Training in progress, step 9000, checkpoint
Browse files- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/rng_state_4.pth +1 -1
- last-checkpoint/rng_state_5.pth +1 -1
- last-checkpoint/rng_state_6.pth +1 -1
- last-checkpoint/rng_state_7.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +783 -3
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 570000656
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:25daec079aa80cd442292d34ae9d873f8a9267ecd9f2aeef7b9a48b444c47ee1
|
| 3 |
size 570000656
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1622091
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69fb1bb9b64b5f4a7bf23282e8dc9ea8c37def5ab69dfdd2f9b3c6bd5c8a324b
|
| 3 |
size 1622091
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7b61ff27b6922c34153fbf03ac386fe0a756cd98733dd45d5a0ef95c8d5649c5
|
| 3 |
size 14469
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7cae364e7cf38097315c3ef3b0f622833c3b1da572a0b8fdaa4146ae71eba341
|
| 3 |
size 14469
|
last-checkpoint/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:de03bb38bd2f03e2b491e76ba5e85ebc6c1c932d6775bf6a719f2932b85d90b7
|
| 3 |
size 14469
|
last-checkpoint/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:badf05b4369ee9be3ff96687da8330f3881cd2466e19a797479011694b1e2fc0
|
| 3 |
size 14469
|
last-checkpoint/rng_state_4.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ebc5b0944514c0301ec9d148f163a685abafbce8607513264faee4a1598baf22
|
| 3 |
size 14469
|
last-checkpoint/rng_state_5.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:63ac613d410082080b68756a552c9d14327f087f1b6a185023aa73b220e29feb
|
| 3 |
size 14469
|
last-checkpoint/rng_state_6.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99e9305aa1cd8522d5a7c8505c182973bb47b262f9d5d1c69921c48050ab7328
|
| 3 |
size 14469
|
last-checkpoint/rng_state_7.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0164255d679a6ba1b8dd248450ea3998324ef7163b69441493c8153561efabb1
|
| 3 |
size 14469
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:459b45dd9801b618e07700636012d164748c45e8cc4fe00a1caecc86da8ca65b
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 100,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": false,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -6248,6 +6248,786 @@
|
|
| 6248 |
"eval_samples_per_second": 16.498,
|
| 6249 |
"eval_steps_per_second": 0.917,
|
| 6250 |
"step": 8000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6251 |
}
|
| 6252 |
],
|
| 6253 |
"logging_steps": 10,
|
|
@@ -6267,7 +7047,7 @@
|
|
| 6267 |
"attributes": {}
|
| 6268 |
}
|
| 6269 |
},
|
| 6270 |
-
"total_flos":
|
| 6271 |
"train_batch_size": 4,
|
| 6272 |
"trial_name": null,
|
| 6273 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.8181818181818182,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 9000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": false,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 6248 |
"eval_samples_per_second": 16.498,
|
| 6249 |
"eval_steps_per_second": 0.917,
|
| 6250 |
"step": 8000
|
| 6251 |
+
},
|
| 6252 |
+
{
|
| 6253 |
+
"epoch": 0.7281818181818182,
|
| 6254 |
+
"grad_norm": 204158569283584.0,
|
| 6255 |
+
"learning_rate": 1.7159942866861132e-07,
|
| 6256 |
+
"loss": 26.29222412109375,
|
| 6257 |
+
"step": 8010
|
| 6258 |
+
},
|
| 6259 |
+
{
|
| 6260 |
+
"epoch": 0.7290909090909091,
|
| 6261 |
+
"grad_norm": 145085589094400.0,
|
| 6262 |
+
"learning_rate": 1.7052396883314152e-07,
|
| 6263 |
+
"loss": 26.959561157226563,
|
| 6264 |
+
"step": 8020
|
| 6265 |
+
},
|
| 6266 |
+
{
|
| 6267 |
+
"epoch": 0.73,
|
| 6268 |
+
"grad_norm": 110825029238784.0,
|
| 6269 |
+
"learning_rate": 1.6945119643228072e-07,
|
| 6270 |
+
"loss": 26.370242309570312,
|
| 6271 |
+
"step": 8030
|
| 6272 |
+
},
|
| 6273 |
+
{
|
| 6274 |
+
"epoch": 0.730909090909091,
|
| 6275 |
+
"grad_norm": 252334294695936.0,
|
| 6276 |
+
"learning_rate": 1.6838112021630356e-07,
|
| 6277 |
+
"loss": 26.4259765625,
|
| 6278 |
+
"step": 8040
|
| 6279 |
+
},
|
| 6280 |
+
{
|
| 6281 |
+
"epoch": 0.7318181818181818,
|
| 6282 |
+
"grad_norm": 193456450306048.0,
|
| 6283 |
+
"learning_rate": 1.673137489134921e-07,
|
| 6284 |
+
"loss": 26.351275634765624,
|
| 6285 |
+
"step": 8050
|
| 6286 |
+
},
|
| 6287 |
+
{
|
| 6288 |
+
"epoch": 0.7327272727272728,
|
| 6289 |
+
"grad_norm": 61600425836544.0,
|
| 6290 |
+
"learning_rate": 1.6624909123006575e-07,
|
| 6291 |
+
"loss": 26.925289916992188,
|
| 6292 |
+
"step": 8060
|
| 6293 |
+
},
|
| 6294 |
+
{
|
| 6295 |
+
"epoch": 0.7336363636363636,
|
| 6296 |
+
"grad_norm": 259621981782016.0,
|
| 6297 |
+
"learning_rate": 1.651871558501099e-07,
|
| 6298 |
+
"loss": 26.745242309570312,
|
| 6299 |
+
"step": 8070
|
| 6300 |
+
},
|
| 6301 |
+
{
|
| 6302 |
+
"epoch": 0.7345454545454545,
|
| 6303 |
+
"grad_norm": 150932784414720.0,
|
| 6304 |
+
"learning_rate": 1.6412795143550427e-07,
|
| 6305 |
+
"loss": 26.458816528320312,
|
| 6306 |
+
"step": 8080
|
| 6307 |
+
},
|
| 6308 |
+
{
|
| 6309 |
+
"epoch": 0.7354545454545455,
|
| 6310 |
+
"grad_norm": 522992664707072.0,
|
| 6311 |
+
"learning_rate": 1.6307148662585362e-07,
|
| 6312 |
+
"loss": 26.513919067382812,
|
| 6313 |
+
"step": 8090
|
| 6314 |
+
},
|
| 6315 |
+
{
|
| 6316 |
+
"epoch": 0.7363636363636363,
|
| 6317 |
+
"grad_norm": 129581780566016.0,
|
| 6318 |
+
"learning_rate": 1.6201777003841633e-07,
|
| 6319 |
+
"loss": 26.6354248046875,
|
| 6320 |
+
"step": 8100
|
| 6321 |
+
},
|
| 6322 |
+
{
|
| 6323 |
+
"epoch": 0.7363636363636363,
|
| 6324 |
+
"eval_loss": 26.260547637939453,
|
| 6325 |
+
"eval_runtime": 2.1543,
|
| 6326 |
+
"eval_samples_per_second": 16.711,
|
| 6327 |
+
"eval_steps_per_second": 0.928,
|
| 6328 |
+
"step": 8100
|
| 6329 |
+
},
|
| 6330 |
+
{
|
| 6331 |
+
"epoch": 0.7372727272727273,
|
| 6332 |
+
"grad_norm": 220818277662720.0,
|
| 6333 |
+
"learning_rate": 1.6096681026803443e-07,
|
| 6334 |
+
"loss": 26.771890258789064,
|
| 6335 |
+
"step": 8110
|
| 6336 |
+
},
|
| 6337 |
+
{
|
| 6338 |
+
"epoch": 0.7381818181818182,
|
| 6339 |
+
"grad_norm": 389124204789760.0,
|
| 6340 |
+
"learning_rate": 1.5991861588706306e-07,
|
| 6341 |
+
"loss": 26.528866577148438,
|
| 6342 |
+
"step": 8120
|
| 6343 |
+
},
|
| 6344 |
+
{
|
| 6345 |
+
"epoch": 0.7390909090909091,
|
| 6346 |
+
"grad_norm": 421828568809472.0,
|
| 6347 |
+
"learning_rate": 1.588731954453018e-07,
|
| 6348 |
+
"loss": 26.294219970703125,
|
| 6349 |
+
"step": 8130
|
| 6350 |
+
},
|
| 6351 |
+
{
|
| 6352 |
+
"epoch": 0.74,
|
| 6353 |
+
"grad_norm": 186962057101312.0,
|
| 6354 |
+
"learning_rate": 1.5783055746992312e-07,
|
| 6355 |
+
"loss": 26.486737060546876,
|
| 6356 |
+
"step": 8140
|
| 6357 |
+
},
|
| 6358 |
+
{
|
| 6359 |
+
"epoch": 0.740909090909091,
|
| 6360 |
+
"grad_norm": 320954349649920.0,
|
| 6361 |
+
"learning_rate": 1.5679071046540432e-07,
|
| 6362 |
+
"loss": 26.416116333007814,
|
| 6363 |
+
"step": 8150
|
| 6364 |
+
},
|
| 6365 |
+
{
|
| 6366 |
+
"epoch": 0.7418181818181818,
|
| 6367 |
+
"grad_norm": 448088032411648.0,
|
| 6368 |
+
"learning_rate": 1.5575366291345759e-07,
|
| 6369 |
+
"loss": 26.369647216796874,
|
| 6370 |
+
"step": 8160
|
| 6371 |
+
},
|
| 6372 |
+
{
|
| 6373 |
+
"epoch": 0.7427272727272727,
|
| 6374 |
+
"grad_norm": 405070747271168.0,
|
| 6375 |
+
"learning_rate": 1.547194232729604e-07,
|
| 6376 |
+
"loss": 26.7183837890625,
|
| 6377 |
+
"step": 8170
|
| 6378 |
+
},
|
| 6379 |
+
{
|
| 6380 |
+
"epoch": 0.7436363636363637,
|
| 6381 |
+
"grad_norm": 392587995250688.0,
|
| 6382 |
+
"learning_rate": 1.536879999798874e-07,
|
| 6383 |
+
"loss": 26.57523193359375,
|
| 6384 |
+
"step": 8180
|
| 6385 |
+
},
|
| 6386 |
+
{
|
| 6387 |
+
"epoch": 0.7445454545454545,
|
| 6388 |
+
"grad_norm": 414031860989952.0,
|
| 6389 |
+
"learning_rate": 1.5265940144724114e-07,
|
| 6390 |
+
"loss": 26.745343017578126,
|
| 6391 |
+
"step": 8190
|
| 6392 |
+
},
|
| 6393 |
+
{
|
| 6394 |
+
"epoch": 0.7454545454545455,
|
| 6395 |
+
"grad_norm": 151567047065600.0,
|
| 6396 |
+
"learning_rate": 1.516336360649834e-07,
|
| 6397 |
+
"loss": 26.580767822265624,
|
| 6398 |
+
"step": 8200
|
| 6399 |
+
},
|
| 6400 |
+
{
|
| 6401 |
+
"epoch": 0.7454545454545455,
|
| 6402 |
+
"eval_loss": 26.26454734802246,
|
| 6403 |
+
"eval_runtime": 2.1459,
|
| 6404 |
+
"eval_samples_per_second": 16.776,
|
| 6405 |
+
"eval_steps_per_second": 0.932,
|
| 6406 |
+
"step": 8200
|
| 6407 |
+
},
|
| 6408 |
+
{
|
| 6409 |
+
"epoch": 0.7463636363636363,
|
| 6410 |
+
"grad_norm": 140819075956736.0,
|
| 6411 |
+
"learning_rate": 1.506107121999664e-07,
|
| 6412 |
+
"loss": 26.278732299804688,
|
| 6413 |
+
"step": 8210
|
| 6414 |
+
},
|
| 6415 |
+
{
|
| 6416 |
+
"epoch": 0.7472727272727273,
|
| 6417 |
+
"grad_norm": 66263875321856.0,
|
| 6418 |
+
"learning_rate": 1.4959063819586593e-07,
|
| 6419 |
+
"loss": 26.610507202148437,
|
| 6420 |
+
"step": 8220
|
| 6421 |
+
},
|
| 6422 |
+
{
|
| 6423 |
+
"epoch": 0.7481818181818182,
|
| 6424 |
+
"grad_norm": 438402814050304.0,
|
| 6425 |
+
"learning_rate": 1.4857342237311138e-07,
|
| 6426 |
+
"loss": 26.609988403320312,
|
| 6427 |
+
"step": 8230
|
| 6428 |
+
},
|
| 6429 |
+
{
|
| 6430 |
+
"epoch": 0.7490909090909091,
|
| 6431 |
+
"grad_norm": 509375739330560.0,
|
| 6432 |
+
"learning_rate": 1.4755907302881927e-07,
|
| 6433 |
+
"loss": 26.591030883789063,
|
| 6434 |
+
"step": 8240
|
| 6435 |
+
},
|
| 6436 |
+
{
|
| 6437 |
+
"epoch": 0.75,
|
| 6438 |
+
"grad_norm": 469766544293888.0,
|
| 6439 |
+
"learning_rate": 1.465475984367253e-07,
|
| 6440 |
+
"loss": 26.608944702148438,
|
| 6441 |
+
"step": 8250
|
| 6442 |
+
},
|
| 6443 |
+
{
|
| 6444 |
+
"epoch": 0.7509090909090909,
|
| 6445 |
+
"grad_norm": 205435751628800.0,
|
| 6446 |
+
"learning_rate": 1.4553900684711623e-07,
|
| 6447 |
+
"loss": 26.571286010742188,
|
| 6448 |
+
"step": 8260
|
| 6449 |
+
},
|
| 6450 |
+
{
|
| 6451 |
+
"epoch": 0.7518181818181818,
|
| 6452 |
+
"grad_norm": 146254860386304.0,
|
| 6453 |
+
"learning_rate": 1.4453330648676338e-07,
|
| 6454 |
+
"loss": 26.6401611328125,
|
| 6455 |
+
"step": 8270
|
| 6456 |
+
},
|
| 6457 |
+
{
|
| 6458 |
+
"epoch": 0.7527272727272727,
|
| 6459 |
+
"grad_norm": 406260688093184.0,
|
| 6460 |
+
"learning_rate": 1.435305055588552e-07,
|
| 6461 |
+
"loss": 26.502618408203126,
|
| 6462 |
+
"step": 8280
|
| 6463 |
+
},
|
| 6464 |
+
{
|
| 6465 |
+
"epoch": 0.7536363636363637,
|
| 6466 |
+
"grad_norm": 163714019885056.0,
|
| 6467 |
+
"learning_rate": 1.425306122429305e-07,
|
| 6468 |
+
"loss": 26.30799560546875,
|
| 6469 |
+
"step": 8290
|
| 6470 |
+
},
|
| 6471 |
+
{
|
| 6472 |
+
"epoch": 0.7545454545454545,
|
| 6473 |
+
"grad_norm": 265974758506496.0,
|
| 6474 |
+
"learning_rate": 1.4153363469481107e-07,
|
| 6475 |
+
"loss": 26.497518920898436,
|
| 6476 |
+
"step": 8300
|
| 6477 |
+
},
|
| 6478 |
+
{
|
| 6479 |
+
"epoch": 0.7545454545454545,
|
| 6480 |
+
"eval_loss": 26.25794792175293,
|
| 6481 |
+
"eval_runtime": 2.1488,
|
| 6482 |
+
"eval_samples_per_second": 16.753,
|
| 6483 |
+
"eval_steps_per_second": 0.931,
|
| 6484 |
+
"step": 8300
|
| 6485 |
+
},
|
| 6486 |
+
{
|
| 6487 |
+
"epoch": 0.7554545454545455,
|
| 6488 |
+
"grad_norm": 144185525010432.0,
|
| 6489 |
+
"learning_rate": 1.4053958104653612e-07,
|
| 6490 |
+
"loss": 26.401416015625,
|
| 6491 |
+
"step": 8310
|
| 6492 |
+
},
|
| 6493 |
+
{
|
| 6494 |
+
"epoch": 0.7563636363636363,
|
| 6495 |
+
"grad_norm": 80316861513728.0,
|
| 6496 |
+
"learning_rate": 1.395484594062954e-07,
|
| 6497 |
+
"loss": 26.3766357421875,
|
| 6498 |
+
"step": 8320
|
| 6499 |
+
},
|
| 6500 |
+
{
|
| 6501 |
+
"epoch": 0.7572727272727273,
|
| 6502 |
+
"grad_norm": 405346531147776.0,
|
| 6503 |
+
"learning_rate": 1.3856027785836333e-07,
|
| 6504 |
+
"loss": 26.344137573242186,
|
| 6505 |
+
"step": 8330
|
| 6506 |
+
},
|
| 6507 |
+
{
|
| 6508 |
+
"epoch": 0.7581818181818182,
|
| 6509 |
+
"grad_norm": 155818091610112.0,
|
| 6510 |
+
"learning_rate": 1.3757504446303248e-07,
|
| 6511 |
+
"loss": 26.66131591796875,
|
| 6512 |
+
"step": 8340
|
| 6513 |
+
},
|
| 6514 |
+
{
|
| 6515 |
+
"epoch": 0.759090909090909,
|
| 6516 |
+
"grad_norm": 264115809419264.0,
|
| 6517 |
+
"learning_rate": 1.365927672565486e-07,
|
| 6518 |
+
"loss": 26.72258605957031,
|
| 6519 |
+
"step": 8350
|
| 6520 |
+
},
|
| 6521 |
+
{
|
| 6522 |
+
"epoch": 0.76,
|
| 6523 |
+
"grad_norm": 289586693537792.0,
|
| 6524 |
+
"learning_rate": 1.356134542510448e-07,
|
| 6525 |
+
"loss": 26.394635009765626,
|
| 6526 |
+
"step": 8360
|
| 6527 |
+
},
|
| 6528 |
+
{
|
| 6529 |
+
"epoch": 0.7609090909090909,
|
| 6530 |
+
"grad_norm": 74874424918016.0,
|
| 6531 |
+
"learning_rate": 1.3463711343447547e-07,
|
| 6532 |
+
"loss": 26.273269653320312,
|
| 6533 |
+
"step": 8370
|
| 6534 |
+
},
|
| 6535 |
+
{
|
| 6536 |
+
"epoch": 0.7618181818181818,
|
| 6537 |
+
"grad_norm": 576587481219072.0,
|
| 6538 |
+
"learning_rate": 1.336637527705529e-07,
|
| 6539 |
+
"loss": 26.7587890625,
|
| 6540 |
+
"step": 8380
|
| 6541 |
+
},
|
| 6542 |
+
{
|
| 6543 |
+
"epoch": 0.7627272727272727,
|
| 6544 |
+
"grad_norm": 132786539200512.0,
|
| 6545 |
+
"learning_rate": 1.3269338019868003e-07,
|
| 6546 |
+
"loss": 26.890203857421874,
|
| 6547 |
+
"step": 8390
|
| 6548 |
+
},
|
| 6549 |
+
{
|
| 6550 |
+
"epoch": 0.7636363636363637,
|
| 6551 |
+
"grad_norm": 113306773749760.0,
|
| 6552 |
+
"learning_rate": 1.3172600363388752e-07,
|
| 6553 |
+
"loss": 26.88782958984375,
|
| 6554 |
+
"step": 8400
|
| 6555 |
+
},
|
| 6556 |
+
{
|
| 6557 |
+
"epoch": 0.7636363636363637,
|
| 6558 |
+
"eval_loss": 26.261314392089844,
|
| 6559 |
+
"eval_runtime": 2.1523,
|
| 6560 |
+
"eval_samples_per_second": 16.726,
|
| 6561 |
+
"eval_steps_per_second": 0.929,
|
| 6562 |
+
"step": 8400
|
| 6563 |
+
},
|
| 6564 |
+
{
|
| 6565 |
+
"epoch": 0.7645454545454545,
|
| 6566 |
+
"grad_norm": 287690733912064.0,
|
| 6567 |
+
"learning_rate": 1.3076163096676828e-07,
|
| 6568 |
+
"loss": 25.92532958984375,
|
| 6569 |
+
"step": 8410
|
| 6570 |
+
},
|
| 6571 |
+
{
|
| 6572 |
+
"epoch": 0.7654545454545455,
|
| 6573 |
+
"grad_norm": 170665340567552.0,
|
| 6574 |
+
"learning_rate": 1.2980027006341371e-07,
|
| 6575 |
+
"loss": 26.285894775390624,
|
| 6576 |
+
"step": 8420
|
| 6577 |
+
},
|
| 6578 |
+
{
|
| 6579 |
+
"epoch": 0.7663636363636364,
|
| 6580 |
+
"grad_norm": 293886157127680.0,
|
| 6581 |
+
"learning_rate": 1.288419287653485e-07,
|
| 6582 |
+
"loss": 26.6665771484375,
|
| 6583 |
+
"step": 8430
|
| 6584 |
+
},
|
| 6585 |
+
{
|
| 6586 |
+
"epoch": 0.7672727272727272,
|
| 6587 |
+
"grad_norm": 211919927508992.0,
|
| 6588 |
+
"learning_rate": 1.2788661488946788e-07,
|
| 6589 |
+
"loss": 26.39056396484375,
|
| 6590 |
+
"step": 8440
|
| 6591 |
+
},
|
| 6592 |
+
{
|
| 6593 |
+
"epoch": 0.7681818181818182,
|
| 6594 |
+
"grad_norm": 160566647717888.0,
|
| 6595 |
+
"learning_rate": 1.2693433622797346e-07,
|
| 6596 |
+
"loss": 26.330914306640626,
|
| 6597 |
+
"step": 8450
|
| 6598 |
+
},
|
| 6599 |
+
{
|
| 6600 |
+
"epoch": 0.769090909090909,
|
| 6601 |
+
"grad_norm": 168821155430400.0,
|
| 6602 |
+
"learning_rate": 1.2598510054830885e-07,
|
| 6603 |
+
"loss": 26.57433166503906,
|
| 6604 |
+
"step": 8460
|
| 6605 |
+
},
|
| 6606 |
+
{
|
| 6607 |
+
"epoch": 0.77,
|
| 6608 |
+
"grad_norm": 341319708835840.0,
|
| 6609 |
+
"learning_rate": 1.250389155930976e-07,
|
| 6610 |
+
"loss": 26.557040405273437,
|
| 6611 |
+
"step": 8470
|
| 6612 |
+
},
|
| 6613 |
+
{
|
| 6614 |
+
"epoch": 0.7709090909090909,
|
| 6615 |
+
"grad_norm": 291579591917568.0,
|
| 6616 |
+
"learning_rate": 1.240957890800793e-07,
|
| 6617 |
+
"loss": 26.230010986328125,
|
| 6618 |
+
"step": 8480
|
| 6619 |
+
},
|
| 6620 |
+
{
|
| 6621 |
+
"epoch": 0.7718181818181818,
|
| 6622 |
+
"grad_norm": 246938574258176.0,
|
| 6623 |
+
"learning_rate": 1.2315572870204665e-07,
|
| 6624 |
+
"loss": 26.432733154296876,
|
| 6625 |
+
"step": 8490
|
| 6626 |
+
},
|
| 6627 |
+
{
|
| 6628 |
+
"epoch": 0.7727272727272727,
|
| 6629 |
+
"grad_norm": 275942924615680.0,
|
| 6630 |
+
"learning_rate": 1.2221874212678295e-07,
|
| 6631 |
+
"loss": 26.655825805664062,
|
| 6632 |
+
"step": 8500
|
| 6633 |
+
},
|
| 6634 |
+
{
|
| 6635 |
+
"epoch": 0.7727272727272727,
|
| 6636 |
+
"eval_loss": 26.262447357177734,
|
| 6637 |
+
"eval_runtime": 2.1593,
|
| 6638 |
+
"eval_samples_per_second": 16.672,
|
| 6639 |
+
"eval_steps_per_second": 0.926,
|
| 6640 |
+
"step": 8500
|
| 6641 |
+
},
|
| 6642 |
+
{
|
| 6643 |
+
"epoch": 0.7736363636363637,
|
| 6644 |
+
"grad_norm": 252627107446784.0,
|
| 6645 |
+
"learning_rate": 1.212848369969996e-07,
|
| 6646 |
+
"loss": 26.5862548828125,
|
| 6647 |
+
"step": 8510
|
| 6648 |
+
},
|
| 6649 |
+
{
|
| 6650 |
+
"epoch": 0.7745454545454545,
|
| 6651 |
+
"grad_norm": 349064071741440.0,
|
| 6652 |
+
"learning_rate": 1.2035402093027298e-07,
|
| 6653 |
+
"loss": 26.767044067382812,
|
| 6654 |
+
"step": 8520
|
| 6655 |
+
},
|
| 6656 |
+
{
|
| 6657 |
+
"epoch": 0.7754545454545455,
|
| 6658 |
+
"grad_norm": 215884400427008.0,
|
| 6659 |
+
"learning_rate": 1.1942630151898353e-07,
|
| 6660 |
+
"loss": 26.34154357910156,
|
| 6661 |
+
"step": 8530
|
| 6662 |
+
},
|
| 6663 |
+
{
|
| 6664 |
+
"epoch": 0.7763636363636364,
|
| 6665 |
+
"grad_norm": 109126453559296.0,
|
| 6666 |
+
"learning_rate": 1.1850168633025321e-07,
|
| 6667 |
+
"loss": 26.6445556640625,
|
| 6668 |
+
"step": 8540
|
| 6669 |
+
},
|
| 6670 |
+
{
|
| 6671 |
+
"epoch": 0.7772727272727272,
|
| 6672 |
+
"grad_norm": 59685176606720.0,
|
| 6673 |
+
"learning_rate": 1.1758018290588328e-07,
|
| 6674 |
+
"loss": 26.464410400390626,
|
| 6675 |
+
"step": 8550
|
| 6676 |
+
},
|
| 6677 |
+
{
|
| 6678 |
+
"epoch": 0.7781818181818182,
|
| 6679 |
+
"grad_norm": 336345364955136.0,
|
| 6680 |
+
"learning_rate": 1.1666179876229382e-07,
|
| 6681 |
+
"loss": 26.180218505859376,
|
| 6682 |
+
"step": 8560
|
| 6683 |
+
},
|
| 6684 |
+
{
|
| 6685 |
+
"epoch": 0.7790909090909091,
|
| 6686 |
+
"grad_norm": 258242475196416.0,
|
| 6687 |
+
"learning_rate": 1.1574654139046169e-07,
|
| 6688 |
+
"loss": 26.63001708984375,
|
| 6689 |
+
"step": 8570
|
| 6690 |
+
},
|
| 6691 |
+
{
|
| 6692 |
+
"epoch": 0.78,
|
| 6693 |
+
"grad_norm": 308295034208256.0,
|
| 6694 |
+
"learning_rate": 1.1483441825585982e-07,
|
| 6695 |
+
"loss": 26.5321044921875,
|
| 6696 |
+
"step": 8580
|
| 6697 |
+
},
|
| 6698 |
+
{
|
| 6699 |
+
"epoch": 0.7809090909090909,
|
| 6700 |
+
"grad_norm": 256095578750976.0,
|
| 6701 |
+
"learning_rate": 1.139254367983955e-07,
|
| 6702 |
+
"loss": 26.616192626953126,
|
| 6703 |
+
"step": 8590
|
| 6704 |
+
},
|
| 6705 |
+
{
|
| 6706 |
+
"epoch": 0.7818181818181819,
|
| 6707 |
+
"grad_norm": 433377970749440.0,
|
| 6708 |
+
"learning_rate": 1.1301960443235137e-07,
|
| 6709 |
+
"loss": 26.703271484375,
|
| 6710 |
+
"step": 8600
|
| 6711 |
+
},
|
| 6712 |
+
{
|
| 6713 |
+
"epoch": 0.7818181818181819,
|
| 6714 |
+
"eval_loss": 26.259233474731445,
|
| 6715 |
+
"eval_runtime": 2.1684,
|
| 6716 |
+
"eval_samples_per_second": 16.602,
|
| 6717 |
+
"eval_steps_per_second": 0.922,
|
| 6718 |
+
"step": 8600
|
| 6719 |
+
},
|
| 6720 |
+
{
|
| 6721 |
+
"epoch": 0.7827272727272727,
|
| 6722 |
+
"grad_norm": 91433688104960.0,
|
| 6723 |
+
"learning_rate": 1.1211692854632292e-07,
|
| 6724 |
+
"loss": 26.381759643554688,
|
| 6725 |
+
"step": 8610
|
| 6726 |
+
},
|
| 6727 |
+
{
|
| 6728 |
+
"epoch": 0.7836363636363637,
|
| 6729 |
+
"grad_norm": 323897308217344.0,
|
| 6730 |
+
"learning_rate": 1.1121741650315975e-07,
|
| 6731 |
+
"loss": 26.4834228515625,
|
| 6732 |
+
"step": 8620
|
| 6733 |
+
},
|
| 6734 |
+
{
|
| 6735 |
+
"epoch": 0.7845454545454545,
|
| 6736 |
+
"grad_norm": 615513239584768.0,
|
| 6737 |
+
"learning_rate": 1.1032107563990489e-07,
|
| 6738 |
+
"loss": 26.638949584960937,
|
| 6739 |
+
"step": 8630
|
| 6740 |
+
},
|
| 6741 |
+
{
|
| 6742 |
+
"epoch": 0.7854545454545454,
|
| 6743 |
+
"grad_norm": 144044713836544.0,
|
| 6744 |
+
"learning_rate": 1.0942791326773481e-07,
|
| 6745 |
+
"loss": 26.390069580078126,
|
| 6746 |
+
"step": 8640
|
| 6747 |
+
},
|
| 6748 |
+
{
|
| 6749 |
+
"epoch": 0.7863636363636364,
|
| 6750 |
+
"grad_norm": 226855877607424.0,
|
| 6751 |
+
"learning_rate": 1.085379366719002e-07,
|
| 6752 |
+
"loss": 26.581057739257812,
|
| 6753 |
+
"step": 8650
|
| 6754 |
+
},
|
| 6755 |
+
{
|
| 6756 |
+
"epoch": 0.7872727272727272,
|
| 6757 |
+
"grad_norm": 750627542007808.0,
|
| 6758 |
+
"learning_rate": 1.0765115311166634e-07,
|
| 6759 |
+
"loss": 26.551568603515626,
|
| 6760 |
+
"step": 8660
|
| 6761 |
+
},
|
| 6762 |
+
{
|
| 6763 |
+
"epoch": 0.7881818181818182,
|
| 6764 |
+
"grad_norm": 386196177944576.0,
|
| 6765 |
+
"learning_rate": 1.0676756982025403e-07,
|
| 6766 |
+
"loss": 26.428280639648438,
|
| 6767 |
+
"step": 8670
|
| 6768 |
+
},
|
| 6769 |
+
{
|
| 6770 |
+
"epoch": 0.7890909090909091,
|
| 6771 |
+
"grad_norm": 159711546572800.0,
|
| 6772 |
+
"learning_rate": 1.0588719400478002e-07,
|
| 6773 |
+
"loss": 26.626437377929687,
|
| 6774 |
+
"step": 8680
|
| 6775 |
+
},
|
| 6776 |
+
{
|
| 6777 |
+
"epoch": 0.79,
|
| 6778 |
+
"grad_norm": 170563821633536.0,
|
| 6779 |
+
"learning_rate": 1.0501003284619908e-07,
|
| 6780 |
+
"loss": 26.475860595703125,
|
| 6781 |
+
"step": 8690
|
| 6782 |
+
},
|
| 6783 |
+
{
|
| 6784 |
+
"epoch": 0.7909090909090909,
|
| 6785 |
+
"grad_norm": 134558968184832.0,
|
| 6786 |
+
"learning_rate": 1.041360934992449e-07,
|
| 6787 |
+
"loss": 26.157559204101563,
|
| 6788 |
+
"step": 8700
|
| 6789 |
+
},
|
| 6790 |
+
{
|
| 6791 |
+
"epoch": 0.7909090909090909,
|
| 6792 |
+
"eval_loss": 26.25774574279785,
|
| 6793 |
+
"eval_runtime": 2.1872,
|
| 6794 |
+
"eval_samples_per_second": 16.46,
|
| 6795 |
+
"eval_steps_per_second": 0.914,
|
| 6796 |
+
"step": 8700
|
| 6797 |
+
},
|
| 6798 |
+
{
|
| 6799 |
+
"epoch": 0.7918181818181819,
|
| 6800 |
+
"grad_norm": 72320622264320.0,
|
| 6801 |
+
"learning_rate": 1.0326538309237198e-07,
|
| 6802 |
+
"loss": 26.85352478027344,
|
| 6803 |
+
"step": 8710
|
| 6804 |
+
},
|
| 6805 |
+
{
|
| 6806 |
+
"epoch": 0.7927272727272727,
|
| 6807 |
+
"grad_norm": 481769501491200.0,
|
| 6808 |
+
"learning_rate": 1.0239790872769688e-07,
|
| 6809 |
+
"loss": 26.115414428710938,
|
| 6810 |
+
"step": 8720
|
| 6811 |
+
},
|
| 6812 |
+
{
|
| 6813 |
+
"epoch": 0.7936363636363636,
|
| 6814 |
+
"grad_norm": 342039317184512.0,
|
| 6815 |
+
"learning_rate": 1.0153367748094126e-07,
|
| 6816 |
+
"loss": 26.53397216796875,
|
| 6817 |
+
"step": 8730
|
| 6818 |
+
},
|
| 6819 |
+
{
|
| 6820 |
+
"epoch": 0.7945454545454546,
|
| 6821 |
+
"grad_norm": 404055557931008.0,
|
| 6822 |
+
"learning_rate": 1.006726964013736e-07,
|
| 6823 |
+
"loss": 26.39664306640625,
|
| 6824 |
+
"step": 8740
|
| 6825 |
+
},
|
| 6826 |
+
{
|
| 6827 |
+
"epoch": 0.7954545454545454,
|
| 6828 |
+
"grad_norm": 81613387988992.0,
|
| 6829 |
+
"learning_rate": 9.98149725117512e-08,
|
| 6830 |
+
"loss": 26.064663696289063,
|
| 6831 |
+
"step": 8750
|
| 6832 |
+
},
|
| 6833 |
+
{
|
| 6834 |
+
"epoch": 0.7963636363636364,
|
| 6835 |
+
"grad_norm": 350235490516992.0,
|
| 6836 |
+
"learning_rate": 9.896051280826451e-08,
|
| 6837 |
+
"loss": 26.903762817382812,
|
| 6838 |
+
"step": 8760
|
| 6839 |
+
},
|
| 6840 |
+
{
|
| 6841 |
+
"epoch": 0.7972727272727272,
|
| 6842 |
+
"grad_norm": 459837653647360.0,
|
| 6843 |
+
"learning_rate": 9.810932426047802e-08,
|
| 6844 |
+
"loss": 26.57601318359375,
|
| 6845 |
+
"step": 8770
|
| 6846 |
+
},
|
| 6847 |
+
{
|
| 6848 |
+
"epoch": 0.7981818181818182,
|
| 6849 |
+
"grad_norm": 147279679848448.0,
|
| 6850 |
+
"learning_rate": 9.72614138112749e-08,
|
| 6851 |
+
"loss": 26.590640258789062,
|
| 6852 |
+
"step": 8780
|
| 6853 |
+
},
|
| 6854 |
+
{
|
| 6855 |
+
"epoch": 0.7990909090909091,
|
| 6856 |
+
"grad_norm": 417567155945472.0,
|
| 6857 |
+
"learning_rate": 9.641678837679983e-08,
|
| 6858 |
+
"loss": 26.296502685546876,
|
| 6859 |
+
"step": 8790
|
| 6860 |
+
},
|
| 6861 |
+
{
|
| 6862 |
+
"epoch": 0.8,
|
| 6863 |
+
"grad_norm": 84892301918208.0,
|
| 6864 |
+
"learning_rate": 9.557545484640273e-08,
|
| 6865 |
+
"loss": 26.586184692382812,
|
| 6866 |
+
"step": 8800
|
| 6867 |
+
},
|
| 6868 |
+
{
|
| 6869 |
+
"epoch": 0.8,
|
| 6870 |
+
"eval_loss": 26.26275634765625,
|
| 6871 |
+
"eval_runtime": 2.1775,
|
| 6872 |
+
"eval_samples_per_second": 16.532,
|
| 6873 |
+
"eval_steps_per_second": 0.918,
|
| 6874 |
+
"step": 8800
|
| 6875 |
+
},
|
| 6876 |
+
{
|
| 6877 |
+
"epoch": 0.8009090909090909,
|
| 6878 |
+
"grad_norm": 709094000295936.0,
|
| 6879 |
+
"learning_rate": 9.4737420082582e-08,
|
| 6880 |
+
"loss": 26.503372192382812,
|
| 6881 |
+
"step": 8810
|
| 6882 |
+
},
|
| 6883 |
+
{
|
| 6884 |
+
"epoch": 0.8018181818181818,
|
| 6885 |
+
"grad_norm": 214880770588672.0,
|
| 6886 |
+
"learning_rate": 9.390269092092956e-08,
|
| 6887 |
+
"loss": 26.727926635742186,
|
| 6888 |
+
"step": 8820
|
| 6889 |
+
},
|
| 6890 |
+
{
|
| 6891 |
+
"epoch": 0.8027272727272727,
|
| 6892 |
+
"grad_norm": 206503084556288.0,
|
| 6893 |
+
"learning_rate": 9.307127417007444e-08,
|
| 6894 |
+
"loss": 26.764566040039064,
|
| 6895 |
+
"step": 8830
|
| 6896 |
+
},
|
| 6897 |
+
{
|
| 6898 |
+
"epoch": 0.8036363636363636,
|
| 6899 |
+
"grad_norm": 105237830434816.0,
|
| 6900 |
+
"learning_rate": 9.224317661162684e-08,
|
| 6901 |
+
"loss": 26.60999755859375,
|
| 6902 |
+
"step": 8840
|
| 6903 |
+
},
|
| 6904 |
+
{
|
| 6905 |
+
"epoch": 0.8045454545454546,
|
| 6906 |
+
"grad_norm": 101385144107008.0,
|
| 6907 |
+
"learning_rate": 9.141840500012437e-08,
|
| 6908 |
+
"loss": 26.486932373046876,
|
| 6909 |
+
"step": 8850
|
| 6910 |
+
},
|
| 6911 |
+
{
|
| 6912 |
+
"epoch": 0.8054545454545454,
|
| 6913 |
+
"grad_norm": 195066056409088.0,
|
| 6914 |
+
"learning_rate": 9.059696606297479e-08,
|
| 6915 |
+
"loss": 26.113775634765624,
|
| 6916 |
+
"step": 8860
|
| 6917 |
+
},
|
| 6918 |
+
{
|
| 6919 |
+
"epoch": 0.8063636363636364,
|
| 6920 |
+
"grad_norm": 116232594391040.0,
|
| 6921 |
+
"learning_rate": 8.977886650040289e-08,
|
| 6922 |
+
"loss": 26.361734008789064,
|
| 6923 |
+
"step": 8870
|
| 6924 |
+
},
|
| 6925 |
+
{
|
| 6926 |
+
"epoch": 0.8072727272727273,
|
| 6927 |
+
"grad_norm": 459080464334848.0,
|
| 6928 |
+
"learning_rate": 8.89641129853952e-08,
|
| 6929 |
+
"loss": 26.5644287109375,
|
| 6930 |
+
"step": 8880
|
| 6931 |
+
},
|
| 6932 |
+
{
|
| 6933 |
+
"epoch": 0.8081818181818182,
|
| 6934 |
+
"grad_norm": 279257364299776.0,
|
| 6935 |
+
"learning_rate": 8.81527121636454e-08,
|
| 6936 |
+
"loss": 26.242138671875,
|
| 6937 |
+
"step": 8890
|
| 6938 |
+
},
|
| 6939 |
+
{
|
| 6940 |
+
"epoch": 0.8090909090909091,
|
| 6941 |
+
"grad_norm": 177674492313600.0,
|
| 6942 |
+
"learning_rate": 8.734467065350021e-08,
|
| 6943 |
+
"loss": 26.649420166015624,
|
| 6944 |
+
"step": 8900
|
| 6945 |
+
},
|
| 6946 |
+
{
|
| 6947 |
+
"epoch": 0.8090909090909091,
|
| 6948 |
+
"eval_loss": 26.255714416503906,
|
| 6949 |
+
"eval_runtime": 2.1572,
|
| 6950 |
+
"eval_samples_per_second": 16.689,
|
| 6951 |
+
"eval_steps_per_second": 0.927,
|
| 6952 |
+
"step": 8900
|
| 6953 |
+
},
|
| 6954 |
+
{
|
| 6955 |
+
"epoch": 0.81,
|
| 6956 |
+
"grad_norm": 408893402382336.0,
|
| 6957 |
+
"learning_rate": 8.65399950459057e-08,
|
| 6958 |
+
"loss": 26.4176025390625,
|
| 6959 |
+
"step": 8910
|
| 6960 |
+
},
|
| 6961 |
+
{
|
| 6962 |
+
"epoch": 0.8109090909090909,
|
| 6963 |
+
"grad_norm": 166783998754816.0,
|
| 6964 |
+
"learning_rate": 8.573869190435335e-08,
|
| 6965 |
+
"loss": 26.493902587890624,
|
| 6966 |
+
"step": 8920
|
| 6967 |
+
},
|
| 6968 |
+
{
|
| 6969 |
+
"epoch": 0.8118181818181818,
|
| 6970 |
+
"grad_norm": 176515757113344.0,
|
| 6971 |
+
"learning_rate": 8.4940767764826e-08,
|
| 6972 |
+
"loss": 26.102850341796874,
|
| 6973 |
+
"step": 8930
|
| 6974 |
+
},
|
| 6975 |
+
{
|
| 6976 |
+
"epoch": 0.8127272727272727,
|
| 6977 |
+
"grad_norm": 542467724148736.0,
|
| 6978 |
+
"learning_rate": 8.414622913574549e-08,
|
| 6979 |
+
"loss": 26.501849365234374,
|
| 6980 |
+
"step": 8940
|
| 6981 |
+
},
|
| 6982 |
+
{
|
| 6983 |
+
"epoch": 0.8136363636363636,
|
| 6984 |
+
"grad_norm": 216477290463232.0,
|
| 6985 |
+
"learning_rate": 8.3355082497919e-08,
|
| 6986 |
+
"loss": 26.50646057128906,
|
| 6987 |
+
"step": 8950
|
| 6988 |
+
},
|
| 6989 |
+
{
|
| 6990 |
+
"epoch": 0.8145454545454546,
|
| 6991 |
+
"grad_norm": 122593944272896.0,
|
| 6992 |
+
"learning_rate": 8.256733430448637e-08,
|
| 6993 |
+
"loss": 26.44368896484375,
|
| 6994 |
+
"step": 8960
|
| 6995 |
+
},
|
| 6996 |
+
{
|
| 6997 |
+
"epoch": 0.8154545454545454,
|
| 6998 |
+
"grad_norm": 182341729255424.0,
|
| 6999 |
+
"learning_rate": 8.178299098086678e-08,
|
| 7000 |
+
"loss": 26.42027893066406,
|
| 7001 |
+
"step": 8970
|
| 7002 |
+
},
|
| 7003 |
+
{
|
| 7004 |
+
"epoch": 0.8163636363636364,
|
| 7005 |
+
"grad_norm": 612390999687168.0,
|
| 7006 |
+
"learning_rate": 8.100205892470785e-08,
|
| 7007 |
+
"loss": 27.214886474609376,
|
| 7008 |
+
"step": 8980
|
| 7009 |
+
},
|
| 7010 |
+
{
|
| 7011 |
+
"epoch": 0.8172727272727273,
|
| 7012 |
+
"grad_norm": 336913676369920.0,
|
| 7013 |
+
"learning_rate": 8.022454450583183e-08,
|
| 7014 |
+
"loss": 26.410107421875,
|
| 7015 |
+
"step": 8990
|
| 7016 |
+
},
|
| 7017 |
+
{
|
| 7018 |
+
"epoch": 0.8181818181818182,
|
| 7019 |
+
"grad_norm": 358478572945408.0,
|
| 7020 |
+
"learning_rate": 7.945045406618462e-08,
|
| 7021 |
+
"loss": 26.294644165039063,
|
| 7022 |
+
"step": 9000
|
| 7023 |
+
},
|
| 7024 |
+
{
|
| 7025 |
+
"epoch": 0.8181818181818182,
|
| 7026 |
+
"eval_loss": 26.25908088684082,
|
| 7027 |
+
"eval_runtime": 2.195,
|
| 7028 |
+
"eval_samples_per_second": 16.401,
|
| 7029 |
+
"eval_steps_per_second": 0.911,
|
| 7030 |
+
"step": 9000
|
| 7031 |
}
|
| 7032 |
],
|
| 7033 |
"logging_steps": 10,
|
|
|
|
| 7047 |
"attributes": {}
|
| 7048 |
}
|
| 7049 |
},
|
| 7050 |
+
"total_flos": 8.1538628714496e+16,
|
| 7051 |
"train_batch_size": 4,
|
| 7052 |
"trial_name": null,
|
| 7053 |
"trial_params": null
|