47 Commits
Author SHA1 Message Date
Tarek DakhranandXuan Son Nguyen d4f1b18b3a mtmd: enable video input support for LFM3 model family (#31283)
* mtmd : add LFM3-VL and LFM3-Omni video input

Co-authored-by: Xuan Son Nguyen <son@huggingface.co>
Assisted-by: Claude Opus 5.5

* mtmd : reserve the soundtrack command to avoid a GCC 11 warning

---------

Co-authored-by: Xuan Son Nguyen <son@huggingface.co>
2026-10-28 12:23:06 +01:00
Tarek DakhranandXuan Son Nguyen 19ff817fb2 model : add LiquidAI/LFM3-Audio-2B and LiquidAI/LFM3-Omni-2B (#31257)
* model : add LiquidAI/LFM3-Audio-2B and LiquidAI/LFM3-Omni-2B

Co-authored-by: Xuan Son Nguyen <son@huggingface.co>
Assisted-by: Claude Opus 5.5

* convert : read the LFM3-Audio config.json as released

A transformers that knows lfm3_audio rewrites the config into another layout (no lfm/encoder keys).

Assisted-by: Claude Opus 5.5

* mtmd : disable flash attention for the LFM3-Audio depthformer

Assisted-by: Claude Opus 5.5

* tts : add a voice option for LFM3-Audio and LFM3-Omni

The model card offers a female and a male voice; --tts-voice for llama-tts, "voice" for the server /tts.

Assisted-by: Claude Opus 5.5

* clean up template handling

* convert : detect LFM3-Audio and LFM3-Omni by model_type

Assisted-by: Claude Opus 5.5

* move banned tokens --> logits bias --> mtmd_helper

* leave mtmd-cli unchanged

* no more template_asr_system

* left over

* llama : keep the LFM3-Audio encoder convs and codec unquantized

Assisted-by: Claude Opus 5.5

* common, mtmd : shorten comments

Assisted-by: Claude Opus 5.5

---------

Co-authored-by: Xuan Son Nguyen <son@huggingface.co>
2026-10-27 18:47:14 +01:00
Hans 9ae73e68f5 mtmd : add Soprano TTS support (#28607)
Add Soprano-80M and Soprano-1.1-80M (https://huggingface.co/ekwek/Soprano-1.1-80M)
as an audio generation model: a Qwen3 backbone plus a Vocos decoder that runs
as a gen-only mmproj.

- convert: detect Soprano checkpoints and export the backbone as qwen3 and the
  decoder.pth Vocos weights as the mmproj
- vocab: add the soprano pre-tokenizer (lowercase, whitespace collapse,
  per-digit split, [UNK] fallback)
- mtmd: add the Vocos graph and ISTFT, allow loading a gen-only mmproj for
  Soprano and include the gen-audio context in mtmd_get_memory_usage
- mtmd-helper: add the Soprano pipeline; the decoder runs over the whole
  utterance, so stream mode sends the audio in one chunk after flush()
- tts: document Soprano usage

Assisted-by: Codex
Assisted-by: Claude Opus 5.5
2026-10-24 01:09:50 +02:00
Xuan-Son NguyenandPascal d814cbb88c server: realtime API: send complete after committed audio drains (#30994)
* server: realtime API: send complete after committed audio drains

* server: realtime API: complete a commit at its end row

Completion is triggered when the stream runs dry, which is a timing
proxy: audio appended while the tail drains leaks into the committed
item, a final commit before the drain merges both items, and back to
back commits collapse into one.

mtmd_audio_stream_flush() records the row count at the end of the
flushed audio (mtmd_audio_stream_get_n_rows_flushed), next to the rows
produced so far (mtmd_audio_stream_get_n_rows). The slot keeps a FIFO
of these end rows and completes each commit once every row up to it is
decoded.

---------

Co-authored-by: Pascal <admin@serveurperso.com>
2026-10-23 00:37:47 +02:00
Xuan-Son Nguyen 0a1ef309cd mtmd: add state object for mtmd_gen (#30902) 2026-10-21 12:59:54 +02:00
Xuan-Son NguyenandPascal 5c0b49a23b server : add POST /tts endpoint for TTS models (#26603)
* mtmd: add audio out stream api

* add /tts endpoint

* add docs

* wire up mtmd_helper_model_can_chat

* upload speaker_ref via form-data

* expose seed to api

* fix header for pcm

* wip

* fix penalty_last_n

* fix server_output_limits

* fix pocket-tts stream

* wip

* add server_schema for tts

* add server routing

* server: coerce seed and stream from multipart form fields

Form fields arrive as strings, so -F seed=42 and -F stream=true were
rejected by the schema. seed joins the integer fields, parsed as a long
long so a full uint32 seed fits, and stream is read as a boolean when
it comes as a string.

---------

Co-authored-by: Pascal <admin@serveurperso.com>
2026-10-21 01:19:42 +02:00
Xuan-Son Nguyen 2851b3bcbf mtmd: add mtmd_audio_stream + support voxtral-realtime (#30629)
* mtmd: add mtmd_audio_stream + support voxtral-realtime

* address review
2026-10-16 01:46:42 +02:00
Xuan-Son Nguyen 552f18f912 mtmd: cap max_image to ubatch for non_causal models (#29773) 2026-10-01 11:55:11 +02:00
Mads Marquart f45576aa86 mtmd : add const in various places (#28307)
* mtmd : mark context as const in more methods

Mark `mtmd_context` as `const` in:
- mtmd_bitmap_init_lazy
- mtmd_tokenize
- mtmd_tokenize_from_parts
- mtmd_helper_support_video
- mtmd_helper_bitmap_init_from_file
- mtmd_helper_bitmap_init_from_buf
- mtmd_helper_video_init
- mtmd_helper_video_init_from_buf
- mtmd_helper_model_can_chat

The tokenization functions in particular are useful to have marked
`const`, as that allows more easily telling the compiler that we can
safely tokenize from multiple threads (`mtmd_tokenize` is already
documented as thread-safe, this just reifies that in the signature).

* mtmd : mark tokenization input pointer as const

Mark the `bitmaps` and `parts` pointers in `mtmd_tokenize` and
`mtmd_tokenize_from_parts` as `const`. This allows more easily calling
these with immutable arrays / vectors.

* mtmd : mark llama_context as const in mtmd_helper_model_can_chat
2026-09-03 12:12:49 +02:00
Xuan-Son Nguyen 7339054744 mtmd: add mtmd_tokenize_from_parts() (#28250)
* add mtmd_tokenize_from_parts

* use it in mtmd-cli

* move add_special to call level
2026-09-02 21:20:10 +02:00
John-Henry LimandXuan Son Nguyen 681c29d36a mtmd: add --mmproj-device argument (#23255)
* feat: add --mmproj-device arg & backwards compatible MTMD_BACKEND_DEVICE env var

* feat: load mmproj device backend immediately, add -mmdev shortflag

* fix: its a pointer now get the name

* clean up

* gen docs

* nits

---------

Co-authored-by: Xuan Son Nguyen <son@huggingface.co>
2026-08-20 18:45:37 +02:00
Xuan-Son Nguyen 95c409c136 mtmd: add mtmd_bitmap_set_mergeable (#27348) 2026-08-19 13:48:22 +02:00
Xuan-Son Nguyen 533b18257b server: save processed mtmd chunks as placeholder (#27278) 2026-08-17 21:29:50 +02:00
Xuan-Son NguyenandPascal 6e62ba5384 mtmd: support pocket-tts (#26871)
* adapt the api

* text model ok

* working impl, need verify and clean up

* mtmd: build the pocket-tts transposed convolutions as GEMM + col2im

ggml_conv_transpose_1d has no grouped mode, so the depthwise upsample
was built as one convolution and one concat per channel, which floods
the graph with small nodes and makes kernel launches dominate the
decoder.

Fold both cases into the column form the seanet decoder already needs:
the general case reshapes the kernel to [IC, K * OC] and matmuls it
with the input, the depthwise case batches a matmul over the channels
so a step scales its own kernel. A single col2im_1d then scatter-adds
the columns back to the signal, with the same shape as before, so the
overlap-add tail, the streaming state and the bias are untouched.

Generation time per frame drops by 80% on CUDA and by 50% on CPU. The
output matches the previous implementation sample for sample, with a
correlation of 0.999994 and identical frame counts.

* flow_temp +  frames_after_eos

* chunking

* mtmd: carry the remaining pocket-tts per-pack settings

The language packs also tune the end-of-speech padding and the padding
of short prompts, next to the temperature already carried in the
mmproj: french_24l asks for 8 tail frames instead of the guessed 3,
english_2026-01 asks for short prompts to be padded with spaces.

Write both in the mmproj as clip.gen.audio.frames_after_eos and
clip.gen.audio.pad_short_text, keyed on the pack in the conversion
script like the temperature. The loader keeps them optional, so a
mmproj without them behaves as before. Map semicolons to commas for
every pack instead, the reference only asks for it on three of them and
it costs nothing elsewhere.

Existing mmproj files must be converted again to carry the two keys.

On a long french text the port now lands within 2% of the reference:
22.96s against 23.44s, with the same peak level and the same amount of
silence.

* clip.gen.audio.model_variant

* clean up code comments

* nit: drop the dead flow_temp hparam, the pack table holds the default

* update docs

* address security problems

* less invasive base.py

* lint

* add mtmd_gen_inp_default

* add docs

* rm gen_flow_temp

---------

Co-authored-by: Pascal <admin@serveurperso.com>
2026-08-11 14:18:30 +02:00
Xuan-Son Nguyen 15586e2d71 mtmd: add chunk save/load function (#26645)
* mtmd: add chunk save/load function

* nits

* add tests

* rn _MAX --> _COUNT
2026-08-06 19:46:40 +02:00
Xuan-Son NguyenandPascal 0713275082 mtmd: support Qwen3-TTS (note: breaking change to llama-tts binary) (#26254)
* convert text model

* main model load ok

* convert encoder ok

* speaker encoder loading ok

* speaker enc graph

* adapt vocab for backbone (with some tricks)

* add suppress_tokens

* poc new mtmd gen api

* convert code_predictor to gguf

* load gen_code model ok

* add clip_encode

* wire up

* code gen cgraph init version

Co-authored-by: Pascal <admin@serveurperso.com>

* code2wav convert to gguf

* code2wav graph ok

* wire up in/out

* (wip) subgraph

* wire up

* wip, correct code2wav

* demo (to be removed)

* code2wav preserve kv between calls

* demo voice clone

* llama: add llama_model_get_tok_embd

* mtmd_helper_gen_audio API

* fix clamp cold prefix

Co-authored-by: Pascal <admin@serveurperso.com>

* fuse snake op

Co-authored-by: Pascal <admin@serveurperso.com>

* demo: use proper sampling

* update dev docs

* polymorphism helper

* revamp llama-tts binary

* update docs

* fix compile

* fix lint

* nits

* add guide + docs

* more timings info

* clean up code comments

* security fixes

* update docs

* use ggml_build_forward_select, clean up comments

* fix ci

* use ISO 639-1 language code

* rename CODE2WAV --> GEN_WAV, update docs

* clean up

* clean up tts.cpp

* add seq_id

* add step_prompt()

* mtmd_helper_model_can_chat

* clean up comments

---------

Co-authored-by: Pascal <admin@serveurperso.com>
2026-08-04 17:26:15 +02:00
PascalandXuan Son Nguyen 4114ba18b2 mtmd: fix silent prompt truncation on embedded NUL (#25548)
* mtmd: fix silent prompt truncation on embedded NUL

mtmd_input_text carried the prompt as a bare const char* with no
length, so a NUL byte in message content cut the prompt at the
tokenizer boundary and dropped every later message plus the assistant
marker, with no log. Add an explicit text_len and thread it through,
matching llama_tokenize and the text only path.

* cleanup

---------

Co-authored-by: Xuan Son Nguyen <son@huggingface.co>
2026-07-13 00:47:25 +02:00
Xuan-Son Nguyen 2f89acc2bc mtmd: add load progress callback (#24865) 2026-06-21 13:40:52 +02:00
Xuan-Son Nguyen e37abd6b5f mtmd: add batching API (#24384)
* mtmd: add batching API

* wip

* first working version (gemma4v)

* add arg

* nits

* wire up support_batch()

* fix 0.0 output embd

* fix audio

* nits

* refactor a bit

* nits

* fix non-batching case

* fix comment
2026-06-13 00:10:29 +02:00
Xuan-Son Nguyen 8f83d6c271 mtmd : add video input support (#24269)
* wip

* ok: lazy bitmap API

* remember to free lazy text

* wip

* add mtmd_helper_video

* support video input on server (base64 input)

* add MTMD_VIDEO config

* add timestamp

* update CLI

* cli: allow auto-completion for video

* add --video arg

* fix build

* update docs

* rename as suggested
2026-06-08 14:40:12 +03:00
Xuan-Son Nguyenandandrewmd5 31e82494c0 mtmd: support "frame merge" for qwen-vl-based models (#21858)
* feat: add video support for Qwen3.5

* various clean up

* revise the design

* fix llava-uhd case

* nits

* nits 2

---------

Co-authored-by: andrewmd5 <1297077+andrewmd5@users.noreply.github.com>
2026-06-06 21:17:25 +02:00
Xuan-Son Nguyen f5c6ae1827 mtmd, server: add "placeholder bitmap" for counting tokens , add */input_tokens API (#23913)
* mtmd: add "placeholder bitmap" for counting tokens w/o preprocessing

* fast path skip preproc for placeholder

* fix build

* correct the api

* add server endpoint + tests

* add object name

* update docs

* add proxy handling

* fix build

* fix audio input path

* use is_placeholder in process_mtmd_prompt()

* nits

* nits (2)

* docs: clarify chat/completions/input_tokens is not official

* fix merge problem
2026-06-06 11:06:51 +02:00
Xuan-Son Nguyen e2b129e1bf mtmd: fit_params now take into account mmproj (#21489)
* mtmd: fit_params now take into account mmproj

* rename alloc_compute_meta to reserve_compute_meta

* rm unused functions

* add ggml_backend_dev_t support

* add debug log
2026-05-20 11:27:44 +02:00
Xuan-Son Nguyen 7bfe120c21 mtmd, server, common: expose modalities to /v1/models (#22952)
* mtmd, server, common: expose modalities to /v1/models

* fix build

* rename to mtmd_caps
2026-05-12 19:08:07 +02:00
Xuan-Son Nguyen 86f8daacfe mtmd: correct get_n_pos / get_decoder_pos (#22175) 2026-04-20 23:29:19 +02:00
Xuan-Son Nguyen 19124078be mtmd: add pos_0 to mtmd_image_tokens_get_decoder_pos (breaking change) (#22082)
* mtmd: add pos_0 to mtmd_image_tokens_get_decoder_pos

* fix build
2026-04-19 11:57:21 +02:00
Xuan-Son Nguyen 408225bb1a server: use random media marker (#21962)
* server: use random media marker

* nits

* remove legacy <__image__> token

* revert special char in random
2026-04-15 23:52:22 +02:00
Xuan-Son Nguyen 707c0b7a6e mtmd: add mtmd_image_tokens_get_decoder_pos() API (#21851)
* mtmd: add mtmd_image_tokens_get_decoder_pos() API

* consistent naming

* fix build
2026-04-14 16:07:41 +02:00
Xuan-Son Nguyen 920b3e78cb mtmd: use causal attn for gemma 4 audio (#21824) 2026-04-13 09:47:55 +02:00
Daniel Bevenius 8f974d2392 mtmd : rename mtmd_get_audio_bitrate to mtmd_get_audio_sample_rate (#20105)
This commit renames the the function `mtmd_get_audio_bitrate` to
`mtmd_get_audio_sample_rate` to better reflect its purpose.

The motivation for this is that the function currently returns the audio
sample rate, not the bitrate (sample_rate × bit_depth × channels), and
that is how it is used in the code as well.

This is a breaking change, but I believe mtmd is still in
experimental/development phase so it might be alright to simply rename.
2026-03-13 12:30:02 +01:00
Xuan-Son Nguyen 9eb5bfec1a mtmd : update docs to use llama_model_n_embd_inp (#18999) 2026-01-22 14:36:32 +01:00
Piotr Wilkin (ilintar)andXuan-Son Nguyen d98b548120 Restore clip's cb() to its rightful glory - extract common debugging elements in llama (#17914)
* Extract common debugging functions; plug eval-callback and mtmd's MTMD_DEBUG_GRAPH with same functionality

* Move to common

* Remove unneeded header

* Unlink from common

* chore: update webui build output

* Cleanup; properly pass params to mtmd without depending on common; factorize debug.cpp to use common debug code.

* Revert change to webapp

* Post-merge adjust

* Apply suggestions from code review

Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>

* Apply code review changes

* Remove changes to server-context

* Remove mtmd.h include

* Remove utility functions from header

* Apply suggestions from code review

Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>

* Rename functions

* Update tools/mtmd/clip.cpp

Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>

* Update tools/mtmd/clip.cpp

Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>

* Update tools/mtmd/clip.cpp

Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>

---------

Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>
2026-01-14 20:29:35 +01:00
Xuan-Son Nguyen cffa5c46ea mtmd: clarify that we no longer accept AI-generated PRs (#18406) 2025-12-28 09:57:04 +01:00
Xuan-Son Nguyen e39a2ce66d clip: move model cgraphs into their own files (#17965)
* clip: move model cgraphs into their own files

* more explicit enums

* fix linux build

* fix naming

* missing headers

* nits: add comments for contributors
2025-12-12 21:14:48 +01:00
Xuan-Son Nguyen ecf74a8417 mtmd: add mtmd_context_params::warmup option (#17652)
* mtmd: add mtmd_context_params::warmup option

* reuse the common_params::warmup
2025-12-01 21:32:25 +01:00
Xuan-Son Nguyen 9b17d74ab7 mtmd: add mtmd_log_set (#17268) 2025-11-14 15:56:19 +01:00
Xuan-Son Nguyen 070ff4d535 mtmd: add --image-min/max-tokens (#16921) 2025-11-03 11:11:18 +01:00
Georgi GerganovandXuan Son Nguyen 2f966b8ed8 clip : use FA (#16837)
* clip : use FA

* cont : add warning about unsupported ops

* implement "auto" mode for clip flash attn

* clip : print more detailed op support info during warmup

* cont : remove obsolete comment [no ci]

* improve debugging message

* trailing space

* metal : remove stray return

---------

Co-authored-by: Xuan Son Nguyen <son@huggingface.co>
2025-11-02 21:21:48 +01:00
Xuan-Son NguyenandGeorgi Gerganov e3af5563bd llama: store mrope data in KV cell (#16825)
* llama: store mrope data in KV cell

* correct x,y ordering

* address review comments

* add consistency checks

* Update src/llama-kv-cache.cpp

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

* add TODO

* fix asan error

* kv-cells : improve ext handling

* cont : fix headers

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2025-10-29 18:09:18 +01:00
Xuan-Son Nguyen 51fa76f172 mtmd : drop _shared from libmtmd name, merge helpers into libmtmd (⚠️ breaking change) (#13917)
* mtmd : fix missing public header

* no object

* apply suggestion from Georgi

* rm mtmd-helper, merge it to mtmd

* missing vendor include dir
2025-05-31 10:14:29 +02:00
Xuan-Son Nguyen 10961339b2 mtmd : move helpers to dedicated library (⚠️ breaking change) (#13866)
* mtmd : move helpers to dedicated library

* fix server build

* rm leftover cmakelist code
2025-05-28 22:35:22 +02:00
Xuan-Son Nguyen 40aaa8a403 mtmd : add support for Qwen2-Audio and SeaLLM-Audio (#13760)
* mtmd : add Qwen2-Audio support

* small clean up

* update discussion link

* clarify mtmd_get_output_embd

* clarification in multimodal.md

* fix ultravox bug

* ggml_cont
2025-05-25 14:06:32 +02:00
Xuan-Son Nguyen 9ecf3e66a3 server : support audio input (#13714)
* server : support audio input

* add audio support on webui
2025-05-23 11:03:47 +02:00
Xuan-Son Nguyen 797990c4bc mtmd : add ultravox audio input (#13623)
* convert ok, load ok

* warmup ok

* test

* still does not work?

* fix padding

* temporary give up

* fix merge conflict

* build_ultravox()

* rm test

* fix merge conflict

* add necessary mtmd APIs

* first working version (only 4s of audio)

* will this monster compile?

* fix compile

* please compile

* fPIC

* fix windows

* various fixes

* clean up audio_helpers

* fix conversion

* add some debug stuff

* long audio input ok

* adapt the api

* add --audio arg

* final touch UX

* add miniaudio to readme

* fix typo

* refactor kv metadata

* mtmd_default_marker()
2025-05-22 20:42:48 +02:00
Xuan-Son Nguyen a634d75d1b mtmd : move helpers to dedicated file (#13442)
* mtmd : move helpers to dedicated file

* fix windows build

* rm redundant include
2025-05-11 11:34:23 +02:00
Matt Clayton f05a6d71a0 mtmd : Expose helper_decode_image_chunk (#13366)
* mtmd: Expose helper_decode_image, output_embd_copy, image_tokens_copy/free

* Slim down

* Cleanups
2025-05-08 20:25:39 +02:00
Xuan-Son Nguyen 9b61acf060 mtmd : rename llava directory to mtmd (#13311)
* mv llava to mtmd

* change ref everywhere
2025-05-05 16:02:55 +02:00