* added cli arg to disable context shift
* reverted precommit
* updated README.md for main
* white space
* allow disabling context shift in the server
* Update common/arg.cpp
no-context-shift only works for main example
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* added server example to --no-context-shift args
* removed server changes
* white space
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* Adding loading page for '/' server requests
* set content when model is loading
* removed loading html file
* updated cmakelist
* updated makefile
* cleaned up whitespace
* cleanup for PR removed error
* updated server test to handle 503 HTML
* updated server test to handle 503 HTML
* ca†ch 503 before parsing json
* revert test
* account for both api and web browser requests
* precommit corrections
* eol fix
* revert changes to pre-commit
* removed print statement
* made loading message more descriptive
* also support .html files
---------
Co-authored-by: VJHack <flymyplane21@gmail.com>
Co-authored-by: Vinesh Janarthanan <36610342+VJHack@users.noreply.github.com>
* llama : llama_perf + option to disable timings during decode
ggml-ci
* common : add llama_arg
* Update src/llama.cpp
Co-authored-by: Xuan Son Nguyen <thichthat@gmail.com>
* perf : separate functions in the API
ggml-ci
* perf : safer pointer handling + naming update
ggml-ci
* minor : better local var name
* perf : abort on invalid sampler pointer
ggml-ci
---------
Co-authored-by: Xuan Son Nguyen <thichthat@gmail.com>
* server : added with_pieces functionality to /tokenize endpoint
* server : Add tokenize with pieces tests to server.feature
* Handle case if tokenizer splits along utf8 continuation bytes
* Add example of token splitting
* Remove trailing ws
* Fix trailing ws
* Maybe fix ci
* maybe this fix windows ci?
---------
Co-authored-by: Xuan Son Nguyen <son@huggingface.co>
* Arm AArch64: Documentation updates
* Update docs/build.md to include information on how to enable the Arm optimized gemm/gemv kernels
* Update examples/quantize/README.md with information on the Q4_0_4_4, Q4_0_4_8 and Q4_0_8_8 formats
* Add newline to the end of docs/build.md
* common : do not add null tokens during warmup
ggml-ci
* llama : check that the input tokens are valid
ggml-ci
* tests : fix batch size of bert model
ggml-ci
- Add `struct llama_sampler` and `struct llama_sampler_i`
- Add `llama_sampler_` API
- Add `llama_sampler_chain_` API for chaining multiple samplers
- Remove `LLAMA_API_INTERNAL`
- Add `llama_perf_` API and remove old `llama_print_timings` and `llama_reset_timings`
* ggml-quants : 1.625 bpw ternary packing for BitNet 1.58b
* ggml-quants : faster 1.625 bpw AVX2 vec_dot
Not using a lookup table anymore makes it match q4_0 speed.
* gguf-py : fix formatting
* llama : remove spaces on empty line
* ggml-quants : subtract 1 when back in epi8
This makes the 1.625 bpw type go faster than q4_0. Still not the fastest.
* ggml-quants : Q2_2 now faster than Q4_K on with AVX2
* ggml-quants : cleanup Q1_3 code formatting
* ggml-quants : ARM NEON vec_dot for q2_2 and q1_3
* ggml-quants : use ceiling division when quantizing q1_3
* convert-hf : simplify BitNet pre-quantization
This still results in the exact same tensor weights and scales,
but it reveals some weirdness in the current algorithm.
* convert-hf : allow converting the weird BitNet 1.3B
Its FFN size is 5460 which is not convenient.
The offending tensors are kept in F16,
which makes the final model 5.01 bpw.
* bitnet : replace 1.58b with b1.58, as in the paper
* ggml-quants : fix build failure on Windows
* ggml-quants : attempt to fix Arm 32-bit support
* ggml : add some informative comments in q1_3 vec_dot
* ggml : add TQ1_0 and TQ2_0 ternary quantization types
* ggml : even faster TQ2_0
* ggml : also faster TQ1_0
Same optimization as for TQ2_0 by offsetting the sum instead of the weights.
This makes TQ1_0 almost as fast as Q8_0 on AVX2.
* ggml : fix build issues in certain environments
* ggml : add NEON vec_dot implementation for TQ1_0 and TQ2_0
* ggml : avoid directly using vmlal_high_s8, for 32-bit ARM compat
The compiler seems smart enough to use the same instruction
even when using vget_high_s8 instead.
* ggml : remove q1_3 and q2_2
No more 1.625 bpw and 2.000 bpw,
now instead using 1.6875 bpw and 2.0625 bpw
with TQ1_0 and TQ2_0, respectively.
* llama : remove the separate scale tensors of BitNet b1.58
They won't be needed, since the remaining ternary quant types have
built-in scales.
* ggml-quants : rename fields of TQ1_0 and TQ2_0 structs for consistency
* ggml-quants : allow using vdotq_s32 in TQ2_0 vec_dot
Not yet tested on hardware which supports it,
might not work or might not even compile. But also it might.
It should make the performance better on recent ARM CPUs.
* ggml-quants : remove comment about possible format change of TQ2_0
Making it slightly more convenient for AVX512
but less convenient for everything else is not worth the trouble.
* gguf-py : Numpy (de)quantization for TQ1_0 and TQ2_0
* ggml-quants : use roundf instead of nearest_int for TQ1_0 and TQ2_0
This does not change anything for ternary models,
since their values should never end up being in halfway cases anyway.
* convert : allow direct conversion to TQ1_0 and TQ2_0
The token embeddings and output tensors are kept in F16
to allow quantizing them to Q4_K and Q6_K with llama-quantize.
* llama : handle fallback for TQ1_0 and TQ2_0 with Q4_0
Q4_0 is not completely symmetric (so not lossless for ternary models),
but it should be good enough.
* ggml-quants : allow using ARM dot product instructions for TQ1_0
* ggml-quants : deduplicate TQ1_0 and TQ2_0 __ARM_FEATURE_DOTPROD support
* ggml : remove unused ggml_mul special case
It would otherwise conflict with the more general
optimization coming with Mamba-2.
* ggml : handle TQ1_0 and TQ2_0 in dequantization-based operators
* test-backend-ops : add TQ1_0 and TQ2_0 comments for later
Not yet adding uncommented, because some backends like SYCL and Metal
do not properly handle unknown types in supports_op for GGML_OP_MUL_MAT.
(and Metal also doesn't handle it with GGML_OP_GET_ROWS)
Support for TQ1_0 and TQ2_0 for other backends than CPU
will be added in follow-up pull requests.
* server : remove multitask from server_task
* refactor completions handler
* fix embeddings
* use res_ok everywhere
* small change for handle_slots_action
* use unordered_set everywhere
* (try) fix test
* no more "mutable" lambda
* Apply suggestions from code review
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* use deque
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* Introduce ggml_compute_threadpool
- OpenMP functional: check
- Vanilla ggml functional: Check
- ggml w/threadpool functional: Check
- OpenMP no regression: No glaring problems
- Vanilla ggml no regression: No glaring problems
- ggml w/threadpool no regression: No glaring problems
* Minor fixes
* fixed use after release bug
* fixed a harmless race condition
* Fix Android bulid issue
* fix more race conditions
* fix deadlock for cases where cgraph.n_nodes == 1
and fix --poll case
* threadpool: use cpu_get_num_math to set the default number of threadpool threads
This way we avoid using E-Cores and Hyperthreaded siblings.
* bench: create fresh threadpool for each test
For benchmarking it's better to start a fresh pool for each test with the exact number of threads
needed for that test. Having larger pools is suboptimal (causes more load, etc).
* atomics: always use stdatomics with clang and use relaxed memory order when polling in ggml_barrier
This also removes sched_yield() calls from ggml_barrier() to match OpenMP behavior.
* threadpool: make polling the default to match openmp behavior
All command line args now allow for setting poll to 0 (false).
* threadpool: do not wakeup threads in already paused threadpool
* fix potential race condition in check_for_work
* threadpool: do not create two threadpools if their params are identical
* threadpool: reduce pause/resume/wakeup overhead in common cases
We now start threadpool in paused state only if we have two.
The resume is now implicit (ie new work) which allows for reduced locking and context-switch overhead.
* threadpool: add support for hybrid polling
poll params (--poll, ...) now specify "polling level", i.e. how aggresively we poll before waiting on cond.var.
poll=0 means no polling, 1 means poll for 128K rounds then wait, 2 for 256K rounds, ...
The default value of 50 (ie 50x128K rounds) seems like a decent default across modern platforms.
We can tune this further as things evolve.
* threadpool: reduce the number of barrier required
New work is now indicated with an atomic counter that is incremented for
each new graph that needs to be computed.
This removes the need for extra barrier for clearing the "new_work" and
removes the special case for trivial graphs.
* threadpool: remove special-casing for disposable threadpools
With the efficient hybrid polling there is no need to make disposable pools any different.
This simplifies the overall logic and reduces branching.
Include n_threads in debug print for disposable threadpool.
Declare pause and stop flags as atomic_bool
This doesn't actually generate any memory barriers and simply informs
the thread sanitizer that these flags can be written & read by different
threads without locking.
* threadpool: do not clear barrier counters between graphs computes (fixes race with small graphs)
This fixes the race condition with very small graphs where the main thread happens to
start a new graph while the workers are just about to exit from barriers.
* threadpool: use relaxed order for chunk sync
Full memory barrier is an overkill for this since each thread works on different chunk
* threadpool: remove abort_callback from threadpool state
* threadpool: better naming for thread/cpumask releated functions
* threadpool: consistent use of int type for n_threads params
* threadpool: add support for ggml_threadpool_params_default/init
Also removes the need for explicit mask_specified param.
all-zero cpumask means use default (usually inherited) cpu affinity mask.
* threadpool: move typedef into ggml.h
* threadpool: fix apply_priority() function name
* threadpool: fix swift wrapper errors due to n_threads int type cleanup
* threadpool: enable --cpu-mask and other threadpool related options only if threadpool is enabled
* threadpool: replace checks for compute_thread ret code with proper status check
* threadpool: simplify threadpool init logic and fix main thread affinity application
Most of the init code is now exactly the same between threadpool and openmp.
* threadpool: update threadpool resume/pause function names
* threadpool: enable openmp by default for now
* threadpool: don't forget to free workers state when omp is enabled
* threadpool: avoid updating process priority on the platforms that do not require it
On Windows we need to change overall process priority class in order to set thread priorities,
but on Linux, Mac, etc we do not need to touch the overall process settings.
* threadpool: update calling thread prio and affinity only at start/resume
This avoids extra syscalls for each graph_compute()
* llama-bench: turn threadpool params into vectors, add output headers, etc
* llama-bench: add support for cool off between tests --delay
This helps for long running tests on platforms that are thermally limited (phones, laptops, etc).
--delay (disabled by default) introduces the sleep for N seconds before starting each test.
* threadpool: move process priority setting into the apps (bench and cli)
This avoids changing the overall process priority on Windows for the apps
that use ggml/llama.cpp directy.
* threadpool: move all pause/resume logic into ggml
* threadpool: futher api cleanup and prep for future refactoring
All threadpool related functions and structs use ggml_threadpool prefix.
* threadpool: minor indent fixes
* threadpool: improve setprioty error message
* Update examples/llama-bench/llama-bench.cpp
Co-authored-by: slaren <slarengh@gmail.com>
* threadpool: fix indent in set_threadpool call
* use int32_t for n_thread type in public llama.cpp API
* threadpool: use _new and _free instead of _create and _release
* fix two more public APIs to use int32_t for n_threads
* build: set _GNU_SOURCE for Adroid
---------
Co-authored-by: Max Krasnyansky <quic_maxk@quicinc.com>
Co-authored-by: fmz <quic_fzaghlou@quic.com>
Co-authored-by: Max Krasnyansky <max.krasnyansky@gmail.com>
Co-authored-by: slaren <slarengh@gmail.com>
This change fixes a bug where replacing text in a very long string could
cause llama.cpp to hang indefinitely. This is because the algorithm used
was quadratic, due to memmove() when s.replace() is called in a loop. It
seems most search results and LLM responses actually provide the O(n**2)
algorithm, which is a great tragedy. Using a builder string fixes things
* llava: Add ACC OP for GPU acceleration to the Vulkan backend in the LLAVA CLIP model.
- The CLIP model now prioritizes the Vulkan backend over the CPU when vulkan available.
- A GGML_OP_ACC shader has been added.
- The encoding performance of the CLIP model improved from 4.2s on the CPU to 0.9s on the GPU.
Signed-off-by: Changyeon Kim <cyzero.kim@samsung.com>
* fix-up coding style.
Signed-off-by: Changyeon Kim <cyzero.kim@samsung.com>
* Fix-up the missing initial parameter to resolve the compilation warning.
Signed-off-by: Changyeon Kim <cyzero.kim@samsung.com>
* [fix] Add missing parameters.
Signed-off-by: Changyeon Kim <cyzero.kim@samsung.com>
* [fix] Use nb1 and nb2 for dst.
Signed-off-by: Changyeon Kim <cyzero.kim@samsung.com>
* Fix check results ggml_acc call
---------
Signed-off-by: Changyeon Kim <cyzero.kim@samsung.com>
Co-authored-by: 0cc4m <picard12@live.de>
* server : refactor middleware and /health endpoint
* move "fail_on_no_slot" to /slots
* Update examples/server/server.cpp
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* fix server tests
* fix CI
* update server docs
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* gguf-py : add T5ENCODER model architecture
* common : call llama_decode() during warmup only if the model has decoder
* convert-hf : add T5EncoderModel
* llama : add llama_model_has_decoder() API function
* llama : split build_t5() into build_t5_encoder() and build_t5_decoder()
* llama : add support for LLM_ARCH_T5ENCODER
* llama-embedding : add support for LLAMA_POOLING_TYPE_NONE
* llama-embedding : add support for encoder-only models
---------
Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com>
This commit adds the `--pooling` option to the README.md file in the
`examples/embedding` directory.
The motivation for adding this options is that currently if the model
used does not specify a pooling type the embedding example will fail
with the following error message:
```console
main: error: pooling type NONE not supported
```
This commit also updates the name of the executable in the examples
section.
* common : Changed tuple to struct (TODO fix)
Use struct `llama_init_result` to replace the previous
std::tuple<struct llama_model *, struct llama_context *>
* delete llama_init_default_params()
* delete the extra whitespace
* [example] batched-bench "segmentation fault"
When `llama-batched-bench` is invoked _without_ setting `-npl`, "number
of parallel prompts", it segfaults.
The segfault is caused by invoking `max_element()` on a zero-length
vector, `n_pl`
This commit addresses that by first checking to see if the number of
parallel prompts is zero, and if so sets the maximum sequence size to 1;
otherwise, sets it to the original, the result of `max_element()`.
Fixes, when running `lldb build/bin/llama-batched-bench -- -m models/Meta-Llama-3-8B.gguf`
```
* thread #1, queue = 'com.apple.main-thread', stop reason = EXC_BAD_ACCESS (code=1, address=0x0)
frame #0: 0x000000010000366c llama-batched-bench`main(argc=3, argv=0x000000016fdff268) at batched-bench.cpp:72:28
69 llama_context_params ctx_params = llama_context_params_from_gpt_params(params);
70
71 // ensure enough sequences are available
-> 72 ctx_params.n_seq_max = *std::max_element(n_pl.begin(), n_pl.end());
```
* Update examples/batched-bench/batched-bench.cpp
Co-authored-by: compilade <git@compilade.net>
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Co-authored-by: compilade <git@compilade.net>
* llama : refactor session file management
* llama : saving and restoring state checks for overflow
The size of the buffers should now be given to the functions working
with them, otherwise a truncated file could cause out of bound reads.
* llama : stream from session file instead of copying into a big buffer
Loading session files should no longer cause a memory usage spike.
* llama : llama_state_get_size returns the actual size instead of max
This is a breaking change, but makes that function *much* easier
to keep up to date, and it also makes it reflect the behavior
of llama_state_seq_get_size.
* llama : share code between whole and seq_id-specific state saving
Both session file types now use a more similar format.
* llama : no longer store all hparams in session files
Instead, the model arch name is stored.
The layer count and the embedding dimensions of the KV cache
are still verified when loading.
Storing all the hparams is not necessary.
* llama : fix uint64_t format type
* llama : various integer type cast and format string fixes
Some platforms use "%lu" and others "%llu" for uint64_t.
Not sure how to handle that, so casting to size_t when displaying errors.
* llama : remove _context suffix for llama_data_context
* llama : fix session file loading
llama_state_get_size cannot be used to get the max size anymore.
* llama : more graceful error handling of invalid session files
* llama : remove LLAMA_MAX_RNG_STATE
It's no longer necessary to limit the size of the RNG state,
because the max size of session files is not estimated anymore.
* llama : cast seq_id in comparison with unsigned n_seq_max
Changes:
- Move each example into its own function. This makes the code much
easier to read and understand.
- Make the program easy to only run one test by commenting out function
calls in main().
- Make the output easy to parse by indenting the output for each example.
- Add shebang and +x bit to make it clear it's an executable.
- Make the host configurable via --host with a default 127.0.0.1:8080.
- Make the code look in the tools list to call the registered tool,
instead of hardcoding the returned values. This makes the code more
copy-pastable.
- Add error checking, so that the program exits 1 if the LLM didn't
returned expected values. It's super useful to check for correctness.
Testing:
- Tested with Mistral-7B-Instruct-v0.3 in F16 and Q5_K_M and
Meta-Llama-3-8B-Instruct in F16 and Q5_K_M.
- I did not observe a failure even once in Mistral-7B-Instruct-v0.3.
- Llama-3 failed about a third of the time in example_concurrent: it
only returned one call instead of 3. Even for F16.
Potential follow ups:
- Do not fix the prompt encoding yet. Surprisingly it mostly works even
if the prompt encoding is not model optimized.
- Add chained answer and response.
Test only change.
* fix continuing generating blank lines after getting EOT token or EOS token from LLM
* change variable name to is_done (variable name suggested by ggerganov)
* minor : fix trailing whitespace
* minor : add space
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Main thing is that the default output filename will take this form
{name}{parameters}{finetune}{version}{encoding}{kind}
In addition this add and remove some entries in the KV store and adds a metadata class with automatic heuristics capability to derive some values based on model card content
* No Change:
- Internal GGUF Spec
- `general.architecture`
- `general.quantization_version`
- `general.alignment`
- `general.file_type`
- General Model Details
- `general.name`
- `general.author`
- `general.version`
- `general.description`
- Licensing details
- `general.license`
- Typically represents the converted GGUF repo (Unless made from scratch)
- `general.url`
- Model Source during conversion
- `general.source.url`
* Removed:
- Model Source during conversion
- `general.source.huggingface.repository`
* Added:
- General Model Details
- `general.organization`
- `general.finetune`
- `general.basename`
- `general.quantized_by`
- `general.size_label`
- Licensing details
- `general.license.name`
- `general.license.link`
- Typically represents the converted GGUF repo (Unless made from scratch)
- `general.doi`
- `general.uuid`
- `general.repo_url`
- Model Source during conversion
- `general.source.doi`
- `general.source.uuid`
- `general.source.repo_url`
- Base Model Source
- `general.base_model.count`
- `general.base_model.{id}.name`
- `general.base_model.{id}.author`
- `general.base_model.{id}.version`
- `general.base_model.{id}.organization`
- `general.base_model.{id}.url` (Model Website/Paper)
- `general.base_model.{id}.doi`
- `general.base_model.{id}.uuid`
- `general.base_model.{id}.repo_url` (Model Source Repository (git/svn/etc...))
- Array based KV stores
- `general.tags`
- `general.languages`
- `general.datasets`
---------
Co-authored-by: compilade <git@compilade.net>
Co-authored-by: Xuan Son Nguyen <thichthat@gmail.com>
* [CANN] Add Ascend NPU backend
Ascend is a full-stack AI computing infrastructure for industry
applications and services based on Huawei Ascend processors and
software.
CANN (Compute Architecture of Neural Networks), developped by
Huawei, is a heterogeneous computing architecture for AI.
Co-authored-by: wangshuai09 <391746016@qq.com>
* delete trailing whitespaces
* Modify the code based on review comment
* Rename LLAMA_CANN to GGML_CANN
* Make ggml-common.h private
* add ggml_cann prefix for acl funcs
* Add logging for CANN backend
* Delete Trailing whitespace
---------
Co-authored-by: wangshuai09 <391746016@qq.com>
* Update clib.json to point to Cyan4973 original xxhash
Convinced Cyan4973 to add clib.json directly to his repo, so can now point the clib package directly to him now. Previously pointed to my fork with the clib.json package metadata
https://github.com/Cyan4973/xxHash/pull/954
* gguf-hash: readme update to point to Cyan4973 xxHash repo [no ci]
The --help option on export-lora isn't accepted as valid. The help still gets displayed by default, but the script exits with an error message and nonzero status.
