baalajimaestro/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
1431 commits 1 branch 0 tags 2.1 MiB
Commit graph

610 commits

Author SHA1 Message Date
randoentity
3f8f276f9f Add bindings for custom_rope 2023-07-10 17:37:46 +02:00
Andrei Betlen
a86bfdf0a5 bugfix: truncate completion max_tokens to fit context length by default 2023-07-09 18:13:29 -04:00
Andrei Betlen
6f70cc4b7d bugfix: pydantic settings missing / changed fields 2023-07-09 18:03:31 -04:00
Andrei
5d756de314
Merge branch 'main' into add_unlimited_max_tokens 2023-07-08 02:37:38 -04:00
Andrei
b8e0bed295
Merge pull request #453 from wu-qing-157/main 
Fix incorrect token_logprobs (due to indexing after sorting)
 2023-07-08 02:31:52 -04:00
Andrei Betlen
d6e6aad927 bugfix: fix compatibility bug with openai api on last token 2023-07-08 00:06:11 -04:00
Andrei Betlen
4f2b5d0b53 Format 2023-07-08 00:05:10 -04:00
Andrei Betlen
34c505edf2 perf: convert pointer to byref 2023-07-07 22:54:07 -04:00
Andrei Betlen
52753b77f5 Upgrade fastapi to 0.100.0 and pydantic v2 2023-07-07 21:38:46 -04:00
Andrei Betlen
11eae75211 perf: avoid allocating new buffers during sampling 2023-07-07 19:28:53 -04:00
Andrei Betlen
a14d8a9b3f perf: assign to candidates data structure instead 2023-07-07 18:58:43 -04:00
wu-qing-157
9e61661518 fix indexing token_logprobs after sorting 2023-07-07 10:18:49 +00:00
Andrei Betlen
57d8ec3899 Add setting to control request interruption 2023-07-07 03:37:23 -04:00
Andrei Betlen
4c7cdcca00 Add interruptible streaming requests for llama-cpp-python server. Closes #183 2023-07-07 03:04:17 -04:00
Andrei Betlen
98ae4e58a3 Update llama.cpp 2023-07-06 17:57:56 -04:00
Andrei Betlen
b994296c75 Update llama.cpp 2023-07-05 01:00:14 -04:00
Andrei Betlen
c67f786360 Update llama.cpp 2023-06-29 01:08:15 -04:00
Andrei Betlen
e34f4414cf Hotfix: logits_all bug 2023-06-29 00:57:27 -04:00
Andrei Betlen
a2ede37bd5 Load logits directly into scores buffer 2023-06-29 00:45:46 -04:00
Andrei Betlen
b95b0ffbeb Use pre-allocated buffers to store input_ids and scores 2023-06-29 00:40:47 -04:00
Andrei Betlen
a5e059c053 Free model when llama is unloaded. Closes #434 2023-06-28 23:58:55 -04:00
Andrei Betlen
3379dc40a1 Merge branch 'main' of github.com:abetlen/llama_cpp_python into main 2023-06-26 08:50:48 -04:00
Andrei Betlen
952228407e Update llama.cpp 2023-06-26 08:50:38 -04:00
Andrei Betlen
b4a3db3e54 Update type signature 2023-06-26 08:50:30 -04:00
Andrei
5eb4ebb041
Merge branch 'main' into fix-state-pickle 2023-06-26 08:45:02 -04:00
samfundev
d788fb49bf
Only concatenate after all batches are done 2023-06-24 15:51:46 -04:00
Andrei
877ca6d016
Merge branch 'main' into fix-state-pickle 2023-06-23 15:13:07 -04:00
Alexey
282698b6d3
server: pass seed param from command line to llama 2023-06-23 00:19:24 +04:00
Andrei Betlen
e37798777e Update llama.cpp 2023-06-20 11:25:10 -04:00
Andrei Betlen
d410f12fae Update docs. Closes #386 2023-06-17 13:38:48 -04:00
Andrei Betlen
9f528f4715 Merge branch 'main' of github.com:abetlen/llama_cpp_python into main 2023-06-17 13:37:17 -04:00
Andrei Betlen
d7153abcf8 Update llama.cpp 2023-06-16 23:11:14 -04:00
imaprogrammer
fd9f294b3a
Update llama.py: Added how many input tokens in ValueError exception 2023-06-16 14:11:57 +05:30
Andrei Betlen
1e20be6d0c Add low_vram to server settings 2023-06-14 22:13:42 -04:00
Andrei Betlen
44b83cada5 Add low_vram parameter 2023-06-14 22:12:33 -04:00
Andrei Betlen
f7c5cfaf50 Format server options 2023-06-14 22:08:28 -04:00
Andrei Betlen
9c41a3e990 Merge branch 'main' of github.com:abetlen/llama_cpp_python into main 2023-06-14 21:50:43 -04:00
Andrei
f568baeef1
Merge pull request #351 from player1537-forks/th/add-logits-bias-parameter 
Add support for `logit_bias` and `logit_bias_type` parameters
 2023-06-14 21:49:56 -04:00
Andrei Betlen
f27393ab7e Add additional verbose logs for cache 2023-06-14 21:46:48 -04:00
Andrei Betlen
4cefb70cd0 Merge branch 'main' of github.com:abetlen/llama_cpp_python into main 2023-06-14 21:40:19 -04:00
Andrei Betlen
715f98c591 Update llama.cpp 2023-06-14 21:40:13 -04:00
Okabintaro
10b0cb727b fix: Make LLamaState pickable for disk cache 
I fixed the issue by making the saved state a bytes object instead of the ctypes one which can't be pickled.
 2023-06-13 12:03:31 +02:00
Gabor
3129a0e7e5 correction to add back environment variable support <3 docker 2023-06-11 01:11:24 +01:00
Gabor
3ea31930e5 fixes abetlen/llama-cpp-python #358 2023-06-11 00:58:08 +01:00
Andrei Betlen
21acd7901f Re-enable cache 2023-06-10 12:22:31 -04:00
Andrei Betlen
6639371407 Update llama.cpp 2023-06-10 12:17:38 -04:00
Tanner Hobson
eb7645b3ba Add support for logit_bias and logit_bias_type parameters 2023-06-09 13:13:08 -04:00
Andrei Betlen
0da655b3be Temporarily disable cache until save state bug is fixed. 2023-06-09 11:10:24 -04:00
Andrei Betlen
556c7edf47 Truncate max_tokens if it exceeds context length 2023-06-09 10:57:36 -04:00
Andrei Betlen
0c42168508 Fix cache implementation breaking changes 2023-06-08 13:19:23 -04:00
Andrei Betlen
607d217caa Allow both .so and .dylib extensions for macos 2023-06-08 00:27:19 -04:00
Andrei
0f0b447fa4
Merge pull request #289 from Maximilian-Winter/main 
Diskcache implementation for llama state.
 2023-06-06 17:03:03 -04:00
Andrei
d508573fb4
Merge pull request #328 from spirilis/mirostat 
Added mirostat support for completions, chat completions API
 2023-06-06 16:58:23 -04:00
Andrei Betlen
aad4b17f52 Update llama.cpp 2023-06-06 16:23:55 -04:00
Andrei Betlen
8b4968ea67 Fix resize issue. Closes #330 2023-06-06 11:37:57 -04:00
Eric B
9b1c9e902c Added mirostat support for completions, chat completions API 2023-06-05 22:37:11 -04:00
Andrei Betlen
7b57420ea9 Update llama.cpp 2023-06-05 18:17:29 -04:00
Maximilian-Winter
29f9c9cca3 Added both LlamaChache classes Disk and RAM. 2023-05-31 22:33:56 +02:00
Maximilian Winter
9ea7a379d3
Merge branch 'abetlen:main' into main 2023-05-31 12:55:51 +02:00
Andrei
49fe9395a1
Merge pull request #277 from abetlen/add-numpy-support 
Use numpy for internal buffers
 2023-05-29 20:59:30 -04:00
Maximilian-Winter
719c3eae0a Diskcache implementation for llama state. 2023-05-28 15:56:38 +02:00
Andrei Betlen
80066f0b80 Use async routes 2023-05-27 09:12:58 -04:00
Andrei Betlen
c2b59a5f59 Import unnused import 2023-05-26 22:59:29 -04:00
Andrei Betlen
8f2b4456ad Format 2023-05-26 22:04:31 -04:00
Andrei Betlen
84e313bd6e Align dtype to match c structs 2023-05-26 22:02:16 -04:00
Andrei Betlen
66bcb8d70d Merge branch 'main' into add-numpy-support 2023-05-26 20:25:03 -04:00
Andrei Betlen
8f35bddd7e Fix stop sequence performance bug. 2023-05-26 20:23:49 -04:00
Andrei Betlen
7fc7bc30e7 Remove usage of eval_tokens for cache check 2023-05-26 20:12:05 -04:00
Andrei Betlen
fe331ec589 Replace eval_logits and eval_tokens with numpy arrays 2023-05-26 20:03:31 -04:00
Andrei Betlen
8eb9769f78 Add support for numpy 2023-05-26 16:12:45 -04:00
Andrei Betlen
4c1b7f7a76 Bugfix for logits_processor and stopping_criteria 2023-05-26 10:25:28 -04:00
Andrei Betlen
433a2e3e8a Add extra logits_processor and stopping_criteria 2023-05-26 03:13:24 -04:00
Andrei Betlen
f74b90ed67 Fix streaming hang on last token when cache is on. 2023-05-26 03:03:01 -04:00
Andrei Betlen
5be8354e11 Added tokenizer 2023-05-26 03:00:51 -04:00
Andrei Betlen
8fa2ef1959 Format 2023-05-26 03:00:35 -04:00
Andrei Betlen
6bd1075291 Merge branch 'Maximilian-Winter/main' into main 2023-05-26 02:56:11 -04:00
Andrei Betlen
ca01f98e09 Add LlamaTokenizer class 2023-05-25 14:11:33 -04:00
Andrei Betlen
1d247e0f35 Add StoppingCriteria and LogitsProcessor to generate to match huggingface API 2023-05-25 14:04:54 -04:00
Maximilian-Winter
c2585b6889 Fixed list elements typing 2023-05-25 10:54:08 +02:00
Maximilian-Winter
da463e6c8c Added types to logit processor list and stop criteria list 2023-05-25 09:07:16 +02:00
Maximilian-Winter
c05fcdf42f Fixed none value of logits processors. 2023-05-24 22:02:06 +02:00
Maximilian-Winter
5bb780d455 Implemented logit processors and stop criteria's 2023-05-24 21:55:44 +02:00
Andrei Betlen
fab064ded9 Remove unnecessary ffi calls 2023-05-23 17:56:21 -04:00
Andrei Betlen
0adb9ec37a Use model_name and index in response 2023-05-21 21:30:03 -04:00
Andrei Betlen
922b5b2bfd Merge branch 'main' into server-embedding 2023-05-21 21:21:38 -04:00
Andrei Betlen
cd102e9da1 Cache shared library function calls for static tokens 2023-05-21 19:18:56 -04:00
Andrei Betlen
b895511cca Fix penalize_nl 2023-05-21 18:38:06 -04:00
Andrei Betlen
03e2947b03 Fix unnecessary memory allocation while sampling 2023-05-21 18:36:34 -04:00
Andrei Betlen
fafe47114c Update llama.cpp 2023-05-21 17:47:21 -04:00
Andrei Betlen
76b1d2cd20 Change properties to functions to match token functions 2023-05-20 08:24:06 -04:00
Andrei Betlen
a7ba85834f Add n_ctx, n_vocab, and n_embd properties 2023-05-20 08:13:41 -04:00
Simon Chabot
e783f1c191 feat: make embedding support list of string as input 
makes the /v1/embedding route similar to OpenAI api.
 2023-05-20 01:23:32 +02:00
Andrei Betlen
01a010be52 Fix llama_cpp and Llama type signatures. Closes #221 2023-05-19 11:59:33 -04:00
Andrei Betlen
a8cd169251 Bugfix: Stop sequences can be strings 2023-05-19 03:15:08 -04:00
Andrei Betlen
17d4271b04 Fix logprobs for completions and implement for streaming logprobs. 2023-05-19 02:20:27 -04:00
Andrei Betlen
a634a2453b Allow first logprob token to be null to match openai api 2023-05-19 02:04:57 -04:00
Andrei Betlen
dc39cc0fa4 Use server sent events function for streaming completion 2023-05-19 02:04:30 -04:00
Andrei Betlen
f0ec6e615e Stream tokens instead of text chunks 2023-05-18 11:35:59 -04:00
Andrei Betlen
21d8f5fa9f Remove unnused union 2023-05-18 11:35:15 -04:00
Andrei Betlen
61d58e7b35 Check for CUDA_PATH before adding 2023-05-17 15:26:38 -04:00
Andrei Betlen
7c95895626 Merge branch 'main' of github.com:abetlen/llama_cpp_python into main 2023-05-17 15:19:32 -04:00
Aneesh Joy
e9794f91f2
Fixd CUBLAS dll load issue in Windows 2023-05-17 18:04:58 +01:00
Andrei Betlen
4f342795e5 Update token checks 2023-05-17 03:35:13 -04:00
Andrei Betlen
f5c2f998ab Format 2023-05-17 02:00:39 -04:00
Andrei Betlen
d28b753ed2 Implement penalize_nl 2023-05-17 01:53:26 -04:00
Andrei Betlen
f11e2a781c Fix last_n_tokens_size 2023-05-17 01:42:51 -04:00
Andrei Betlen
7e55244540 Fix top_k value. Closes #220 2023-05-17 01:41:42 -04:00
Andrei Betlen
a7c9e38287 Update variable name 2023-05-16 18:07:25 -04:00
Andrei Betlen
a3352923c7 Add model_alias option to override model_path in completions. Closes #39 2023-05-16 17:22:00 -04:00
Andrei Betlen
a65125c0bd Add sampling defaults for generate 2023-05-16 09:35:50 -04:00
Andrei Betlen
cbac19bf24 Add winmode arg only on windows if python version supports it 2023-05-15 09:15:01 -04:00
Andrei Betlen
c804efe3f0 Fix obscure Wndows DLL issue. Closes #208 2023-05-14 22:08:11 -04:00
Andrei Betlen
cdf59768f5 Update llama.cpp 2023-05-14 00:04:22 -04:00
Andrei Betlen
7a536e86c2 Allow model to tokenize strings longer than context length and set add_bos. Closes #92 2023-05-12 14:28:22 -04:00
Andrei Betlen
8740ddc58e Only support generating one prompt at a time. 2023-05-12 07:21:46 -04:00
Andrei Betlen
8895b9002a Revert "llama_cpp server: prompt is a string". Closes #187 
This reverts commit b9098b0ef7.
 2023-05-12 07:16:57 -04:00
Andrei Betlen
7be584fe82 Add missing tfs_z paramter 2023-05-11 21:56:19 -04:00
Andrei Betlen
cdeaded251 Bugfix: Ensure logs are printed when streaming 2023-05-10 16:12:17 -04:00
Lucas Doyle
02e8a018ae llama_cpp server: document presence_penalty and frequency_penalty, mark as supported 2023-05-09 16:25:00 -07:00
Andrei Betlen
d957422bf4 Implement sampling as in llama.cpp main example 2023-05-08 21:21:25 -04:00
Andrei Betlen
93a9019bb1 Merge branch 'main' of github.com:abetlen/llama_cpp_python into Maximilian-Winter/main 2023-05-08 19:57:09 -04:00
Andrei Betlen
82d138fe54 Fix: default repeat_penalty 2023-05-08 18:49:11 -04:00
Andrei Betlen
29f094bbcf Bugfix: not falling back to environment variables when default is value is set. 2023-05-08 14:46:25 -04:00
Andrei Betlen
0d6c60097a Show default value when --help is called 2023-05-08 14:21:15 -04:00
Andrei Betlen
022e9ebcb8 Use environment variable if parsed cli arg is None 2023-05-08 14:20:53 -04:00
Andrei Betlen
0d751a69a7 Set repeat_penalty to 0 by default 2023-05-08 01:50:43 -04:00
Andrei Betlen
65d9cc050c Add openai frequency and presence penalty parameters. Closes #169 2023-05-08 01:30:18 -04:00
Andrei Betlen
a0b61ea2a7 Bugfix for models endpoint 2023-05-07 20:17:52 -04:00
Andrei Betlen
e72f58614b Change pointer to lower overhead byref 2023-05-07 20:01:34 -04:00
Andrei Betlen
14da46f16e Added cache size to settins object. 2023-05-07 19:33:17 -04:00
Andrei Betlen
0e94a70de1 Add in-memory longest prefix cache. Closes #158 2023-05-07 19:31:26 -04:00
Andrei Betlen
8dfde63255 Fix return type 2023-05-07 19:30:14 -04:00
Andrei Betlen
2753b85321 Format 2023-05-07 13:19:56 -04:00
Andrei Betlen
627811ea83 Add verbose flag to server 2023-05-07 05:09:10 -04:00
Andrei Betlen
3fbda71790 Fix mlock_supported and mmap_supported return type 2023-05-07 03:04:22 -04:00
Andrei Betlen
5a3413eee3 Update cpu_count 2023-05-07 03:03:57 -04:00
Andrei Betlen
1a00e452ea Update settings fields and defaults 2023-05-07 02:52:20 -04:00
Andrei Betlen
86753976c4 Revert "llama_cpp server: delete some ignored / unused parameters" 
This reverts commit b47b9549d5.
 2023-05-07 02:02:34 -04:00
Andrei Betlen
c382d8f86a Revert "llama_cpp server: mark model as required" 
This reverts commit e40fcb0575.
 2023-05-07 02:00:22 -04:00
Andrei Betlen
d8fddcce73 Merge branch 'main' of github.com:abetlen/llama_cpp_python into better-server-params-and-fields 2023-05-07 01:54:00 -04:00
Andrei Betlen
7c3743fe5f Update llama.cpp 2023-05-07 00:12:47 -04:00
Andrei Betlen
bc853e3742 Fix type for eval_logits in LlamaState object 2023-05-06 21:32:50 -04:00
Maximilian Winter
515d9bde7e Fixed somethings and activated cublas 2023-05-06 23:40:19 +02:00
Maximilian Winter
aa203a0d65 Added mirostat sampling to the high level API. 2023-05-06 22:47:47 +02:00
Andrei Betlen
98bbd1c6a8 Fix eval logits type 2023-05-05 14:23:14 -04:00
Andrei Betlen
b5f3e74627 Add return type annotations for embeddings and logits 2023-05-05 14:22:55 -04:00
Andrei Betlen
3e28e0e50c Fix: runtime type errors 2023-05-05 14:12:26 -04:00
Andrei Betlen
e24c3d7447 Prefer explicit imports 2023-05-05 14:05:31 -04:00
Andrei Betlen
40501435c1 Fix: types 2023-05-05 14:04:12 -04:00
Andrei Betlen
66e28eb548 Fix temperature bug 2023-05-05 14:00:41 -04:00
Andrei Betlen
6702d2abfd Fix candidates type 2023-05-05 14:00:30 -04:00
Andrei Betlen
5e7ddfc3d6 Fix llama_cpp types 2023-05-05 13:54:22 -04:00
Andrei Betlen
b6a9a0b6ba Add types for all low-level api functions 2023-05-05 12:22:27 -04:00
Andrei Betlen
5be0efa5f8 Cache should raise KeyError when key is missing 2023-05-05 12:21:49 -04:00
Andrei Betlen
24fc38754b Add cli options to server. Closes #37 2023-05-05 12:08:28 -04:00
Andrei Betlen
853dc711cc Format 2023-05-04 21:58:36 -04:00
Andrei Betlen
97c6372350 Rewind model to longest prefix. 2023-05-04 21:58:27 -04:00
Andrei Betlen
329297fafb Bugfix: Missing logits_to_logprobs 2023-05-04 12:18:40 -04:00
Lucas Doyle
3008a954c1 Merge branch 'main' of github.com:abetlen/llama-cpp-python into better-server-params-and-fields 2023-05-03 13:10:03 -07:00
Andrei Betlen
9e5b6d675a Improve logging messages 2023-05-03 10:28:10 -04:00
Andrei Betlen
43f2907e3a Support smaller state sizes 2023-05-03 09:33:50 -04:00
Andrei Betlen
1d47cce222 Update llama.cpp 2023-05-03 09:33:30 -04:00
Lucas Doyle
b9098b0ef7 llama_cpp server: prompt is a string 
Not sure why this union type was here but taking a look at llama.py, prompt is only ever processed as a string for completion

This was breaking types when generating an openapi client
 2023-05-02 14:47:07 -07:00
Matt Hoffner
f97ff3c5bb
Update llama_cpp.py 2023-05-01 20:40:06 -07:00
Andrei
7ab08b8d10
Merge branch 'main' into better-server-params-and-fields 2023-05-01 22:45:57 -04:00
Andrei Betlen
9eafc4c49a Refactor server to use factory 2023-05-01 22:38:46 -04:00
Andrei Betlen
dd9ad1c759 Formatting 2023-05-01 21:51:16 -04:00
Lucas Doyle
dbbfc4ba2f llama_cpp server: fix to ChatCompletionRequestMessage 
When I generate a client, it breaks because it fails to process the schema of ChatCompletionRequestMessage

These fix that:
- I think `Union[Literal["user"], Literal["channel"], ...]` is the same as Literal["user", "channel", ...]
- Turns out default value `Literal["user"]` isn't JSON serializable, so replace with "user"
 2023-05-01 15:38:19 -07:00
Lucas Doyle
fa2a61e065 llama_cpp server: fields for the embedding endpoint 2023-05-01 15:38:19 -07:00
Lucas Doyle
8dcbf65a45 llama_cpp server: define fields for chat completions 
Slight refactor for common fields shared between completion and chat completion
 2023-05-01 15:38:19 -07:00
Lucas Doyle
978b6daf93 llama_cpp server: add some more information to fields for completions 2023-05-01 15:38:19 -07:00
Lucas Doyle
a5aa6c1478 llama_cpp server: add missing top_k param to CreateChatCompletionRequest 
`llama.create_chat_completion` definitely has a `top_k` argument, but its missing from `CreateChatCompletionRequest`. decision: add it
 2023-05-01 15:38:19 -07:00
Lucas Doyle
1e42913599 llama_cpp server: move logprobs to supported 
I think this is actually supported (its in the arguments of `LLama.__call__`, which is how the completion is invoked). decision: mark as supported
 2023-05-01 15:38:19 -07:00
Lucas Doyle
b47b9549d5 llama_cpp server: delete some ignored / unused parameters 
`n`, `presence_penalty`, `frequency_penalty`, `best_of`, `logit_bias`, `user`: not supported, excluded from the calls into llama. decision: delete it
 2023-05-01 15:38:19 -07:00
Lucas Doyle
e40fcb0575 llama_cpp server: mark model as required 
`model` is ignored, but currently marked "optional"... on the one hand could mark "required" to make it explicit in case the server supports multiple llama's at the same time, but also could delete it since its ignored. decision: mark it required for the sake of openai api compatibility.

I think out of all parameters, `model` is probably the most important one for people to keep using even if its ignored for now.
 2023-05-01 15:38:19 -07:00
Andrei Betlen
b6747f722e Fix logprob calculation. Fixes #134 2023-05-01 17:45:08 -04:00
Andrei Betlen
9ff9cdd7fc Fix import error 2023-05-01 15:11:15 -04:00
Andrei Betlen
350a1769e1 Update sampling api 2023-05-01 14:47:55 -04:00
Andrei Betlen
7837c3fdc7 Fix return types and import comments 2023-05-01 14:02:06 -04:00
Andrei Betlen
ccf1ed54ae Merge branch 'main' of github.com:abetlen/llama_cpp_python into main 2023-05-01 11:35:14 -04:00
Andrei Betlen
80184a286c Update llama.cpp 2023-05-01 10:44:28 -04:00
Lucas Doyle
efe8e6f879 llama_cpp server: slight refactor to init_llama function 
Define an init_llama function that starts llama with supplied settings instead of just doing it in the global context of app.py

This allows the test to be less brittle by not needing to mess with os.environ, then importing the app
 2023-04-29 11:42:23 -07:00
Lucas Doyle
6d8db9d017 tests: simple test for server module 2023-04-29 11:42:20 -07:00
Lucas Doyle
468377b0e2 llama_cpp server: app is now importable, still runnable as a module 2023-04-29 11:41:25 -07:00
Andrei
755f9fa455
Merge pull request #118 from SagsMug/main 
Fix UnicodeDecodeError permanently
 2023-04-29 07:19:01 -04:00
Mug
18a0c10032 Remove excessive errors="ignore" and add utf8 test 2023-04-29 12:19:22 +02:00
Andrei Betlen
ea0faabae1 Update llama.cpp 2023-04-28 15:32:43 -04:00
Mug
b7d14efc8b Python weirdness 2023-04-28 13:20:31 +02:00
Mug
eed61289b6 Dont detect off tokens, detect off detokenized utf8 2023-04-28 13:16:18 +02:00
Mug
3a98747026 One day, i'll fix off by 1 errors permanently too 2023-04-28 12:54:28 +02:00
Mug
c39547a986 Detect multi-byte responses and wait 2023-04-28 12:50:30 +02:00
Andrei Betlen
9339929f56 Update llama.cpp 2023-04-26 20:00:54 -04:00
Mug
5f81400fcb Also ignore errors on input prompts 2023-04-26 14:45:51 +02:00
Mug
be2c961bc9 Merge branch 'main' of https://github.com/abetlen/llama-cpp-python 2023-04-26 14:38:09 +02:00
Mug
c4a8491d42 Fix decode errors permanently 2023-04-26 14:37:06 +02:00
Andrei Betlen
cbd26fdcc1 Update llama.cpp 2023-04-25 19:03:41 -04:00
Andrei Betlen
3cab3ef4cb Update n_batch for server 2023-04-25 09:11:32 -04:00
Andrei Betlen
cc706fb944 Add ctx check and re-order __init__. Closes #112 2023-04-25 09:00:53 -04:00
Andrei Betlen
d484c5634e Bugfix: Check cache keys as prefix to prompt tokens 2023-04-24 22:18:54 -04:00
Andrei Betlen
cbe95bbb75 Add cache implementation using llama state 2023-04-24 19:54:41 -04:00
Andrei Betlen
2c359a28ff Merge branch 'main' of github.com:abetlen/llama_cpp_python into main 2023-04-24 17:51:27 -04:00
Andrei Betlen
197cf80601 Add save/load state api for Llama class 2023-04-24 17:51:25 -04:00
Andrei Betlen
86f8e5ad91 Refactor internal state for Llama class 2023-04-24 15:47:54 -04:00
Andrei
f37456133a
Merge pull request #108 from eiery/main 
Update n_batch default to 512 to match upstream llama.cpp
 2023-04-24 13:48:09 -04:00
Andrei Betlen
02cf881317 Update llama.cpp 2023-04-24 09:30:10 -04:00
eiery
aa12d8a81f
Update llama.py 
update n_batch default to 512 to match upstream llama.cpp
 2023-04-23 20:56:40 -04:00
Andrei Betlen
7230599593 Disable mmap when applying lora weights. Closes #107 2023-04-23 14:53:17 -04:00
Andrei Betlen
e99caedbbd Update llama.cpp 2023-04-22 19:50:28 -04:00
Andrei Betlen
1eb130a6b2 Update llama.cpp 2023-04-21 17:40:27 -04:00
Andrei Betlen
e4647c75ec Add use_mmap flag to server 2023-04-19 15:57:46 -04:00
Andrei Betlen
0df4d69c20 If lora base is not set avoid re-loading the model by passing NULL 2023-04-18 23:45:25 -04:00
Andrei Betlen
95c0dc134e Update type signature to allow for null pointer to be passed. 2023-04-18 23:44:46 -04:00
Andrei Betlen
453e517fd5 Add seperate lora_base path for applying LoRA to quantized models using original unquantized model weights. 2023-04-18 10:20:46 -04:00
Andrei Betlen
eb7f278cc6 Add lora_path parameter to Llama model 2023-04-18 01:43:44 -04:00
Andrei Betlen
35abf89552 Add bindings for LoRA adapters. Closes #88 2023-04-18 01:30:04 -04:00
Andrei Betlen
89856ef00d Bugfix: only eval new tokens 2023-04-15 17:32:53 -04:00
Andrei Betlen
92c077136d Add experimental cache 2023-04-15 12:03:09 -04:00
Andrei Betlen
a6372a7ae5 Update stop sequences for chat 2023-04-15 12:02:48 -04:00
Andrei Betlen
83b2be6dc4 Update chat parameters 2023-04-15 11:58:43 -04:00
Andrei Betlen
62087514c6 Update chat prompt 2023-04-15 11:58:19 -04:00
Andrei Betlen
02f9fb82fb Bugfix 2023-04-15 11:39:52 -04:00
Andrei Betlen
3cd67c7bd7 Add type annotations 2023-04-15 11:39:21 -04:00
Andrei Betlen
d7de0e8014 Bugfix 2023-04-15 00:08:04 -04:00
Andrei Betlen
e90e122f2a Use clear 2023-04-14 23:33:18 -04:00
Andrei Betlen
ac7068a469 Track generated tokens internally 2023-04-14 23:33:00 -04:00
Andrei Betlen
6e298d8fca Set kv cache size to f16 by default 2023-04-14 22:21:19 -04:00
Andrei Betlen
6c7cec0c65 Fix completion request 2023-04-14 10:01:15 -04:00
Andrei Betlen
6153baab2d Clean up logprobs implementation 2023-04-14 09:59:33 -04:00
Andrei Betlen
26cc4ee029 Fix signature for stop parameter 2023-04-14 09:59:08 -04:00
Andrei Betlen
6595ad84bf Add field to disable reseting between generations 2023-04-13 00:28:00 -04:00
Andrei Betlen
22fa5a621f Revert "Deprecate generate method" 
This reverts commit 6cf5876538.
 2023-04-13 00:19:55 -04:00
Andrei Betlen
4f5f99ef2a Formatting 2023-04-12 22:40:12 -04:00
Andrei Betlen
0daf16defc Enable logprobs on completion endpoint 2023-04-12 19:08:11 -04:00
Andrei Betlen
19598ac4e8 Fix threading bug. Closes #62 2023-04-12 19:07:53 -04:00
Andrei Betlen
005c78d26c Update llama.cpp 2023-04-12 14:29:00 -04:00
Andrei Betlen
c854c2564b Don't serialize stateful parameters 2023-04-12 14:07:14 -04:00
Andrei Betlen
2f9b649005 Style fix 2023-04-12 14:06:22 -04:00
Andrei Betlen
6cf5876538 Deprecate generate method 2023-04-12 14:06:04 -04:00
Andrei Betlen
b3805bb9cc Implement logprobs parameter for text completion. Closes #2 2023-04-12 14:05:11 -04:00
Andrei Betlen
9f1e565594 Update llama.cpp 2023-04-11 11:59:03 -04:00
Andrei Betlen
213cc5c340 Remove async from function signature to avoid blocking the server 2023-04-11 11:54:31 -04:00
jm12138
90e1021154 Add unlimited max_tokens 2023-04-10 15:56:05 +00:00
Mug
2559e5af9b Changed the environment variable name into "LLAMA_CPP_LIB" 2023-04-10 17:27:17 +02:00
Mug
ee71ce8ab7 Make windows users happy (hopefully) 2023-04-10 17:12:25 +02:00
Mug
cf339c9b3c Better custom library debugging 2023-04-10 17:06:58 +02:00
Mug
4132293d2d Merge branch 'main' of https://github.com/abetlen/llama-cpp-python into local-lib 2023-04-10 17:00:42 +02:00
Mug
76131d5bb8 Use environment variable for library override 2023-04-10 17:00:35 +02:00
Andrei Betlen
1f67ad2a0b Add use_mmap option 2023-04-10 02:11:35 -04:00
Andrei Betlen
c3c2623e8b Update llama.cpp 2023-04-09 22:01:33 -04:00
Andrei Betlen
314ce7d1cc Fix cpu count default 2023-04-08 19:54:04 -04:00
Andrei Betlen
3fbc06361f Formatting 2023-04-08 16:01:45 -04:00
Andrei Betlen
0067c1a588 Formatting 2023-04-08 16:01:18 -04:00
Andrei Betlen
38f442deb0 Bugfix: Wrong size of embeddings. Closes #47 2023-04-08 15:05:33 -04:00
Andrei Betlen
ae3e9c3d6f Update shared library extension for macos 2023-04-08 02:45:21 -04:00
Andrei Betlen
da539cc2ee Safer calculation of default n_threads 2023-04-06 21:22:19 -04:00
Andrei Betlen
930db37dd2 Merge branch 'main' of github.com:abetlen/llama_cpp_python into main 2023-04-06 21:07:38 -04:00
Andrei Betlen
55279b679d Handle prompt list 2023-04-06 21:07:35 -04:00
MillionthOdin16
c283edd7f2 Set n_batch to default values and reduce thread count: 
Change batch size to the llama.cpp default of 8. I've seen issues in llama.cpp where batch size affects quality of generations. (It shouldn't) But in case that's still an issue I changed to default.

Set auto-determined num of threads to 1/2 system count. ggml will sometimes lock cores at 100% while doing nothing. This is being addressed, but can cause bad experience for user if pegged at 100%
 2023-04-05 18:17:29 -04:00
MillionthOdin16
76a82babef Set n_batch to the default value of 8. I think this is leftover from when n_ctx was missing and n_batch was 2048. 2023-04-05 17:44:53 -04:00
Andrei Betlen
44448fb3a8 Add server as a subpackage 2023-04-05 16:23:25 -04:00
Mug
e3ea354547 Allow local llama library usage 2023-04-05 14:23:01 +02:00
Andrei Betlen
e96a5c5722 Make Llama instance pickleable. Closes #27 2023-04-05 06:52:17 -04:00
Andrei Betlen
7643f6677d Bugfix for Python3.7 2023-04-05 04:37:33 -04:00
Andrei Betlen
cefc69ea43 Add runtime check to ensure embedding is enabled if trying to generate embeddings 2023-04-05 03:25:37 -04:00
Andrei Betlen
5c50af7462 Remove workaround 2023-04-05 03:25:09 -04:00
Andrei Betlen
51dbcf2693 Bugfix: wrong signature for quantize function 2023-04-04 22:36:59 -04:00
Andrei Betlen
c137789143 Add verbose flag. Closes #19 2023-04-04 13:09:24 -04:00
Andrei Betlen
5075c16fcc Bugfix: n_batch should always be <= n_ctx 2023-04-04 13:08:21 -04:00
Andrei Betlen
caf3c0362b Add return type for default __call__ method 2023-04-03 20:26:08 -04:00
Andrei Betlen
4aa349d777 Add docstring for create_chat_completion 2023-04-03 20:24:20 -04:00
Andrei Betlen
7fedf16531 Add support for chat completion 2023-04-03 20:12:44 -04:00
Andrei Betlen
3dec778c90 Update to more sensible return signature 2023-04-03 20:12:14 -04:00
Andrei Betlen
ae004eb69e Fix #16 2023-04-03 18:46:19 -04:00
MillionthOdin16
a0758f0077
Update llama_cpp.py with PR requests 
lib_base_name and load_shared_library
to 
_lib_base_name and _load_shared_library
 2023-04-03 13:06:50 -04:00
MillionthOdin16
a40476e299
Update llama_cpp.py 
Make shared library code more robust with some platform specific functionality and more descriptive errors when failures occur
 2023-04-02 21:50:13 -04:00
Andrei Betlen
1ed8cd023d Update llama_cpp and add kv_cache api support 2023-04-02 13:33:49 -04:00
Andrei Betlen
4f509b963e Bugfix: Stop sequences and missing max_tokens check 2023-04-02 03:59:19 -04:00
Andrei Betlen
353e18a781 Move workaround to new sample method 2023-04-02 00:06:34 -04:00
Andrei Betlen
a4a1bbeaa9 Update api to allow for easier interactive mode 2023-04-02 00:02:47 -04:00
Andrei Betlen
eef627c09c Fix example documentation 2023-04-01 17:39:35 -04:00
Andrei Betlen
1e4346307c Add documentation for generate method 2023-04-01 17:36:30 -04:00
Andrei Betlen
67c70cc8eb Add static methods for beginning and end of sequence tokens. 2023-04-01 17:29:30 -04:00
Andrei Betlen
318eae237e Update high-level api 2023-04-01 13:01:27 -04:00
Andrei Betlen
69e7d9f60e Add type definitions 2023-04-01 12:59:58 -04:00
Andrei Betlen
49c8df369a Fix type signature of token_to_str 2023-03-31 03:25:12 -04:00
Andrei Betlen
670d390001 Fix ctypes typing issue for Arrays 2023-03-31 03:20:15 -04:00
Andrei Betlen
1545b22727 Fix array type signatures 2023-03-31 02:08:20 -04:00
Andrei Betlen
c928e0afc8 Formatting 2023-03-31 00:00:27 -04:00
Andrei Betlen
8908f4614c Update llama.cpp 2023-03-28 21:10:23 -04:00
Andrei Betlen
70b8a1ef75 Add support to get embeddings from high-level api. Closes #4 2023-03-28 04:59:54 -04:00
Andrei Betlen
3dbb3fd3f6 Add support for stream parameter. Closes #1 2023-03-28 04:03:57 -04:00
Andrei Betlen
30fc0f3866 Extract generate method 2023-03-28 02:42:22 -04:00
Andrei Betlen
1c823f6d0f Refactor Llama class and add tokenize / detokenize methods Closes #3 2023-03-28 01:45:37 -04:00
Andrei Betlen
8ae3beda9c Update Llama to add params 2023-03-25 16:26:23 -04:00
Andrei Betlen
4525236214 Update llama.cpp 2023-03-25 16:26:03 -04:00
Andrei Betlen
b121b7c05b Update docstring 2023-03-25 12:33:18 -04:00
Andrei Betlen
fa92740a10 Update llama.cpp 2023-03-25 12:12:09 -04:00
Andrei Betlen
df15caa877 Add mkdocs 2023-03-24 18:57:59 -04:00
Andrei Betlen
4da5faa28b Bugfix: cross-platform method to find shared lib 2023-03-24 18:43:29 -04:00
Andrei Betlen
b93675608a Handle errors returned by llama.cpp 2023-03-24 15:47:17 -04:00
Andrei Betlen
7786edb0f9 Black formatting 2023-03-24 14:59:29 -04:00
Andrei Betlen
c784d83131 Update llama.cpp and re-organize low-level api 2023-03-24 14:58:42 -04:00
Andrei Betlen
b9c53b88a1 Use n_ctx provided from actual context not params 2023-03-24 14:58:10 -04:00
Andrei Betlen
2cc499512c Black formatting 2023-03-24 14:35:41 -04:00
Andrei Betlen
e24c581b5a Implement prompt batch processing as in main.cpp 2023-03-24 14:33:38 -04:00
Andrei Betlen
a28cb92d8f Remove model_name param 2023-03-24 04:04:29 -04:00
Andrei Betlen
eec9256a42 Bugfix: avoid decoding partial utf-8 characters 2023-03-23 16:25:13 -04:00
Andrei Betlen
e63ea4dbbc Add support for logprobs 2023-03-23 15:51:05 -04:00
Andrei Betlen
465238b179 Updated package to build with skbuild 2023-03-23 13:54:14 -04:00
Andrei Betlen
79b304c9d4 Initial commit 2023-03-23 05:33:06 -04:00