Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +88 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,88 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.0032,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0,
|
| 13 |
+
"logps_train/chosen": -73.01441955566406,
|
| 14 |
+
"logps_train/ref_chosen": -73.0,
|
| 15 |
+
"logps_train/ref_rejected": -72.0,
|
| 16 |
+
"logps_train/rejected": -71.96434020996094,
|
| 17 |
+
"rewards_train/accuracies": 0.375,
|
| 18 |
+
"rewards_train/chosen": -0.007691763807088137,
|
| 19 |
+
"rewards_train/margins": 0.004367972258478403,
|
| 20 |
+
"rewards_train/rejected": -0.01205973606556654,
|
| 21 |
+
"step": 0
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.0,
|
| 25 |
+
"logps_train/chosen": -79.28456115722656,
|
| 26 |
+
"logps_train/ref_chosen": -79.5,
|
| 27 |
+
"logps_train/ref_rejected": -78.5,
|
| 28 |
+
"logps_train/rejected": -78.5880126953125,
|
| 29 |
+
"rewards_train/accuracies": 0.625,
|
| 30 |
+
"rewards_train/chosen": -0.00033102091401815414,
|
| 31 |
+
"rewards_train/margins": 0.017845441587269306,
|
| 32 |
+
"rewards_train/rejected": -0.01817646250128746,
|
| 33 |
+
"step": 1
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.0,
|
| 37 |
+
"learning_rate": 0.0,
|
| 38 |
+
"loss": 0.6846,
|
| 39 |
+
"step": 2
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.0,
|
| 43 |
+
"logps_train/chosen": -81.11515808105469,
|
| 44 |
+
"logps_train/ref_chosen": -81.0,
|
| 45 |
+
"logps_train/ref_rejected": -83.0,
|
| 46 |
+
"logps_train/rejected": -83.31285095214844,
|
| 47 |
+
"rewards_train/accuracies": 0.5,
|
| 48 |
+
"rewards_train/chosen": -0.008390522561967373,
|
| 49 |
+
"rewards_train/margins": -0.0021054279059171677,
|
| 50 |
+
"rewards_train/rejected": -0.006285094656050205,
|
| 51 |
+
"step": 2
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.0,
|
| 55 |
+
"logps_train/chosen": -73.90677642822266,
|
| 56 |
+
"logps_train/ref_chosen": -74.0,
|
| 57 |
+
"logps_train/ref_rejected": -76.0,
|
| 58 |
+
"logps_train/rejected": -76.30572509765625,
|
| 59 |
+
"rewards_train/accuracies": 0.5,
|
| 60 |
+
"rewards_train/chosen": -0.012552643194794655,
|
| 61 |
+
"rewards_train/margins": -0.0038555627688765526,
|
| 62 |
+
"rewards_train/rejected": -0.008697080425918102,
|
| 63 |
+
"step": 3
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.0,
|
| 67 |
+
"learning_rate": 0.0,
|
| 68 |
+
"loss": 0.6952,
|
| 69 |
+
"step": 4
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.0,
|
| 73 |
+
"step": 4,
|
| 74 |
+
"total_flos": 0.0,
|
| 75 |
+
"train_loss": 0.17379045486450195,
|
| 76 |
+
"train_runtime": 88.9643,
|
| 77 |
+
"train_samples_per_second": 0.09,
|
| 78 |
+
"train_steps_per_second": 0.011
|
| 79 |
+
}
|
| 80 |
+
],
|
| 81 |
+
"logging_steps": 2,
|
| 82 |
+
"max_steps": 1,
|
| 83 |
+
"num_train_epochs": 1,
|
| 84 |
+
"save_steps": 1.0,
|
| 85 |
+
"total_flos": 0.0,
|
| 86 |
+
"trial_name": null,
|
| 87 |
+
"trial_params": null
|
| 88 |
+
}
|