Training in progress, step 2100, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 847599616
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0f95d257a75a0818c3b281479bc5fe7f314c5f2fa671c8fdc9cffd9baa2e332c
|
| 3 |
size 847599616
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 350603
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45922bcb488761f8b4d89b01e053e03390d7dfff01b2448374ff5b9f516bf29d
|
| 3 |
size 350603
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8af8b3a1a646a87663f3ba47feec659ef3c74660d21434727f5930534bbd3ce4
|
| 3 |
size 14645
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a9c87bf073c670e4a83c28ad54cbc759acbb83ba035f81c4710f8a0ea2d39a0b
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 150,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -2636,6 +2636,444 @@
|
|
| 2636 |
"eval_samples_per_second": 9.684,
|
| 2637 |
"eval_steps_per_second": 1.646,
|
| 2638 |
"step": 1800
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2639 |
}
|
| 2640 |
],
|
| 2641 |
"logging_steps": 5,
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.42,
|
| 6 |
"eval_steps": 150,
|
| 7 |
+
"global_step": 2100,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 2636 |
"eval_samples_per_second": 9.684,
|
| 2637 |
"eval_steps_per_second": 1.646,
|
| 2638 |
"step": 1800
|
| 2639 |
+
},
|
| 2640 |
+
{
|
| 2641 |
+
"epoch": 0.361,
|
| 2642 |
+
"grad_norm": 0.06729965656995773,
|
| 2643 |
+
"learning_rate": 3.033412723959422e-06,
|
| 2644 |
+
"loss": 6.357138442993164,
|
| 2645 |
+
"step": 1805
|
| 2646 |
+
},
|
| 2647 |
+
{
|
| 2648 |
+
"epoch": 0.362,
|
| 2649 |
+
"grad_norm": 0.12906977534294128,
|
| 2650 |
+
"learning_rate": 3.027744527057284e-06,
|
| 2651 |
+
"loss": 6.433673858642578,
|
| 2652 |
+
"step": 1810
|
| 2653 |
+
},
|
| 2654 |
+
{
|
| 2655 |
+
"epoch": 0.363,
|
| 2656 |
+
"grad_norm": 0.07005853205919266,
|
| 2657 |
+
"learning_rate": 3.0220650909056574e-06,
|
| 2658 |
+
"loss": 6.411043548583985,
|
| 2659 |
+
"step": 1815
|
| 2660 |
+
},
|
| 2661 |
+
{
|
| 2662 |
+
"epoch": 0.364,
|
| 2663 |
+
"grad_norm": 0.07833214104175568,
|
| 2664 |
+
"learning_rate": 3.0163744776139457e-06,
|
| 2665 |
+
"loss": 6.432377624511719,
|
| 2666 |
+
"step": 1820
|
| 2667 |
+
},
|
| 2668 |
+
{
|
| 2669 |
+
"epoch": 0.365,
|
| 2670 |
+
"grad_norm": 0.08222617954015732,
|
| 2671 |
+
"learning_rate": 3.010672749413785e-06,
|
| 2672 |
+
"loss": 6.4166709899902346,
|
| 2673 |
+
"step": 1825
|
| 2674 |
+
},
|
| 2675 |
+
{
|
| 2676 |
+
"epoch": 0.366,
|
| 2677 |
+
"grad_norm": 0.07129046320915222,
|
| 2678 |
+
"learning_rate": 3.00495996865836e-06,
|
| 2679 |
+
"loss": 6.371786880493164,
|
| 2680 |
+
"step": 1830
|
| 2681 |
+
},
|
| 2682 |
+
{
|
| 2683 |
+
"epoch": 0.367,
|
| 2684 |
+
"grad_norm": 0.07886160910129547,
|
| 2685 |
+
"learning_rate": 2.999236197821727e-06,
|
| 2686 |
+
"loss": 6.4869743347167965,
|
| 2687 |
+
"step": 1835
|
| 2688 |
+
},
|
| 2689 |
+
{
|
| 2690 |
+
"epoch": 0.368,
|
| 2691 |
+
"grad_norm": 0.0706927478313446,
|
| 2692 |
+
"learning_rate": 2.9935014994981264e-06,
|
| 2693 |
+
"loss": 6.370818328857422,
|
| 2694 |
+
"step": 1840
|
| 2695 |
+
},
|
| 2696 |
+
{
|
| 2697 |
+
"epoch": 0.369,
|
| 2698 |
+
"grad_norm": 0.06873773038387299,
|
| 2699 |
+
"learning_rate": 2.9877559364013003e-06,
|
| 2700 |
+
"loss": 6.457371520996094,
|
| 2701 |
+
"step": 1845
|
| 2702 |
+
},
|
| 2703 |
+
{
|
| 2704 |
+
"epoch": 0.37,
|
| 2705 |
+
"grad_norm": 0.06985519826412201,
|
| 2706 |
+
"learning_rate": 2.981999571363806e-06,
|
| 2707 |
+
"loss": 6.5107368469238285,
|
| 2708 |
+
"step": 1850
|
| 2709 |
+
},
|
| 2710 |
+
{
|
| 2711 |
+
"epoch": 0.371,
|
| 2712 |
+
"grad_norm": 0.07739956676959991,
|
| 2713 |
+
"learning_rate": 2.9762324673363294e-06,
|
| 2714 |
+
"loss": 6.502996826171875,
|
| 2715 |
+
"step": 1855
|
| 2716 |
+
},
|
| 2717 |
+
{
|
| 2718 |
+
"epoch": 0.372,
|
| 2719 |
+
"grad_norm": 0.07745375484228134,
|
| 2720 |
+
"learning_rate": 2.9704546873869973e-06,
|
| 2721 |
+
"loss": 6.466319274902344,
|
| 2722 |
+
"step": 1860
|
| 2723 |
+
},
|
| 2724 |
+
{
|
| 2725 |
+
"epoch": 0.373,
|
| 2726 |
+
"grad_norm": 0.07050798833370209,
|
| 2727 |
+
"learning_rate": 2.964666294700683e-06,
|
| 2728 |
+
"loss": 6.417623901367188,
|
| 2729 |
+
"step": 1865
|
| 2730 |
+
},
|
| 2731 |
+
{
|
| 2732 |
+
"epoch": 0.374,
|
| 2733 |
+
"grad_norm": 0.08847381919622421,
|
| 2734 |
+
"learning_rate": 2.9588673525783243e-06,
|
| 2735 |
+
"loss": 6.471012878417969,
|
| 2736 |
+
"step": 1870
|
| 2737 |
+
},
|
| 2738 |
+
{
|
| 2739 |
+
"epoch": 0.375,
|
| 2740 |
+
"grad_norm": 0.06258077919483185,
|
| 2741 |
+
"learning_rate": 2.953057924436222e-06,
|
| 2742 |
+
"loss": 6.468167114257812,
|
| 2743 |
+
"step": 1875
|
| 2744 |
+
},
|
| 2745 |
+
{
|
| 2746 |
+
"epoch": 0.376,
|
| 2747 |
+
"grad_norm": 0.10755794495344162,
|
| 2748 |
+
"learning_rate": 2.9472380738053507e-06,
|
| 2749 |
+
"loss": 6.4053596496582035,
|
| 2750 |
+
"step": 1880
|
| 2751 |
+
},
|
| 2752 |
+
{
|
| 2753 |
+
"epoch": 0.377,
|
| 2754 |
+
"grad_norm": 0.07566042244434357,
|
| 2755 |
+
"learning_rate": 2.9414078643306656e-06,
|
| 2756 |
+
"loss": 6.423107147216797,
|
| 2757 |
+
"step": 1885
|
| 2758 |
+
},
|
| 2759 |
+
{
|
| 2760 |
+
"epoch": 0.378,
|
| 2761 |
+
"grad_norm": 0.06413055211305618,
|
| 2762 |
+
"learning_rate": 2.935567359770402e-06,
|
| 2763 |
+
"loss": 6.366802215576172,
|
| 2764 |
+
"step": 1890
|
| 2765 |
+
},
|
| 2766 |
+
{
|
| 2767 |
+
"epoch": 0.379,
|
| 2768 |
+
"grad_norm": 0.07258564233779907,
|
| 2769 |
+
"learning_rate": 2.929716623995384e-06,
|
| 2770 |
+
"loss": 6.4979804992675785,
|
| 2771 |
+
"step": 1895
|
| 2772 |
+
},
|
| 2773 |
+
{
|
| 2774 |
+
"epoch": 0.38,
|
| 2775 |
+
"grad_norm": 0.07089866697788239,
|
| 2776 |
+
"learning_rate": 2.92385572098832e-06,
|
| 2777 |
+
"loss": 6.426543426513672,
|
| 2778 |
+
"step": 1900
|
| 2779 |
+
},
|
| 2780 |
+
{
|
| 2781 |
+
"epoch": 0.381,
|
| 2782 |
+
"grad_norm": 0.1062646433711052,
|
| 2783 |
+
"learning_rate": 2.917984714843106e-06,
|
| 2784 |
+
"loss": 6.410246276855469,
|
| 2785 |
+
"step": 1905
|
| 2786 |
+
},
|
| 2787 |
+
{
|
| 2788 |
+
"epoch": 0.382,
|
| 2789 |
+
"grad_norm": 0.06609494239091873,
|
| 2790 |
+
"learning_rate": 2.9121036697641246e-06,
|
| 2791 |
+
"loss": 6.521483612060547,
|
| 2792 |
+
"step": 1910
|
| 2793 |
+
},
|
| 2794 |
+
{
|
| 2795 |
+
"epoch": 0.383,
|
| 2796 |
+
"grad_norm": 0.10069076716899872,
|
| 2797 |
+
"learning_rate": 2.9062126500655423e-06,
|
| 2798 |
+
"loss": 6.415631103515625,
|
| 2799 |
+
"step": 1915
|
| 2800 |
+
},
|
| 2801 |
+
{
|
| 2802 |
+
"epoch": 0.384,
|
| 2803 |
+
"grad_norm": 0.09040419012308121,
|
| 2804 |
+
"learning_rate": 2.900311720170607e-06,
|
| 2805 |
+
"loss": 6.462349700927734,
|
| 2806 |
+
"step": 1920
|
| 2807 |
+
},
|
| 2808 |
+
{
|
| 2809 |
+
"epoch": 0.385,
|
| 2810 |
+
"grad_norm": 0.07208941876888275,
|
| 2811 |
+
"learning_rate": 2.8944009446109422e-06,
|
| 2812 |
+
"loss": 6.416206359863281,
|
| 2813 |
+
"step": 1925
|
| 2814 |
+
},
|
| 2815 |
+
{
|
| 2816 |
+
"epoch": 0.386,
|
| 2817 |
+
"grad_norm": 0.07884255051612854,
|
| 2818 |
+
"learning_rate": 2.888480388025844e-06,
|
| 2819 |
+
"loss": 6.340870666503906,
|
| 2820 |
+
"step": 1930
|
| 2821 |
+
},
|
| 2822 |
+
{
|
| 2823 |
+
"epoch": 0.387,
|
| 2824 |
+
"grad_norm": 0.0692768543958664,
|
| 2825 |
+
"learning_rate": 2.8825501151615683e-06,
|
| 2826 |
+
"loss": 6.401695251464844,
|
| 2827 |
+
"step": 1935
|
| 2828 |
+
},
|
| 2829 |
+
{
|
| 2830 |
+
"epoch": 0.388,
|
| 2831 |
+
"grad_norm": 0.07506253570318222,
|
| 2832 |
+
"learning_rate": 2.8766101908706305e-06,
|
| 2833 |
+
"loss": 6.508678436279297,
|
| 2834 |
+
"step": 1940
|
| 2835 |
+
},
|
| 2836 |
+
{
|
| 2837 |
+
"epoch": 0.389,
|
| 2838 |
+
"grad_norm": 0.0661938264966011,
|
| 2839 |
+
"learning_rate": 2.8706606801110913e-06,
|
| 2840 |
+
"loss": 6.381354141235351,
|
| 2841 |
+
"step": 1945
|
| 2842 |
+
},
|
| 2843 |
+
{
|
| 2844 |
+
"epoch": 0.39,
|
| 2845 |
+
"grad_norm": 0.06145844608545303,
|
| 2846 |
+
"learning_rate": 2.8647016479458466e-06,
|
| 2847 |
+
"loss": 6.404244232177734,
|
| 2848 |
+
"step": 1950
|
| 2849 |
+
},
|
| 2850 |
+
{
|
| 2851 |
+
"epoch": 0.39,
|
| 2852 |
+
"eval_accuracy": 0.14162881562881563,
|
| 2853 |
+
"eval_loss": 6.339434623718262,
|
| 2854 |
+
"eval_runtime": 10.2887,
|
| 2855 |
+
"eval_samples_per_second": 9.719,
|
| 2856 |
+
"eval_steps_per_second": 1.652,
|
| 2857 |
+
"step": 1950
|
| 2858 |
+
},
|
| 2859 |
+
{
|
| 2860 |
+
"epoch": 0.391,
|
| 2861 |
+
"grad_norm": 0.07143565267324448,
|
| 2862 |
+
"learning_rate": 2.858733159541917e-06,
|
| 2863 |
+
"loss": 6.414325714111328,
|
| 2864 |
+
"step": 1955
|
| 2865 |
+
},
|
| 2866 |
+
{
|
| 2867 |
+
"epoch": 0.392,
|
| 2868 |
+
"grad_norm": 0.09751276671886444,
|
| 2869 |
+
"learning_rate": 2.8527552801697353e-06,
|
| 2870 |
+
"loss": 6.454493713378906,
|
| 2871 |
+
"step": 1960
|
| 2872 |
+
},
|
| 2873 |
+
{
|
| 2874 |
+
"epoch": 0.393,
|
| 2875 |
+
"grad_norm": 0.07358941435813904,
|
| 2876 |
+
"learning_rate": 2.846768075202434e-06,
|
| 2877 |
+
"loss": 6.389128494262695,
|
| 2878 |
+
"step": 1965
|
| 2879 |
+
},
|
| 2880 |
+
{
|
| 2881 |
+
"epoch": 0.394,
|
| 2882 |
+
"grad_norm": 0.10593241453170776,
|
| 2883 |
+
"learning_rate": 2.8407716101151263e-06,
|
| 2884 |
+
"loss": 6.536373138427734,
|
| 2885 |
+
"step": 1970
|
| 2886 |
+
},
|
| 2887 |
+
{
|
| 2888 |
+
"epoch": 0.395,
|
| 2889 |
+
"grad_norm": 0.08145789057016373,
|
| 2890 |
+
"learning_rate": 2.834765950484194e-06,
|
| 2891 |
+
"loss": 6.53704605102539,
|
| 2892 |
+
"step": 1975
|
| 2893 |
+
},
|
| 2894 |
+
{
|
| 2895 |
+
"epoch": 0.396,
|
| 2896 |
+
"grad_norm": 0.08613098412752151,
|
| 2897 |
+
"learning_rate": 2.8287511619865686e-06,
|
| 2898 |
+
"loss": 6.46817626953125,
|
| 2899 |
+
"step": 1980
|
| 2900 |
+
},
|
| 2901 |
+
{
|
| 2902 |
+
"epoch": 0.397,
|
| 2903 |
+
"grad_norm": 0.0659852847456932,
|
| 2904 |
+
"learning_rate": 2.8227273103990125e-06,
|
| 2905 |
+
"loss": 6.42239990234375,
|
| 2906 |
+
"step": 1985
|
| 2907 |
+
},
|
| 2908 |
+
{
|
| 2909 |
+
"epoch": 0.398,
|
| 2910 |
+
"grad_norm": 0.07967545092105865,
|
| 2911 |
+
"learning_rate": 2.8166944615974036e-06,
|
| 2912 |
+
"loss": 6.402915191650391,
|
| 2913 |
+
"step": 1990
|
| 2914 |
+
},
|
| 2915 |
+
{
|
| 2916 |
+
"epoch": 0.399,
|
| 2917 |
+
"grad_norm": 0.08329648524522781,
|
| 2918 |
+
"learning_rate": 2.810652681556008e-06,
|
| 2919 |
+
"loss": 6.374594116210938,
|
| 2920 |
+
"step": 1995
|
| 2921 |
+
},
|
| 2922 |
+
{
|
| 2923 |
+
"epoch": 0.4,
|
| 2924 |
+
"grad_norm": 0.11066129803657532,
|
| 2925 |
+
"learning_rate": 2.804602036346765e-06,
|
| 2926 |
+
"loss": 6.40692138671875,
|
| 2927 |
+
"step": 2000
|
| 2928 |
+
},
|
| 2929 |
+
{
|
| 2930 |
+
"epoch": 0.401,
|
| 2931 |
+
"grad_norm": 0.07232162356376648,
|
| 2932 |
+
"learning_rate": 2.7985425921385614e-06,
|
| 2933 |
+
"loss": 6.517178344726562,
|
| 2934 |
+
"step": 2005
|
| 2935 |
+
},
|
| 2936 |
+
{
|
| 2937 |
+
"epoch": 0.402,
|
| 2938 |
+
"grad_norm": 0.08892683684825897,
|
| 2939 |
+
"learning_rate": 2.7924744151965082e-06,
|
| 2940 |
+
"loss": 6.389472198486328,
|
| 2941 |
+
"step": 2010
|
| 2942 |
+
},
|
| 2943 |
+
{
|
| 2944 |
+
"epoch": 0.403,
|
| 2945 |
+
"grad_norm": 0.07714874297380447,
|
| 2946 |
+
"learning_rate": 2.7863975718812157e-06,
|
| 2947 |
+
"loss": 6.390988159179687,
|
| 2948 |
+
"step": 2015
|
| 2949 |
+
},
|
| 2950 |
+
{
|
| 2951 |
+
"epoch": 0.404,
|
| 2952 |
+
"grad_norm": 0.07850048691034317,
|
| 2953 |
+
"learning_rate": 2.7803121286480694e-06,
|
| 2954 |
+
"loss": 6.434590148925781,
|
| 2955 |
+
"step": 2020
|
| 2956 |
+
},
|
| 2957 |
+
{
|
| 2958 |
+
"epoch": 0.405,
|
| 2959 |
+
"grad_norm": 0.09433537721633911,
|
| 2960 |
+
"learning_rate": 2.774218152046501e-06,
|
| 2961 |
+
"loss": 6.42537841796875,
|
| 2962 |
+
"step": 2025
|
| 2963 |
+
},
|
| 2964 |
+
{
|
| 2965 |
+
"epoch": 0.406,
|
| 2966 |
+
"grad_norm": 0.08464869856834412,
|
| 2967 |
+
"learning_rate": 2.7681157087192627e-06,
|
| 2968 |
+
"loss": 6.396466827392578,
|
| 2969 |
+
"step": 2030
|
| 2970 |
+
},
|
| 2971 |
+
{
|
| 2972 |
+
"epoch": 0.407,
|
| 2973 |
+
"grad_norm": 0.07773975282907486,
|
| 2974 |
+
"learning_rate": 2.7620048654016962e-06,
|
| 2975 |
+
"loss": 6.33466796875,
|
| 2976 |
+
"step": 2035
|
| 2977 |
+
},
|
| 2978 |
+
{
|
| 2979 |
+
"epoch": 0.408,
|
| 2980 |
+
"grad_norm": 0.07614398002624512,
|
| 2981 |
+
"learning_rate": 2.7558856889210064e-06,
|
| 2982 |
+
"loss": 6.456300354003906,
|
| 2983 |
+
"step": 2040
|
| 2984 |
+
},
|
| 2985 |
+
{
|
| 2986 |
+
"epoch": 0.409,
|
| 2987 |
+
"grad_norm": 0.13619555532932281,
|
| 2988 |
+
"learning_rate": 2.749758246195526e-06,
|
| 2989 |
+
"loss": 6.360931396484375,
|
| 2990 |
+
"step": 2045
|
| 2991 |
+
},
|
| 2992 |
+
{
|
| 2993 |
+
"epoch": 0.41,
|
| 2994 |
+
"grad_norm": 0.06665536761283875,
|
| 2995 |
+
"learning_rate": 2.7436226042339875e-06,
|
| 2996 |
+
"loss": 6.435082244873047,
|
| 2997 |
+
"step": 2050
|
| 2998 |
+
},
|
| 2999 |
+
{
|
| 3000 |
+
"epoch": 0.411,
|
| 3001 |
+
"grad_norm": 0.07151633501052856,
|
| 3002 |
+
"learning_rate": 2.737478830134787e-06,
|
| 3003 |
+
"loss": 6.470430755615235,
|
| 3004 |
+
"step": 2055
|
| 3005 |
+
},
|
| 3006 |
+
{
|
| 3007 |
+
"epoch": 0.412,
|
| 3008 |
+
"grad_norm": 0.06864432990550995,
|
| 3009 |
+
"learning_rate": 2.7313269910852557e-06,
|
| 3010 |
+
"loss": 6.375028228759765,
|
| 3011 |
+
"step": 2060
|
| 3012 |
+
},
|
| 3013 |
+
{
|
| 3014 |
+
"epoch": 0.413,
|
| 3015 |
+
"grad_norm": 0.07884984463453293,
|
| 3016 |
+
"learning_rate": 2.725167154360918e-06,
|
| 3017 |
+
"loss": 6.419892120361328,
|
| 3018 |
+
"step": 2065
|
| 3019 |
+
},
|
| 3020 |
+
{
|
| 3021 |
+
"epoch": 0.414,
|
| 3022 |
+
"grad_norm": 0.09694062173366547,
|
| 3023 |
+
"learning_rate": 2.7189993873247627e-06,
|
| 3024 |
+
"loss": 6.416704559326172,
|
| 3025 |
+
"step": 2070
|
| 3026 |
+
},
|
| 3027 |
+
{
|
| 3028 |
+
"epoch": 0.415,
|
| 3029 |
+
"grad_norm": 0.08397479355335236,
|
| 3030 |
+
"learning_rate": 2.712823757426501e-06,
|
| 3031 |
+
"loss": 6.319974517822265,
|
| 3032 |
+
"step": 2075
|
| 3033 |
+
},
|
| 3034 |
+
{
|
| 3035 |
+
"epoch": 0.416,
|
| 3036 |
+
"grad_norm": 0.07720082998275757,
|
| 3037 |
+
"learning_rate": 2.706640332201834e-06,
|
| 3038 |
+
"loss": 6.391875839233398,
|
| 3039 |
+
"step": 2080
|
| 3040 |
+
},
|
| 3041 |
+
{
|
| 3042 |
+
"epoch": 0.417,
|
| 3043 |
+
"grad_norm": 0.07131435722112656,
|
| 3044 |
+
"learning_rate": 2.700449179271706e-06,
|
| 3045 |
+
"loss": 6.3818107604980465,
|
| 3046 |
+
"step": 2085
|
| 3047 |
+
},
|
| 3048 |
+
{
|
| 3049 |
+
"epoch": 0.418,
|
| 3050 |
+
"grad_norm": 0.22894398868083954,
|
| 3051 |
+
"learning_rate": 2.694250366341577e-06,
|
| 3052 |
+
"loss": 6.468603515625,
|
| 3053 |
+
"step": 2090
|
| 3054 |
+
},
|
| 3055 |
+
{
|
| 3056 |
+
"epoch": 0.419,
|
| 3057 |
+
"grad_norm": 0.08977403491735458,
|
| 3058 |
+
"learning_rate": 2.688043961200669e-06,
|
| 3059 |
+
"loss": 6.402335357666016,
|
| 3060 |
+
"step": 2095
|
| 3061 |
+
},
|
| 3062 |
+
{
|
| 3063 |
+
"epoch": 0.42,
|
| 3064 |
+
"grad_norm": 0.06067352741956711,
|
| 3065 |
+
"learning_rate": 2.681830031721236e-06,
|
| 3066 |
+
"loss": 6.368752288818359,
|
| 3067 |
+
"step": 2100
|
| 3068 |
+
},
|
| 3069 |
+
{
|
| 3070 |
+
"epoch": 0.42,
|
| 3071 |
+
"eval_accuracy": 0.1413943833943834,
|
| 3072 |
+
"eval_loss": 6.312870979309082,
|
| 3073 |
+
"eval_runtime": 10.3742,
|
| 3074 |
+
"eval_samples_per_second": 9.639,
|
| 3075 |
+
"eval_steps_per_second": 1.639,
|
| 3076 |
+
"step": 2100
|
| 3077 |
}
|
| 3078 |
],
|
| 3079 |
"logging_steps": 5,
|