Hugging Face's logo

Spaces:
Dovakiins
/
qwerrwe
Build error

App Files Community
qwerrwe / examples

Commit History

Feat: Add example for Mistral (#644)
eb41f76
unverified

Nanobit commited on

eval_table isn't quite stable enough to be in default llama configs (#637)
d887ad8
unverified

winglian commited on

Feat: Add support for upstream FA2 (#626)
19a600a
unverified

Nanobit commited on

default model changed
4fecbfe

mhenrichsen commited on

support to disable exllama for gptq (#604)
faecff9
unverified

winglian commited on

more sane defaults for openllama 3b used for quickstarts (#602)
674c576
unverified

winglian commited on

btlm and falcon monkey patches for flash attn (#566)
6b9b229
unverified

winglian commited on

make phi training work with Loras (#588)
62eaee7
unverified

winglian commited on

Support Sample packing for phi arch (#586)
12a2dbb
unverified

winglian commited on

Fix Codellama examples (#582)
1aa4007
unverified

Doan Minh Phuong commited on

Phi examples (#569)
2284209
unverified

winglian commited on

Add training callback to send predictions to WandB table (#521)
5b67ea9
unverified

Glavin001 commited on

recommend padding when using sample packing (#531)
3437149
unverified

winglian commited on

Add support for GPTQ using native transformers/peft (#468)
3355706
unverified

winglian commited on

pad_to_worst_case_seq_len boolean, for testing memory limits (#498)
8e197f6
unverified

Birch-san tmm1 commited on

Feat(cfg): Add code-llama configs for all sizes (#479)
3513071
unverified

mhenrichsen mhenrichsen commited on

Add example Llama 2 ReLoRA config (#471)
fe4d6ba
unverified

chargoddard commited on

improve llama pad token handling (#475)
cb9797e
unverified

winglian commited on

don't use mask expansion for inference (#392)
1687be6
unverified

winglian commited on

new llama-2 default settings (#370)
fdffef5
unverified

mhenrichsen Mads Henrichsen commited on

Add wandb_entity to wandb options, update example configs, update README (#361)
7019509
unverified

Morgan McGuire Morgan McGuire winglian commited on

set group_by_length to false in examples
36fefcf

tmm1 commited on

feat/llama-2 examples (#319)
dc71d88
unverified

mhenrichsen Mads Henrichsen commited on

Add XGen info to README and example config
3881143

ethanhs commited on

Use AutoTokenizer for redpajama example
945c419

sroecker commited on

Merge pull request #92 from OpenAccess-AI-Collective/flash-optimum
16bb627
unverified

winglian commited on

Merge branch 'main' into flash-optimum
fd2c981
unverified

winglian commited on

tweak config to work
2ba4ae8

winglian commited on

Merge pull request #193 from OpenAccess-AI-Collective/config-fixes-20230612
94f310c
unverified

winglian commited on

Fix config path after config moved
52cde69

Nanobit commited on

config fixes
9a58e99

winglian commited on

forgot to add this file
6b3f509

winglian commited on

update openllama and clean up paths
d0d7eaa

winglian commited on

more pruning
effbbf6

winglian commited on

more config pruning and migrating
c530e4b

winglian commited on

get rid of some configs, formalize pythioa lora config
77762a5

winglian commited on

address PR feedback
0c6f928

winglian commited on

linting fix
1db46a9

winglian commited on

use pythia-12b, neox-20b is flaky
3961902

winglian commited on

Merge pull request #132 from utensil/falcon-7b-qlora
c8242de
unverified

Nanobit commited on

Trim trailing whitespace
79a8f52
unverified

utensil commited on

Default `wandb_project` to empty as suggested
a52f481
unverified

utensil Nanobit commited on

Default micro_batch_size to 1 for a safer start
c9c0503
unverified

utensil commited on

Add comments/alternatives for falcon-qlora configs
ca11ae9
unverified

utensil commited on

falcon + qlora + xformer mbs 40 gas 2 on A6000
fb3d40f
unverified

utensil commited on

Create config-7b-qlora.yml
72bf8aa
unverified

utensil commited on

swap batch size for gradient accumulation steps to decouple from num gpu
c2a0792

winglian commited on

Merge pull request #105 from viktoriussuwandi/viktoriussuwandi-patch
4df9da7
unverified

winglian commited on

Merge pull request #106 from fearnworks/qlora-openllama-3b-example
2531ea2
unverified

winglian commited on

Lint and format
392dfd9

Nanobit commited on