Upload folder using huggingface_hub
Browse files
slots/15/checkpoint-2597/config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"TwinyForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"dtype": "float32",
|
| 7 |
+
"hidden_dropout": 0.0,
|
| 8 |
+
"hidden_size": 768,
|
| 9 |
+
"initializer_range": 0.02,
|
| 10 |
+
"intermediate_size": 3072,
|
| 11 |
+
"max_position_embeddings": 128,
|
| 12 |
+
"model_type": "twiny",
|
| 13 |
+
"neftune_alpha": 0.0,
|
| 14 |
+
"num_attention_heads": 12,
|
| 15 |
+
"num_hidden_layers": 1,
|
| 16 |
+
"num_key_value_heads": 3,
|
| 17 |
+
"qk_norm": true,
|
| 18 |
+
"rezero_init": 1.0,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_theta": 10000.0,
|
| 21 |
+
"transformers_version": "5.0.0",
|
| 22 |
+
"use_cache": false,
|
| 23 |
+
"vocab_size": 32000
|
| 24 |
+
}
|
slots/15/checkpoint-2597/trainer_state.json
ADDED
|
@@ -0,0 +1,944 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.6524545031064028,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2597,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 2.617016627215141e-05,
|
| 14 |
+
"grad_norm": 122.33340454101562,
|
| 15 |
+
"learning_rate": 5e-05,
|
| 16 |
+
"loss": 419.9996032714844,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0005234033254430282,
|
| 21 |
+
"grad_norm": 90.82410430908203,
|
| 22 |
+
"learning_rate": 4.999999511980259e-05,
|
| 23 |
+
"loss": 428.18197471217104,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0010468066508860564,
|
| 28 |
+
"grad_norm": 59.52641296386719,
|
| 29 |
+
"learning_rate": 4.999997943828395e-05,
|
| 30 |
+
"loss": 366.3947265625,
|
| 31 |
+
"step": 40
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0015702099763290846,
|
| 35 |
+
"grad_norm": 41.61277389526367,
|
| 36 |
+
"learning_rate": 4.999995294193233e-05,
|
| 37 |
+
"loss": 325.6545166015625,
|
| 38 |
+
"step": 60
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.0020936133017721128,
|
| 42 |
+
"grad_norm": 46.67556381225586,
|
| 43 |
+
"learning_rate": 4.999991563075919e-05,
|
| 44 |
+
"loss": 310.3648193359375,
|
| 45 |
+
"step": 80
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.002617016627215141,
|
| 49 |
+
"grad_norm": 47.65824890136719,
|
| 50 |
+
"learning_rate": 4.999986750478065e-05,
|
| 51 |
+
"loss": 303.49443359375,
|
| 52 |
+
"step": 100
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.003140419952658169,
|
| 56 |
+
"grad_norm": 62.748138427734375,
|
| 57 |
+
"learning_rate": 4.999980856401755e-05,
|
| 58 |
+
"loss": 322.389990234375,
|
| 59 |
+
"step": 120
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.007327646556202395,
|
| 63 |
+
"grad_norm": 48.20169448852539,
|
| 64 |
+
"learning_rate": 4.999895523943923e-05,
|
| 65 |
+
"loss": 288.82497336647725,
|
| 66 |
+
"step": 140
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.008374453207088451,
|
| 70 |
+
"grad_norm": 68.66603088378906,
|
| 71 |
+
"learning_rate": 4.99986329623214e-05,
|
| 72 |
+
"loss": 293.87509765625,
|
| 73 |
+
"step": 160
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.009421259857974507,
|
| 77 |
+
"grad_norm": 61.69401931762695,
|
| 78 |
+
"learning_rate": 4.999826742820622e-05,
|
| 79 |
+
"loss": 324.10361328125,
|
| 80 |
+
"step": 180
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.010468066508860565,
|
| 84 |
+
"grad_norm": 56.778385162353516,
|
| 85 |
+
"learning_rate": 4.999785863772618e-05,
|
| 86 |
+
"loss": 274.875537109375,
|
| 87 |
+
"step": 200
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.01151487315974662,
|
| 91 |
+
"grad_norm": 71.3816146850586,
|
| 92 |
+
"learning_rate": 4.999740659158865e-05,
|
| 93 |
+
"loss": 281.357080078125,
|
| 94 |
+
"step": 220
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.012561679810632677,
|
| 98 |
+
"grad_norm": 75.08697509765625,
|
| 99 |
+
"learning_rate": 4.9996911290575844e-05,
|
| 100 |
+
"loss": 263.852783203125,
|
| 101 |
+
"step": 240
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.013608486461518732,
|
| 105 |
+
"grad_norm": 53.91341781616211,
|
| 106 |
+
"learning_rate": 4.999637273554481e-05,
|
| 107 |
+
"loss": 258.546630859375,
|
| 108 |
+
"step": 260
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.01465529311240479,
|
| 112 |
+
"grad_norm": 56.39162063598633,
|
| 113 |
+
"learning_rate": 4.9995790927427446e-05,
|
| 114 |
+
"loss": 270.5470703125,
|
| 115 |
+
"step": 280
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.015702099763290844,
|
| 119 |
+
"grad_norm": 62.93777084350586,
|
| 120 |
+
"learning_rate": 4.99951658672305e-05,
|
| 121 |
+
"loss": 252.8383056640625,
|
| 122 |
+
"step": 300
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.016748906414176902,
|
| 126 |
+
"grad_norm": 61.51578903198242,
|
| 127 |
+
"learning_rate": 4.9994497556035567e-05,
|
| 128 |
+
"loss": 287.8167724609375,
|
| 129 |
+
"step": 320
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.01779571306506296,
|
| 133 |
+
"grad_norm": 62.75235366821289,
|
| 134 |
+
"learning_rate": 4.9993785994999074e-05,
|
| 135 |
+
"loss": 258.673779296875,
|
| 136 |
+
"step": 340
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.018842519715949014,
|
| 140 |
+
"grad_norm": 56.2258186340332,
|
| 141 |
+
"learning_rate": 4.999303118535229e-05,
|
| 142 |
+
"loss": 252.9609375,
|
| 143 |
+
"step": 360
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.01988932636683507,
|
| 147 |
+
"grad_norm": 54.998130798339844,
|
| 148 |
+
"learning_rate": 4.9992233128401314e-05,
|
| 149 |
+
"loss": 241.9411376953125,
|
| 150 |
+
"step": 380
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.02093613301772113,
|
| 154 |
+
"grad_norm": 63.37522888183594,
|
| 155 |
+
"learning_rate": 4.999139182552709e-05,
|
| 156 |
+
"loss": 249.7280517578125,
|
| 157 |
+
"step": 400
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.021982939668607183,
|
| 161 |
+
"grad_norm": 60.007144927978516,
|
| 162 |
+
"learning_rate": 4.9990507278185374e-05,
|
| 163 |
+
"loss": 236.941015625,
|
| 164 |
+
"step": 420
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.02302974631949324,
|
| 168 |
+
"grad_norm": 66.06304168701172,
|
| 169 |
+
"learning_rate": 4.998957948790677e-05,
|
| 170 |
+
"loss": 240.55878363715277,
|
| 171 |
+
"step": 440
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.0240765529703793,
|
| 175 |
+
"grad_norm": 68.9736099243164,
|
| 176 |
+
"learning_rate": 4.998860845629671e-05,
|
| 177 |
+
"loss": 247.988720703125,
|
| 178 |
+
"step": 460
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.025123359621265353,
|
| 182 |
+
"grad_norm": 61.50578308105469,
|
| 183 |
+
"learning_rate": 4.998759418503544e-05,
|
| 184 |
+
"loss": 233.025537109375,
|
| 185 |
+
"step": 480
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.02617016627215141,
|
| 189 |
+
"grad_norm": 54.96660614013672,
|
| 190 |
+
"learning_rate": 4.998653667587801e-05,
|
| 191 |
+
"loss": 243.004736328125,
|
| 192 |
+
"step": 500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.06804234327362181,
|
| 196 |
+
"grad_norm": 50.733734130859375,
|
| 197 |
+
"learning_rate": 4.990903999473118e-05,
|
| 198 |
+
"loss": 250.31749378551137,
|
| 199 |
+
"step": 520
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.07065935647645342,
|
| 203 |
+
"grad_norm": 48.06721496582031,
|
| 204 |
+
"learning_rate": 4.9901899191179905e-05,
|
| 205 |
+
"loss": 243.69541015625,
|
| 206 |
+
"step": 540
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.07327636967928504,
|
| 210 |
+
"grad_norm": 54.863311767578125,
|
| 211 |
+
"learning_rate": 4.989448913413268e-05,
|
| 212 |
+
"loss": 225.0319580078125,
|
| 213 |
+
"step": 560
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.07589338288211664,
|
| 217 |
+
"grad_norm": 60.36458969116211,
|
| 218 |
+
"learning_rate": 4.988680990371125e-05,
|
| 219 |
+
"loss": 222.7537353515625,
|
| 220 |
+
"step": 580
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.07851039608494825,
|
| 224 |
+
"grad_norm": 57.877986907958984,
|
| 225 |
+
"learning_rate": 4.987886158294783e-05,
|
| 226 |
+
"loss": 241.103515625,
|
| 227 |
+
"step": 600
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.08112740928777985,
|
| 231 |
+
"grad_norm": 64.6783676147461,
|
| 232 |
+
"learning_rate": 4.987064425778417e-05,
|
| 233 |
+
"loss": 215.971142578125,
|
| 234 |
+
"step": 620
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.08374442249061147,
|
| 238 |
+
"grad_norm": 50.828765869140625,
|
| 239 |
+
"learning_rate": 4.986215801707066e-05,
|
| 240 |
+
"loss": 213.7265380859375,
|
| 241 |
+
"step": 640
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.08636143569344307,
|
| 245 |
+
"grad_norm": 55.68048858642578,
|
| 246 |
+
"learning_rate": 4.9853402952565383e-05,
|
| 247 |
+
"loss": 216.94306640625,
|
| 248 |
+
"step": 660
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.08897844889627468,
|
| 252 |
+
"grad_norm": 54.125267028808594,
|
| 253 |
+
"learning_rate": 4.984437915893305e-05,
|
| 254 |
+
"loss": 204.139990234375,
|
| 255 |
+
"step": 680
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.0915954620991063,
|
| 259 |
+
"grad_norm": 57.27487564086914,
|
| 260 |
+
"learning_rate": 4.9835086733744074e-05,
|
| 261 |
+
"loss": 202.9946533203125,
|
| 262 |
+
"step": 700
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.0942124753019379,
|
| 266 |
+
"grad_norm": 65.67598724365234,
|
| 267 |
+
"learning_rate": 4.9825525777473424e-05,
|
| 268 |
+
"loss": 207.21923828125,
|
| 269 |
+
"step": 720
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.09682948850476951,
|
| 273 |
+
"grad_norm": 71.1091537475586,
|
| 274 |
+
"learning_rate": 4.98156963934996e-05,
|
| 275 |
+
"loss": 204.4235107421875,
|
| 276 |
+
"step": 740
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.09944650170760111,
|
| 280 |
+
"grad_norm": 60.16761779785156,
|
| 281 |
+
"learning_rate": 4.9805598688103496e-05,
|
| 282 |
+
"loss": 213.841748046875,
|
| 283 |
+
"step": 760
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.19596246146437973,
|
| 287 |
+
"grad_norm": 373.96917724609375,
|
| 288 |
+
"learning_rate": 4.99999986487106e-05,
|
| 289 |
+
"loss": 470.0482177734375,
|
| 290 |
+
"step": 780
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.2009871399634664,
|
| 294 |
+
"grad_norm": 97.38092803955078,
|
| 295 |
+
"learning_rate": 4.9999459486180746e-05,
|
| 296 |
+
"loss": 494.169580078125,
|
| 297 |
+
"step": 800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.20601181846255306,
|
| 301 |
+
"grad_norm": Infinity,
|
| 302 |
+
"learning_rate": 4.9997837968095394e-05,
|
| 303 |
+
"loss": 435.49521484375,
|
| 304 |
+
"step": 820
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.21103649696163973,
|
| 308 |
+
"grad_norm": 102.23469543457031,
|
| 309 |
+
"learning_rate": 4.999529630904343e-05,
|
| 310 |
+
"loss": 447.469873046875,
|
| 311 |
+
"step": 840
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.21606117546072637,
|
| 315 |
+
"grad_norm": 104.21376037597656,
|
| 316 |
+
"learning_rate": 4.999156707688316e-05,
|
| 317 |
+
"loss": 439.9337890625,
|
| 318 |
+
"step": 860
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.22108585395981303,
|
| 322 |
+
"grad_norm": 100.12609100341797,
|
| 323 |
+
"learning_rate": 4.9986757181733314e-05,
|
| 324 |
+
"loss": 471.3021484375,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.2261105324588997,
|
| 329 |
+
"grad_norm": 102.25228881835938,
|
| 330 |
+
"learning_rate": 4.998086683157904e-05,
|
| 331 |
+
"loss": 416.210498046875,
|
| 332 |
+
"step": 900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.23113521095798636,
|
| 336 |
+
"grad_norm": 131.55450439453125,
|
| 337 |
+
"learning_rate": 4.997389628112563e-05,
|
| 338 |
+
"loss": 426.757763671875,
|
| 339 |
+
"step": 920
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.23615988945707303,
|
| 343 |
+
"grad_norm": 118.51521301269531,
|
| 344 |
+
"learning_rate": 4.996584583178737e-05,
|
| 345 |
+
"loss": 430.067724609375,
|
| 346 |
+
"step": 940
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.24118456795615967,
|
| 350 |
+
"grad_norm": 97.93070220947266,
|
| 351 |
+
"learning_rate": 4.995671583167459e-05,
|
| 352 |
+
"loss": 452.283056640625,
|
| 353 |
+
"step": 960
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.24620924645524633,
|
| 357 |
+
"grad_norm": 87.021484375,
|
| 358 |
+
"learning_rate": 4.9946506675578595e-05,
|
| 359 |
+
"loss": 384.961181640625,
|
| 360 |
+
"step": 980
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.251233924954333,
|
| 364 |
+
"grad_norm": 104.47926330566406,
|
| 365 |
+
"learning_rate": 4.993521880495458e-05,
|
| 366 |
+
"loss": 435.11455078125,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.25625860345341966,
|
| 371 |
+
"grad_norm": 88.34175109863281,
|
| 372 |
+
"learning_rate": 4.992285270790253e-05,
|
| 373 |
+
"loss": 387.83525390625,
|
| 374 |
+
"step": 1020
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.26128328195250633,
|
| 378 |
+
"grad_norm": 97.24592590332031,
|
| 379 |
+
"learning_rate": 4.990940891914619e-05,
|
| 380 |
+
"loss": 377.01767578125,
|
| 381 |
+
"step": 1040
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.266307960451593,
|
| 385 |
+
"grad_norm": 93.01044464111328,
|
| 386 |
+
"learning_rate": 4.989488802000982e-05,
|
| 387 |
+
"loss": 392.9261962890625,
|
| 388 |
+
"step": 1060
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.27133263895067966,
|
| 392 |
+
"grad_norm": 87.1651611328125,
|
| 393 |
+
"learning_rate": 4.987929063839317e-05,
|
| 394 |
+
"loss": 378.5872802734375,
|
| 395 |
+
"step": 1080
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.2763573174497663,
|
| 399 |
+
"grad_norm": 106.6820068359375,
|
| 400 |
+
"learning_rate": 4.9862617448744245e-05,
|
| 401 |
+
"loss": 362.4338623046875,
|
| 402 |
+
"step": 1100
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.28138199594885294,
|
| 406 |
+
"grad_norm": 87.04276275634766,
|
| 407 |
+
"learning_rate": 4.984486917203022e-05,
|
| 408 |
+
"loss": 347.4080810546875,
|
| 409 |
+
"step": 1120
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.2864066744479396,
|
| 413 |
+
"grad_norm": 95.84716033935547,
|
| 414 |
+
"learning_rate": 4.982604657570619e-05,
|
| 415 |
+
"loss": 360.5807861328125,
|
| 416 |
+
"step": 1140
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.29143135294702627,
|
| 420 |
+
"grad_norm": 93.88927459716797,
|
| 421 |
+
"learning_rate": 4.980615047368202e-05,
|
| 422 |
+
"loss": 318.7136474609375,
|
| 423 |
+
"step": 1160
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.29645603144611293,
|
| 427 |
+
"grad_norm": 98.77674865722656,
|
| 428 |
+
"learning_rate": 4.9785181726287174e-05,
|
| 429 |
+
"loss": 347.99814453125,
|
| 430 |
+
"step": 1180
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.3014807099451996,
|
| 434 |
+
"grad_norm": 79.85491943359375,
|
| 435 |
+
"learning_rate": 4.976314124023346e-05,
|
| 436 |
+
"loss": 355.9615234375,
|
| 437 |
+
"step": 1200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.30650538844428626,
|
| 441 |
+
"grad_norm": 91.22853088378906,
|
| 442 |
+
"learning_rate": 4.9740029968575874e-05,
|
| 443 |
+
"loss": 335.508837890625,
|
| 444 |
+
"step": 1220
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.31153006694337293,
|
| 448 |
+
"grad_norm": 97.69166564941406,
|
| 449 |
+
"learning_rate": 4.971584891067134e-05,
|
| 450 |
+
"loss": 327.9396240234375,
|
| 451 |
+
"step": 1240
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.3165547454424596,
|
| 455 |
+
"grad_norm": 87.74190521240234,
|
| 456 |
+
"learning_rate": 4.969059911213554e-05,
|
| 457 |
+
"loss": 288.4083251953125,
|
| 458 |
+
"step": 1260
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.32157942394154626,
|
| 462 |
+
"grad_norm": 85.86646270751953,
|
| 463 |
+
"learning_rate": 4.966428166479768e-05,
|
| 464 |
+
"loss": 285.6180419921875,
|
| 465 |
+
"step": 1280
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.3266041024406329,
|
| 469 |
+
"grad_norm": 91.06046295166016,
|
| 470 |
+
"learning_rate": 4.963689770665327e-05,
|
| 471 |
+
"loss": 305.8553955078125,
|
| 472 |
+
"step": 1300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.33162878093971954,
|
| 476 |
+
"grad_norm": 81.2901611328125,
|
| 477 |
+
"learning_rate": 4.960844842181494e-05,
|
| 478 |
+
"loss": 263.1835205078125,
|
| 479 |
+
"step": 1320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.3366534594388062,
|
| 483 |
+
"grad_norm": 85.06604766845703,
|
| 484 |
+
"learning_rate": 4.957893504046123e-05,
|
| 485 |
+
"loss": 293.3139404296875,
|
| 486 |
+
"step": 1340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.34167813793789287,
|
| 490 |
+
"grad_norm": 88.67339324951172,
|
| 491 |
+
"learning_rate": 4.9548358838783354e-05,
|
| 492 |
+
"loss": 275.2658203125,
|
| 493 |
+
"step": 1360
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.34670281643697953,
|
| 497 |
+
"grad_norm": 92.95565795898438,
|
| 498 |
+
"learning_rate": 4.951672113893009e-05,
|
| 499 |
+
"loss": 272.006298828125,
|
| 500 |
+
"step": 1380
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.3517274949360662,
|
| 504 |
+
"grad_norm": 77.20344543457031,
|
| 505 |
+
"learning_rate": 4.948402330895056e-05,
|
| 506 |
+
"loss": 264.4460693359375,
|
| 507 |
+
"step": 1400
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.35675217343515286,
|
| 511 |
+
"grad_norm": 88.61670684814453,
|
| 512 |
+
"learning_rate": 4.9450266762735066e-05,
|
| 513 |
+
"loss": 264.2577392578125,
|
| 514 |
+
"step": 1420
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.36177685193423953,
|
| 518 |
+
"grad_norm": 78.82662200927734,
|
| 519 |
+
"learning_rate": 4.9415452959954e-05,
|
| 520 |
+
"loss": 247.95751953125,
|
| 521 |
+
"step": 1440
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.3668015304333262,
|
| 525 |
+
"grad_norm": 91.61518096923828,
|
| 526 |
+
"learning_rate": 4.937958340599467e-05,
|
| 527 |
+
"loss": 260.0285888671875,
|
| 528 |
+
"step": 1460
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.37182620893241286,
|
| 532 |
+
"grad_norm": 77.36851501464844,
|
| 533 |
+
"learning_rate": 4.934265965189626e-05,
|
| 534 |
+
"loss": 233.4987060546875,
|
| 535 |
+
"step": 1480
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.3768508874314995,
|
| 539 |
+
"grad_norm": 83.2024154663086,
|
| 540 |
+
"learning_rate": 4.930468329428271e-05,
|
| 541 |
+
"loss": 219.914892578125,
|
| 542 |
+
"step": 1500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.38187556593058614,
|
| 546 |
+
"grad_norm": 76.80755615234375,
|
| 547 |
+
"learning_rate": 4.9265655975293693e-05,
|
| 548 |
+
"loss": 257.715625,
|
| 549 |
+
"step": 1520
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.3869002444296728,
|
| 553 |
+
"grad_norm": 74.73920440673828,
|
| 554 |
+
"learning_rate": 4.9225579382513656e-05,
|
| 555 |
+
"loss": 230.7301513671875,
|
| 556 |
+
"step": 1540
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.39192492292875947,
|
| 560 |
+
"grad_norm": 80.25586700439453,
|
| 561 |
+
"learning_rate": 4.918445524889876e-05,
|
| 562 |
+
"loss": 218.6532470703125,
|
| 563 |
+
"step": 1560
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.39694960142784613,
|
| 567 |
+
"grad_norm": 75.49822235107422,
|
| 568 |
+
"learning_rate": 4.914228535270201e-05,
|
| 569 |
+
"loss": 216.0865966796875,
|
| 570 |
+
"step": 1580
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.4019742799269328,
|
| 574 |
+
"grad_norm": 78.31658172607422,
|
| 575 |
+
"learning_rate": 4.909907151739633e-05,
|
| 576 |
+
"loss": 230.8989990234375,
|
| 577 |
+
"step": 1600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.40699895842601946,
|
| 581 |
+
"grad_norm": 82.01393127441406,
|
| 582 |
+
"learning_rate": 4.905481561159575e-05,
|
| 583 |
+
"loss": 198.04837646484376,
|
| 584 |
+
"step": 1620
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.41202363692510613,
|
| 588 |
+
"grad_norm": 112.93590545654297,
|
| 589 |
+
"learning_rate": 4.900951954897456e-05,
|
| 590 |
+
"loss": 192.21912841796876,
|
| 591 |
+
"step": 1640
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.4170483154241928,
|
| 595 |
+
"grad_norm": 69.93424224853516,
|
| 596 |
+
"learning_rate": 4.8963185288184574e-05,
|
| 597 |
+
"loss": 221.135107421875,
|
| 598 |
+
"step": 1660
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.42207299392327946,
|
| 602 |
+
"grad_norm": 67.13924407958984,
|
| 603 |
+
"learning_rate": 4.891581483277046e-05,
|
| 604 |
+
"loss": 184.96981201171874,
|
| 605 |
+
"step": 1680
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.4270976724223661,
|
| 609 |
+
"grad_norm": 69.87117004394531,
|
| 610 |
+
"learning_rate": 4.886741023108307e-05,
|
| 611 |
+
"loss": 185.549267578125,
|
| 612 |
+
"step": 1700
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.43212235092145274,
|
| 616 |
+
"grad_norm": 76.50455474853516,
|
| 617 |
+
"learning_rate": 4.8817973576190915e-05,
|
| 618 |
+
"loss": 208.102685546875,
|
| 619 |
+
"step": 1720
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.4371470294205394,
|
| 623 |
+
"grad_norm": 88.16585540771484,
|
| 624 |
+
"learning_rate": 4.876750700578959e-05,
|
| 625 |
+
"loss": 178.43084716796875,
|
| 626 |
+
"step": 1740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.44217170791962607,
|
| 630 |
+
"grad_norm": 66.18992614746094,
|
| 631 |
+
"learning_rate": 4.871601270210939e-05,
|
| 632 |
+
"loss": 190.35350341796874,
|
| 633 |
+
"step": 1760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.44719638641871273,
|
| 637 |
+
"grad_norm": 69.53152465820312,
|
| 638 |
+
"learning_rate": 4.866349289182095e-05,
|
| 639 |
+
"loss": 190.474609375,
|
| 640 |
+
"step": 1780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.4522210649177994,
|
| 644 |
+
"grad_norm": 71.9003677368164,
|
| 645 |
+
"learning_rate": 4.860994984593892e-05,
|
| 646 |
+
"loss": 179.31239013671876,
|
| 647 |
+
"step": 1800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.45724574341688606,
|
| 651 |
+
"grad_norm": 64.94371032714844,
|
| 652 |
+
"learning_rate": 4.855538587972379e-05,
|
| 653 |
+
"loss": 188.06192626953126,
|
| 654 |
+
"step": 1820
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.46227042191597273,
|
| 658 |
+
"grad_norm": 65.83982849121094,
|
| 659 |
+
"learning_rate": 4.8499803352581795e-05,
|
| 660 |
+
"loss": 160.46326904296876,
|
| 661 |
+
"step": 1840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.4672951004150594,
|
| 665 |
+
"grad_norm": 65.22266387939453,
|
| 666 |
+
"learning_rate": 4.844320466796286e-05,
|
| 667 |
+
"loss": 175.50235595703126,
|
| 668 |
+
"step": 1860
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.47231977891414606,
|
| 672 |
+
"grad_norm": 72.8160171508789,
|
| 673 |
+
"learning_rate": 4.8385592273256675e-05,
|
| 674 |
+
"loss": 180.21500244140626,
|
| 675 |
+
"step": 1880
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.4773444574132327,
|
| 679 |
+
"grad_norm": 63.11011505126953,
|
| 680 |
+
"learning_rate": 4.832696865968686e-05,
|
| 681 |
+
"loss": 162.31240234375,
|
| 682 |
+
"step": 1900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.48236913591231934,
|
| 686 |
+
"grad_norm": 65.32222747802734,
|
| 687 |
+
"learning_rate": 4.826733636220332e-05,
|
| 688 |
+
"loss": 151.82484130859376,
|
| 689 |
+
"step": 1920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.487393814411406,
|
| 693 |
+
"grad_norm": 72.49311828613281,
|
| 694 |
+
"learning_rate": 4.820669795937248e-05,
|
| 695 |
+
"loss": 162.365576171875,
|
| 696 |
+
"step": 1940
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.49241849291049267,
|
| 700 |
+
"grad_norm": 73.10563659667969,
|
| 701 |
+
"learning_rate": 4.8145056073265945e-05,
|
| 702 |
+
"loss": 154.83590087890624,
|
| 703 |
+
"step": 1960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.49744317140957933,
|
| 707 |
+
"grad_norm": 65.19477081298828,
|
| 708 |
+
"learning_rate": 4.8082413369347003e-05,
|
| 709 |
+
"loss": 151.2623291015625,
|
| 710 |
+
"step": 1980
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.502467849908666,
|
| 714 |
+
"grad_norm": 72.08782958984375,
|
| 715 |
+
"learning_rate": 4.8018772556355436e-05,
|
| 716 |
+
"loss": 162.5435791015625,
|
| 717 |
+
"step": 2000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.5074925284077526,
|
| 721 |
+
"grad_norm": 64.37730407714844,
|
| 722 |
+
"learning_rate": 4.795413638619035e-05,
|
| 723 |
+
"loss": 151.4502197265625,
|
| 724 |
+
"step": 2020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.5125172069068393,
|
| 728 |
+
"grad_norm": 57.9962272644043,
|
| 729 |
+
"learning_rate": 4.788850765379119e-05,
|
| 730 |
+
"loss": 134.33564453125,
|
| 731 |
+
"step": 2040
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.5175418854059259,
|
| 735 |
+
"grad_norm": 67.57404327392578,
|
| 736 |
+
"learning_rate": 4.782188919701692e-05,
|
| 737 |
+
"loss": 134.29805908203124,
|
| 738 |
+
"step": 2060
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.5225665639050127,
|
| 742 |
+
"grad_norm": 89.81603240966797,
|
| 743 |
+
"learning_rate": 4.775428389652326e-05,
|
| 744 |
+
"loss": 154.1037841796875,
|
| 745 |
+
"step": 2080
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.5275912424040993,
|
| 749 |
+
"grad_norm": 79.56139373779297,
|
| 750 |
+
"learning_rate": 4.768569467563812e-05,
|
| 751 |
+
"loss": 141.463916015625,
|
| 752 |
+
"step": 2100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.532615920903186,
|
| 756 |
+
"grad_norm": 62.466373443603516,
|
| 757 |
+
"learning_rate": 4.761612450023526e-05,
|
| 758 |
+
"loss": 134.26026611328126,
|
| 759 |
+
"step": 2120
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.5376405994022726,
|
| 763 |
+
"grad_norm": 83.5448226928711,
|
| 764 |
+
"learning_rate": 4.754557637860597e-05,
|
| 765 |
+
"loss": 139.75875244140624,
|
| 766 |
+
"step": 2140
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.5426652779013593,
|
| 770 |
+
"grad_norm": 61.366981506347656,
|
| 771 |
+
"learning_rate": 4.747405336132903e-05,
|
| 772 |
+
"loss": 134.8552001953125,
|
| 773 |
+
"step": 2160
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.5476899564004459,
|
| 777 |
+
"grad_norm": 53.69393539428711,
|
| 778 |
+
"learning_rate": 4.740155854113877e-05,
|
| 779 |
+
"loss": 131.57630615234376,
|
| 780 |
+
"step": 2180
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.5527146348995327,
|
| 784 |
+
"grad_norm": 64.62442016601562,
|
| 785 |
+
"learning_rate": 4.732809505279136e-05,
|
| 786 |
+
"loss": 129.31090087890624,
|
| 787 |
+
"step": 2200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.5577393133986193,
|
| 791 |
+
"grad_norm": 57.51287841796875,
|
| 792 |
+
"learning_rate": 4.725366607292926e-05,
|
| 793 |
+
"loss": 134.6401611328125,
|
| 794 |
+
"step": 2220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.5627639918977059,
|
| 798 |
+
"grad_norm": 78.75836181640625,
|
| 799 |
+
"learning_rate": 4.717827481994384e-05,
|
| 800 |
+
"loss": 123.9548583984375,
|
| 801 |
+
"step": 2240
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.5677886703967926,
|
| 805 |
+
"grad_norm": 52.389793395996094,
|
| 806 |
+
"learning_rate": 4.710192455383623e-05,
|
| 807 |
+
"loss": 131.11953125,
|
| 808 |
+
"step": 2260
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.5728133488958792,
|
| 812 |
+
"grad_norm": 60.91038131713867,
|
| 813 |
+
"learning_rate": 4.702461857607634e-05,
|
| 814 |
+
"loss": 139.98529052734375,
|
| 815 |
+
"step": 2280
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.5778380273949659,
|
| 819 |
+
"grad_norm": 61.93384552001953,
|
| 820 |
+
"learning_rate": 4.694636022946012e-05,
|
| 821 |
+
"loss": 116.9314453125,
|
| 822 |
+
"step": 2300
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.5828627058940525,
|
| 826 |
+
"grad_norm": 72.56891632080078,
|
| 827 |
+
"learning_rate": 4.6867152897964984e-05,
|
| 828 |
+
"loss": 113.78642578125,
|
| 829 |
+
"step": 2320
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.5878873843931393,
|
| 833 |
+
"grad_norm": 65.45015716552734,
|
| 834 |
+
"learning_rate": 4.678700000660352e-05,
|
| 835 |
+
"loss": 122.62408447265625,
|
| 836 |
+
"step": 2340
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.5929120628922259,
|
| 840 |
+
"grad_norm": 47.3505859375,
|
| 841 |
+
"learning_rate": 4.670590502127537e-05,
|
| 842 |
+
"loss": 109.8542724609375,
|
| 843 |
+
"step": 2360
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.5979367413913126,
|
| 847 |
+
"grad_norm": 50.940364837646484,
|
| 848 |
+
"learning_rate": 4.662387144861734e-05,
|
| 849 |
+
"loss": 105.6350830078125,
|
| 850 |
+
"step": 2380
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.6029614198903992,
|
| 854 |
+
"grad_norm": 63.25798797607422,
|
| 855 |
+
"learning_rate": 4.654090283585181e-05,
|
| 856 |
+
"loss": 105.77552490234375,
|
| 857 |
+
"step": 2400
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.6079860983894858,
|
| 861 |
+
"grad_norm": 51.82925033569336,
|
| 862 |
+
"learning_rate": 4.645700277063333e-05,
|
| 863 |
+
"loss": 104.80977783203124,
|
| 864 |
+
"step": 2420
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.6130107768885725,
|
| 868 |
+
"grad_norm": 68.77577209472656,
|
| 869 |
+
"learning_rate": 4.637217488089345e-05,
|
| 870 |
+
"loss": 101.56100463867188,
|
| 871 |
+
"step": 2440
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.6180354553876591,
|
| 875 |
+
"grad_norm": 65.48799133300781,
|
| 876 |
+
"learning_rate": 4.628642283468393e-05,
|
| 877 |
+
"loss": 105.1793212890625,
|
| 878 |
+
"step": 2460
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.6230601338867459,
|
| 882 |
+
"grad_norm": 58.55667495727539,
|
| 883 |
+
"learning_rate": 4.6199750340018033e-05,
|
| 884 |
+
"loss": 102.6951171875,
|
| 885 |
+
"step": 2480
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.6280848123858325,
|
| 889 |
+
"grad_norm": 63.48239517211914,
|
| 890 |
+
"learning_rate": 4.6112161144710255e-05,
|
| 891 |
+
"loss": 101.88885498046875,
|
| 892 |
+
"step": 2500
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.6331094908849192,
|
| 896 |
+
"grad_norm": 61.78999710083008,
|
| 897 |
+
"learning_rate": 4.602365903621423e-05,
|
| 898 |
+
"loss": 98.3158203125,
|
| 899 |
+
"step": 2520
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.6381341693840058,
|
| 903 |
+
"grad_norm": 59.46126174926758,
|
| 904 |
+
"learning_rate": 4.5934247841458976e-05,
|
| 905 |
+
"loss": 92.30280151367188,
|
| 906 |
+
"step": 2540
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.6431588478830925,
|
| 910 |
+
"grad_norm": 50.481239318847656,
|
| 911 |
+
"learning_rate": 4.584393142668341e-05,
|
| 912 |
+
"loss": 109.47457275390624,
|
| 913 |
+
"step": 2560
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.6481835263821791,
|
| 917 |
+
"grad_norm": 64.15777587890625,
|
| 918 |
+
"learning_rate": 4.5752713697269135e-05,
|
| 919 |
+
"loss": 93.06622314453125,
|
| 920 |
+
"step": 2580
|
| 921 |
+
}
|
| 922 |
+
],
|
| 923 |
+
"logging_steps": 20,
|
| 924 |
+
"max_steps": 9555,
|
| 925 |
+
"num_input_tokens_seen": 0,
|
| 926 |
+
"num_train_epochs": 3,
|
| 927 |
+
"save_steps": 1000000000,
|
| 928 |
+
"stateful_callbacks": {
|
| 929 |
+
"TrainerControl": {
|
| 930 |
+
"args": {
|
| 931 |
+
"should_epoch_stop": false,
|
| 932 |
+
"should_evaluate": false,
|
| 933 |
+
"should_log": false,
|
| 934 |
+
"should_save": true,
|
| 935 |
+
"should_training_stop": false
|
| 936 |
+
},
|
| 937 |
+
"attributes": {}
|
| 938 |
+
}
|
| 939 |
+
},
|
| 940 |
+
"total_flos": 5331328824545280.0,
|
| 941 |
+
"train_batch_size": 1,
|
| 942 |
+
"trial_name": null,
|
| 943 |
+
"trial_params": null
|
| 944 |
+
}
|