Commit
•
9896419
1
Parent(s):
d1f9513
Training in progress, step 100900, checkpoint
Browse files
last-checkpoint/adapter_config.json
CHANGED
|
@@ -20,13 +20,13 @@
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
-
"
|
| 24 |
"k_proj",
|
| 25 |
"gate_proj",
|
| 26 |
"down_proj",
|
| 27 |
-
"
|
| 28 |
-
"
|
| 29 |
-
"
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"use_dora": false,
|
|
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
+
"v_proj",
|
| 24 |
"k_proj",
|
| 25 |
"gate_proj",
|
| 26 |
"down_proj",
|
| 27 |
+
"q_proj",
|
| 28 |
+
"up_proj",
|
| 29 |
+
"o_proj"
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"use_dora": false,
|
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5544997664
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ff50e9a0eef14c00f32c5e550257295427f2d666e009aac32472aef43b0c78f
|
| 3 |
size 5544997664
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 674093138
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4b1f3c8b9d1c8514057a760685418307853b2172387d395e371d81516debcc99
|
| 3 |
size 674093138
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3d52e9778ae961a843d4efe5adba669832146332ec663eac9df46d71427724e3
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b1dd725a3e5295711459643d6e1204a1d04a7f905cc6416544fa87ecdfb18228
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 200,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -6657,6 +6657,419 @@
|
|
| 6657 |
"learning_rate": 1.9990150014305462e-05,
|
| 6658 |
"loss": 1.5194,
|
| 6659 |
"step": 95000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6660 |
}
|
| 6661 |
],
|
| 6662 |
"logging_steps": 100,
|
|
@@ -6676,7 +7089,7 @@
|
|
| 6676 |
"attributes": {}
|
| 6677 |
}
|
| 6678 |
},
|
| 6679 |
-
"total_flos": 1.
|
| 6680 |
"train_batch_size": 1,
|
| 6681 |
"trial_name": null,
|
| 6682 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.07503755596210195,
|
| 5 |
"eval_steps": 200,
|
| 6 |
+
"global_step": 100900,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 6657 |
"learning_rate": 1.9990150014305462e-05,
|
| 6658 |
"loss": 1.5194,
|
| 6659 |
"step": 95000
|
| 6660 |
+
},
|
| 6661 |
+
{
|
| 6662 |
+
"epoch": 0.07072419793851234,
|
| 6663 |
+
"grad_norm": 0.7444689273834229,
|
| 6664 |
+
"learning_rate": 1.9990129268900848e-05,
|
| 6665 |
+
"loss": 1.5198,
|
| 6666 |
+
"step": 95100
|
| 6667 |
+
},
|
| 6668 |
+
{
|
| 6669 |
+
"epoch": 0.07079856618029837,
|
| 6670 |
+
"grad_norm": 0.9299377202987671,
|
| 6671 |
+
"learning_rate": 1.9990108501683685e-05,
|
| 6672 |
+
"loss": 1.5393,
|
| 6673 |
+
"step": 95200
|
| 6674 |
+
},
|
| 6675 |
+
{
|
| 6676 |
+
"epoch": 0.0708729344220844,
|
| 6677 |
+
"grad_norm": 0.6611402630805969,
|
| 6678 |
+
"learning_rate": 1.999008771265401e-05,
|
| 6679 |
+
"loss": 1.5351,
|
| 6680 |
+
"step": 95300
|
| 6681 |
+
},
|
| 6682 |
+
{
|
| 6683 |
+
"epoch": 0.07094730266387042,
|
| 6684 |
+
"grad_norm": 0.4772530496120453,
|
| 6685 |
+
"learning_rate": 1.9990066901811876e-05,
|
| 6686 |
+
"loss": 1.5243,
|
| 6687 |
+
"step": 95400
|
| 6688 |
+
},
|
| 6689 |
+
{
|
| 6690 |
+
"epoch": 0.07102167090565645,
|
| 6691 |
+
"grad_norm": 0.42998188734054565,
|
| 6692 |
+
"learning_rate": 1.9990046069157322e-05,
|
| 6693 |
+
"loss": 1.5877,
|
| 6694 |
+
"step": 95500
|
| 6695 |
+
},
|
| 6696 |
+
{
|
| 6697 |
+
"epoch": 0.07109603914744247,
|
| 6698 |
+
"grad_norm": 0.7415347099304199,
|
| 6699 |
+
"learning_rate": 1.9990025214690396e-05,
|
| 6700 |
+
"loss": 1.5633,
|
| 6701 |
+
"step": 95600
|
| 6702 |
+
},
|
| 6703 |
+
{
|
| 6704 |
+
"epoch": 0.0711704073892285,
|
| 6705 |
+
"grad_norm": 0.657112717628479,
|
| 6706 |
+
"learning_rate": 1.999000433841114e-05,
|
| 6707 |
+
"loss": 1.4555,
|
| 6708 |
+
"step": 95700
|
| 6709 |
+
},
|
| 6710 |
+
{
|
| 6711 |
+
"epoch": 0.07124477563101453,
|
| 6712 |
+
"grad_norm": 0.9188429713249207,
|
| 6713 |
+
"learning_rate": 1.998998344031961e-05,
|
| 6714 |
+
"loss": 1.4329,
|
| 6715 |
+
"step": 95800
|
| 6716 |
+
},
|
| 6717 |
+
{
|
| 6718 |
+
"epoch": 0.07131914387280056,
|
| 6719 |
+
"grad_norm": 0.8823667168617249,
|
| 6720 |
+
"learning_rate": 1.9989962520415836e-05,
|
| 6721 |
+
"loss": 1.4754,
|
| 6722 |
+
"step": 95900
|
| 6723 |
+
},
|
| 6724 |
+
{
|
| 6725 |
+
"epoch": 0.07139351211458658,
|
| 6726 |
+
"grad_norm": 0.7276200652122498,
|
| 6727 |
+
"learning_rate": 1.9989941578699878e-05,
|
| 6728 |
+
"loss": 1.5286,
|
| 6729 |
+
"step": 96000
|
| 6730 |
+
},
|
| 6731 |
+
{
|
| 6732 |
+
"epoch": 0.07146788035637261,
|
| 6733 |
+
"grad_norm": 0.941512405872345,
|
| 6734 |
+
"learning_rate": 1.998992061517177e-05,
|
| 6735 |
+
"loss": 1.5087,
|
| 6736 |
+
"step": 96100
|
| 6737 |
+
},
|
| 6738 |
+
{
|
| 6739 |
+
"epoch": 0.07154224859815864,
|
| 6740 |
+
"grad_norm": 1.0310442447662354,
|
| 6741 |
+
"learning_rate": 1.998989962983157e-05,
|
| 6742 |
+
"loss": 1.5895,
|
| 6743 |
+
"step": 96200
|
| 6744 |
+
},
|
| 6745 |
+
{
|
| 6746 |
+
"epoch": 0.07161661683994466,
|
| 6747 |
+
"grad_norm": 1.3620883226394653,
|
| 6748 |
+
"learning_rate": 1.9989878622679317e-05,
|
| 6749 |
+
"loss": 1.474,
|
| 6750 |
+
"step": 96300
|
| 6751 |
+
},
|
| 6752 |
+
{
|
| 6753 |
+
"epoch": 0.0716909850817307,
|
| 6754 |
+
"grad_norm": 0.5119801163673401,
|
| 6755 |
+
"learning_rate": 1.998985759371505e-05,
|
| 6756 |
+
"loss": 1.5112,
|
| 6757 |
+
"step": 96400
|
| 6758 |
+
},
|
| 6759 |
+
{
|
| 6760 |
+
"epoch": 0.07176535332351673,
|
| 6761 |
+
"grad_norm": 0.8966123461723328,
|
| 6762 |
+
"learning_rate": 1.998983654293883e-05,
|
| 6763 |
+
"loss": 1.4903,
|
| 6764 |
+
"step": 96500
|
| 6765 |
+
},
|
| 6766 |
+
{
|
| 6767 |
+
"epoch": 0.07183972156530276,
|
| 6768 |
+
"grad_norm": 0.5336944460868835,
|
| 6769 |
+
"learning_rate": 1.998981547035069e-05,
|
| 6770 |
+
"loss": 1.5673,
|
| 6771 |
+
"step": 96600
|
| 6772 |
+
},
|
| 6773 |
+
{
|
| 6774 |
+
"epoch": 0.07191408980708879,
|
| 6775 |
+
"grad_norm": 1.2533961534500122,
|
| 6776 |
+
"learning_rate": 1.9989794375950688e-05,
|
| 6777 |
+
"loss": 1.5039,
|
| 6778 |
+
"step": 96700
|
| 6779 |
+
},
|
| 6780 |
+
{
|
| 6781 |
+
"epoch": 0.07198845804887481,
|
| 6782 |
+
"grad_norm": 1.3317081928253174,
|
| 6783 |
+
"learning_rate": 1.9989773259738858e-05,
|
| 6784 |
+
"loss": 1.567,
|
| 6785 |
+
"step": 96800
|
| 6786 |
+
},
|
| 6787 |
+
{
|
| 6788 |
+
"epoch": 0.07206282629066084,
|
| 6789 |
+
"grad_norm": 0.49700722098350525,
|
| 6790 |
+
"learning_rate": 1.998975212171525e-05,
|
| 6791 |
+
"loss": 1.542,
|
| 6792 |
+
"step": 96900
|
| 6793 |
+
},
|
| 6794 |
+
{
|
| 6795 |
+
"epoch": 0.07213719453244687,
|
| 6796 |
+
"grad_norm": 0.5809246301651001,
|
| 6797 |
+
"learning_rate": 1.9989730961879913e-05,
|
| 6798 |
+
"loss": 1.5097,
|
| 6799 |
+
"step": 97000
|
| 6800 |
+
},
|
| 6801 |
+
{
|
| 6802 |
+
"epoch": 0.07221156277423289,
|
| 6803 |
+
"grad_norm": 0.6107625365257263,
|
| 6804 |
+
"learning_rate": 1.9989709780232894e-05,
|
| 6805 |
+
"loss": 1.536,
|
| 6806 |
+
"step": 97100
|
| 6807 |
+
},
|
| 6808 |
+
{
|
| 6809 |
+
"epoch": 0.07228593101601892,
|
| 6810 |
+
"grad_norm": 0.5271338820457458,
|
| 6811 |
+
"learning_rate": 1.9989688576774234e-05,
|
| 6812 |
+
"loss": 1.5819,
|
| 6813 |
+
"step": 97200
|
| 6814 |
+
},
|
| 6815 |
+
{
|
| 6816 |
+
"epoch": 0.07236029925780495,
|
| 6817 |
+
"grad_norm": 0.6692411303520203,
|
| 6818 |
+
"learning_rate": 1.9989667351503988e-05,
|
| 6819 |
+
"loss": 1.4833,
|
| 6820 |
+
"step": 97300
|
| 6821 |
+
},
|
| 6822 |
+
{
|
| 6823 |
+
"epoch": 0.07243466749959097,
|
| 6824 |
+
"grad_norm": 1.0627728700637817,
|
| 6825 |
+
"learning_rate": 1.998964610442219e-05,
|
| 6826 |
+
"loss": 1.5404,
|
| 6827 |
+
"step": 97400
|
| 6828 |
+
},
|
| 6829 |
+
{
|
| 6830 |
+
"epoch": 0.072509035741377,
|
| 6831 |
+
"grad_norm": 0.5696298480033875,
|
| 6832 |
+
"learning_rate": 1.9989624835528896e-05,
|
| 6833 |
+
"loss": 1.4491,
|
| 6834 |
+
"step": 97500
|
| 6835 |
+
},
|
| 6836 |
+
{
|
| 6837 |
+
"epoch": 0.07258340398316303,
|
| 6838 |
+
"grad_norm": 0.5105301141738892,
|
| 6839 |
+
"learning_rate": 1.998960354482415e-05,
|
| 6840 |
+
"loss": 1.5188,
|
| 6841 |
+
"step": 97600
|
| 6842 |
+
},
|
| 6843 |
+
{
|
| 6844 |
+
"epoch": 0.07265777222494905,
|
| 6845 |
+
"grad_norm": 0.53251713514328,
|
| 6846 |
+
"learning_rate": 1.9989582232307998e-05,
|
| 6847 |
+
"loss": 1.5367,
|
| 6848 |
+
"step": 97700
|
| 6849 |
+
},
|
| 6850 |
+
{
|
| 6851 |
+
"epoch": 0.07273214046673508,
|
| 6852 |
+
"grad_norm": 0.6559078693389893,
|
| 6853 |
+
"learning_rate": 1.9989560897980485e-05,
|
| 6854 |
+
"loss": 1.4773,
|
| 6855 |
+
"step": 97800
|
| 6856 |
+
},
|
| 6857 |
+
{
|
| 6858 |
+
"epoch": 0.07280650870852111,
|
| 6859 |
+
"grad_norm": 0.39833974838256836,
|
| 6860 |
+
"learning_rate": 1.998953954184166e-05,
|
| 6861 |
+
"loss": 1.6063,
|
| 6862 |
+
"step": 97900
|
| 6863 |
+
},
|
| 6864 |
+
{
|
| 6865 |
+
"epoch": 0.07288087695030714,
|
| 6866 |
+
"grad_norm": 1.0479645729064941,
|
| 6867 |
+
"learning_rate": 1.9989518163891566e-05,
|
| 6868 |
+
"loss": 1.565,
|
| 6869 |
+
"step": 98000
|
| 6870 |
+
},
|
| 6871 |
+
{
|
| 6872 |
+
"epoch": 0.07295524519209316,
|
| 6873 |
+
"grad_norm": 0.7905478477478027,
|
| 6874 |
+
"learning_rate": 1.9989496764130253e-05,
|
| 6875 |
+
"loss": 1.5266,
|
| 6876 |
+
"step": 98100
|
| 6877 |
+
},
|
| 6878 |
+
{
|
| 6879 |
+
"epoch": 0.07302961343387919,
|
| 6880 |
+
"grad_norm": 0.4569951295852661,
|
| 6881 |
+
"learning_rate": 1.998947534255777e-05,
|
| 6882 |
+
"loss": 1.5295,
|
| 6883 |
+
"step": 98200
|
| 6884 |
+
},
|
| 6885 |
+
{
|
| 6886 |
+
"epoch": 0.07310398167566523,
|
| 6887 |
+
"grad_norm": 0.5308849215507507,
|
| 6888 |
+
"learning_rate": 1.9989453899174158e-05,
|
| 6889 |
+
"loss": 1.5203,
|
| 6890 |
+
"step": 98300
|
| 6891 |
+
},
|
| 6892 |
+
{
|
| 6893 |
+
"epoch": 0.07317834991745126,
|
| 6894 |
+
"grad_norm": 0.906802773475647,
|
| 6895 |
+
"learning_rate": 1.998943243397947e-05,
|
| 6896 |
+
"loss": 1.556,
|
| 6897 |
+
"step": 98400
|
| 6898 |
+
},
|
| 6899 |
+
{
|
| 6900 |
+
"epoch": 0.07325271815923728,
|
| 6901 |
+
"grad_norm": 0.5071494579315186,
|
| 6902 |
+
"learning_rate": 1.9989410946973747e-05,
|
| 6903 |
+
"loss": 1.5627,
|
| 6904 |
+
"step": 98500
|
| 6905 |
+
},
|
| 6906 |
+
{
|
| 6907 |
+
"epoch": 0.07332708640102331,
|
| 6908 |
+
"grad_norm": 0.5252199172973633,
|
| 6909 |
+
"learning_rate": 1.9989389438157037e-05,
|
| 6910 |
+
"loss": 1.5181,
|
| 6911 |
+
"step": 98600
|
| 6912 |
+
},
|
| 6913 |
+
{
|
| 6914 |
+
"epoch": 0.07340145464280934,
|
| 6915 |
+
"grad_norm": 0.5738980174064636,
|
| 6916 |
+
"learning_rate": 1.9989367907529394e-05,
|
| 6917 |
+
"loss": 1.6101,
|
| 6918 |
+
"step": 98700
|
| 6919 |
+
},
|
| 6920 |
+
{
|
| 6921 |
+
"epoch": 0.07347582288459537,
|
| 6922 |
+
"grad_norm": 0.6898683309555054,
|
| 6923 |
+
"learning_rate": 1.9989346355090853e-05,
|
| 6924 |
+
"loss": 1.579,
|
| 6925 |
+
"step": 98800
|
| 6926 |
+
},
|
| 6927 |
+
{
|
| 6928 |
+
"epoch": 0.07355019112638139,
|
| 6929 |
+
"grad_norm": 0.5396860241889954,
|
| 6930 |
+
"learning_rate": 1.998932478084147e-05,
|
| 6931 |
+
"loss": 1.5645,
|
| 6932 |
+
"step": 98900
|
| 6933 |
+
},
|
| 6934 |
+
{
|
| 6935 |
+
"epoch": 0.07362455936816742,
|
| 6936 |
+
"grad_norm": 0.5482293367385864,
|
| 6937 |
+
"learning_rate": 1.998930318478129e-05,
|
| 6938 |
+
"loss": 1.5453,
|
| 6939 |
+
"step": 99000
|
| 6940 |
+
},
|
| 6941 |
+
{
|
| 6942 |
+
"epoch": 0.07369892760995345,
|
| 6943 |
+
"grad_norm": 0.8394240736961365,
|
| 6944 |
+
"learning_rate": 1.9989281566910363e-05,
|
| 6945 |
+
"loss": 1.5025,
|
| 6946 |
+
"step": 99100
|
| 6947 |
+
},
|
| 6948 |
+
{
|
| 6949 |
+
"epoch": 0.07377329585173947,
|
| 6950 |
+
"grad_norm": 0.9409950971603394,
|
| 6951 |
+
"learning_rate": 1.9989259927228725e-05,
|
| 6952 |
+
"loss": 1.5489,
|
| 6953 |
+
"step": 99200
|
| 6954 |
+
},
|
| 6955 |
+
{
|
| 6956 |
+
"epoch": 0.0738476640935255,
|
| 6957 |
+
"grad_norm": 0.5597321391105652,
|
| 6958 |
+
"learning_rate": 1.9989238265736437e-05,
|
| 6959 |
+
"loss": 1.5994,
|
| 6960 |
+
"step": 99300
|
| 6961 |
+
},
|
| 6962 |
+
{
|
| 6963 |
+
"epoch": 0.07392203233531153,
|
| 6964 |
+
"grad_norm": 0.5139235258102417,
|
| 6965 |
+
"learning_rate": 1.9989216582433538e-05,
|
| 6966 |
+
"loss": 1.5478,
|
| 6967 |
+
"step": 99400
|
| 6968 |
+
},
|
| 6969 |
+
{
|
| 6970 |
+
"epoch": 0.07399640057709755,
|
| 6971 |
+
"grad_norm": 0.6312362551689148,
|
| 6972 |
+
"learning_rate": 1.998919487732008e-05,
|
| 6973 |
+
"loss": 1.4989,
|
| 6974 |
+
"step": 99500
|
| 6975 |
+
},
|
| 6976 |
+
{
|
| 6977 |
+
"epoch": 0.07407076881888358,
|
| 6978 |
+
"grad_norm": 0.6924223303794861,
|
| 6979 |
+
"learning_rate": 1.9989173150396105e-05,
|
| 6980 |
+
"loss": 1.4491,
|
| 6981 |
+
"step": 99600
|
| 6982 |
+
},
|
| 6983 |
+
{
|
| 6984 |
+
"epoch": 0.07414513706066961,
|
| 6985 |
+
"grad_norm": 0.5490585565567017,
|
| 6986 |
+
"learning_rate": 1.9989151401661666e-05,
|
| 6987 |
+
"loss": 1.538,
|
| 6988 |
+
"step": 99700
|
| 6989 |
+
},
|
| 6990 |
+
{
|
| 6991 |
+
"epoch": 0.07421950530245564,
|
| 6992 |
+
"grad_norm": 0.630455732345581,
|
| 6993 |
+
"learning_rate": 1.998912963111681e-05,
|
| 6994 |
+
"loss": 1.5286,
|
| 6995 |
+
"step": 99800
|
| 6996 |
+
},
|
| 6997 |
+
{
|
| 6998 |
+
"epoch": 0.07429387354424166,
|
| 6999 |
+
"grad_norm": 0.8591504693031311,
|
| 7000 |
+
"learning_rate": 1.998910783876158e-05,
|
| 7001 |
+
"loss": 1.5612,
|
| 7002 |
+
"step": 99900
|
| 7003 |
+
},
|
| 7004 |
+
{
|
| 7005 |
+
"epoch": 0.07436824178602769,
|
| 7006 |
+
"grad_norm": 1.0016669034957886,
|
| 7007 |
+
"learning_rate": 1.9989086024596027e-05,
|
| 7008 |
+
"loss": 1.5154,
|
| 7009 |
+
"step": 100000
|
| 7010 |
+
},
|
| 7011 |
+
{
|
| 7012 |
+
"epoch": 0.07444261002781372,
|
| 7013 |
+
"grad_norm": 0.6513885259628296,
|
| 7014 |
+
"learning_rate": 1.9989064188620197e-05,
|
| 7015 |
+
"loss": 1.5446,
|
| 7016 |
+
"step": 100100
|
| 7017 |
+
},
|
| 7018 |
+
{
|
| 7019 |
+
"epoch": 0.07451697826959976,
|
| 7020 |
+
"grad_norm": 0.6838514804840088,
|
| 7021 |
+
"learning_rate": 1.998904233083414e-05,
|
| 7022 |
+
"loss": 1.5336,
|
| 7023 |
+
"step": 100200
|
| 7024 |
+
},
|
| 7025 |
+
{
|
| 7026 |
+
"epoch": 0.07459134651138578,
|
| 7027 |
+
"grad_norm": 0.46571242809295654,
|
| 7028 |
+
"learning_rate": 1.9989020451237903e-05,
|
| 7029 |
+
"loss": 1.4838,
|
| 7030 |
+
"step": 100300
|
| 7031 |
+
},
|
| 7032 |
+
{
|
| 7033 |
+
"epoch": 0.07466571475317181,
|
| 7034 |
+
"grad_norm": 0.9936356544494629,
|
| 7035 |
+
"learning_rate": 1.998899854983153e-05,
|
| 7036 |
+
"loss": 1.5929,
|
| 7037 |
+
"step": 100400
|
| 7038 |
+
},
|
| 7039 |
+
{
|
| 7040 |
+
"epoch": 0.07474008299495784,
|
| 7041 |
+
"grad_norm": 0.6591018438339233,
|
| 7042 |
+
"learning_rate": 1.9988976626615075e-05,
|
| 7043 |
+
"loss": 1.54,
|
| 7044 |
+
"step": 100500
|
| 7045 |
+
},
|
| 7046 |
+
{
|
| 7047 |
+
"epoch": 0.07481445123674386,
|
| 7048 |
+
"grad_norm": 0.8453909754753113,
|
| 7049 |
+
"learning_rate": 1.998895468158858e-05,
|
| 7050 |
+
"loss": 1.5191,
|
| 7051 |
+
"step": 100600
|
| 7052 |
+
},
|
| 7053 |
+
{
|
| 7054 |
+
"epoch": 0.07488881947852989,
|
| 7055 |
+
"grad_norm": 0.6555935144424438,
|
| 7056 |
+
"learning_rate": 1.9988932714752095e-05,
|
| 7057 |
+
"loss": 1.5734,
|
| 7058 |
+
"step": 100700
|
| 7059 |
+
},
|
| 7060 |
+
{
|
| 7061 |
+
"epoch": 0.07496318772031592,
|
| 7062 |
+
"grad_norm": 0.6445733308792114,
|
| 7063 |
+
"learning_rate": 1.998891072610567e-05,
|
| 7064 |
+
"loss": 1.5516,
|
| 7065 |
+
"step": 100800
|
| 7066 |
+
},
|
| 7067 |
+
{
|
| 7068 |
+
"epoch": 0.07503755596210195,
|
| 7069 |
+
"grad_norm": 0.534389078617096,
|
| 7070 |
+
"learning_rate": 1.9988888715649357e-05,
|
| 7071 |
+
"loss": 1.5441,
|
| 7072 |
+
"step": 100900
|
| 7073 |
}
|
| 7074 |
],
|
| 7075 |
"logging_steps": 100,
|
|
|
|
| 7089 |
"attributes": {}
|
| 7090 |
}
|
| 7091 |
},
|
| 7092 |
+
"total_flos": 1.3747108667853128e+18,
|
| 7093 |
"train_batch_size": 1,
|
| 7094 |
"trial_name": null,
|
| 7095 |
"trial_params": null
|