rename jamba example (#1846) [skip ci]
* rename jamba example * feat: change readme --------- Co-authored-by: Gal Cohen <galc@ai21.com>
This commit is contained in:
committed by
GitHub
parent
f07802f9fa
commit
957c956f89
@@ -6,5 +6,5 @@
|
||||
- ✅ qlora w/ deepspeed Zero-3 needs at least 2x GPUs and 67GiB VRAM (wtf?)
|
||||
- ✅ qlora single-gpu, ~51GiB VRAM
|
||||
- ✅ multipack
|
||||
- ❓ FSDP
|
||||
- ✅ FSDP
|
||||
- ❓ 8-bit LoRA
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
base_model: ai21labs/Jamba-v0.1
|
||||
base_model: ai21labs/AI21-Jamba-1.5-Large
|
||||
tokenizer_type: AutoTokenizer
|
||||
|
||||
load_in_4bit: true
|
||||
@@ -11,7 +11,7 @@ datasets:
|
||||
drop_system_message: true
|
||||
dataset_prepared_path: last_run_prepared
|
||||
val_set_size: 0.0
|
||||
output_dir: jamba-fsdp-qlora-ft
|
||||
output_dir: jamba-large-fsdp-qlora-ft
|
||||
save_safetensors: true
|
||||
adapter: qlora
|
||||
sequence_len: 2048
|
||||
Reference in New Issue
Block a user