End of training
Browse files- README.md +3 -2
- all_results.json +12 -0
- eval_results.json +7 -0
- train_results.json +8 -0
- trainer_state.json +615 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: llama3
|
|
| 4 |
base_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: sft_mc
|
|
@@ -15,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# sft_mc
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 2.
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
|
|
| 4 |
base_model: meta-llama/Meta-Llama-3-8B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: sft_mc
|
|
|
|
| 16 |
|
| 17 |
# sft_mc
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the identity and the data_mc datasets.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 2.3274
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_loss": 2.3273847103118896,
|
| 4 |
+
"eval_runtime": 0.9866,
|
| 5 |
+
"eval_samples_per_second": 22.299,
|
| 6 |
+
"eval_steps_per_second": 3.041,
|
| 7 |
+
"total_flos": 51054080163840.0,
|
| 8 |
+
"train_loss": 0.28787647419011414,
|
| 9 |
+
"train_runtime": 2129.2799,
|
| 10 |
+
"train_samples_per_second": 5.02,
|
| 11 |
+
"train_steps_per_second": 0.315
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_loss": 2.3273847103118896,
|
| 4 |
+
"eval_runtime": 0.9866,
|
| 5 |
+
"eval_samples_per_second": 22.299,
|
| 6 |
+
"eval_steps_per_second": 3.041
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"total_flos": 51054080163840.0,
|
| 4 |
+
"train_loss": 0.28787647419011414,
|
| 5 |
+
"train_runtime": 2129.2799,
|
| 6 |
+
"train_samples_per_second": 5.02,
|
| 7 |
+
"train_steps_per_second": 0.315
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,615 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 50,
|
| 6 |
+
"global_step": 670,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.14925373134328357,
|
| 13 |
+
"grad_norm": 35.90125075274422,
|
| 14 |
+
"learning_rate": 1.4925373134328358e-06,
|
| 15 |
+
"loss": 1.921,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.29850746268656714,
|
| 20 |
+
"grad_norm": 21.611909970996276,
|
| 21 |
+
"learning_rate": 2.9850746268656716e-06,
|
| 22 |
+
"loss": 1.1511,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.44776119402985076,
|
| 27 |
+
"grad_norm": 17.54986781724882,
|
| 28 |
+
"learning_rate": 4.477611940298508e-06,
|
| 29 |
+
"loss": 0.9965,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.5970149253731343,
|
| 34 |
+
"grad_norm": 12.86435917115879,
|
| 35 |
+
"learning_rate": 5.970149253731343e-06,
|
| 36 |
+
"loss": 1.0133,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.746268656716418,
|
| 41 |
+
"grad_norm": 12.804255629397026,
|
| 42 |
+
"learning_rate": 7.46268656716418e-06,
|
| 43 |
+
"loss": 1.053,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.746268656716418,
|
| 48 |
+
"eval_loss": 1.2667393684387207,
|
| 49 |
+
"eval_runtime": 0.9884,
|
| 50 |
+
"eval_samples_per_second": 22.259,
|
| 51 |
+
"eval_steps_per_second": 3.035,
|
| 52 |
+
"step": 50
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.8955223880597015,
|
| 56 |
+
"grad_norm": 13.275489475520203,
|
| 57 |
+
"learning_rate": 8.955223880597016e-06,
|
| 58 |
+
"loss": 1.0093,
|
| 59 |
+
"step": 60
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.044776119402985,
|
| 63 |
+
"grad_norm": 10.227125744078247,
|
| 64 |
+
"learning_rate": 9.999389284703265e-06,
|
| 65 |
+
"loss": 0.9774,
|
| 66 |
+
"step": 70
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.1940298507462686,
|
| 70 |
+
"grad_norm": 10.433290314777752,
|
| 71 |
+
"learning_rate": 9.988536273658876e-06,
|
| 72 |
+
"loss": 0.7151,
|
| 73 |
+
"step": 80
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 1.3432835820895521,
|
| 77 |
+
"grad_norm": 10.688663360597172,
|
| 78 |
+
"learning_rate": 9.964145714351633e-06,
|
| 79 |
+
"loss": 0.7292,
|
| 80 |
+
"step": 90
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.4925373134328357,
|
| 84 |
+
"grad_norm": 14.363632603556267,
|
| 85 |
+
"learning_rate": 9.926283796211796e-06,
|
| 86 |
+
"loss": 0.8178,
|
| 87 |
+
"step": 100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.4925373134328357,
|
| 91 |
+
"eval_loss": 1.3229804039001465,
|
| 92 |
+
"eval_runtime": 0.9845,
|
| 93 |
+
"eval_samples_per_second": 22.346,
|
| 94 |
+
"eval_steps_per_second": 3.047,
|
| 95 |
+
"step": 100
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.6417910447761193,
|
| 99 |
+
"grad_norm": 10.45476204806484,
|
| 100 |
+
"learning_rate": 9.87505326632108e-06,
|
| 101 |
+
"loss": 0.8059,
|
| 102 |
+
"step": 110
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.7910447761194028,
|
| 106 |
+
"grad_norm": 9.865796302977193,
|
| 107 |
+
"learning_rate": 9.810593150584658e-06,
|
| 108 |
+
"loss": 0.8046,
|
| 109 |
+
"step": 120
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.9402985074626866,
|
| 113 |
+
"grad_norm": 9.16685167044992,
|
| 114 |
+
"learning_rate": 9.733078376452172e-06,
|
| 115 |
+
"loss": 0.7796,
|
| 116 |
+
"step": 130
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.08955223880597,
|
| 120 |
+
"grad_norm": 8.577950674483414,
|
| 121 |
+
"learning_rate": 9.642719298211602e-06,
|
| 122 |
+
"loss": 0.5014,
|
| 123 |
+
"step": 140
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.2388059701492535,
|
| 127 |
+
"grad_norm": 9.564526041646893,
|
| 128 |
+
"learning_rate": 9.539761126144193e-06,
|
| 129 |
+
"loss": 0.3803,
|
| 130 |
+
"step": 150
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.2388059701492535,
|
| 134 |
+
"eval_loss": 1.5490957498550415,
|
| 135 |
+
"eval_runtime": 0.9879,
|
| 136 |
+
"eval_samples_per_second": 22.27,
|
| 137 |
+
"eval_steps_per_second": 3.037,
|
| 138 |
+
"step": 150
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 2.388059701492537,
|
| 142 |
+
"grad_norm": 6.6979250460920055,
|
| 143 |
+
"learning_rate": 9.424483261089584e-06,
|
| 144 |
+
"loss": 0.4,
|
| 145 |
+
"step": 160
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 2.5373134328358207,
|
| 149 |
+
"grad_norm": 7.411435193851966,
|
| 150 |
+
"learning_rate": 9.297198536226927e-06,
|
| 151 |
+
"loss": 0.4118,
|
| 152 |
+
"step": 170
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 2.6865671641791042,
|
| 156 |
+
"grad_norm": 8.596999087873268,
|
| 157 |
+
"learning_rate": 9.158252368129628e-06,
|
| 158 |
+
"loss": 0.4074,
|
| 159 |
+
"step": 180
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 2.835820895522388,
|
| 163 |
+
"grad_norm": 7.821138995440488,
|
| 164 |
+
"learning_rate": 9.008021819397488e-06,
|
| 165 |
+
"loss": 0.438,
|
| 166 |
+
"step": 190
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 2.9850746268656714,
|
| 170 |
+
"grad_norm": 8.079644227754834,
|
| 171 |
+
"learning_rate": 8.846914575410035e-06,
|
| 172 |
+
"loss": 0.436,
|
| 173 |
+
"step": 200
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 2.9850746268656714,
|
| 177 |
+
"eval_loss": 1.5159422159194946,
|
| 178 |
+
"eval_runtime": 0.9922,
|
| 179 |
+
"eval_samples_per_second": 22.173,
|
| 180 |
+
"eval_steps_per_second": 3.024,
|
| 181 |
+
"step": 200
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 3.1343283582089554,
|
| 185 |
+
"grad_norm": 7.1129105753354125,
|
| 186 |
+
"learning_rate": 8.675367837977848e-06,
|
| 187 |
+
"loss": 0.2353,
|
| 188 |
+
"step": 210
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 3.283582089552239,
|
| 192 |
+
"grad_norm": 7.345277153510642,
|
| 193 |
+
"learning_rate": 8.49384713889421e-06,
|
| 194 |
+
"loss": 0.2442,
|
| 195 |
+
"step": 220
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 3.4328358208955225,
|
| 199 |
+
"grad_norm": 9.63151530312459,
|
| 200 |
+
"learning_rate": 8.302845076606786e-06,
|
| 201 |
+
"loss": 0.264,
|
| 202 |
+
"step": 230
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 3.582089552238806,
|
| 206 |
+
"grad_norm": 6.18137195728966,
|
| 207 |
+
"learning_rate": 8.10287997943769e-06,
|
| 208 |
+
"loss": 0.2486,
|
| 209 |
+
"step": 240
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 3.7313432835820897,
|
| 213 |
+
"grad_norm": 7.558744812430462,
|
| 214 |
+
"learning_rate": 7.894494498979558e-06,
|
| 215 |
+
"loss": 0.2529,
|
| 216 |
+
"step": 250
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 3.7313432835820897,
|
| 220 |
+
"eval_loss": 1.6716222763061523,
|
| 221 |
+
"eval_runtime": 0.9844,
|
| 222 |
+
"eval_samples_per_second": 22.349,
|
| 223 |
+
"eval_steps_per_second": 3.048,
|
| 224 |
+
"step": 250
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 3.8805970149253732,
|
| 228 |
+
"grad_norm": 7.589509313359255,
|
| 229 |
+
"learning_rate": 7.678254137484797e-06,
|
| 230 |
+
"loss": 0.2382,
|
| 231 |
+
"step": 260
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 4.029850746268656,
|
| 235 |
+
"grad_norm": 4.292329479116102,
|
| 236 |
+
"learning_rate": 7.4547457132442895e-06,
|
| 237 |
+
"loss": 0.2137,
|
| 238 |
+
"step": 270
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 4.17910447761194,
|
| 242 |
+
"grad_norm": 4.93005317817384,
|
| 243 |
+
"learning_rate": 7.2245757681200835e-06,
|
| 244 |
+
"loss": 0.1201,
|
| 245 |
+
"step": 280
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 4.3283582089552235,
|
| 249 |
+
"grad_norm": 6.340158286084005,
|
| 250 |
+
"learning_rate": 6.988368921553601e-06,
|
| 251 |
+
"loss": 0.1246,
|
| 252 |
+
"step": 290
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.477611940298507,
|
| 256 |
+
"grad_norm": 7.035143744951474,
|
| 257 |
+
"learning_rate": 6.746766175516159e-06,
|
| 258 |
+
"loss": 0.1555,
|
| 259 |
+
"step": 300
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 4.477611940298507,
|
| 263 |
+
"eval_loss": 1.847908616065979,
|
| 264 |
+
"eval_runtime": 0.9862,
|
| 265 |
+
"eval_samples_per_second": 22.307,
|
| 266 |
+
"eval_steps_per_second": 3.042,
|
| 267 |
+
"step": 300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 4.6268656716417915,
|
| 271 |
+
"grad_norm": 5.665676328156013,
|
| 272 |
+
"learning_rate": 6.500423175001705e-06,
|
| 273 |
+
"loss": 0.1391,
|
| 274 |
+
"step": 310
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 4.776119402985074,
|
| 278 |
+
"grad_norm": 6.086585562332805,
|
| 279 |
+
"learning_rate": 6.2500084287822925e-06,
|
| 280 |
+
"loss": 0.1576,
|
| 281 |
+
"step": 320
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 4.925373134328359,
|
| 285 |
+
"grad_norm": 5.964957579344617,
|
| 286 |
+
"learning_rate": 5.996201495254757e-06,
|
| 287 |
+
"loss": 0.1441,
|
| 288 |
+
"step": 330
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 5.074626865671641,
|
| 292 |
+
"grad_norm": 2.2272515089974716,
|
| 293 |
+
"learning_rate": 5.73969113830165e-06,
|
| 294 |
+
"loss": 0.0898,
|
| 295 |
+
"step": 340
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 5.223880597014926,
|
| 299 |
+
"grad_norm": 4.927338192771822,
|
| 300 |
+
"learning_rate": 5.481173458170952e-06,
|
| 301 |
+
"loss": 0.078,
|
| 302 |
+
"step": 350
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 5.223880597014926,
|
| 306 |
+
"eval_loss": 2.03751277923584,
|
| 307 |
+
"eval_runtime": 0.989,
|
| 308 |
+
"eval_samples_per_second": 22.245,
|
| 309 |
+
"eval_steps_per_second": 3.033,
|
| 310 |
+
"step": 350
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.373134328358209,
|
| 314 |
+
"grad_norm": 4.519214108822612,
|
| 315 |
+
"learning_rate": 5.221350002446882e-06,
|
| 316 |
+
"loss": 0.0828,
|
| 317 |
+
"step": 360
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.522388059701493,
|
| 321 |
+
"grad_norm": 3.298936360185895,
|
| 322 |
+
"learning_rate": 4.96092586223808e-06,
|
| 323 |
+
"loss": 0.0734,
|
| 324 |
+
"step": 370
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.6716417910447765,
|
| 328 |
+
"grad_norm": 6.294477283779623,
|
| 329 |
+
"learning_rate": 4.700607758749626e-06,
|
| 330 |
+
"loss": 0.0873,
|
| 331 |
+
"step": 380
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.82089552238806,
|
| 335 |
+
"grad_norm": 4.501734589626586,
|
| 336 |
+
"learning_rate": 4.441102125431398e-06,
|
| 337 |
+
"loss": 0.0647,
|
| 338 |
+
"step": 390
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.970149253731344,
|
| 342 |
+
"grad_norm": 6.080468816741591,
|
| 343 |
+
"learning_rate": 4.183113190907349e-06,
|
| 344 |
+
"loss": 0.0726,
|
| 345 |
+
"step": 400
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.970149253731344,
|
| 349 |
+
"eval_loss": 1.8965630531311035,
|
| 350 |
+
"eval_runtime": 0.9838,
|
| 351 |
+
"eval_samples_per_second": 22.363,
|
| 352 |
+
"eval_steps_per_second": 3.049,
|
| 353 |
+
"step": 400
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 6.119402985074627,
|
| 357 |
+
"grad_norm": 3.2916344262433874,
|
| 358 |
+
"learning_rate": 3.927341067888065e-06,
|
| 359 |
+
"loss": 0.0471,
|
| 360 |
+
"step": 410
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 6.268656716417911,
|
| 364 |
+
"grad_norm": 3.847276551426768,
|
| 365 |
+
"learning_rate": 3.6744798532528137e-06,
|
| 366 |
+
"loss": 0.0332,
|
| 367 |
+
"step": 420
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 6.417910447761194,
|
| 371 |
+
"grad_norm": 2.2055620155467532,
|
| 372 |
+
"learning_rate": 3.4252157444569478e-06,
|
| 373 |
+
"loss": 0.0302,
|
| 374 |
+
"step": 430
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 6.567164179104478,
|
| 378 |
+
"grad_norm": 7.540352868440021,
|
| 379 |
+
"learning_rate": 3.1802251773762294e-06,
|
| 380 |
+
"loss": 0.048,
|
| 381 |
+
"step": 440
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 6.7164179104477615,
|
| 385 |
+
"grad_norm": 3.3252060274641644,
|
| 386 |
+
"learning_rate": 2.9401729906414385e-06,
|
| 387 |
+
"loss": 0.0411,
|
| 388 |
+
"step": 450
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 6.7164179104477615,
|
| 392 |
+
"eval_loss": 2.1223175525665283,
|
| 393 |
+
"eval_runtime": 0.9929,
|
| 394 |
+
"eval_samples_per_second": 22.157,
|
| 395 |
+
"eval_steps_per_second": 3.021,
|
| 396 |
+
"step": 450
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 6.865671641791045,
|
| 400 |
+
"grad_norm": 1.3769658356628893,
|
| 401 |
+
"learning_rate": 2.7057106214448216e-06,
|
| 402 |
+
"loss": 0.0388,
|
| 403 |
+
"step": 460
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 7.014925373134329,
|
| 407 |
+
"grad_norm": 0.5978691678502914,
|
| 408 |
+
"learning_rate": 2.4774743377144265e-06,
|
| 409 |
+
"loss": 0.0395,
|
| 410 |
+
"step": 470
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 7.164179104477612,
|
| 414 |
+
"grad_norm": 0.6955205479840721,
|
| 415 |
+
"learning_rate": 2.256083511453747e-06,
|
| 416 |
+
"loss": 0.0106,
|
| 417 |
+
"step": 480
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 7.313432835820896,
|
| 421 |
+
"grad_norm": 5.0523793409280495,
|
| 422 |
+
"learning_rate": 2.042138937932388e-06,
|
| 423 |
+
"loss": 0.0252,
|
| 424 |
+
"step": 490
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 7.462686567164179,
|
| 428 |
+
"grad_norm": 1.442828007683923,
|
| 429 |
+
"learning_rate": 1.8362212052889827e-06,
|
| 430 |
+
"loss": 0.0127,
|
| 431 |
+
"step": 500
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 7.462686567164179,
|
| 435 |
+
"eval_loss": 2.1807539463043213,
|
| 436 |
+
"eval_runtime": 0.9861,
|
| 437 |
+
"eval_samples_per_second": 22.31,
|
| 438 |
+
"eval_steps_per_second": 3.042,
|
| 439 |
+
"step": 500
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 7.611940298507463,
|
| 443 |
+
"grad_norm": 3.231146614611382,
|
| 444 |
+
"learning_rate": 1.63888911897084e-06,
|
| 445 |
+
"loss": 0.017,
|
| 446 |
+
"step": 510
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 7.7611940298507465,
|
| 450 |
+
"grad_norm": 0.9690088812134732,
|
| 451 |
+
"learning_rate": 1.4506781852859836e-06,
|
| 452 |
+
"loss": 0.0192,
|
| 453 |
+
"step": 520
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 7.91044776119403,
|
| 457 |
+
"grad_norm": 0.15130005364803242,
|
| 458 |
+
"learning_rate": 1.2720991581827852e-06,
|
| 459 |
+
"loss": 0.0102,
|
| 460 |
+
"step": 530
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 8.059701492537313,
|
| 464 |
+
"grad_norm": 0.06663781297906822,
|
| 465 |
+
"learning_rate": 1.1036366532008552e-06,
|
| 466 |
+
"loss": 0.0119,
|
| 467 |
+
"step": 540
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 8.208955223880597,
|
| 471 |
+
"grad_norm": 0.904607664250542,
|
| 472 |
+
"learning_rate": 9.457478323545749e-07,
|
| 473 |
+
"loss": 0.0083,
|
| 474 |
+
"step": 550
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 8.208955223880597,
|
| 478 |
+
"eval_loss": 2.2392308712005615,
|
| 479 |
+
"eval_runtime": 0.9906,
|
| 480 |
+
"eval_samples_per_second": 22.209,
|
| 481 |
+
"eval_steps_per_second": 3.029,
|
| 482 |
+
"step": 550
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 8.35820895522388,
|
| 486 |
+
"grad_norm": 0.4615723348056157,
|
| 487 |
+
"learning_rate": 7.988611635181099e-07,
|
| 488 |
+
"loss": 0.0059,
|
| 489 |
+
"step": 560
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 8.507462686567164,
|
| 493 |
+
"grad_norm": 0.4272952034809819,
|
| 494 |
+
"learning_rate": 6.633752576786251e-07,
|
| 495 |
+
"loss": 0.012,
|
| 496 |
+
"step": 570
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 8.656716417910447,
|
| 500 |
+
"grad_norm": 0.5145191101097174,
|
| 501 |
+
"learning_rate": 5.396577872130676e-07,
|
| 502 |
+
"loss": 0.0071,
|
| 503 |
+
"step": 580
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 8.805970149253731,
|
| 507 |
+
"grad_norm": 0.12341593830842175,
|
| 508 |
+
"learning_rate": 4.2804448812404754e-07,
|
| 509 |
+
"loss": 0.0036,
|
| 510 |
+
"step": 590
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 8.955223880597014,
|
| 514 |
+
"grad_norm": 0.7095059409594088,
|
| 515 |
+
"learning_rate": 3.288382489424502e-07,
|
| 516 |
+
"loss": 0.0049,
|
| 517 |
+
"step": 600
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 8.955223880597014,
|
| 521 |
+
"eval_loss": 2.3008365631103516,
|
| 522 |
+
"eval_runtime": 0.9934,
|
| 523 |
+
"eval_samples_per_second": 22.146,
|
| 524 |
+
"eval_steps_per_second": 3.02,
|
| 525 |
+
"step": 600
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 9.104477611940299,
|
| 529 |
+
"grad_norm": 1.4842604841678675,
|
| 530 |
+
"learning_rate": 2.4230828876927293e-07,
|
| 531 |
+
"loss": 0.0051,
|
| 532 |
+
"step": 610
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 9.253731343283581,
|
| 536 |
+
"grad_norm": 1.0220912222856482,
|
| 537 |
+
"learning_rate": 1.6868942668726408e-07,
|
| 538 |
+
"loss": 0.0032,
|
| 539 |
+
"step": 620
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 9.402985074626866,
|
| 543 |
+
"grad_norm": 0.8790141026127328,
|
| 544 |
+
"learning_rate": 1.0818144452496293e-07,
|
| 545 |
+
"loss": 0.0063,
|
| 546 |
+
"step": 630
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 9.552238805970148,
|
| 550 |
+
"grad_norm": 0.04185510117018065,
|
| 551 |
+
"learning_rate": 6.094854470245326e-08,
|
| 552 |
+
"loss": 0.0024,
|
| 553 |
+
"step": 640
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 9.701492537313433,
|
| 557 |
+
"grad_norm": 0.05125416943229584,
|
| 558 |
+
"learning_rate": 2.711890463007405e-08,
|
| 559 |
+
"loss": 0.0035,
|
| 560 |
+
"step": 650
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 9.701492537313433,
|
| 564 |
+
"eval_loss": 2.3232414722442627,
|
| 565 |
+
"eval_runtime": 0.9838,
|
| 566 |
+
"eval_samples_per_second": 22.362,
|
| 567 |
+
"eval_steps_per_second": 3.049,
|
| 568 |
+
"step": 650
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 9.850746268656717,
|
| 572 |
+
"grad_norm": 0.8613851942534393,
|
| 573 |
+
"learning_rate": 6.784328869339218e-09,
|
| 574 |
+
"loss": 0.0035,
|
| 575 |
+
"step": 660
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 10.0,
|
| 579 |
+
"grad_norm": 0.1820002975830766,
|
| 580 |
+
"learning_rate": 0.0,
|
| 581 |
+
"loss": 0.0021,
|
| 582 |
+
"step": 670
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 10.0,
|
| 586 |
+
"step": 670,
|
| 587 |
+
"total_flos": 51054080163840.0,
|
| 588 |
+
"train_loss": 0.28787647419011414,
|
| 589 |
+
"train_runtime": 2129.2799,
|
| 590 |
+
"train_samples_per_second": 5.02,
|
| 591 |
+
"train_steps_per_second": 0.315
|
| 592 |
+
}
|
| 593 |
+
],
|
| 594 |
+
"logging_steps": 10,
|
| 595 |
+
"max_steps": 670,
|
| 596 |
+
"num_input_tokens_seen": 0,
|
| 597 |
+
"num_train_epochs": 10,
|
| 598 |
+
"save_steps": 500,
|
| 599 |
+
"stateful_callbacks": {
|
| 600 |
+
"TrainerControl": {
|
| 601 |
+
"args": {
|
| 602 |
+
"should_epoch_stop": false,
|
| 603 |
+
"should_evaluate": false,
|
| 604 |
+
"should_log": false,
|
| 605 |
+
"should_save": true,
|
| 606 |
+
"should_training_stop": true
|
| 607 |
+
},
|
| 608 |
+
"attributes": {}
|
| 609 |
+
}
|
| 610 |
+
},
|
| 611 |
+
"total_flos": 51054080163840.0,
|
| 612 |
+
"train_batch_size": 2,
|
| 613 |
+
"trial_name": null,
|
| 614 |
+
"trial_params": null
|
| 615 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|