Windows support and bug fixes for BF16 model #1
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This was referenced May 3, 2024
* Introduce bfloat16 support
Many models on Hugging Face (e.g. Mistral, TinyLLaMA) use bfloat16 as
their canonical floating point format.
┌sign
│
│ ┌exponent
│ │
│ │ ┌mantissa
│ │ │
│┌──┴───┐┌─┴───┐
0b0000000000000000 brain16
This encoding has the same number of exponent bits as float32. That
makes conversion relatively straightforward, even in the absence of
hardware support. For example, converting brain16 to binary32 means
simply shifting 16 bits to the left.
┌sign
│
│ ┌exponent
│ │
│ │ ┌mantissa
│ │ │
│┌──┴───┐┌─┴───────────────────┐
0b00000000000000000000000000000000 IEEE binary32
The issue is that converting bf16 to fp16 can result in information
loss. Only 13% of bf16 numbers can be precisely represented in fp16
which in practice ends up being 99.71% of Mistral 7b v0.2's weights
however there is currently no way other than fp32 to get the others
┌sign
│
│ ┌exponent
│ │
│ │ ┌mantissa
│ │ │
│┌─┴─┐┌─┴──────┐
0b0000000000000000 IEEE binary16
This change fixes that, by adding a bf16 data type to GGML. Support
for CPU inference has been implemented along with optimizations for
the AVX2, AVX512, and AVX512BF16 ISAs. Perplexity on Mistral 7b 0.2
improves somewhere around -0.0024 to -0.0046 compared to using fp16
* Remove GGML code that's not needed
* Minimize the GGML API surface area for BF16
* Remove bf16 luts
* Make the GGML header look nicer
* Fix documentation
* Apply ggerganov's fixes for test-backend-ops
* Add BF16 code for new ggml_validate_row_data() function
* compare-llama-bench.py: add missing basicConfig * compare-llama-bench.py: Add line break between error message and print_help() * Add regular print() markdown table
* Add BPE pre-tokenization for DBRX. * Add vocab GGUFs. * Remove test. * Remove GGUFs.
* Add BPE pre-tokenization for Qwen2. * minor : fixes --------- Co-authored-by: Ren Xuancheng <[email protected]> Co-authored-by: Georgi Gerganov <[email protected]>
…7027) An example of how this might be used in the style of baby-llama will be attached with this PR.
* basic avx implementation * style * combine denibble with load * reduce 256 to 128 (and back!) conversions * sse load * Update sgemm.cpp * oops oops
…org#7078) * fix: use `malloc` instead of `posix_memalign` in `ggml-metal.m` to make it not crash Electron proccesses * fix: typo * fix: use `vm_allocate` instead of `posix_memalign` * fix: don't call `newBufferWithBytesNoCopy` with `NULL` when `ggml_metal_host_malloc` returns `NULL` * fix: use `vm_allocate` only on macOS
* Added themes support with two sample themes and a favicon. * Newline * Newline * Newline * Trailing whitespace * Increased opacity for contrast * Increase opacity. Check actions cancelled for some other priority job and I can't seem to manually re-run them, so MOAR OPACITY * Opacity action trigger. Trying to re-trigger the cancelled action. * One more opacity adjustment This Actions pipeline is failing for random issues. * Delete examples/server/themes/buttons_top/completion.js This will be served from the static string built-in to server. * Delete examples/server/themes/buttons_top/index.js This will be served from the static string built-in to server. * Delete examples/server/themes/wild/completion.js This will be served from the static string built-in to server. * Delete examples/server/themes/buttons_top/json-schema-to-grammar.mjs This will be served from the static string built-in to server. * Delete examples/server/themes/wild/index.js This will be served from the static string built-in to server. * Delete examples/server/themes/wild/json-schema-to-grammar.mjs This will be served from the static string built-in to server. * Replaced underscore.
…-org#6899)" This reverts commit 46e12c4.
* DRAFT: Introduction of CUDA Graphs to LLama.cpp * FIx issues raised in comments * Tidied to now only use CUDA runtime (not mixed with driver calls) * disable for multi-gpu and batch size > 1 * Disable CUDA graphs for old GPU arch and with env var * added missing CUDA_CHECKs * Addressed comments * further addressed comments * limit to GGML_ALLOW_CUDA_GRAPHS defined in llama.cpp cmake * Added more comprehensive graph node checking * With mechanism to fall back if graph capture fails * Revert "With mechanism to fall back if graph capture fails" This reverts commit eb9f15f. * Fall back if graph capture fails and address other comments * - renamed GGML_ALLOW_CUDA_GRAPHS to GGML_CUDA_USE_GRAPHS - rename env variable to disable CUDA graphs to GGML_CUDA_DISABLE_GRAPHS - updated Makefile build to enable CUDA graphs - removed graph capture failure checking in ggml_cuda_error using a global variable to track this is not thread safe, but I am also not safistied with checking an error by string if this is necessary to workaround some issues with graph capture with eg. cuBLAS, we can pass the ggml_backend_cuda_context to the error checking macro and store the result in the context - fixed several resource leaks - fixed issue with zero node graphs - changed fixed size arrays to vectors - removed the count of number of evaluations before start capturing, and instead changed the capture mode to relaxed - removed the check for multiple devices so that it is still possible to use a single device, instead checks for split buffers to disable cuda graphs with -sm row - changed the op for checking batch size to GGML_OP_ADD, should be more reliable than GGML_OP_SOFT_MAX - code style fixes - things to look into - VRAM usage of the cudaGraphExec_t, if it is significant we may need to make it optional - possibility of using cudaStreamBeginCaptureToGraph to keep track of which ggml graph nodes correspond to which cuda graph nodes * fix build without cuda graphs * remove outdated comment * replace minimum cc value with a constant --------- Co-authored-by: slaren <[email protected]>
* convert-hf : begin refactoring write_tensor * convert : upgrade to sentencepiece v0.2.0 * convert-hf : remove unused n_dims in extra_*_tensors * convert-hf : simplify MoE weights stacking * convert-hf : flake8 linter doesn't like semicolons * convert-hf : allow unusual model part names For example, loading `model-00001-of-00001.safetensors` now works. * convert-hf : fix stacking MoE expert tensors `torch.stack` and `torch.cat` don't do the same thing. * convert-hf : fix Mamba conversion Tested to work even with a SentencePiece-based tokenizer. * convert : use a string for the SentencePiece tokenizer path * convert-hf : display tensor shape * convert-hf : convert norms to f32 by default * convert-hf : sort model part names `os.listdir` is said to list files in arbitrary order. Sorting the file names should let "model-00009-of-00042.safetensors" be loaded before "model-00010-of-00042.safetensors". * convert-hf : use an ABC for Model again It seems Protocol can't be used as a statically type-checked ABC, because its subclasses also can't be instantiated. (why did it seem to work?) At least there's still a way to throw an error when forgetting to define the `model_arch` property of any registered Model subclasses. * convert-hf : use a plain class for Model, and forbid direct instantiation There are no abstract methods used anyway, so using ABC isn't really necessary. * convert-hf : more consistent formatting of cmdline args * convert-hf : align the message logged for converted tensors * convert-hf : fix Refact conversion * convert-hf : save memory with lazy evaluation * convert-hf : flake8 doesn't like lowercase L as a variable name * convert-hf : remove einops requirement for InternLM2 * convert-hf : faster model parts loading Instead of pre-loading them all into a dict, iterate on the tensors in the model parts progressively as needed in Model.write_tensors Conversion for some architectures relies on checking for the presence of specific tensor names, so for multi-part models, the weight map is read from the relevant json file to quickly get these names up-front. * convert-hf : minor changes for consistency * gguf-py : add tqdm as a dependency It's small, and used for a progress bar in GGUFWriter.write_tensors_to_file
* opencl alignment size should be converted from bits to bytes Reference: https://registry.khronos.org/OpenCL/specs/3.0-unified/html/OpenCL_API.html#CL_DEVICE_MEM_BASE_ADDR_ALIGN > Alignment requirement (in bits) for sub-buffer offsets. * Update ggml-opencl.cpp for readability using division instead of shift Co-authored-by: Jared Van Bortel <[email protected]> --------- Co-authored-by: Jared Van Bortel <[email protected]>
* Add special token modification capability To be able to fix/amend special tokens in a GGUF let's add two new arguments: * `--special-token <name> <value>` where `<name>` can be bos, eos, prefix, middle, etc. while `<value>` is the token value, f.ex. `"<｜fim▁begin｜>"` * `--special-token-by-id <name> <id>` where `<id>` is the ID of the token, f.ex. 32006 So, in order to f.ex. add fill-in-middle tokens to a GGUF you would do the following: ```bash python3 gguf-new-metadata.py input.gguf output.gguf --special-token prefix "<｜fim▁begin｜>" --special-token middle "<｜fim▁hole｜>" --special-token suffix "<｜fim▁end｜>" ``` * improve help text * flake-- * fix multiple tokens warning * make script executable * switch to namedtuple, no need to dataclass * typing++ * add progress bar * Add special token modification capability To be able to fix/amend special tokens in a GGUF let's add two new arguments: * `--special-token <name> <value>` where `<name>` can be bos, eos, prefix, middle, etc. while `<value>` is the token value, f.ex. `"<｜fim▁begin｜>"` * `--special-token-by-id <name> <id>` where `<id>` is the ID of the token, f.ex. 32006 So, in order to f.ex. add fill-in-middle tokens to a GGUF you would do the following: ```bash gguf-new-metadata.py input.gguf output.gguf --special-token prefix "<｜fim▁begin｜>" --special-token middle "<｜fim▁end｜>" --special-token suffix "<｜fim▁hole｜>" ``` (yes, fim_end is the `middle` token, because completion is a `prefix`/`suffix`/`middle` sequence (where `middle` is unfilled)) or ```bash gguf-new-metadata.py input.gguf output.gguf --special-token prefix "<fim_prefix>" --special-token middle "<fim_middle>" --special-token suffix "<fim_suffix>" ``` etc... NB: The tokens have to exist already, trying to add non-existent token name/IDs will be ignored (with a warning), while non-existent values will fail (with an error). * improve help text * flake-- * fix multiple tokens warning * make script executable * switch to namedtuple, no need to dataclass * typing++ * add progress bar * fail on invalid token id
This commit changes the value assigned to llama_timings.n_p_eval when ctx->n_p_eval is 0 to be 1 instead of 1 which is the current value. The motivation for this change is that if session caching is enabled, for example using the `--prompt-cache main-session.txt` command line argument for the main example, and if the same prompt is used then on subsequent runs, the prompt tokens will not actually be passed to llama_decode, and n_p_eval will not be updated by llama_synchoronize. But the value of n_p_eval will be set 1 by llama_get_timings because ctx->n_p_eval will be 0. This could be interpreted as 1 token was evaluated for the prompt which could be misleading for applications using this value. Signed-off-by: Daniel Bevenius <[email protected]>
* CUDA: generalize FP16 fattn vec kernel * disable unsupported head sizes for AMD in test * try AMD fix * fix batch size 2-8 * partially revert changes
* merged the changes from deepseeker models to main branch
* Moved regex patterns to unicode.cpp and updated unicode.h
* Moved header files
* Resolved issues
* added and refactored unicode_regex_split and related functions
* Updated/merged the deepseek coder pr
* Refactored code
* Adding unicode regex mappings
* Adding unicode regex function
* Added needed functionality, testing remains
* Fixed issues
* Fixed issue with gpt2 regex custom preprocessor
* unicode : fix? unicode_wstring_to_utf8
* lint : fix whitespaces
* tests : add tokenizer tests for numbers
* unicode : remove redundant headers
* tests : remove and rename tokenizer test scripts
* tests : add sample usage
* gguf-py : reader prints warnings on duplicate keys
* llama : towards llama3 tokenization support (wip)
* unicode : shot in the dark to fix tests on Windows
* unicode : first try custom implementations
* convert : add "tokenizer.ggml.pre" GGUF KV (wip)
* llama : use new pre-tokenizer type
* convert : fix pre-tokenizer type writing
* lint : fix
* make : add test-tokenizer-0-llama-v3
* wip
* models : add llama v3 vocab file
* llama : adapt punctuation regex + add llama 3 regex
* minor
* unicode : set bomb
* unicode : set bomb
* unicode : always use std::wregex
* unicode : support \p{N}, \p{L} and \p{P} natively
* unicode : try fix windows
* unicode : category support via std::regex
* unicode : clean-up
* unicode : simplify
* llama3 custom regex split
* convert : add convert-hf-to-gguf-update.py
ggml-ci
* lint : update
* convert : add falcon
ggml-ci
* unicode : normalize signatures
* lint : fix
* lint : fix
* convert : remove unused functions
* convert : add comments
* convert : exercise contractions
ggml-ci
* Using char32_t for codepoints
* lint : fix
* already exists unicode_tolower()
* Typing
* Restore BOM
* cmake : refactor test targets
* tests : refactor vocab tests
ggml-ci
* tests : add more vocabs and tests
ggml-ci
* unicode : cleanup
* scripts : ignore new update script in check-requirements.sh
* Fix merge
* models : add phi-3, mpt, gpt-2, starcoder
* tests : disable obsolete
ggml-ci
* tests : use faster bpe test
ggml-ci
* llama : more prominent warning for old BPE models
* tests : disable test-tokenizer-1-bpe due to slowness
ggml-ci
* Move unused variable value
* GPT2 custom regex split
* Add alternative regex for custom aplit llama3
Co-authored-by: Georgi Gerganov <[email protected]>
* Style
* Add bruteforce random tests for token encoding
* wip: fixing unicode codepoint ranges
* Fix merge
* Unicode tables: separator, lowercase, uppercase and whitespace
* llama3 custom regex split: fix \s
* Restore BOM
* Style
* wip: generate NDF table
* Ignore special tokens for testing
* Clean gen-unicode-data.py
* Refactor random tokenizer test
* lint : fix
* tests : add fail test for llama-bpe
---------
Co-authored-by: Jaggzh <[email protected]>
Co-authored-by: Kazim Abrar Mahi <[email protected]>
Co-authored-by: Georgi Gerganov <[email protected]>
Co-authored-by: jaime-m-p <>
* added Layla to supported UIs * Update README.md
* ggml : full ALiBi support * ggml : update ggml_soft_max_ext() CUDA, SYCL * ggml : ggml_flash_attn_ext() support ALiBi (CPU) * ggml : ggml_flash_attn_ext() support ALiBi (Metal) * ggml : fix warning * ggml : ggml_flash_attn_ext() support ALiBi (CUDA) ggml-ci * ggml : fix assert message * vulkan : add dev notes * ggml : require mask when using ALiBi ggml-ci * convert : fix convert for refact models
* feat: first things to do * feat: create tensors for Jina architecture * fix: use other tensors * feat: embedding gets results * fix: fix usage of ALIBI * fix: clean prints * fix: do some cleanup unused vars * fix: revert changes to Makefile and CMakeLists * fix: revert some changes * fix: fix small detail * fix: fix convert formatting * fix: fix linting and editor * feat: set proper vocab settings * fix: JinaBertForMaskedLM registration * feat: support q_normalization and k_normalization in Jina arch * feat: handle gpt2 tokenizer with Jina architecture * feat: example comments in embedding * feat: rename Jina Bert to Jina Bert V2 * fix: add some changes as per review * feat: proper KQ_pos for Jina embeddings * feat: add capacity to load models ES and DE for Spanish * llama : fix pre-tokenizers * ggml : full ALiBi support * ggml : update ggml_soft_max_ext() CUDA, SYCL * ggml : ggml_flash_attn_ext() support ALiBi (CPU) * ggml : ggml_flash_attn_ext() support ALiBi (Metal) * ggml : fix warning * ggml : ggml_flash_attn_ext() support ALiBi (CUDA) ggml-ci * minor : clean-up * embedding : add warning about missing SEP --------- Co-authored-by: Georgi Gerganov <[email protected]>
* fix: llama-3 ignore_merges * test: add test for llama-3 bpe ignore_merges * fix: set ignore_merges only for llama-3 * fix: test-tokenizer-1-bpe --ingore-merges detection * fix: copy to fix fallthrough * fix: change ignore_merges to bool * fix: add ignore merges tests to cmake * llama : alternative merge ignore logic --------- Co-authored-by: Haoxiang Fei <[email protected]> Co-authored-by: Georgi Gerganov <[email protected]>
* [server] Cleanup a memory leak on exit There are a couple memory leaks on exit of the server. This hides others. After cleaning this up, you can see leaks on slots. But that is another patch to be sent after this. * make tab into spaces
* added sigmoid function * implemented metal kernel for sigmoid * implemented cuda kernel for sigmoid * added sigmoid unary op and incremented count
ggml-ci
* convert-hf : support bfloat16 conversion * gguf-py : flake8 fixes * convert-hf : add missing space after comma * convert-hf : get bit-exact same output as ./quantize The quantization version was missing. * convert-hf : don't round bf16 NANs * convert-hf : save some memory with np.int16 intermediate bf16 weights * convert-hf : more closely match llama.cpp with which weights to keep in f32 * convert-hf : add --outtype auto-f16 A reason for this to exist is for model quantizers who want an initial GGUF with the most fidelity to the original model while still using a 16-bit float type instead of 32-bit floats. * convert-hf : remove a semicolon because flake8 doesn't like it It's a reflex from when programming in C/C++, I guess. * convert-hf : support outtype templating in outfile name * convert-hf : rename --outtype auto-f16 to --outtype auto
…he loop. (ggml-org#7096) * A little documentation that shares my quick tips for working in the repository. * Update startup-testing-debugging.md * script that shows a menu of tests to pick from & run the debugger on * debug-test.sh: Refactor CLI help message * debug-test.sh: documentation update * debug-test.sh: CLI Help output corrections * debug-test.sh: minor doc fix --------- authored-by: Josh Ramer <[email protected]> Assisted-by: brian khuu <[email protected]>
ggml-ci
* CUDA: add FP32 FlashAttention vector kernel * fixup! CUDA: add FP32 FlashAttention vector kernel * fixup! fixup! CUDA: add FP32 FlashAttention vector kernel * fixup! fixup! fixup! CUDA: add FP32 FlashAttention vector kernel
Co-authored-by: Zhang <[email protected]>
…7241) * add oneapi running time dlls to release package * fix path * fix path * fix path * fix path * fix path --------- Co-authored-by: Zhang <[email protected]>
* change default temperature of OAI compat API from 0 to 1 * make tests explicitly send temperature to OAI API
…rt (ggml-org#4858) * convert.py: Outfile default name change and additional metadata support * convert.py: don't stringify Metadata load method output * convert.py: typo fix * convert.py: fix metadata format to sync with LLM_KV_NAMES in llama.cpp
* refactor: rename jina tokenizers to v2 * refactor: keep refactoring non-breaking
…ision for enabling AVX512_BF16
Srihari-mcw
force-pushed
the
bf16_cmake_plus_windows_build
branch
from
4e57aa6 to
29d5012
Compare
jart
pushed a commit
that referenced
this pull request
Aug 5, 2024
* [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 <[email protected]> --------- Co-authored-by: Georgi Gerganov <[email protected]> Co-authored-by: compilade <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The PR fixes build issues that were found while building the BF16 model PR in windows platform.