mpasila's picture
Upload 6 files
7475ed1 verified
raw
history blame contribute delete
492 Bytes
{
"base_model_name": "LumiOpen_Viking-7B",
"base_model_class": "LlamaForCausalLM",
"base_loaded_in_4bit": true,
"base_loaded_in_8bit": false,
"projections": "q, v",
"loss": 1.6425,
"grad_norm": 0.27980342507362366,
"learning_rate": 2.9999999999999997e-06,
"epoch": 1.0,
"current_steps": 12927,
"train_runtime": 25093.9843,
"train_samples_per_second": 2.063,
"train_steps_per_second": 0.016,
"total_flos": 1.1247635250680955e+18,
"train_loss": 1.712549373655036
}