Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- checkpoints-d3.1/checkpoint-16384/eval_state.json +3 -0
- checkpoints-d3.1/checkpoint-16384/model.safetensors +3 -0
- checkpoints-d3.1/checkpoint-16384/optimizer.pt +3 -0
- checkpoints-d3.1/checkpoint-16384/rng_state.pth +3 -0
- checkpoints-d3.1/checkpoint-16384/scaler.pt +3 -0
- checkpoints-d3.1/checkpoint-16384/scheduler.pt +3 -0
- checkpoints-d3.1/checkpoint-16384/trainer_state.json +754 -0
- checkpoints-d3.1/checkpoint-16384/training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -84,3 +84,4 @@ checkpoints-d1.3/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs
|
|
| 84 |
checkpoints-d1.4/checkpoint-20480/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 85 |
checkpoints-d3.0/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 86 |
checkpoints-d3.1/checkpoint-8192/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 84 |
checkpoints-d1.4/checkpoint-20480/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 85 |
checkpoints-d3.0/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 86 |
checkpoints-d3.1/checkpoint-8192/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 87 |
+
checkpoints-d3.1/checkpoint-16384/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
checkpoints-d3.1/checkpoint-16384/eval_state.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ffac4d18aba5992c0f5faf9e8f08f1cfd7d54e34ffcb0e4b62c3d5a40b41e14
|
| 3 |
+
size 49167191
|
checkpoints-d3.1/checkpoint-16384/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd161bfdcddaa6b6bd07773a1ee838153a63f54a821fe9885d724a5e92728a47
|
| 3 |
+
size 32318104
|
checkpoints-d3.1/checkpoint-16384/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7ffeefe9703965210eb7fb3022b2f71b60eb297e751b6ef3e331555997c85433
|
| 3 |
+
size 10010635
|
checkpoints-d3.1/checkpoint-16384/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:113f4d47b8fa81f8482ed213c7b753551a1d58489dc79b30cae2e7914bd7f28f
|
| 3 |
+
size 14645
|
checkpoints-d3.1/checkpoint-16384/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a853d1eb1783b9d23f4749ba311d8645d1dd2f985cba56ad5f654b00bf5c0f9
|
| 3 |
+
size 1383
|
checkpoints-d3.1/checkpoint-16384/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:76fcb7a86d1f7aa11e64b46f7be2ae365f3eb77336967019eb5d88410e83c463
|
| 3 |
+
size 1465
|
checkpoints-d3.1/checkpoint-16384/trainer_state.json
ADDED
|
@@ -0,0 +1,754 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.7567317906794143,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 16384,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011823934229365849,
|
| 14 |
+
"grad_norm": 0.0007770129013806581,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 0.638597846031189,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.0036502168513834476,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 0.6255820393562317,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.00819552130997181,
|
| 29 |
+
"learning_rate": 0.000999640996023194,
|
| 30 |
+
"loss": 0.5880061388015747,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.014988946728408337,
|
| 36 |
+
"learning_rate": 0.0009985588674043958,
|
| 37 |
+
"loss": 0.5405965447425842,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_bleu": 0.020278462916643203,
|
| 43 |
+
"eval_ce_loss": 0.4954925418988755,
|
| 44 |
+
"eval_loss": 0.4954925418988755,
|
| 45 |
+
"step": 1024
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.047295736917463395,
|
| 49 |
+
"eval_bleu": 0.020278462916643203,
|
| 50 |
+
"eval_ce_loss": 0.4954925418988755,
|
| 51 |
+
"eval_loss": 0.4954925418988755,
|
| 52 |
+
"eval_runtime": 131.1195,
|
| 53 |
+
"eval_samples_per_second": 213.492,
|
| 54 |
+
"eval_steps_per_second": 3.34,
|
| 55 |
+
"step": 1024
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.05911967114682925,
|
| 59 |
+
"grad_norm": 0.018301010131835938,
|
| 60 |
+
"learning_rate": 0.0009967551747861387,
|
| 61 |
+
"loss": 0.4193720519542694,
|
| 62 |
+
"step": 1280
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.0709436053761951,
|
| 66 |
+
"grad_norm": 0.02350299246609211,
|
| 67 |
+
"learning_rate": 0.000994232528651847,
|
| 68 |
+
"loss": 0.32118111848831177,
|
| 69 |
+
"step": 1536
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.08276753960556095,
|
| 73 |
+
"grad_norm": 0.024243628606200218,
|
| 74 |
+
"learning_rate": 0.0009909945800260092,
|
| 75 |
+
"loss": 0.27625754475593567,
|
| 76 |
+
"step": 1792
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.09459147383492679,
|
| 80 |
+
"grad_norm": 0.022370202466845512,
|
| 81 |
+
"learning_rate": 0.0009870460151900522,
|
| 82 |
+
"loss": 0.25225532054901123,
|
| 83 |
+
"step": 2048
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.09459147383492679,
|
| 87 |
+
"eval_bleu": 0.16173067052738044,
|
| 88 |
+
"eval_ce_loss": 0.24119392792655997,
|
| 89 |
+
"eval_loss": 0.24119392792655997,
|
| 90 |
+
"step": 2048
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.09459147383492679,
|
| 94 |
+
"eval_bleu": 0.16173067052738044,
|
| 95 |
+
"eval_ce_loss": 0.24119392792655997,
|
| 96 |
+
"eval_loss": 0.24119392792655997,
|
| 97 |
+
"eval_runtime": 124.4827,
|
| 98 |
+
"eval_samples_per_second": 224.875,
|
| 99 |
+
"eval_steps_per_second": 3.519,
|
| 100 |
+
"step": 2048
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.10641540806429264,
|
| 104 |
+
"grad_norm": 0.025183985009789467,
|
| 105 |
+
"learning_rate": 0.0009823925488998885,
|
| 106 |
+
"loss": 0.23278099298477173,
|
| 107 |
+
"step": 2304
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.1182393422936585,
|
| 111 |
+
"grad_norm": 0.025138691067695618,
|
| 112 |
+
"learning_rate": 0.0009770409161149525,
|
| 113 |
+
"loss": 0.21929927170276642,
|
| 114 |
+
"step": 2560
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.13006327652302435,
|
| 118 |
+
"grad_norm": 0.02561408467590809,
|
| 119 |
+
"learning_rate": 0.0009709988622506973,
|
| 120 |
+
"loss": 0.20501913130283356,
|
| 121 |
+
"step": 2816
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1418872107523902,
|
| 125 |
+
"grad_norm": 0.028821036219596863,
|
| 126 |
+
"learning_rate": 0.000964275131968659,
|
| 127 |
+
"loss": 0.19426275789737701,
|
| 128 |
+
"step": 3072
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1418872107523902,
|
| 132 |
+
"eval_bleu": 0.19048343372542764,
|
| 133 |
+
"eval_ce_loss": 0.18918462373214226,
|
| 134 |
+
"eval_loss": 0.18918462373214226,
|
| 135 |
+
"step": 3072
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1418872107523902,
|
| 139 |
+
"eval_bleu": 0.19048343372542764,
|
| 140 |
+
"eval_ce_loss": 0.18918462373214226,
|
| 141 |
+
"eval_loss": 0.18918462373214226,
|
| 142 |
+
"eval_runtime": 125.1618,
|
| 143 |
+
"eval_samples_per_second": 223.654,
|
| 144 |
+
"eval_steps_per_second": 3.499,
|
| 145 |
+
"step": 3072
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.15371114498175603,
|
| 149 |
+
"grad_norm": 0.030421536415815353,
|
| 150 |
+
"learning_rate": 0.0009568794565203123,
|
| 151 |
+
"loss": 0.1839032918214798,
|
| 152 |
+
"step": 3328
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.1655350792111219,
|
| 156 |
+
"grad_norm": 0.032538577914237976,
|
| 157 |
+
"learning_rate": 0.0009488225396630347,
|
| 158 |
+
"loss": 0.17660538852214813,
|
| 159 |
+
"step": 3584
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.17735901344048774,
|
| 163 |
+
"grad_norm": 0.030282698571681976,
|
| 164 |
+
"learning_rate": 0.0009401160421685646,
|
| 165 |
+
"loss": 0.1695183515548706,
|
| 166 |
+
"step": 3840
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.18918294766985358,
|
| 170 |
+
"grad_norm": 0.029157549142837524,
|
| 171 |
+
"learning_rate": 0.0009307725649463714,
|
| 172 |
+
"loss": 0.16315339505672455,
|
| 173 |
+
"step": 4096
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.18918294766985358,
|
| 177 |
+
"eval_bleu": 0.20316795180629035,
|
| 178 |
+
"eval_ce_loss": 0.16114722188885353,
|
| 179 |
+
"eval_loss": 0.16114722188885353,
|
| 180 |
+
"step": 4096
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 0.18918294766985358,
|
| 184 |
+
"eval_bleu": 0.20316795180629035,
|
| 185 |
+
"eval_ce_loss": 0.16114722188885353,
|
| 186 |
+
"eval_loss": 0.16114722188885353,
|
| 187 |
+
"eval_runtime": 124.8908,
|
| 188 |
+
"eval_samples_per_second": 224.14,
|
| 189 |
+
"eval_steps_per_second": 3.507,
|
| 190 |
+
"step": 4096
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.20100688189921945,
|
| 194 |
+
"grad_norm": 0.033927738666534424,
|
| 195 |
+
"learning_rate": 0.0009208056308063659,
|
| 196 |
+
"loss": 0.15908151865005493,
|
| 197 |
+
"step": 4352
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.2128308161285853,
|
| 201 |
+
"grad_norm": 0.037180621176958084,
|
| 202 |
+
"learning_rate": 0.0009102296648873445,
|
| 203 |
+
"loss": 0.15225762128829956,
|
| 204 |
+
"step": 4608
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.22465475035795113,
|
| 208 |
+
"grad_norm": 0.037015024572610855,
|
| 209 |
+
"learning_rate": 0.0008990599737794927,
|
| 210 |
+
"loss": 0.1476851999759674,
|
| 211 |
+
"step": 4864
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.236478684587317,
|
| 215 |
+
"grad_norm": 0.03508929908275604,
|
| 216 |
+
"learning_rate": 0.0008873127233711644,
|
| 217 |
+
"loss": 0.14329124987125397,
|
| 218 |
+
"step": 5120
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.236478684587317,
|
| 222 |
+
"eval_bleu": 0.21566794230701403,
|
| 223 |
+
"eval_ce_loss": 0.14114715776418987,
|
| 224 |
+
"eval_loss": 0.14114715776418987,
|
| 225 |
+
"step": 5120
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.236478684587317,
|
| 229 |
+
"eval_bleu": 0.21566794230701403,
|
| 230 |
+
"eval_ce_loss": 0.14114715776418987,
|
| 231 |
+
"eval_loss": 0.14114715776418987,
|
| 232 |
+
"eval_runtime": 125.5711,
|
| 233 |
+
"eval_samples_per_second": 222.926,
|
| 234 |
+
"eval_steps_per_second": 3.488,
|
| 235 |
+
"step": 5120
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.24830261881668284,
|
| 239 |
+
"grad_norm": 0.03854870796203613,
|
| 240 |
+
"learning_rate": 0.0008750049154520011,
|
| 241 |
+
"loss": 0.14005360007286072,
|
| 242 |
+
"step": 5376
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.2601265530460487,
|
| 246 |
+
"grad_norm": 0.03188326954841614,
|
| 247 |
+
"learning_rate": 0.0008621543631062487,
|
| 248 |
+
"loss": 0.13561338186264038,
|
| 249 |
+
"step": 5632
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.27195048727541454,
|
| 253 |
+
"grad_norm": 0.03463684767484665,
|
| 254 |
+
"learning_rate": 0.0008487796649318904,
|
| 255 |
+
"loss": 0.1329711377620697,
|
| 256 |
+
"step": 5888
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.2837744215047804,
|
| 260 |
+
"grad_norm": 0.0358298160135746,
|
| 261 |
+
"learning_rate": 0.0008349001781229053,
|
| 262 |
+
"loss": 0.1301661729812622,
|
| 263 |
+
"step": 6144
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.2837744215047804,
|
| 267 |
+
"eval_bleu": 0.22173654323693984,
|
| 268 |
+
"eval_ce_loss": 0.12855397619000852,
|
| 269 |
+
"eval_loss": 0.12855397619000852,
|
| 270 |
+
"step": 6144
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.2837744215047804,
|
| 274 |
+
"eval_bleu": 0.22173654323693984,
|
| 275 |
+
"eval_ce_loss": 0.12855397619000852,
|
| 276 |
+
"eval_loss": 0.12855397619000852,
|
| 277 |
+
"eval_runtime": 126.9306,
|
| 278 |
+
"eval_samples_per_second": 220.538,
|
| 279 |
+
"eval_steps_per_second": 3.451,
|
| 280 |
+
"step": 6144
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.2955983557341462,
|
| 284 |
+
"grad_norm": 0.037201233208179474,
|
| 285 |
+
"learning_rate": 0.0008205359904536107,
|
| 286 |
+
"loss": 0.1271919459104538,
|
| 287 |
+
"step": 6400
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.30742228996351206,
|
| 291 |
+
"grad_norm": 0.0340813584625721,
|
| 292 |
+
"learning_rate": 0.0008057078912056363,
|
| 293 |
+
"loss": 0.12491822242736816,
|
| 294 |
+
"step": 6656
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.3192462241928779,
|
| 298 |
+
"grad_norm": 0.031119871884584427,
|
| 299 |
+
"learning_rate": 0.0007904373410796086,
|
| 300 |
+
"loss": 0.12252337485551834,
|
| 301 |
+
"step": 6912
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 0.3310701584222438,
|
| 305 |
+
"grad_norm": 0.036261022090911865,
|
| 306 |
+
"learning_rate": 0.0007747464411350876,
|
| 307 |
+
"loss": 0.12051380425691605,
|
| 308 |
+
"step": 7168
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.3310701584222438,
|
| 312 |
+
"eval_bleu": 0.228738826346753,
|
| 313 |
+
"eval_ce_loss": 0.12006643695169933,
|
| 314 |
+
"eval_loss": 0.12006643695169933,
|
| 315 |
+
"step": 7168
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.3310701584222438,
|
| 319 |
+
"eval_bleu": 0.228738826346753,
|
| 320 |
+
"eval_ce_loss": 0.12006643695169933,
|
| 321 |
+
"eval_loss": 0.12006643695169933,
|
| 322 |
+
"eval_runtime": 128.0461,
|
| 323 |
+
"eval_samples_per_second": 218.617,
|
| 324 |
+
"eval_steps_per_second": 3.421,
|
| 325 |
+
"step": 7168
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.34289409265160964,
|
| 329 |
+
"grad_norm": 0.03925757110118866,
|
| 330 |
+
"learning_rate": 0.000758657900803716,
|
| 331 |
+
"loss": 0.11830627918243408,
|
| 332 |
+
"step": 7424
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.3547180268809755,
|
| 336 |
+
"grad_norm": 0.03783626854419708,
|
| 337 |
+
"learning_rate": 0.000742195005021869,
|
| 338 |
+
"loss": 0.11684451997280121,
|
| 339 |
+
"step": 7680
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.3665419611103413,
|
| 343 |
+
"grad_norm": 0.04456119239330292,
|
| 344 |
+
"learning_rate": 0.0007253815805303786,
|
| 345 |
+
"loss": 0.11479435861110687,
|
| 346 |
+
"step": 7936
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.37836589533970716,
|
| 350 |
+
"grad_norm": 0.0406075082719326,
|
| 351 |
+
"learning_rate": 0.0007082419613901028,
|
| 352 |
+
"loss": 0.11298805475234985,
|
| 353 |
+
"step": 8192
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.37836589533970716,
|
| 357 |
+
"eval_bleu": 0.23316787828393487,
|
| 358 |
+
"eval_ce_loss": 0.11301100160501319,
|
| 359 |
+
"eval_loss": 0.11301100160501319,
|
| 360 |
+
"step": 8192
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.37836589533970716,
|
| 364 |
+
"eval_bleu": 0.23316787828393487,
|
| 365 |
+
"eval_ce_loss": 0.11301100160501319,
|
| 366 |
+
"eval_loss": 0.11301100160501319,
|
| 367 |
+
"eval_runtime": 129.7081,
|
| 368 |
+
"eval_samples_per_second": 215.815,
|
| 369 |
+
"eval_steps_per_second": 3.377,
|
| 370 |
+
"step": 8192
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.390189829569073,
|
| 374 |
+
"grad_norm": 0.04124804586172104,
|
| 375 |
+
"learning_rate": 0.0006908009537632514,
|
| 376 |
+
"loss": 0.11198760569095612,
|
| 377 |
+
"step": 8448
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.4020137637984389,
|
| 381 |
+
"grad_norm": 0.036579761654138565,
|
| 382 |
+
"learning_rate": 0.0006730838000114403,
|
| 383 |
+
"loss": 0.11078736931085587,
|
| 384 |
+
"step": 8704
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.41383769802780473,
|
| 388 |
+
"grad_norm": 0.042201053351163864,
|
| 389 |
+
"learning_rate": 0.0006551161421624341,
|
| 390 |
+
"loss": 0.10881925374269485,
|
| 391 |
+
"step": 8960
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.4256616322571706,
|
| 395 |
+
"grad_norm": 0.032566700130701065,
|
| 396 |
+
"learning_rate": 0.0006369239847984517,
|
| 397 |
+
"loss": 0.10761021077632904,
|
| 398 |
+
"step": 9216
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.4256616322571706,
|
| 402 |
+
"eval_bleu": 0.23706037291352922,
|
| 403 |
+
"eval_ce_loss": 0.10713473501610973,
|
| 404 |
+
"eval_loss": 0.10713473501610973,
|
| 405 |
+
"step": 9216
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.4256616322571706,
|
| 409 |
+
"eval_bleu": 0.23706037291352922,
|
| 410 |
+
"eval_ce_loss": 0.10713473501610973,
|
| 411 |
+
"eval_loss": 0.10713473501610973,
|
| 412 |
+
"eval_runtime": 136.316,
|
| 413 |
+
"eval_samples_per_second": 205.354,
|
| 414 |
+
"eval_steps_per_second": 3.213,
|
| 415 |
+
"step": 9216
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.4374855664865364,
|
| 419 |
+
"grad_norm": 0.03181539848446846,
|
| 420 |
+
"learning_rate": 0.0006185336574197479,
|
| 421 |
+
"loss": 0.10633238404989243,
|
| 422 |
+
"step": 9472
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.44930950071590225,
|
| 426 |
+
"grad_norm": 0.039569102227687836,
|
| 427 |
+
"learning_rate": 0.0005999717763379407,
|
| 428 |
+
"loss": 0.10582899302244186,
|
| 429 |
+
"step": 9728
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.4611334349452681,
|
| 433 |
+
"grad_norm": 0.03662619739770889,
|
| 434 |
+
"learning_rate": 0.0005812652061542363,
|
| 435 |
+
"loss": 0.10484181344509125,
|
| 436 |
+
"step": 9984
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.472957369174634,
|
| 440 |
+
"grad_norm": 0.037710174918174744,
|
| 441 |
+
"learning_rate": 0.0005624410208783071,
|
| 442 |
+
"loss": 0.10320662707090378,
|
| 443 |
+
"step": 10240
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.472957369174634,
|
| 447 |
+
"eval_bleu": 0.2394408170470903,
|
| 448 |
+
"eval_ce_loss": 0.10344514005804716,
|
| 449 |
+
"eval_loss": 0.10344514005804716,
|
| 450 |
+
"step": 10240
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.472957369174634,
|
| 454 |
+
"eval_bleu": 0.2394408170470903,
|
| 455 |
+
"eval_ce_loss": 0.10344514005804716,
|
| 456 |
+
"eval_loss": 0.10344514005804716,
|
| 457 |
+
"eval_runtime": 130.5108,
|
| 458 |
+
"eval_samples_per_second": 214.488,
|
| 459 |
+
"eval_steps_per_second": 3.356,
|
| 460 |
+
"step": 10240
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 0.48478130340399983,
|
| 464 |
+
"grad_norm": 0.035840343683958054,
|
| 465 |
+
"learning_rate": 0.0005435264647440881,
|
| 466 |
+
"loss": 0.10283803939819336,
|
| 467 |
+
"step": 10496
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.49660523763336567,
|
| 471 |
+
"grad_norm": 0.04071688652038574,
|
| 472 |
+
"learning_rate": 0.000524548912779213,
|
| 473 |
+
"loss": 0.1021987721323967,
|
| 474 |
+
"step": 10752
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.5084291718627315,
|
| 478 |
+
"grad_norm": 0.036369238048791885,
|
| 479 |
+
"learning_rate": 0.0005055358311851499,
|
| 480 |
+
"loss": 0.10196679085493088,
|
| 481 |
+
"step": 11008
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 0.5202531060920974,
|
| 485 |
+
"grad_norm": 0.035920336842536926,
|
| 486 |
+
"learning_rate": 0.0004865147375853812,
|
| 487 |
+
"loss": 0.10061318427324295,
|
| 488 |
+
"step": 11264
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.5202531060920974,
|
| 492 |
+
"eval_bleu": 0.2421649824093764,
|
| 493 |
+
"eval_ce_loss": 0.09971236017400815,
|
| 494 |
+
"eval_loss": 0.09971236017400815,
|
| 495 |
+
"step": 11264
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.5202531060920974,
|
| 499 |
+
"eval_bleu": 0.2421649824093764,
|
| 500 |
+
"eval_ce_loss": 0.09971236017400815,
|
| 501 |
+
"eval_loss": 0.09971236017400815,
|
| 502 |
+
"eval_runtime": 124.8326,
|
| 503 |
+
"eval_samples_per_second": 224.244,
|
| 504 |
+
"eval_steps_per_second": 3.509,
|
| 505 |
+
"step": 11264
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.5320770403214632,
|
| 509 |
+
"grad_norm": 0.035242412239313126,
|
| 510 |
+
"learning_rate": 0.0004675131611991607,
|
| 511 |
+
"loss": 0.10020274668931961,
|
| 512 |
+
"step": 11520
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.5439009745508291,
|
| 516 |
+
"grad_norm": 0.03995301201939583,
|
| 517 |
+
"learning_rate": 0.0004485586029984899,
|
| 518 |
+
"loss": 0.09980272501707077,
|
| 519 |
+
"step": 11776
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.5557249087801949,
|
| 523 |
+
"grad_norm": 0.0381169356405735,
|
| 524 |
+
"learning_rate": 0.00042967849590597266,
|
| 525 |
+
"loss": 0.09913930296897888,
|
| 526 |
+
"step": 12032
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.5675488430095608,
|
| 530 |
+
"grad_norm": 0.03928132355213165,
|
| 531 |
+
"learning_rate": 0.0004109001650911621,
|
| 532 |
+
"loss": 0.09818586707115173,
|
| 533 |
+
"step": 12288
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.5675488430095608,
|
| 537 |
+
"eval_bleu": 0.24165027099086678,
|
| 538 |
+
"eval_ce_loss": 0.09799085275087183,
|
| 539 |
+
"eval_loss": 0.09799085275087183,
|
| 540 |
+
"step": 12288
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.5675488430095608,
|
| 544 |
+
"eval_bleu": 0.24165027099086678,
|
| 545 |
+
"eval_ce_loss": 0.09799085275087183,
|
| 546 |
+
"eval_loss": 0.09799085275087183,
|
| 547 |
+
"eval_runtime": 128.0046,
|
| 548 |
+
"eval_samples_per_second": 218.687,
|
| 549 |
+
"eval_steps_per_second": 3.422,
|
| 550 |
+
"step": 12288
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.5793727772389267,
|
| 554 |
+
"grad_norm": 0.038329754024744034,
|
| 555 |
+
"learning_rate": 0.0003922507884228551,
|
| 556 |
+
"loss": 0.09753931313753128,
|
| 557 |
+
"step": 12544
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.5911967114682924,
|
| 561 |
+
"grad_norm": 0.03905988112092018,
|
| 562 |
+
"learning_rate": 0.00037375735713457723,
|
| 563 |
+
"loss": 0.09739061444997787,
|
| 564 |
+
"step": 12800
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.6030206456976583,
|
| 568 |
+
"grad_norm": 0.03997718170285225,
|
| 569 |
+
"learning_rate": 0.00035544663676018276,
|
| 570 |
+
"loss": 0.09687047451734543,
|
| 571 |
+
"step": 13056
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.6148445799270241,
|
| 575 |
+
"grad_norm": 0.037294477224349976,
|
| 576 |
+
"learning_rate": 0.00033734512839611255,
|
| 577 |
+
"loss": 0.09683524817228317,
|
| 578 |
+
"step": 13312
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.6148445799270241,
|
| 582 |
+
"eval_bleu": 0.244063034514669,
|
| 583 |
+
"eval_ce_loss": 0.0961443424803209,
|
| 584 |
+
"eval_loss": 0.0961443424803209,
|
| 585 |
+
"step": 13312
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.6148445799270241,
|
| 589 |
+
"eval_bleu": 0.244063034514669,
|
| 590 |
+
"eval_ce_loss": 0.0961443424803209,
|
| 591 |
+
"eval_loss": 0.0961443424803209,
|
| 592 |
+
"eval_runtime": 127.2652,
|
| 593 |
+
"eval_samples_per_second": 219.958,
|
| 594 |
+
"eval_steps_per_second": 3.442,
|
| 595 |
+
"step": 13312
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 0.62666851415639,
|
| 599 |
+
"grad_norm": 0.04453560709953308,
|
| 600 |
+
"learning_rate": 0.0003194790303463687,
|
| 601 |
+
"loss": 0.09584437310695648,
|
| 602 |
+
"step": 13568
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.6384924483857558,
|
| 606 |
+
"grad_norm": 0.03323407843708992,
|
| 607 |
+
"learning_rate": 0.00030187420020572406,
|
| 608 |
+
"loss": 0.09573940187692642,
|
| 609 |
+
"step": 13824
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.6503163826151217,
|
| 613 |
+
"grad_norm": 0.03651966527104378,
|
| 614 |
+
"learning_rate": 0.00028455611743603626,
|
| 615 |
+
"loss": 0.09577071666717529,
|
| 616 |
+
"step": 14080
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 0.6621403168444876,
|
| 620 |
+
"grad_norm": 0.03129328042268753,
|
| 621 |
+
"learning_rate": 0.0002675498464898373,
|
| 622 |
+
"loss": 0.09538843482732773,
|
| 623 |
+
"step": 14336
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 0.6621403168444876,
|
| 627 |
+
"eval_bleu": 0.24573789493234413,
|
| 628 |
+
"eval_ce_loss": 0.09470595116620739,
|
| 629 |
+
"eval_loss": 0.09470595116620739,
|
| 630 |
+
"step": 14336
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 0.6621403168444876,
|
| 634 |
+
"eval_bleu": 0.24573789493234413,
|
| 635 |
+
"eval_ce_loss": 0.09470595116620739,
|
| 636 |
+
"eval_loss": 0.09470595116620739,
|
| 637 |
+
"eval_runtime": 128.6793,
|
| 638 |
+
"eval_samples_per_second": 217.541,
|
| 639 |
+
"eval_steps_per_second": 3.404,
|
| 640 |
+
"step": 14336
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.6739642510738534,
|
| 644 |
+
"grad_norm": 0.03330269455909729,
|
| 645 |
+
"learning_rate": 0.0002508800005345623,
|
| 646 |
+
"loss": 0.09491758048534393,
|
| 647 |
+
"step": 14592
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.6857881853032193,
|
| 651 |
+
"grad_norm": 0.03013775125145912,
|
| 652 |
+
"learning_rate": 0.00023457070582992562,
|
| 653 |
+
"loss": 0.09489757567644119,
|
| 654 |
+
"step": 14848
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.6976121195325851,
|
| 658 |
+
"grad_norm": 0.03159448131918907,
|
| 659 |
+
"learning_rate": 0.00021864556680999692,
|
| 660 |
+
"loss": 0.09417803585529327,
|
| 661 |
+
"step": 15104
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.709436053761951,
|
| 665 |
+
"grad_norm": 0.028872501105070114,
|
| 666 |
+
"learning_rate": 0.0002031276319205152,
|
| 667 |
+
"loss": 0.094399593770504,
|
| 668 |
+
"step": 15360
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.709436053761951,
|
| 672 |
+
"eval_bleu": 0.242515567114501,
|
| 673 |
+
"eval_ce_loss": 0.09441615134205449,
|
| 674 |
+
"eval_loss": 0.09441615134205449,
|
| 675 |
+
"step": 15360
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.709436053761951,
|
| 679 |
+
"eval_bleu": 0.242515567114501,
|
| 680 |
+
"eval_ce_loss": 0.09441615134205449,
|
| 681 |
+
"eval_loss": 0.09441615134205449,
|
| 682 |
+
"eval_runtime": 128.6332,
|
| 683 |
+
"eval_samples_per_second": 217.619,
|
| 684 |
+
"eval_steps_per_second": 3.405,
|
| 685 |
+
"step": 15360
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.7212599879913169,
|
| 689 |
+
"grad_norm": 0.03194403648376465,
|
| 690 |
+
"learning_rate": 0.00018803936026088542,
|
| 691 |
+
"loss": 0.09390134364366531,
|
| 692 |
+
"step": 15616
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.7330839222206826,
|
| 696 |
+
"grad_norm": 0.03124140575528145,
|
| 697 |
+
"learning_rate": 0.00017340258907913464,
|
| 698 |
+
"loss": 0.09355755150318146,
|
| 699 |
+
"step": 15872
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.7449078564500485,
|
| 703 |
+
"grad_norm": 0.027917077764868736,
|
| 704 |
+
"learning_rate": 0.0001592385021668743,
|
| 705 |
+
"loss": 0.09316008538007736,
|
| 706 |
+
"step": 16128
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.7567317906794143,
|
| 710 |
+
"grad_norm": 0.02921847254037857,
|
| 711 |
+
"learning_rate": 0.0001455675992000087,
|
| 712 |
+
"loss": 0.09386321902275085,
|
| 713 |
+
"step": 16384
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.7567317906794143,
|
| 717 |
+
"eval_bleu": 0.24441726952160492,
|
| 718 |
+
"eval_ce_loss": 0.0937359631061554,
|
| 719 |
+
"eval_loss": 0.0937359631061554,
|
| 720 |
+
"step": 16384
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.7567317906794143,
|
| 724 |
+
"eval_bleu": 0.24441726952160492,
|
| 725 |
+
"eval_ce_loss": 0.0937359631061554,
|
| 726 |
+
"eval_loss": 0.0937359631061554,
|
| 727 |
+
"eval_runtime": 130.1985,
|
| 728 |
+
"eval_samples_per_second": 215.002,
|
| 729 |
+
"eval_steps_per_second": 3.364,
|
| 730 |
+
"step": 16384
|
| 731 |
+
}
|
| 732 |
+
],
|
| 733 |
+
"logging_steps": 256,
|
| 734 |
+
"max_steps": 21651,
|
| 735 |
+
"num_input_tokens_seen": 0,
|
| 736 |
+
"num_train_epochs": 1,
|
| 737 |
+
"save_steps": 1024,
|
| 738 |
+
"stateful_callbacks": {
|
| 739 |
+
"TrainerControl": {
|
| 740 |
+
"args": {
|
| 741 |
+
"should_epoch_stop": false,
|
| 742 |
+
"should_evaluate": false,
|
| 743 |
+
"should_log": false,
|
| 744 |
+
"should_save": true,
|
| 745 |
+
"should_training_stop": false
|
| 746 |
+
},
|
| 747 |
+
"attributes": {}
|
| 748 |
+
}
|
| 749 |
+
},
|
| 750 |
+
"total_flos": 0.0,
|
| 751 |
+
"train_batch_size": 64,
|
| 752 |
+
"trial_name": null,
|
| 753 |
+
"trial_params": null
|
| 754 |
+
}
|
checkpoints-d3.1/checkpoint-16384/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61a16bb839f687414b8e48611327c4b9cfddeefe38c031ca70808f9a97c476b7
|
| 3 |
+
size 5137
|