Training in progress, step 21000, checkpoint
Browse files
checkpoint-21000/config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "openai/clip-vit-large-patch14-336",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"CLIPModel"
|
| 5 |
+
],
|
| 6 |
+
"initializer_factor": 1.0,
|
| 7 |
+
"logit_scale_init_value": 2.6592,
|
| 8 |
+
"model_type": "clip",
|
| 9 |
+
"projection_dim": 768,
|
| 10 |
+
"text_config": {
|
| 11 |
+
"dropout": 0.0,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"model_type": "clip_text_model",
|
| 15 |
+
"num_attention_heads": 12,
|
| 16 |
+
"projection_dim": 768
|
| 17 |
+
},
|
| 18 |
+
"torch_dtype": "float32",
|
| 19 |
+
"transformers_version": "4.45.0.dev0",
|
| 20 |
+
"vision_config": {
|
| 21 |
+
"dropout": 0.0,
|
| 22 |
+
"hidden_size": 1024,
|
| 23 |
+
"image_size": 336,
|
| 24 |
+
"intermediate_size": 4096,
|
| 25 |
+
"model_type": "clip_vision_model",
|
| 26 |
+
"num_attention_heads": 16,
|
| 27 |
+
"num_hidden_layers": 24,
|
| 28 |
+
"patch_size": 14,
|
| 29 |
+
"projection_dim": 768
|
| 30 |
+
}
|
| 31 |
+
}
|
checkpoint-21000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:468fa020eb3c3831b6d8f38373c8e4dd6ca8e91ae2791941a0e386a7209d6fb1
|
| 3 |
+
size 1711848436
|
checkpoint-21000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bef758d4143ce49398b56b08f46fc14c06656efc8d1d70eaa04069aa7ded1fc0
|
| 3 |
+
size 3424043887
|
checkpoint-21000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:551bd9fefe21a0210b8f009f7e379124ea37132a2e134634bfee8fc680d70fd3
|
| 3 |
+
size 14503
|
checkpoint-21000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6131bb70c0cb06d6b4cdb440d05c9970696397b8de9205389e588e7b79cfa4a1
|
| 3 |
+
size 623
|
checkpoint-21000/trainer_state.json
ADDED
|
@@ -0,0 +1,663 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7303625345230103,
|
| 3 |
+
"best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e3l17-l/checkpoint-21000",
|
| 4 |
+
"epoch": 2.2397610921501707,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 21000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.05332764505119454,
|
| 13 |
+
"grad_norm": 246.1711883544922,
|
| 14 |
+
"learning_rate": 9.822241183162683e-08,
|
| 15 |
+
"loss": 0.4515,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05332764505119454,
|
| 20 |
+
"eval_loss": 1.3850886821746826,
|
| 21 |
+
"eval_runtime": 63.8439,
|
| 22 |
+
"eval_samples_per_second": 15.46,
|
| 23 |
+
"eval_steps_per_second": 1.942,
|
| 24 |
+
"step": 500
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.10665529010238908,
|
| 28 |
+
"grad_norm": 81.48503875732422,
|
| 29 |
+
"learning_rate": 9.64448236632537e-08,
|
| 30 |
+
"loss": 0.4148,
|
| 31 |
+
"step": 1000
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.10665529010238908,
|
| 35 |
+
"eval_loss": 1.284144401550293,
|
| 36 |
+
"eval_runtime": 63.853,
|
| 37 |
+
"eval_samples_per_second": 15.457,
|
| 38 |
+
"eval_steps_per_second": 1.942,
|
| 39 |
+
"step": 1000
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.1599829351535836,
|
| 43 |
+
"grad_norm": 718.367919921875,
|
| 44 |
+
"learning_rate": 9.466723549488054e-08,
|
| 45 |
+
"loss": 0.3281,
|
| 46 |
+
"step": 1500
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.1599829351535836,
|
| 50 |
+
"eval_loss": 1.2112621068954468,
|
| 51 |
+
"eval_runtime": 64.127,
|
| 52 |
+
"eval_samples_per_second": 15.391,
|
| 53 |
+
"eval_steps_per_second": 1.934,
|
| 54 |
+
"step": 1500
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.21331058020477817,
|
| 58 |
+
"grad_norm": 4.5923566818237305,
|
| 59 |
+
"learning_rate": 9.288964732650739e-08,
|
| 60 |
+
"loss": 0.2912,
|
| 61 |
+
"step": 2000
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.21331058020477817,
|
| 65 |
+
"eval_loss": 1.159011960029602,
|
| 66 |
+
"eval_runtime": 63.7062,
|
| 67 |
+
"eval_samples_per_second": 15.493,
|
| 68 |
+
"eval_steps_per_second": 1.946,
|
| 69 |
+
"step": 2000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.2666382252559727,
|
| 73 |
+
"grad_norm": 319.4483642578125,
|
| 74 |
+
"learning_rate": 9.111205915813424e-08,
|
| 75 |
+
"loss": 0.3073,
|
| 76 |
+
"step": 2500
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.2666382252559727,
|
| 80 |
+
"eval_loss": 1.1095446348190308,
|
| 81 |
+
"eval_runtime": 63.9925,
|
| 82 |
+
"eval_samples_per_second": 15.424,
|
| 83 |
+
"eval_steps_per_second": 1.938,
|
| 84 |
+
"step": 2500
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.3199658703071672,
|
| 88 |
+
"grad_norm": 0.00027541385497897863,
|
| 89 |
+
"learning_rate": 8.933447098976109e-08,
|
| 90 |
+
"loss": 0.2446,
|
| 91 |
+
"step": 3000
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.3199658703071672,
|
| 95 |
+
"eval_loss": 1.069318175315857,
|
| 96 |
+
"eval_runtime": 64.1107,
|
| 97 |
+
"eval_samples_per_second": 15.395,
|
| 98 |
+
"eval_steps_per_second": 1.934,
|
| 99 |
+
"step": 3000
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.37329351535836175,
|
| 103 |
+
"grad_norm": 5.2035698890686035,
|
| 104 |
+
"learning_rate": 8.755688282138793e-08,
|
| 105 |
+
"loss": 0.273,
|
| 106 |
+
"step": 3500
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.37329351535836175,
|
| 110 |
+
"eval_loss": 1.043074369430542,
|
| 111 |
+
"eval_runtime": 63.9447,
|
| 112 |
+
"eval_samples_per_second": 15.435,
|
| 113 |
+
"eval_steps_per_second": 1.939,
|
| 114 |
+
"step": 3500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.42662116040955633,
|
| 118 |
+
"grad_norm": 3.6792104244232178,
|
| 119 |
+
"learning_rate": 8.57792946530148e-08,
|
| 120 |
+
"loss": 0.193,
|
| 121 |
+
"step": 4000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.42662116040955633,
|
| 125 |
+
"eval_loss": 1.0195527076721191,
|
| 126 |
+
"eval_runtime": 63.8076,
|
| 127 |
+
"eval_samples_per_second": 15.468,
|
| 128 |
+
"eval_steps_per_second": 1.943,
|
| 129 |
+
"step": 4000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.47994880546075086,
|
| 133 |
+
"grad_norm": 111.2447280883789,
|
| 134 |
+
"learning_rate": 8.400170648464163e-08,
|
| 135 |
+
"loss": 0.2419,
|
| 136 |
+
"step": 4500
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.47994880546075086,
|
| 140 |
+
"eval_loss": 0.9946303963661194,
|
| 141 |
+
"eval_runtime": 63.8873,
|
| 142 |
+
"eval_samples_per_second": 15.449,
|
| 143 |
+
"eval_steps_per_second": 1.941,
|
| 144 |
+
"step": 4500
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.5332764505119454,
|
| 148 |
+
"grad_norm": 0.024256089702248573,
|
| 149 |
+
"learning_rate": 8.222411831626848e-08,
|
| 150 |
+
"loss": 0.2577,
|
| 151 |
+
"step": 5000
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.5332764505119454,
|
| 155 |
+
"eval_loss": 0.9679088592529297,
|
| 156 |
+
"eval_runtime": 62.896,
|
| 157 |
+
"eval_samples_per_second": 15.693,
|
| 158 |
+
"eval_steps_per_second": 1.972,
|
| 159 |
+
"step": 5000
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.58660409556314,
|
| 163 |
+
"grad_norm": 0.05403488501906395,
|
| 164 |
+
"learning_rate": 8.044653014789534e-08,
|
| 165 |
+
"loss": 0.2468,
|
| 166 |
+
"step": 5500
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.58660409556314,
|
| 170 |
+
"eval_loss": 0.9470070004463196,
|
| 171 |
+
"eval_runtime": 64.2367,
|
| 172 |
+
"eval_samples_per_second": 15.365,
|
| 173 |
+
"eval_steps_per_second": 1.93,
|
| 174 |
+
"step": 5500
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.6399317406143344,
|
| 178 |
+
"grad_norm": 56.23408508300781,
|
| 179 |
+
"learning_rate": 7.866894197952219e-08,
|
| 180 |
+
"loss": 0.2148,
|
| 181 |
+
"step": 6000
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 0.6399317406143344,
|
| 185 |
+
"eval_loss": 0.9275079369544983,
|
| 186 |
+
"eval_runtime": 62.7425,
|
| 187 |
+
"eval_samples_per_second": 15.731,
|
| 188 |
+
"eval_steps_per_second": 1.976,
|
| 189 |
+
"step": 6000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.693259385665529,
|
| 193 |
+
"grad_norm": 335.7691955566406,
|
| 194 |
+
"learning_rate": 7.689135381114903e-08,
|
| 195 |
+
"loss": 0.179,
|
| 196 |
+
"step": 6500
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.693259385665529,
|
| 200 |
+
"eval_loss": 0.9107733964920044,
|
| 201 |
+
"eval_runtime": 63.711,
|
| 202 |
+
"eval_samples_per_second": 15.492,
|
| 203 |
+
"eval_steps_per_second": 1.946,
|
| 204 |
+
"step": 6500
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.7465870307167235,
|
| 208 |
+
"grad_norm": 0.0303946640342474,
|
| 209 |
+
"learning_rate": 7.511376564277589e-08,
|
| 210 |
+
"loss": 0.1992,
|
| 211 |
+
"step": 7000
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.7465870307167235,
|
| 215 |
+
"eval_loss": 0.8946455121040344,
|
| 216 |
+
"eval_runtime": 62.7465,
|
| 217 |
+
"eval_samples_per_second": 15.73,
|
| 218 |
+
"eval_steps_per_second": 1.976,
|
| 219 |
+
"step": 7000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.7999146757679181,
|
| 223 |
+
"grad_norm": 24.27122688293457,
|
| 224 |
+
"learning_rate": 7.333617747440273e-08,
|
| 225 |
+
"loss": 0.2264,
|
| 226 |
+
"step": 7500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.7999146757679181,
|
| 230 |
+
"eval_loss": 0.8790715932846069,
|
| 231 |
+
"eval_runtime": 62.625,
|
| 232 |
+
"eval_samples_per_second": 15.76,
|
| 233 |
+
"eval_steps_per_second": 1.98,
|
| 234 |
+
"step": 7500
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.8532423208191127,
|
| 238 |
+
"grad_norm": 485.13018798828125,
|
| 239 |
+
"learning_rate": 7.155858930602957e-08,
|
| 240 |
+
"loss": 0.2349,
|
| 241 |
+
"step": 8000
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.8532423208191127,
|
| 245 |
+
"eval_loss": 0.8648595213890076,
|
| 246 |
+
"eval_runtime": 63.4718,
|
| 247 |
+
"eval_samples_per_second": 15.55,
|
| 248 |
+
"eval_steps_per_second": 1.954,
|
| 249 |
+
"step": 8000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.9065699658703071,
|
| 253 |
+
"grad_norm": 39.18803787231445,
|
| 254 |
+
"learning_rate": 6.978100113765643e-08,
|
| 255 |
+
"loss": 0.1869,
|
| 256 |
+
"step": 8500
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.9065699658703071,
|
| 260 |
+
"eval_loss": 0.8521347641944885,
|
| 261 |
+
"eval_runtime": 62.649,
|
| 262 |
+
"eval_samples_per_second": 15.754,
|
| 263 |
+
"eval_steps_per_second": 1.979,
|
| 264 |
+
"step": 8500
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.9598976109215017,
|
| 268 |
+
"grad_norm": 15.91203784942627,
|
| 269 |
+
"learning_rate": 6.800341296928327e-08,
|
| 270 |
+
"loss": 0.1859,
|
| 271 |
+
"step": 9000
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.9598976109215017,
|
| 275 |
+
"eval_loss": 0.8395125269889832,
|
| 276 |
+
"eval_runtime": 63.1398,
|
| 277 |
+
"eval_samples_per_second": 15.632,
|
| 278 |
+
"eval_steps_per_second": 1.964,
|
| 279 |
+
"step": 9000
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.0132252559726962,
|
| 283 |
+
"grad_norm": 545.9375610351562,
|
| 284 |
+
"learning_rate": 6.622582480091012e-08,
|
| 285 |
+
"loss": 0.1943,
|
| 286 |
+
"step": 9500
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 1.0132252559726962,
|
| 290 |
+
"eval_loss": 0.8300319910049438,
|
| 291 |
+
"eval_runtime": 63.2047,
|
| 292 |
+
"eval_samples_per_second": 15.616,
|
| 293 |
+
"eval_steps_per_second": 1.962,
|
| 294 |
+
"step": 9500
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.0665529010238908,
|
| 298 |
+
"grad_norm": 451.04473876953125,
|
| 299 |
+
"learning_rate": 6.444823663253697e-08,
|
| 300 |
+
"loss": 0.1419,
|
| 301 |
+
"step": 10000
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.0665529010238908,
|
| 305 |
+
"eval_loss": 0.8227387070655823,
|
| 306 |
+
"eval_runtime": 62.7357,
|
| 307 |
+
"eval_samples_per_second": 15.733,
|
| 308 |
+
"eval_steps_per_second": 1.977,
|
| 309 |
+
"step": 10000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.1198805460750854,
|
| 313 |
+
"grad_norm": 12.075934410095215,
|
| 314 |
+
"learning_rate": 6.267064846416382e-08,
|
| 315 |
+
"loss": 0.1558,
|
| 316 |
+
"step": 10500
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.1198805460750854,
|
| 320 |
+
"eval_loss": 0.8159177303314209,
|
| 321 |
+
"eval_runtime": 62.8297,
|
| 322 |
+
"eval_samples_per_second": 15.709,
|
| 323 |
+
"eval_steps_per_second": 1.974,
|
| 324 |
+
"step": 10500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.17320819112628,
|
| 328 |
+
"grad_norm": 363.5056457519531,
|
| 329 |
+
"learning_rate": 6.089306029579066e-08,
|
| 330 |
+
"loss": 0.15,
|
| 331 |
+
"step": 11000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.17320819112628,
|
| 335 |
+
"eval_loss": 0.8096186518669128,
|
| 336 |
+
"eval_runtime": 63.3529,
|
| 337 |
+
"eval_samples_per_second": 15.579,
|
| 338 |
+
"eval_steps_per_second": 1.957,
|
| 339 |
+
"step": 11000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.2265358361774745,
|
| 343 |
+
"grad_norm": 6.587007646885468e-06,
|
| 344 |
+
"learning_rate": 5.911547212741752e-08,
|
| 345 |
+
"loss": 0.1919,
|
| 346 |
+
"step": 11500
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.2265358361774745,
|
| 350 |
+
"eval_loss": 0.8023524284362793,
|
| 351 |
+
"eval_runtime": 62.9393,
|
| 352 |
+
"eval_samples_per_second": 15.682,
|
| 353 |
+
"eval_steps_per_second": 1.97,
|
| 354 |
+
"step": 11500
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.2798634812286689,
|
| 358 |
+
"grad_norm": 0.11630824208259583,
|
| 359 |
+
"learning_rate": 5.7337883959044365e-08,
|
| 360 |
+
"loss": 0.1298,
|
| 361 |
+
"step": 12000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.2798634812286689,
|
| 365 |
+
"eval_loss": 0.7956322431564331,
|
| 366 |
+
"eval_runtime": 63.0726,
|
| 367 |
+
"eval_samples_per_second": 15.649,
|
| 368 |
+
"eval_steps_per_second": 1.966,
|
| 369 |
+
"step": 12000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.3331911262798635,
|
| 373 |
+
"grad_norm": 0.0007865954539738595,
|
| 374 |
+
"learning_rate": 5.556029579067122e-08,
|
| 375 |
+
"loss": 0.1385,
|
| 376 |
+
"step": 12500
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 1.3331911262798635,
|
| 380 |
+
"eval_loss": 0.7920646667480469,
|
| 381 |
+
"eval_runtime": 63.3155,
|
| 382 |
+
"eval_samples_per_second": 15.589,
|
| 383 |
+
"eval_steps_per_second": 1.958,
|
| 384 |
+
"step": 12500
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.386518771331058,
|
| 388 |
+
"grad_norm": 0.00023369144764728844,
|
| 389 |
+
"learning_rate": 5.378270762229806e-08,
|
| 390 |
+
"loss": 0.1414,
|
| 391 |
+
"step": 13000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 1.386518771331058,
|
| 395 |
+
"eval_loss": 0.7884665131568909,
|
| 396 |
+
"eval_runtime": 63.002,
|
| 397 |
+
"eval_samples_per_second": 15.666,
|
| 398 |
+
"eval_steps_per_second": 1.968,
|
| 399 |
+
"step": 13000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.4398464163822526,
|
| 403 |
+
"grad_norm": 4.277670383453369,
|
| 404 |
+
"learning_rate": 5.200511945392492e-08,
|
| 405 |
+
"loss": 0.1549,
|
| 406 |
+
"step": 13500
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.4398464163822526,
|
| 410 |
+
"eval_loss": 0.7809606194496155,
|
| 411 |
+
"eval_runtime": 63.0606,
|
| 412 |
+
"eval_samples_per_second": 15.652,
|
| 413 |
+
"eval_steps_per_second": 1.966,
|
| 414 |
+
"step": 13500
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 1.493174061433447,
|
| 418 |
+
"grad_norm": 0.04092055931687355,
|
| 419 |
+
"learning_rate": 5.022753128555176e-08,
|
| 420 |
+
"loss": 0.126,
|
| 421 |
+
"step": 14000
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 1.493174061433447,
|
| 425 |
+
"eval_loss": 0.7771912217140198,
|
| 426 |
+
"eval_runtime": 63.2952,
|
| 427 |
+
"eval_samples_per_second": 15.594,
|
| 428 |
+
"eval_steps_per_second": 1.959,
|
| 429 |
+
"step": 14000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.5465017064846416,
|
| 433 |
+
"grad_norm": 213.26644897460938,
|
| 434 |
+
"learning_rate": 4.844994311717861e-08,
|
| 435 |
+
"loss": 0.1518,
|
| 436 |
+
"step": 14500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.5465017064846416,
|
| 440 |
+
"eval_loss": 0.7739039659500122,
|
| 441 |
+
"eval_runtime": 63.1919,
|
| 442 |
+
"eval_samples_per_second": 15.619,
|
| 443 |
+
"eval_steps_per_second": 1.962,
|
| 444 |
+
"step": 14500
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.5998293515358362,
|
| 448 |
+
"grad_norm": 550.1605834960938,
|
| 449 |
+
"learning_rate": 4.667235494880546e-08,
|
| 450 |
+
"loss": 0.1143,
|
| 451 |
+
"step": 15000
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.5998293515358362,
|
| 455 |
+
"eval_loss": 0.7702728509902954,
|
| 456 |
+
"eval_runtime": 62.9504,
|
| 457 |
+
"eval_samples_per_second": 15.679,
|
| 458 |
+
"eval_steps_per_second": 1.97,
|
| 459 |
+
"step": 15000
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.6531569965870307,
|
| 463 |
+
"grad_norm": 6.588364124298096,
|
| 464 |
+
"learning_rate": 4.4894766780432306e-08,
|
| 465 |
+
"loss": 0.161,
|
| 466 |
+
"step": 15500
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.6531569965870307,
|
| 470 |
+
"eval_loss": 0.7638877034187317,
|
| 471 |
+
"eval_runtime": 63.3812,
|
| 472 |
+
"eval_samples_per_second": 15.572,
|
| 473 |
+
"eval_steps_per_second": 1.956,
|
| 474 |
+
"step": 15500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 1.7064846416382253,
|
| 478 |
+
"grad_norm": 245.7973175048828,
|
| 479 |
+
"learning_rate": 4.311717861205916e-08,
|
| 480 |
+
"loss": 0.1277,
|
| 481 |
+
"step": 16000
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 1.7064846416382253,
|
| 485 |
+
"eval_loss": 0.7590143084526062,
|
| 486 |
+
"eval_runtime": 63.157,
|
| 487 |
+
"eval_samples_per_second": 15.628,
|
| 488 |
+
"eval_steps_per_second": 1.963,
|
| 489 |
+
"step": 16000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.75981228668942,
|
| 493 |
+
"grad_norm": 0.013761636801064014,
|
| 494 |
+
"learning_rate": 4.133959044368601e-08,
|
| 495 |
+
"loss": 0.1394,
|
| 496 |
+
"step": 16500
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 1.75981228668942,
|
| 500 |
+
"eval_loss": 0.7563473582267761,
|
| 501 |
+
"eval_runtime": 63.006,
|
| 502 |
+
"eval_samples_per_second": 15.665,
|
| 503 |
+
"eval_steps_per_second": 1.968,
|
| 504 |
+
"step": 16500
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 1.8131399317406145,
|
| 508 |
+
"grad_norm": 0.0008418822544626892,
|
| 509 |
+
"learning_rate": 3.9562002275312854e-08,
|
| 510 |
+
"loss": 0.1253,
|
| 511 |
+
"step": 17000
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 1.8131399317406145,
|
| 515 |
+
"eval_loss": 0.7549872398376465,
|
| 516 |
+
"eval_runtime": 63.1042,
|
| 517 |
+
"eval_samples_per_second": 15.641,
|
| 518 |
+
"eval_steps_per_second": 1.965,
|
| 519 |
+
"step": 17000
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.8664675767918089,
|
| 523 |
+
"grad_norm": 58.008296966552734,
|
| 524 |
+
"learning_rate": 3.7784414106939705e-08,
|
| 525 |
+
"loss": 0.2146,
|
| 526 |
+
"step": 17500
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 1.8664675767918089,
|
| 530 |
+
"eval_loss": 0.7488264441490173,
|
| 531 |
+
"eval_runtime": 63.1754,
|
| 532 |
+
"eval_samples_per_second": 15.623,
|
| 533 |
+
"eval_steps_per_second": 1.963,
|
| 534 |
+
"step": 17500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.9197952218430034,
|
| 538 |
+
"grad_norm": 0.0007694943924434483,
|
| 539 |
+
"learning_rate": 3.600682593856656e-08,
|
| 540 |
+
"loss": 0.1267,
|
| 541 |
+
"step": 18000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.9197952218430034,
|
| 545 |
+
"eval_loss": 0.7465928792953491,
|
| 546 |
+
"eval_runtime": 62.8126,
|
| 547 |
+
"eval_samples_per_second": 15.713,
|
| 548 |
+
"eval_steps_per_second": 1.974,
|
| 549 |
+
"step": 18000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.9731228668941978,
|
| 553 |
+
"grad_norm": 515.19580078125,
|
| 554 |
+
"learning_rate": 3.4229237770193395e-08,
|
| 555 |
+
"loss": 0.1564,
|
| 556 |
+
"step": 18500
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.9731228668941978,
|
| 560 |
+
"eval_loss": 0.7438480854034424,
|
| 561 |
+
"eval_runtime": 63.2698,
|
| 562 |
+
"eval_samples_per_second": 15.6,
|
| 563 |
+
"eval_steps_per_second": 1.96,
|
| 564 |
+
"step": 18500
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.0264505119453924,
|
| 568 |
+
"grad_norm": 27.036834716796875,
|
| 569 |
+
"learning_rate": 3.2451649601820246e-08,
|
| 570 |
+
"loss": 0.1355,
|
| 571 |
+
"step": 19000
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 2.0264505119453924,
|
| 575 |
+
"eval_loss": 0.7395957112312317,
|
| 576 |
+
"eval_runtime": 63.3696,
|
| 577 |
+
"eval_samples_per_second": 15.575,
|
| 578 |
+
"eval_steps_per_second": 1.957,
|
| 579 |
+
"step": 19000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 2.079778156996587,
|
| 583 |
+
"grad_norm": 6.471228122711182,
|
| 584 |
+
"learning_rate": 3.06740614334471e-08,
|
| 585 |
+
"loss": 0.1228,
|
| 586 |
+
"step": 19500
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 2.079778156996587,
|
| 590 |
+
"eval_loss": 0.7349324226379395,
|
| 591 |
+
"eval_runtime": 62.8971,
|
| 592 |
+
"eval_samples_per_second": 15.692,
|
| 593 |
+
"eval_steps_per_second": 1.971,
|
| 594 |
+
"step": 19500
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.1331058020477816,
|
| 598 |
+
"grad_norm": 0.05623076856136322,
|
| 599 |
+
"learning_rate": 2.8896473265073946e-08,
|
| 600 |
+
"loss": 0.1082,
|
| 601 |
+
"step": 20000
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.1331058020477816,
|
| 605 |
+
"eval_loss": 0.7336575388908386,
|
| 606 |
+
"eval_runtime": 63.3128,
|
| 607 |
+
"eval_samples_per_second": 15.589,
|
| 608 |
+
"eval_steps_per_second": 1.959,
|
| 609 |
+
"step": 20000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 2.186433447098976,
|
| 613 |
+
"grad_norm": 0.013808301649987698,
|
| 614 |
+
"learning_rate": 2.7118885096700794e-08,
|
| 615 |
+
"loss": 0.1353,
|
| 616 |
+
"step": 20500
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 2.186433447098976,
|
| 620 |
+
"eval_loss": 0.7329619526863098,
|
| 621 |
+
"eval_runtime": 63.2838,
|
| 622 |
+
"eval_samples_per_second": 15.596,
|
| 623 |
+
"eval_steps_per_second": 1.959,
|
| 624 |
+
"step": 20500
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 2.2397610921501707,
|
| 628 |
+
"grad_norm": 1.4853014945983887,
|
| 629 |
+
"learning_rate": 2.5341296928327643e-08,
|
| 630 |
+
"loss": 0.0925,
|
| 631 |
+
"step": 21000
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 2.2397610921501707,
|
| 635 |
+
"eval_loss": 0.7303625345230103,
|
| 636 |
+
"eval_runtime": 62.7838,
|
| 637 |
+
"eval_samples_per_second": 15.721,
|
| 638 |
+
"eval_steps_per_second": 1.975,
|
| 639 |
+
"step": 21000
|
| 640 |
+
}
|
| 641 |
+
],
|
| 642 |
+
"logging_steps": 500,
|
| 643 |
+
"max_steps": 28128,
|
| 644 |
+
"num_input_tokens_seen": 0,
|
| 645 |
+
"num_train_epochs": 3,
|
| 646 |
+
"save_steps": 500,
|
| 647 |
+
"stateful_callbacks": {
|
| 648 |
+
"TrainerControl": {
|
| 649 |
+
"args": {
|
| 650 |
+
"should_epoch_stop": false,
|
| 651 |
+
"should_evaluate": false,
|
| 652 |
+
"should_log": false,
|
| 653 |
+
"should_save": true,
|
| 654 |
+
"should_training_stop": false
|
| 655 |
+
},
|
| 656 |
+
"attributes": {}
|
| 657 |
+
}
|
| 658 |
+
},
|
| 659 |
+
"total_flos": 7555644893162520.0,
|
| 660 |
+
"train_batch_size": 2,
|
| 661 |
+
"trial_name": null,
|
| 662 |
+
"trial_params": null
|
| 663 |
+
}
|
checkpoint-21000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:52c8db40c27633302368150be011282e1650a78d29cdfc5e366df6966e8e6c70
|
| 3 |
+
size 4847
|