mono repo structure

2025-09-24 21:10:22 +00:00 · 2023-05-01 15:45:23 -04:00
parent b00d338c1e
commit 02d1bdb0be
53 changed files with 20 additions and 254 deletions
--- a/gpt4all-training/configs/inference/gptj.yaml
+++ b/gpt4all-training/configs/inference/gptj.yaml
@@ -0,0 +1,14 @@
+# 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
+