End of training
Browse files- README.md +3 -3
- all_results.json +16 -0
- eval_results.json +10 -0
- train_results.json +10 -0
- trainer_state.json +477 -0
README.md
CHANGED
|
@@ -18,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
This model is a fine-tuned version of [EleutherAI/pythia-410m](https://huggingface.co/EleutherAI/pythia-410m) on an unknown dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 1.
|
| 22 |
-
- Accuracy: 0.
|
| 23 |
-
- Num Input Tokens Seen:
|
| 24 |
|
| 25 |
## Model description
|
| 26 |
|
|
|
|
| 18 |
|
| 19 |
This model is a fine-tuned version of [EleutherAI/pythia-410m](https://huggingface.co/EleutherAI/pythia-410m) on an unknown dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 1.6346
|
| 22 |
+
- Accuracy: 0.2113
|
| 23 |
+
- Num Input Tokens Seen: 1941504
|
| 24 |
|
| 25 |
## Model description
|
| 26 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.2112676056338028,
|
| 4 |
+
"eval_loss": 1.6345868110656738,
|
| 5 |
+
"eval_runtime": 0.3769,
|
| 6 |
+
"eval_samples": 71,
|
| 7 |
+
"eval_samples_per_second": 188.367,
|
| 8 |
+
"eval_steps_per_second": 23.877,
|
| 9 |
+
"num_input_tokens_seen": 1941504,
|
| 10 |
+
"total_flos": 3521692676653056.0,
|
| 11 |
+
"train_loss": 1.8175702477306253,
|
| 12 |
+
"train_runtime": 153.0605,
|
| 13 |
+
"train_samples": 631,
|
| 14 |
+
"train_samples_per_second": 12.368,
|
| 15 |
+
"train_steps_per_second": 1.548
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.2112676056338028,
|
| 4 |
+
"eval_loss": 1.6345868110656738,
|
| 5 |
+
"eval_runtime": 0.3769,
|
| 6 |
+
"eval_samples": 71,
|
| 7 |
+
"eval_samples_per_second": 188.367,
|
| 8 |
+
"eval_steps_per_second": 23.877,
|
| 9 |
+
"num_input_tokens_seen": 1941504
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"num_input_tokens_seen": 1941504,
|
| 4 |
+
"total_flos": 3521692676653056.0,
|
| 5 |
+
"train_loss": 1.8175702477306253,
|
| 6 |
+
"train_runtime": 153.0605,
|
| 7 |
+
"train_samples": 631,
|
| 8 |
+
"train_samples_per_second": 12.368,
|
| 9 |
+
"train_steps_per_second": 1.548
|
| 10 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,477 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.6345868110656738,
|
| 3 |
+
"best_model_checkpoint": "/data/user_data/gonilude/python_and_text_pythia_410m/checkpoint-150",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 50,
|
| 6 |
+
"global_step": 237,
|
| 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 |
+
"eval_accuracy": 0.23943661971830985,
|
| 14 |
+
"eval_loss": 2.654968023300171,
|
| 15 |
+
"eval_runtime": 0.502,
|
| 16 |
+
"eval_samples_per_second": 141.43,
|
| 17 |
+
"eval_steps_per_second": 17.928,
|
| 18 |
+
"num_input_tokens_seen": 0,
|
| 19 |
+
"step": 0
|
| 20 |
+
},
|
| 21 |
+
{
|
| 22 |
+
"epoch": 0.012658227848101266,
|
| 23 |
+
"grad_norm": NaN,
|
| 24 |
+
"learning_rate": 0.0,
|
| 25 |
+
"loss": 2.196,
|
| 26 |
+
"num_input_tokens_seen": 8192,
|
| 27 |
+
"step": 1
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.06329113924050633,
|
| 31 |
+
"grad_norm": Infinity,
|
| 32 |
+
"learning_rate": 0.0,
|
| 33 |
+
"loss": 2.6755,
|
| 34 |
+
"num_input_tokens_seen": 40960,
|
| 35 |
+
"step": 5
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 0.12658227848101267,
|
| 39 |
+
"grad_norm": 403.4698181152344,
|
| 40 |
+
"learning_rate": 1.25e-05,
|
| 41 |
+
"loss": 3.1957,
|
| 42 |
+
"num_input_tokens_seen": 81920,
|
| 43 |
+
"step": 10
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.189873417721519,
|
| 47 |
+
"grad_norm": 226.32037353515625,
|
| 48 |
+
"learning_rate": 1.9999058994907564e-05,
|
| 49 |
+
"loss": 3.0085,
|
| 50 |
+
"num_input_tokens_seen": 122880,
|
| 51 |
+
"step": 15
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.25316455696202533,
|
| 55 |
+
"grad_norm": 128.1693878173828,
|
| 56 |
+
"learning_rate": 1.99661424082419e-05,
|
| 57 |
+
"loss": 2.7458,
|
| 58 |
+
"num_input_tokens_seen": 163840,
|
| 59 |
+
"step": 20
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.31645569620253167,
|
| 63 |
+
"grad_norm": 231.92202758789062,
|
| 64 |
+
"learning_rate": 1.9886352515311134e-05,
|
| 65 |
+
"loss": 2.0427,
|
| 66 |
+
"num_input_tokens_seen": 204800,
|
| 67 |
+
"step": 25
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 0.379746835443038,
|
| 71 |
+
"grad_norm": 62.582794189453125,
|
| 72 |
+
"learning_rate": 1.9760064588305347e-05,
|
| 73 |
+
"loss": 1.7791,
|
| 74 |
+
"num_input_tokens_seen": 245760,
|
| 75 |
+
"step": 30
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.4430379746835443,
|
| 79 |
+
"grad_norm": 87.08434295654297,
|
| 80 |
+
"learning_rate": 1.9587872591512583e-05,
|
| 81 |
+
"loss": 2.2494,
|
| 82 |
+
"num_input_tokens_seen": 286720,
|
| 83 |
+
"step": 35
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.5063291139240507,
|
| 87 |
+
"grad_norm": 63.20038604736328,
|
| 88 |
+
"learning_rate": 1.9370586387753532e-05,
|
| 89 |
+
"loss": 2.0989,
|
| 90 |
+
"num_input_tokens_seen": 327680,
|
| 91 |
+
"step": 40
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.569620253164557,
|
| 95 |
+
"grad_norm": 32.26015090942383,
|
| 96 |
+
"learning_rate": 1.9109227929390378e-05,
|
| 97 |
+
"loss": 2.1393,
|
| 98 |
+
"num_input_tokens_seen": 368640,
|
| 99 |
+
"step": 45
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.6329113924050633,
|
| 103 |
+
"grad_norm": 51.239463806152344,
|
| 104 |
+
"learning_rate": 1.8805026451824547e-05,
|
| 105 |
+
"loss": 1.8381,
|
| 106 |
+
"num_input_tokens_seen": 409600,
|
| 107 |
+
"step": 50
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.6329113924050633,
|
| 111 |
+
"eval_accuracy": 0.1267605633802817,
|
| 112 |
+
"eval_loss": 2.1300408840179443,
|
| 113 |
+
"eval_runtime": 0.3672,
|
| 114 |
+
"eval_samples_per_second": 193.345,
|
| 115 |
+
"eval_steps_per_second": 24.509,
|
| 116 |
+
"num_input_tokens_seen": 409600,
|
| 117 |
+
"step": 50
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.6962025316455697,
|
| 121 |
+
"grad_norm": 45.65270233154297,
|
| 122 |
+
"learning_rate": 1.8459412692089497e-05,
|
| 123 |
+
"loss": 1.9123,
|
| 124 |
+
"num_input_tokens_seen": 450560,
|
| 125 |
+
"step": 55
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.759493670886076,
|
| 129 |
+
"grad_norm": 33.92768096923828,
|
| 130 |
+
"learning_rate": 1.8074012159730034e-05,
|
| 131 |
+
"loss": 1.6314,
|
| 132 |
+
"num_input_tokens_seen": 491520,
|
| 133 |
+
"step": 60
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 0.8227848101265823,
|
| 137 |
+
"grad_norm": 34.99441146850586,
|
| 138 |
+
"learning_rate": 1.765063749161688e-05,
|
| 139 |
+
"loss": 1.8143,
|
| 140 |
+
"num_input_tokens_seen": 532480,
|
| 141 |
+
"step": 65
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.8860759493670886,
|
| 145 |
+
"grad_norm": 47.59036636352539,
|
| 146 |
+
"learning_rate": 1.719127992665376e-05,
|
| 147 |
+
"loss": 1.8323,
|
| 148 |
+
"num_input_tokens_seen": 573440,
|
| 149 |
+
"step": 70
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.9493670886075949,
|
| 153 |
+
"grad_norm": 19.0941162109375,
|
| 154 |
+
"learning_rate": 1.6698099940473644e-05,
|
| 155 |
+
"loss": 1.7099,
|
| 156 |
+
"num_input_tokens_seen": 614400,
|
| 157 |
+
"step": 75
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.0126582278481013,
|
| 161 |
+
"grad_norm": 56.19622039794922,
|
| 162 |
+
"learning_rate": 1.6173417084171537e-05,
|
| 163 |
+
"loss": 1.7048,
|
| 164 |
+
"num_input_tokens_seen": 655360,
|
| 165 |
+
"step": 80
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.0759493670886076,
|
| 169 |
+
"grad_norm": 50.44072723388672,
|
| 170 |
+
"learning_rate": 1.5619699074864864e-05,
|
| 171 |
+
"loss": 1.817,
|
| 172 |
+
"num_input_tokens_seen": 696320,
|
| 173 |
+
"step": 85
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 1.139240506329114,
|
| 177 |
+
"grad_norm": 47.04330825805664,
|
| 178 |
+
"learning_rate": 1.50395501893913e-05,
|
| 179 |
+
"loss": 1.8287,
|
| 180 |
+
"num_input_tokens_seen": 737280,
|
| 181 |
+
"step": 90
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 1.2025316455696202,
|
| 185 |
+
"grad_norm": 29.618629455566406,
|
| 186 |
+
"learning_rate": 1.4435699015731449e-05,
|
| 187 |
+
"loss": 1.6956,
|
| 188 |
+
"num_input_tokens_seen": 778240,
|
| 189 |
+
"step": 95
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.2658227848101267,
|
| 193 |
+
"grad_norm": 22.177780151367188,
|
| 194 |
+
"learning_rate": 1.3810985619764573e-05,
|
| 195 |
+
"loss": 1.8597,
|
| 196 |
+
"num_input_tokens_seen": 819200,
|
| 197 |
+
"step": 100
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 1.2658227848101267,
|
| 201 |
+
"eval_accuracy": 0.19718309859154928,
|
| 202 |
+
"eval_loss": 1.6786352396011353,
|
| 203 |
+
"eval_runtime": 0.3701,
|
| 204 |
+
"eval_samples_per_second": 191.836,
|
| 205 |
+
"eval_steps_per_second": 24.317,
|
| 206 |
+
"num_input_tokens_seen": 819200,
|
| 207 |
+
"step": 100
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.3291139240506329,
|
| 211 |
+
"grad_norm": 32.201202392578125,
|
| 212 |
+
"learning_rate": 1.3168348187715353e-05,
|
| 213 |
+
"loss": 1.6832,
|
| 214 |
+
"num_input_tokens_seen": 860160,
|
| 215 |
+
"step": 105
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 1.3924050632911391,
|
| 219 |
+
"grad_norm": 31.627992630004883,
|
| 220 |
+
"learning_rate": 1.2510809207115666e-05,
|
| 221 |
+
"loss": 1.663,
|
| 222 |
+
"num_input_tokens_seen": 901120,
|
| 223 |
+
"step": 110
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 1.4556962025316456,
|
| 227 |
+
"grad_norm": 20.219005584716797,
|
| 228 |
+
"learning_rate": 1.1841461251275868e-05,
|
| 229 |
+
"loss": 1.7544,
|
| 230 |
+
"num_input_tokens_seen": 942080,
|
| 231 |
+
"step": 115
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.518987341772152,
|
| 235 |
+
"grad_norm": 22.077184677124023,
|
| 236 |
+
"learning_rate": 1.1163452434124773e-05,
|
| 237 |
+
"loss": 1.6455,
|
| 238 |
+
"num_input_tokens_seen": 983040,
|
| 239 |
+
"step": 120
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.5822784810126582,
|
| 243 |
+
"grad_norm": 24.04667091369629,
|
| 244 |
+
"learning_rate": 1.0479971603828001e-05,
|
| 245 |
+
"loss": 1.6255,
|
| 246 |
+
"num_input_tokens_seen": 1024000,
|
| 247 |
+
"step": 125
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.6455696202531644,
|
| 251 |
+
"grad_norm": 24.609832763671875,
|
| 252 |
+
"learning_rate": 9.79423334482279e-06,
|
| 253 |
+
"loss": 1.8148,
|
| 254 |
+
"num_input_tokens_seen": 1064960,
|
| 255 |
+
"step": 130
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.7088607594936709,
|
| 259 |
+
"grad_norm": 16.160457611083984,
|
| 260 |
+
"learning_rate": 9.109462858808586e-06,
|
| 261 |
+
"loss": 1.6357,
|
| 262 |
+
"num_input_tokens_seen": 1105920,
|
| 263 |
+
"step": 135
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.7721518987341773,
|
| 267 |
+
"grad_norm": 23.06271743774414,
|
| 268 |
+
"learning_rate": 8.428880795801965e-06,
|
| 269 |
+
"loss": 1.6064,
|
| 270 |
+
"num_input_tokens_seen": 1146880,
|
| 271 |
+
"step": 140
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 1.8354430379746836,
|
| 275 |
+
"grad_norm": 27.028121948242188,
|
| 276 |
+
"learning_rate": 7.75568810659924e-06,
|
| 277 |
+
"loss": 1.6333,
|
| 278 |
+
"num_input_tokens_seen": 1187840,
|
| 279 |
+
"step": 145
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.8987341772151898,
|
| 283 |
+
"grad_norm": 48.31101608276367,
|
| 284 |
+
"learning_rate": 7.093050987889547e-06,
|
| 285 |
+
"loss": 1.694,
|
| 286 |
+
"num_input_tokens_seen": 1228800,
|
| 287 |
+
"step": 150
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.8987341772151898,
|
| 291 |
+
"eval_accuracy": 0.2112676056338028,
|
| 292 |
+
"eval_loss": 1.6345868110656738,
|
| 293 |
+
"eval_runtime": 0.3747,
|
| 294 |
+
"eval_samples_per_second": 189.475,
|
| 295 |
+
"eval_steps_per_second": 24.018,
|
| 296 |
+
"num_input_tokens_seen": 1228800,
|
| 297 |
+
"step": 150
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.9620253164556962,
|
| 301 |
+
"grad_norm": 12.064297676086426,
|
| 302 |
+
"learning_rate": 6.444085990825338e-06,
|
| 303 |
+
"loss": 1.5869,
|
| 304 |
+
"num_input_tokens_seen": 1269760,
|
| 305 |
+
"step": 155
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.0253164556962027,
|
| 309 |
+
"grad_norm": 21.297130584716797,
|
| 310 |
+
"learning_rate": 5.811845363088477e-06,
|
| 311 |
+
"loss": 1.5669,
|
| 312 |
+
"num_input_tokens_seen": 1310720,
|
| 313 |
+
"step": 160
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 2.088607594936709,
|
| 317 |
+
"grad_norm": 26.849414825439453,
|
| 318 |
+
"learning_rate": 5.199302693391958e-06,
|
| 319 |
+
"loss": 1.6514,
|
| 320 |
+
"num_input_tokens_seen": 1351680,
|
| 321 |
+
"step": 165
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 2.151898734177215,
|
| 325 |
+
"grad_norm": 28.412883758544922,
|
| 326 |
+
"learning_rate": 4.609338925934743e-06,
|
| 327 |
+
"loss": 1.6545,
|
| 328 |
+
"num_input_tokens_seen": 1392640,
|
| 329 |
+
"step": 170
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 2.2151898734177213,
|
| 333 |
+
"grad_norm": 38.271270751953125,
|
| 334 |
+
"learning_rate": 4.044728810587406e-06,
|
| 335 |
+
"loss": 1.5853,
|
| 336 |
+
"num_input_tokens_seen": 1433600,
|
| 337 |
+
"step": 175
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 2.278481012658228,
|
| 341 |
+
"grad_norm": 48.123687744140625,
|
| 342 |
+
"learning_rate": 3.508127852536698e-06,
|
| 343 |
+
"loss": 1.6596,
|
| 344 |
+
"num_input_tokens_seen": 1474560,
|
| 345 |
+
"step": 180
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.3417721518987342,
|
| 349 |
+
"grad_norm": 30.174638748168945,
|
| 350 |
+
"learning_rate": 3.0020598227682794e-06,
|
| 351 |
+
"loss": 1.6385,
|
| 352 |
+
"num_input_tokens_seen": 1515520,
|
| 353 |
+
"step": 185
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 2.4050632911392404,
|
| 357 |
+
"grad_norm": 15.243988037109375,
|
| 358 |
+
"learning_rate": 2.5289048881289256e-06,
|
| 359 |
+
"loss": 1.5851,
|
| 360 |
+
"num_input_tokens_seen": 1556480,
|
| 361 |
+
"step": 190
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.4683544303797467,
|
| 365 |
+
"grad_norm": 50.41596603393555,
|
| 366 |
+
"learning_rate": 2.090888416795582e-06,
|
| 367 |
+
"loss": 1.6231,
|
| 368 |
+
"num_input_tokens_seen": 1597440,
|
| 369 |
+
"step": 195
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 2.5316455696202533,
|
| 373 |
+
"grad_norm": 17.154624938964844,
|
| 374 |
+
"learning_rate": 1.69007051180199e-06,
|
| 375 |
+
"loss": 1.586,
|
| 376 |
+
"num_input_tokens_seen": 1638400,
|
| 377 |
+
"step": 200
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 2.5316455696202533,
|
| 381 |
+
"eval_accuracy": 0.14084507042253522,
|
| 382 |
+
"eval_loss": 1.6395212411880493,
|
| 383 |
+
"eval_runtime": 0.4282,
|
| 384 |
+
"eval_samples_per_second": 165.806,
|
| 385 |
+
"eval_steps_per_second": 21.018,
|
| 386 |
+
"num_input_tokens_seen": 1638400,
|
| 387 |
+
"step": 200
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 2.5949367088607596,
|
| 391 |
+
"grad_norm": 25.179792404174805,
|
| 392 |
+
"learning_rate": 1.3283363218493962e-06,
|
| 393 |
+
"loss": 1.5881,
|
| 394 |
+
"num_input_tokens_seen": 1679360,
|
| 395 |
+
"step": 205
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 2.6582278481012658,
|
| 399 |
+
"grad_norm": 14.442487716674805,
|
| 400 |
+
"learning_rate": 1.0073871749720221e-06,
|
| 401 |
+
"loss": 1.5929,
|
| 402 |
+
"num_input_tokens_seen": 1720320,
|
| 403 |
+
"step": 210
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 2.721518987341772,
|
| 407 |
+
"grad_norm": 24.72681999206543,
|
| 408 |
+
"learning_rate": 7.287325767579756e-07,
|
| 409 |
+
"loss": 1.5963,
|
| 410 |
+
"num_input_tokens_seen": 1761280,
|
| 411 |
+
"step": 215
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 2.7848101265822782,
|
| 415 |
+
"grad_norm": 9.948562622070312,
|
| 416 |
+
"learning_rate": 4.936831107599749e-07,
|
| 417 |
+
"loss": 1.5503,
|
| 418 |
+
"num_input_tokens_seen": 1802240,
|
| 419 |
+
"step": 220
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 2.848101265822785,
|
| 423 |
+
"grad_norm": 30.714181900024414,
|
| 424 |
+
"learning_rate": 3.033442744870685e-07,
|
| 425 |
+
"loss": 1.6379,
|
| 426 |
+
"num_input_tokens_seen": 1843200,
|
| 427 |
+
"step": 225
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 2.911392405063291,
|
| 431 |
+
"grad_norm": 20.53106117248535,
|
| 432 |
+
"learning_rate": 1.5861127996827597e-07,
|
| 433 |
+
"loss": 1.5714,
|
| 434 |
+
"num_input_tokens_seen": 1884160,
|
| 435 |
+
"step": 230
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 2.9746835443037973,
|
| 439 |
+
"grad_norm": 37.22603225708008,
|
| 440 |
+
"learning_rate": 6.016484334238515e-08,
|
| 441 |
+
"loss": 1.5922,
|
| 442 |
+
"num_input_tokens_seen": 1925120,
|
| 443 |
+
"step": 235
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 3.0,
|
| 447 |
+
"num_input_tokens_seen": 1941504,
|
| 448 |
+
"step": 237,
|
| 449 |
+
"total_flos": 3521692676653056.0,
|
| 450 |
+
"train_loss": 1.8175702477306253,
|
| 451 |
+
"train_runtime": 153.0605,
|
| 452 |
+
"train_samples_per_second": 12.368,
|
| 453 |
+
"train_steps_per_second": 1.548
|
| 454 |
+
}
|
| 455 |
+
],
|
| 456 |
+
"logging_steps": 5,
|
| 457 |
+
"max_steps": 237,
|
| 458 |
+
"num_input_tokens_seen": 1941504,
|
| 459 |
+
"num_train_epochs": 3,
|
| 460 |
+
"save_steps": 50,
|
| 461 |
+
"stateful_callbacks": {
|
| 462 |
+
"TrainerControl": {
|
| 463 |
+
"args": {
|
| 464 |
+
"should_epoch_stop": false,
|
| 465 |
+
"should_evaluate": false,
|
| 466 |
+
"should_log": false,
|
| 467 |
+
"should_save": true,
|
| 468 |
+
"should_training_stop": true
|
| 469 |
+
},
|
| 470 |
+
"attributes": {}
|
| 471 |
+
}
|
| 472 |
+
},
|
| 473 |
+
"total_flos": 3521692676653056.0,
|
| 474 |
+
"train_batch_size": 4,
|
| 475 |
+
"trial_name": null,
|
| 476 |
+
"trial_params": null
|
| 477 |
+
}
|