Upload folder using huggingface_hub
#1
by
krishnakalyan3
- opened
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +801 -0
- training_args.bin +3 -0
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4478730f732fbb53d7545e380e448ccf7a073e15b3c68d977022f1cec04fd383
|
| 3 |
+
size 379087640
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:35035d2d7cb8ef5e57ca52cab364265dfeca6e1699a9cf1217983e177504fb40
|
| 3 |
+
size 3152668
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:58046cb749995f2c22dddb6fde638ff19504b2e6af986ac8ab7832bfb7fd7979
|
| 3 |
+
size 14308
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:13840319dca442e4fb1af18b14c487350ba9d7925330764edc88119d995c869d
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,801 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.21792557956921524,
|
| 3 |
+
"best_model_checkpoint": "/workspace/disk2/krishna/checkpoints/checkpoint-480",
|
| 4 |
+
"epoch": 0.048,
|
| 5 |
+
"eval_steps": 10,
|
| 6 |
+
"global_step": 480,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.001,
|
| 13 |
+
"grad_norm": 1.7856025695800781,
|
| 14 |
+
"learning_rate": 5.000000000000001e-07,
|
| 15 |
+
"loss": 0.682,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.001,
|
| 20 |
+
"eval_cos_sim": 0.30435964465141296,
|
| 21 |
+
"eval_loss": 0.6965229106162732,
|
| 22 |
+
"eval_runtime": 181.358,
|
| 23 |
+
"eval_samples_per_second": 22.056,
|
| 24 |
+
"eval_steps_per_second": 0.347,
|
| 25 |
+
"step": 10
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.002,
|
| 29 |
+
"grad_norm": 1.632306694984436,
|
| 30 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 31 |
+
"loss": 0.6902,
|
| 32 |
+
"step": 20
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.002,
|
| 36 |
+
"eval_cos_sim": 0.3060647249221802,
|
| 37 |
+
"eval_loss": 0.6948180051063245,
|
| 38 |
+
"eval_runtime": 158.9409,
|
| 39 |
+
"eval_samples_per_second": 25.167,
|
| 40 |
+
"eval_steps_per_second": 0.396,
|
| 41 |
+
"step": 20
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.003,
|
| 45 |
+
"grad_norm": 1.4781064987182617,
|
| 46 |
+
"learning_rate": 1.5e-06,
|
| 47 |
+
"loss": 0.6879,
|
| 48 |
+
"step": 30
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.003,
|
| 52 |
+
"eval_cos_sim": 0.30893880128860474,
|
| 53 |
+
"eval_loss": 0.6919441828941053,
|
| 54 |
+
"eval_runtime": 163.5656,
|
| 55 |
+
"eval_samples_per_second": 24.455,
|
| 56 |
+
"eval_steps_per_second": 0.385,
|
| 57 |
+
"step": 30
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.004,
|
| 61 |
+
"grad_norm": 1.4996154308319092,
|
| 62 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 63 |
+
"loss": 0.6805,
|
| 64 |
+
"step": 40
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.004,
|
| 68 |
+
"eval_cos_sim": 0.3130052089691162,
|
| 69 |
+
"eval_loss": 0.6878782258247084,
|
| 70 |
+
"eval_runtime": 160.8804,
|
| 71 |
+
"eval_samples_per_second": 24.863,
|
| 72 |
+
"eval_steps_per_second": 0.392,
|
| 73 |
+
"step": 40
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.005,
|
| 77 |
+
"grad_norm": 1.7979084253311157,
|
| 78 |
+
"learning_rate": 2.5e-06,
|
| 79 |
+
"loss": 0.6795,
|
| 80 |
+
"step": 50
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.005,
|
| 84 |
+
"eval_cos_sim": 0.3181459903717041,
|
| 85 |
+
"eval_loss": 0.6827379918311781,
|
| 86 |
+
"eval_runtime": 159.4021,
|
| 87 |
+
"eval_samples_per_second": 25.094,
|
| 88 |
+
"eval_steps_per_second": 0.395,
|
| 89 |
+
"step": 50
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.006,
|
| 93 |
+
"grad_norm": 1.4049359560012817,
|
| 94 |
+
"learning_rate": 3e-06,
|
| 95 |
+
"loss": 0.6781,
|
| 96 |
+
"step": 60
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.006,
|
| 100 |
+
"eval_cos_sim": 0.32444360852241516,
|
| 101 |
+
"eval_loss": 0.6764405532096571,
|
| 102 |
+
"eval_runtime": 159.9515,
|
| 103 |
+
"eval_samples_per_second": 25.008,
|
| 104 |
+
"eval_steps_per_second": 0.394,
|
| 105 |
+
"step": 60
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.007,
|
| 109 |
+
"grad_norm": 1.7637789249420166,
|
| 110 |
+
"learning_rate": 3.5000000000000004e-06,
|
| 111 |
+
"loss": 0.6718,
|
| 112 |
+
"step": 70
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.007,
|
| 116 |
+
"eval_cos_sim": 0.331728994846344,
|
| 117 |
+
"eval_loss": 0.6691558842872327,
|
| 118 |
+
"eval_runtime": 160.8921,
|
| 119 |
+
"eval_samples_per_second": 24.861,
|
| 120 |
+
"eval_steps_per_second": 0.392,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.008,
|
| 125 |
+
"grad_norm": 1.7560211420059204,
|
| 126 |
+
"learning_rate": 4.000000000000001e-06,
|
| 127 |
+
"loss": 0.6551,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.008,
|
| 132 |
+
"eval_cos_sim": 0.3400111794471741,
|
| 133 |
+
"eval_loss": 0.6608743510459608,
|
| 134 |
+
"eval_runtime": 172.4109,
|
| 135 |
+
"eval_samples_per_second": 23.2,
|
| 136 |
+
"eval_steps_per_second": 0.365,
|
| 137 |
+
"step": 80
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.009,
|
| 141 |
+
"grad_norm": 1.7577630281448364,
|
| 142 |
+
"learning_rate": 4.5e-06,
|
| 143 |
+
"loss": 0.6566,
|
| 144 |
+
"step": 90
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.009,
|
| 148 |
+
"eval_cos_sim": 0.34923624992370605,
|
| 149 |
+
"eval_loss": 0.6516506343101209,
|
| 150 |
+
"eval_runtime": 161.459,
|
| 151 |
+
"eval_samples_per_second": 24.774,
|
| 152 |
+
"eval_steps_per_second": 0.39,
|
| 153 |
+
"step": 90
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.01,
|
| 157 |
+
"grad_norm": 1.7723366022109985,
|
| 158 |
+
"learning_rate": 5e-06,
|
| 159 |
+
"loss": 0.6332,
|
| 160 |
+
"step": 100
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.01,
|
| 164 |
+
"eval_cos_sim": 0.35936808586120605,
|
| 165 |
+
"eval_loss": 0.6415200128768629,
|
| 166 |
+
"eval_runtime": 158.9596,
|
| 167 |
+
"eval_samples_per_second": 25.164,
|
| 168 |
+
"eval_steps_per_second": 0.396,
|
| 169 |
+
"step": 100
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.011,
|
| 173 |
+
"grad_norm": 1.7818301916122437,
|
| 174 |
+
"learning_rate": 5.500000000000001e-06,
|
| 175 |
+
"loss": 0.6297,
|
| 176 |
+
"step": 110
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.011,
|
| 180 |
+
"eval_cos_sim": 0.3702617287635803,
|
| 181 |
+
"eval_loss": 0.6306272282813734,
|
| 182 |
+
"eval_runtime": 161.4739,
|
| 183 |
+
"eval_samples_per_second": 24.772,
|
| 184 |
+
"eval_steps_per_second": 0.39,
|
| 185 |
+
"step": 110
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.012,
|
| 189 |
+
"grad_norm": 1.4825223684310913,
|
| 190 |
+
"learning_rate": 6e-06,
|
| 191 |
+
"loss": 0.6473,
|
| 192 |
+
"step": 120
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.012,
|
| 196 |
+
"eval_cos_sim": 0.381907194852829,
|
| 197 |
+
"eval_loss": 0.6189834599708265,
|
| 198 |
+
"eval_runtime": 162.0142,
|
| 199 |
+
"eval_samples_per_second": 24.689,
|
| 200 |
+
"eval_steps_per_second": 0.389,
|
| 201 |
+
"step": 120
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.013,
|
| 205 |
+
"grad_norm": 1.7424395084381104,
|
| 206 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 207 |
+
"loss": 0.611,
|
| 208 |
+
"step": 130
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.013,
|
| 212 |
+
"eval_cos_sim": 0.3940889537334442,
|
| 213 |
+
"eval_loss": 0.6068024029945082,
|
| 214 |
+
"eval_runtime": 161.0467,
|
| 215 |
+
"eval_samples_per_second": 24.838,
|
| 216 |
+
"eval_steps_per_second": 0.391,
|
| 217 |
+
"step": 130
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.014,
|
| 221 |
+
"grad_norm": 1.7474108934402466,
|
| 222 |
+
"learning_rate": 7.000000000000001e-06,
|
| 223 |
+
"loss": 0.5932,
|
| 224 |
+
"step": 140
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.014,
|
| 228 |
+
"eval_cos_sim": 0.4073576033115387,
|
| 229 |
+
"eval_loss": 0.5935340962623304,
|
| 230 |
+
"eval_runtime": 161.0017,
|
| 231 |
+
"eval_samples_per_second": 24.844,
|
| 232 |
+
"eval_steps_per_second": 0.391,
|
| 233 |
+
"step": 140
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.015,
|
| 237 |
+
"grad_norm": 1.6058062314987183,
|
| 238 |
+
"learning_rate": 7.5e-06,
|
| 239 |
+
"loss": 0.5824,
|
| 240 |
+
"step": 150
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.015,
|
| 244 |
+
"eval_cos_sim": 0.42149588465690613,
|
| 245 |
+
"eval_loss": 0.5793958764289564,
|
| 246 |
+
"eval_runtime": 167.7429,
|
| 247 |
+
"eval_samples_per_second": 23.846,
|
| 248 |
+
"eval_steps_per_second": 0.376,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.016,
|
| 253 |
+
"grad_norm": 1.5788137912750244,
|
| 254 |
+
"learning_rate": 8.000000000000001e-06,
|
| 255 |
+
"loss": 0.5634,
|
| 256 |
+
"step": 160
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.016,
|
| 260 |
+
"eval_cos_sim": 0.43618762493133545,
|
| 261 |
+
"eval_loss": 0.564703805944795,
|
| 262 |
+
"eval_runtime": 159.9946,
|
| 263 |
+
"eval_samples_per_second": 25.001,
|
| 264 |
+
"eval_steps_per_second": 0.394,
|
| 265 |
+
"step": 160
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.017,
|
| 269 |
+
"grad_norm": 1.6113280057907104,
|
| 270 |
+
"learning_rate": 8.500000000000002e-06,
|
| 271 |
+
"loss": 0.5519,
|
| 272 |
+
"step": 170
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.017,
|
| 276 |
+
"eval_cos_sim": 0.4508235454559326,
|
| 277 |
+
"eval_loss": 0.5500685830329604,
|
| 278 |
+
"eval_runtime": 157.8438,
|
| 279 |
+
"eval_samples_per_second": 25.342,
|
| 280 |
+
"eval_steps_per_second": 0.399,
|
| 281 |
+
"step": 170
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.018,
|
| 285 |
+
"grad_norm": 1.6156154870986938,
|
| 286 |
+
"learning_rate": 9e-06,
|
| 287 |
+
"loss": 0.5407,
|
| 288 |
+
"step": 180
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.018,
|
| 292 |
+
"eval_cos_sim": 0.46566617488861084,
|
| 293 |
+
"eval_loss": 0.5352267394279188,
|
| 294 |
+
"eval_runtime": 159.9444,
|
| 295 |
+
"eval_samples_per_second": 25.009,
|
| 296 |
+
"eval_steps_per_second": 0.394,
|
| 297 |
+
"step": 180
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.019,
|
| 301 |
+
"grad_norm": 1.5958607196807861,
|
| 302 |
+
"learning_rate": 9.5e-06,
|
| 303 |
+
"loss": 0.5117,
|
| 304 |
+
"step": 190
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.019,
|
| 308 |
+
"eval_cos_sim": 0.4806951582431793,
|
| 309 |
+
"eval_loss": 0.5202000065063185,
|
| 310 |
+
"eval_runtime": 157.8685,
|
| 311 |
+
"eval_samples_per_second": 25.338,
|
| 312 |
+
"eval_steps_per_second": 0.399,
|
| 313 |
+
"step": 190
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.02,
|
| 317 |
+
"grad_norm": 1.6012651920318604,
|
| 318 |
+
"learning_rate": 1e-05,
|
| 319 |
+
"loss": 0.5116,
|
| 320 |
+
"step": 200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.02,
|
| 324 |
+
"eval_cos_sim": 0.49558621644973755,
|
| 325 |
+
"eval_loss": 0.5053114848350233,
|
| 326 |
+
"eval_runtime": 155.395,
|
| 327 |
+
"eval_samples_per_second": 25.741,
|
| 328 |
+
"eval_steps_per_second": 0.405,
|
| 329 |
+
"step": 200
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.021,
|
| 333 |
+
"grad_norm": 1.7965919971466064,
|
| 334 |
+
"learning_rate": 1.05e-05,
|
| 335 |
+
"loss": 0.4782,
|
| 336 |
+
"step": 210
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 0.021,
|
| 340 |
+
"eval_cos_sim": 0.5108910202980042,
|
| 341 |
+
"eval_loss": 0.49000780775296154,
|
| 342 |
+
"eval_runtime": 159.6303,
|
| 343 |
+
"eval_samples_per_second": 25.058,
|
| 344 |
+
"eval_steps_per_second": 0.395,
|
| 345 |
+
"step": 210
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.022,
|
| 349 |
+
"grad_norm": 1.791284441947937,
|
| 350 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 351 |
+
"loss": 0.4802,
|
| 352 |
+
"step": 220
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.022,
|
| 356 |
+
"eval_cos_sim": 0.5271998643875122,
|
| 357 |
+
"eval_loss": 0.4737012028907484,
|
| 358 |
+
"eval_runtime": 169.963,
|
| 359 |
+
"eval_samples_per_second": 23.535,
|
| 360 |
+
"eval_steps_per_second": 0.371,
|
| 361 |
+
"step": 220
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.023,
|
| 365 |
+
"grad_norm": 1.7923716306686401,
|
| 366 |
+
"learning_rate": 1.1500000000000002e-05,
|
| 367 |
+
"loss": 0.4513,
|
| 368 |
+
"step": 230
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.023,
|
| 372 |
+
"eval_cos_sim": 0.5435076355934143,
|
| 373 |
+
"eval_loss": 0.45739728691327036,
|
| 374 |
+
"eval_runtime": 160.753,
|
| 375 |
+
"eval_samples_per_second": 24.883,
|
| 376 |
+
"eval_steps_per_second": 0.392,
|
| 377 |
+
"step": 230
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.024,
|
| 381 |
+
"grad_norm": 1.7997322082519531,
|
| 382 |
+
"learning_rate": 1.2e-05,
|
| 383 |
+
"loss": 0.4288,
|
| 384 |
+
"step": 240
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.024,
|
| 388 |
+
"eval_cos_sim": 0.5595658421516418,
|
| 389 |
+
"eval_loss": 0.44134216930615366,
|
| 390 |
+
"eval_runtime": 163.2069,
|
| 391 |
+
"eval_samples_per_second": 24.509,
|
| 392 |
+
"eval_steps_per_second": 0.386,
|
| 393 |
+
"step": 240
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.025,
|
| 397 |
+
"grad_norm": 1.7595293521881104,
|
| 398 |
+
"learning_rate": 1.25e-05,
|
| 399 |
+
"loss": 0.4365,
|
| 400 |
+
"step": 250
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.025,
|
| 404 |
+
"eval_cos_sim": 0.5746485590934753,
|
| 405 |
+
"eval_loss": 0.4262613997672743,
|
| 406 |
+
"eval_runtime": 160.8892,
|
| 407 |
+
"eval_samples_per_second": 24.862,
|
| 408 |
+
"eval_steps_per_second": 0.392,
|
| 409 |
+
"step": 250
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.026,
|
| 413 |
+
"grad_norm": 1.614985704421997,
|
| 414 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 415 |
+
"loss": 0.4007,
|
| 416 |
+
"step": 260
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.026,
|
| 420 |
+
"eval_cos_sim": 0.5887113809585571,
|
| 421 |
+
"eval_loss": 0.4121987219070143,
|
| 422 |
+
"eval_runtime": 157.6029,
|
| 423 |
+
"eval_samples_per_second": 25.38,
|
| 424 |
+
"eval_steps_per_second": 0.4,
|
| 425 |
+
"step": 260
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.027,
|
| 429 |
+
"grad_norm": 1.6005358695983887,
|
| 430 |
+
"learning_rate": 1.3500000000000001e-05,
|
| 431 |
+
"loss": 0.4048,
|
| 432 |
+
"step": 270
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.027,
|
| 436 |
+
"eval_cos_sim": 0.6018706560134888,
|
| 437 |
+
"eval_loss": 0.39904179241406385,
|
| 438 |
+
"eval_runtime": 163.2521,
|
| 439 |
+
"eval_samples_per_second": 24.502,
|
| 440 |
+
"eval_steps_per_second": 0.386,
|
| 441 |
+
"step": 270
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.028,
|
| 445 |
+
"grad_norm": 1.7216198444366455,
|
| 446 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 447 |
+
"loss": 0.3793,
|
| 448 |
+
"step": 280
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.028,
|
| 452 |
+
"eval_cos_sim": 0.6152481436729431,
|
| 453 |
+
"eval_loss": 0.385668321153993,
|
| 454 |
+
"eval_runtime": 163.9341,
|
| 455 |
+
"eval_samples_per_second": 24.4,
|
| 456 |
+
"eval_steps_per_second": 0.384,
|
| 457 |
+
"step": 280
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.029,
|
| 461 |
+
"grad_norm": 1.7449027299880981,
|
| 462 |
+
"learning_rate": 1.45e-05,
|
| 463 |
+
"loss": 0.3614,
|
| 464 |
+
"step": 290
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.029,
|
| 468 |
+
"eval_cos_sim": 0.6277655959129333,
|
| 469 |
+
"eval_loss": 0.3731525359367079,
|
| 470 |
+
"eval_runtime": 170.8176,
|
| 471 |
+
"eval_samples_per_second": 23.417,
|
| 472 |
+
"eval_steps_per_second": 0.369,
|
| 473 |
+
"step": 290
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.03,
|
| 477 |
+
"grad_norm": 1.5057954788208008,
|
| 478 |
+
"learning_rate": 1.5e-05,
|
| 479 |
+
"loss": 0.3557,
|
| 480 |
+
"step": 300
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.03,
|
| 484 |
+
"eval_cos_sim": 0.6390805244445801,
|
| 485 |
+
"eval_loss": 0.36183924534069956,
|
| 486 |
+
"eval_runtime": 160.3353,
|
| 487 |
+
"eval_samples_per_second": 24.948,
|
| 488 |
+
"eval_steps_per_second": 0.393,
|
| 489 |
+
"step": 300
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.031,
|
| 493 |
+
"grad_norm": 1.5245262384414673,
|
| 494 |
+
"learning_rate": 1.55e-05,
|
| 495 |
+
"loss": 0.3729,
|
| 496 |
+
"step": 310
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.031,
|
| 500 |
+
"eval_cos_sim": 0.6503457427024841,
|
| 501 |
+
"eval_loss": 0.3505764603828138,
|
| 502 |
+
"eval_runtime": 161.447,
|
| 503 |
+
"eval_samples_per_second": 24.776,
|
| 504 |
+
"eval_steps_per_second": 0.39,
|
| 505 |
+
"step": 310
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.032,
|
| 509 |
+
"grad_norm": 1.4508600234985352,
|
| 510 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 511 |
+
"loss": 0.3476,
|
| 512 |
+
"step": 320
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.032,
|
| 516 |
+
"eval_cos_sim": 0.664078414440155,
|
| 517 |
+
"eval_loss": 0.33684703972088753,
|
| 518 |
+
"eval_runtime": 162.7312,
|
| 519 |
+
"eval_samples_per_second": 24.58,
|
| 520 |
+
"eval_steps_per_second": 0.387,
|
| 521 |
+
"step": 320
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.033,
|
| 525 |
+
"grad_norm": 1.545847773551941,
|
| 526 |
+
"learning_rate": 1.65e-05,
|
| 527 |
+
"loss": 0.3335,
|
| 528 |
+
"step": 330
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.033,
|
| 532 |
+
"eval_cos_sim": 0.6734265089035034,
|
| 533 |
+
"eval_loss": 0.32750138094174325,
|
| 534 |
+
"eval_runtime": 162.4883,
|
| 535 |
+
"eval_samples_per_second": 24.617,
|
| 536 |
+
"eval_steps_per_second": 0.388,
|
| 537 |
+
"step": 330
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.034,
|
| 541 |
+
"grad_norm": 1.5199618339538574,
|
| 542 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 543 |
+
"loss": 0.3369,
|
| 544 |
+
"step": 340
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.034,
|
| 548 |
+
"eval_cos_sim": 0.6810038685798645,
|
| 549 |
+
"eval_loss": 0.31992570402371345,
|
| 550 |
+
"eval_runtime": 161.4031,
|
| 551 |
+
"eval_samples_per_second": 24.783,
|
| 552 |
+
"eval_steps_per_second": 0.39,
|
| 553 |
+
"step": 340
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 0.035,
|
| 557 |
+
"grad_norm": 1.2642650604248047,
|
| 558 |
+
"learning_rate": 1.75e-05,
|
| 559 |
+
"loss": 0.3396,
|
| 560 |
+
"step": 350
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.035,
|
| 564 |
+
"eval_cos_sim": 0.6911267638206482,
|
| 565 |
+
"eval_loss": 0.3098063874458021,
|
| 566 |
+
"eval_runtime": 165.9265,
|
| 567 |
+
"eval_samples_per_second": 24.107,
|
| 568 |
+
"eval_steps_per_second": 0.38,
|
| 569 |
+
"step": 350
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.036,
|
| 573 |
+
"grad_norm": 1.2891690731048584,
|
| 574 |
+
"learning_rate": 1.8e-05,
|
| 575 |
+
"loss": 0.3015,
|
| 576 |
+
"step": 360
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.036,
|
| 580 |
+
"eval_cos_sim": 0.6987318992614746,
|
| 581 |
+
"eval_loss": 0.3022033410285658,
|
| 582 |
+
"eval_runtime": 173.6837,
|
| 583 |
+
"eval_samples_per_second": 23.03,
|
| 584 |
+
"eval_steps_per_second": 0.363,
|
| 585 |
+
"step": 360
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.037,
|
| 589 |
+
"grad_norm": 1.2396866083145142,
|
| 590 |
+
"learning_rate": 1.85e-05,
|
| 591 |
+
"loss": 0.3077,
|
| 592 |
+
"step": 370
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.037,
|
| 596 |
+
"eval_cos_sim": 0.7082959413528442,
|
| 597 |
+
"eval_loss": 0.2926409912322706,
|
| 598 |
+
"eval_runtime": 158.0401,
|
| 599 |
+
"eval_samples_per_second": 25.31,
|
| 600 |
+
"eval_steps_per_second": 0.399,
|
| 601 |
+
"step": 370
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 0.038,
|
| 605 |
+
"grad_norm": 1.2055168151855469,
|
| 606 |
+
"learning_rate": 1.9e-05,
|
| 607 |
+
"loss": 0.2895,
|
| 608 |
+
"step": 380
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.038,
|
| 612 |
+
"eval_cos_sim": 0.7168179750442505,
|
| 613 |
+
"eval_loss": 0.284124677202577,
|
| 614 |
+
"eval_runtime": 160.1742,
|
| 615 |
+
"eval_samples_per_second": 24.973,
|
| 616 |
+
"eval_steps_per_second": 0.393,
|
| 617 |
+
"step": 380
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.039,
|
| 621 |
+
"grad_norm": 1.1191834211349487,
|
| 622 |
+
"learning_rate": 1.9500000000000003e-05,
|
| 623 |
+
"loss": 0.2944,
|
| 624 |
+
"step": 390
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.039,
|
| 628 |
+
"eval_cos_sim": 0.7234470248222351,
|
| 629 |
+
"eval_loss": 0.2775011363242811,
|
| 630 |
+
"eval_runtime": 165.4196,
|
| 631 |
+
"eval_samples_per_second": 24.181,
|
| 632 |
+
"eval_steps_per_second": 0.381,
|
| 633 |
+
"step": 390
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.04,
|
| 637 |
+
"grad_norm": 1.097105860710144,
|
| 638 |
+
"learning_rate": 2e-05,
|
| 639 |
+
"loss": 0.2898,
|
| 640 |
+
"step": 400
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.04,
|
| 644 |
+
"eval_cos_sim": 0.7314009666442871,
|
| 645 |
+
"eval_loss": 0.2695479884360975,
|
| 646 |
+
"eval_runtime": 159.5742,
|
| 647 |
+
"eval_samples_per_second": 25.067,
|
| 648 |
+
"eval_steps_per_second": 0.395,
|
| 649 |
+
"step": 400
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.041,
|
| 653 |
+
"grad_norm": 1.040711760520935,
|
| 654 |
+
"learning_rate": 2.05e-05,
|
| 655 |
+
"loss": 0.2573,
|
| 656 |
+
"step": 410
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.041,
|
| 660 |
+
"eval_cos_sim": 0.7389957904815674,
|
| 661 |
+
"eval_loss": 0.26195128538357676,
|
| 662 |
+
"eval_runtime": 160.3231,
|
| 663 |
+
"eval_samples_per_second": 24.95,
|
| 664 |
+
"eval_steps_per_second": 0.393,
|
| 665 |
+
"step": 410
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.042,
|
| 669 |
+
"grad_norm": 1.0322086811065674,
|
| 670 |
+
"learning_rate": 2.1e-05,
|
| 671 |
+
"loss": 0.241,
|
| 672 |
+
"step": 420
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 0.042,
|
| 676 |
+
"eval_cos_sim": 0.7471259236335754,
|
| 677 |
+
"eval_loss": 0.25382448627697884,
|
| 678 |
+
"eval_runtime": 161.5365,
|
| 679 |
+
"eval_samples_per_second": 24.762,
|
| 680 |
+
"eval_steps_per_second": 0.39,
|
| 681 |
+
"step": 420
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.043,
|
| 685 |
+
"grad_norm": 0.9718736410140991,
|
| 686 |
+
"learning_rate": 2.15e-05,
|
| 687 |
+
"loss": 0.26,
|
| 688 |
+
"step": 430
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.043,
|
| 692 |
+
"eval_cos_sim": 0.7534591555595398,
|
| 693 |
+
"eval_loss": 0.24748690702664317,
|
| 694 |
+
"eval_runtime": 169.2487,
|
| 695 |
+
"eval_samples_per_second": 23.634,
|
| 696 |
+
"eval_steps_per_second": 0.372,
|
| 697 |
+
"step": 430
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.044,
|
| 701 |
+
"grad_norm": 0.8780993223190308,
|
| 702 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 703 |
+
"loss": 0.2255,
|
| 704 |
+
"step": 440
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.044,
|
| 708 |
+
"eval_cos_sim": 0.7604101896286011,
|
| 709 |
+
"eval_loss": 0.2405370173667616,
|
| 710 |
+
"eval_runtime": 161.7354,
|
| 711 |
+
"eval_samples_per_second": 24.732,
|
| 712 |
+
"eval_steps_per_second": 0.39,
|
| 713 |
+
"step": 440
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.045,
|
| 717 |
+
"grad_norm": 0.8193601369857788,
|
| 718 |
+
"learning_rate": 2.25e-05,
|
| 719 |
+
"loss": 0.2182,
|
| 720 |
+
"step": 450
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.045,
|
| 724 |
+
"eval_cos_sim": 0.7673761248588562,
|
| 725 |
+
"eval_loss": 0.2335676951621718,
|
| 726 |
+
"eval_runtime": 159.2952,
|
| 727 |
+
"eval_samples_per_second": 25.111,
|
| 728 |
+
"eval_steps_per_second": 0.395,
|
| 729 |
+
"step": 450
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.046,
|
| 733 |
+
"grad_norm": 0.8820355534553528,
|
| 734 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 735 |
+
"loss": 0.2153,
|
| 736 |
+
"step": 460
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 0.046,
|
| 740 |
+
"eval_cos_sim": 0.7723870277404785,
|
| 741 |
+
"eval_loss": 0.228554349443788,
|
| 742 |
+
"eval_runtime": 161.8756,
|
| 743 |
+
"eval_samples_per_second": 24.71,
|
| 744 |
+
"eval_steps_per_second": 0.389,
|
| 745 |
+
"step": 460
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.047,
|
| 749 |
+
"grad_norm": 0.9045171141624451,
|
| 750 |
+
"learning_rate": 2.35e-05,
|
| 751 |
+
"loss": 0.2245,
|
| 752 |
+
"step": 470
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.047,
|
| 756 |
+
"eval_cos_sim": 0.7763780951499939,
|
| 757 |
+
"eval_loss": 0.22456564476239152,
|
| 758 |
+
"eval_runtime": 160.6855,
|
| 759 |
+
"eval_samples_per_second": 24.893,
|
| 760 |
+
"eval_steps_per_second": 0.392,
|
| 761 |
+
"step": 470
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.048,
|
| 765 |
+
"grad_norm": 0.7304402589797974,
|
| 766 |
+
"learning_rate": 2.4e-05,
|
| 767 |
+
"loss": 0.2069,
|
| 768 |
+
"step": 480
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.048,
|
| 772 |
+
"eval_cos_sim": 0.7830195426940918,
|
| 773 |
+
"eval_loss": 0.21792557956921524,
|
| 774 |
+
"eval_runtime": 162.6536,
|
| 775 |
+
"eval_samples_per_second": 24.592,
|
| 776 |
+
"eval_steps_per_second": 0.387,
|
| 777 |
+
"step": 480
|
| 778 |
+
}
|
| 779 |
+
],
|
| 780 |
+
"logging_steps": 10,
|
| 781 |
+
"max_steps": 10000,
|
| 782 |
+
"num_input_tokens_seen": 0,
|
| 783 |
+
"num_train_epochs": 9223372036854775807,
|
| 784 |
+
"save_steps": 10,
|
| 785 |
+
"stateful_callbacks": {
|
| 786 |
+
"TrainerControl": {
|
| 787 |
+
"args": {
|
| 788 |
+
"should_epoch_stop": false,
|
| 789 |
+
"should_evaluate": false,
|
| 790 |
+
"should_log": false,
|
| 791 |
+
"should_save": true,
|
| 792 |
+
"should_training_stop": false
|
| 793 |
+
},
|
| 794 |
+
"attributes": {}
|
| 795 |
+
}
|
| 796 |
+
},
|
| 797 |
+
"total_flos": 0.0,
|
| 798 |
+
"train_batch_size": 160,
|
| 799 |
+
"trial_name": null,
|
| 800 |
+
"trial_params": null
|
| 801 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:228943860476aef422f429bf676bbb7f95e79b4b2c6e641a22176011b8628860
|
| 3 |
+
size 5176
|