Georgi Gerganov fd1234cb46 llama : add gpt-oss (#15091) ...

* oai moe

* compat with new checkpoint

* add attn sink impl

* add rope scaling yarn

* logits match with latest transformers code

* wip chat template

* rm trailing space

* use ggml_scale_bias

* rm redundant is_swa_all

* convert interleaved gate_up

* graph : fix activation function to match reference (#7)

* vocab : handle o200k_harmony special tokens

* ggml : add attention sinks support (#1)

* llama : add attn sinks

* ggml : add attn sinks

* cuda : add attn sinks

* vulkan : add support for sinks in softmax

remove unnecessary return

* ggml : add fused swiglu_oai op (#11)

* ggml : add fused swiglu_oai op

* Update ggml/src/ggml-cpu/ops.cpp

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

* update CUDA impl

* cont : metal impl

* add vulkan impl

* test-backend-ops : more test cases, clean up

* llama : remove unfused impl

* remove extra lines

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

---------

Co-authored-by: slaren <slarengh@gmail.com>

* repack mxfp4 upon conversion

* clean up a bit

* enable thinking

* add quick hack to render only some special tokens

* fix bf16 conversion

* remove vocab hack

* webui ok

* support chat parsing for gpt-oss

* fix webui

* direct mapping mxfp4, FINALLY

* force using mxfp4

* properly use lazy tensor

* ggml : add mxfp4

ggml : use e8m0 conversion instead of powf

Co-authored-by: Diego Devesa <slarengh@gmail.com>

change kvalues_mxfp4 table to match e2m1 (#6)

metal : remove quantization for now (not used)

cuda : fix disabled CUDA graphs due to ffn moe bias

vulkan : add support for mxfp4

cont : add cm2 dequant

* ggml : add ggml_add_id (#13)

* ggml : add ggml_add_id

* add cuda impl

* llama : add weight support check for add_id

* perf opt

* add vulkan impl

* rename cuda files

* add metal impl

* allow in-place ggml_add_id

* llama : keep biases on CPU with --cpu-moe

* llama : fix compile error

ggml-ci

* cuda : add fallback for __nv_cvt_e8m0_to_bf16raw

ggml-ci

* cleanup

ggml-ci

* sycl : fix supports_op for MXFP4

ggml-ci

* fix Unknown reasoning format

* ggml-cpu : fix AVX build

ggml-ci

* fix hip build

ggml-ci

* cuda : add mxfp4 dequantization support for cuBLAS

ggml-ci

* ggml-cpu : fix mxfp4 fallback definitions for some architectures

ggml-ci

* cuda : fix version required for __nv_cvt_e8m0_to_bf16raw

---------

Co-authored-by: Xuan Son Nguyen <son@huggingface.co>
Co-authored-by: slaren <slarengh@gmail.com>

2025-08-05 22:10:36 +03:00

..

.gitignore

tests : gitignore ggml-common.h

2024-03-09 14:17:11 +02:00

CMakeLists.txt

tests : update for LLAMA_SET_ROWS=1 (#14961)

2025-07-30 15:12:02 +03:00

get-model.cpp

ci : add model tests + script wrapper (#4586)

2024-01-26 14:18:00 +02:00

get-model.h

ci : add model tests + script wrapper (#4586)

2024-01-26 14:18:00 +02:00

run-json-schema-to-grammar.mjs

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

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

test-arg-parser.cpp

tests : avoid github urls due to throttling (#13654)

2025-05-20 12:03:17 +02:00

test-autorelease.cpp

llama : add llama_vocab, functions -> methods, naming (#11110)

2025-01-12 11:32:42 +02:00

test-backend-ops.cpp

llama : add gpt-oss (#15091)

2025-08-05 22:10:36 +03:00

test-barrier.cpp

ggml : move CPU backend to a separate file (#10144)

2024-11-03 19:34:08 +01:00

test-c.c

ggml : remove kompute backend (#14501)

2025-07-03 07:48:32 +03:00

test-chat-parser.cpp

tests : remove json.hpp from a test (#13880)

2025-05-29 12:17:16 +03:00

test-chat-template.cpp

chat : only remove double bos/eos if added (#15086)

2025-08-05 20:43:36 +02:00

test-chat.cpp

chat : fix multiple tool_calls on hermes-2-pro (#14962)

2025-08-02 18:04:48 +08:00

test-double-float.cpp

ggml : minor naming changes (#8433)

2024-07-12 10:46:02 +03:00

test-gbnf-validator.cpp

cmake : do not include ./src as public for libllama (#13062)

2025-04-24 16:00:10 +03:00

test-gguf.cpp

gguf: fix failure on version == 0 (#13956)

2025-06-01 18:08:05 +02:00

test-grammar-integration.cpp

sync : vendor (#13901)

2025-05-30 16:25:45 +03:00

test-grammar-llguidance.cpp

cmake : do not include ./src as public for libllama (#13062)

2025-04-24 16:00:10 +03:00

test-grammar-parser.cpp

cmake : do not include ./src as public for libllama (#13062)

2025-04-24 16:00:10 +03:00

test-json-partial.cpp

server: streaming of tool calls and thoughts when --jinja is on (#12379)

2025-05-25 01:48:08 +01:00

test-json-schema-to-grammar.cpp

sync : vendor (#13901)

2025-05-30 16:25:45 +03:00

test-llama-grammar.cpp

cmake : do not include ./src as public for libllama (#13062)

2025-04-24 16:00:10 +03:00

test-log.cpp

common : use common_ prefix for common library functions (#9805)

2024-10-10 22:57:42 +02:00

test-lora-conversion-inference.sh

scripts : make the shell scripts cross-platform (#14341)

2025-06-30 10:17:18 +02:00

test-model-load-cancel.cpp

llama : update llama_model API names (#11063)

2025-01-06 10:55:18 +02:00

test-mtmd-c-api.c

mtmd : add C public API (#13184)

2025-05-04 23:43:42 +02:00

test-opt.cpp

llama/ggml: add LLM training support (#10544)

2025-05-12 14:44:49 +02:00

test-quantize-fns.cpp

tests : fix test-quantize-fns to init the CPU backend (#12306)

2025-03-10 14:07:15 +02:00

test-quantize-perf.cpp

ggml : inttypes.h -> cinttypes (#0)

2024-11-17 08:30:29 +02:00

test-quantize-stats.cpp

docker : do not build tests (#13204)

2025-04-30 10:44:07 +02:00

test-regex-partial.cpp

common: add partial regex support (#12808)

2025-05-14 19:50:57 +01:00

test-rope.cpp

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

2024-12-14 14:43:46 +02:00

test-sampling.cpp

sampling : make sure samplers return at least 1 token (#13822)

2025-05-27 12:07:52 +03:00

test-thread-safety.cpp

tests : update for LLAMA_SET_ROWS=1 (#14961)

2025-07-30 15:12:02 +03:00

test-tokenizer-0.cpp

llama : add llama_vocab, functions -> methods, naming (#11110)

2025-01-12 11:32:42 +02:00

test-tokenizer-0.py

py : logging and flake8 suppression refactoring (#7081)

2024-05-05 08:07:48 +03:00

test-tokenizer-0.sh

scripts : make the shell scripts cross-platform (#14341)

2025-06-30 10:17:18 +02:00

test-tokenizer-1-bpe.cpp

cmake : do not include ./src as public for libllama (#13062)

2025-04-24 16:00:10 +03:00

test-tokenizer-1-spm.cpp

cmake : do not include ./src as public for libllama (#13062)

2025-04-24 16:00:10 +03:00

test-tokenizer-random.py

llama : add llama_vocab, functions -> methods, naming (#11110)

2025-01-12 11:32:42 +02:00

test-tokenizers-repo.sh

scripts : make the shell scripts cross-platform (#14341)

2025-06-30 10:17:18 +02:00