Convert checkpoints and configurations#


We have designed the deep learning models in the Cerebras Model Zoo as highly generalizable, allowing users to easily make architectural modifications from a single configuration file. Unfortunately, the general model implementations make it difficult to directly use model configs and checkpoints from other code repositories (e.g., Hugging Face) with Cerebras Model Zoo. Therefore, we offer you the Checkpoint and Config Converter tool, that was built to allow users to easily convert model implementations between Cerebras Model Zoo and other code repositories.

Some use cases include:

  • Take a pretrained checkpoint from another code repository and convert it into the equivalent Cerebras Model Zoo compatible config & checkpoint so that you can continue training on the CS system.

  • Train a model within the Cerebras ecosystem using the Cerebras Model Zoo, then convert to another equivalent implementation (e.g., Hugging Face) to run inference.

  • “Upgrade” an old Cerebras Model Zoo config and checkpoint to a new one (e.g., convert Cerebras Model Zoo rel 1.6 checkpoints to 1.7) to ensure that old checkpoints can continue to be used in new releases if our model implementations evolve.


We only support conversions between Hugging Face (HF) and Cerebras Model Zoo (CS) implementations.




The tool offers three commands:

Table 7 :header-rows: 1#




Displays all the available conversions (models and formats)


Performs only config conversion. If you intend to convert from Cerebras Model Zoo to another repository at any point, we recommend running config conversion before training the model. This way, you can determine whether your configuration within Cerebras Model Zoo is a candidate for conversion. Conversions are only sometimes possible as other repositories are less general than our Cerebras Model Zoo implementations (e.g., many Hugging Face NLP model implementations support a limited range of positional embeddings).


Performs both config and checkpoint conversion. In other words, the tool is supplied the old config and checkpoint and produces a new config and checkpoint.


Cerebras configuration files contain model parameters and configurations for the optimizer, train_input, eval_input, and runconfig. Most other open-source repositories (ex: Hugging Face) need this information. Since these cannot be inferred by the converter tool, you will need to modify the output config with these additional properties. As a starter, you can look at the example configs in Cerebras Model Zoo.


The three commands introduced above are used as follows:

  1. To get a list of all models/conversions that we support, use the following command:

(venv_cerebras_pt) $ python <modelzoo path>/modelzoo/common/pytorch/model_utils/ \


It is essential that you read the notes section of the list command output before using the converter! This section explains the exact model classes that are being converted from/to. It also lists any caveats about the conversion process. For example, many NLP models offer -headless variants which are missing a language model head.

  1. To convert a config file only, use the following command:

(venv_cerebras_pt) $ python <modelzoo path>/modelzoo/common/pytorch/model_utils/ \
  convert-config \
  --model <model name> \
  --src-fmt <format of input config> \
  --tgt-fmt <format of output config> \
  --output-dir <location to save output config> \
  <config file path>
  1. To convert a checkpoint and its corresponding config, use the following command:

(venv_cerebras_pt) $ python <modelzoo path>/modelzoo/common/pytorch/model_utils/ \
  convert \
  --model <model name> \
  --src-fmt <format of input checkpoint> \
  --tgt-fmt <format of output checkpoint> \
  --output-dir <location to save output checkpoint> \
  <input checkpoint file path> \
  --config <input config file path>

To learn more about usage and optional parameters about a particular subcommand, you can pass the -h flag. For example:

(venv_cerebras_pt) $ python <modelzoo path>/modelzoo/common/pytorch/model_utils/ \
  convert -h

Models supported#

The following is a list of models supported by the Checkpoint and Config Converter tool:























Converting Eleuther AI GPT-J 6B (from model card) to Cerebras Model Zoo#

Eleuther’s final GPT-J checkpoint can be accessed on Hugging Face at EleutherAI/gpt-j-6B. Rather than manually entering the values from the model architecture table into a config file and writing a script to convert their checkpoint, we can auto-generate these with a single command.

First, we need to download the config and checkpoint files from the model card locally:

$ mkdir opensource_checkpoints
$ wget -P opensource_checkpoints
$ wget -P opensource_checkpoints


Use the appropriate https link when downloading files from Hugging Face model card pages. For config files, use the path that contains …/raw/…; for checkpoint files, use the path that contains …/resolve/….

Hugging Face configs contain the architecture property, which specifies the class with which the checkpoint was generated. According to config.json, the HF checkpoint is from the GPTJForCausalLM class. Using this information, we can use the checkpoint converter tool’s list command to find the appropriate converter. In this case, we want to use the gptj model, with a source format of hf, and a target format of cs-1.9.

Now to convert the config & checkpoint, run the following command:

(venv_cerebras_pt) $ python <modelzoo path>/modelzoo/common/pytorch/model_utils/ \
   convert \
   --model gptj \
   --src-fmt hf \
   --tgt-fmt cs-1.9 \
   --output-dir opensource_checkpoints/ \
   opensource_checkpoints/pytorch_model.bin \
   --config opensource_checkpoints/config.json

This produces two files:

  • opensource_checkpoints/pytorch_model_to_cs-1.9.mdl

  • opensource_checkpoints/config_to_cs-1.9.yaml

The output YAML config file contains the auto-generated model parameters from the Eleuther implementation. Before you can train/eval the model on the Cerebras cluster, add the train_input, eval_input, optimizer, and runconfig parameters to the YAML. Examples for these parameters can be found in the configs/ folder for each model within Model Zoo. In this case, we can copy the missing information from modelzoo/transformers/pytorch/gptj/configs/params_gptj_6B.yaml into opensource_checkpoints/config_to_cs-1.9.yaml. Make sure you modify the dataset paths under train_input and eval_input if they are stored elsewhere.

The following command demonstrates using the converted config and checkpoint for continuous pretraining:

(venv_cerebras_pt) $ python <modelzoo path>/modelzoo/transformers/pytorch/gptj/ \
  CSX \
  --mode train \
  --params opensource_checkpoints/config_to_cs-1.9.yaml \
  --checkpoint_path opensource_checkpoints/pytorch_model_to_cs-1.9.mdl \
  --model_dir gptj6b_continuous_pretraining \
  --mount_dirs {paths to modelzoo and to data} \
  --python_paths {paths to modelzoo and other python code if used}


First navigate to the directory of the model (GPT-J in this case) before executing Additional details about the command can be found on the Launch your job page.

Converting a Hugging Face model without a model card to Cerebras Model Zoo#

Not all pretrained checkpoints on Hugging Face have corresponding model card web pages. You can still download these checkpoints and configs to convert them into a Model Zoo-compatible format.

For example, Hugging Face has a model card for BertForMaskedLM accessible through the name bert-base-uncased. However, it doesn’t have a webpage for BertForPreTraining, which we’re interested in.

We can manually get the config and checkpoint for this model as follows:

>>> from transformers import BertForPreTraining
>>> model = BertForPreTraining.from_pretrained("bert-base-uncased")
>>> model.save_pretrained("bert_checkpoint")

This saves two files: bert_checkpoint/config.json and bert_checkpoint/pytorch_model.bin

Now that you have downloaded the required files, you can convert the checkpoints. Use the --model bert flag since the Hugging Face checkpoint is from the BertForPreTraining class. If you want to use another checkpoint from a different variant (such as a finetuning model), see the other bert- model converters.

The final conversion command is:

(venv_cerebras_pt) $ python <modelzoo path>/modelzoo/common/pytorch/model_utils/ \
   convert \
   --model bert \
   --src-fmt hf \
   --tgt-fmt cs-1.9 \
   bert_checkpoint/pytorch_model.bin \
   --config bert_checkpoint/config.json
Checkpoint saved to bert_checkpoint/pytorch_model_to_cs-1.9.mdl
Config saved to bert_checkpoint/config_to_cs-1.9.yaml

Converting Cerebras Model Zoo GPT-2 checkpoint to Hugging Face#

Suppose you just finished training GPT-2 on CS and want to run the model within the Hugging Face ecosystem. In this example, the configuration file is saved at model_dir/train/params_train.yaml and the checkpoint (corresponding to step 10k) is at model_dir/checkpoint_10000.mdl

To convert the Hugging Face, run the following command:

(venv_cerebras_pt) $ python <modelzoo path>/modelzoo/common/pytorch/model_utils/ \
  convert \
  --model gpt2 \
  --src-fmt cs-1.9 \
  --tgt-fmt hf \
  model_dir/checkpoint_10000.mdl \
  --config model_dir/train/params_train.yaml

Since the --output-dir flag is omitted, the two output files are saved to the same directories as the original files: model_dir/train/params_train_to_hf.json and model_dir/checkpoint_10000_to_hf.bin

Converting a GPT2 muP checkpoint to Hugging Face#

Hugging Face does not support the muP model. However, if you have a Cerebras GPT2/3 checkpoint that uses muP, it is possible to convert it to a Hugging Face model to run inference. This process only works for models that can be converted to a Hugging Face GPT2 model, in particular, it is not compatible with models that use Alibi or Swiglu.

Proceed with the following steps to convert:

  1. Use the transformers/pytorch/gpt2/scripts/ script to fold the muP scaling constants into the weights of the model. This script takes a path to a muP checkpoint and its associated config file and outputs a folded checkpoint and a matching params file. For example,

# Create /path/to/sP/output/checkpoint.mdl and /path/to/sP/output/params.yaml
$ python --src /path/to/muP/checkpoint.mdl --params /path/to/muP/params.yaml --dest /path/to/sP/output/checkpoint.mdl
  1. Once you have folded the muP constants into the weights of the model, use the existing monolith checkpoint conversion scripts to convert. For example,

$ python convert /path/to/sp/output/checkpoint.mdl --config /path/to/sP/output/params.yaml --src-fmt cs-1.9 --tgt-fmt hf --output-dir /path/to/hf/output/dir

Upgrading Checkpoints and Configs to the current release#

As our Model Zoo implementations evolve over time, the changes may sometimes break out-of-the-box compatibility when moving to a new release. To ensure that you can continue using your old checkpoints, we offer converters that allow you to “upgrade” configs and checkpoints when necessary. The section below covers conversions that are required when moving to a particular release. If a converter doesn’t exist, no explicit conversion is necessary.

Release 1.9#

All configs & checkpoints from 1.8 can continue to be used in 1.9 without any conversion.

Release 1.8#

T5 / Vanilla Transformer

As described in the release notes, the behavior of the use_pre_encoder_decoder_layer_norm flag has been flipped. In order to continue using rel 1.7 checkpoints in rel 1.8, you’ll need to update the config to reflect this change. You can do this automatically using the config converter tool as follows:

(venv_cerebras_pt) $ python <modelzoo path>/modelzoo/common/pytorch/model_utils/ \
   convert-config \
   --model <model type> \
   --src-fmt cs-1.7 \
   --tgt-fmt cs-1.8 \
   <config file path>

In the command above, --model should be either t5 or transformer depending on which model you’re using. The config file path should point to the train/params_train.yaml file within your model directory.


As described in the release notes, we expanded the BERT model configurations to expose two additional parameters: pooler_nonlinearity and mlm_nonlinearity. Due to a change in the default value of the mlm_nonlinearity parameter, you will need to update the config when using a rel 1.7 checkpoint in rel 1.8. You can do this automatically using the config converter tool as follows:

(venv_cerebras_pt) $ python <modelzoo path>/modelzoo/common/pytorch/model_utils/ \
  convert-config \
  --model bert \
  --src-fmt cs-1.7 \
  --tgt-fmt cs-1.8 \
  <config file path>

The config file path should point to the train/params_train.yaml file within your model directory.

Frequently Asked Questions#

Table 8 :widths: 40 60 :header-rows: 1#



Which models, formats, classes, etc, are supported?

See the list command under the usage section.

Which frameworks are supported?

PyTorch only.

Does the optimizer state get converted?

No. Hugging Face checkpoints contain model state information only; unlike CS, they do not contain optimizer state information.

Sometimes, when I run the checkpoint converter tool, it runs for a while before saying Killed. What happened?

The program hit the memory limit. Unfortunately, PyTorch pickling works by storing whole checkpoints in the same file, forcing everything to be read into memory at once. Ensure that the system you’re running on has at least as much RAM as the size of the checkpoint file.

Conversion failed with a ConfigConversionError. Why did this happen?

Conversions are only sometimes possible as other repositories are less general than our Model Zoo implementations (ex: many Hugging Face NLP model implementations support limited types of positional embeddings while Model Zoo includes an expanded range). For this reason, we recommend that you run config conversion before training the model if you intend to convert a Model Zoo model to another repository at any time. This will allow you to determine if the configuration you are using within Model Zoo can be converted before you train the model.

Additionally, you can use the information in the error message to modify the config file to generate a configuration that can be converted.

Sometimes during config conversion, I see the following:
WARNING:root:Key not matched:
Should I be concerned?

No. Not all keys in one config format must be converted to another. This warning message is simply printing out the keys that will be ignored.

For example HF configs contain the use_cache parameter which isn’t relevant on CS.

Model conversion failed with the following error:
AssertionError: Unable to match all keys. If you want to proceed by dropping keys that couldn't be matched, rerun with --drop-unmatched-keys
What should I do?

The checkpoint contains keys that weren’t expected, and, therefore, couldn’t be converted. The converters are heavily tested, so this error message highlights an issue with the input checkpoint or the command being run, not the converter itself.

Make sure that you are using the correct --model and --src-fmt flags corresponding to the checkpoint you are using. To double-check, look at the notes column displayed by the checkpoint converter tool’s list command. A misspecified --model or --src-fmt will lead to this error.

All unexpected keys in the checkpoint are displayed with WARNING:root:Key not matched:. If these keys do not need to be converted, you can bypass the assertion using the --drop-unmatched-keys flag. It would help if you never had to use this feature unless you’re using a custom checkpoint that deviates from the --src-fmt format.

I am unable to use a converted checkpoint because I get the following errors:
Error(s) in loading state_dict for <model name>:
Missing key(s) in state_dict:...
Unexpected key(s) in state_dict:...
What should I do?

There is a discrepancy between the format of the converted checkpoint and the expected format that you’re loading the model into. This is caused by a misspecified --model or --tgt-fmt flags.

To double-check that you’re using the correct flags, look at the notes column displayed by the checkpoint converter tool’s list command.

I have a sharded checkpoint. How do I use the checkpoint converter tool?

Starting 1.9, the checkpoint & config converter tool support sharded Hugging Face checkpoints. To convert from a sharded HF checkpoint, download all shards (Pytorch .bin files) and the weight index file (pytorch_model.bin.index.json), which contains the mapping between model layers and shard files. The only difference when using the checkpoint converter tool’s CLI with a sharded checkpoint is that the checkpoint path argument must point to the .index.json file rather than the singular .bin file as before. By doing so, the tool can find all the corresponding shards, merge them, and convert the model.

An alternate, manual approach for converting sharded checkpoints is to merge them into a single file:
import torch
shard1 = torch.load("<shard 1>.bin")
shard2 = torch.load("<shard 2>.bin")
shard1.update(shard2), "full_checkpoint.bin")
# now use converter tool on full checkpoint.