed 281 days ago

Verified b40eb848

llava : zero-initialize clip_ctx structure fields with aggregate initialization 908)

fairydreaming committed 281 days ago

Verified f63f603c

llama : std::move llm_bigram_bpe from work_queue (#9062)

danbev committed 281 days ago

Verified 8455340b

llama : add llama_model_is_recurrent to simplify figuring that out

compilade committed 281 days ago

1be5ea7d

llama : fix Mamba pooled embeddings with multiple sequences

compilade committed 281 days ago

b264eddb

llama : fix T5 segfault again

compilade committed 281 days ago

652e9b0d

llava: Add ACC OP for GPU acceleration to the Vulkan backend in the LLAVA CLIP model. (#8984)

cyzero-kim committed 281 days ago

Verified 2f3c1466

[SYCL] fallback mmvq (#9088)

airMeng committed 282 days ago

Verified 50addec9

[SYCL] Fix SYCL `im2col` and `convert` Overflow with Large Dims (#9052)

zhentaoyu committed 282 days ago

Verified 4f8d19ff

tests : add missing comma in grammar integration tests (#9099)

fairydreaming committed 282 days ago

Verified 90db8146

cann: add doc for cann backend (#8867)

wangshuai09 committed 283 days ago

Verified cfac111e

rpc : print error message when failed to connect endpoint (#9042)

rgerganov committed 283 days ago

Verified 1b6ff90f

rpc : prevent crashes on invalid input (#9040)

rgerganov committed 283 days ago

Verified 18eaf29f

flake.lock: Update (#9068)

ggerganov committed 284 days ago

Verified 554b0490

tests : add integration test for lora adapters (#8957)

ltoniazzi committed 284 days ago

Verified 2339a0be

Fix incorrect use of ctx_split for bias tensors (#9063)

suhara committed 285 days ago

Verified 2fb92678

server : refactor middleware and /health endpoint (#9056)

ngxson committed 286 days ago

Verified 8b3befc0

llava : support MiniCPM-V-2.6 (#8967)

tc-mb committed 286 days ago

Verified d565bb2f

py : fix wrong input type for raw_dtype in ggml to gguf scripts (#8928)

farbodbj committed 286 days ago

Verified ee2984bd

Fix inference example lacks required parameters (#9035)

Aisuko committed 286 days ago

Verified c8ddce85

gguf-py : bump version from 0.9.1 to 0.10.0 (#9051)

compilade committed 286 days ago

Verified 23fd4535

llama : add EXAONE model support (#9025)

mscheong01 committed 286 days ago

Verified c679e0cb

common : add support for cpu_get_num_physical_cores() on Windows (#8771)

Septa2112 committed 286 days ago

Verified fb487bb5

Add Nemotron/Minitron GGUF Conversion & Inference Support (#8922)

suhara committed 286 days ago

Verified 2a24c8ca

ggml : dynamic ggml_sched_max_splits based on graph_size (#9047)

nicoboss committed 286 days ago

Verified e3f6fd56

retrieval : fix memory leak in retrieval query handling (#8955)

gtygo committed 287 days ago

Verified 4b9afbbe

server : fix duplicated n_predict key in the generation_settings (#8994)

snowyu committed 287 days ago

Verified 37501d9c

common : remove duplicate function llama_should_add_bos_token (#8778)

kylo5aby committed 287 days ago

Verified 4af8420a

Older