Upload folder using huggingface_hub
Browse files- checkpoints-v1.0-discrete-semantc/checkpoint-20480/eval_state.json +0 -0
- checkpoints-v1.0-discrete-semantc/checkpoint-20480/model.safetensors +3 -0
- checkpoints-v1.0-discrete-semantc/checkpoint-20480/optimizer.pt +3 -0
- checkpoints-v1.0-discrete-semantc/checkpoint-20480/rng_state.pth +3 -0
- checkpoints-v1.0-discrete-semantc/checkpoint-20480/scaler.pt +3 -0
- checkpoints-v1.0-discrete-semantc/checkpoint-20480/scheduler.pt +3 -0
- checkpoints-v1.0-discrete-semantc/checkpoint-20480/trainer_state.json +824 -0
- checkpoints-v1.0-discrete-semantc/checkpoint-20480/training_args.bin +3 -0
checkpoints-v1.0-discrete-semantc/checkpoint-20480/eval_state.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoints-v1.0-discrete-semantc/checkpoint-20480/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b767a63f41c3869eb7c2eef6745d19a9610b830f0f773532d674d05ebc4a0747
|
| 3 |
+
size 9318340
|
checkpoints-v1.0-discrete-semantc/checkpoint-20480/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9847a8b1c64368508a4fc840883b4636601af3b5266ad17f564a9177e52c119b
|
| 3 |
+
size 1641995
|
checkpoints-v1.0-discrete-semantc/checkpoint-20480/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:66027b053d92554ecf06a4e17396590e01102bcf3b25d43bc5e6c98a2e5fd648
|
| 3 |
+
size 14645
|
checkpoints-v1.0-discrete-semantc/checkpoint-20480/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e1654750e91fad9c84d48b2b56e72818b509aa308aa50c4ccf5496b389664c5b
|
| 3 |
+
size 1383
|
checkpoints-v1.0-discrete-semantc/checkpoint-20480/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ebd523f2f266d8d9e2fe88e1870b08c91dfa6ae1dc725a1649aff91c0b251fbb
|
| 3 |
+
size 1465
|
checkpoints-v1.0-discrete-semantc/checkpoint-20480/trainer_state.json
ADDED
|
@@ -0,0 +1,824 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.2084382903582024,
|
| 6 |
+
"eval_steps": 2048,
|
| 7 |
+
"global_step": 20480,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0026054786294775305,
|
| 14 |
+
"grad_norm": 0.25278976559638977,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 0.18901292979717255,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.005210957258955061,
|
| 21 |
+
"grad_norm": 0.3028739094734192,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 0.13242575526237488,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.007816435888432591,
|
| 28 |
+
"grad_norm": 0.23641528189182281,
|
| 29 |
+
"learning_rate": 0.0009999832063013975,
|
| 30 |
+
"loss": 0.1147228330373764,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.010421914517910122,
|
| 35 |
+
"grad_norm": 0.1741982400417328,
|
| 36 |
+
"learning_rate": 0.0009999325626552273,
|
| 37 |
+
"loss": 0.10675819963216782,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.01302739314738765,
|
| 42 |
+
"grad_norm": 0.23829564452171326,
|
| 43 |
+
"learning_rate": 0.000999848072231934,
|
| 44 |
+
"loss": 0.10438411682844162,
|
| 45 |
+
"step": 1280
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.015632871776865183,
|
| 49 |
+
"grad_norm": 0.13794861733913422,
|
| 50 |
+
"learning_rate": 0.0009997297407517455,
|
| 51 |
+
"loss": 0.1017049103975296,
|
| 52 |
+
"step": 1536
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.018238350406342713,
|
| 56 |
+
"grad_norm": 0.17629992961883545,
|
| 57 |
+
"learning_rate": 0.0009995775762260215,
|
| 58 |
+
"loss": 0.0993618294596672,
|
| 59 |
+
"step": 1792
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.020843829035820244,
|
| 63 |
+
"grad_norm": 0.23240557312965393,
|
| 64 |
+
"learning_rate": 0.0009993915889567086,
|
| 65 |
+
"loss": 0.09926237165927887,
|
| 66 |
+
"step": 2048
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.020843829035820244,
|
| 70 |
+
"eval_bce_loss": 1.5750247291156223,
|
| 71 |
+
"eval_bleu": 0.0008629714973265223,
|
| 72 |
+
"eval_cos_loss": 0.23970603900296347,
|
| 73 |
+
"eval_loss": 0.10751142757279532,
|
| 74 |
+
"eval_mean_score": 0.6800641741071428,
|
| 75 |
+
"eval_mse_loss": 0.09176118011985507,
|
| 76 |
+
"step": 2048
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.020843829035820244,
|
| 80 |
+
"eval_bce_loss": 1.5750247291156223,
|
| 81 |
+
"eval_bleu": 0.0008629714973265223,
|
| 82 |
+
"eval_cos_loss": 0.23970603900296347,
|
| 83 |
+
"eval_loss": 0.10751142757279532,
|
| 84 |
+
"eval_mean_score": 0.6800641741071428,
|
| 85 |
+
"eval_mse_loss": 0.09176118011985507,
|
| 86 |
+
"eval_runtime": 18.5003,
|
| 87 |
+
"eval_samples_per_second": 118.917,
|
| 88 |
+
"eval_steps_per_second": 1.892,
|
| 89 |
+
"step": 2048
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.023449307665297774,
|
| 93 |
+
"grad_norm": 0.16738885641098022,
|
| 94 |
+
"learning_rate": 0.0009991717915356446,
|
| 95 |
+
"loss": 0.09865972399711609,
|
| 96 |
+
"step": 2304
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.0260547862947753,
|
| 100 |
+
"grad_norm": 0.16121487319469452,
|
| 101 |
+
"learning_rate": 0.0009989181988437053,
|
| 102 |
+
"loss": 0.09649977833032608,
|
| 103 |
+
"step": 2560
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.028660264924252832,
|
| 107 |
+
"grad_norm": 0.24666129052639008,
|
| 108 |
+
"learning_rate": 0.0009986308280497966,
|
| 109 |
+
"loss": 0.09620590507984161,
|
| 110 |
+
"step": 2816
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.031265743553730366,
|
| 114 |
+
"grad_norm": 0.1150430217385292,
|
| 115 |
+
"learning_rate": 0.0009983096986096933,
|
| 116 |
+
"loss": 0.0973711758852005,
|
| 117 |
+
"step": 3072
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.03387122218320789,
|
| 121 |
+
"grad_norm": 0.22624613344669342,
|
| 122 |
+
"learning_rate": 0.000997954832264721,
|
| 123 |
+
"loss": 0.09581801295280457,
|
| 124 |
+
"step": 3328
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.03647670081268543,
|
| 128 |
+
"grad_norm": 0.23274144530296326,
|
| 129 |
+
"learning_rate": 0.0009975662530402843,
|
| 130 |
+
"loss": 0.09428834170103073,
|
| 131 |
+
"step": 3584
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.039082179442162954,
|
| 135 |
+
"grad_norm": 0.2979954183101654,
|
| 136 |
+
"learning_rate": 0.0009971439872442397,
|
| 137 |
+
"loss": 0.0952843651175499,
|
| 138 |
+
"step": 3840
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.04168765807164049,
|
| 142 |
+
"grad_norm": 0.13622763752937317,
|
| 143 |
+
"learning_rate": 0.0009966880634651165,
|
| 144 |
+
"loss": 0.09436459839344025,
|
| 145 |
+
"step": 4096
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.04168765807164049,
|
| 149 |
+
"eval_bce_loss": 1.2727542996406556,
|
| 150 |
+
"eval_bleu": 0.0006777012438118088,
|
| 151 |
+
"eval_cos_loss": 0.2344500047819955,
|
| 152 |
+
"eval_loss": 0.10118420294352941,
|
| 153 |
+
"eval_mean_score": 0.71640625,
|
| 154 |
+
"eval_mse_loss": 0.0884566605091095,
|
| 155 |
+
"step": 4096
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.04168765807164049,
|
| 159 |
+
"eval_bce_loss": 1.2727542996406556,
|
| 160 |
+
"eval_bleu": 0.0006777012438118088,
|
| 161 |
+
"eval_cos_loss": 0.2344500047819955,
|
| 162 |
+
"eval_loss": 0.10118420294352941,
|
| 163 |
+
"eval_mean_score": 0.71640625,
|
| 164 |
+
"eval_mse_loss": 0.0884566605091095,
|
| 165 |
+
"eval_runtime": 19.465,
|
| 166 |
+
"eval_samples_per_second": 113.023,
|
| 167 |
+
"eval_steps_per_second": 1.798,
|
| 168 |
+
"step": 4096
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 0.044293136701118015,
|
| 172 |
+
"grad_norm": 0.14671610295772552,
|
| 173 |
+
"learning_rate": 0.0009961985125701786,
|
| 174 |
+
"loss": 0.09386898577213287,
|
| 175 |
+
"step": 4352
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.04689861533059555,
|
| 179 |
+
"grad_norm": 0.11165767163038254,
|
| 180 |
+
"learning_rate": 0.0009956753677033364,
|
| 181 |
+
"loss": 0.09496171772480011,
|
| 182 |
+
"step": 4608
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.049504093960073076,
|
| 186 |
+
"grad_norm": 0.19689643383026123,
|
| 187 |
+
"learning_rate": 0.000995118664282902,
|
| 188 |
+
"loss": 0.09313815832138062,
|
| 189 |
+
"step": 4864
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.0521095725895506,
|
| 193 |
+
"grad_norm": 0.12060590833425522,
|
| 194 |
+
"learning_rate": 0.0009945284399991924,
|
| 195 |
+
"loss": 0.09367529302835464,
|
| 196 |
+
"step": 5120
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.05471505121902814,
|
| 200 |
+
"grad_norm": 0.13436441123485565,
|
| 201 |
+
"learning_rate": 0.0009939047348119768,
|
| 202 |
+
"loss": 0.0925508588552475,
|
| 203 |
+
"step": 5376
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.057320529848505664,
|
| 207 |
+
"grad_norm": 0.1394355297088623,
|
| 208 |
+
"learning_rate": 0.0009932475909477712,
|
| 209 |
+
"loss": 0.092799112200737,
|
| 210 |
+
"step": 5632
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 0.0599260084779832,
|
| 214 |
+
"grad_norm": 0.16785988211631775,
|
| 215 |
+
"learning_rate": 0.0009925570528969802,
|
| 216 |
+
"loss": 0.09231170266866684,
|
| 217 |
+
"step": 5888
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.06253148710746073,
|
| 221 |
+
"grad_norm": 0.12323057651519775,
|
| 222 |
+
"learning_rate": 0.0009918331674108843,
|
| 223 |
+
"loss": 0.09121806174516678,
|
| 224 |
+
"step": 6144
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.06253148710746073,
|
| 228 |
+
"eval_bce_loss": 1.37689631155559,
|
| 229 |
+
"eval_bleu": 0.0007894363314727929,
|
| 230 |
+
"eval_cos_loss": 0.23060905975954873,
|
| 231 |
+
"eval_loss": 0.10043682392154421,
|
| 232 |
+
"eval_mean_score": 0.6661830357142857,
|
| 233 |
+
"eval_mse_loss": 0.0866678610444069,
|
| 234 |
+
"step": 6144
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.06253148710746073,
|
| 238 |
+
"eval_bce_loss": 1.37689631155559,
|
| 239 |
+
"eval_bleu": 0.0007894363314727929,
|
| 240 |
+
"eval_cos_loss": 0.23060905975954873,
|
| 241 |
+
"eval_loss": 0.10043682392154421,
|
| 242 |
+
"eval_mean_score": 0.6661830357142857,
|
| 243 |
+
"eval_mse_loss": 0.0866678610444069,
|
| 244 |
+
"eval_runtime": 18.447,
|
| 245 |
+
"eval_samples_per_second": 119.261,
|
| 246 |
+
"eval_steps_per_second": 1.897,
|
| 247 |
+
"step": 6144
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.06513696573693825,
|
| 251 |
+
"grad_norm": 0.11431664973497391,
|
| 252 |
+
"learning_rate": 0.000991075983498475,
|
| 253 |
+
"loss": 0.09111239016056061,
|
| 254 |
+
"step": 6400
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.06774244436641579,
|
| 258 |
+
"grad_norm": 0.15795110166072845,
|
| 259 |
+
"learning_rate": 0.0009902855524231366,
|
| 260 |
+
"loss": 0.09170184284448624,
|
| 261 |
+
"step": 6656
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.07034792299589332,
|
| 265 |
+
"grad_norm": 0.1189679428935051,
|
| 266 |
+
"learning_rate": 0.0009894619276991758,
|
| 267 |
+
"loss": 0.09055215865373611,
|
| 268 |
+
"step": 6912
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.07295340162537085,
|
| 272 |
+
"grad_norm": 0.12559665739536285,
|
| 273 |
+
"learning_rate": 0.0009886051650881986,
|
| 274 |
+
"loss": 0.09139999002218246,
|
| 275 |
+
"step": 7168
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.07555888025484837,
|
| 279 |
+
"grad_norm": 0.10260415077209473,
|
| 280 |
+
"learning_rate": 0.0009877153225953342,
|
| 281 |
+
"loss": 0.09160503000020981,
|
| 282 |
+
"step": 7424
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.07816435888432591,
|
| 286 |
+
"grad_norm": 0.14531247317790985,
|
| 287 |
+
"learning_rate": 0.0009867924604653093,
|
| 288 |
+
"loss": 0.08971104025840759,
|
| 289 |
+
"step": 7680
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.08076983751380344,
|
| 293 |
+
"grad_norm": 0.135915145277977,
|
| 294 |
+
"learning_rate": 0.0009858366411783688,
|
| 295 |
+
"loss": 0.09080137312412262,
|
| 296 |
+
"step": 7936
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.08337531614328098,
|
| 300 |
+
"grad_norm": 0.0951119214296341,
|
| 301 |
+
"learning_rate": 0.0009848479294460454,
|
| 302 |
+
"loss": 0.08976119756698608,
|
| 303 |
+
"step": 8192
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.08337531614328098,
|
| 307 |
+
"eval_bce_loss": 1.1331708482333591,
|
| 308 |
+
"eval_bleu": 0.0,
|
| 309 |
+
"eval_cos_loss": 0.22396006413868497,
|
| 310 |
+
"eval_loss": 0.09644954183271953,
|
| 311 |
+
"eval_mean_score": 0.7250139508928571,
|
| 312 |
+
"eval_mse_loss": 0.08511783331632614,
|
| 313 |
+
"step": 8192
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.08337531614328098,
|
| 317 |
+
"eval_bce_loss": 1.1331708482333591,
|
| 318 |
+
"eval_bleu": 0.0,
|
| 319 |
+
"eval_cos_loss": 0.22396006413868497,
|
| 320 |
+
"eval_loss": 0.09644954183271953,
|
| 321 |
+
"eval_mean_score": 0.7250139508928571,
|
| 322 |
+
"eval_mse_loss": 0.08511783331632614,
|
| 323 |
+
"eval_runtime": 18.5283,
|
| 324 |
+
"eval_samples_per_second": 118.738,
|
| 325 |
+
"eval_steps_per_second": 1.889,
|
| 326 |
+
"step": 8192
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.0859807947727585,
|
| 330 |
+
"grad_norm": 0.2771708369255066,
|
| 331 |
+
"learning_rate": 0.0009838263922067783,
|
| 332 |
+
"loss": 0.09032878279685974,
|
| 333 |
+
"step": 8448
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.08858627340223603,
|
| 337 |
+
"grad_norm": 0.1163194477558136,
|
| 338 |
+
"learning_rate": 0.0009827720986213824,
|
| 339 |
+
"loss": 0.09021615236997604,
|
| 340 |
+
"step": 8704
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.09119175203171356,
|
| 344 |
+
"grad_norm": 0.1319112330675125,
|
| 345 |
+
"learning_rate": 0.000981685120068365,
|
| 346 |
+
"loss": 0.08950018137693405,
|
| 347 |
+
"step": 8960
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.0937972306611911,
|
| 351 |
+
"grad_norm": 0.17556321620941162,
|
| 352 |
+
"learning_rate": 0.0009805655301390929,
|
| 353 |
+
"loss": 0.08931967616081238,
|
| 354 |
+
"step": 9216
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.09640270929066862,
|
| 358 |
+
"grad_norm": 0.10782352089881897,
|
| 359 |
+
"learning_rate": 0.0009794134046328114,
|
| 360 |
+
"loss": 0.08995676040649414,
|
| 361 |
+
"step": 9472
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.09900818792014615,
|
| 365 |
+
"grad_norm": 0.33878061175346375,
|
| 366 |
+
"learning_rate": 0.0009782288215515113,
|
| 367 |
+
"loss": 0.08961572498083115,
|
| 368 |
+
"step": 9728
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.10161366654962369,
|
| 372 |
+
"grad_norm": 0.18630877137184143,
|
| 373 |
+
"learning_rate": 0.0009770118610946488,
|
| 374 |
+
"loss": 0.09020468592643738,
|
| 375 |
+
"step": 9984
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.1042191451791012,
|
| 379 |
+
"grad_norm": 0.10592332482337952,
|
| 380 |
+
"learning_rate": 0.0009757626056537148,
|
| 381 |
+
"loss": 0.08932031691074371,
|
| 382 |
+
"step": 10240
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.1042191451791012,
|
| 386 |
+
"eval_bce_loss": 1.0940844322953904,
|
| 387 |
+
"eval_bleu": 0.0,
|
| 388 |
+
"eval_cos_loss": 0.22045842707157134,
|
| 389 |
+
"eval_loss": 0.0946416518517903,
|
| 390 |
+
"eval_mean_score": 0.761328125,
|
| 391 |
+
"eval_mse_loss": 0.08370080781834466,
|
| 392 |
+
"step": 10240
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.1042191451791012,
|
| 396 |
+
"eval_bce_loss": 1.0940844322953904,
|
| 397 |
+
"eval_bleu": 0.0,
|
| 398 |
+
"eval_cos_loss": 0.22045842707157134,
|
| 399 |
+
"eval_loss": 0.0946416518517903,
|
| 400 |
+
"eval_mean_score": 0.761328125,
|
| 401 |
+
"eval_mse_loss": 0.08370080781834466,
|
| 402 |
+
"eval_runtime": 18.6175,
|
| 403 |
+
"eval_samples_per_second": 118.168,
|
| 404 |
+
"eval_steps_per_second": 1.88,
|
| 405 |
+
"step": 10240
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.10682462380857874,
|
| 409 |
+
"grad_norm": 0.09680312126874924,
|
| 410 |
+
"learning_rate": 0.000974481139806658,
|
| 411 |
+
"loss": 0.08943861722946167,
|
| 412 |
+
"step": 10496
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.10943010243805627,
|
| 416 |
+
"grad_norm": 0.2351679801940918,
|
| 417 |
+
"learning_rate": 0.0009731675503121576,
|
| 418 |
+
"loss": 0.08960960805416107,
|
| 419 |
+
"step": 10752
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.11203558106753381,
|
| 423 |
+
"grad_norm": 0.12805607914924622,
|
| 424 |
+
"learning_rate": 0.0009718219261037504,
|
| 425 |
+
"loss": 0.08869561553001404,
|
| 426 |
+
"step": 11008
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.11464105969701133,
|
| 430 |
+
"grad_norm": 0.11389618366956711,
|
| 431 |
+
"learning_rate": 0.0009704443582838089,
|
| 432 |
+
"loss": 0.08700637519359589,
|
| 433 |
+
"step": 11264
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.11724653832648886,
|
| 437 |
+
"grad_norm": 0.11915703117847443,
|
| 438 |
+
"learning_rate": 0.0009690349401173742,
|
| 439 |
+
"loss": 0.08796984702348709,
|
| 440 |
+
"step": 11520
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.1198520169559664,
|
| 444 |
+
"grad_norm": 0.08544740080833435,
|
| 445 |
+
"learning_rate": 0.0009675937670258413,
|
| 446 |
+
"loss": 0.087662473320961,
|
| 447 |
+
"step": 11776
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.12245749558544393,
|
| 451 |
+
"grad_norm": 0.10790608078241348,
|
| 452 |
+
"learning_rate": 0.000966120936580499,
|
| 453 |
+
"loss": 0.08814739435911179,
|
| 454 |
+
"step": 12032
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 0.12506297421492146,
|
| 458 |
+
"grad_norm": 0.13961143791675568,
|
| 459 |
+
"learning_rate": 0.0009646165484959241,
|
| 460 |
+
"loss": 0.08711596578359604,
|
| 461 |
+
"step": 12288
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.12506297421492146,
|
| 465 |
+
"eval_bce_loss": 1.1939687473433358,
|
| 466 |
+
"eval_bleu": 0.0,
|
| 467 |
+
"eval_cos_loss": 0.21508082662309919,
|
| 468 |
+
"eval_loss": 0.0935181138770921,
|
| 469 |
+
"eval_mean_score": 0.73857421875,
|
| 470 |
+
"eval_mse_loss": 0.08157842648880823,
|
| 471 |
+
"step": 12288
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.12506297421492146,
|
| 475 |
+
"eval_bce_loss": 1.1939687473433358,
|
| 476 |
+
"eval_bleu": 0.0,
|
| 477 |
+
"eval_cos_loss": 0.21508082662309919,
|
| 478 |
+
"eval_loss": 0.0935181138770921,
|
| 479 |
+
"eval_mean_score": 0.73857421875,
|
| 480 |
+
"eval_mse_loss": 0.08157842648880823,
|
| 481 |
+
"eval_runtime": 18.583,
|
| 482 |
+
"eval_samples_per_second": 118.388,
|
| 483 |
+
"eval_steps_per_second": 1.883,
|
| 484 |
+
"step": 12288
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.12766845284439898,
|
| 488 |
+
"grad_norm": 0.09527724236249924,
|
| 489 |
+
"learning_rate": 0.0009630807046232299,
|
| 490 |
+
"loss": 0.08747082203626633,
|
| 491 |
+
"step": 12544
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.1302739314738765,
|
| 495 |
+
"grad_norm": 0.09321276098489761,
|
| 496 |
+
"learning_rate": 0.0009615135089431713,
|
| 497 |
+
"loss": 0.08775044977664948,
|
| 498 |
+
"step": 12800
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 0.13287941010335405,
|
| 502 |
+
"grad_norm": 0.08462223410606384,
|
| 503 |
+
"learning_rate": 0.0009599150675591049,
|
| 504 |
+
"loss": 0.08798608183860779,
|
| 505 |
+
"step": 13056
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.13548488873283157,
|
| 509 |
+
"grad_norm": 0.09068363904953003,
|
| 510 |
+
"learning_rate": 0.0009582854886898051,
|
| 511 |
+
"loss": 0.08702214062213898,
|
| 512 |
+
"step": 13312
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.13809036736230912,
|
| 516 |
+
"grad_norm": 0.13726262748241425,
|
| 517 |
+
"learning_rate": 0.0009566248826621377,
|
| 518 |
+
"loss": 0.08681564033031464,
|
| 519 |
+
"step": 13568
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.14069584599178664,
|
| 523 |
+
"grad_norm": 0.10372263193130493,
|
| 524 |
+
"learning_rate": 0.0009549333619035909,
|
| 525 |
+
"loss": 0.08741886168718338,
|
| 526 |
+
"step": 13824
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.14330132462126416,
|
| 530 |
+
"grad_norm": 0.21570627391338348,
|
| 531 |
+
"learning_rate": 0.0009532110409346625,
|
| 532 |
+
"loss": 0.08581440150737762,
|
| 533 |
+
"step": 14080
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.1459068032507417,
|
| 537 |
+
"grad_norm": 0.11762681603431702,
|
| 538 |
+
"learning_rate": 0.0009514580363611076,
|
| 539 |
+
"loss": 0.08681235462427139,
|
| 540 |
+
"step": 14336
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.1459068032507417,
|
| 544 |
+
"eval_bce_loss": 0.9678310785974775,
|
| 545 |
+
"eval_bleu": 0.0007685106705860851,
|
| 546 |
+
"eval_cos_loss": 0.22033600551741464,
|
| 547 |
+
"eval_loss": 0.092972403551851,
|
| 548 |
+
"eval_mean_score": 0.7832589285714285,
|
| 549 |
+
"eval_mse_loss": 0.083294092970235,
|
| 550 |
+
"step": 14336
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.1459068032507417,
|
| 554 |
+
"eval_bce_loss": 0.9678310785974775,
|
| 555 |
+
"eval_bleu": 0.0007685106705860851,
|
| 556 |
+
"eval_cos_loss": 0.22033600551741464,
|
| 557 |
+
"eval_loss": 0.092972403551851,
|
| 558 |
+
"eval_mean_score": 0.7832589285714285,
|
| 559 |
+
"eval_mse_loss": 0.083294092970235,
|
| 560 |
+
"eval_runtime": 18.8884,
|
| 561 |
+
"eval_samples_per_second": 116.474,
|
| 562 |
+
"eval_steps_per_second": 1.853,
|
| 563 |
+
"step": 14336
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.14851228188021923,
|
| 567 |
+
"grad_norm": 0.11540628969669342,
|
| 568 |
+
"learning_rate": 0.0009496744668660439,
|
| 569 |
+
"loss": 0.08700872212648392,
|
| 570 |
+
"step": 14592
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.15111776050969675,
|
| 574 |
+
"grad_norm": 0.22385740280151367,
|
| 575 |
+
"learning_rate": 0.0009478604532019162,
|
| 576 |
+
"loss": 0.08554290980100632,
|
| 577 |
+
"step": 14848
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.1537232391391743,
|
| 581 |
+
"grad_norm": 0.08197021484375,
|
| 582 |
+
"learning_rate": 0.0009460161181823213,
|
| 583 |
+
"loss": 0.08630763739347458,
|
| 584 |
+
"step": 15104
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.15632871776865181,
|
| 588 |
+
"grad_norm": 0.12326359003782272,
|
| 589 |
+
"learning_rate": 0.0009441415866736933,
|
| 590 |
+
"loss": 0.08561107516288757,
|
| 591 |
+
"step": 15360
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.15893419639812933,
|
| 595 |
+
"grad_norm": 0.10007017850875854,
|
| 596 |
+
"learning_rate": 0.0009422369855868494,
|
| 597 |
+
"loss": 0.08561205118894577,
|
| 598 |
+
"step": 15616
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.16153967502760688,
|
| 602 |
+
"grad_norm": 0.10633770376443863,
|
| 603 |
+
"learning_rate": 0.0009403024438683983,
|
| 604 |
+
"loss": 0.08642981946468353,
|
| 605 |
+
"step": 15872
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.1641451536570844,
|
| 609 |
+
"grad_norm": 0.07606809586286545,
|
| 610 |
+
"learning_rate": 0.0009383380924920099,
|
| 611 |
+
"loss": 0.08625014871358871,
|
| 612 |
+
"step": 16128
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.16675063228656195,
|
| 616 |
+
"grad_norm": 0.1325465440750122,
|
| 617 |
+
"learning_rate": 0.0009363440644495478,
|
| 618 |
+
"loss": 0.08541498333215714,
|
| 619 |
+
"step": 16384
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.16675063228656195,
|
| 623 |
+
"eval_bce_loss": 1.0740072871957507,
|
| 624 |
+
"eval_bleu": 0.0007592252289275898,
|
| 625 |
+
"eval_cos_loss": 0.22020856738090516,
|
| 626 |
+
"eval_loss": 0.0942037650517055,
|
| 627 |
+
"eval_mean_score": 0.7717494419642857,
|
| 628 |
+
"eval_mse_loss": 0.08346369223935264,
|
| 629 |
+
"step": 16384
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 0.16675063228656195,
|
| 633 |
+
"eval_bce_loss": 1.0740072871957507,
|
| 634 |
+
"eval_bleu": 0.0007592252289275898,
|
| 635 |
+
"eval_cos_loss": 0.22020856738090516,
|
| 636 |
+
"eval_loss": 0.0942037650517055,
|
| 637 |
+
"eval_mean_score": 0.7717494419642857,
|
| 638 |
+
"eval_mse_loss": 0.08346369223935264,
|
| 639 |
+
"eval_runtime": 18.6586,
|
| 640 |
+
"eval_samples_per_second": 117.908,
|
| 641 |
+
"eval_steps_per_second": 1.876,
|
| 642 |
+
"step": 16384
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.16935611091603947,
|
| 646 |
+
"grad_norm": 0.091303251683712,
|
| 647 |
+
"learning_rate": 0.0009343204947420658,
|
| 648 |
+
"loss": 0.08596814423799515,
|
| 649 |
+
"step": 16640
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.171961589545517,
|
| 653 |
+
"grad_norm": 0.12643784284591675,
|
| 654 |
+
"learning_rate": 0.0009322675203706674,
|
| 655 |
+
"loss": 0.08583420515060425,
|
| 656 |
+
"step": 16896
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.17456706817499454,
|
| 660 |
+
"grad_norm": 0.07990458607673645,
|
| 661 |
+
"learning_rate": 0.0009301852803272315,
|
| 662 |
+
"loss": 0.08544512093067169,
|
| 663 |
+
"step": 17152
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.17717254680447206,
|
| 667 |
+
"grad_norm": 0.08775606006383896,
|
| 668 |
+
"learning_rate": 0.0009280739155850008,
|
| 669 |
+
"loss": 0.08551637083292007,
|
| 670 |
+
"step": 17408
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 0.17977802543394958,
|
| 674 |
+
"grad_norm": 0.1105438694357872,
|
| 675 |
+
"learning_rate": 0.0009259335690890387,
|
| 676 |
+
"loss": 0.08535689860582352,
|
| 677 |
+
"step": 17664
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.18238350406342713,
|
| 681 |
+
"grad_norm": 0.20936648547649384,
|
| 682 |
+
"learning_rate": 0.0009237643857465513,
|
| 683 |
+
"loss": 0.08509912341833115,
|
| 684 |
+
"step": 17920
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.18498898269290465,
|
| 688 |
+
"grad_norm": 0.14499887824058533,
|
| 689 |
+
"learning_rate": 0.0009215665124170764,
|
| 690 |
+
"loss": 0.08604084700345993,
|
| 691 |
+
"step": 18176
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.1875944613223822,
|
| 695 |
+
"grad_norm": 0.12544849514961243,
|
| 696 |
+
"learning_rate": 0.0009193400979025412,
|
| 697 |
+
"loss": 0.08515645563602448,
|
| 698 |
+
"step": 18432
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.1875944613223822,
|
| 702 |
+
"eval_bce_loss": 1.1422579322542463,
|
| 703 |
+
"eval_bleu": 0.0,
|
| 704 |
+
"eval_cos_loss": 0.2127199751990182,
|
| 705 |
+
"eval_loss": 0.09097630232572555,
|
| 706 |
+
"eval_mean_score": 0.7556780133928571,
|
| 707 |
+
"eval_mse_loss": 0.0795537236545767,
|
| 708 |
+
"step": 18432
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 0.1875944613223822,
|
| 712 |
+
"eval_bce_loss": 1.1422579322542463,
|
| 713 |
+
"eval_bleu": 0.0,
|
| 714 |
+
"eval_cos_loss": 0.2127199751990182,
|
| 715 |
+
"eval_loss": 0.09097630232572555,
|
| 716 |
+
"eval_mean_score": 0.7556780133928571,
|
| 717 |
+
"eval_mse_loss": 0.0795537236545767,
|
| 718 |
+
"eval_runtime": 18.7909,
|
| 719 |
+
"eval_samples_per_second": 117.078,
|
| 720 |
+
"eval_steps_per_second": 1.863,
|
| 721 |
+
"step": 18432
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 0.19019993995185971,
|
| 725 |
+
"grad_norm": 0.08133585005998611,
|
| 726 |
+
"learning_rate": 0.0009170852929371873,
|
| 727 |
+
"loss": 0.08439277112483978,
|
| 728 |
+
"step": 18688
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.19280541858133723,
|
| 732 |
+
"grad_norm": 0.06131386756896973,
|
| 733 |
+
"learning_rate": 0.0009148022501773669,
|
| 734 |
+
"loss": 0.08581775426864624,
|
| 735 |
+
"step": 18944
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.19541089721081478,
|
| 739 |
+
"grad_norm": 0.07301361113786697,
|
| 740 |
+
"learning_rate": 0.0009124911241912059,
|
| 741 |
+
"loss": 0.08452904224395752,
|
| 742 |
+
"step": 19200
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 0.1980163758402923,
|
| 746 |
+
"grad_norm": 0.19900229573249817,
|
| 747 |
+
"learning_rate": 0.0009101520714481405,
|
| 748 |
+
"loss": 0.08590829372406006,
|
| 749 |
+
"step": 19456
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.20062185446976982,
|
| 753 |
+
"grad_norm": 0.09695785492658615,
|
| 754 |
+
"learning_rate": 0.0009077852503083233,
|
| 755 |
+
"loss": 0.08517937362194061,
|
| 756 |
+
"step": 19712
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 0.20322733309924737,
|
| 760 |
+
"grad_norm": 0.10248035192489624,
|
| 761 |
+
"learning_rate": 0.0009053908210119015,
|
| 762 |
+
"loss": 0.08495298027992249,
|
| 763 |
+
"step": 19968
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 0.2058328117287249,
|
| 767 |
+
"grad_norm": 0.07756716758012772,
|
| 768 |
+
"learning_rate": 0.0009029689456681695,
|
| 769 |
+
"loss": 0.08380474150180817,
|
| 770 |
+
"step": 20224
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.2084382903582024,
|
| 774 |
+
"grad_norm": 0.12057510763406754,
|
| 775 |
+
"learning_rate": 0.000900519788244592,
|
| 776 |
+
"loss": 0.08552172780036926,
|
| 777 |
+
"step": 20480
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.2084382903582024,
|
| 781 |
+
"eval_bce_loss": 1.0288238414696285,
|
| 782 |
+
"eval_bleu": 0.0007499294204948674,
|
| 783 |
+
"eval_cos_loss": 0.21279827271189009,
|
| 784 |
+
"eval_loss": 0.0905734583735466,
|
| 785 |
+
"eval_mean_score": 0.7749860491071429,
|
| 786 |
+
"eval_mse_loss": 0.08028522048677716,
|
| 787 |
+
"step": 20480
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.2084382903582024,
|
| 791 |
+
"eval_bce_loss": 1.0288238414696285,
|
| 792 |
+
"eval_bleu": 0.0007499294204948674,
|
| 793 |
+
"eval_cos_loss": 0.21279827271189009,
|
| 794 |
+
"eval_loss": 0.0905734583735466,
|
| 795 |
+
"eval_mean_score": 0.7749860491071429,
|
| 796 |
+
"eval_mse_loss": 0.08028522048677716,
|
| 797 |
+
"eval_runtime": 18.3827,
|
| 798 |
+
"eval_samples_per_second": 119.678,
|
| 799 |
+
"eval_steps_per_second": 1.904,
|
| 800 |
+
"step": 20480
|
| 801 |
+
}
|
| 802 |
+
],
|
| 803 |
+
"logging_steps": 256,
|
| 804 |
+
"max_steps": 98255,
|
| 805 |
+
"num_input_tokens_seen": 0,
|
| 806 |
+
"num_train_epochs": 1,
|
| 807 |
+
"save_steps": 2048,
|
| 808 |
+
"stateful_callbacks": {
|
| 809 |
+
"TrainerControl": {
|
| 810 |
+
"args": {
|
| 811 |
+
"should_epoch_stop": false,
|
| 812 |
+
"should_evaluate": false,
|
| 813 |
+
"should_log": false,
|
| 814 |
+
"should_save": true,
|
| 815 |
+
"should_training_stop": false
|
| 816 |
+
},
|
| 817 |
+
"attributes": {}
|
| 818 |
+
}
|
| 819 |
+
},
|
| 820 |
+
"total_flos": 0.0,
|
| 821 |
+
"train_batch_size": 64,
|
| 822 |
+
"trial_name": null,
|
| 823 |
+
"trial_params": null
|
| 824 |
+
}
|
checkpoints-v1.0-discrete-semantc/checkpoint-20480/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7a20a7dfaef1ebb92c146eb8a8691f98d5b3e68f0184ad28ef70ddfe5a160a49
|
| 3 |
+
size 5137
|