From 9a58e99e812f1e4074da02ff56529b4986563931 Mon Sep 17 00:00:00 2001 From: Wing Lian Date: Mon, 12 Jun 2023 01:52:58 -0400 Subject: [PATCH] config fixes --- examples/falcon/config-7b-lora.yml | 2 +- examples/falcon/config-7b.yml | 2 +- examples/openllama-3b/config.yml | 4 ++-- examples/openllama-3b/lora.yml | 2 +- 4 files changed, 5 insertions(+), 5 deletions(-) diff --git a/examples/falcon/config-7b-lora.yml b/examples/falcon/config-7b-lora.yml index 090cc6bcf..8aa585851 100644 --- a/examples/falcon/config-7b-lora.yml +++ b/examples/falcon/config-7b-lora.yml @@ -23,7 +23,7 @@ lora_dropout: 0.0 lora_target_modules: lora_target_linear: true lora_fan_in_fan_out: -wandb_project: falcon-7b +wandb_project: wandb_watch: wandb_run_id: wandb_log_model: diff --git a/examples/falcon/config-7b.yml b/examples/falcon/config-7b.yml index dc67d6125..b267566ce 100644 --- a/examples/falcon/config-7b.yml +++ b/examples/falcon/config-7b.yml @@ -23,7 +23,7 @@ lora_dropout: 0.0 lora_target_modules: lora_target_linear: true lora_fan_in_fan_out: -wandb_project: falcon-7b +wandb_project: wandb_watch: wandb_run_id: wandb_log_model: diff --git a/examples/openllama-3b/config.yml b/examples/openllama-3b/config.yml index 248b740ff..6fd704ffc 100644 --- a/examples/openllama-3b/config.yml +++ b/examples/openllama-3b/config.yml @@ -25,7 +25,7 @@ wandb_project: wandb_watch: wandb_run_id: wandb_log_model: -output_dir: ./lora-out +output_dir: ./openllama-out batch_size: 16 micro_batch_size: 4 num_epochs: 3 @@ -43,7 +43,7 @@ early_stopping_patience: resume_from_checkpoint: local_rank: logging_steps: 1 -xformers_attention: +xformers_attention: true flash_attention: gptq_groupsize: gptq_model_v1: diff --git a/examples/openllama-3b/lora.yml b/examples/openllama-3b/lora.yml index 98e2c2adc..d1f252455 100644 --- a/examples/openllama-3b/lora.yml +++ b/examples/openllama-3b/lora.yml @@ -49,7 +49,7 @@ early_stopping_patience: resume_from_checkpoint: local_rank: logging_steps: 1 -xformers_attention: +xformers_attention: true flash_attention: gptq_groupsize: gptq_model_v1: