baalajimaestro/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
1775 commits 1 branch 0 tags 2.1 MiB
Commit graph

338 commits

Author SHA1 Message Date
nullname
d634efcdd9
feat: adding rpc_servers parameter to Llama class (#1477) 
* passthru rpc_servers params

wip

* enable llama rpc by default

* convert string to byte

* add rpc package

* Revert "enable llama rpc by default"

This reverts commit 832c6dd56c979514cec5df224bf2d2014dccd790.

* update readme

* Only set rpc_servers when provided

* Add rpc servers to server options

---------

Co-authored-by: Andrei Betlen <abetlen@gmail.com>
 2024-06-04 10:38:21 -04:00
Asghar Ghorbani
6e0642ca19
fix: fix logprobs when BOS is not present (#1471) 
* Fix lobprobs when BOS is not present

* Fix logprobs when bos is not available
 2024-06-04 10:18:38 -04:00
Sigbjørn Skjæret
027f7bc678
fix: Avoid duplicate special tokens in chat formats (#1439) 
* Templates sometimes have BOS in them, remove duplicate

* tokenize chat format prompts before completion

This is to ensure that we don't duplicate any special tokens.

Hopefully I amended the existing formats correctly?

* updated comment

* corrected a few

* add some missing internals

* proper bos/eos detection

* just let tokenizer do the job

* typo--

* align test with new response

* changed to a warning

* move to another PR

* Use python warnings module

---------

Co-authored-by: Andrei Betlen <abetlen@gmail.com>
 2024-06-04 10:15:41 -04:00
Andrei Betlen
df45a4b3fe fix: fix string value kv_overrides. Closes #1487 2024-05-29 02:02:22 -04:00
Linghan Zhong
5cae1040e3
feat: Improve Llama.eval performance by avoiding list conversion (#1476) 
Co-authored-by: Andrei <abetlen@gmail.com>
 2024-05-24 01:49:44 -04:00
Andrei Betlen
d99a6ba607 fix: segfault for models without eos / bos tokens. Closes #1463 2024-05-16 00:37:27 -04:00
twaka
5212fb08ae
feat: add MinTokensLogitProcessor and min_tokens argument to server (#1333) 
* implement min_tokens

* set default to 0

* pass min_tokens

* fix

* remove copy

* implement MinTokensLogitsProcessor

* format

* fix condition
 2024-05-14 09:50:53 -04:00
Sigbjørn Skjæret
389e09c2f5
misc: Remove unnecessary metadata lookups (#1448) 
Special tokens are already mapped from metadata by llama.cpp
 2024-05-14 09:44:09 -04:00
Sigbjørn Skjæret
5ab40e6167
feat: Support multiple chat templates - step 1 (#1396) 
* Support multiple chat templates - step 1

As a first step, allow user to to select template from metadata with chat_format parameter in the form of `chat_template.name`.

* register chat templates to self.chat_formats instead of globally

* Don't expose internal chat handlers yet

---------

Co-authored-by: Andrei <abetlen@gmail.com>
 2024-05-09 09:49:09 -04:00
Sigbjørn Skjæret
4a7122d22f
feat: fill-in-middle support (#1386) 
* Proper fill-in-middle support

Use prefix/middle/suffix tokens when metadata is present in GGUF, like f.ex. in [this](https://huggingface.co/CISCai/CodeQwen1.5-7B-Chat-SOTA-GGUF) one.

* fall back to internal prefix/middle/suffix id

In some cases llama.cpp will make a guess at fim tokens, use them if there's no metadata.

* typo--

* don't insert special tokens that are not there in suffix

Note: add_bos is misnamed, it's actually add_special and can cause several special tokens to be added to the token list (the special parameter is actually parse_special).

* don't add/parse any special tokens when using fim

I've left original behavior when no fim tokens are found, but this should perhaps be re-evaluated.

* don't append suffix to prompt_tokens unless fim tokens are detected

* make sure we only do this for fim

---------

Co-authored-by: Andrei <abetlen@gmail.com>
 2024-05-08 02:26:22 -04:00
Bruno Alvisio
a50d24e3a7
fix: chat_format log where auto-detected format prints None (#1434) 2024-05-08 02:19:35 -04:00
Andrei Betlen
9f7a85571a fix: Use memmove to copy str_value kv_override. Closes #1417 2024-05-03 19:07:50 -04:00
Andrei Betlen
29b6e9a5c8 fix: wrong parameter for flash attention in pickle __getstate__ 2024-04-30 09:32:47 -04:00
Andrei Betlen
22d77eefd2 feat: Add option to enable flash_attn to Lllama params and ModelSettings 2024-04-30 09:29:16 -04:00
Andrei Betlen
a411612b38 feat: Add support for str type kv_overrides 2024-04-27 23:42:19 -04:00
Douglas Hanley
f6ed21f9a2
feat: Allow for possibly non-pooled embeddings (#1380) 
* allow for possibly non-pooled embeddings

* add more to embeddings section in README.md

---------

Co-authored-by: Andrei <abetlen@gmail.com>
 2024-04-25 21:32:44 -04:00
Andrei Betlen
d40a250ef3 feat: Use new llama_token_is_eog in create_completions 2024-04-22 00:35:47 -04:00
Andrei Betlen
cc81afebf0 feat: Add stopping_criteria to ChatFormatter, allow stopping on arbitrary token ids, fixes llama3 instruct 2024-04-20 00:00:53 -04:00
tc-wolf
4924455dec
feat: Make saved state more compact on-disk (#1296) 
* State load/save changes

- Only store up to `n_tokens` logits instead of full `(n_ctx, n_vocab)`
  sized array.
  - Difference between ~350MB and ~1500MB for example prompt with ~300
    tokens (makes sense lol)
- Auto-formatting changes

* Back out formatting changes
 2024-04-17 10:06:50 -04:00
ddh0
c96b2daebf feat: Use all available CPUs for batch processing (#1345) 2024-04-17 10:05:54 -04:00
Andrei Betlen
bb65b4d764 fix: pass correct type to chat handlers for chat completion logprobs 2024-04-10 03:41:55 -04:00
Andrei Betlen
8649d7671b fix: segfault when logits_all=False. Closes #1319 2024-04-03 15:30:31 -04:00
Limour
f165048a69
feat: add support for KV cache quantization options (#1307) 
* add KV cache quantization options

https://github.com/abetlen/llama-cpp-python/discussions/1220
https://github.com/abetlen/llama-cpp-python/issues/1305

* Add ggml_type

* Use ggml_type instead of string for quantization

* Add server support

---------

Co-authored-by: Andrei Betlen <abetlen@gmail.com>
 2024-04-01 10:19:28 -04:00
windspirit95
aa9f1ae011
feat: Add logprobs support to chat completions (#1311) 
* Add logprobs return in ChatCompletionResponse

* Fix duplicate field

* Set default to false

* Simplify check

* Add server example

---------

Co-authored-by: Andrei Betlen <abetlen@gmail.com>
 2024-03-31 13:30:13 -04:00
Andrei Betlen
4084aabe86 fix: set default pooling type to unspecified 2024-03-14 10:04:57 -04:00
Andrei Betlen
d318cc8b83 fix: Set default pooling_type to mean, check for null pointer. 2024-03-14 09:17:41 -04:00
Douglas Hanley
2811014bae
feat: Switch embed to llama_get_embeddings_seq (#1263) 
* switch to llama_get_embeddings_seq

* Remove duplicate definition of llama_get_embeddings_seq

Co-authored-by: Andrei <abetlen@gmail.com>

---------

Co-authored-by: Andrei <abetlen@gmail.com>
 2024-03-08 20:59:35 -05:00
Andrei Betlen
93dc56ace8 Update llama.cpp 2024-03-06 01:32:00 -05:00
Andrei Betlen
97aa3a153d docs: Add information re: auto chat formats. Closes #1236 2024-03-01 13:10:25 -05:00
Andrei Betlen
f062a7f51d feat: Update llama.cpp 2024-03-01 12:57:16 -05:00
Sigbjørn Skjæret
c36ab15e68
fix: eos/bos_token set correctly for Jinja2ChatFormatter and automatic chat formatter (#1230) 
The token strings were not correctly retrieved (empty).
 2024-02-28 01:30:31 -05:00
Andrei Betlen
2292af5796 feat: Update llama.cpp 2024-02-25 16:53:58 -05:00
Andrei Betlen
47bad30dd7 fix: LlamaHFTokenizer now receives pre_tokens 2024-02-23 12:23:24 -05:00
Andrei Betlen
db776a885c fix: module 'llama_cpp.llama_cpp' has no attribute 'c_uint8' 2024-02-23 11:24:53 -05:00
Andrei Betlen
e6d6260a91 fix: Update from_pretrained defaults to match hf_hub_download 2024-02-22 00:10:23 -05:00
Andrei
7f51b6071f
feat(low-level-api): Improve API static type-safety and performance (#1205) 2024-02-21 16:25:38 -05:00
Andrei
0f8aa4ab5c
feat: Pull models directly from huggingface (#1206) 
* Add from_pretrained method to Llama class

* Update docs

* Merge filename and pattern
 2024-02-21 16:25:10 -05:00
Andrei Betlen
53f6f5f415 fix: self.numa missing 2024-02-17 01:02:33 -05:00
Andrei Betlen
fdce078cb9 feat: Update llama.cpp 2024-02-17 00:37:51 -05:00
Andrei Betlen
0ce66bc080 fix: create_embedding broken response for input type str 2024-02-15 16:09:48 -05:00
Douglas Hanley
7bb91f025f
fix: Incorporate embedding pooling layer fixes (#1194) 
* remove division by token count

* truncate to n_batch, not n_ctx
 2024-02-15 15:16:30 -05:00
Douglas Hanley
d7a67917ba
feat: Support batch embeddings (#1186) 
* handle batched embeddings

* fix normalization issue

* fix type hints, ensure no breaking changes to embed

* Clear kv cache / reset internal state after embedding complete

---------

Co-authored-by: Andrei <abetlen@gmail.com>
 2024-02-14 04:26:09 -05:00
Andrew Lapp
d6be5333e1
fix: sample idx off-by-one error for logit_processors (#1179) 
* fix sample_idx off-by-one error

* self._scores is indexed differently, only modify the index within self._input_ids

---------

Co-authored-by: Andrew Lapp <andrew@rew.la>
Co-authored-by: Andrei <abetlen@gmail.com>
 2024-02-13 12:26:07 -05:00
Andrei Betlen
cb791716b4 fix: Always set logits_all = True when using speculative decoding 2024-02-12 16:19:05 -05:00
Andrei
153a0049d9
feat: Generic chatml Function Calling (#957) 
* Add demo notebook

* Add initial chat handler

* Update OpenAI types

* Add generic chatml function calling (wip)

* Update chatml generic function calling.

* Progress on auto-tool calls

* fix streaming functions

* Remove print statements

* fix: Suppress output from llama.cpp init and grammar creation

* Add OpenAI v1 python api compatible chat completion function

* Support non-streaming multi-tool calls

* Format

* Include function_call in response.
 2024-02-12 15:56:07 -05:00
Andrei Betlen
4abb8c9386 Merge branch 'main' of github.com:abetlen/llama_cpp_python into main 2024-02-09 13:32:31 -05:00
Andrei Betlen
e16f06e6eb fix: revert _create_completions. 2024-02-09 02:02:13 -05:00
Andrei Betlen
b5fca911b5 feat: Move tokenizer to own module 2024-02-08 01:08:18 -05:00
Jeffrey Fong
901827013b
feat: Integrate functionary v1.4 and v2 models + add custom tokenizer support to Llama class (#1078) 
* convert functionary-v1 chat handler to use hf autotokenizer

* add hf_tokenizer + inteegrate functionary-v1.4 prompt template

* integrate functionary v2 prompt template

* update readme

* set up parallel function calling wip

* set up parallel function calling

* Update README.md

* Update README.md

* refactor tokenizers

* include old functionary handler for backward compatibility

* add hf_tokenizer_path in server ModelSettings

* convert functionary-v1 chat handler to use hf autotokenizer

* add hf_tokenizer + inteegrate functionary-v1.4 prompt template

* integrate functionary v2 prompt template

* update readme

* set up parallel function calling wip

* resolve merge conflict

* Update README.md

* Update README.md

* refactor tokenizers

* include old functionary handler for backward compatibility

* add hf_tokenizer_path in server ModelSettings

* Cleanup PR, fix breaking changes

* Use hf_pretrained_model_name_or_path for tokenizer

* fix hf tokenizer in streaming

* update README

* refactor offset mapping

---------

Co-authored-by: Andrei <abetlen@gmail.com>
 2024-02-07 20:07:03 -05:00
Andrei Betlen
59760c85ed fix: Use llama_log_callback to avoid suppress_stdout_stderr 2024-02-05 21:52:12 -05:00