Upload folder using huggingface_hub
Browse files
slots/5/checkpoint-2369/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/5/checkpoint-2369/trainer_state.json
ADDED
|
@@ -0,0 +1,867 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.5951716106407757,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2369,
|
| 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": 0.19596194862803082,
|
| 287 |
+
"grad_norm": 334.94952392578125,
|
| 288 |
+
"learning_rate": 4.99999986487106e-05,
|
| 289 |
+
"loss": 444.81829833984375,
|
| 290 |
+
"step": 780
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.2009866139774675,
|
| 294 |
+
"grad_norm": 117.09273529052734,
|
| 295 |
+
"learning_rate": 4.9999459486180746e-05,
|
| 296 |
+
"loss": 458.396142578125,
|
| 297 |
+
"step": 800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.20601127932690422,
|
| 301 |
+
"grad_norm": 108.73420715332031,
|
| 302 |
+
"learning_rate": 4.9997837968095394e-05,
|
| 303 |
+
"loss": 500.8515625,
|
| 304 |
+
"step": 820
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.2110359446763409,
|
| 308 |
+
"grad_norm": 111.36239624023438,
|
| 309 |
+
"learning_rate": 4.999513551586019e-05,
|
| 310 |
+
"loss": 420.47216796875,
|
| 311 |
+
"step": 840
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.2160606100257776,
|
| 315 |
+
"grad_norm": 112.1257553100586,
|
| 316 |
+
"learning_rate": 4.999135224633215e-05,
|
| 317 |
+
"loss": 408.894921875,
|
| 318 |
+
"step": 860
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.22108527537521427,
|
| 322 |
+
"grad_norm": 96.66751098632812,
|
| 323 |
+
"learning_rate": 4.998648832310402e-05,
|
| 324 |
+
"loss": 463.8513671875,
|
| 325 |
+
"step": 880
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.22610994072465096,
|
| 329 |
+
"grad_norm": 113.46974182128906,
|
| 330 |
+
"learning_rate": 4.9980543956497236e-05,
|
| 331 |
+
"loss": 387.7029052734375,
|
| 332 |
+
"step": 900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.23113460607408765,
|
| 336 |
+
"grad_norm": 109.18795013427734,
|
| 337 |
+
"learning_rate": 4.997351940355277e-05,
|
| 338 |
+
"loss": 423.491064453125,
|
| 339 |
+
"step": 920
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.23615927142352433,
|
| 343 |
+
"grad_norm": 95.84347534179688,
|
| 344 |
+
"learning_rate": 4.996541496802006e-05,
|
| 345 |
+
"loss": 439.167529296875,
|
| 346 |
+
"step": 940
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.24118393677296102,
|
| 350 |
+
"grad_norm": 96.51708221435547,
|
| 351 |
+
"learning_rate": 4.995623100034386e-05,
|
| 352 |
+
"loss": 364.4853271484375,
|
| 353 |
+
"step": 960
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.2462086021223977,
|
| 357 |
+
"grad_norm": 101.59738159179688,
|
| 358 |
+
"learning_rate": 4.994596789764909e-05,
|
| 359 |
+
"loss": 380.7418701171875,
|
| 360 |
+
"step": 980
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.2512332674718344,
|
| 364 |
+
"grad_norm": 104.19242095947266,
|
| 365 |
+
"learning_rate": 4.993462610372364e-05,
|
| 366 |
+
"loss": 387.2640869140625,
|
| 367 |
+
"step": 1000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.2562579328212711,
|
| 371 |
+
"grad_norm": 95.20624542236328,
|
| 372 |
+
"learning_rate": 4.992220610899924e-05,
|
| 373 |
+
"loss": 347.985693359375,
|
| 374 |
+
"step": 1020
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.26128259817070776,
|
| 378 |
+
"grad_norm": 105.89356994628906,
|
| 379 |
+
"learning_rate": 4.990870845053017e-05,
|
| 380 |
+
"loss": 358.2223876953125,
|
| 381 |
+
"step": 1040
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.26630726352014444,
|
| 385 |
+
"grad_norm": 100.86421203613281,
|
| 386 |
+
"learning_rate": 4.989413371197013e-05,
|
| 387 |
+
"loss": 399.116796875,
|
| 388 |
+
"step": 1060
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.27133192886958113,
|
| 392 |
+
"grad_norm": 93.12042236328125,
|
| 393 |
+
"learning_rate": 4.987848252354691e-05,
|
| 394 |
+
"loss": 327.8339111328125,
|
| 395 |
+
"step": 1080
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.2763565942190178,
|
| 399 |
+
"grad_norm": 93.19499969482422,
|
| 400 |
+
"learning_rate": 4.986175556203521e-05,
|
| 401 |
+
"loss": 348.908642578125,
|
| 402 |
+
"step": 1100
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.2813812595684545,
|
| 406 |
+
"grad_norm": 94.92535400390625,
|
| 407 |
+
"learning_rate": 4.984395355072732e-05,
|
| 408 |
+
"loss": 315.4578857421875,
|
| 409 |
+
"step": 1120
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.2864059249178912,
|
| 413 |
+
"grad_norm": 86.75041198730469,
|
| 414 |
+
"learning_rate": 4.982507725940191e-05,
|
| 415 |
+
"loss": 308.7882080078125,
|
| 416 |
+
"step": 1140
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.29143059026732787,
|
| 420 |
+
"grad_norm": 96.52128601074219,
|
| 421 |
+
"learning_rate": 4.980512750429068e-05,
|
| 422 |
+
"loss": 321.39384765625,
|
| 423 |
+
"step": 1160
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.2964552556167646,
|
| 427 |
+
"grad_norm": 84.87345123291016,
|
| 428 |
+
"learning_rate": 4.97841051480431e-05,
|
| 429 |
+
"loss": 326.368408203125,
|
| 430 |
+
"step": 1180
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.3014799209662013,
|
| 434 |
+
"grad_norm": 80.8378677368164,
|
| 435 |
+
"learning_rate": 4.976201109968908e-05,
|
| 436 |
+
"loss": 289.3224365234375,
|
| 437 |
+
"step": 1200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.306504586315638,
|
| 441 |
+
"grad_norm": 97.03463745117188,
|
| 442 |
+
"learning_rate": 4.973884631459971e-05,
|
| 443 |
+
"loss": 286.0260009765625,
|
| 444 |
+
"step": 1220
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.31152925166507467,
|
| 448 |
+
"grad_norm": 76.3429183959961,
|
| 449 |
+
"learning_rate": 4.9714611794445906e-05,
|
| 450 |
+
"loss": 309.6694091796875,
|
| 451 |
+
"step": 1240
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.31655391701451135,
|
| 455 |
+
"grad_norm": 89.26808166503906,
|
| 456 |
+
"learning_rate": 4.968930858715511e-05,
|
| 457 |
+
"loss": 277.4451416015625,
|
| 458 |
+
"step": 1260
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.32157858236394804,
|
| 462 |
+
"grad_norm": 92.4432601928711,
|
| 463 |
+
"learning_rate": 4.966293778686598e-05,
|
| 464 |
+
"loss": 282.81572265625,
|
| 465 |
+
"step": 1280
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.3266032477133847,
|
| 469 |
+
"grad_norm": 85.95075225830078,
|
| 470 |
+
"learning_rate": 4.963550053388106e-05,
|
| 471 |
+
"loss": 305.7855224609375,
|
| 472 |
+
"step": 1300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.3316279130628214,
|
| 476 |
+
"grad_norm": 92.0710678100586,
|
| 477 |
+
"learning_rate": 4.9606998014617535e-05,
|
| 478 |
+
"loss": 263.8234130859375,
|
| 479 |
+
"step": 1320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.3366525784122581,
|
| 483 |
+
"grad_norm": 79.8038101196289,
|
| 484 |
+
"learning_rate": 4.957743146155581e-05,
|
| 485 |
+
"loss": 267.8830078125,
|
| 486 |
+
"step": 1340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.3416772437616948,
|
| 490 |
+
"grad_norm": 85.92285919189453,
|
| 491 |
+
"learning_rate": 4.954680215318635e-05,
|
| 492 |
+
"loss": 291.7197021484375,
|
| 493 |
+
"step": 1360
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.34670190911113147,
|
| 497 |
+
"grad_norm": 77.87874603271484,
|
| 498 |
+
"learning_rate": 4.951672113893009e-05,
|
| 499 |
+
"loss": 328.0381591796875,
|
| 500 |
+
"step": 1380
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.35172657446056815,
|
| 504 |
+
"grad_norm": 88.03858947753906,
|
| 505 |
+
"learning_rate": 4.948402330895056e-05,
|
| 506 |
+
"loss": 263.4621826171875,
|
| 507 |
+
"step": 1400
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.35675123981000484,
|
| 511 |
+
"grad_norm": 80.90335845947266,
|
| 512 |
+
"learning_rate": 4.9450266762735066e-05,
|
| 513 |
+
"loss": 265.0658203125,
|
| 514 |
+
"step": 1420
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.3617759051594415,
|
| 518 |
+
"grad_norm": 85.93920135498047,
|
| 519 |
+
"learning_rate": 4.9415452959954e-05,
|
| 520 |
+
"loss": 230.0629638671875,
|
| 521 |
+
"step": 1440
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.3668005705088782,
|
| 525 |
+
"grad_norm": 79.39418029785156,
|
| 526 |
+
"learning_rate": 4.937958340599467e-05,
|
| 527 |
+
"loss": 237.785498046875,
|
| 528 |
+
"step": 1460
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.3718252358583149,
|
| 532 |
+
"grad_norm": 77.78357696533203,
|
| 533 |
+
"learning_rate": 4.934265965189626e-05,
|
| 534 |
+
"loss": 247.1903564453125,
|
| 535 |
+
"step": 1480
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.3768499012077516,
|
| 539 |
+
"grad_norm": 78.9517593383789,
|
| 540 |
+
"learning_rate": 4.930468329428271e-05,
|
| 541 |
+
"loss": 218.24853515625,
|
| 542 |
+
"step": 1500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.38187456655718827,
|
| 546 |
+
"grad_norm": 77.0859146118164,
|
| 547 |
+
"learning_rate": 4.9265655975293693e-05,
|
| 548 |
+
"loss": 216.1979248046875,
|
| 549 |
+
"step": 1520
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.38689923190662495,
|
| 553 |
+
"grad_norm": 79.377685546875,
|
| 554 |
+
"learning_rate": 4.9225579382513656e-05,
|
| 555 |
+
"loss": 211.2160888671875,
|
| 556 |
+
"step": 1540
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.39192389725606164,
|
| 560 |
+
"grad_norm": 74.0326156616211,
|
| 561 |
+
"learning_rate": 4.918445524889876e-05,
|
| 562 |
+
"loss": 217.993310546875,
|
| 563 |
+
"step": 1560
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.3969485626054983,
|
| 567 |
+
"grad_norm": 80.3754653930664,
|
| 568 |
+
"learning_rate": 4.914228535270201e-05,
|
| 569 |
+
"loss": 222.884765625,
|
| 570 |
+
"step": 1580
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.401973227954935,
|
| 574 |
+
"grad_norm": 68.28597259521484,
|
| 575 |
+
"learning_rate": 4.909907151739633e-05,
|
| 576 |
+
"loss": 208.1251220703125,
|
| 577 |
+
"step": 1600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.4069978933043717,
|
| 581 |
+
"grad_norm": 62.552974700927734,
|
| 582 |
+
"learning_rate": 4.905481561159575e-05,
|
| 583 |
+
"loss": 206.8597900390625,
|
| 584 |
+
"step": 1620
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.41202255865380844,
|
| 588 |
+
"grad_norm": 73.88465118408203,
|
| 589 |
+
"learning_rate": 4.900951954897456e-05,
|
| 590 |
+
"loss": 200.8366455078125,
|
| 591 |
+
"step": 1640
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.4170472240032451,
|
| 595 |
+
"grad_norm": 66.97296905517578,
|
| 596 |
+
"learning_rate": 4.8963185288184574e-05,
|
| 597 |
+
"loss": 187.9607421875,
|
| 598 |
+
"step": 1660
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.4220718893526818,
|
| 602 |
+
"grad_norm": 73.1982192993164,
|
| 603 |
+
"learning_rate": 4.891581483277046e-05,
|
| 604 |
+
"loss": 202.24766845703124,
|
| 605 |
+
"step": 1680
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.4270965547021185,
|
| 609 |
+
"grad_norm": 67.89013671875,
|
| 610 |
+
"learning_rate": 4.886741023108307e-05,
|
| 611 |
+
"loss": 192.30718994140625,
|
| 612 |
+
"step": 1700
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.4321212200515552,
|
| 616 |
+
"grad_norm": 78.48082733154297,
|
| 617 |
+
"learning_rate": 4.8817973576190915e-05,
|
| 618 |
+
"loss": 178.83857421875,
|
| 619 |
+
"step": 1720
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.43714588540099186,
|
| 623 |
+
"grad_norm": 74.65118408203125,
|
| 624 |
+
"learning_rate": 4.876750700578959e-05,
|
| 625 |
+
"loss": 185.63944091796876,
|
| 626 |
+
"step": 1740
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.44217055075042855,
|
| 630 |
+
"grad_norm": 69.2575912475586,
|
| 631 |
+
"learning_rate": 4.871601270210939e-05,
|
| 632 |
+
"loss": 171.92572021484375,
|
| 633 |
+
"step": 1760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.44719521609986523,
|
| 637 |
+
"grad_norm": 69.49695587158203,
|
| 638 |
+
"learning_rate": 4.866349289182095e-05,
|
| 639 |
+
"loss": 174.3166259765625,
|
| 640 |
+
"step": 1780
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.4522198814493019,
|
| 644 |
+
"grad_norm": 65.37928009033203,
|
| 645 |
+
"learning_rate": 4.860994984593892e-05,
|
| 646 |
+
"loss": 185.66673583984374,
|
| 647 |
+
"step": 1800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.4572445467987386,
|
| 651 |
+
"grad_norm": 80.81356048583984,
|
| 652 |
+
"learning_rate": 4.855538587972379e-05,
|
| 653 |
+
"loss": 167.90670166015624,
|
| 654 |
+
"step": 1820
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.4622692121481753,
|
| 658 |
+
"grad_norm": 75.3848876953125,
|
| 659 |
+
"learning_rate": 4.8499803352581795e-05,
|
| 660 |
+
"loss": 163.5643310546875,
|
| 661 |
+
"step": 1840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.467293877497612,
|
| 665 |
+
"grad_norm": 62.15745162963867,
|
| 666 |
+
"learning_rate": 4.844320466796286e-05,
|
| 667 |
+
"loss": 161.65322265625,
|
| 668 |
+
"step": 1860
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.47231854284704866,
|
| 672 |
+
"grad_norm": 60.19189453125,
|
| 673 |
+
"learning_rate": 4.8385592273256675e-05,
|
| 674 |
+
"loss": 163.45,
|
| 675 |
+
"step": 1880
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.47734320819648535,
|
| 679 |
+
"grad_norm": 77.19636535644531,
|
| 680 |
+
"learning_rate": 4.832696865968686e-05,
|
| 681 |
+
"loss": 169.5025390625,
|
| 682 |
+
"step": 1900
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.48236787354592203,
|
| 686 |
+
"grad_norm": 65.1484146118164,
|
| 687 |
+
"learning_rate": 4.826733636220332e-05,
|
| 688 |
+
"loss": 157.4214111328125,
|
| 689 |
+
"step": 1920
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.4873925388953587,
|
| 693 |
+
"grad_norm": 76.47470092773438,
|
| 694 |
+
"learning_rate": 4.820669795937248e-05,
|
| 695 |
+
"loss": 145.504833984375,
|
| 696 |
+
"step": 1940
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.4924172042447954,
|
| 700 |
+
"grad_norm": 106.30007934570312,
|
| 701 |
+
"learning_rate": 4.8145056073265945e-05,
|
| 702 |
+
"loss": 144.782275390625,
|
| 703 |
+
"step": 1960
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.4974418695942321,
|
| 707 |
+
"grad_norm": 74.23037719726562,
|
| 708 |
+
"learning_rate": 4.8082413369347003e-05,
|
| 709 |
+
"loss": 147.2016357421875,
|
| 710 |
+
"step": 1980
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.5024665349436688,
|
| 714 |
+
"grad_norm": 61.561580657958984,
|
| 715 |
+
"learning_rate": 4.8018772556355436e-05,
|
| 716 |
+
"loss": 146.63094482421874,
|
| 717 |
+
"step": 2000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.5074912002931055,
|
| 721 |
+
"grad_norm": 66.47526550292969,
|
| 722 |
+
"learning_rate": 4.795413638619035e-05,
|
| 723 |
+
"loss": 147.86654052734374,
|
| 724 |
+
"step": 2020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.5125158656425421,
|
| 728 |
+
"grad_norm": 81.40436553955078,
|
| 729 |
+
"learning_rate": 4.788850765379119e-05,
|
| 730 |
+
"loss": 143.17598876953124,
|
| 731 |
+
"step": 2040
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.5175405309919788,
|
| 735 |
+
"grad_norm": 66.76620483398438,
|
| 736 |
+
"learning_rate": 4.782188919701692e-05,
|
| 737 |
+
"loss": 147.62774658203125,
|
| 738 |
+
"step": 2060
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.5225651963414155,
|
| 742 |
+
"grad_norm": 58.314029693603516,
|
| 743 |
+
"learning_rate": 4.775428389652326e-05,
|
| 744 |
+
"loss": 132.99815673828124,
|
| 745 |
+
"step": 2080
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.5275898616908522,
|
| 749 |
+
"grad_norm": 50.07378005981445,
|
| 750 |
+
"learning_rate": 4.768569467563812e-05,
|
| 751 |
+
"loss": 146.73978271484376,
|
| 752 |
+
"step": 2100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.5326145270402889,
|
| 756 |
+
"grad_norm": 61.799163818359375,
|
| 757 |
+
"learning_rate": 4.761612450023526e-05,
|
| 758 |
+
"loss": 123.45181884765626,
|
| 759 |
+
"step": 2120
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.5376391923897256,
|
| 763 |
+
"grad_norm": 55.920928955078125,
|
| 764 |
+
"learning_rate": 4.754557637860597e-05,
|
| 765 |
+
"loss": 130.98118896484374,
|
| 766 |
+
"step": 2140
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.5426638577391623,
|
| 770 |
+
"grad_norm": 54.5310173034668,
|
| 771 |
+
"learning_rate": 4.747405336132903e-05,
|
| 772 |
+
"loss": 123.3492431640625,
|
| 773 |
+
"step": 2160
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.5476885230885989,
|
| 777 |
+
"grad_norm": 60.343509674072266,
|
| 778 |
+
"learning_rate": 4.740155854113877e-05,
|
| 779 |
+
"loss": 125.5275390625,
|
| 780 |
+
"step": 2180
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.5527131884380356,
|
| 784 |
+
"grad_norm": 57.162322998046875,
|
| 785 |
+
"learning_rate": 4.732809505279136e-05,
|
| 786 |
+
"loss": 124.96339111328125,
|
| 787 |
+
"step": 2200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.5577378537874723,
|
| 791 |
+
"grad_norm": 60.92855453491211,
|
| 792 |
+
"learning_rate": 4.725366607292926e-05,
|
| 793 |
+
"loss": 119.469873046875,
|
| 794 |
+
"step": 2220
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.562762519136909,
|
| 798 |
+
"grad_norm": 58.21229934692383,
|
| 799 |
+
"learning_rate": 4.717827481994384e-05,
|
| 800 |
+
"loss": 114.98289794921875,
|
| 801 |
+
"step": 2240
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.5677871844863457,
|
| 805 |
+
"grad_norm": 51.71220779418945,
|
| 806 |
+
"learning_rate": 4.710192455383623e-05,
|
| 807 |
+
"loss": 116.884716796875,
|
| 808 |
+
"step": 2260
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.5728118498357824,
|
| 812 |
+
"grad_norm": 51.0583381652832,
|
| 813 |
+
"learning_rate": 4.702461857607634e-05,
|
| 814 |
+
"loss": 124.8749755859375,
|
| 815 |
+
"step": 2280
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.5778365151852191,
|
| 819 |
+
"grad_norm": 51.23188400268555,
|
| 820 |
+
"learning_rate": 4.694636022946012e-05,
|
| 821 |
+
"loss": 112.71795654296875,
|
| 822 |
+
"step": 2300
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.5828611805346557,
|
| 826 |
+
"grad_norm": 59.2495231628418,
|
| 827 |
+
"learning_rate": 4.6867152897964984e-05,
|
| 828 |
+
"loss": 108.90137939453125,
|
| 829 |
+
"step": 2320
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.5878858458840925,
|
| 833 |
+
"grad_norm": 51.4527702331543,
|
| 834 |
+
"learning_rate": 4.678700000660352e-05,
|
| 835 |
+
"loss": 114.8643798828125,
|
| 836 |
+
"step": 2340
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.5929105112335292,
|
| 840 |
+
"grad_norm": 55.94075012207031,
|
| 841 |
+
"learning_rate": 4.670590502127537e-05,
|
| 842 |
+
"loss": 106.931298828125,
|
| 843 |
+
"step": 2360
|
| 844 |
+
}
|
| 845 |
+
],
|
| 846 |
+
"logging_steps": 20,
|
| 847 |
+
"max_steps": 9555,
|
| 848 |
+
"num_input_tokens_seen": 0,
|
| 849 |
+
"num_train_epochs": 3,
|
| 850 |
+
"save_steps": 1000000000,
|
| 851 |
+
"stateful_callbacks": {
|
| 852 |
+
"TrainerControl": {
|
| 853 |
+
"args": {
|
| 854 |
+
"should_epoch_stop": false,
|
| 855 |
+
"should_evaluate": false,
|
| 856 |
+
"should_log": false,
|
| 857 |
+
"should_save": true,
|
| 858 |
+
"should_training_stop": false
|
| 859 |
+
},
|
| 860 |
+
"attributes": {}
|
| 861 |
+
}
|
| 862 |
+
},
|
| 863 |
+
"total_flos": 4766908314710016.0,
|
| 864 |
+
"train_batch_size": 1,
|
| 865 |
+
"trial_name": null,
|
| 866 |
+
"trial_params": null
|
| 867 |
+
}
|