Upload 8 files
Browse files- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +964 -4
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 18494040
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9fe830f215593241d5a06be7b7382d26df2df95fa5877818c55602cc0aaaec7c
|
| 3 |
size 18494040
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 37035002
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ffc9e9ba9737e7047e65caa20e5526ed8da4c213c4ce3f2cca848b1ac8ecdbd
|
| 3 |
size 37035002
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1ff264f99d31b522cc7e2a4eac9d38606d0c58a34c0adc74d71e0ca8b371dc36
|
| 3 |
size 14244
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9c4c29cdd3791b07f27eccf7f1e479362eae64a2df7b8ed21d32b1b0f2e78f0d
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -4351,11 +4351,971 @@
|
|
| 4351 |
"loss": 4.4091,
|
| 4352 |
"num_input_tokens_seen": 475882624,
|
| 4353 |
"step": 81450
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4354 |
}
|
| 4355 |
],
|
| 4356 |
"logging_steps": 150,
|
| 4357 |
"max_steps": 272232,
|
| 4358 |
-
"num_input_tokens_seen":
|
| 4359 |
"num_train_epochs": 3,
|
| 4360 |
"save_steps": 500,
|
| 4361 |
"stateful_callbacks": {
|
|
@@ -4370,7 +5330,7 @@
|
|
| 4370 |
"attributes": {}
|
| 4371 |
}
|
| 4372 |
},
|
| 4373 |
-
"total_flos":
|
| 4374 |
"train_batch_size": 32,
|
| 4375 |
"trial_name": null,
|
| 4376 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0964851864300316,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 99500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 4351 |
"loss": 4.4091,
|
| 4352 |
"num_input_tokens_seen": 475882624,
|
| 4353 |
"step": 81450
|
| 4354 |
+
},
|
| 4355 |
+
{
|
| 4356 |
+
"epoch": 0.8992280523888501,
|
| 4357 |
+
"grad_norm": 1.8798983097076416,
|
| 4358 |
+
"learning_rate": 0.00011205900935980135,
|
| 4359 |
+
"loss": 4.4113,
|
| 4360 |
+
"num_input_tokens_seen": 476749664,
|
| 4361 |
+
"step": 81600
|
| 4362 |
+
},
|
| 4363 |
+
{
|
| 4364 |
+
"epoch": 0.9008810451322119,
|
| 4365 |
+
"grad_norm": 1.8249197006225586,
|
| 4366 |
+
"learning_rate": 0.00011197084796567584,
|
| 4367 |
+
"loss": 4.4059,
|
| 4368 |
+
"num_input_tokens_seen": 477650784,
|
| 4369 |
+
"step": 81750
|
| 4370 |
+
},
|
| 4371 |
+
{
|
| 4372 |
+
"epoch": 0.9025340378755737,
|
| 4373 |
+
"grad_norm": 1.9157739877700806,
|
| 4374 |
+
"learning_rate": 0.00011188268657155033,
|
| 4375 |
+
"loss": 4.4109,
|
| 4376 |
+
"num_input_tokens_seen": 478523840,
|
| 4377 |
+
"step": 81900
|
| 4378 |
+
},
|
| 4379 |
+
{
|
| 4380 |
+
"epoch": 0.9041870306189356,
|
| 4381 |
+
"grad_norm": 1.9503858089447021,
|
| 4382 |
+
"learning_rate": 0.00011179452517742482,
|
| 4383 |
+
"loss": 4.4139,
|
| 4384 |
+
"num_input_tokens_seen": 479399296,
|
| 4385 |
+
"step": 82050
|
| 4386 |
+
},
|
| 4387 |
+
{
|
| 4388 |
+
"epoch": 0.9058400233622974,
|
| 4389 |
+
"grad_norm": 1.8298823833465576,
|
| 4390 |
+
"learning_rate": 0.0001117063637832993,
|
| 4391 |
+
"loss": 4.4123,
|
| 4392 |
+
"num_input_tokens_seen": 480262240,
|
| 4393 |
+
"step": 82200
|
| 4394 |
+
},
|
| 4395 |
+
{
|
| 4396 |
+
"epoch": 0.9074930161056592,
|
| 4397 |
+
"grad_norm": 1.9161386489868164,
|
| 4398 |
+
"learning_rate": 0.0001116182023891738,
|
| 4399 |
+
"loss": 4.4058,
|
| 4400 |
+
"num_input_tokens_seen": 481141056,
|
| 4401 |
+
"step": 82350
|
| 4402 |
+
},
|
| 4403 |
+
{
|
| 4404 |
+
"epoch": 0.9091460088490212,
|
| 4405 |
+
"grad_norm": 1.872722864151001,
|
| 4406 |
+
"learning_rate": 0.00011153004099504828,
|
| 4407 |
+
"loss": 4.4279,
|
| 4408 |
+
"num_input_tokens_seen": 482014112,
|
| 4409 |
+
"step": 82500
|
| 4410 |
+
},
|
| 4411 |
+
{
|
| 4412 |
+
"epoch": 0.910799001592383,
|
| 4413 |
+
"grad_norm": 1.8831090927124023,
|
| 4414 |
+
"learning_rate": 0.00011144187960092277,
|
| 4415 |
+
"loss": 4.4121,
|
| 4416 |
+
"num_input_tokens_seen": 482898336,
|
| 4417 |
+
"step": 82650
|
| 4418 |
+
},
|
| 4419 |
+
{
|
| 4420 |
+
"epoch": 0.9124519943357449,
|
| 4421 |
+
"grad_norm": 1.8128923177719116,
|
| 4422 |
+
"learning_rate": 0.00011135371820679725,
|
| 4423 |
+
"loss": 4.4105,
|
| 4424 |
+
"num_input_tokens_seen": 483773760,
|
| 4425 |
+
"step": 82800
|
| 4426 |
+
},
|
| 4427 |
+
{
|
| 4428 |
+
"epoch": 0.9141049870791067,
|
| 4429 |
+
"grad_norm": 1.8982397317886353,
|
| 4430 |
+
"learning_rate": 0.00011126555681267175,
|
| 4431 |
+
"loss": 4.4234,
|
| 4432 |
+
"num_input_tokens_seen": 484671008,
|
| 4433 |
+
"step": 82950
|
| 4434 |
+
},
|
| 4435 |
+
{
|
| 4436 |
+
"epoch": 0.9157579798224685,
|
| 4437 |
+
"grad_norm": 1.8295831680297852,
|
| 4438 |
+
"learning_rate": 0.00011117739541854622,
|
| 4439 |
+
"loss": 4.4227,
|
| 4440 |
+
"num_input_tokens_seen": 485553984,
|
| 4441 |
+
"step": 83100
|
| 4442 |
+
},
|
| 4443 |
+
{
|
| 4444 |
+
"epoch": 0.9174109725658305,
|
| 4445 |
+
"grad_norm": 1.8975720405578613,
|
| 4446 |
+
"learning_rate": 0.00011108923402442071,
|
| 4447 |
+
"loss": 4.4176,
|
| 4448 |
+
"num_input_tokens_seen": 486416672,
|
| 4449 |
+
"step": 83250
|
| 4450 |
+
},
|
| 4451 |
+
{
|
| 4452 |
+
"epoch": 0.9190639653091923,
|
| 4453 |
+
"grad_norm": 1.8207321166992188,
|
| 4454 |
+
"learning_rate": 0.0001110010726302952,
|
| 4455 |
+
"loss": 4.424,
|
| 4456 |
+
"num_input_tokens_seen": 487286816,
|
| 4457 |
+
"step": 83400
|
| 4458 |
+
},
|
| 4459 |
+
{
|
| 4460 |
+
"epoch": 0.9207169580525542,
|
| 4461 |
+
"grad_norm": 1.9241523742675781,
|
| 4462 |
+
"learning_rate": 0.00011091291123616969,
|
| 4463 |
+
"loss": 4.4129,
|
| 4464 |
+
"num_input_tokens_seen": 488157088,
|
| 4465 |
+
"step": 83550
|
| 4466 |
+
},
|
| 4467 |
+
{
|
| 4468 |
+
"epoch": 0.922369950795916,
|
| 4469 |
+
"grad_norm": 1.8391443490982056,
|
| 4470 |
+
"learning_rate": 0.00011082474984204418,
|
| 4471 |
+
"loss": 4.4036,
|
| 4472 |
+
"num_input_tokens_seen": 489008320,
|
| 4473 |
+
"step": 83700
|
| 4474 |
+
},
|
| 4475 |
+
{
|
| 4476 |
+
"epoch": 0.9240229435392778,
|
| 4477 |
+
"grad_norm": 1.9244701862335205,
|
| 4478 |
+
"learning_rate": 0.00011073658844791865,
|
| 4479 |
+
"loss": 4.4215,
|
| 4480 |
+
"num_input_tokens_seen": 489887328,
|
| 4481 |
+
"step": 83850
|
| 4482 |
+
},
|
| 4483 |
+
{
|
| 4484 |
+
"epoch": 0.9256759362826397,
|
| 4485 |
+
"grad_norm": 1.8949611186981201,
|
| 4486 |
+
"learning_rate": 0.00011064842705379315,
|
| 4487 |
+
"loss": 4.4205,
|
| 4488 |
+
"num_input_tokens_seen": 490765504,
|
| 4489 |
+
"step": 84000
|
| 4490 |
+
},
|
| 4491 |
+
{
|
| 4492 |
+
"epoch": 0.9273289290260016,
|
| 4493 |
+
"grad_norm": 1.810594081878662,
|
| 4494 |
+
"learning_rate": 0.00011056026565966763,
|
| 4495 |
+
"loss": 4.4149,
|
| 4496 |
+
"num_input_tokens_seen": 491650144,
|
| 4497 |
+
"step": 84150
|
| 4498 |
+
},
|
| 4499 |
+
{
|
| 4500 |
+
"epoch": 0.9289819217693635,
|
| 4501 |
+
"grad_norm": 1.8556066751480103,
|
| 4502 |
+
"learning_rate": 0.00011047210426554213,
|
| 4503 |
+
"loss": 4.4102,
|
| 4504 |
+
"num_input_tokens_seen": 492539968,
|
| 4505 |
+
"step": 84300
|
| 4506 |
+
},
|
| 4507 |
+
{
|
| 4508 |
+
"epoch": 0.9306349145127253,
|
| 4509 |
+
"grad_norm": 1.8486409187316895,
|
| 4510 |
+
"learning_rate": 0.00011038394287141661,
|
| 4511 |
+
"loss": 4.4231,
|
| 4512 |
+
"num_input_tokens_seen": 493424352,
|
| 4513 |
+
"step": 84450
|
| 4514 |
+
},
|
| 4515 |
+
{
|
| 4516 |
+
"epoch": 0.9322879072560871,
|
| 4517 |
+
"grad_norm": 1.8193395137786865,
|
| 4518 |
+
"learning_rate": 0.00011029578147729111,
|
| 4519 |
+
"loss": 4.4195,
|
| 4520 |
+
"num_input_tokens_seen": 494301152,
|
| 4521 |
+
"step": 84600
|
| 4522 |
+
},
|
| 4523 |
+
{
|
| 4524 |
+
"epoch": 0.933940899999449,
|
| 4525 |
+
"grad_norm": 1.8344619274139404,
|
| 4526 |
+
"learning_rate": 0.00011020762008316558,
|
| 4527 |
+
"loss": 4.4075,
|
| 4528 |
+
"num_input_tokens_seen": 495177600,
|
| 4529 |
+
"step": 84750
|
| 4530 |
+
},
|
| 4531 |
+
{
|
| 4532 |
+
"epoch": 0.9355938927428109,
|
| 4533 |
+
"grad_norm": 1.781654953956604,
|
| 4534 |
+
"learning_rate": 0.00011012004643166758,
|
| 4535 |
+
"loss": 4.4075,
|
| 4536 |
+
"num_input_tokens_seen": 496047680,
|
| 4537 |
+
"step": 84900
|
| 4538 |
+
},
|
| 4539 |
+
{
|
| 4540 |
+
"epoch": 0.9372468854861727,
|
| 4541 |
+
"grad_norm": 1.935810923576355,
|
| 4542 |
+
"learning_rate": 0.00011003188503754206,
|
| 4543 |
+
"loss": 4.408,
|
| 4544 |
+
"num_input_tokens_seen": 496919488,
|
| 4545 |
+
"step": 85050
|
| 4546 |
+
},
|
| 4547 |
+
{
|
| 4548 |
+
"epoch": 0.9388998782295346,
|
| 4549 |
+
"grad_norm": 1.8130574226379395,
|
| 4550 |
+
"learning_rate": 0.00010994372364341655,
|
| 4551 |
+
"loss": 4.4152,
|
| 4552 |
+
"num_input_tokens_seen": 497821280,
|
| 4553 |
+
"step": 85200
|
| 4554 |
+
},
|
| 4555 |
+
{
|
| 4556 |
+
"epoch": 0.9405528709728964,
|
| 4557 |
+
"grad_norm": 1.9481176137924194,
|
| 4558 |
+
"learning_rate": 0.00010985556224929104,
|
| 4559 |
+
"loss": 4.4115,
|
| 4560 |
+
"num_input_tokens_seen": 498694560,
|
| 4561 |
+
"step": 85350
|
| 4562 |
+
},
|
| 4563 |
+
{
|
| 4564 |
+
"epoch": 0.9422058637162583,
|
| 4565 |
+
"grad_norm": 1.8938475847244263,
|
| 4566 |
+
"learning_rate": 0.00010976740085516553,
|
| 4567 |
+
"loss": 4.4077,
|
| 4568 |
+
"num_input_tokens_seen": 499570016,
|
| 4569 |
+
"step": 85500
|
| 4570 |
+
},
|
| 4571 |
+
{
|
| 4572 |
+
"epoch": 0.9438588564596201,
|
| 4573 |
+
"grad_norm": 1.8449296951293945,
|
| 4574 |
+
"learning_rate": 0.00010967923946104002,
|
| 4575 |
+
"loss": 4.4043,
|
| 4576 |
+
"num_input_tokens_seen": 500436288,
|
| 4577 |
+
"step": 85650
|
| 4578 |
+
},
|
| 4579 |
+
{
|
| 4580 |
+
"epoch": 0.945511849202982,
|
| 4581 |
+
"grad_norm": 1.775891661643982,
|
| 4582 |
+
"learning_rate": 0.00010959166580954201,
|
| 4583 |
+
"loss": 4.4112,
|
| 4584 |
+
"num_input_tokens_seen": 501298944,
|
| 4585 |
+
"step": 85800
|
| 4586 |
+
},
|
| 4587 |
+
{
|
| 4588 |
+
"epoch": 0.9471648419463439,
|
| 4589 |
+
"grad_norm": 1.827708125114441,
|
| 4590 |
+
"learning_rate": 0.00010950350441541651,
|
| 4591 |
+
"loss": 4.4088,
|
| 4592 |
+
"num_input_tokens_seen": 502206976,
|
| 4593 |
+
"step": 85950
|
| 4594 |
+
},
|
| 4595 |
+
{
|
| 4596 |
+
"epoch": 0.9488178346897057,
|
| 4597 |
+
"grad_norm": 1.8833259344100952,
|
| 4598 |
+
"learning_rate": 0.00010941534302129099,
|
| 4599 |
+
"loss": 4.4107,
|
| 4600 |
+
"num_input_tokens_seen": 503083488,
|
| 4601 |
+
"step": 86100
|
| 4602 |
+
},
|
| 4603 |
+
{
|
| 4604 |
+
"epoch": 0.9504708274330675,
|
| 4605 |
+
"grad_norm": 1.8116602897644043,
|
| 4606 |
+
"learning_rate": 0.00010932718162716549,
|
| 4607 |
+
"loss": 4.4109,
|
| 4608 |
+
"num_input_tokens_seen": 503978240,
|
| 4609 |
+
"step": 86250
|
| 4610 |
+
},
|
| 4611 |
+
{
|
| 4612 |
+
"epoch": 0.9521238201764294,
|
| 4613 |
+
"grad_norm": 1.8248368501663208,
|
| 4614 |
+
"learning_rate": 0.00010923902023303996,
|
| 4615 |
+
"loss": 4.4041,
|
| 4616 |
+
"num_input_tokens_seen": 504859744,
|
| 4617 |
+
"step": 86400
|
| 4618 |
+
},
|
| 4619 |
+
{
|
| 4620 |
+
"epoch": 0.9537768129197913,
|
| 4621 |
+
"grad_norm": 1.862371802330017,
|
| 4622 |
+
"learning_rate": 0.00010915085883891444,
|
| 4623 |
+
"loss": 4.4221,
|
| 4624 |
+
"num_input_tokens_seen": 505740576,
|
| 4625 |
+
"step": 86550
|
| 4626 |
+
},
|
| 4627 |
+
{
|
| 4628 |
+
"epoch": 0.9554298056631532,
|
| 4629 |
+
"grad_norm": 1.8358848094940186,
|
| 4630 |
+
"learning_rate": 0.00010906269744478894,
|
| 4631 |
+
"loss": 4.4051,
|
| 4632 |
+
"num_input_tokens_seen": 506615680,
|
| 4633 |
+
"step": 86700
|
| 4634 |
+
},
|
| 4635 |
+
{
|
| 4636 |
+
"epoch": 0.957082798406515,
|
| 4637 |
+
"grad_norm": 1.8686786890029907,
|
| 4638 |
+
"learning_rate": 0.00010897453605066342,
|
| 4639 |
+
"loss": 4.41,
|
| 4640 |
+
"num_input_tokens_seen": 507477824,
|
| 4641 |
+
"step": 86850
|
| 4642 |
+
},
|
| 4643 |
+
{
|
| 4644 |
+
"epoch": 0.9587357911498768,
|
| 4645 |
+
"grad_norm": 1.833525538444519,
|
| 4646 |
+
"learning_rate": 0.00010888637465653792,
|
| 4647 |
+
"loss": 4.4188,
|
| 4648 |
+
"num_input_tokens_seen": 508371904,
|
| 4649 |
+
"step": 87000
|
| 4650 |
+
},
|
| 4651 |
+
{
|
| 4652 |
+
"epoch": 0.9603887838932387,
|
| 4653 |
+
"grad_norm": 1.9611468315124512,
|
| 4654 |
+
"learning_rate": 0.00010879821326241239,
|
| 4655 |
+
"loss": 4.4034,
|
| 4656 |
+
"num_input_tokens_seen": 509250272,
|
| 4657 |
+
"step": 87150
|
| 4658 |
+
},
|
| 4659 |
+
{
|
| 4660 |
+
"epoch": 0.9620417766366006,
|
| 4661 |
+
"grad_norm": 1.6934946775436401,
|
| 4662 |
+
"learning_rate": 0.00010871005186828688,
|
| 4663 |
+
"loss": 4.4067,
|
| 4664 |
+
"num_input_tokens_seen": 510129568,
|
| 4665 |
+
"step": 87300
|
| 4666 |
+
},
|
| 4667 |
+
{
|
| 4668 |
+
"epoch": 0.9636947693799625,
|
| 4669 |
+
"grad_norm": 1.8400757312774658,
|
| 4670 |
+
"learning_rate": 0.00010862189047416137,
|
| 4671 |
+
"loss": 4.4,
|
| 4672 |
+
"num_input_tokens_seen": 510995328,
|
| 4673 |
+
"step": 87450
|
| 4674 |
+
},
|
| 4675 |
+
{
|
| 4676 |
+
"epoch": 0.9653477621233243,
|
| 4677 |
+
"grad_norm": 1.8682547807693481,
|
| 4678 |
+
"learning_rate": 0.00010853372908003586,
|
| 4679 |
+
"loss": 4.4026,
|
| 4680 |
+
"num_input_tokens_seen": 511880416,
|
| 4681 |
+
"step": 87600
|
| 4682 |
+
},
|
| 4683 |
+
{
|
| 4684 |
+
"epoch": 0.9670007548666861,
|
| 4685 |
+
"grad_norm": 1.8408825397491455,
|
| 4686 |
+
"learning_rate": 0.00010844556768591035,
|
| 4687 |
+
"loss": 4.4019,
|
| 4688 |
+
"num_input_tokens_seen": 512759072,
|
| 4689 |
+
"step": 87750
|
| 4690 |
+
},
|
| 4691 |
+
{
|
| 4692 |
+
"epoch": 0.968653747610048,
|
| 4693 |
+
"grad_norm": 1.9082870483398438,
|
| 4694 |
+
"learning_rate": 0.00010835740629178484,
|
| 4695 |
+
"loss": 4.4076,
|
| 4696 |
+
"num_input_tokens_seen": 513640896,
|
| 4697 |
+
"step": 87900
|
| 4698 |
+
},
|
| 4699 |
+
{
|
| 4700 |
+
"epoch": 0.9703067403534098,
|
| 4701 |
+
"grad_norm": 1.9512287378311157,
|
| 4702 |
+
"learning_rate": 0.00010826983264028683,
|
| 4703 |
+
"loss": 4.4217,
|
| 4704 |
+
"num_input_tokens_seen": 514532256,
|
| 4705 |
+
"step": 88050
|
| 4706 |
+
},
|
| 4707 |
+
{
|
| 4708 |
+
"epoch": 0.9719597330967717,
|
| 4709 |
+
"grad_norm": 1.9278032779693604,
|
| 4710 |
+
"learning_rate": 0.0001081816712461613,
|
| 4711 |
+
"loss": 4.4217,
|
| 4712 |
+
"num_input_tokens_seen": 515412864,
|
| 4713 |
+
"step": 88200
|
| 4714 |
+
},
|
| 4715 |
+
{
|
| 4716 |
+
"epoch": 0.9736127258401336,
|
| 4717 |
+
"grad_norm": 1.814817190170288,
|
| 4718 |
+
"learning_rate": 0.0001080935098520358,
|
| 4719 |
+
"loss": 4.4178,
|
| 4720 |
+
"num_input_tokens_seen": 516301408,
|
| 4721 |
+
"step": 88350
|
| 4722 |
+
},
|
| 4723 |
+
{
|
| 4724 |
+
"epoch": 0.9752657185834954,
|
| 4725 |
+
"grad_norm": 2.0495548248291016,
|
| 4726 |
+
"learning_rate": 0.00010800534845791028,
|
| 4727 |
+
"loss": 4.4101,
|
| 4728 |
+
"num_input_tokens_seen": 517179648,
|
| 4729 |
+
"step": 88500
|
| 4730 |
+
},
|
| 4731 |
+
{
|
| 4732 |
+
"epoch": 0.9769187113268573,
|
| 4733 |
+
"grad_norm": 1.8378854990005493,
|
| 4734 |
+
"learning_rate": 0.00010791718706378478,
|
| 4735 |
+
"loss": 4.4031,
|
| 4736 |
+
"num_input_tokens_seen": 518064288,
|
| 4737 |
+
"step": 88650
|
| 4738 |
+
},
|
| 4739 |
+
{
|
| 4740 |
+
"epoch": 0.9785717040702191,
|
| 4741 |
+
"grad_norm": 1.8407827615737915,
|
| 4742 |
+
"learning_rate": 0.00010782902566965926,
|
| 4743 |
+
"loss": 4.4135,
|
| 4744 |
+
"num_input_tokens_seen": 518947776,
|
| 4745 |
+
"step": 88800
|
| 4746 |
+
},
|
| 4747 |
+
{
|
| 4748 |
+
"epoch": 0.980224696813581,
|
| 4749 |
+
"grad_norm": 1.845199704170227,
|
| 4750 |
+
"learning_rate": 0.00010774086427553376,
|
| 4751 |
+
"loss": 4.4,
|
| 4752 |
+
"num_input_tokens_seen": 519822560,
|
| 4753 |
+
"step": 88950
|
| 4754 |
+
},
|
| 4755 |
+
{
|
| 4756 |
+
"epoch": 0.9818776895569429,
|
| 4757 |
+
"grad_norm": 1.8627071380615234,
|
| 4758 |
+
"learning_rate": 0.00010765270288140823,
|
| 4759 |
+
"loss": 4.4125,
|
| 4760 |
+
"num_input_tokens_seen": 520709888,
|
| 4761 |
+
"step": 89100
|
| 4762 |
+
},
|
| 4763 |
+
{
|
| 4764 |
+
"epoch": 0.9835306823003047,
|
| 4765 |
+
"grad_norm": 1.826648235321045,
|
| 4766 |
+
"learning_rate": 0.00010756454148728274,
|
| 4767 |
+
"loss": 4.4006,
|
| 4768 |
+
"num_input_tokens_seen": 521586080,
|
| 4769 |
+
"step": 89250
|
| 4770 |
+
},
|
| 4771 |
+
{
|
| 4772 |
+
"epoch": 0.9851836750436666,
|
| 4773 |
+
"grad_norm": 1.9315438270568848,
|
| 4774 |
+
"learning_rate": 0.00010747638009315721,
|
| 4775 |
+
"loss": 4.4104,
|
| 4776 |
+
"num_input_tokens_seen": 522450944,
|
| 4777 |
+
"step": 89400
|
| 4778 |
+
},
|
| 4779 |
+
{
|
| 4780 |
+
"epoch": 0.9868366677870284,
|
| 4781 |
+
"grad_norm": 1.8507201671600342,
|
| 4782 |
+
"learning_rate": 0.0001073882186990317,
|
| 4783 |
+
"loss": 4.412,
|
| 4784 |
+
"num_input_tokens_seen": 523335744,
|
| 4785 |
+
"step": 89550
|
| 4786 |
+
},
|
| 4787 |
+
{
|
| 4788 |
+
"epoch": 0.9884896605303902,
|
| 4789 |
+
"grad_norm": 1.8950568437576294,
|
| 4790 |
+
"learning_rate": 0.00010730005730490619,
|
| 4791 |
+
"loss": 4.4106,
|
| 4792 |
+
"num_input_tokens_seen": 524216960,
|
| 4793 |
+
"step": 89700
|
| 4794 |
+
},
|
| 4795 |
+
{
|
| 4796 |
+
"epoch": 0.9901426532737522,
|
| 4797 |
+
"grad_norm": 1.92140531539917,
|
| 4798 |
+
"learning_rate": 0.00010721189591078068,
|
| 4799 |
+
"loss": 4.4001,
|
| 4800 |
+
"num_input_tokens_seen": 525081088,
|
| 4801 |
+
"step": 89850
|
| 4802 |
+
},
|
| 4803 |
+
{
|
| 4804 |
+
"epoch": 0.991795646017114,
|
| 4805 |
+
"grad_norm": 1.9179210662841797,
|
| 4806 |
+
"learning_rate": 0.00010712373451665516,
|
| 4807 |
+
"loss": 4.4044,
|
| 4808 |
+
"num_input_tokens_seen": 525964384,
|
| 4809 |
+
"step": 90000
|
| 4810 |
+
},
|
| 4811 |
+
{
|
| 4812 |
+
"epoch": 0.9934486387604758,
|
| 4813 |
+
"grad_norm": 1.8168158531188965,
|
| 4814 |
+
"learning_rate": 0.00010703557312252965,
|
| 4815 |
+
"loss": 4.4082,
|
| 4816 |
+
"num_input_tokens_seen": 526852256,
|
| 4817 |
+
"step": 90150
|
| 4818 |
+
},
|
| 4819 |
+
{
|
| 4820 |
+
"epoch": 0.9951016315038377,
|
| 4821 |
+
"grad_norm": 2.0058753490448,
|
| 4822 |
+
"learning_rate": 0.00010694741172840414,
|
| 4823 |
+
"loss": 4.4061,
|
| 4824 |
+
"num_input_tokens_seen": 527721152,
|
| 4825 |
+
"step": 90300
|
| 4826 |
+
},
|
| 4827 |
+
{
|
| 4828 |
+
"epoch": 0.9967546242471995,
|
| 4829 |
+
"grad_norm": 2.0036473274230957,
|
| 4830 |
+
"learning_rate": 0.00010685925033427863,
|
| 4831 |
+
"loss": 4.4013,
|
| 4832 |
+
"num_input_tokens_seen": 528601472,
|
| 4833 |
+
"step": 90450
|
| 4834 |
+
},
|
| 4835 |
+
{
|
| 4836 |
+
"epoch": 0.9984076169905615,
|
| 4837 |
+
"grad_norm": 1.8912723064422607,
|
| 4838 |
+
"learning_rate": 0.0001067710889401531,
|
| 4839 |
+
"loss": 4.408,
|
| 4840 |
+
"num_input_tokens_seen": 529485344,
|
| 4841 |
+
"step": 90600
|
| 4842 |
+
},
|
| 4843 |
+
{
|
| 4844 |
+
"epoch": 1.0000606097339233,
|
| 4845 |
+
"grad_norm": 1.8539482355117798,
|
| 4846 |
+
"learning_rate": 0.00010668292754602761,
|
| 4847 |
+
"loss": 4.4072,
|
| 4848 |
+
"num_input_tokens_seen": 530388834,
|
| 4849 |
+
"step": 90750
|
| 4850 |
+
},
|
| 4851 |
+
{
|
| 4852 |
+
"epoch": 1.001713602477285,
|
| 4853 |
+
"grad_norm": 1.9648711681365967,
|
| 4854 |
+
"learning_rate": 0.00010659476615190208,
|
| 4855 |
+
"loss": 4.4023,
|
| 4856 |
+
"num_input_tokens_seen": 531261986,
|
| 4857 |
+
"step": 90900
|
| 4858 |
+
},
|
| 4859 |
+
{
|
| 4860 |
+
"epoch": 1.003366595220647,
|
| 4861 |
+
"grad_norm": 1.8683501482009888,
|
| 4862 |
+
"learning_rate": 0.00010650660475777658,
|
| 4863 |
+
"loss": 4.3953,
|
| 4864 |
+
"num_input_tokens_seen": 532138626,
|
| 4865 |
+
"step": 91050
|
| 4866 |
+
},
|
| 4867 |
+
{
|
| 4868 |
+
"epoch": 1.005019587964009,
|
| 4869 |
+
"grad_norm": 1.9149645566940308,
|
| 4870 |
+
"learning_rate": 0.00010641844336365106,
|
| 4871 |
+
"loss": 4.3907,
|
| 4872 |
+
"num_input_tokens_seen": 533008418,
|
| 4873 |
+
"step": 91200
|
| 4874 |
+
},
|
| 4875 |
+
{
|
| 4876 |
+
"epoch": 1.0066725807073706,
|
| 4877 |
+
"grad_norm": 1.804408073425293,
|
| 4878 |
+
"learning_rate": 0.00010633028196952556,
|
| 4879 |
+
"loss": 4.4053,
|
| 4880 |
+
"num_input_tokens_seen": 533881122,
|
| 4881 |
+
"step": 91350
|
| 4882 |
+
},
|
| 4883 |
+
{
|
| 4884 |
+
"epoch": 1.0083255734507326,
|
| 4885 |
+
"grad_norm": 1.8145511150360107,
|
| 4886 |
+
"learning_rate": 0.00010624212057540004,
|
| 4887 |
+
"loss": 4.3993,
|
| 4888 |
+
"num_input_tokens_seen": 534752738,
|
| 4889 |
+
"step": 91500
|
| 4890 |
+
},
|
| 4891 |
+
{
|
| 4892 |
+
"epoch": 1.0099785661940943,
|
| 4893 |
+
"grad_norm": 1.8206557035446167,
|
| 4894 |
+
"learning_rate": 0.00010615395918127454,
|
| 4895 |
+
"loss": 4.3954,
|
| 4896 |
+
"num_input_tokens_seen": 535622338,
|
| 4897 |
+
"step": 91650
|
| 4898 |
+
},
|
| 4899 |
+
{
|
| 4900 |
+
"epoch": 1.0116315589374563,
|
| 4901 |
+
"grad_norm": 1.880231261253357,
|
| 4902 |
+
"learning_rate": 0.00010606579778714901,
|
| 4903 |
+
"loss": 4.4002,
|
| 4904 |
+
"num_input_tokens_seen": 536488386,
|
| 4905 |
+
"step": 91800
|
| 4906 |
+
},
|
| 4907 |
+
{
|
| 4908 |
+
"epoch": 1.0132845516808182,
|
| 4909 |
+
"grad_norm": 1.8914505243301392,
|
| 4910 |
+
"learning_rate": 0.00010597763639302352,
|
| 4911 |
+
"loss": 4.3961,
|
| 4912 |
+
"num_input_tokens_seen": 537374658,
|
| 4913 |
+
"step": 91950
|
| 4914 |
+
},
|
| 4915 |
+
{
|
| 4916 |
+
"epoch": 1.01493754442418,
|
| 4917 |
+
"grad_norm": 1.9163919687271118,
|
| 4918 |
+
"learning_rate": 0.0001058900627415255,
|
| 4919 |
+
"loss": 4.3973,
|
| 4920 |
+
"num_input_tokens_seen": 538244194,
|
| 4921 |
+
"step": 92100
|
| 4922 |
+
},
|
| 4923 |
+
{
|
| 4924 |
+
"epoch": 1.0165905371675419,
|
| 4925 |
+
"grad_norm": 1.9003725051879883,
|
| 4926 |
+
"learning_rate": 0.00010580190134739998,
|
| 4927 |
+
"loss": 4.3957,
|
| 4928 |
+
"num_input_tokens_seen": 539130914,
|
| 4929 |
+
"step": 92250
|
| 4930 |
+
},
|
| 4931 |
+
{
|
| 4932 |
+
"epoch": 1.0182435299109036,
|
| 4933 |
+
"grad_norm": 1.838493824005127,
|
| 4934 |
+
"learning_rate": 0.00010571373995327447,
|
| 4935 |
+
"loss": 4.3878,
|
| 4936 |
+
"num_input_tokens_seen": 540022850,
|
| 4937 |
+
"step": 92400
|
| 4938 |
+
},
|
| 4939 |
+
{
|
| 4940 |
+
"epoch": 1.0198965226542656,
|
| 4941 |
+
"grad_norm": 1.9080275297164917,
|
| 4942 |
+
"learning_rate": 0.00010562557855914896,
|
| 4943 |
+
"loss": 4.3917,
|
| 4944 |
+
"num_input_tokens_seen": 540895874,
|
| 4945 |
+
"step": 92550
|
| 4946 |
+
},
|
| 4947 |
+
{
|
| 4948 |
+
"epoch": 1.0215495153976275,
|
| 4949 |
+
"grad_norm": 1.8060060739517212,
|
| 4950 |
+
"learning_rate": 0.00010553741716502345,
|
| 4951 |
+
"loss": 4.3953,
|
| 4952 |
+
"num_input_tokens_seen": 541762658,
|
| 4953 |
+
"step": 92700
|
| 4954 |
+
},
|
| 4955 |
+
{
|
| 4956 |
+
"epoch": 1.0232025081409892,
|
| 4957 |
+
"grad_norm": 1.903151273727417,
|
| 4958 |
+
"learning_rate": 0.00010544925577089792,
|
| 4959 |
+
"loss": 4.3952,
|
| 4960 |
+
"num_input_tokens_seen": 542643138,
|
| 4961 |
+
"step": 92850
|
| 4962 |
+
},
|
| 4963 |
+
{
|
| 4964 |
+
"epoch": 1.0248555008843512,
|
| 4965 |
+
"grad_norm": 1.9957008361816406,
|
| 4966 |
+
"learning_rate": 0.00010536109437677243,
|
| 4967 |
+
"loss": 4.3952,
|
| 4968 |
+
"num_input_tokens_seen": 543505570,
|
| 4969 |
+
"step": 93000
|
| 4970 |
+
},
|
| 4971 |
+
{
|
| 4972 |
+
"epoch": 1.026508493627713,
|
| 4973 |
+
"grad_norm": 1.8897976875305176,
|
| 4974 |
+
"learning_rate": 0.0001052729329826469,
|
| 4975 |
+
"loss": 4.395,
|
| 4976 |
+
"num_input_tokens_seen": 544378466,
|
| 4977 |
+
"step": 93150
|
| 4978 |
+
},
|
| 4979 |
+
{
|
| 4980 |
+
"epoch": 1.0281614863710749,
|
| 4981 |
+
"grad_norm": 1.895654320716858,
|
| 4982 |
+
"learning_rate": 0.0001051847715885214,
|
| 4983 |
+
"loss": 4.4016,
|
| 4984 |
+
"num_input_tokens_seen": 545256738,
|
| 4985 |
+
"step": 93300
|
| 4986 |
+
},
|
| 4987 |
+
{
|
| 4988 |
+
"epoch": 1.0298144791144366,
|
| 4989 |
+
"grad_norm": 1.9977262020111084,
|
| 4990 |
+
"learning_rate": 0.00010509661019439588,
|
| 4991 |
+
"loss": 4.3994,
|
| 4992 |
+
"num_input_tokens_seen": 546150498,
|
| 4993 |
+
"step": 93450
|
| 4994 |
+
},
|
| 4995 |
+
{
|
| 4996 |
+
"epoch": 1.0314674718577985,
|
| 4997 |
+
"grad_norm": 1.82341468334198,
|
| 4998 |
+
"learning_rate": 0.00010500844880027038,
|
| 4999 |
+
"loss": 4.4003,
|
| 5000 |
+
"num_input_tokens_seen": 547021922,
|
| 5001 |
+
"step": 93600
|
| 5002 |
+
},
|
| 5003 |
+
{
|
| 5004 |
+
"epoch": 1.0331204646011605,
|
| 5005 |
+
"grad_norm": 1.7573907375335693,
|
| 5006 |
+
"learning_rate": 0.00010492028740614485,
|
| 5007 |
+
"loss": 4.3936,
|
| 5008 |
+
"num_input_tokens_seen": 547888450,
|
| 5009 |
+
"step": 93750
|
| 5010 |
+
},
|
| 5011 |
+
{
|
| 5012 |
+
"epoch": 1.0347734573445222,
|
| 5013 |
+
"grad_norm": 2.011516571044922,
|
| 5014 |
+
"learning_rate": 0.00010483212601201936,
|
| 5015 |
+
"loss": 4.3861,
|
| 5016 |
+
"num_input_tokens_seen": 548752514,
|
| 5017 |
+
"step": 93900
|
| 5018 |
+
},
|
| 5019 |
+
{
|
| 5020 |
+
"epoch": 1.0364264500878841,
|
| 5021 |
+
"grad_norm": 1.8368171453475952,
|
| 5022 |
+
"learning_rate": 0.00010474396461789383,
|
| 5023 |
+
"loss": 4.3975,
|
| 5024 |
+
"num_input_tokens_seen": 549641218,
|
| 5025 |
+
"step": 94050
|
| 5026 |
+
},
|
| 5027 |
+
{
|
| 5028 |
+
"epoch": 1.0380794428312459,
|
| 5029 |
+
"grad_norm": 2.0658929347991943,
|
| 5030 |
+
"learning_rate": 0.00010465639096639582,
|
| 5031 |
+
"loss": 4.4042,
|
| 5032 |
+
"num_input_tokens_seen": 550521378,
|
| 5033 |
+
"step": 94200
|
| 5034 |
+
},
|
| 5035 |
+
{
|
| 5036 |
+
"epoch": 1.0397324355746078,
|
| 5037 |
+
"grad_norm": 1.8516744375228882,
|
| 5038 |
+
"learning_rate": 0.00010456822957227031,
|
| 5039 |
+
"loss": 4.3937,
|
| 5040 |
+
"num_input_tokens_seen": 551403138,
|
| 5041 |
+
"step": 94350
|
| 5042 |
+
},
|
| 5043 |
+
{
|
| 5044 |
+
"epoch": 1.0413854283179698,
|
| 5045 |
+
"grad_norm": 1.9704523086547852,
|
| 5046 |
+
"learning_rate": 0.0001044800681781448,
|
| 5047 |
+
"loss": 4.3892,
|
| 5048 |
+
"num_input_tokens_seen": 552268866,
|
| 5049 |
+
"step": 94500
|
| 5050 |
+
},
|
| 5051 |
+
{
|
| 5052 |
+
"epoch": 1.0430384210613315,
|
| 5053 |
+
"grad_norm": 1.8856583833694458,
|
| 5054 |
+
"learning_rate": 0.00010439190678401929,
|
| 5055 |
+
"loss": 4.3969,
|
| 5056 |
+
"num_input_tokens_seen": 553139522,
|
| 5057 |
+
"step": 94650
|
| 5058 |
+
},
|
| 5059 |
+
{
|
| 5060 |
+
"epoch": 1.0446914138046934,
|
| 5061 |
+
"grad_norm": 1.9823240041732788,
|
| 5062 |
+
"learning_rate": 0.00010430374538989378,
|
| 5063 |
+
"loss": 4.3937,
|
| 5064 |
+
"num_input_tokens_seen": 554009858,
|
| 5065 |
+
"step": 94800
|
| 5066 |
+
},
|
| 5067 |
+
{
|
| 5068 |
+
"epoch": 1.0463444065480552,
|
| 5069 |
+
"grad_norm": 1.8391404151916504,
|
| 5070 |
+
"learning_rate": 0.00010421558399576827,
|
| 5071 |
+
"loss": 4.3891,
|
| 5072 |
+
"num_input_tokens_seen": 554896962,
|
| 5073 |
+
"step": 94950
|
| 5074 |
+
},
|
| 5075 |
+
{
|
| 5076 |
+
"epoch": 1.047997399291417,
|
| 5077 |
+
"grad_norm": 1.829777717590332,
|
| 5078 |
+
"learning_rate": 0.00010412742260164274,
|
| 5079 |
+
"loss": 4.3996,
|
| 5080 |
+
"num_input_tokens_seen": 555778274,
|
| 5081 |
+
"step": 95100
|
| 5082 |
+
},
|
| 5083 |
+
{
|
| 5084 |
+
"epoch": 1.049650392034779,
|
| 5085 |
+
"grad_norm": 1.884190320968628,
|
| 5086 |
+
"learning_rate": 0.00010403926120751724,
|
| 5087 |
+
"loss": 4.3899,
|
| 5088 |
+
"num_input_tokens_seen": 556658210,
|
| 5089 |
+
"step": 95250
|
| 5090 |
+
},
|
| 5091 |
+
{
|
| 5092 |
+
"epoch": 1.0513033847781408,
|
| 5093 |
+
"grad_norm": 1.8368123769760132,
|
| 5094 |
+
"learning_rate": 0.00010395109981339172,
|
| 5095 |
+
"loss": 4.3989,
|
| 5096 |
+
"num_input_tokens_seen": 557549442,
|
| 5097 |
+
"step": 95400
|
| 5098 |
+
},
|
| 5099 |
+
{
|
| 5100 |
+
"epoch": 1.0529563775215027,
|
| 5101 |
+
"grad_norm": 1.7985849380493164,
|
| 5102 |
+
"learning_rate": 0.00010386293841926622,
|
| 5103 |
+
"loss": 4.3868,
|
| 5104 |
+
"num_input_tokens_seen": 558417634,
|
| 5105 |
+
"step": 95550
|
| 5106 |
+
},
|
| 5107 |
+
{
|
| 5108 |
+
"epoch": 1.0546093702648645,
|
| 5109 |
+
"grad_norm": 1.8913172483444214,
|
| 5110 |
+
"learning_rate": 0.0001037747770251407,
|
| 5111 |
+
"loss": 4.4031,
|
| 5112 |
+
"num_input_tokens_seen": 559314882,
|
| 5113 |
+
"step": 95700
|
| 5114 |
+
},
|
| 5115 |
+
{
|
| 5116 |
+
"epoch": 1.0562623630082264,
|
| 5117 |
+
"grad_norm": 1.9179192781448364,
|
| 5118 |
+
"learning_rate": 0.0001036866156310152,
|
| 5119 |
+
"loss": 4.3812,
|
| 5120 |
+
"num_input_tokens_seen": 560207298,
|
| 5121 |
+
"step": 95850
|
| 5122 |
+
},
|
| 5123 |
+
{
|
| 5124 |
+
"epoch": 1.0579153557515883,
|
| 5125 |
+
"grad_norm": 1.8890949487686157,
|
| 5126 |
+
"learning_rate": 0.00010359845423688967,
|
| 5127 |
+
"loss": 4.3916,
|
| 5128 |
+
"num_input_tokens_seen": 561097570,
|
| 5129 |
+
"step": 96000
|
| 5130 |
+
},
|
| 5131 |
+
{
|
| 5132 |
+
"epoch": 1.05956834849495,
|
| 5133 |
+
"grad_norm": 1.7995752096176147,
|
| 5134 |
+
"learning_rate": 0.00010351088058539166,
|
| 5135 |
+
"loss": 4.3933,
|
| 5136 |
+
"num_input_tokens_seen": 561973218,
|
| 5137 |
+
"step": 96150
|
| 5138 |
+
},
|
| 5139 |
+
{
|
| 5140 |
+
"epoch": 1.061221341238312,
|
| 5141 |
+
"grad_norm": 1.928031086921692,
|
| 5142 |
+
"learning_rate": 0.00010342330693389365,
|
| 5143 |
+
"loss": 4.3914,
|
| 5144 |
+
"num_input_tokens_seen": 562851074,
|
| 5145 |
+
"step": 96300
|
| 5146 |
+
},
|
| 5147 |
+
{
|
| 5148 |
+
"epoch": 1.0628743339816737,
|
| 5149 |
+
"grad_norm": 1.94650137424469,
|
| 5150 |
+
"learning_rate": 0.00010333514553976814,
|
| 5151 |
+
"loss": 4.3766,
|
| 5152 |
+
"num_input_tokens_seen": 563725506,
|
| 5153 |
+
"step": 96450
|
| 5154 |
+
},
|
| 5155 |
+
{
|
| 5156 |
+
"epoch": 1.0645273267250357,
|
| 5157 |
+
"grad_norm": 1.8535209894180298,
|
| 5158 |
+
"learning_rate": 0.00010324698414564263,
|
| 5159 |
+
"loss": 4.3916,
|
| 5160 |
+
"num_input_tokens_seen": 564576034,
|
| 5161 |
+
"step": 96600
|
| 5162 |
+
},
|
| 5163 |
+
{
|
| 5164 |
+
"epoch": 1.0661803194683976,
|
| 5165 |
+
"grad_norm": 1.9456048011779785,
|
| 5166 |
+
"learning_rate": 0.00010315882275151712,
|
| 5167 |
+
"loss": 4.3975,
|
| 5168 |
+
"num_input_tokens_seen": 565446626,
|
| 5169 |
+
"step": 96750
|
| 5170 |
+
},
|
| 5171 |
+
{
|
| 5172 |
+
"epoch": 1.0678333122117594,
|
| 5173 |
+
"grad_norm": 1.8319114446640015,
|
| 5174 |
+
"learning_rate": 0.0001030706613573916,
|
| 5175 |
+
"loss": 4.3977,
|
| 5176 |
+
"num_input_tokens_seen": 566321378,
|
| 5177 |
+
"step": 96900
|
| 5178 |
+
},
|
| 5179 |
+
{
|
| 5180 |
+
"epoch": 1.0694863049551213,
|
| 5181 |
+
"grad_norm": 2.1267592906951904,
|
| 5182 |
+
"learning_rate": 0.0001029824999632661,
|
| 5183 |
+
"loss": 4.3896,
|
| 5184 |
+
"num_input_tokens_seen": 567203778,
|
| 5185 |
+
"step": 97050
|
| 5186 |
+
},
|
| 5187 |
+
{
|
| 5188 |
+
"epoch": 1.071139297698483,
|
| 5189 |
+
"grad_norm": 1.8523855209350586,
|
| 5190 |
+
"learning_rate": 0.00010289433856914057,
|
| 5191 |
+
"loss": 4.3868,
|
| 5192 |
+
"num_input_tokens_seen": 568083906,
|
| 5193 |
+
"step": 97200
|
| 5194 |
+
},
|
| 5195 |
+
{
|
| 5196 |
+
"epoch": 1.072792290441845,
|
| 5197 |
+
"grad_norm": 1.882645606994629,
|
| 5198 |
+
"learning_rate": 0.00010280617717501507,
|
| 5199 |
+
"loss": 4.3976,
|
| 5200 |
+
"num_input_tokens_seen": 568970690,
|
| 5201 |
+
"step": 97350
|
| 5202 |
+
},
|
| 5203 |
+
{
|
| 5204 |
+
"epoch": 1.0744452831852067,
|
| 5205 |
+
"grad_norm": 1.9347394704818726,
|
| 5206 |
+
"learning_rate": 0.00010271801578088955,
|
| 5207 |
+
"loss": 4.3905,
|
| 5208 |
+
"num_input_tokens_seen": 569844482,
|
| 5209 |
+
"step": 97500
|
| 5210 |
+
},
|
| 5211 |
+
{
|
| 5212 |
+
"epoch": 1.0760982759285687,
|
| 5213 |
+
"grad_norm": 1.855491280555725,
|
| 5214 |
+
"learning_rate": 0.00010262985438676405,
|
| 5215 |
+
"loss": 4.3886,
|
| 5216 |
+
"num_input_tokens_seen": 570717602,
|
| 5217 |
+
"step": 97650
|
| 5218 |
+
},
|
| 5219 |
+
{
|
| 5220 |
+
"epoch": 1.0777512686719306,
|
| 5221 |
+
"grad_norm": 1.8031153678894043,
|
| 5222 |
+
"learning_rate": 0.00010254169299263853,
|
| 5223 |
+
"loss": 4.3807,
|
| 5224 |
+
"num_input_tokens_seen": 571591586,
|
| 5225 |
+
"step": 97800
|
| 5226 |
+
},
|
| 5227 |
+
{
|
| 5228 |
+
"epoch": 1.0794042614152923,
|
| 5229 |
+
"grad_norm": 1.9792248010635376,
|
| 5230 |
+
"learning_rate": 0.00010245353159851303,
|
| 5231 |
+
"loss": 4.389,
|
| 5232 |
+
"num_input_tokens_seen": 572476162,
|
| 5233 |
+
"step": 97950
|
| 5234 |
+
},
|
| 5235 |
+
{
|
| 5236 |
+
"epoch": 1.0810572541586543,
|
| 5237 |
+
"grad_norm": 1.9110472202301025,
|
| 5238 |
+
"learning_rate": 0.0001023653702043875,
|
| 5239 |
+
"loss": 4.3889,
|
| 5240 |
+
"num_input_tokens_seen": 573353346,
|
| 5241 |
+
"step": 98100
|
| 5242 |
+
},
|
| 5243 |
+
{
|
| 5244 |
+
"epoch": 1.082710246902016,
|
| 5245 |
+
"grad_norm": 1.9655945301055908,
|
| 5246 |
+
"learning_rate": 0.000102277208810262,
|
| 5247 |
+
"loss": 4.3808,
|
| 5248 |
+
"num_input_tokens_seen": 574237986,
|
| 5249 |
+
"step": 98250
|
| 5250 |
+
},
|
| 5251 |
+
{
|
| 5252 |
+
"epoch": 1.084363239645378,
|
| 5253 |
+
"grad_norm": 1.806372880935669,
|
| 5254 |
+
"learning_rate": 0.00010218904741613648,
|
| 5255 |
+
"loss": 4.389,
|
| 5256 |
+
"num_input_tokens_seen": 575113346,
|
| 5257 |
+
"step": 98400
|
| 5258 |
+
},
|
| 5259 |
+
{
|
| 5260 |
+
"epoch": 1.08601623238874,
|
| 5261 |
+
"grad_norm": 1.9266657829284668,
|
| 5262 |
+
"learning_rate": 0.00010210147376463847,
|
| 5263 |
+
"loss": 4.3957,
|
| 5264 |
+
"num_input_tokens_seen": 576010818,
|
| 5265 |
+
"step": 98550
|
| 5266 |
+
},
|
| 5267 |
+
{
|
| 5268 |
+
"epoch": 1.0876692251321016,
|
| 5269 |
+
"grad_norm": 1.8409209251403809,
|
| 5270 |
+
"learning_rate": 0.00010201331237051296,
|
| 5271 |
+
"loss": 4.3949,
|
| 5272 |
+
"num_input_tokens_seen": 576872610,
|
| 5273 |
+
"step": 98700
|
| 5274 |
+
},
|
| 5275 |
+
{
|
| 5276 |
+
"epoch": 1.0893222178754636,
|
| 5277 |
+
"grad_norm": 1.7804383039474487,
|
| 5278 |
+
"learning_rate": 0.00010192515097638745,
|
| 5279 |
+
"loss": 4.3835,
|
| 5280 |
+
"num_input_tokens_seen": 577737090,
|
| 5281 |
+
"step": 98850
|
| 5282 |
+
},
|
| 5283 |
+
{
|
| 5284 |
+
"epoch": 1.0909752106188253,
|
| 5285 |
+
"grad_norm": 1.8269861936569214,
|
| 5286 |
+
"learning_rate": 0.00010183698958226194,
|
| 5287 |
+
"loss": 4.3967,
|
| 5288 |
+
"num_input_tokens_seen": 578610178,
|
| 5289 |
+
"step": 99000
|
| 5290 |
+
},
|
| 5291 |
+
{
|
| 5292 |
+
"epoch": 1.0926282033621872,
|
| 5293 |
+
"grad_norm": 1.9065062999725342,
|
| 5294 |
+
"learning_rate": 0.00010174882818813641,
|
| 5295 |
+
"loss": 4.395,
|
| 5296 |
+
"num_input_tokens_seen": 579483074,
|
| 5297 |
+
"step": 99150
|
| 5298 |
+
},
|
| 5299 |
+
{
|
| 5300 |
+
"epoch": 1.0942811961055492,
|
| 5301 |
+
"grad_norm": 1.8511546850204468,
|
| 5302 |
+
"learning_rate": 0.00010166066679401092,
|
| 5303 |
+
"loss": 4.3847,
|
| 5304 |
+
"num_input_tokens_seen": 580339138,
|
| 5305 |
+
"step": 99300
|
| 5306 |
+
},
|
| 5307 |
+
{
|
| 5308 |
+
"epoch": 1.095934188848911,
|
| 5309 |
+
"grad_norm": 1.9003854990005493,
|
| 5310 |
+
"learning_rate": 0.00010157250539988539,
|
| 5311 |
+
"loss": 4.3878,
|
| 5312 |
+
"num_input_tokens_seen": 581214146,
|
| 5313 |
+
"step": 99450
|
| 5314 |
}
|
| 5315 |
],
|
| 5316 |
"logging_steps": 150,
|
| 5317 |
"max_steps": 272232,
|
| 5318 |
+
"num_input_tokens_seen": 581503010,
|
| 5319 |
"num_train_epochs": 3,
|
| 5320 |
"save_steps": 500,
|
| 5321 |
"stateful_callbacks": {
|
|
|
|
| 5330 |
"attributes": {}
|
| 5331 |
}
|
| 5332 |
},
|
| 5333 |
+
"total_flos": 8978778636326400.0,
|
| 5334 |
"train_batch_size": 32,
|
| 5335 |
"trial_name": null,
|
| 5336 |
"trial_params": null
|