PIPPA_128_StableLM / training_log.json
jeiku's picture
Upload 5 files
4260f75 verified
raw
history blame contribute delete
No virus
469 Bytes
{
"base_model_name": "TheBloke_rocket-3B-GPTQ",
"base_model_class": "StableLMEpochForCausalLM",
"base_loaded_in_4bit": false,
"base_loaded_in_8bit": false,
"projections": "q, v",
"loss": 1.6475,
"learning_rate": 6.7779087612229726e-09,
"epoch": 2.97,
"current_steps": 1771,
"train_runtime": 25834.4978,
"train_samples_per_second": 4.44,
"train_steps_per_second": 0.009,
"total_flos": 5.968590834499584e+16,
"train_loss": 1.8597017387012105
}