End of training
Browse files- README.md +2 -1
- all_results.json +12 -0
- eval_results.json +7 -0
- train_results.json +8 -0
- trainer_state.json +591 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: mistralai/Mistral-7B-v0.3
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: mistral_7b_0-3_webinstruct_original_750k_uniform_v2
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# mistral_7b_0-3_webinstruct_original_750k_uniform_v2
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.3](https://huggingface.co/mistralai/Mistral-7B-v0.3) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 0.3717
|
| 21 |
|
|
|
|
| 4 |
base_model: mistralai/Mistral-7B-v0.3
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: mistral_7b_0-3_webinstruct_original_750k_uniform_v2
|
|
|
|
| 16 |
|
| 17 |
# mistral_7b_0-3_webinstruct_original_750k_uniform_v2
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.3](https://huggingface.co/mistralai/Mistral-7B-v0.3) on the mlfoundations-dev/webinstruct_original_750k_uniform dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.3717
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 0.37169188261032104,
|
| 4 |
+
"eval_runtime": 23.4247,
|
| 5 |
+
"eval_samples_per_second": 286.577,
|
| 6 |
+
"eval_steps_per_second": 1.153,
|
| 7 |
+
"total_flos": 1256277934080000.0,
|
| 8 |
+
"train_loss": 0.3125666214625041,
|
| 9 |
+
"train_runtime": 4624.742,
|
| 10 |
+
"train_samples_per_second": 82.737,
|
| 11 |
+
"train_steps_per_second": 0.162
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 0.37169188261032104,
|
| 4 |
+
"eval_runtime": 23.4247,
|
| 5 |
+
"eval_samples_per_second": 286.577,
|
| 6 |
+
"eval_steps_per_second": 1.153
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1256277934080000.0,
|
| 4 |
+
"train_loss": 0.3125666214625041,
|
| 5 |
+
"train_runtime": 4624.742,
|
| 6 |
+
"train_samples_per_second": 82.737,
|
| 7 |
+
"train_steps_per_second": 0.162
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,591 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 750,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.04,
|
| 13 |
+
"grad_norm": 7.405676394065885,
|
| 14 |
+
"learning_rate": 1.3157894736842106e-06,
|
| 15 |
+
"loss": 0.562,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.08,
|
| 20 |
+
"grad_norm": 1.4808664279431412,
|
| 21 |
+
"learning_rate": 2.631578947368421e-06,
|
| 22 |
+
"loss": 0.4745,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.12,
|
| 27 |
+
"grad_norm": 2.1412237798795983,
|
| 28 |
+
"learning_rate": 3.947368421052632e-06,
|
| 29 |
+
"loss": 0.4401,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.16,
|
| 34 |
+
"grad_norm": 1.779775771320836,
|
| 35 |
+
"learning_rate": 4.9999123908526755e-06,
|
| 36 |
+
"loss": 0.4181,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.2,
|
| 41 |
+
"grad_norm": 1.8098388619283128,
|
| 42 |
+
"learning_rate": 4.996846787004002e-06,
|
| 43 |
+
"loss": 0.4075,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.24,
|
| 48 |
+
"grad_norm": 1.3956362585237865,
|
| 49 |
+
"learning_rate": 4.989407545945897e-06,
|
| 50 |
+
"loss": 0.3981,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.28,
|
| 55 |
+
"grad_norm": 1.3390011122743337,
|
| 56 |
+
"learning_rate": 4.977609148657522e-06,
|
| 57 |
+
"loss": 0.3953,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.32,
|
| 62 |
+
"grad_norm": 1.6165361734987194,
|
| 63 |
+
"learning_rate": 4.961474561506542e-06,
|
| 64 |
+
"loss": 0.3906,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.36,
|
| 69 |
+
"grad_norm": 1.0867017331160276,
|
| 70 |
+
"learning_rate": 4.941035191543551e-06,
|
| 71 |
+
"loss": 0.3896,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.4,
|
| 76 |
+
"grad_norm": 1.0086678350055638,
|
| 77 |
+
"learning_rate": 4.9163308253661615e-06,
|
| 78 |
+
"loss": 0.3831,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.44,
|
| 83 |
+
"grad_norm": 1.072846551894899,
|
| 84 |
+
"learning_rate": 4.88740955167173e-06,
|
| 85 |
+
"loss": 0.3765,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.48,
|
| 90 |
+
"grad_norm": 1.221763749203504,
|
| 91 |
+
"learning_rate": 4.854327667649502e-06,
|
| 92 |
+
"loss": 0.3775,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.52,
|
| 97 |
+
"grad_norm": 1.0092621110027236,
|
| 98 |
+
"learning_rate": 4.8171495693943766e-06,
|
| 99 |
+
"loss": 0.375,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.56,
|
| 104 |
+
"grad_norm": 1.208561990347492,
|
| 105 |
+
"learning_rate": 4.775947626555608e-06,
|
| 106 |
+
"loss": 0.3739,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.6,
|
| 111 |
+
"grad_norm": 1.176554376995382,
|
| 112 |
+
"learning_rate": 4.730802041464456e-06,
|
| 113 |
+
"loss": 0.3793,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.64,
|
| 118 |
+
"grad_norm": 1.1369349974571963,
|
| 119 |
+
"learning_rate": 4.681800693014995e-06,
|
| 120 |
+
"loss": 0.3725,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.68,
|
| 125 |
+
"grad_norm": 0.9231187737297902,
|
| 126 |
+
"learning_rate": 4.62903896560198e-06,
|
| 127 |
+
"loss": 0.3737,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.72,
|
| 132 |
+
"grad_norm": 1.0454964740112676,
|
| 133 |
+
"learning_rate": 4.572619563448765e-06,
|
| 134 |
+
"loss": 0.3691,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.76,
|
| 139 |
+
"grad_norm": 0.9928273265656697,
|
| 140 |
+
"learning_rate": 4.5126523106866685e-06,
|
| 141 |
+
"loss": 0.3683,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.8,
|
| 146 |
+
"grad_norm": 0.9575987290306286,
|
| 147 |
+
"learning_rate": 4.449253937574979e-06,
|
| 148 |
+
"loss": 0.3649,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.84,
|
| 153 |
+
"grad_norm": 0.787890419318379,
|
| 154 |
+
"learning_rate": 4.382547853277711e-06,
|
| 155 |
+
"loss": 0.3626,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.88,
|
| 160 |
+
"grad_norm": 0.8612414019861957,
|
| 161 |
+
"learning_rate": 4.312663905639441e-06,
|
| 162 |
+
"loss": 0.3694,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.92,
|
| 167 |
+
"grad_norm": 0.8262841994453005,
|
| 168 |
+
"learning_rate": 4.239738128427801e-06,
|
| 169 |
+
"loss": 0.3645,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.96,
|
| 174 |
+
"grad_norm": 0.8368718170451225,
|
| 175 |
+
"learning_rate": 4.16391247653468e-06,
|
| 176 |
+
"loss": 0.3594,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.0,
|
| 181 |
+
"grad_norm": 0.9215074005834266,
|
| 182 |
+
"learning_rate": 4.085334549651552e-06,
|
| 183 |
+
"loss": 0.3646,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.0,
|
| 188 |
+
"eval_loss": 0.35972926020622253,
|
| 189 |
+
"eval_runtime": 23.2519,
|
| 190 |
+
"eval_samples_per_second": 288.707,
|
| 191 |
+
"eval_steps_per_second": 1.161,
|
| 192 |
+
"step": 250
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.04,
|
| 196 |
+
"grad_norm": 1.165050524786145,
|
| 197 |
+
"learning_rate": 4.004157304956824e-06,
|
| 198 |
+
"loss": 0.3087,
|
| 199 |
+
"step": 260
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.08,
|
| 203 |
+
"grad_norm": 1.008240758670673,
|
| 204 |
+
"learning_rate": 3.920538759374485e-06,
|
| 205 |
+
"loss": 0.3054,
|
| 206 |
+
"step": 270
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.12,
|
| 210 |
+
"grad_norm": 1.0221095175436803,
|
| 211 |
+
"learning_rate": 3.834641681983617e-06,
|
| 212 |
+
"loss": 0.3052,
|
| 213 |
+
"step": 280
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.16,
|
| 217 |
+
"grad_norm": 1.0240271073473286,
|
| 218 |
+
"learning_rate": 3.7466332771775184e-06,
|
| 219 |
+
"loss": 0.3073,
|
| 220 |
+
"step": 290
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.2,
|
| 224 |
+
"grad_norm": 0.9412340740751259,
|
| 225 |
+
"learning_rate": 3.656684859189198e-06,
|
| 226 |
+
"loss": 0.304,
|
| 227 |
+
"step": 300
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.24,
|
| 231 |
+
"grad_norm": 1.0757903728934715,
|
| 232 |
+
"learning_rate": 3.564971518616786e-06,
|
| 233 |
+
"loss": 0.3074,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.28,
|
| 238 |
+
"grad_norm": 0.8788588031740697,
|
| 239 |
+
"learning_rate": 3.4716717815980004e-06,
|
| 240 |
+
"loss": 0.3025,
|
| 241 |
+
"step": 320
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.32,
|
| 245 |
+
"grad_norm": 0.8596918070645869,
|
| 246 |
+
"learning_rate": 3.376967262297108e-06,
|
| 247 |
+
"loss": 0.3033,
|
| 248 |
+
"step": 330
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.3599999999999999,
|
| 252 |
+
"grad_norm": 1.0313995367592261,
|
| 253 |
+
"learning_rate": 3.2810423093808307e-06,
|
| 254 |
+
"loss": 0.3079,
|
| 255 |
+
"step": 340
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.4,
|
| 259 |
+
"grad_norm": 0.8667828516508759,
|
| 260 |
+
"learning_rate": 3.184083647171368e-06,
|
| 261 |
+
"loss": 0.3055,
|
| 262 |
+
"step": 350
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.44,
|
| 266 |
+
"grad_norm": 0.8867289748518105,
|
| 267 |
+
"learning_rate": 3.08628001217504e-06,
|
| 268 |
+
"loss": 0.3089,
|
| 269 |
+
"step": 360
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.48,
|
| 273 |
+
"grad_norm": 0.9208383212214953,
|
| 274 |
+
"learning_rate": 2.9878217856940844e-06,
|
| 275 |
+
"loss": 0.3074,
|
| 276 |
+
"step": 370
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.52,
|
| 280 |
+
"grad_norm": 0.9978552238621926,
|
| 281 |
+
"learning_rate": 2.8889006232367455e-06,
|
| 282 |
+
"loss": 0.3034,
|
| 283 |
+
"step": 380
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 1.56,
|
| 287 |
+
"grad_norm": 0.8760456045025133,
|
| 288 |
+
"learning_rate": 2.7897090814470402e-06,
|
| 289 |
+
"loss": 0.3029,
|
| 290 |
+
"step": 390
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.6,
|
| 294 |
+
"grad_norm": 0.8636869853498937,
|
| 295 |
+
"learning_rate": 2.6904402432803968e-06,
|
| 296 |
+
"loss": 0.3035,
|
| 297 |
+
"step": 400
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.6400000000000001,
|
| 301 |
+
"grad_norm": 0.8906545485164509,
|
| 302 |
+
"learning_rate": 2.5912873421548e-06,
|
| 303 |
+
"loss": 0.3024,
|
| 304 |
+
"step": 410
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.6800000000000002,
|
| 308 |
+
"grad_norm": 0.8378263474774879,
|
| 309 |
+
"learning_rate": 2.4924433858090513e-06,
|
| 310 |
+
"loss": 0.3028,
|
| 311 |
+
"step": 420
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.72,
|
| 315 |
+
"grad_norm": 0.8807828593678831,
|
| 316 |
+
"learning_rate": 2.3941007806003163e-06,
|
| 317 |
+
"loss": 0.3029,
|
| 318 |
+
"step": 430
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.76,
|
| 322 |
+
"grad_norm": 0.866045737856776,
|
| 323 |
+
"learning_rate": 2.2964509569723175e-06,
|
| 324 |
+
"loss": 0.3039,
|
| 325 |
+
"step": 440
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 1.8,
|
| 329 |
+
"grad_norm": 0.8045213693263892,
|
| 330 |
+
"learning_rate": 2.1996839968232002e-06,
|
| 331 |
+
"loss": 0.2988,
|
| 332 |
+
"step": 450
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.8399999999999999,
|
| 336 |
+
"grad_norm": 0.8985482585858041,
|
| 337 |
+
"learning_rate": 2.1039882634984314e-06,
|
| 338 |
+
"loss": 0.3032,
|
| 339 |
+
"step": 460
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.88,
|
| 343 |
+
"grad_norm": 0.8026724494445288,
|
| 344 |
+
"learning_rate": 2.0095500351289776e-06,
|
| 345 |
+
"loss": 0.3033,
|
| 346 |
+
"step": 470
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.92,
|
| 350 |
+
"grad_norm": 0.8098584424025912,
|
| 351 |
+
"learning_rate": 1.9165531420284942e-06,
|
| 352 |
+
"loss": 0.299,
|
| 353 |
+
"step": 480
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.96,
|
| 357 |
+
"grad_norm": 0.8418880488102612,
|
| 358 |
+
"learning_rate": 1.825178608855351e-06,
|
| 359 |
+
"loss": 0.3018,
|
| 360 |
+
"step": 490
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 2.0,
|
| 364 |
+
"grad_norm": 1.2621791449191886,
|
| 365 |
+
"learning_rate": 1.7356043022360534e-06,
|
| 366 |
+
"loss": 0.2994,
|
| 367 |
+
"step": 500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 2.0,
|
| 371 |
+
"eval_loss": 0.3540443778038025,
|
| 372 |
+
"eval_runtime": 23.427,
|
| 373 |
+
"eval_samples_per_second": 286.549,
|
| 374 |
+
"eval_steps_per_second": 1.153,
|
| 375 |
+
"step": 500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.04,
|
| 379 |
+
"grad_norm": 1.1125408865085207,
|
| 380 |
+
"learning_rate": 1.6480045845359866e-06,
|
| 381 |
+
"loss": 0.2452,
|
| 382 |
+
"step": 510
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 2.08,
|
| 386 |
+
"grad_norm": 0.9550266586265046,
|
| 387 |
+
"learning_rate": 1.5625499744514327e-06,
|
| 388 |
+
"loss": 0.2392,
|
| 389 |
+
"step": 520
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.12,
|
| 393 |
+
"grad_norm": 0.9636896936192534,
|
| 394 |
+
"learning_rate": 1.4794068150835547e-06,
|
| 395 |
+
"loss": 0.2441,
|
| 396 |
+
"step": 530
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 2.16,
|
| 400 |
+
"grad_norm": 0.990740891225938,
|
| 401 |
+
"learning_rate": 1.3987369501404421e-06,
|
| 402 |
+
"loss": 0.2426,
|
| 403 |
+
"step": 540
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 2.2,
|
| 407 |
+
"grad_norm": 0.9541502720129523,
|
| 408 |
+
"learning_rate": 1.3206974088975384e-06,
|
| 409 |
+
"loss": 0.2417,
|
| 410 |
+
"step": 550
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.24,
|
| 414 |
+
"grad_norm": 0.9287304371229521,
|
| 415 |
+
"learning_rate": 1.245440100529675e-06,
|
| 416 |
+
"loss": 0.2441,
|
| 417 |
+
"step": 560
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 2.2800000000000002,
|
| 421 |
+
"grad_norm": 0.962119628291641,
|
| 422 |
+
"learning_rate": 1.1731115184097258e-06,
|
| 423 |
+
"loss": 0.2419,
|
| 424 |
+
"step": 570
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 2.32,
|
| 428 |
+
"grad_norm": 0.9741055322033145,
|
| 429 |
+
"learning_rate": 1.1038524549494814e-06,
|
| 430 |
+
"loss": 0.2426,
|
| 431 |
+
"step": 580
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 2.36,
|
| 435 |
+
"grad_norm": 0.979048696685581,
|
| 436 |
+
"learning_rate": 1.0377977275378227e-06,
|
| 437 |
+
"loss": 0.237,
|
| 438 |
+
"step": 590
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 2.4,
|
| 442 |
+
"grad_norm": 0.9374109657605778,
|
| 443 |
+
"learning_rate": 9.75075916109678e-07,
|
| 444 |
+
"loss": 0.239,
|
| 445 |
+
"step": 600
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 2.44,
|
| 449 |
+
"grad_norm": 0.9213868496179243,
|
| 450 |
+
"learning_rate": 9.158091128565985e-07,
|
| 451 |
+
"loss": 0.2414,
|
| 452 |
+
"step": 610
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.48,
|
| 456 |
+
"grad_norm": 0.9362610185478109,
|
| 457 |
+
"learning_rate": 8.601126845661592e-07,
|
| 458 |
+
"loss": 0.2435,
|
| 459 |
+
"step": 620
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.52,
|
| 463 |
+
"grad_norm": 0.9594230633417852,
|
| 464 |
+
"learning_rate": 8.080950480528108e-07,
|
| 465 |
+
"loss": 0.2411,
|
| 466 |
+
"step": 630
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 2.56,
|
| 470 |
+
"grad_norm": 0.9587708118572509,
|
| 471 |
+
"learning_rate": 7.598574591173123e-07,
|
| 472 |
+
"loss": 0.2389,
|
| 473 |
+
"step": 640
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 2.6,
|
| 477 |
+
"grad_norm": 0.9632573104844587,
|
| 478 |
+
"learning_rate": 7.154938154455593e-07,
|
| 479 |
+
"loss": 0.2407,
|
| 480 |
+
"step": 650
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 2.64,
|
| 484 |
+
"grad_norm": 0.9937798751831849,
|
| 485 |
+
"learning_rate": 6.750904738304712e-07,
|
| 486 |
+
"loss": 0.2444,
|
| 487 |
+
"step": 660
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 2.68,
|
| 491 |
+
"grad_norm": 0.9538408777366829,
|
| 492 |
+
"learning_rate": 6.387260820727336e-07,
|
| 493 |
+
"loss": 0.2442,
|
| 494 |
+
"step": 670
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.7199999999999998,
|
| 498 |
+
"grad_norm": 0.9865719301819308,
|
| 499 |
+
"learning_rate": 6.064714258876026e-07,
|
| 500 |
+
"loss": 0.244,
|
| 501 |
+
"step": 680
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 2.76,
|
| 505 |
+
"grad_norm": 0.9447856849767184,
|
| 506 |
+
"learning_rate": 5.783892911157884e-07,
|
| 507 |
+
"loss": 0.2379,
|
| 508 |
+
"step": 690
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.8,
|
| 512 |
+
"grad_norm": 0.9255210583910624,
|
| 513 |
+
"learning_rate": 5.545343415066254e-07,
|
| 514 |
+
"loss": 0.2388,
|
| 515 |
+
"step": 700
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.84,
|
| 519 |
+
"grad_norm": 0.92833568809518,
|
| 520 |
+
"learning_rate": 5.349530123114375e-07,
|
| 521 |
+
"loss": 0.244,
|
| 522 |
+
"step": 710
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 2.88,
|
| 526 |
+
"grad_norm": 0.986492057017101,
|
| 527 |
+
"learning_rate": 5.196834198942211e-07,
|
| 528 |
+
"loss": 0.2396,
|
| 529 |
+
"step": 720
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 2.92,
|
| 533 |
+
"grad_norm": 1.0041806350073348,
|
| 534 |
+
"learning_rate": 5.087552875355995e-07,
|
| 535 |
+
"loss": 0.2399,
|
| 536 |
+
"step": 730
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.96,
|
| 540 |
+
"grad_norm": 0.9294612829305497,
|
| 541 |
+
"learning_rate": 5.021898875744752e-07,
|
| 542 |
+
"loss": 0.2402,
|
| 543 |
+
"step": 740
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 3.0,
|
| 547 |
+
"grad_norm": 1.1188607975055915,
|
| 548 |
+
"learning_rate": 5e-07,
|
| 549 |
+
"loss": 0.2354,
|
| 550 |
+
"step": 750
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 3.0,
|
| 554 |
+
"eval_loss": 0.37169188261032104,
|
| 555 |
+
"eval_runtime": 23.4243,
|
| 556 |
+
"eval_samples_per_second": 286.583,
|
| 557 |
+
"eval_steps_per_second": 1.153,
|
| 558 |
+
"step": 750
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 3.0,
|
| 562 |
+
"step": 750,
|
| 563 |
+
"total_flos": 1256277934080000.0,
|
| 564 |
+
"train_loss": 0.3125666214625041,
|
| 565 |
+
"train_runtime": 4624.742,
|
| 566 |
+
"train_samples_per_second": 82.737,
|
| 567 |
+
"train_steps_per_second": 0.162
|
| 568 |
+
}
|
| 569 |
+
],
|
| 570 |
+
"logging_steps": 10,
|
| 571 |
+
"max_steps": 750,
|
| 572 |
+
"num_input_tokens_seen": 0,
|
| 573 |
+
"num_train_epochs": 3,
|
| 574 |
+
"save_steps": 500,
|
| 575 |
+
"stateful_callbacks": {
|
| 576 |
+
"TrainerControl": {
|
| 577 |
+
"args": {
|
| 578 |
+
"should_epoch_stop": false,
|
| 579 |
+
"should_evaluate": false,
|
| 580 |
+
"should_log": false,
|
| 581 |
+
"should_save": true,
|
| 582 |
+
"should_training_stop": true
|
| 583 |
+
},
|
| 584 |
+
"attributes": {}
|
| 585 |
+
}
|
| 586 |
+
},
|
| 587 |
+
"total_flos": 1256277934080000.0,
|
| 588 |
+
"train_batch_size": 16,
|
| 589 |
+
"trial_name": null,
|
| 590 |
+
"trial_params": null
|
| 591 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|