llama.cpp

ggml : more perfo with llamafile tinyblas on x86_64 (#10714)

#1569:Commit 2cd43f4900 pushed by root

master

2024-12-25 02:44:36 +00:00

0s

server : fix missing model id in /model endpoint (#10957)

#1564:Commit 14b699ecde pushed by root

master

2024-12-24 18:34:36 +00:00

0s

llama : support InfiniAI Megrez 3b (#10893)

#1553:Commit b92a14a841 pushed by root

master

2024-12-23 18:04:35 +00:00

0s

convert : add BertForMaskedLM (#10919)

#1547:Commit 5cd85b5e00 pushed by root

master

2024-12-22 14:44:32 +00:00

0s

SYCL: Migrate away from deprecated ggml_tensor->backend (#10840)

#1540:Commit eb5c3dc64b pushed by root

master

2024-12-21 08:54:36 +00:00

0s

llama : minor grammar refactor (#10897)

#1529:Commit 5cab3e4aaa pushed by root

master

2024-12-20 16:34:36 +00:00

0s

ggml : fix arm build (#10890)

#1524:Commit 9177484f58 pushed by root

master

2024-12-19 16:04:37 +00:00

0s

server: avoid overwriting Authorization header (#10878)

#1519:Commit 7bbb5acf12 pushed by root

master

2024-12-18 23:44:35 +00:00

0s

Revert "llama : add Falcon3 support (#10864)" (#10876)

#1514:Commit 4da69d1abd pushed by root

master

2024-12-18 15:34:37 +00:00

0s

Use model->gguf_kv for loading the template instead of using the C API. (#10868)

#1504:Commit d62b532c52 pushed by root

master

2024-12-18 07:24:36 +00:00

0s

llava : Allow locally downloaded models for QwenVL (#10833)

#1490:Commit 4ddd199f6f pushed by root

master

2024-12-17 02:44:32 +00:00

0s

server: Fix `has_next_line` in JSON response (#10818)

#1483:Commit 89d604f2c8 pushed by root

master

2024-12-15 22:14:36 +00:00

0s

llama : add Qwen2VL support + multimodal RoPE (#10361)

#1477:Commit ba1cb19cdd pushed by root

master

2024-12-15 05:54:36 +00:00

0s

Introducing experimental OpenCL backend with support for Qualcomm Adreno GPUs (#10693)

#1471:Commit a76c56fa1a pushed by root

master

2024-12-14 13:34:35 +00:00

0s

ci : build docker images only once daily (#10503)

#1467:Commit 50d5cecbda pushed by root

b4172

2024-12-13 14:09:49 +00:00

0s

ggml : optimize Q4_0 into Q4_0_X_Y repack (#10324)

#1466:Commit 1e58ee1318 pushed by root

b4096

2024-12-13 14:09:49 +00:00

0s

scripts: update compare-llama-bench.py (#10319)

#1464:Commit 4047be74da pushed by root

b4093

2024-12-13 14:09:49 +00:00

0s

server : (web UI) add copy button for code block, fix api key (#10242)

#1463:Commit 9901068ac7 pushed by root

b4085

2024-12-13 14:09:49 +00:00

0s

docs : update bindings list (#10261)

#1462:Commit 1ee9eea094 pushed by root

b4073

2024-12-13 14:09:49 +00:00

0s

sync : ggml

#1461:Commit 190a37d797 pushed by root

b3969

2024-12-13 14:09:49 +00:00

0s

llama.vim : add classic vim support (#9995)

#1460:Commit ac113a0fee pushed by root

b3965

2024-12-13 14:09:49 +00:00

0s

lora : warn user if new token is added in the adapter (#9948)

#1458:Commit c421ac072d pushed by root

b3959

2024-12-13 14:09:49 +00:00

0s

readme : remove --memory-f32 references (#9925)

#1455:Commit 17bb928080 pushed by root

b3937

2024-12-13 14:09:49 +00:00

0s

grammar : fix JSON Schema for string regex with top-level alt. (#9903)

#1453:Commit 66c2c93082 pushed by root

b3928

2024-12-13 14:09:49 +00:00

0s

readme : add programmable prompt engine language CLI (#9599)

#1452:Commit 1d48e98e4f pushed by root

b3810

2024-12-13 14:09:49 +00:00

0s

quantize : improve type name parsing (#9570)

#1451:Commit 63351143b2 pushed by root

b3796

2024-12-13 14:09:49 +00:00

0s

convert : identify missing model files (#9397)

#1449:Commit d54c21df7e pushed by root

b3769

2024-12-13 14:09:49 +00:00

0s

py : add "LLaMAForCausalLM" conversion support (#9485)

#1446:Commit 3c7989fd29 pushed by root

b3758

2024-12-13 14:09:49 +00:00

0s

readme : update tools list (#9475)

#1444:Commit d6b37c881f pushed by root

b3757

2024-12-13 14:09:49 +00:00

0s

server : Add option to return token pieces in /tokenize endpoint (#9108)

#1442:Commit 78203641fe pushed by root

b3748

2024-12-13 14:09:49 +00:00

0s