Archives/gpt4all
2
0
Fork 0
mirror of https://github.com/nomic-ai/gpt4all synced 2024-11-08 07:10:32 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix: gptj multinode

Browse Source
This commit is contained in:
Zach Nussbaum 2023-04-05 02:52:44 +00:00
parent 0c8666f635
commit 98ceac34e1
2 changed files with 9 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
configs/deepspeed/ds_config_gptj.json
View File

@ -1,6 +1,6 @@
{ {
"train_batch_size": "auto", "train_batch_size": "auto",
"gradient_accumulation_steps": 4, "gradient_accumulation_steps": "auto",
"train_micro_batch_size_per_gpu": "auto", "train_micro_batch_size_per_gpu": "auto",
"fp16": { "fp16": {
"enabled": "auto", "enabled": "auto",
@ -19,7 +19,7 @@
"device": "none" "device": "none"
}, },
"offload_optimizer": { "offload_optimizer": {
"device": "none" "device": "cpu"
}, },
"allgather_partitions": true, "allgather_partitions": true,
"allgather_bucket_size": 5e8, "allgather_bucket_size": 5e8,

14
configs/train/finetune_gptj.yaml
View File

@ -2,24 +2,24 @@
model_name: "EleutherAI/gpt-j-6B" model_name: "EleutherAI/gpt-j-6B"
tokenizer_name: "EleutherAI/gpt-j-6B" tokenizer_name: "EleutherAI/gpt-j-6B"
gradient_checkpointing: true gradient_checkpointing: true
save_name: "nomic-ai/gpt4all-gptj-multiturn-lr-aggressive" save_name: "nomic-ai/gpt4all-gptj-multinode"
# dataset # dataset
streaming: false streaming: false
num_proc: 64 num_proc: 64
dataset_path: "data_multiplus" dataset_path: "data_multiplus"
max_length: 1024 max_length: 1024
batch_size: 8 batch_size: 32
# train dynamics # train dynamics
lr: 2.0e-5 lr: 4.0e-5
min_lr: 0 min_lr: 0
weight_decay: 0.0 weight_decay: 0.0
eval_every: 200 eval_every: 100
eval_steps: 105 eval_steps: 105
save_every: 400 save_every: 100
log_grads_every: 200 log_grads_every: 100
output_dir: "ckpts/gpt4all-gptj-full-multiturn-lr-aggreive" output_dir: "ckpts/gpt4all-gptj-multinode"
checkpoint: null checkpoint: null
lora: false lora: false
warmup_steps: 500 warmup_steps: 500