Skip to content

Latest commit

 

History

History
317 lines (254 loc) · 13.1 KB

download_model_weights.md

File metadata and controls

317 lines (254 loc) · 13.1 KB
Raw

Download Model Weights with LitGPT

LitGPT supports a variety of LLM architectures with publicly available weights. You can download model weights and access a list of supported models using the LitGPT download.py script.

Model Model size Reference
CodeGemma by Google 7B Google Team, Google Deepmind
Code Llama by Meta AI 7B, 13B, 34B, 70B Rozière et al. 2023
Danube2 by H2O.ai 1.8B H2O.ai
Dolly by Databricks 3B, 7B, 12B Conover et al. 2023
Falcon by TII UAE 7B, 40B, 180B TII 2023
FreeWilly2 (Stable Beluga 2) by Stability AI 70B Stability AI 2023
Function Calling Llama 2 by Trelis 7B Trelis et al. 2023
Gemma by Google 2B, 7B Google Team, Google Deepmind
Llama 2 by Meta AI 7B, 13B, 70B Touvron et al. 2023
Llama 3 by Meta AI 8B, 70B Meta AI 2024
LongChat by LMSYS 7B, 13B LongChat Team 2023
Mistral and Mixtral by Mistral AI 7B Mistral website
Nous-Hermes by NousResearch 7B, 13B, 70B Org page
OpenLLaMA by OpenLM Research 3B, 7B, 13B Geng & Liu 2023
Phi by Microsoft Research 1.3B, 2.7B Li et al. 2023
Platypus by Lee at el. 7B, 13B, 70B Lee, Hunter, and Ruiz 2023
Pythia by EleutherAI {14,31,70,160,410}M, {1,1.4,2.8,6.9,12}B Biderman et al. 2023
RedPajama-INCITE by Together 3B, 7B Together 2023
StableCode by Stability AI 3B Stability AI 2023
StableLM by Stability AI 3B, 7B Stability AI 2023
StableLM Zephyr by Stability AI 3B Stability AI 2023
TinyLlama by Zhang et al. 1.1B Zhang et al. 2023
Vicuna by LMSYS 7B, 13B, 33B Li et al. 2023

 

General Instructions

1. List Available Models

To see all supported models, run the following command without arguments:

litgpt download

The output is shown below:

codellama/CodeLlama-13b-hf
codellama/CodeLlama-13b-Instruct-hf
codellama/CodeLlama-13b-Python-hf
codellama/CodeLlama-34b-hf
codellama/CodeLlama-34b-Instruct-hf
codellama/CodeLlama-34b-Python-hf
codellama/CodeLlama-70b-hf
codellama/CodeLlama-70b-Instruct-hf
codellama/CodeLlama-70b-Python-hf
codellama/CodeLlama-7b-hf
codellama/CodeLlama-7b-Instruct-hf
codellama/CodeLlama-7b-Python-hf
databricks/dolly-v2-12b
databricks/dolly-v2-3b
databricks/dolly-v2-7b
EleutherAI/pythia-1.4b
EleutherAI/pythia-1.4b-deduped
EleutherAI/pythia-12b
EleutherAI/pythia-12b-deduped
EleutherAI/pythia-14m
EleutherAI/pythia-160m
EleutherAI/pythia-160m-deduped
EleutherAI/pythia-1b
EleutherAI/pythia-1b-deduped
EleutherAI/pythia-2.8b
EleutherAI/pythia-2.8b-deduped
EleutherAI/pythia-31m
EleutherAI/pythia-410m
EleutherAI/pythia-410m-deduped
EleutherAI/pythia-6.9b
EleutherAI/pythia-6.9b-deduped
EleutherAI/pythia-70m
EleutherAI/pythia-70m-deduped
garage-bAInd/Camel-Platypus2-13B
garage-bAInd/Camel-Platypus2-70B
garage-bAInd/Platypus-30B
garage-bAInd/Platypus2-13B
garage-bAInd/Platypus2-70B
garage-bAInd/Platypus2-70B-instruct
garage-bAInd/Platypus2-7B
garage-bAInd/Stable-Platypus2-13B
google/codegemma-7b-it
google/gemma-2b
google/gemma-2b-it
google/gemma-7b
google/gemma-7b-it
h2oai/h2o-danube2-1.8b-chat
lmsys/longchat-13b-16k
lmsys/longchat-7b-16k
lmsys/vicuna-13b-v1.3
lmsys/vicuna-13b-v1.5
lmsys/vicuna-13b-v1.5-16k
lmsys/vicuna-33b-v1.3
lmsys/vicuna-7b-v1.3
lmsys/vicuna-7b-v1.5
lmsys/vicuna-7b-v1.5-16k
meta-llama/Llama-2-13b-chat-hf
meta-llama/Llama-2-13b-hf
meta-llama/Llama-2-70b-chat-hf
meta-llama/Llama-2-70b-hf
meta-llama/Llama-2-7b-chat-hf
meta-llama/Llama-2-7b-hf
meta-llama/Meta-Llama-3-70B
meta-llama/Meta-Llama-3-70B-Instruct
meta-llama/Meta-Llama-3-8B
meta-llama/Meta-Llama-3-8B-Instruct
microsoft/phi-1_5
microsoft/phi-2
mistralai/Mistral-7B-Instruct-v0.1
mistralai/Mistral-7B-Instruct-v0.2
mistralai/Mistral-7B-v0.1
mistralai/Mixtral-8x7B-Instruct-v0.1
mistralai/Mixtral-8x7B-v0.1
NousResearch/Nous-Hermes-13b
NousResearch/Nous-Hermes-llama-2-7b
NousResearch/Nous-Hermes-Llama2-13b
openlm-research/open_llama_13b
openlm-research/open_llama_3b
openlm-research/open_llama_7b
stabilityai/FreeWilly2
stabilityai/stable-code-3b
stabilityai/stablecode-completion-alpha-3b
stabilityai/stablecode-completion-alpha-3b-4k
stabilityai/stablecode-instruct-alpha-3b
stabilityai/stablelm-3b-4e1t
stabilityai/stablelm-base-alpha-3b
stabilityai/stablelm-base-alpha-7b
stabilityai/stablelm-tuned-alpha-3b
stabilityai/stablelm-tuned-alpha-7b
stabilityai/stablelm-zephyr-3b
tiiuae/falcon-180B
tiiuae/falcon-180B-chat
tiiuae/falcon-40b
tiiuae/falcon-40b-instruct
tiiuae/falcon-7b
tiiuae/falcon-7b-instruct
TinyLlama/TinyLlama-1.1B-Chat-v1.0
TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T
togethercomputer/LLaMA-2-7B-32K
togethercomputer/RedPajama-INCITE-7B-Base
togethercomputer/RedPajama-INCITE-7B-Chat
togethercomputer/RedPajama-INCITE-7B-Instruct
togethercomputer/RedPajama-INCITE-Base-3B-v1
togethercomputer/RedPajama-INCITE-Base-7B-v0.1
togethercomputer/RedPajama-INCITE-Chat-3B-v1
togethercomputer/RedPajama-INCITE-Chat-7B-v0.1
togethercomputer/RedPajama-INCITE-Instruct-3B-v1
togethercomputer/RedPajama-INCITE-Instruct-7B-v0.1
Trelis/Llama-2-7b-chat-hf-function-calling-v2
unsloth/Mistral-7B-v0.2

 

Tip

To sort the list above by model name after the /, use litgpt download | sort -f -t'/' -k2.

 

Note

If you want to adopt a model variant that is not listed in the table above but has a similar architecture as one of the supported models, you can use this model by by using the --model_name argument as shown below:

litgpt download \
 --repo_id NousResearch/Hermes-2-Pro-Mistral-7B \
 --model_name Mistral-7B-v0.1

 

2. Download Model Weights

To download the weights for a specific model, use the --repo_id argument. Replace <repo_id> with the model's repository ID. For example:

litgpt download --repo_id <repo_id>

This command downloads the model checkpoint into the checkpoints/ directory.

 

3. Additional Help

For more options, add the --help flag when running the script:

litgpt download --help

 

4. Run the Model

After conversion, run the model with the --checkpoint_dir flag, adjusting repo_id accordingly:

litgpt chat --checkpoint_dir checkpoints/<repo_id>

 

Tinyllama Example

This section shows a typical end-to-end example for downloading and using TinyLlama:

  1. List available TinyLlama checkpoints:
litgpt download | grep Tiny
TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T
TinyLlama/TinyLlama-1.1B-Chat-v1.0
  1. Download a TinyLlama checkpoint:
export repo_id=TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T
litgpt download --repo_id $repo_id
  1. Use the TinyLlama model:
litgpt chat --checkpoint_dir checkpoints/$repo_id

 

Specific models and access tokens

Note that certain models require that you've been granted access to the weights on the Hugging Face Hub.

For example, to get access to the Gemma 2B model, you can do so by following the steps at https://huggingface.co/google/gemma-2b. After access is granted, you can find your HF hub token in https://huggingface.co/settings/tokens.

Once you've been granted access and obtained the access token you need to pass the additional --access_token:

litgpt download \
  --repo_id google/gemma-2b \
  --access_token your_hf_token

 

Finetunes and Other Model Variants

Sometimes you want to download the weights of a finetune of one of the models listed above. To do this, you need to manually specify the model_name associated to the config to use. For example:

litgpt download \
  --repo_id NousResearch/Hermes-2-Pro-Mistral-7B \
  --model_name Mistral-7B-v0.1

 

Tips for GPU Memory Limitations

The download.py script will automatically convert the downloaded model checkpoint into a LitGPT-compatible format. In case this conversion fails due to GPU memory constraints, you can try to reduce the memory requirements by passing the --dtype bf16-true flag to convert all parameters into this smaller precision (however, note that most model weights are already in a bfloat16 format, so it may not have any effect):

litgpt download \
  --repo_id <repo_id>
  --dtype bf16-true

(If your GPU does not support the bfloat16 format, you can also try a regular 16-bit float format via --dtype 16-true.)

 

Converting Checkpoints Manually

For development purposes, for example, when adding or experimenting with new model configurations, it may be beneficial to split the weight download and model conversion into two separate steps.

You can do this by passing the --convert_checkpoint false option to the download script:

litgpt download \
  --repo_id <repo_id> \
  --convert_checkpoint false

and then calling the convert_hf_checkpoint.py script:

litgpt convert to_litgpt \
  --checkpoint_dir checkpoint_dir/<repo_id>

 

Downloading Tokenizers Only

In some cases we don't need the model weight, for example, when we are pretraining a model from scratch instead of finetuning it. For cases like this, you can use the --tokenizer_only flag to only download a model's tokenizer, which can then be used in the pretraining scripts:

litgpt download \
  --repo_id TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T \
  --tokenizer_only true

and

litgpt pretrain \
  --data ... \
  --model_name tiny-llama-1.1b \
  --tokenizer_dir checkpoints/TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T/