Training in progress, step 600, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 250422888
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:214e6c8896d85eafd81d130569d64f61acdea940387997e5b23cca2b87ef8b10
|
| 3 |
size 250422888
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 126034975
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ff03043d6dabbea09af926cdcb31c7aaa770550057581349afa1f9724c33b2d5
|
| 3 |
size 126034975
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5d6f4fc1c930b54bcb56fd2abfe1a33e2eb76c1cc7f94aa49f35f3d4ceb6ca2
|
| 3 |
size 14575
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9c1a2ac1c11599601eeac95feb1dbfd49ec5c625e61dcce18b3f094491f9cf2d
|
| 3 |
size 627
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 200,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -33,13 +33,26 @@
|
|
| 33 |
"eval_samples_per_second": 4.266,
|
| 34 |
"eval_steps_per_second": 0.171,
|
| 35 |
"step": 400
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 36 |
}
|
| 37 |
],
|
| 38 |
"logging_steps": 200,
|
| 39 |
"max_steps": 960,
|
| 40 |
"num_train_epochs": 16,
|
| 41 |
"save_steps": 200,
|
| 42 |
-
"total_flos":
|
| 43 |
"trial_name": null,
|
| 44 |
"trial_params": null
|
| 45 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
"eval_steps": 200,
|
| 6 |
+
"global_step": 600,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 33 |
"eval_samples_per_second": 4.266,
|
| 34 |
"eval_steps_per_second": 0.171,
|
| 35 |
"step": 400
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 10.0,
|
| 39 |
+
"learning_rate": 0.0002,
|
| 40 |
+
"loss": 0.0211,
|
| 41 |
+
"step": 600
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 10.0,
|
| 45 |
+
"eval_runtime": 23.4523,
|
| 46 |
+
"eval_samples_per_second": 4.264,
|
| 47 |
+
"eval_steps_per_second": 0.171,
|
| 48 |
+
"step": 600
|
| 49 |
}
|
| 50 |
],
|
| 51 |
"logging_steps": 200,
|
| 52 |
"max_steps": 960,
|
| 53 |
"num_train_epochs": 16,
|
| 54 |
"save_steps": 200,
|
| 55 |
+
"total_flos": 3.835201585152e+17,
|
| 56 |
"trial_name": null,
|
| 57 |
"trial_params": null
|
| 58 |
}
|