gpt4all/gpt4all-training/configs/inference/gptj.yaml

# model/tokenizer
model_name: "nomic-ai/gpt4all-warmup-lr-epoch_1"
tokenizer_name: "EleutherAI/gpt-j-6B"

# dataset
streaming: false
num_proc: 64
dataset_path: "nomic-ai/turbo-500k-multi" 
max_length: 1024
batch_size: 32 

# logging
seed: 42
feat: inference for embedding plots 2023-04-06 21:40:39 -04:00			`# model/tokenizer`
feat: evals on new gptj models 2023-04-09 22:14:20 -04:00			`model_name: "nomic-ai/gpt4all-warmup-lr-epoch_1"`
feat: inference for embedding plots 2023-04-06 21:40:39 -04:00			`tokenizer_name: "EleutherAI/gpt-j-6B"`

			`# dataset`
			`streaming: false`
			`num_proc: 64`
feat: evals on new gptj models 2023-04-09 22:14:20 -04:00			`dataset_path: "nomic-ai/turbo-500k-multi"`
feat: inference for embedding plots 2023-04-06 21:40:39 -04:00			`max_length: 1024`
			`batch_size: 32`

			`# logging`
			`seed: 42`