1362 Commits

Author	SHA1	Message	Date
oobabooga	cf0697936a	Optimize StreamingLLM by over 10x	2024-03-08 21:48:28 -08:00
oobabooga	afb51bd5d6	Add StreamingLLM for llamacpp & llamacpp_HF (2nd attempt) (#5669)	2024-03-09 00:25:33 -03:00
oobabooga	549bb88975	Increase height of "Custom stopping strings" UI field	2024-03-08 12:54:30 -08:00
oobabooga	238f69accc	Move "Command for chat-instruct mode" to the main chat tab (closes #5634)	2024-03-08 12:52:52 -08:00
oobabooga	bae14c8f13	Right-truncate long chat completion prompts instead of left-truncating ... Instructions are usually at the beginning of the prompt.	2024-03-07 08:50:24 -08:00
Bartowski	104573f7d4	Update cache_4bit documentation (#5649) ... --------- Co-authored-by: oobabooga <112222186+oobabooga@users.noreply.github.com>	2024-03-07 13:08:21 -03:00
oobabooga	2ec1d96c91	Add cache_4bit option for ExLlamaV2 (#5645)	2024-03-06 23:02:25 -03:00
oobabooga	2174958362	Revert gradio to 3.50.2 (#5640)	2024-03-06 11:52:46 -03:00
oobabooga	d61e31e182	Save the extensions after Gradio 4 (#5632)	2024-03-05 07:54:34 -03:00
oobabooga	63a1d4afc8	Bump gradio to 4.19 (#5522)	2024-03-05 07:32:28 -03:00
oobabooga	f697cb4609	Move update_wizard_windows.sh to update_wizard_windows.bat (oops)	2024-03-04 19:26:24 -08:00
kalomaze	cfb25c9b3f	Cubic sampling w/ curve param (#5551) ... --------- Co-authored-by: oobabooga <112222186+oobabooga@users.noreply.github.com>	2024-03-03 13:22:21 -03:00
oobabooga	09b13acfb2	Perplexity evaluation: print to terminal after calculation is finished	2024-02-28 19:58:21 -08:00
oobabooga	4164e29416	Block the "To create a public link, set share=True" gradio message	2024-02-25 15:06:08 -08:00
oobabooga	d34126255d	Fix loading extensions with "-" in the name (closes #5557)	2024-02-25 09:24:52 -08:00
oobabooga	10aedc329f	Logging: more readable messages when renaming chat histories	2024-02-22 07:57:06 -08:00
oobabooga	faf3bf2503	Perplexity evaluation: make UI events more robust (attempt)	2024-02-22 07:13:22 -08:00
oobabooga	ac5a7a26ea	Perplexity evaluation: add some informative error messages	2024-02-21 20:20:52 -08:00
oobabooga	59032140b5	Fix CFG with llamacpp_HF (2nd attempt)	2024-02-19 18:35:42 -08:00
oobabooga	c203c57c18	Fix CFG with llamacpp_HF	2024-02-19 18:09:49 -08:00
oobabooga	ae05d9830f	Replace {{char}}, {{user}} in the chat template itself	2024-02-18 19:57:54 -08:00
oobabooga	1f27bef71b	Move chat UI elements to the right on desktop (#5538)	2024-02-18 14:32:05 -03:00
oobabooga	d6bd71db7f	ExLlamaV2: fix loading when autosplit is not set	2024-02-17 12:54:37 -08:00
oobabooga	af0bbf5b13	Lint	2024-02-17 09:01:04 -08:00
oobabooga	a6730f88f7	Add --autosplit flag for ExLlamaV2 (#5524)	2024-02-16 15:26:10 -03:00
oobabooga	4039999be5	Autodetect llamacpp_HF loader when tokenizer exists	2024-02-16 09:29:26 -08:00
oobabooga	76d28eaa9e	Add a menu for customizing the instruction template for the model (#5521)	2024-02-16 14:21:17 -03:00
oobabooga	0e1d8d5601	Instruction template: make "Send to default/notebook" work without a tokenizer	2024-02-16 08:01:07 -08:00
oobabooga	44018c2f69	Add a "llamacpp_HF creator" menu (#5519)	2024-02-16 12:43:24 -03:00
oobabooga	b2b74c83a6	Fix Qwen1.5 in llamacpp_HF	2024-02-15 19:04:19 -08:00
oobabooga	080f7132c0	Revert gradio to 3.50.2 (#5513)	2024-02-15 20:40:23 -03:00
oobabooga	7123ac3f77	Remove "Maximum UI updates/second" parameter (#5507)	2024-02-14 23:34:30 -03:00
DominikKowalczyk	33c4ce0720	Bump gradio to 4.19 (#5419) ... --------- Co-authored-by: oobabooga <112222186+oobabooga@users.noreply.github.com>	2024-02-14 23:28:26 -03:00
oobabooga	b16958575f	Minor bug fix	2024-02-13 19:48:32 -08:00
oobabooga	d47182d9d1	llamacpp_HF: do not use oobabooga/llama-tokenizer (#5499)	2024-02-14 00:28:51 -03:00
oobabooga	069ed7c6ef	Lint	2024-02-13 16:05:41 -08:00
oobabooga	86c320ab5a	llama.cpp: add a progress bar for prompt evaluation	2024-02-07 21:56:10 -08:00
oobabooga	c55b8ce932	Improved random preset generation	2024-02-06 08:51:52 -08:00
oobabooga	4e34ae0587	Minor logging improvements	2024-02-06 08:22:08 -08:00
oobabooga	3add2376cd	Better warpers logging	2024-02-06 07:09:21 -08:00
oobabooga	494cc3c5b0	Handle empty sampler priority field, use default values	2024-02-06 07:05:32 -08:00
oobabooga	775902c1f2	Sampler priority: better logging, always save to presets	2024-02-06 06:49:22 -08:00
oobabooga	acfbe6b3b3	Minor doc changes	2024-02-06 06:35:01 -08:00
oobabooga	8ee3cea7cb	Improve some log messages	2024-02-06 06:31:27 -08:00
oobabooga	8a6d9abb41	Small fixes	2024-02-06 06:26:27 -08:00
oobabooga	2a1063eff5	Revert "Remove non-HF ExLlamaV2 loader (#5431)" ... This reverts commit cde000d478.	2024-02-06 06:21:36 -08:00
oobabooga	8c35fefb3b	Add custom sampler order support (#5443)	2024-02-06 11:20:10 -03:00
oobabooga	7301c7618f	Minor change to Models tab	2024-02-04 21:49:58 -08:00
oobabooga	f234fbe83f	Improve a log message after previous commit	2024-02-04 21:44:53 -08:00
oobabooga	7073665a10	Truncate long chat completions inputs (#5439)	2024-02-05 02:31:24 -03:00
oobabooga	9033fa5eee	Organize the Model tab	2024-02-04 19:30:22 -08:00
Forkoz	2a45620c85	Split by rows instead of layers for llama.cpp multi-gpu (#5435)	2024-02-04 23:36:40 -03:00
Badis Ghoubali	3df7e151f7	fix the n_batch slider (#5436)	2024-02-04 18:15:30 -03:00
oobabooga	4e188eeb80	Lint	2024-02-03 20:40:10 -08:00
oobabooga	cde000d478	Remove non-HF ExLlamaV2 loader (#5431)	2024-02-04 01:15:51 -03:00
kalomaze	b6077b02e4	Quadratic sampling (#5403) ... --------- Co-authored-by: oobabooga <112222186+oobabooga@users.noreply.github.com>	2024-02-04 00:20:02 -03:00
Badis Ghoubali	40c7977f9b	Add roleplay.gbnf grammar (#5368)	2024-01-28 21:41:28 -03:00
sam-ngu	c0bdcee646	added trust_remote_code to deepspeed init loaderClass (#5237)	2024-01-26 11:10:57 -03:00
oobabooga	87dc421ee8	Bump exllamav2 to 0.0.12 (#5352)	2024-01-22 22:40:12 -03:00
oobabooga	aad73667af	Lint	2024-01-22 03:25:55 -08:00
lmg-anon	db1da9f98d	Fix logprobs tokens in OpenAI API (#5339)	2024-01-22 08:07:42 -03:00
Forkoz	5c5ef4cef7	UI: change n_gpu_layers maximum to 256 for larger models. (#5262)	2024-01-17 17:13:16 -03:00
ilya sheprut	4d14eb8b82	LoRA: Fix error "Attempting to unscale FP16 gradients" when training (#5268)	2024-01-17 17:11:49 -03:00
oobabooga	e055967974	Add prompt_lookup_num_tokens parameter (#5296)	2024-01-17 17:09:36 -03:00
oobabooga	b3fc2cd887	UI: Do not save unchanged extension settings to settings.yaml	2024-01-10 03:48:30 -08:00
oobabooga	53dc1d8197	UI: Do not save unchanged settings to settings.yaml	2024-01-09 18:59:04 -08:00
oobabooga	89e7e107fc	Lint	2024-01-09 16:27:50 -08:00
mamei16	bec4e0a1ce	Fix update event in refresh buttons (#5197)	2024-01-09 14:49:37 -03:00
oobabooga	4333d82b9d	Minor bug fix	2024-01-09 06:55:18 -08:00
oobabooga	953343cced	Improve the file saving/deletion menus	2024-01-09 06:33:47 -08:00
oobabooga	123f27a3c5	Load the nearest character after deleting a character ... Instead of the first.	2024-01-09 06:24:27 -08:00
oobabooga	b908ed318d	Revert "Rename past chats -> chat history" ... This reverts commit aac93a1fd6.	2024-01-09 05:26:07 -08:00
oobabooga	4ca82a4df9	Save light/dark theme on "Save UI defaults to settings.yaml"	2024-01-09 04:20:10 -08:00
oobabooga	7af50ede94	Reorder some buttons	2024-01-09 04:11:50 -08:00
oobabooga	a9f49a7574	Confirm the chat history rename with enter	2024-01-09 04:00:53 -08:00
oobabooga	7bdd2118a2	Change some log messages when deleting files	2024-01-09 03:32:01 -08:00
oobabooga	aac93a1fd6	Rename past chats -> chat history	2024-01-09 03:14:30 -08:00
oobabooga	615fa11af8	Move new chat button, improve history deletion handling	2024-01-08 21:22:37 -08:00
oobabooga	4f7e1eeafd	Past chat histories in a side bar on desktop (#5098) ... Lots of room for improvement, but that's a start.	2024-01-09 01:57:29 -03:00
oobabooga	372ef5e2d8	Fix dynatemp parameters always visible	2024-01-08 19:42:31 -08:00
oobabooga	29c2693ea0	dynatemp_low, dynatemp_high, dynatemp_exponent parameters (#5209)	2024-01-08 23:28:35 -03:00
oobabooga	c4e005efec	Fix dropdown menus sometimes failing to refresh	2024-01-08 17:49:54 -08:00
oobabooga	9cd2106303	Revert "Add dynamic temperature to the random preset button" ... This reverts commit 4365fb890f.	2024-01-08 16:46:24 -08:00
oobabooga	4365fb890f	Add dynamic temperature to the random preset button	2024-01-07 13:08:15 -08:00
oobabooga	0d07b3a6a1	Add dynamic_temperature_low parameter (#5198)	2024-01-07 17:03:47 -03:00
oobabooga	b8a0b3f925	Don't print torch tensors with --verbose	2024-01-07 10:35:55 -08:00
oobabooga	cf820c69c5	Print generation parameters with --verbose (HF only)	2024-01-07 10:06:23 -08:00
oobabooga	c4c7fc4ab3	Lint	2024-01-07 09:36:56 -08:00
kalomaze	48327cc5c4	Dynamic Temperature HF loader support (#5174) ... --------- Co-authored-by: oobabooga <112222186+oobabooga@users.noreply.github.com>	2024-01-07 10:36:26 -03:00
oobabooga	248742df1c	Save extension fields to settings.yaml on "Save UI defaults"	2024-01-04 20:33:42 -08:00
oobabooga	c9d814592e	Increase maximum temperature value to 5	2024-01-04 17:28:15 -08:00
oobabooga	e4d724eb3f	Fix cache_folder bug introduced in 37eff915d6	2024-01-04 07:49:40 -08:00
Alberto Cano	37eff915d6	Use --disk-cache-dir for all caches	2024-01-04 00:27:26 -03:00
Lounger	7965f6045e	Fix loading latest history for file names with dots (#5162)	2024-01-03 22:39:41 -03:00
AstrisCantCode	b80e6365d0	Fix various bugs for LoRA training (#5161)	2024-01-03 20:42:20 -03:00
oobabooga	7cce88c403	Rmove an unncecessary exception	2024-01-02 07:20:59 -08:00
oobabooga	94afa0f9cf	Minor style changes	2024-01-01 16:00:22 -08:00
oobabooga	cbf6f9e695	Update some UI messages	2023-12-30 21:31:17 -08:00
oobabooga	2aad91f3c9	Remove deprecated command-line flags (#5131)	2023-12-31 02:07:48 -03:00
oobabooga	2734ce3e4c	Remove RWKV loader (#5130)	2023-12-31 02:01:40 -03:00
oobabooga	0e54a09bcb	Remove exllamav1 loaders (#5128)	2023-12-31 01:57:06 -03:00
oobabooga	8e397915c9	Remove --sdp-attention, --xformers flags (#5126)	2023-12-31 01:36:51 -03:00
B611	b7dd1f9542	Specify utf-8 encoding for model metadata file open (#5125)	2023-12-31 01:34:32 -03:00
oobabooga	c06f630bcc	Increase max_updates_second maximum value	2023-12-24 13:29:47 -08:00
oobabooga	8c60495878	UI: add "Maximum UI updates/second" parameter	2023-12-24 09:17:40 -08:00
zhangningboo	1b8b61b928	Fix output_ids decoding for Qwen/Qwen-7B-Chat (#5045)	2023-12-22 23:11:02 -03:00
Yiximail	afc91edcb2	Reset the model_name after unloading the model (#5051)	2023-12-22 22:18:24 -03:00
oobabooga	2706149c65	Organize the CMD arguments by group (#5027)	2023-12-21 00:33:55 -03:00
oobabooga	c727a70572	Remove redundancy from modules/loaders.py	2023-12-20 19:18:07 -08:00
luna	6efbe3009f	let exllama v1 models load safetensor loras (#4854)	2023-12-20 13:29:19 -03:00
oobabooga	bcba200790	Fix EOS being ignored in ExLlamav2 after previous commit	2023-12-20 07:54:06 -08:00
oobabooga	f0f6d9bdf9	Add HQQ back & update version ... This reverts commit 2289e9031e.	2023-12-20 07:46:09 -08:00
oobabooga	b15f510154	Optimize ExLlamav2 (non-HF) loader	2023-12-20 07:31:42 -08:00
oobabooga	fadb295d4d	Lint	2023-12-19 21:36:57 -08:00
oobabooga	fb8ee9f7ff	Add a specific error if HQQ is missing	2023-12-19 21:32:58 -08:00
oobabooga	9992f7d8c0	Improve several log messages	2023-12-19 20:54:32 -08:00
oobabooga	23818dc098	Better logger ... Credits: vladmandic/automatic	2023-12-19 20:38:33 -08:00
oobabooga	95600073bc	Add an informative error when extension requirements are missing	2023-12-19 20:20:45 -08:00
oobabooga	d8279dc710	Replace character name placeholders in chat context (closes #5007)	2023-12-19 17:31:46 -08:00
oobabooga	e83e6cedbe	Organize the model menu	2023-12-19 13:18:26 -08:00
oobabooga	f4ae0075e8	Fix conversion from old template format to jinja2	2023-12-19 13:16:52 -08:00
oobabooga	de138b8ba6	Add llama-cpp-python wheels with tensor cores support (#5003)	2023-12-19 17:30:53 -03:00
oobabooga	0a299d5959	Bump llama-cpp-python to 0.2.24 (#5001)	2023-12-19 15:22:21 -03:00
oobabooga	83cf1a6b67	Fix Yi space issue (closes #4996)	2023-12-19 07:54:19 -08:00
oobabooga	9847809a7a	Add a warning about ppl evaluation without --no_use_fast	2023-12-18 18:09:24 -08:00
oobabooga	f6d701624c	UI: mention that QuIP# does not work on Windows	2023-12-18 18:05:02 -08:00
oobabooga	a23a004434	Update the example template	2023-12-18 17:47:35 -08:00
Water	674be9a09a	Add HQQ quant loader (#4888) ... --------- Co-authored-by: oobabooga <112222186+oobabooga@users.noreply.github.com>	2023-12-18 21:23:16 -03:00
oobabooga	1f9e25e76a	UI: update "Saved instruction templates" dropdown after loading template	2023-12-17 21:19:06 -08:00
oobabooga	da1c8d77ea	Merge remote-tracking branch 'refs/remotes/origin/dev' into dev	2023-12-17 21:05:10 -08:00
oobabooga	cac89df97b	Instruction templates: better handle unwanted bos tokens	2023-12-17 21:04:30 -08:00
oobabooga	f0d6ead877	llama.cpp: read instruction template from GGUF metadata (#4975)	2023-12-18 01:51:58 -03:00
oobabooga	f1f2c4c3f4	Add --num_experts_per_token parameter (ExLlamav2) (#4955)	2023-12-17 12:08:33 -03:00
oobabooga	12690d3ffc	Better HF grammar implementation (#4953)	2023-12-17 02:01:23 -03:00
oobabooga	f8079d067d	UI: save the sent chat message on "no model is loaded" error	2023-12-16 10:52:41 -08:00
oobabooga	3bbf6c601d	AutoGPTQ: Add --disable_exllamav2 flag (Mixtral CPU offloading needs this)	2023-12-15 06:46:13 -08:00
oobabooga	2cb5b68ad9	Bug fix: when generation fails, save the sent message (#4915)	2023-12-15 01:01:45 -03:00
Kim Jaewon	e53f99faa0	[OpenAI Extension] Add 'max_logits' parameter in logits endpoint (#4916)	2023-12-15 00:22:43 -03:00
Lounger	5754f0c357	Fix deleting chat logs (#4914)	2023-12-13 21:54:43 -03:00
Bartowski	f51156705d	Allow symlinked folder within root directory (#4863)	2023-12-13 18:08:21 -03:00
Ixion	3f3960dbfb	Fixed invalid Jinja2 syntax in instruction templates (#4911)	2023-12-13 15:46:23 -03:00
oobabooga	fcf5512364	Jinja templates: fix a potential small bug	2023-12-13 10:19:39 -08:00
oobabooga	7f1a6a70e3	Update the llamacpp_HF comment	2023-12-12 21:04:20 -08:00
oobabooga	1c531a3713	Minor cleanup	2023-12-12 13:25:21 -08:00
oobabooga	8513028968	Fix lag in the chat tab during streaming	2023-12-12 13:01:25 -08:00
oobabooga	39d2fe1ed9	Jinja templates for Instruct and Chat (#4874)	2023-12-12 17:23:14 -03:00
oobabooga	aab0dd962d	Revert "Update callbacks.py to show tracebacks on ValueError (#4892)" ... This reverts commit 993ca51a65.	2023-12-12 11:47:11 -08:00
Nehereus	993ca51a65	Update callbacks.py to show tracebacks on ValueError (#4892)	2023-12-12 02:29:27 -03:00
Morgan Schweers	602b8c6210	Make new browser reloads recognize current model. (#4865)	2023-12-11 02:51:01 -03:00
oobabooga	8c8825b777	Add QuIP# to README	2023-12-08 08:40:42 -08:00