Training in progress, step 560, 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 80013120
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86e4cb8ff828d940202480f4bf809c5a8c28e8bda594058004fd6fedcf340deb
|
| 3 |
size 80013120
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 41120084
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b58206039ff4a896ccc8cb3fdaa8d57b843c989eb4832ff2a606082e29c6812e
|
| 3 |
size 41120084
|
last-checkpoint/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:d2a7a94af35907f76baad16bf73c2fcb56ea77ed482ac94a96e55df62acb9e2d
|
| 3 |
size 14244
|
last-checkpoint/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 sha256:9fd920ca7c276a3eaed9fc64f5a8671269635687c93d3a6383990a00e0334d1c
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0
|
| 5 |
"eval_steps": 140,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -2971,6 +2971,994 @@
|
|
| 2971 |
"eval_samples_per_second": 17.637,
|
| 2972 |
"eval_steps_per_second": 8.819,
|
| 2973 |
"step": 420
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2974 |
}
|
| 2975 |
],
|
| 2976 |
"logging_steps": 1,
|
|
@@ -2985,12 +3973,12 @@
|
|
| 2985 |
"should_evaluate": false,
|
| 2986 |
"should_log": false,
|
| 2987 |
"should_save": true,
|
| 2988 |
-
"should_training_stop":
|
| 2989 |
},
|
| 2990 |
"attributes": {}
|
| 2991 |
}
|
| 2992 |
},
|
| 2993 |
-
"total_flos": 1.
|
| 2994 |
"train_batch_size": 2,
|
| 2995 |
"trial_name": null,
|
| 2996 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
"eval_steps": 140,
|
| 6 |
+
"global_step": 560,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 2971 |
"eval_samples_per_second": 17.637,
|
| 2972 |
"eval_steps_per_second": 8.819,
|
| 2973 |
"step": 420
|
| 2974 |
+
},
|
| 2975 |
+
{
|
| 2976 |
+
"epoch": 0.7517857142857143,
|
| 2977 |
+
"grad_norm": 0.2867870330810547,
|
| 2978 |
+
"learning_rate": 2.9897757571668905e-05,
|
| 2979 |
+
"loss": 1.5308,
|
| 2980 |
+
"step": 421
|
| 2981 |
+
},
|
| 2982 |
+
{
|
| 2983 |
+
"epoch": 0.7535714285714286,
|
| 2984 |
+
"grad_norm": 0.26840099692344666,
|
| 2985 |
+
"learning_rate": 2.949155921388943e-05,
|
| 2986 |
+
"loss": 1.8933,
|
| 2987 |
+
"step": 422
|
| 2988 |
+
},
|
| 2989 |
+
{
|
| 2990 |
+
"epoch": 0.7553571428571428,
|
| 2991 |
+
"grad_norm": 0.32162272930145264,
|
| 2992 |
+
"learning_rate": 2.9087661314045366e-05,
|
| 2993 |
+
"loss": 1.6497,
|
| 2994 |
+
"step": 423
|
| 2995 |
+
},
|
| 2996 |
+
{
|
| 2997 |
+
"epoch": 0.7571428571428571,
|
| 2998 |
+
"grad_norm": 0.2842392921447754,
|
| 2999 |
+
"learning_rate": 2.8686077049993287e-05,
|
| 3000 |
+
"loss": 1.521,
|
| 3001 |
+
"step": 424
|
| 3002 |
+
},
|
| 3003 |
+
{
|
| 3004 |
+
"epoch": 0.7589285714285714,
|
| 3005 |
+
"grad_norm": 0.28831931948661804,
|
| 3006 |
+
"learning_rate": 2.828681952410366e-05,
|
| 3007 |
+
"loss": 1.2619,
|
| 3008 |
+
"step": 425
|
| 3009 |
+
},
|
| 3010 |
+
{
|
| 3011 |
+
"epoch": 0.7607142857142857,
|
| 3012 |
+
"grad_norm": 0.2879536747932434,
|
| 3013 |
+
"learning_rate": 2.7889901762833083e-05,
|
| 3014 |
+
"loss": 1.1705,
|
| 3015 |
+
"step": 426
|
| 3016 |
+
},
|
| 3017 |
+
{
|
| 3018 |
+
"epoch": 0.7625,
|
| 3019 |
+
"grad_norm": 0.32446354627609253,
|
| 3020 |
+
"learning_rate": 2.7495336716299313e-05,
|
| 3021 |
+
"loss": 1.1636,
|
| 3022 |
+
"step": 427
|
| 3023 |
+
},
|
| 3024 |
+
{
|
| 3025 |
+
"epoch": 0.7642857142857142,
|
| 3026 |
+
"grad_norm": 0.2487679123878479,
|
| 3027 |
+
"learning_rate": 2.7103137257858868e-05,
|
| 3028 |
+
"loss": 0.7092,
|
| 3029 |
+
"step": 428
|
| 3030 |
+
},
|
| 3031 |
+
{
|
| 3032 |
+
"epoch": 0.7660714285714286,
|
| 3033 |
+
"grad_norm": 0.3402203321456909,
|
| 3034 |
+
"learning_rate": 2.671331618368682e-05,
|
| 3035 |
+
"loss": 0.8279,
|
| 3036 |
+
"step": 429
|
| 3037 |
+
},
|
| 3038 |
+
{
|
| 3039 |
+
"epoch": 0.7678571428571429,
|
| 3040 |
+
"grad_norm": 0.3117457330226898,
|
| 3041 |
+
"learning_rate": 2.6325886212359498e-05,
|
| 3042 |
+
"loss": 1.2101,
|
| 3043 |
+
"step": 430
|
| 3044 |
+
},
|
| 3045 |
+
{
|
| 3046 |
+
"epoch": 0.7696428571428572,
|
| 3047 |
+
"grad_norm": 0.28996577858924866,
|
| 3048 |
+
"learning_rate": 2.5940859984439424e-05,
|
| 3049 |
+
"loss": 1.1556,
|
| 3050 |
+
"step": 431
|
| 3051 |
+
},
|
| 3052 |
+
{
|
| 3053 |
+
"epoch": 0.7714285714285715,
|
| 3054 |
+
"grad_norm": 0.31623905897140503,
|
| 3055 |
+
"learning_rate": 2.5558250062062828e-05,
|
| 3056 |
+
"loss": 1.1324,
|
| 3057 |
+
"step": 432
|
| 3058 |
+
},
|
| 3059 |
+
{
|
| 3060 |
+
"epoch": 0.7732142857142857,
|
| 3061 |
+
"grad_norm": 0.29438769817352295,
|
| 3062 |
+
"learning_rate": 2.5178068928529864e-05,
|
| 3063 |
+
"loss": 1.4183,
|
| 3064 |
+
"step": 433
|
| 3065 |
+
},
|
| 3066 |
+
{
|
| 3067 |
+
"epoch": 0.775,
|
| 3068 |
+
"grad_norm": 0.31944945454597473,
|
| 3069 |
+
"learning_rate": 2.4800328987897427e-05,
|
| 3070 |
+
"loss": 1.1763,
|
| 3071 |
+
"step": 434
|
| 3072 |
+
},
|
| 3073 |
+
{
|
| 3074 |
+
"epoch": 0.7767857142857143,
|
| 3075 |
+
"grad_norm": 0.32895660400390625,
|
| 3076 |
+
"learning_rate": 2.4425042564574184e-05,
|
| 3077 |
+
"loss": 0.8288,
|
| 3078 |
+
"step": 435
|
| 3079 |
+
},
|
| 3080 |
+
{
|
| 3081 |
+
"epoch": 0.7785714285714286,
|
| 3082 |
+
"grad_norm": 0.46789219975471497,
|
| 3083 |
+
"learning_rate": 2.4052221902918725e-05,
|
| 3084 |
+
"loss": 0.889,
|
| 3085 |
+
"step": 436
|
| 3086 |
+
},
|
| 3087 |
+
{
|
| 3088 |
+
"epoch": 0.7803571428571429,
|
| 3089 |
+
"grad_norm": 0.35038378834724426,
|
| 3090 |
+
"learning_rate": 2.368187916683997e-05,
|
| 3091 |
+
"loss": 1.1607,
|
| 3092 |
+
"step": 437
|
| 3093 |
+
},
|
| 3094 |
+
{
|
| 3095 |
+
"epoch": 0.7821428571428571,
|
| 3096 |
+
"grad_norm": 0.3326191008090973,
|
| 3097 |
+
"learning_rate": 2.3314026439400217e-05,
|
| 3098 |
+
"loss": 1.3321,
|
| 3099 |
+
"step": 438
|
| 3100 |
+
},
|
| 3101 |
+
{
|
| 3102 |
+
"epoch": 0.7839285714285714,
|
| 3103 |
+
"grad_norm": 0.3683798015117645,
|
| 3104 |
+
"learning_rate": 2.2948675722421086e-05,
|
| 3105 |
+
"loss": 0.831,
|
| 3106 |
+
"step": 439
|
| 3107 |
+
},
|
| 3108 |
+
{
|
| 3109 |
+
"epoch": 0.7857142857142857,
|
| 3110 |
+
"grad_norm": 0.3587857186794281,
|
| 3111 |
+
"learning_rate": 2.2585838936091754e-05,
|
| 3112 |
+
"loss": 1.0373,
|
| 3113 |
+
"step": 440
|
| 3114 |
+
},
|
| 3115 |
+
{
|
| 3116 |
+
"epoch": 0.7875,
|
| 3117 |
+
"grad_norm": 0.341905802488327,
|
| 3118 |
+
"learning_rate": 2.2225527918580204e-05,
|
| 3119 |
+
"loss": 1.2294,
|
| 3120 |
+
"step": 441
|
| 3121 |
+
},
|
| 3122 |
+
{
|
| 3123 |
+
"epoch": 0.7892857142857143,
|
| 3124 |
+
"grad_norm": 0.36167776584625244,
|
| 3125 |
+
"learning_rate": 2.1867754425646926e-05,
|
| 3126 |
+
"loss": 1.06,
|
| 3127 |
+
"step": 442
|
| 3128 |
+
},
|
| 3129 |
+
{
|
| 3130 |
+
"epoch": 0.7910714285714285,
|
| 3131 |
+
"grad_norm": 0.35555702447891235,
|
| 3132 |
+
"learning_rate": 2.151253013026121e-05,
|
| 3133 |
+
"loss": 1.5072,
|
| 3134 |
+
"step": 443
|
| 3135 |
+
},
|
| 3136 |
+
{
|
| 3137 |
+
"epoch": 0.7928571428571428,
|
| 3138 |
+
"grad_norm": 0.34384891390800476,
|
| 3139 |
+
"learning_rate": 2.115986662222058e-05,
|
| 3140 |
+
"loss": 1.3456,
|
| 3141 |
+
"step": 444
|
| 3142 |
+
},
|
| 3143 |
+
{
|
| 3144 |
+
"epoch": 0.7946428571428571,
|
| 3145 |
+
"grad_norm": 0.3673519492149353,
|
| 3146 |
+
"learning_rate": 2.0809775407772503e-05,
|
| 3147 |
+
"loss": 1.3223,
|
| 3148 |
+
"step": 445
|
| 3149 |
+
},
|
| 3150 |
+
{
|
| 3151 |
+
"epoch": 0.7964285714285714,
|
| 3152 |
+
"grad_norm": 0.400193452835083,
|
| 3153 |
+
"learning_rate": 2.0462267909238896e-05,
|
| 3154 |
+
"loss": 1.562,
|
| 3155 |
+
"step": 446
|
| 3156 |
+
},
|
| 3157 |
+
{
|
| 3158 |
+
"epoch": 0.7982142857142858,
|
| 3159 |
+
"grad_norm": 0.4361307621002197,
|
| 3160 |
+
"learning_rate": 2.011735546464365e-05,
|
| 3161 |
+
"loss": 1.1107,
|
| 3162 |
+
"step": 447
|
| 3163 |
+
},
|
| 3164 |
+
{
|
| 3165 |
+
"epoch": 0.8,
|
| 3166 |
+
"grad_norm": 0.38674113154411316,
|
| 3167 |
+
"learning_rate": 1.9775049327342486e-05,
|
| 3168 |
+
"loss": 1.2813,
|
| 3169 |
+
"step": 448
|
| 3170 |
+
},
|
| 3171 |
+
{
|
| 3172 |
+
"epoch": 0.8017857142857143,
|
| 3173 |
+
"grad_norm": 0.40437138080596924,
|
| 3174 |
+
"learning_rate": 1.943536066565603e-05,
|
| 3175 |
+
"loss": 1.5164,
|
| 3176 |
+
"step": 449
|
| 3177 |
+
},
|
| 3178 |
+
{
|
| 3179 |
+
"epoch": 0.8035714285714286,
|
| 3180 |
+
"grad_norm": 0.5968692898750305,
|
| 3181 |
+
"learning_rate": 1.9098300562505266e-05,
|
| 3182 |
+
"loss": 2.0903,
|
| 3183 |
+
"step": 450
|
| 3184 |
+
},
|
| 3185 |
+
{
|
| 3186 |
+
"epoch": 0.8053571428571429,
|
| 3187 |
+
"grad_norm": 0.16759343445301056,
|
| 3188 |
+
"learning_rate": 1.876388001504995e-05,
|
| 3189 |
+
"loss": 1.4236,
|
| 3190 |
+
"step": 451
|
| 3191 |
+
},
|
| 3192 |
+
{
|
| 3193 |
+
"epoch": 0.8071428571428572,
|
| 3194 |
+
"grad_norm": 0.1815633326768875,
|
| 3195 |
+
"learning_rate": 1.8432109934329834e-05,
|
| 3196 |
+
"loss": 1.6486,
|
| 3197 |
+
"step": 452
|
| 3198 |
+
},
|
| 3199 |
+
{
|
| 3200 |
+
"epoch": 0.8089285714285714,
|
| 3201 |
+
"grad_norm": 0.18043053150177002,
|
| 3202 |
+
"learning_rate": 1.810300114490875e-05,
|
| 3203 |
+
"loss": 1.4118,
|
| 3204 |
+
"step": 453
|
| 3205 |
+
},
|
| 3206 |
+
{
|
| 3207 |
+
"epoch": 0.8107142857142857,
|
| 3208 |
+
"grad_norm": 0.19016335904598236,
|
| 3209 |
+
"learning_rate": 1.777656438452129e-05,
|
| 3210 |
+
"loss": 1.4844,
|
| 3211 |
+
"step": 454
|
| 3212 |
+
},
|
| 3213 |
+
{
|
| 3214 |
+
"epoch": 0.8125,
|
| 3215 |
+
"grad_norm": 0.2066669464111328,
|
| 3216 |
+
"learning_rate": 1.74528103037226e-05,
|
| 3217 |
+
"loss": 1.583,
|
| 3218 |
+
"step": 455
|
| 3219 |
+
},
|
| 3220 |
+
{
|
| 3221 |
+
"epoch": 0.8142857142857143,
|
| 3222 |
+
"grad_norm": 0.21698947250843048,
|
| 3223 |
+
"learning_rate": 1.713174946554086e-05,
|
| 3224 |
+
"loss": 1.6314,
|
| 3225 |
+
"step": 456
|
| 3226 |
+
},
|
| 3227 |
+
{
|
| 3228 |
+
"epoch": 0.8160714285714286,
|
| 3229 |
+
"grad_norm": 0.21937011182308197,
|
| 3230 |
+
"learning_rate": 1.6813392345132518e-05,
|
| 3231 |
+
"loss": 1.6342,
|
| 3232 |
+
"step": 457
|
| 3233 |
+
},
|
| 3234 |
+
{
|
| 3235 |
+
"epoch": 0.8178571428571428,
|
| 3236 |
+
"grad_norm": 0.24016696214675903,
|
| 3237 |
+
"learning_rate": 1.649774932944075e-05,
|
| 3238 |
+
"loss": 1.7726,
|
| 3239 |
+
"step": 458
|
| 3240 |
+
},
|
| 3241 |
+
{
|
| 3242 |
+
"epoch": 0.8196428571428571,
|
| 3243 |
+
"grad_norm": 0.23045389354228973,
|
| 3244 |
+
"learning_rate": 1.6184830716856347e-05,
|
| 3245 |
+
"loss": 1.8679,
|
| 3246 |
+
"step": 459
|
| 3247 |
+
},
|
| 3248 |
+
{
|
| 3249 |
+
"epoch": 0.8214285714285714,
|
| 3250 |
+
"grad_norm": 0.2114153504371643,
|
| 3251 |
+
"learning_rate": 1.587464671688187e-05,
|
| 3252 |
+
"loss": 1.6223,
|
| 3253 |
+
"step": 460
|
| 3254 |
+
},
|
| 3255 |
+
{
|
| 3256 |
+
"epoch": 0.8232142857142857,
|
| 3257 |
+
"grad_norm": 0.22911155223846436,
|
| 3258 |
+
"learning_rate": 1.5567207449798515e-05,
|
| 3259 |
+
"loss": 1.6179,
|
| 3260 |
+
"step": 461
|
| 3261 |
+
},
|
| 3262 |
+
{
|
| 3263 |
+
"epoch": 0.825,
|
| 3264 |
+
"grad_norm": 0.22748011350631714,
|
| 3265 |
+
"learning_rate": 1.5262522946335755e-05,
|
| 3266 |
+
"loss": 1.7988,
|
| 3267 |
+
"step": 462
|
| 3268 |
+
},
|
| 3269 |
+
{
|
| 3270 |
+
"epoch": 0.8267857142857142,
|
| 3271 |
+
"grad_norm": 0.22931312024593353,
|
| 3272 |
+
"learning_rate": 1.4960603147344343e-05,
|
| 3273 |
+
"loss": 1.6515,
|
| 3274 |
+
"step": 463
|
| 3275 |
+
},
|
| 3276 |
+
{
|
| 3277 |
+
"epoch": 0.8285714285714286,
|
| 3278 |
+
"grad_norm": 0.24372749030590057,
|
| 3279 |
+
"learning_rate": 1.466145790347183e-05,
|
| 3280 |
+
"loss": 1.855,
|
| 3281 |
+
"step": 464
|
| 3282 |
+
},
|
| 3283 |
+
{
|
| 3284 |
+
"epoch": 0.8303571428571429,
|
| 3285 |
+
"grad_norm": 0.24426500499248505,
|
| 3286 |
+
"learning_rate": 1.4365096974841108e-05,
|
| 3287 |
+
"loss": 1.7367,
|
| 3288 |
+
"step": 465
|
| 3289 |
+
},
|
| 3290 |
+
{
|
| 3291 |
+
"epoch": 0.8321428571428572,
|
| 3292 |
+
"grad_norm": 0.22962844371795654,
|
| 3293 |
+
"learning_rate": 1.4071530030732095e-05,
|
| 3294 |
+
"loss": 1.871,
|
| 3295 |
+
"step": 466
|
| 3296 |
+
},
|
| 3297 |
+
{
|
| 3298 |
+
"epoch": 0.8339285714285715,
|
| 3299 |
+
"grad_norm": 0.25352221727371216,
|
| 3300 |
+
"learning_rate": 1.3780766649266242e-05,
|
| 3301 |
+
"loss": 1.816,
|
| 3302 |
+
"step": 467
|
| 3303 |
+
},
|
| 3304 |
+
{
|
| 3305 |
+
"epoch": 0.8357142857142857,
|
| 3306 |
+
"grad_norm": 0.2572453022003174,
|
| 3307 |
+
"learning_rate": 1.3492816317093893e-05,
|
| 3308 |
+
"loss": 1.857,
|
| 3309 |
+
"step": 468
|
| 3310 |
+
},
|
| 3311 |
+
{
|
| 3312 |
+
"epoch": 0.8375,
|
| 3313 |
+
"grad_norm": 0.26578372716903687,
|
| 3314 |
+
"learning_rate": 1.3207688429084974e-05,
|
| 3315 |
+
"loss": 1.8242,
|
| 3316 |
+
"step": 469
|
| 3317 |
+
},
|
| 3318 |
+
{
|
| 3319 |
+
"epoch": 0.8392857142857143,
|
| 3320 |
+
"grad_norm": 0.2806295156478882,
|
| 3321 |
+
"learning_rate": 1.2925392288022298e-05,
|
| 3322 |
+
"loss": 1.9345,
|
| 3323 |
+
"step": 470
|
| 3324 |
+
},
|
| 3325 |
+
{
|
| 3326 |
+
"epoch": 0.8410714285714286,
|
| 3327 |
+
"grad_norm": 0.2791413962841034,
|
| 3328 |
+
"learning_rate": 1.2645937104298111e-05,
|
| 3329 |
+
"loss": 1.4595,
|
| 3330 |
+
"step": 471
|
| 3331 |
+
},
|
| 3332 |
+
{
|
| 3333 |
+
"epoch": 0.8428571428571429,
|
| 3334 |
+
"grad_norm": 0.25607559084892273,
|
| 3335 |
+
"learning_rate": 1.2369331995613665e-05,
|
| 3336 |
+
"loss": 1.493,
|
| 3337 |
+
"step": 472
|
| 3338 |
+
},
|
| 3339 |
+
{
|
| 3340 |
+
"epoch": 0.8446428571428571,
|
| 3341 |
+
"grad_norm": 0.2600158452987671,
|
| 3342 |
+
"learning_rate": 1.2095585986681535e-05,
|
| 3343 |
+
"loss": 1.4394,
|
| 3344 |
+
"step": 473
|
| 3345 |
+
},
|
| 3346 |
+
{
|
| 3347 |
+
"epoch": 0.8464285714285714,
|
| 3348 |
+
"grad_norm": 0.27362141013145447,
|
| 3349 |
+
"learning_rate": 1.1824708008931418e-05,
|
| 3350 |
+
"loss": 1.0637,
|
| 3351 |
+
"step": 474
|
| 3352 |
+
},
|
| 3353 |
+
{
|
| 3354 |
+
"epoch": 0.8482142857142857,
|
| 3355 |
+
"grad_norm": 0.27769190073013306,
|
| 3356 |
+
"learning_rate": 1.1556706900218572e-05,
|
| 3357 |
+
"loss": 1.2003,
|
| 3358 |
+
"step": 475
|
| 3359 |
+
},
|
| 3360 |
+
{
|
| 3361 |
+
"epoch": 0.85,
|
| 3362 |
+
"grad_norm": 0.2426546961069107,
|
| 3363 |
+
"learning_rate": 1.1291591404535462e-05,
|
| 3364 |
+
"loss": 0.8119,
|
| 3365 |
+
"step": 476
|
| 3366 |
+
},
|
| 3367 |
+
{
|
| 3368 |
+
"epoch": 0.8517857142857143,
|
| 3369 |
+
"grad_norm": 0.2766360938549042,
|
| 3370 |
+
"learning_rate": 1.1029370171726571e-05,
|
| 3371 |
+
"loss": 1.0697,
|
| 3372 |
+
"step": 477
|
| 3373 |
+
},
|
| 3374 |
+
{
|
| 3375 |
+
"epoch": 0.8535714285714285,
|
| 3376 |
+
"grad_norm": 0.31444060802459717,
|
| 3377 |
+
"learning_rate": 1.0770051757206079e-05,
|
| 3378 |
+
"loss": 1.3142,
|
| 3379 |
+
"step": 478
|
| 3380 |
+
},
|
| 3381 |
+
{
|
| 3382 |
+
"epoch": 0.8553571428571428,
|
| 3383 |
+
"grad_norm": 0.33098921179771423,
|
| 3384 |
+
"learning_rate": 1.051364462167881e-05,
|
| 3385 |
+
"loss": 1.0656,
|
| 3386 |
+
"step": 479
|
| 3387 |
+
},
|
| 3388 |
+
{
|
| 3389 |
+
"epoch": 0.8571428571428571,
|
| 3390 |
+
"grad_norm": 0.327269583940506,
|
| 3391 |
+
"learning_rate": 1.026015713086418e-05,
|
| 3392 |
+
"loss": 1.1335,
|
| 3393 |
+
"step": 480
|
| 3394 |
+
},
|
| 3395 |
+
{
|
| 3396 |
+
"epoch": 0.8589285714285714,
|
| 3397 |
+
"grad_norm": 0.30396682024002075,
|
| 3398 |
+
"learning_rate": 1.0009597555223128e-05,
|
| 3399 |
+
"loss": 1.2085,
|
| 3400 |
+
"step": 481
|
| 3401 |
+
},
|
| 3402 |
+
{
|
| 3403 |
+
"epoch": 0.8607142857142858,
|
| 3404 |
+
"grad_norm": 0.30377310514450073,
|
| 3405 |
+
"learning_rate": 9.761974069688461e-06,
|
| 3406 |
+
"loss": 1.3687,
|
| 3407 |
+
"step": 482
|
| 3408 |
+
},
|
| 3409 |
+
{
|
| 3410 |
+
"epoch": 0.8625,
|
| 3411 |
+
"grad_norm": 0.3333049714565277,
|
| 3412 |
+
"learning_rate": 9.517294753398064e-06,
|
| 3413 |
+
"loss": 1.0564,
|
| 3414 |
+
"step": 483
|
| 3415 |
+
},
|
| 3416 |
+
{
|
| 3417 |
+
"epoch": 0.8642857142857143,
|
| 3418 |
+
"grad_norm": 0.3406635820865631,
|
| 3419 |
+
"learning_rate": 9.275567589431178e-06,
|
| 3420 |
+
"loss": 1.1551,
|
| 3421 |
+
"step": 484
|
| 3422 |
+
},
|
| 3423 |
+
{
|
| 3424 |
+
"epoch": 0.8660714285714286,
|
| 3425 |
+
"grad_norm": 0.35342979431152344,
|
| 3426 |
+
"learning_rate": 9.036800464548157e-06,
|
| 3427 |
+
"loss": 0.8579,
|
| 3428 |
+
"step": 485
|
| 3429 |
+
},
|
| 3430 |
+
{
|
| 3431 |
+
"epoch": 0.8678571428571429,
|
| 3432 |
+
"grad_norm": 0.338238000869751,
|
| 3433 |
+
"learning_rate": 8.80100116893301e-06,
|
| 3434 |
+
"loss": 1.2404,
|
| 3435 |
+
"step": 486
|
| 3436 |
+
},
|
| 3437 |
+
{
|
| 3438 |
+
"epoch": 0.8696428571428572,
|
| 3439 |
+
"grad_norm": 0.32925620675086975,
|
| 3440 |
+
"learning_rate": 8.568177395939215e-06,
|
| 3441 |
+
"loss": 0.9557,
|
| 3442 |
+
"step": 487
|
| 3443 |
+
},
|
| 3444 |
+
{
|
| 3445 |
+
"epoch": 0.8714285714285714,
|
| 3446 |
+
"grad_norm": 0.36271733045578003,
|
| 3447 |
+
"learning_rate": 8.338336741838838e-06,
|
| 3448 |
+
"loss": 1.1323,
|
| 3449 |
+
"step": 488
|
| 3450 |
+
},
|
| 3451 |
+
{
|
| 3452 |
+
"epoch": 0.8732142857142857,
|
| 3453 |
+
"grad_norm": 0.373471736907959,
|
| 3454 |
+
"learning_rate": 8.111486705574534e-06,
|
| 3455 |
+
"loss": 0.9512,
|
| 3456 |
+
"step": 489
|
| 3457 |
+
},
|
| 3458 |
+
{
|
| 3459 |
+
"epoch": 0.875,
|
| 3460 |
+
"grad_norm": 0.3971388041973114,
|
| 3461 |
+
"learning_rate": 7.887634688515e-06,
|
| 3462 |
+
"loss": 1.0311,
|
| 3463 |
+
"step": 490
|
| 3464 |
+
},
|
| 3465 |
+
{
|
| 3466 |
+
"epoch": 0.8767857142857143,
|
| 3467 |
+
"grad_norm": 0.35106372833251953,
|
| 3468 |
+
"learning_rate": 7.666787994213453e-06,
|
| 3469 |
+
"loss": 1.0705,
|
| 3470 |
+
"step": 491
|
| 3471 |
+
},
|
| 3472 |
+
{
|
| 3473 |
+
"epoch": 0.8785714285714286,
|
| 3474 |
+
"grad_norm": 0.3654085099697113,
|
| 3475 |
+
"learning_rate": 7.448953828169314e-06,
|
| 3476 |
+
"loss": 1.2338,
|
| 3477 |
+
"step": 492
|
| 3478 |
+
},
|
| 3479 |
+
{
|
| 3480 |
+
"epoch": 0.8803571428571428,
|
| 3481 |
+
"grad_norm": 0.3616616725921631,
|
| 3482 |
+
"learning_rate": 7.2341392975931785e-06,
|
| 3483 |
+
"loss": 1.0171,
|
| 3484 |
+
"step": 493
|
| 3485 |
+
},
|
| 3486 |
+
{
|
| 3487 |
+
"epoch": 0.8821428571428571,
|
| 3488 |
+
"grad_norm": 0.39053040742874146,
|
| 3489 |
+
"learning_rate": 7.022351411174866e-06,
|
| 3490 |
+
"loss": 1.2481,
|
| 3491 |
+
"step": 494
|
| 3492 |
+
},
|
| 3493 |
+
{
|
| 3494 |
+
"epoch": 0.8839285714285714,
|
| 3495 |
+
"grad_norm": 0.36361098289489746,
|
| 3496 |
+
"learning_rate": 6.813597078854772e-06,
|
| 3497 |
+
"loss": 1.3712,
|
| 3498 |
+
"step": 495
|
| 3499 |
+
},
|
| 3500 |
+
{
|
| 3501 |
+
"epoch": 0.8857142857142857,
|
| 3502 |
+
"grad_norm": 0.3919769823551178,
|
| 3503 |
+
"learning_rate": 6.607883111598445e-06,
|
| 3504 |
+
"loss": 1.2261,
|
| 3505 |
+
"step": 496
|
| 3506 |
+
},
|
| 3507 |
+
{
|
| 3508 |
+
"epoch": 0.8875,
|
| 3509 |
+
"grad_norm": 0.3980182707309723,
|
| 3510 |
+
"learning_rate": 6.405216221174326e-06,
|
| 3511 |
+
"loss": 1.2769,
|
| 3512 |
+
"step": 497
|
| 3513 |
+
},
|
| 3514 |
+
{
|
| 3515 |
+
"epoch": 0.8892857142857142,
|
| 3516 |
+
"grad_norm": 0.4138076901435852,
|
| 3517 |
+
"learning_rate": 6.205603019934791e-06,
|
| 3518 |
+
"loss": 1.4143,
|
| 3519 |
+
"step": 498
|
| 3520 |
+
},
|
| 3521 |
+
{
|
| 3522 |
+
"epoch": 0.8910714285714286,
|
| 3523 |
+
"grad_norm": 0.47321876883506775,
|
| 3524 |
+
"learning_rate": 6.009050020600459e-06,
|
| 3525 |
+
"loss": 1.687,
|
| 3526 |
+
"step": 499
|
| 3527 |
+
},
|
| 3528 |
+
{
|
| 3529 |
+
"epoch": 0.8928571428571429,
|
| 3530 |
+
"grad_norm": 0.532382607460022,
|
| 3531 |
+
"learning_rate": 5.8155636360475385e-06,
|
| 3532 |
+
"loss": 2.0251,
|
| 3533 |
+
"step": 500
|
| 3534 |
+
},
|
| 3535 |
+
{
|
| 3536 |
+
"epoch": 0.8946428571428572,
|
| 3537 |
+
"grad_norm": 0.16291013360023499,
|
| 3538 |
+
"learning_rate": 5.625150179098804e-06,
|
| 3539 |
+
"loss": 1.3453,
|
| 3540 |
+
"step": 501
|
| 3541 |
+
},
|
| 3542 |
+
{
|
| 3543 |
+
"epoch": 0.8964285714285715,
|
| 3544 |
+
"grad_norm": 0.19601905345916748,
|
| 3545 |
+
"learning_rate": 5.437815862317519e-06,
|
| 3546 |
+
"loss": 1.6226,
|
| 3547 |
+
"step": 502
|
| 3548 |
+
},
|
| 3549 |
+
{
|
| 3550 |
+
"epoch": 0.8982142857142857,
|
| 3551 |
+
"grad_norm": 0.20232488214969635,
|
| 3552 |
+
"learning_rate": 5.25356679780471e-06,
|
| 3553 |
+
"loss": 1.7009,
|
| 3554 |
+
"step": 503
|
| 3555 |
+
},
|
| 3556 |
+
{
|
| 3557 |
+
"epoch": 0.9,
|
| 3558 |
+
"grad_norm": 0.21668295562267303,
|
| 3559 |
+
"learning_rate": 5.072408996999844e-06,
|
| 3560 |
+
"loss": 1.674,
|
| 3561 |
+
"step": 504
|
| 3562 |
+
},
|
| 3563 |
+
{
|
| 3564 |
+
"epoch": 0.9017857142857143,
|
| 3565 |
+
"grad_norm": 0.208509162068367,
|
| 3566 |
+
"learning_rate": 4.8943483704846475e-06,
|
| 3567 |
+
"loss": 1.7305,
|
| 3568 |
+
"step": 505
|
| 3569 |
+
},
|
| 3570 |
+
{
|
| 3571 |
+
"epoch": 0.9035714285714286,
|
| 3572 |
+
"grad_norm": 0.22677063941955566,
|
| 3573 |
+
"learning_rate": 4.719390727790218e-06,
|
| 3574 |
+
"loss": 1.9097,
|
| 3575 |
+
"step": 506
|
| 3576 |
+
},
|
| 3577 |
+
{
|
| 3578 |
+
"epoch": 0.9053571428571429,
|
| 3579 |
+
"grad_norm": 0.2320660501718521,
|
| 3580 |
+
"learning_rate": 4.547541777207565e-06,
|
| 3581 |
+
"loss": 1.6309,
|
| 3582 |
+
"step": 507
|
| 3583 |
+
},
|
| 3584 |
+
{
|
| 3585 |
+
"epoch": 0.9071428571428571,
|
| 3586 |
+
"grad_norm": 0.21626609563827515,
|
| 3587 |
+
"learning_rate": 4.378807125601303e-06,
|
| 3588 |
+
"loss": 1.8275,
|
| 3589 |
+
"step": 508
|
| 3590 |
+
},
|
| 3591 |
+
{
|
| 3592 |
+
"epoch": 0.9089285714285714,
|
| 3593 |
+
"grad_norm": 0.2348957508802414,
|
| 3594 |
+
"learning_rate": 4.2131922782267405e-06,
|
| 3595 |
+
"loss": 1.6609,
|
| 3596 |
+
"step": 509
|
| 3597 |
+
},
|
| 3598 |
+
{
|
| 3599 |
+
"epoch": 0.9107142857142857,
|
| 3600 |
+
"grad_norm": 0.21976341307163239,
|
| 3601 |
+
"learning_rate": 4.050702638550275e-06,
|
| 3602 |
+
"loss": 1.6476,
|
| 3603 |
+
"step": 510
|
| 3604 |
+
},
|
| 3605 |
+
{
|
| 3606 |
+
"epoch": 0.9125,
|
| 3607 |
+
"grad_norm": 0.25860780477523804,
|
| 3608 |
+
"learning_rate": 3.891343508073053e-06,
|
| 3609 |
+
"loss": 1.7382,
|
| 3610 |
+
"step": 511
|
| 3611 |
+
},
|
| 3612 |
+
{
|
| 3613 |
+
"epoch": 0.9142857142857143,
|
| 3614 |
+
"grad_norm": 0.24138076603412628,
|
| 3615 |
+
"learning_rate": 3.7351200861580617e-06,
|
| 3616 |
+
"loss": 1.7668,
|
| 3617 |
+
"step": 512
|
| 3618 |
+
},
|
| 3619 |
+
{
|
| 3620 |
+
"epoch": 0.9160714285714285,
|
| 3621 |
+
"grad_norm": 0.23734422028064728,
|
| 3622 |
+
"learning_rate": 3.5820374698604555e-06,
|
| 3623 |
+
"loss": 1.794,
|
| 3624 |
+
"step": 513
|
| 3625 |
+
},
|
| 3626 |
+
{
|
| 3627 |
+
"epoch": 0.9178571428571428,
|
| 3628 |
+
"grad_norm": 0.23496972024440765,
|
| 3629 |
+
"learning_rate": 3.4321006537612165e-06,
|
| 3630 |
+
"loss": 1.7674,
|
| 3631 |
+
"step": 514
|
| 3632 |
+
},
|
| 3633 |
+
{
|
| 3634 |
+
"epoch": 0.9196428571428571,
|
| 3635 |
+
"grad_norm": 0.24626807868480682,
|
| 3636 |
+
"learning_rate": 3.2853145298042953e-06,
|
| 3637 |
+
"loss": 2.0287,
|
| 3638 |
+
"step": 515
|
| 3639 |
+
},
|
| 3640 |
+
{
|
| 3641 |
+
"epoch": 0.9214285714285714,
|
| 3642 |
+
"grad_norm": 0.24339932203292847,
|
| 3643 |
+
"learning_rate": 3.1416838871368924e-06,
|
| 3644 |
+
"loss": 1.8533,
|
| 3645 |
+
"step": 516
|
| 3646 |
+
},
|
| 3647 |
+
{
|
| 3648 |
+
"epoch": 0.9232142857142858,
|
| 3649 |
+
"grad_norm": 0.253841370344162,
|
| 3650 |
+
"learning_rate": 3.0012134119532964e-06,
|
| 3651 |
+
"loss": 1.6682,
|
| 3652 |
+
"step": 517
|
| 3653 |
+
},
|
| 3654 |
+
{
|
| 3655 |
+
"epoch": 0.925,
|
| 3656 |
+
"grad_norm": 0.26810407638549805,
|
| 3657 |
+
"learning_rate": 2.863907687341949e-06,
|
| 3658 |
+
"loss": 1.4686,
|
| 3659 |
+
"step": 518
|
| 3660 |
+
},
|
| 3661 |
+
{
|
| 3662 |
+
"epoch": 0.9267857142857143,
|
| 3663 |
+
"grad_norm": 0.24126453697681427,
|
| 3664 |
+
"learning_rate": 2.7297711931358993e-06,
|
| 3665 |
+
"loss": 1.0048,
|
| 3666 |
+
"step": 519
|
| 3667 |
+
},
|
| 3668 |
+
{
|
| 3669 |
+
"epoch": 0.9285714285714286,
|
| 3670 |
+
"grad_norm": 0.3192795515060425,
|
| 3671 |
+
"learning_rate": 2.5988083057666533e-06,
|
| 3672 |
+
"loss": 1.4609,
|
| 3673 |
+
"step": 520
|
| 3674 |
+
},
|
| 3675 |
+
{
|
| 3676 |
+
"epoch": 0.9303571428571429,
|
| 3677 |
+
"grad_norm": 0.2680839002132416,
|
| 3678 |
+
"learning_rate": 2.471023298121422e-06,
|
| 3679 |
+
"loss": 1.4287,
|
| 3680 |
+
"step": 521
|
| 3681 |
+
},
|
| 3682 |
+
{
|
| 3683 |
+
"epoch": 0.9321428571428572,
|
| 3684 |
+
"grad_norm": 0.34683963656425476,
|
| 3685 |
+
"learning_rate": 2.3464203394036322e-06,
|
| 3686 |
+
"loss": 1.1019,
|
| 3687 |
+
"step": 522
|
| 3688 |
+
},
|
| 3689 |
+
{
|
| 3690 |
+
"epoch": 0.9339285714285714,
|
| 3691 |
+
"grad_norm": 0.31293728947639465,
|
| 3692 |
+
"learning_rate": 2.2250034949969913e-06,
|
| 3693 |
+
"loss": 1.5812,
|
| 3694 |
+
"step": 523
|
| 3695 |
+
},
|
| 3696 |
+
{
|
| 3697 |
+
"epoch": 0.9357142857142857,
|
| 3698 |
+
"grad_norm": 0.27005112171173096,
|
| 3699 |
+
"learning_rate": 2.1067767263327933e-06,
|
| 3700 |
+
"loss": 0.8701,
|
| 3701 |
+
"step": 524
|
| 3702 |
+
},
|
| 3703 |
+
{
|
| 3704 |
+
"epoch": 0.9375,
|
| 3705 |
+
"grad_norm": 0.2739523649215698,
|
| 3706 |
+
"learning_rate": 1.9917438907606556e-06,
|
| 3707 |
+
"loss": 1.1165,
|
| 3708 |
+
"step": 525
|
| 3709 |
+
},
|
| 3710 |
+
{
|
| 3711 |
+
"epoch": 0.9392857142857143,
|
| 3712 |
+
"grad_norm": 0.2776015102863312,
|
| 3713 |
+
"learning_rate": 1.87990874142272e-06,
|
| 3714 |
+
"loss": 1.1448,
|
| 3715 |
+
"step": 526
|
| 3716 |
+
},
|
| 3717 |
+
{
|
| 3718 |
+
"epoch": 0.9410714285714286,
|
| 3719 |
+
"grad_norm": 0.30518367886543274,
|
| 3720 |
+
"learning_rate": 1.771274927131139e-06,
|
| 3721 |
+
"loss": 1.0695,
|
| 3722 |
+
"step": 527
|
| 3723 |
+
},
|
| 3724 |
+
{
|
| 3725 |
+
"epoch": 0.9428571428571428,
|
| 3726 |
+
"grad_norm": 0.31975796818733215,
|
| 3727 |
+
"learning_rate": 1.665845992249071e-06,
|
| 3728 |
+
"loss": 1.1755,
|
| 3729 |
+
"step": 528
|
| 3730 |
+
},
|
| 3731 |
+
{
|
| 3732 |
+
"epoch": 0.9446428571428571,
|
| 3733 |
+
"grad_norm": 0.3304663598537445,
|
| 3734 |
+
"learning_rate": 1.5636253765750508e-06,
|
| 3735 |
+
"loss": 1.0217,
|
| 3736 |
+
"step": 529
|
| 3737 |
+
},
|
| 3738 |
+
{
|
| 3739 |
+
"epoch": 0.9464285714285714,
|
| 3740 |
+
"grad_norm": 0.3353167474269867,
|
| 3741 |
+
"learning_rate": 1.4646164152307018e-06,
|
| 3742 |
+
"loss": 1.2216,
|
| 3743 |
+
"step": 530
|
| 3744 |
+
},
|
| 3745 |
+
{
|
| 3746 |
+
"epoch": 0.9482142857142857,
|
| 3747 |
+
"grad_norm": 0.3266999423503876,
|
| 3748 |
+
"learning_rate": 1.3688223385519672e-06,
|
| 3749 |
+
"loss": 0.8658,
|
| 3750 |
+
"step": 531
|
| 3751 |
+
},
|
| 3752 |
+
{
|
| 3753 |
+
"epoch": 0.95,
|
| 3754 |
+
"grad_norm": 0.34619686007499695,
|
| 3755 |
+
"learning_rate": 1.2762462719837275e-06,
|
| 3756 |
+
"loss": 1.2839,
|
| 3757 |
+
"step": 532
|
| 3758 |
+
},
|
| 3759 |
+
{
|
| 3760 |
+
"epoch": 0.9517857142857142,
|
| 3761 |
+
"grad_norm": 0.32094186544418335,
|
| 3762 |
+
"learning_rate": 1.1868912359777607e-06,
|
| 3763 |
+
"loss": 1.1571,
|
| 3764 |
+
"step": 533
|
| 3765 |
+
},
|
| 3766 |
+
{
|
| 3767 |
+
"epoch": 0.9535714285714286,
|
| 3768 |
+
"grad_norm": 0.33895429968833923,
|
| 3769 |
+
"learning_rate": 1.1007601458942752e-06,
|
| 3770 |
+
"loss": 1.0222,
|
| 3771 |
+
"step": 534
|
| 3772 |
+
},
|
| 3773 |
+
{
|
| 3774 |
+
"epoch": 0.9553571428571429,
|
| 3775 |
+
"grad_norm": 0.34143325686454773,
|
| 3776 |
+
"learning_rate": 1.0178558119067315e-06,
|
| 3777 |
+
"loss": 0.9386,
|
| 3778 |
+
"step": 535
|
| 3779 |
+
},
|
| 3780 |
+
{
|
| 3781 |
+
"epoch": 0.9571428571428572,
|
| 3782 |
+
"grad_norm": 0.3505565822124481,
|
| 3783 |
+
"learning_rate": 9.381809389101825e-07,
|
| 3784 |
+
"loss": 1.1175,
|
| 3785 |
+
"step": 536
|
| 3786 |
+
},
|
| 3787 |
+
{
|
| 3788 |
+
"epoch": 0.9589285714285715,
|
| 3789 |
+
"grad_norm": 0.36876025795936584,
|
| 3790 |
+
"learning_rate": 8.617381264330426e-07,
|
| 3791 |
+
"loss": 1.0065,
|
| 3792 |
+
"step": 537
|
| 3793 |
+
},
|
| 3794 |
+
{
|
| 3795 |
+
"epoch": 0.9607142857142857,
|
| 3796 |
+
"grad_norm": 0.3824593722820282,
|
| 3797 |
+
"learning_rate": 7.885298685522235e-07,
|
| 3798 |
+
"loss": 1.1445,
|
| 3799 |
+
"step": 538
|
| 3800 |
+
},
|
| 3801 |
+
{
|
| 3802 |
+
"epoch": 0.9625,
|
| 3803 |
+
"grad_norm": 0.3944648504257202,
|
| 3804 |
+
"learning_rate": 7.185585538117657e-07,
|
| 3805 |
+
"loss": 0.9773,
|
| 3806 |
+
"step": 539
|
| 3807 |
+
},
|
| 3808 |
+
{
|
| 3809 |
+
"epoch": 0.9642857142857143,
|
| 3810 |
+
"grad_norm": 0.32271888852119446,
|
| 3811 |
+
"learning_rate": 6.518264651449779e-07,
|
| 3812 |
+
"loss": 1.073,
|
| 3813 |
+
"step": 540
|
| 3814 |
+
},
|
| 3815 |
+
{
|
| 3816 |
+
"epoch": 0.9660714285714286,
|
| 3817 |
+
"grad_norm": 0.36668431758880615,
|
| 3818 |
+
"learning_rate": 5.883357797998757e-07,
|
| 3819 |
+
"loss": 1.3001,
|
| 3820 |
+
"step": 541
|
| 3821 |
+
},
|
| 3822 |
+
{
|
| 3823 |
+
"epoch": 0.9678571428571429,
|
| 3824 |
+
"grad_norm": 0.36653202772140503,
|
| 3825 |
+
"learning_rate": 5.280885692681592e-07,
|
| 3826 |
+
"loss": 1.0388,
|
| 3827 |
+
"step": 542
|
| 3828 |
+
},
|
| 3829 |
+
{
|
| 3830 |
+
"epoch": 0.9696428571428571,
|
| 3831 |
+
"grad_norm": 0.3570370674133301,
|
| 3832 |
+
"learning_rate": 4.710867992176682e-07,
|
| 3833 |
+
"loss": 1.0728,
|
| 3834 |
+
"step": 543
|
| 3835 |
+
},
|
| 3836 |
+
{
|
| 3837 |
+
"epoch": 0.9714285714285714,
|
| 3838 |
+
"grad_norm": 0.3680841624736786,
|
| 3839 |
+
"learning_rate": 4.173323294281994e-07,
|
| 3840 |
+
"loss": 1.3424,
|
| 3841 |
+
"step": 544
|
| 3842 |
+
},
|
| 3843 |
+
{
|
| 3844 |
+
"epoch": 0.9732142857142857,
|
| 3845 |
+
"grad_norm": 0.37311863899230957,
|
| 3846 |
+
"learning_rate": 3.6682691373086665e-07,
|
| 3847 |
+
"loss": 1.3634,
|
| 3848 |
+
"step": 545
|
| 3849 |
+
},
|
| 3850 |
+
{
|
| 3851 |
+
"epoch": 0.975,
|
| 3852 |
+
"grad_norm": 0.38995441794395447,
|
| 3853 |
+
"learning_rate": 3.195721999508461e-07,
|
| 3854 |
+
"loss": 1.4773,
|
| 3855 |
+
"step": 546
|
| 3856 |
+
},
|
| 3857 |
+
{
|
| 3858 |
+
"epoch": 0.9767857142857143,
|
| 3859 |
+
"grad_norm": 0.3970450162887573,
|
| 3860 |
+
"learning_rate": 2.7556972985363085e-07,
|
| 3861 |
+
"loss": 1.3839,
|
| 3862 |
+
"step": 547
|
| 3863 |
+
},
|
| 3864 |
+
{
|
| 3865 |
+
"epoch": 0.9785714285714285,
|
| 3866 |
+
"grad_norm": 0.5267580151557922,
|
| 3867 |
+
"learning_rate": 2.3482093909473756e-07,
|
| 3868 |
+
"loss": 1.9768,
|
| 3869 |
+
"step": 548
|
| 3870 |
+
},
|
| 3871 |
+
{
|
| 3872 |
+
"epoch": 0.9803571428571428,
|
| 3873 |
+
"grad_norm": 0.4860612154006958,
|
| 3874 |
+
"learning_rate": 1.973271571728441e-07,
|
| 3875 |
+
"loss": 1.7893,
|
| 3876 |
+
"step": 549
|
| 3877 |
+
},
|
| 3878 |
+
{
|
| 3879 |
+
"epoch": 0.9821428571428571,
|
| 3880 |
+
"grad_norm": 0.9272521734237671,
|
| 3881 |
+
"learning_rate": 1.630896073864352e-07,
|
| 3882 |
+
"loss": 2.1017,
|
| 3883 |
+
"step": 550
|
| 3884 |
+
},
|
| 3885 |
+
{
|
| 3886 |
+
"epoch": 0.9839285714285714,
|
| 3887 |
+
"grad_norm": 0.19919352233409882,
|
| 3888 |
+
"learning_rate": 1.3210940679385664e-07,
|
| 3889 |
+
"loss": 1.5237,
|
| 3890 |
+
"step": 551
|
| 3891 |
+
},
|
| 3892 |
+
{
|
| 3893 |
+
"epoch": 0.9857142857142858,
|
| 3894 |
+
"grad_norm": 0.20294204354286194,
|
| 3895 |
+
"learning_rate": 1.0438756617691115e-07,
|
| 3896 |
+
"loss": 1.597,
|
| 3897 |
+
"step": 552
|
| 3898 |
+
},
|
| 3899 |
+
{
|
| 3900 |
+
"epoch": 0.9875,
|
| 3901 |
+
"grad_norm": 0.2281683087348938,
|
| 3902 |
+
"learning_rate": 7.992499000785136e-08,
|
| 3903 |
+
"loss": 1.7265,
|
| 3904 |
+
"step": 553
|
| 3905 |
+
},
|
| 3906 |
+
{
|
| 3907 |
+
"epoch": 0.9892857142857143,
|
| 3908 |
+
"grad_norm": 0.2398226112127304,
|
| 3909 |
+
"learning_rate": 5.872247641987016e-08,
|
| 3910 |
+
"loss": 1.7301,
|
| 3911 |
+
"step": 554
|
| 3912 |
+
},
|
| 3913 |
+
{
|
| 3914 |
+
"epoch": 0.9910714285714286,
|
| 3915 |
+
"grad_norm": 0.2581394612789154,
|
| 3916 |
+
"learning_rate": 4.078071718107701e-08,
|
| 3917 |
+
"loss": 1.6294,
|
| 3918 |
+
"step": 555
|
| 3919 |
+
},
|
| 3920 |
+
{
|
| 3921 |
+
"epoch": 0.9928571428571429,
|
| 3922 |
+
"grad_norm": 0.2676229774951935,
|
| 3923 |
+
"learning_rate": 2.610029767191602e-08,
|
| 3924 |
+
"loss": 0.9741,
|
| 3925 |
+
"step": 556
|
| 3926 |
+
},
|
| 3927 |
+
{
|
| 3928 |
+
"epoch": 0.9946428571428572,
|
| 3929 |
+
"grad_norm": 0.30877920985221863,
|
| 3930 |
+
"learning_rate": 1.4681696866081229e-08,
|
| 3931 |
+
"loss": 1.1485,
|
| 3932 |
+
"step": 557
|
| 3933 |
+
},
|
| 3934 |
+
{
|
| 3935 |
+
"epoch": 0.9964285714285714,
|
| 3936 |
+
"grad_norm": 0.3330385386943817,
|
| 3937 |
+
"learning_rate": 6.525287314851358e-09,
|
| 3938 |
+
"loss": 0.906,
|
| 3939 |
+
"step": 558
|
| 3940 |
+
},
|
| 3941 |
+
{
|
| 3942 |
+
"epoch": 0.9982142857142857,
|
| 3943 |
+
"grad_norm": 0.364859402179718,
|
| 3944 |
+
"learning_rate": 1.6313351349883655e-09,
|
| 3945 |
+
"loss": 1.0913,
|
| 3946 |
+
"step": 559
|
| 3947 |
+
},
|
| 3948 |
+
{
|
| 3949 |
+
"epoch": 1.0,
|
| 3950 |
+
"grad_norm": 0.5055291652679443,
|
| 3951 |
+
"learning_rate": 0.0,
|
| 3952 |
+
"loss": 1.6353,
|
| 3953 |
+
"step": 560
|
| 3954 |
+
},
|
| 3955 |
+
{
|
| 3956 |
+
"epoch": 1.0,
|
| 3957 |
+
"eval_loss": 1.4418776035308838,
|
| 3958 |
+
"eval_runtime": 13.3713,
|
| 3959 |
+
"eval_samples_per_second": 17.65,
|
| 3960 |
+
"eval_steps_per_second": 8.825,
|
| 3961 |
+
"step": 560
|
| 3962 |
}
|
| 3963 |
],
|
| 3964 |
"logging_steps": 1,
|
|
|
|
| 3973 |
"should_evaluate": false,
|
| 3974 |
"should_log": false,
|
| 3975 |
"should_save": true,
|
| 3976 |
+
"should_training_stop": true
|
| 3977 |
},
|
| 3978 |
"attributes": {}
|
| 3979 |
}
|
| 3980 |
},
|
| 3981 |
+
"total_flos": 1.826642274948219e+17,
|
| 3982 |
"train_batch_size": 2,
|
| 3983 |
"trial_name": null,
|
| 3984 |
"trial_params": null
|