From 5de1c02a7a4eaa72ff074003a972796b270f8986 Mon Sep 17 00:00:00 2001 From: "jack.butler" Date: Fri, 10 Feb 2023 15:18:03 +0000 Subject: [PATCH 1/4] add dummy yml config for reward --- model/reward/instructor/configs/dummy.yml | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) create mode 100644 model/reward/instructor/configs/dummy.yml diff --git a/model/reward/instructor/configs/dummy.yml b/model/reward/instructor/configs/dummy.yml new file mode 100644 index 00000000..6c2c9619 --- /dev/null +++ b/model/reward/instructor/configs/dummy.yml @@ -0,0 +1,21 @@ +model_name: X +tokenizer_name: X +max_length: X +num_train_epochs: X +warmup_steps: X +scheduler: X +learning_rate: X +deepspeed: X +fp16: X +local_rank: X +gradient_checkpointing: X +gradient_accumulation_steps: X +per_device_train_batch_size: X +per_device_eval_batch_size: X +weight_decay: X +max_grad_norm: X +eval_steps: X +save_steps: X +wandb_entity: X +datasets: + - X From 2ebb336141b0f41d8e0325e8fce68afd2dbf251e Mon Sep 17 00:00:00 2001 From: "jack.butler" Date: Fri, 10 Feb 2023 15:18:31 +0000 Subject: [PATCH 2/4] update readme --- model/reward/instructor/README.md | 25 +++++++++++++------------ 1 file changed, 13 insertions(+), 12 deletions(-) diff --git a/model/reward/instructor/README.md b/model/reward/instructor/README.md index 655d6469..1613da27 100644 --- a/model/reward/instructor/README.md +++ b/model/reward/instructor/README.md @@ -2,20 +2,21 @@ Trainer code based on huggingface. Compatible with deepspeed or accelerate -Requirements - -``` -wandb -evaluate -datasets -transformers -torch==1.12 -``` - -Start training reward model +Install Python requirements ```bash -python trainer.py configs/electra-base-dis-webgpt.yml +pip install -r requirements.txt +``` + +Write or inherit a `configs/.yml` file to store training +configuration details. + +> The configuration file must have _at least_ all the keys present in dummy.yml + +Run training procedure + +```bash +python trainer.py configs/.yml ``` Additional axis labeling, this outputs a 4 summary quality evaluation metrics From 3f9e2c31ac4c35e429be0f27838c4d96ce86a56a Mon Sep 17 00:00:00 2001 From: "jack.butler" Date: Fri, 10 Feb 2023 15:22:27 +0000 Subject: [PATCH 3/4] add hyperlink to dummy yml file --- model/reward/instructor/README.md | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/model/reward/instructor/README.md b/model/reward/instructor/README.md index 1613da27..4eb2fb2d 100644 --- a/model/reward/instructor/README.md +++ b/model/reward/instructor/README.md @@ -11,7 +11,8 @@ pip install -r requirements.txt Write or inherit a `configs/.yml` file to store training configuration details. -> The configuration file must have _at least_ all the keys present in dummy.yml +> The configuration file must have _at least_ all the keys present in +> [`configs/dummy.yml`](model/reward/instructor/configs/dummy.yml) Run training procedure From 7036df8dc077b2d58d9ebf2726aeca52b30e70f0 Mon Sep 17 00:00:00 2001 From: "jack.butler" Date: Fri, 10 Feb 2023 15:23:06 +0000 Subject: [PATCH 4/4] update link to relative link --- model/reward/instructor/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/model/reward/instructor/README.md b/model/reward/instructor/README.md index 4eb2fb2d..57e27483 100644 --- a/model/reward/instructor/README.md +++ b/model/reward/instructor/README.md @@ -12,7 +12,7 @@ Write or inherit a `configs/.yml` file to store training configuration details. > The configuration file must have _at least_ all the keys present in -> [`configs/dummy.yml`](model/reward/instructor/configs/dummy.yml) +> [`configs/dummy.yml`](configs/dummy.yml) Run training procedure