147 Commits

Author SHA1 Message Date
Wing Lian
f3a5d119af fix env var extraction (#2043) [skip ci] 2024-11-14 12:58:06 -05:00
Wing Lian
76883851d2 add warning that sharegpt will be deprecated (#1957)
* add warning that sharegpt will be deprecated

* add helper script for chat_templates and document deprecation

* Update src/axolotl/prompt_strategies/sharegpt.py

Co-authored-by: NanoCode012 <nano@axolotl.ai>

---------

Co-authored-by: NanoCode012 <nano@axolotl.ai>
2024-10-11 13:33:20 -04:00
mhenrichsen
1194c2e0b1 github urls (#1734)
Co-authored-by: Henrichsen, Mads (ext) <mads.henrichsen.ext@siemens-energy.com>
2024-07-11 09:19:29 -04:00
Wing Lian
0c49ecc429 more fixes to work with runpod + skypilot (#1629) 2024-05-16 00:05:56 -04:00
Wing Lian
2501a371c6 fix setting the authorized keys when there are more than one in the env var (#1626) 2024-05-15 20:48:56 -04:00
Wing Lian
e6937e884b fix symlinks for axolotl outputs (#1625) 2024-05-15 19:41:45 -04:00
Wing Lian
120b809465 fix for jupyterlab on cloud start (#1594) 2024-05-05 10:08:43 -04:00
Wing Lian
d113331e9a add a helpful motd for cloud image (#1235) [skip ci] 2024-01-31 10:26:02 -05:00
Wing Lian
eaaeefce55 jupyter lab fixes (#1139) [skip ci]
* add a basic notebook for lab users in the root

* update notebook and fix cors for jupyter

* cell is code

* fix eval batch size check

* remove intro notebook
2024-01-22 18:42:40 -05:00
Wing Lian
cbecf3e62a fix check for env var (#1151) 2024-01-18 23:58:11 -05:00
Wing Lian
729740df81 Dockerfile cloud ports (#1148)
* explicitly expose ports 8888 and 22

* support for SSH_KEY from latitude
2024-01-18 22:04:25 -05:00
Wing Lian
ece0211996 Agnostic cloud gpu docker image and Jupyter lab (#1097) 2024-01-15 22:37:54 -05:00
Casper
e50ab072e2 Create preprocess CLI (#785)
* Create preprocess CLI

* Print prompt template if debugging

* Add print for unsupported prompters

* Formatting

* Formatting

* Refactor variables

* Formatting

* Formatting

* Formatting

* Formatting
2023-10-26 09:35:42 -04:00
Napuh
85b0be2ba7 Warn users to login to HuggingFace (#645)
* added warning if user is not logged in HF

* updated doc to suggest logging in to HF
2023-09-27 17:43:35 -04:00
Wing Lian
861cecac2a refactor scripts/finetune.py into new cli modules (#550)
* refactor scripts/finetune.py into new cli modules

* continue to support scripts/finetune.py

* update readme with updated cli commands

* Update scripts/finetune.py

Co-authored-by: NanoCode012 <kevinvong@rocketmail.com>

---------

Co-authored-by: NanoCode012 <kevinvong@rocketmail.com>
2023-09-15 01:43:52 -04:00
Wing Lian
fdb777bc06 check for the existence of the default accelerate config that can create headaches (#561) 2023-09-13 10:38:28 -04:00
Wing Lian
bf0804447c fix wandb so mypy doesn't complain (#562)
* fix wandb so mypy doesn't complain

* fix wandb so mypy doesn't complain

* no need for mypy override anymore
2023-09-13 10:36:16 -04:00
Aman Gupta Karmani
44454ae4c4 move is_llama_derived_model into normalize_config (#524) 2023-09-04 00:19:03 -04:00
Tom Jobbins
48434bec54 Debug tokenization output: Add ability to output text only (no tokens), and/or specify num samples to see (#511) 2023-08-31 14:26:52 -07:00
Wing Lian
b21e4a20fe split train from other cli options (#503) 2023-08-30 22:01:47 -07:00
Wing Lian
548787daae customizable ascii art (#506) 2023-08-29 10:13:42 -07:00
Maxime
36b2e1cfee tweak: use default config file when only one file is present (#501) 2023-08-29 06:17:10 -07:00
Wing Lian
125cccb786 Refactor train cfg cli (#499)
* wip to cleanup cfg cli options

* fix launcher

* fix cli args
2023-08-29 05:37:53 -07:00
Maxime
17605b85d8 fix: inference did not move the model to the correct device (#483) 2023-08-26 16:40:56 -04:00
Charles O. Goddard
bde3c5a478 ReLoRA implementation (with quantization) (#322)
* Experimental ReLoRA (+qlora) implementation

* Add CPU offload

* Remove local config

* Fix saving logic

* Remove redundant assert

* Fix logic errors

* Move ReLoRA into its own trainer class with a method override to create the proper scheduler

* Formatting & typing fixes

* Use safe_serialization

* Don't allow fsdp/deepspeed with ReLoRA

* Fix cpu-offload logic, enable multi gpu

* Document parameters and add comment

* Fix merge issue

* Smooth over some sharp edges

* Implement resume from checkpoint for relora

* Address review comments

* Fix saving logic

* Add necessary metadata to safetensors

---------

Co-authored-by: Wing Lian <wing.lian@gmail.com>
2023-08-23 23:07:18 -04:00
NanoCode012
04a42b6db1 feat(docs): improve user customized prompts (#443)
* feat(docs): improve user customized prompts

* feat(doc): add custom pretokenized instructions

* chore: clean old data folder

* chore: add new line
2023-08-20 23:59:43 -04:00
Wing Lian
29241cf1e4 Ax art (#405)
* axolotl text art :D

* only print art on rank0

* lint and pr feedback
2023-08-15 08:34:30 -04:00
Aman Karmani
2e22404d2d add utils.data.prepare_dataset 2023-08-14 21:28:29 -07:00
Wing Lian
fc2d6be96d use context manager to run things on rank0 before others (#397) 2023-08-15 00:10:47 -04:00
Gabriel Puliatti
3c2ad00d07 Feat(config): add max steps (#387) 2023-08-14 11:19:29 -04:00
Wing Lian
86a91e260b save tokenizer before training starts (#380) 2023-08-13 11:28:58 -04:00
Aman Karmani
efb3b2c95e simplify load_tokenizer 2023-08-12 18:55:06 -07:00
Aman Karmani
7b55fe6419 improve GPU logging to break out pytorch cache and system mem 2023-08-12 18:52:57 -07:00
Aman Karmani
8cec513447 extract module for working with cfg 2023-08-12 18:25:27 -07:00
Wing Lian
2bb0b78975 Attention mask and position id fixes for packing (#285)
* fix attetion mask with packing

* set position ids and use block diagonal attn mask

* fix expand mask for multiple batch items, make sure we pad position_ids

* don't move masks to cpu

* use multi pack dataloader w random sampler

* add position_ids back

* more fixes for dataloader integration

* est total tokens, fix field loop

* more fixes, position_ids seems broken

* more fixes for sample packing

* use distributed sampler, avoid accelerate prepare

* use accelerator prepare for dataloader

* fix for position_ids w packing

* Update src/axolotl/utils/dataloader.py

* validation for sample packing and doc

* more fixes for 4k and optimizations

* optimized expand mask fn

* better handling of variance in multipack dataloader length and trainer hanging when it runs out of data

* fix rounding of len of batches to int

* better handling so that all devices have the same dataloader len

* fix step calc for packing

* pass sample packing efficiency to training args

* add a test for the mask expansion for sequence packing

* only process eval dataset for packing if not None

* don't split batches when packing

* weighted CE losses

* weighted CEL fixes

* limit packing to sequences of max seq len

* seq_len_multiple for packing

* make sure the chunk size is an int

* sample_packing_seq_len_multiplier config

* use cumulative seq len with var len flash attn v2 w packing

* properly calculate max len

* fix flash-attn, xformers, packing, support chatml

* fix chatml system prompt for openorca, legacy tokenizer opts

* add chatml

* add unit tests for cum seq lens, add ability to build cu_seq_lens from positional ids, fix prompt test

* fix test and pylint checks

* more packing and dataset optimizations and fixes

* filter w multiple cpus

* more fixes and optimizations

* fixes and go back to distributed sampler since batch sampler won't work

* fix counts by accounting for num devices

* fix steps calculation

* previous accelerate is still most performant

* add numba to requirements.

* use custom distributed checks

* fix sampler to prevent overfit w new epochs

* let's not cleanup the cached datasets

* calculate cum seq lens with pos_ids instead of mask, simplify packing params, fix distributed barrier

* speed optimizations and set accelerate fsdp env vars

* optimize dataset concatenation?

* more optimizations for dataset handling

* fix import for annotation

* manual pre-commit fixes

* another sum optimization and bug fix for calc steps

* fix packing estimations

* fix formatting

* pylint problems

* add back flash attention branch for handling unpacked sequences seperately

* Address PR feedback

* add optional sample packing config params to readme
2023-08-12 15:14:56 -04:00
NanoCode012
a276c9c88d Fix(save): Save as safetensors (#363) 2023-08-13 01:22:52 +09:00
NanoCode012
289d5c403d feat(merge): save tokenizer on merge (#362) 2023-08-13 00:18:10 +09:00
Aman Gupta Karmani
11ddccb80f Merge pull request #356 from tmm1/load_model-args
simplify `load_model` signature
2023-08-09 18:24:34 -07:00
Aman Karmani
718102271f simplify load_model signature 2023-08-09 22:36:02 +00:00
Aman Karmani
e303d64728 log GPU memory usage 2023-08-09 18:26:28 +00:00
Wing Lian
894cba09f3 fix FSDP save of final model (#329) 2023-07-30 21:46:44 -04:00
Wing Lian
cf62cfd661 add runpod envs to .bashrc, fix bnb env (#316)
* hopper support for base dockerfile, add runpod envs to .bashrc

* set BNB_CUDA_VERSION env for latest bnb

* don't support hopper yet w 118
2023-07-22 10:09:38 -04:00
Wing Lian
d75adb9835 misc fixes 2023-07-17 03:00:27 -04:00
theobjectivedad
b1f4f7a34d Fixed pre-commit problems, fixed small bug in logging_config to handle LOG_LEVEL env var 2023-07-15 12:29:35 +00:00
theobjectivedad
553a86b52c Adding logging enhancement 2023-07-14 07:26:19 -05:00
Wing Lian
16bb6276a5 Merge pull request #92 from OpenAccess-AI-Collective/flash-optimum
add support for opimum bettertransformers
2023-06-14 07:50:15 -04:00
NanoCode012
dc77c8ebce chore: Refactor inf_kwargs out 2023-06-13 12:01:46 +09:00
Wing Lian
fd2c9814c9 Merge branch 'main' into flash-optimum 2023-06-12 13:12:15 -04:00
Wing Lian
8002ffb41f Merge pull request #177 from NanoCode012/fix/landmark-patch
Fix landmark attention patch
2023-06-12 08:27:12 -04:00
NanoCode012
8e568bbdae Merge pull request #159 from AngainorDev/patch-1
Fix training over existing lora
2023-06-12 20:27:11 +09:00