Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- checkpoints-v2/checkpoint-12288/eval_state.json +3 -0
- checkpoints-v2/checkpoint-12288/model.safetensors +3 -0
- checkpoints-v2/checkpoint-12288/optimizer.pt +3 -0
- checkpoints-v2/checkpoint-12288/rng_state.pth +3 -0
- checkpoints-v2/checkpoint-12288/scaler.pt +3 -0
- checkpoints-v2/checkpoint-12288/scheduler.pt +3 -0
- checkpoints-v2/checkpoint-12288/trainer_state.json +742 -0
- checkpoints-v2/checkpoint-12288/training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -34,3 +34,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
checkpoints/checkpoint-4096/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
checkpoints/checkpoint-4096/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
checkpoints-v2/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
checkpoints-v2/checkpoint-12288/eval_state.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0cf110e8f9defe21952aa74d2f942936cc1463bc3ea3a4ca960b3bf6ce4ce01
|
| 3 |
+
size 44791006
|
checkpoints-v2/checkpoint-12288/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2be6124b598c6d326fee282b30336e861057de54d73b75b247d610e59498626f
|
| 3 |
+
size 37402680
|
checkpoints-v2/checkpoint-12288/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0c10552fc0d3f58ede2ad0f7b0c827a12ebaf931f9c1cbceb648e88b243cb16b
|
| 3 |
+
size 512267
|
checkpoints-v2/checkpoint-12288/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9f36e6c28b77555b6da6de84681647b558ac8ebc553a1b458e45112e416a213c
|
| 3 |
+
size 14645
|
checkpoints-v2/checkpoint-12288/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7ffa55f380ad889d60cf02c4b3fd1de071b4f6c3761db430565fff2faa03051d
|
| 3 |
+
size 1383
|
checkpoints-v2/checkpoint-12288/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61352dad588adf3c8b7c226a2fc0b9396b5d25813338cf48e50e2cba3dc1f3cb
|
| 3 |
+
size 1465
|
checkpoints-v2/checkpoint-12288/trainer_state.json
ADDED
|
@@ -0,0 +1,742 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.5675488430095608,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 12288,
|
| 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": 1.1337779760360718,
|
| 15 |
+
"learning_rate": 1.9615384615384617e-05,
|
| 16 |
+
"loss": 10.3487,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 1.2268586158752441,
|
| 22 |
+
"learning_rate": 3.930769230769231e-05,
|
| 23 |
+
"loss": 7.8944,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 1.0524195432662964,
|
| 29 |
+
"learning_rate": 4.999617095521894e-05,
|
| 30 |
+
"loss": 5.6243,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.627362072467804,
|
| 36 |
+
"learning_rate": 4.9961092368776736e-05,
|
| 37 |
+
"loss": 3.7534,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_across_var": 0.0030250440715263558,
|
| 43 |
+
"eval_bleu": 0.5480678351145061,
|
| 44 |
+
"eval_ce_loss": 2.361846002813888,
|
| 45 |
+
"eval_cos_loss": 0.9333138501263101,
|
| 46 |
+
"eval_cov": 0.07032504147046233,
|
| 47 |
+
"eval_global_var": 0.2857983732876712,
|
| 48 |
+
"eval_loss": 2.7384781478202505,
|
| 49 |
+
"eval_mse_loss": 1.9327595538744644,
|
| 50 |
+
"eval_per_var": 0.2780265633918379,
|
| 51 |
+
"eval_within_var": 0.2828233463170866,
|
| 52 |
+
"step": 1024
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.047295736917463395,
|
| 56 |
+
"eval_across_var": 0.0030250440715263558,
|
| 57 |
+
"eval_bleu": 0.5480678351145061,
|
| 58 |
+
"eval_ce_loss": 2.361846002813888,
|
| 59 |
+
"eval_cos_loss": 0.9333138501263101,
|
| 60 |
+
"eval_cov": 0.07032504147046233,
|
| 61 |
+
"eval_global_var": 0.2857983732876712,
|
| 62 |
+
"eval_loss": 2.7384781478202505,
|
| 63 |
+
"eval_mse_loss": 1.9327595538744644,
|
| 64 |
+
"eval_per_var": 0.2780265633918379,
|
| 65 |
+
"eval_runtime": 151.7756,
|
| 66 |
+
"eval_samples_per_second": 184.437,
|
| 67 |
+
"eval_steps_per_second": 2.886,
|
| 68 |
+
"eval_within_var": 0.2828233463170866,
|
| 69 |
+
"step": 1024
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.05911967114682925,
|
| 73 |
+
"grad_norm": 0.3857733905315399,
|
| 74 |
+
"learning_rate": 4.988941132556799e-05,
|
| 75 |
+
"loss": 2.6721,
|
| 76 |
+
"step": 1280
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.0709436053761951,
|
| 80 |
+
"grad_norm": 0.3293663561344147,
|
| 81 |
+
"learning_rate": 4.9781232937269974e-05,
|
| 82 |
+
"loss": 2.0559,
|
| 83 |
+
"step": 1536
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.08276753960556095,
|
| 87 |
+
"grad_norm": 0.2576558291912079,
|
| 88 |
+
"learning_rate": 4.963671583455164e-05,
|
| 89 |
+
"loss": 1.6534,
|
| 90 |
+
"step": 1792
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.09459147383492679,
|
| 94 |
+
"grad_norm": 0.24152210354804993,
|
| 95 |
+
"learning_rate": 4.945607193446079e-05,
|
| 96 |
+
"loss": 1.3739,
|
| 97 |
+
"step": 2048
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.09459147383492679,
|
| 101 |
+
"eval_across_var": 0.0044681047000093,
|
| 102 |
+
"eval_bleu": 0.6452645235714047,
|
| 103 |
+
"eval_ce_loss": 0.7001769141247284,
|
| 104 |
+
"eval_cos_loss": 0.7716340200813938,
|
| 105 |
+
"eval_cov": 0.07174250633204908,
|
| 106 |
+
"eval_global_var": 0.4126723476740868,
|
| 107 |
+
"eval_loss": 1.0148827189996363,
|
| 108 |
+
"eval_mse_loss": 1.6643117905751754,
|
| 109 |
+
"eval_per_var": 0.4014117793949772,
|
| 110 |
+
"eval_within_var": 0.40839217984240894,
|
| 111 |
+
"step": 2048
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.09459147383492679,
|
| 115 |
+
"eval_across_var": 0.0044681047000093,
|
| 116 |
+
"eval_bleu": 0.6452645235714047,
|
| 117 |
+
"eval_ce_loss": 0.7001769141247284,
|
| 118 |
+
"eval_cos_loss": 0.7716340200813938,
|
| 119 |
+
"eval_cov": 0.07174250633204908,
|
| 120 |
+
"eval_global_var": 0.4126723476740868,
|
| 121 |
+
"eval_loss": 1.0148827189996363,
|
| 122 |
+
"eval_mse_loss": 1.6643117905751754,
|
| 123 |
+
"eval_per_var": 0.4014117793949772,
|
| 124 |
+
"eval_runtime": 150.792,
|
| 125 |
+
"eval_samples_per_second": 185.64,
|
| 126 |
+
"eval_steps_per_second": 2.905,
|
| 127 |
+
"eval_within_var": 0.40839217984240894,
|
| 128 |
+
"step": 2048
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.10641540806429264,
|
| 132 |
+
"grad_norm": 0.18892866373062134,
|
| 133 |
+
"learning_rate": 4.923956612967301e-05,
|
| 134 |
+
"loss": 1.1708,
|
| 135 |
+
"step": 2304
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1182393422936585,
|
| 139 |
+
"grad_norm": 0.18101266026496887,
|
| 140 |
+
"learning_rate": 4.898751590005826e-05,
|
| 141 |
+
"loss": 1.0095,
|
| 142 |
+
"step": 2560
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.13006327652302435,
|
| 146 |
+
"grad_norm": 0.15639935433864594,
|
| 147 |
+
"learning_rate": 4.870029084713462e-05,
|
| 148 |
+
"loss": 0.8841,
|
| 149 |
+
"step": 2816
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.1418872107523902,
|
| 153 |
+
"grad_norm": 0.14697179198265076,
|
| 154 |
+
"learning_rate": 4.837831215209188e-05,
|
| 155 |
+
"loss": 0.7841,
|
| 156 |
+
"step": 3072
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1418872107523902,
|
| 160 |
+
"eval_across_var": 0.0055386394200650935,
|
| 161 |
+
"eval_bleu": 0.7978306727589363,
|
| 162 |
+
"eval_ce_loss": 0.32526156282452146,
|
| 163 |
+
"eval_cos_loss": 0.6155619193974151,
|
| 164 |
+
"eval_cov": 0.0707132696561073,
|
| 165 |
+
"eval_global_var": 0.508560529038242,
|
| 166 |
+
"eval_loss": 0.5787542584143817,
|
| 167 |
+
"eval_mse_loss": 1.3764822866818676,
|
| 168 |
+
"eval_per_var": 0.4943646992722603,
|
| 169 |
+
"eval_within_var": 0.5033302515336912,
|
| 170 |
+
"step": 3072
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.1418872107523902,
|
| 174 |
+
"eval_across_var": 0.0055386394200650935,
|
| 175 |
+
"eval_bleu": 0.7978306727589363,
|
| 176 |
+
"eval_ce_loss": 0.32526156282452146,
|
| 177 |
+
"eval_cos_loss": 0.6155619193974151,
|
| 178 |
+
"eval_cov": 0.0707132696561073,
|
| 179 |
+
"eval_global_var": 0.508560529038242,
|
| 180 |
+
"eval_loss": 0.5787542584143817,
|
| 181 |
+
"eval_mse_loss": 1.3764822866818676,
|
| 182 |
+
"eval_per_var": 0.4943646992722603,
|
| 183 |
+
"eval_runtime": 150.3038,
|
| 184 |
+
"eval_samples_per_second": 186.243,
|
| 185 |
+
"eval_steps_per_second": 2.914,
|
| 186 |
+
"eval_within_var": 0.5033302515336912,
|
| 187 |
+
"step": 3072
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 0.15371114498175603,
|
| 191 |
+
"grad_norm": 0.13465629518032074,
|
| 192 |
+
"learning_rate": 4.802205195817963e-05,
|
| 193 |
+
"loss": 0.7042,
|
| 194 |
+
"step": 3328
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.1655350792111219,
|
| 198 |
+
"grad_norm": 0.12159380316734314,
|
| 199 |
+
"learning_rate": 4.763203267836576e-05,
|
| 200 |
+
"loss": 0.6359,
|
| 201 |
+
"step": 3584
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.17735901344048774,
|
| 205 |
+
"grad_norm": 0.11759158223867416,
|
| 206 |
+
"learning_rate": 4.720882622928019e-05,
|
| 207 |
+
"loss": 0.5793,
|
| 208 |
+
"step": 3840
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.18918294766985358,
|
| 212 |
+
"grad_norm": 0.11546578258275986,
|
| 213 |
+
"learning_rate": 4.675305319256765e-05,
|
| 214 |
+
"loss": 0.5323,
|
| 215 |
+
"step": 4096
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.18918294766985358,
|
| 219 |
+
"eval_across_var": 0.006188327843136235,
|
| 220 |
+
"eval_bleu": 0.830698384342328,
|
| 221 |
+
"eval_ce_loss": 0.18860791746067673,
|
| 222 |
+
"eval_cos_loss": 0.4954083457793275,
|
| 223 |
+
"eval_cov": 0.0701905690371718,
|
| 224 |
+
"eval_global_var": 0.574416069135274,
|
| 225 |
+
"eval_loss": 0.3946296600582393,
|
| 226 |
+
"eval_mse_loss": 1.1479846312575144,
|
| 227 |
+
"eval_per_var": 0.558419841609589,
|
| 228 |
+
"eval_within_var": 0.5686683095481297,
|
| 229 |
+
"step": 4096
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.18918294766985358,
|
| 233 |
+
"eval_across_var": 0.006188327843136235,
|
| 234 |
+
"eval_bleu": 0.830698384342328,
|
| 235 |
+
"eval_ce_loss": 0.18860791746067673,
|
| 236 |
+
"eval_cos_loss": 0.4954083457793275,
|
| 237 |
+
"eval_cov": 0.0701905690371718,
|
| 238 |
+
"eval_global_var": 0.574416069135274,
|
| 239 |
+
"eval_loss": 0.3946296600582393,
|
| 240 |
+
"eval_mse_loss": 1.1479846312575144,
|
| 241 |
+
"eval_per_var": 0.558419841609589,
|
| 242 |
+
"eval_runtime": 149.9304,
|
| 243 |
+
"eval_samples_per_second": 186.707,
|
| 244 |
+
"eval_steps_per_second": 2.921,
|
| 245 |
+
"eval_within_var": 0.5686683095481297,
|
| 246 |
+
"step": 4096
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.20100688189921945,
|
| 250 |
+
"grad_norm": 0.11826465278863907,
|
| 251 |
+
"learning_rate": 4.6265381904878854e-05,
|
| 252 |
+
"loss": 0.4896,
|
| 253 |
+
"step": 4352
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.2128308161285853,
|
| 257 |
+
"grad_norm": 0.10021153837442398,
|
| 258 |
+
"learning_rate": 4.57465274778347e-05,
|
| 259 |
+
"loss": 0.4576,
|
| 260 |
+
"step": 4608
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.22465475035795113,
|
| 264 |
+
"grad_norm": 0.10055450350046158,
|
| 265 |
+
"learning_rate": 4.519725074940068e-05,
|
| 266 |
+
"loss": 0.4235,
|
| 267 |
+
"step": 4864
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.236478684587317,
|
| 271 |
+
"grad_norm": 0.0877346396446228,
|
| 272 |
+
"learning_rate": 4.461835716820895e-05,
|
| 273 |
+
"loss": 0.3965,
|
| 274 |
+
"step": 5120
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.236478684587317,
|
| 278 |
+
"eval_across_var": 0.006611716227614444,
|
| 279 |
+
"eval_bleu": 0.9513684417284821,
|
| 280 |
+
"eval_ce_loss": 0.12323284007983121,
|
| 281 |
+
"eval_cos_loss": 0.4073775019411627,
|
| 282 |
+
"eval_cov": 0.06992845230450913,
|
| 283 |
+
"eval_global_var": 0.621496191852169,
|
| 284 |
+
"eval_loss": 0.2943976874969321,
|
| 285 |
+
"eval_mse_loss": 0.9790318277600694,
|
| 286 |
+
"eval_per_var": 0.604084171660959,
|
| 287 |
+
"eval_within_var": 0.615415731113251,
|
| 288 |
+
"step": 5120
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.236478684587317,
|
| 292 |
+
"eval_across_var": 0.006611716227614444,
|
| 293 |
+
"eval_bleu": 0.9513684417284821,
|
| 294 |
+
"eval_ce_loss": 0.12323284007983121,
|
| 295 |
+
"eval_cos_loss": 0.4073775019411627,
|
| 296 |
+
"eval_cov": 0.06992845230450913,
|
| 297 |
+
"eval_global_var": 0.621496191852169,
|
| 298 |
+
"eval_loss": 0.2943976874969321,
|
| 299 |
+
"eval_mse_loss": 0.9790318277600694,
|
| 300 |
+
"eval_per_var": 0.604084171660959,
|
| 301 |
+
"eval_runtime": 150.6714,
|
| 302 |
+
"eval_samples_per_second": 185.788,
|
| 303 |
+
"eval_steps_per_second": 2.907,
|
| 304 |
+
"eval_within_var": 0.615415731113251,
|
| 305 |
+
"step": 5120
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.24830261881668284,
|
| 309 |
+
"grad_norm": 0.0914112776517868,
|
| 310 |
+
"learning_rate": 4.401069561246422e-05,
|
| 311 |
+
"loss": 0.3725,
|
| 312 |
+
"step": 5376
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.2601265530460487,
|
| 316 |
+
"grad_norm": 0.10334540158510208,
|
| 317 |
+
"learning_rate": 4.337515714516545e-05,
|
| 318 |
+
"loss": 0.3507,
|
| 319 |
+
"step": 5632
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 0.27195048727541454,
|
| 323 |
+
"grad_norm": 0.09348654747009277,
|
| 324 |
+
"learning_rate": 4.2712673707468434e-05,
|
| 325 |
+
"loss": 0.335,
|
| 326 |
+
"step": 5888
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.2837744215047804,
|
| 330 |
+
"grad_norm": 0.08614271879196167,
|
| 331 |
+
"learning_rate": 4.202421675210565e-05,
|
| 332 |
+
"loss": 0.3182,
|
| 333 |
+
"step": 6144
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.2837744215047804,
|
| 337 |
+
"eval_across_var": 0.006925537744239313,
|
| 338 |
+
"eval_bleu": 0.9630394430269803,
|
| 339 |
+
"eval_ce_loss": 0.08711560829181104,
|
| 340 |
+
"eval_cos_loss": 0.34602166433312576,
|
| 341 |
+
"eval_cov": 0.06972263283925513,
|
| 342 |
+
"eval_global_var": 0.6587248501712328,
|
| 343 |
+
"eval_loss": 0.23407978237900015,
|
| 344 |
+
"eval_mse_loss": 0.8631533832038374,
|
| 345 |
+
"eval_per_var": 0.6400832084760274,
|
| 346 |
+
"eval_within_var": 0.6524244369981496,
|
| 347 |
+
"step": 6144
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.2837744215047804,
|
| 351 |
+
"eval_across_var": 0.006925537744239313,
|
| 352 |
+
"eval_bleu": 0.9630394430269803,
|
| 353 |
+
"eval_ce_loss": 0.08711560829181104,
|
| 354 |
+
"eval_cos_loss": 0.34602166433312576,
|
| 355 |
+
"eval_cov": 0.06972263283925513,
|
| 356 |
+
"eval_global_var": 0.6587248501712328,
|
| 357 |
+
"eval_loss": 0.23407978237900015,
|
| 358 |
+
"eval_mse_loss": 0.8631533832038374,
|
| 359 |
+
"eval_per_var": 0.6400832084760274,
|
| 360 |
+
"eval_runtime": 149.2005,
|
| 361 |
+
"eval_samples_per_second": 187.62,
|
| 362 |
+
"eval_steps_per_second": 2.936,
|
| 363 |
+
"eval_within_var": 0.6524244369981496,
|
| 364 |
+
"step": 6144
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 0.2955983557341462,
|
| 368 |
+
"grad_norm": 0.08157943934202194,
|
| 369 |
+
"learning_rate": 4.131079581886694e-05,
|
| 370 |
+
"loss": 0.3007,
|
| 371 |
+
"step": 6400
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 0.30742228996351206,
|
| 375 |
+
"grad_norm": 0.08066653460264206,
|
| 376 |
+
"learning_rate": 4.057345705423016e-05,
|
| 377 |
+
"loss": 0.2877,
|
| 378 |
+
"step": 6656
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.3192462241928779,
|
| 382 |
+
"grad_norm": 0.08315503597259521,
|
| 383 |
+
"learning_rate": 3.981328167731251e-05,
|
| 384 |
+
"loss": 0.277,
|
| 385 |
+
"step": 6912
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 0.3310701584222438,
|
| 389 |
+
"grad_norm": 0.09480269253253937,
|
| 390 |
+
"learning_rate": 3.9031384394391954e-05,
|
| 391 |
+
"loss": 0.263,
|
| 392 |
+
"step": 7168
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.3310701584222438,
|
| 396 |
+
"eval_across_var": 0.007185437002104439,
|
| 397 |
+
"eval_bleu": 0.7796619623453035,
|
| 398 |
+
"eval_ce_loss": 0.06453313103549557,
|
| 399 |
+
"eval_cos_loss": 0.3031023931285562,
|
| 400 |
+
"eval_cov": 0.06950845239369292,
|
| 401 |
+
"eval_global_var": 0.6862046054509132,
|
| 402 |
+
"eval_loss": 0.19471138987911346,
|
| 403 |
+
"eval_mse_loss": 0.7849507213455357,
|
| 404 |
+
"eval_per_var": 0.6667993275542238,
|
| 405 |
+
"eval_within_var": 0.679719850761161,
|
| 406 |
+
"step": 7168
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.3310701584222438,
|
| 410 |
+
"eval_across_var": 0.007185437002104439,
|
| 411 |
+
"eval_bleu": 0.7796619623453035,
|
| 412 |
+
"eval_ce_loss": 0.06453313103549557,
|
| 413 |
+
"eval_cos_loss": 0.3031023931285562,
|
| 414 |
+
"eval_cov": 0.06950845239369292,
|
| 415 |
+
"eval_global_var": 0.6862046054509132,
|
| 416 |
+
"eval_loss": 0.19471138987911346,
|
| 417 |
+
"eval_mse_loss": 0.7849507213455357,
|
| 418 |
+
"eval_per_var": 0.6667993275542238,
|
| 419 |
+
"eval_runtime": 153.3468,
|
| 420 |
+
"eval_samples_per_second": 182.547,
|
| 421 |
+
"eval_steps_per_second": 2.856,
|
| 422 |
+
"eval_within_var": 0.679719850761161,
|
| 423 |
+
"step": 7168
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.34289409265160964,
|
| 427 |
+
"grad_norm": 0.08880002796649933,
|
| 428 |
+
"learning_rate": 3.822891176432382e-05,
|
| 429 |
+
"loss": 0.2551,
|
| 430 |
+
"step": 7424
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.3547180268809755,
|
| 434 |
+
"grad_norm": 0.0781559944152832,
|
| 435 |
+
"learning_rate": 3.7407040517249335e-05,
|
| 436 |
+
"loss": 0.2457,
|
| 437 |
+
"step": 7680
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.3665419611103413,
|
| 441 |
+
"grad_norm": 0.07617169618606567,
|
| 442 |
+
"learning_rate": 3.6566975829061614e-05,
|
| 443 |
+
"loss": 0.2364,
|
| 444 |
+
"step": 7936
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.37836589533970716,
|
| 448 |
+
"grad_norm": 0.06962994486093521,
|
| 449 |
+
"learning_rate": 3.5709949554159355e-05,
|
| 450 |
+
"loss": 0.2276,
|
| 451 |
+
"step": 8192
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.37836589533970716,
|
| 455 |
+
"eval_across_var": 0.007326101608889084,
|
| 456 |
+
"eval_bleu": 0.8712716886983086,
|
| 457 |
+
"eval_ce_loss": 0.05016031793688666,
|
| 458 |
+
"eval_cos_loss": 0.2737333217644256,
|
| 459 |
+
"eval_cov": 0.06918223481200057,
|
| 460 |
+
"eval_global_var": 0.7070301352026256,
|
| 461 |
+
"eval_loss": 0.16903498234678077,
|
| 462 |
+
"eval_mse_loss": 0.7350932731203836,
|
| 463 |
+
"eval_per_var": 0.6870072595605022,
|
| 464 |
+
"eval_within_var": 0.7004638244844463,
|
| 465 |
+
"step": 8192
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.37836589533970716,
|
| 469 |
+
"eval_across_var": 0.007326101608889084,
|
| 470 |
+
"eval_bleu": 0.8712716886983086,
|
| 471 |
+
"eval_ce_loss": 0.05016031793688666,
|
| 472 |
+
"eval_cos_loss": 0.2737333217644256,
|
| 473 |
+
"eval_cov": 0.06918223481200057,
|
| 474 |
+
"eval_global_var": 0.7070301352026256,
|
| 475 |
+
"eval_loss": 0.16903498234678077,
|
| 476 |
+
"eval_mse_loss": 0.7350932731203836,
|
| 477 |
+
"eval_per_var": 0.6870072595605022,
|
| 478 |
+
"eval_runtime": 253.015,
|
| 479 |
+
"eval_samples_per_second": 110.638,
|
| 480 |
+
"eval_steps_per_second": 1.731,
|
| 481 |
+
"eval_within_var": 0.7004638244844463,
|
| 482 |
+
"step": 8192
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.390189829569073,
|
| 486 |
+
"grad_norm": 0.07793359458446503,
|
| 487 |
+
"learning_rate": 3.483721841907964e-05,
|
| 488 |
+
"loss": 0.2218,
|
| 489 |
+
"step": 8448
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.4020137637984389,
|
| 493 |
+
"grad_norm": 0.09409929066896439,
|
| 494 |
+
"learning_rate": 3.395006217965885e-05,
|
| 495 |
+
"loss": 0.2157,
|
| 496 |
+
"step": 8704
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.41383769802780473,
|
| 500 |
+
"grad_norm": 0.07621984928846359,
|
| 501 |
+
"learning_rate": 3.3049781744423665e-05,
|
| 502 |
+
"loss": 0.2083,
|
| 503 |
+
"step": 8960
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.4256616322571706,
|
| 507 |
+
"grad_norm": 0.07183349132537842,
|
| 508 |
+
"learning_rate": 3.213769726696439e-05,
|
| 509 |
+
"loss": 0.2036,
|
| 510 |
+
"step": 9216
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 0.4256616322571706,
|
| 514 |
+
"eval_across_var": 0.007464996506098539,
|
| 515 |
+
"eval_bleu": 0.8745999403603312,
|
| 516 |
+
"eval_ce_loss": 0.04043501646568378,
|
| 517 |
+
"eval_cos_loss": 0.2525300460114871,
|
| 518 |
+
"eval_cov": 0.0689811532355879,
|
| 519 |
+
"eval_global_var": 0.7244934360730594,
|
| 520 |
+
"eval_loss": 0.15128187069745913,
|
| 521 |
+
"eval_mse_loss": 0.7017567277498985,
|
| 522 |
+
"eval_per_var": 0.703834011130137,
|
| 523 |
+
"eval_within_var": 0.717845245039082,
|
| 524 |
+
"step": 9216
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.4256616322571706,
|
| 528 |
+
"eval_across_var": 0.007464996506098539,
|
| 529 |
+
"eval_bleu": 0.8745999403603312,
|
| 530 |
+
"eval_ce_loss": 0.04043501646568378,
|
| 531 |
+
"eval_cos_loss": 0.2525300460114871,
|
| 532 |
+
"eval_cov": 0.0689811532355879,
|
| 533 |
+
"eval_global_var": 0.7244934360730594,
|
| 534 |
+
"eval_loss": 0.15128187069745913,
|
| 535 |
+
"eval_mse_loss": 0.7017567277498985,
|
| 536 |
+
"eval_per_var": 0.703834011130137,
|
| 537 |
+
"eval_runtime": 147.2422,
|
| 538 |
+
"eval_samples_per_second": 190.115,
|
| 539 |
+
"eval_steps_per_second": 2.975,
|
| 540 |
+
"eval_within_var": 0.717845245039082,
|
| 541 |
+
"step": 9216
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.4374855664865364,
|
| 545 |
+
"grad_norm": 0.07419689744710922,
|
| 546 |
+
"learning_rate": 3.121514621008757e-05,
|
| 547 |
+
"loss": 0.1988,
|
| 548 |
+
"step": 9472
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.44930950071590225,
|
| 552 |
+
"grad_norm": 0.07019222527742386,
|
| 553 |
+
"learning_rate": 3.0283481384586697e-05,
|
| 554 |
+
"loss": 0.1953,
|
| 555 |
+
"step": 9728
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.4611334349452681,
|
| 559 |
+
"grad_norm": 0.08114204555749893,
|
| 560 |
+
"learning_rate": 2.9344068965507027e-05,
|
| 561 |
+
"loss": 0.1903,
|
| 562 |
+
"step": 9984
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.472957369174634,
|
| 566 |
+
"grad_norm": 0.07998168468475342,
|
| 567 |
+
"learning_rate": 2.839828648881323e-05,
|
| 568 |
+
"loss": 0.1878,
|
| 569 |
+
"step": 10240
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.472957369174634,
|
| 573 |
+
"eval_across_var": 0.00757191778101095,
|
| 574 |
+
"eval_bleu": 0.8768266780689754,
|
| 575 |
+
"eval_ce_loss": 0.033580983376564226,
|
| 576 |
+
"eval_cos_loss": 0.23719462929250987,
|
| 577 |
+
"eval_cov": 0.06886953414847317,
|
| 578 |
+
"eval_global_var": 0.7387516944920092,
|
| 579 |
+
"eval_loss": 0.1387409362856928,
|
| 580 |
+
"eval_mse_loss": 0.6800311979365675,
|
| 581 |
+
"eval_per_var": 0.7177277308076484,
|
| 582 |
+
"eval_within_var": 0.7320368596135753,
|
| 583 |
+
"step": 10240
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.472957369174634,
|
| 587 |
+
"eval_across_var": 0.00757191778101095,
|
| 588 |
+
"eval_bleu": 0.8768266780689754,
|
| 589 |
+
"eval_ce_loss": 0.033580983376564226,
|
| 590 |
+
"eval_cos_loss": 0.23719462929250987,
|
| 591 |
+
"eval_cov": 0.06886953414847317,
|
| 592 |
+
"eval_global_var": 0.7387516944920092,
|
| 593 |
+
"eval_loss": 0.1387409362856928,
|
| 594 |
+
"eval_mse_loss": 0.6800311979365675,
|
| 595 |
+
"eval_per_var": 0.7177277308076484,
|
| 596 |
+
"eval_runtime": 156.8381,
|
| 597 |
+
"eval_samples_per_second": 178.483,
|
| 598 |
+
"eval_steps_per_second": 2.793,
|
| 599 |
+
"eval_within_var": 0.7320368596135753,
|
| 600 |
+
"step": 10240
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.48478130340399983,
|
| 604 |
+
"grad_norm": 0.08175234496593475,
|
| 605 |
+
"learning_rate": 2.7447520831397623e-05,
|
| 606 |
+
"loss": 0.1832,
|
| 607 |
+
"step": 10496
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.49660523763336567,
|
| 611 |
+
"grad_norm": 0.07749126106500626,
|
| 612 |
+
"learning_rate": 2.6493166177391138e-05,
|
| 613 |
+
"loss": 0.1793,
|
| 614 |
+
"step": 10752
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.5084291718627315,
|
| 618 |
+
"grad_norm": 0.06436038762331009,
|
| 619 |
+
"learning_rate": 2.5536621973758952e-05,
|
| 620 |
+
"loss": 0.1781,
|
| 621 |
+
"step": 11008
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5202531060920974,
|
| 625 |
+
"grad_norm": 0.08284164220094681,
|
| 626 |
+
"learning_rate": 2.4579290878178904e-05,
|
| 627 |
+
"loss": 0.1743,
|
| 628 |
+
"step": 11264
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 0.5202531060920974,
|
| 632 |
+
"eval_across_var": 0.00768219434962391,
|
| 633 |
+
"eval_bleu": 0.87828493142736,
|
| 634 |
+
"eval_ce_loss": 0.028931678281288984,
|
| 635 |
+
"eval_cos_loss": 0.22550881647219942,
|
| 636 |
+
"eval_cov": 0.06867876444777397,
|
| 637 |
+
"eval_global_var": 0.7501025613584474,
|
| 638 |
+
"eval_loss": 0.1298226385492168,
|
| 639 |
+
"eval_mse_loss": 0.6647662283623055,
|
| 640 |
+
"eval_per_var": 0.7285691352739726,
|
| 641 |
+
"eval_within_var": 0.7433075608183805,
|
| 642 |
+
"step": 11264
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.5202531060920974,
|
| 646 |
+
"eval_across_var": 0.00768219434962391,
|
| 647 |
+
"eval_bleu": 0.87828493142736,
|
| 648 |
+
"eval_ce_loss": 0.028931678281288984,
|
| 649 |
+
"eval_cos_loss": 0.22550881647219942,
|
| 650 |
+
"eval_cov": 0.06867876444777397,
|
| 651 |
+
"eval_global_var": 0.7501025613584474,
|
| 652 |
+
"eval_loss": 0.1298226385492168,
|
| 653 |
+
"eval_mse_loss": 0.6647662283623055,
|
| 654 |
+
"eval_per_var": 0.7285691352739726,
|
| 655 |
+
"eval_runtime": 147.0377,
|
| 656 |
+
"eval_samples_per_second": 190.38,
|
| 657 |
+
"eval_steps_per_second": 2.979,
|
| 658 |
+
"eval_within_var": 0.7433075608183805,
|
| 659 |
+
"step": 11264
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 0.5320770403214632,
|
| 663 |
+
"grad_norm": 0.08053084462881088,
|
| 664 |
+
"learning_rate": 2.362257670221181e-05,
|
| 665 |
+
"loss": 0.1714,
|
| 666 |
+
"step": 11520
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 0.5439009745508291,
|
| 670 |
+
"grad_norm": 0.06910347938537598,
|
| 671 |
+
"learning_rate": 2.2667882352779608e-05,
|
| 672 |
+
"loss": 0.1699,
|
| 673 |
+
"step": 11776
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 0.5557249087801949,
|
| 677 |
+
"grad_norm": 0.06619574874639511,
|
| 678 |
+
"learning_rate": 2.17166077749702e-05,
|
| 679 |
+
"loss": 0.1677,
|
| 680 |
+
"step": 12032
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.5675488430095608,
|
| 684 |
+
"grad_norm": 0.07791094481945038,
|
| 685 |
+
"learning_rate": 2.0770147899185404e-05,
|
| 686 |
+
"loss": 0.1665,
|
| 687 |
+
"step": 12288
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.5675488430095608,
|
| 691 |
+
"eval_across_var": 0.007755262901598256,
|
| 692 |
+
"eval_bleu": 0.8793172460885812,
|
| 693 |
+
"eval_ce_loss": 0.02565789532520251,
|
| 694 |
+
"eval_cos_loss": 0.2167755546022768,
|
| 695 |
+
"eval_cov": 0.06850555280572203,
|
| 696 |
+
"eval_global_var": 0.759103435359589,
|
| 697 |
+
"eval_loss": 0.12342031796773274,
|
| 698 |
+
"eval_mse_loss": 0.6545950485963256,
|
| 699 |
+
"eval_per_var": 0.73747079230879,
|
| 700 |
+
"eval_within_var": 0.7522586970568792,
|
| 701 |
+
"step": 12288
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.5675488430095608,
|
| 705 |
+
"eval_across_var": 0.007755262901598256,
|
| 706 |
+
"eval_bleu": 0.8793172460885812,
|
| 707 |
+
"eval_ce_loss": 0.02565789532520251,
|
| 708 |
+
"eval_cos_loss": 0.2167755546022768,
|
| 709 |
+
"eval_cov": 0.06850555280572203,
|
| 710 |
+
"eval_global_var": 0.759103435359589,
|
| 711 |
+
"eval_loss": 0.12342031796773274,
|
| 712 |
+
"eval_mse_loss": 0.6545950485963256,
|
| 713 |
+
"eval_per_var": 0.73747079230879,
|
| 714 |
+
"eval_runtime": 145.6821,
|
| 715 |
+
"eval_samples_per_second": 192.151,
|
| 716 |
+
"eval_steps_per_second": 3.007,
|
| 717 |
+
"eval_within_var": 0.7522586970568792,
|
| 718 |
+
"step": 12288
|
| 719 |
+
}
|
| 720 |
+
],
|
| 721 |
+
"logging_steps": 256,
|
| 722 |
+
"max_steps": 21651,
|
| 723 |
+
"num_input_tokens_seen": 0,
|
| 724 |
+
"num_train_epochs": 1,
|
| 725 |
+
"save_steps": 1024,
|
| 726 |
+
"stateful_callbacks": {
|
| 727 |
+
"TrainerControl": {
|
| 728 |
+
"args": {
|
| 729 |
+
"should_epoch_stop": false,
|
| 730 |
+
"should_evaluate": false,
|
| 731 |
+
"should_log": false,
|
| 732 |
+
"should_save": true,
|
| 733 |
+
"should_training_stop": false
|
| 734 |
+
},
|
| 735 |
+
"attributes": {}
|
| 736 |
+
}
|
| 737 |
+
},
|
| 738 |
+
"total_flos": 0.0,
|
| 739 |
+
"train_batch_size": 64,
|
| 740 |
+
"trial_name": null,
|
| 741 |
+
"trial_params": null
|
| 742 |
+
}
|
checkpoints-v2/checkpoint-12288/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3d8cc76487b7771febf234ca067facf7ab46c34066dc0aabf5f33ed9d26b2a5f
|
| 3 |
+
size 5777
|