Upload folder using huggingface_hub
Browse files- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +831 -6
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 738232680
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b50522bbf97853647b3e7dd976a7b77143060a9d6ae9f84a139704035d1b37e
|
| 3 |
size 738232680
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1476611275
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5db096db6ef6cca049587aa70a5476910a9efc056a7da2ade2fd3a3e127d7b24
|
| 3 |
size 1476611275
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cc40fda2dc7baa0544c9cb1a8fdefc63f3880e97dda10b6d0305d014ae7c30be
|
| 3 |
size 14645
|
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 sha256:600c82ad1c2a74114e7bf057a4638d5bdb5e6d378a5ca371abf2b3eea8ba107f
|
| 3 |
size 1465
|
trainer_state.json
CHANGED
|
@@ -1,10 +1,10 @@
|
|
| 1 |
{
|
| 2 |
-
"best_global_step":
|
| 3 |
-
"best_metric": 0.
|
| 4 |
-
"best_model_checkpoint": "/content/drive/MyDrive/lora_model/outputs/task15_microsoft/Phi-4-mini-instruct/checkpoint-
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 1,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -908,6 +908,831 @@
|
|
| 908 |
"eval_samples_per_second": 33.566,
|
| 909 |
"eval_steps_per_second": 4.475,
|
| 910 |
"step": 60
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 911 |
}
|
| 912 |
],
|
| 913 |
"logging_steps": 1,
|
|
@@ -927,7 +1752,7 @@
|
|
| 927 |
"attributes": {}
|
| 928 |
}
|
| 929 |
},
|
| 930 |
-
"total_flos":
|
| 931 |
"train_batch_size": 1,
|
| 932 |
"trial_name": null,
|
| 933 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 115,
|
| 3 |
+
"best_metric": 0.09458151459693909,
|
| 4 |
+
"best_model_checkpoint": "/content/drive/MyDrive/lora_model/outputs/task15_microsoft/Phi-4-mini-instruct/checkpoint-115",
|
| 5 |
+
"epoch": 6.052631578947368,
|
| 6 |
"eval_steps": 1,
|
| 7 |
+
"global_step": 115,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 908 |
"eval_samples_per_second": 33.566,
|
| 909 |
"eval_steps_per_second": 4.475,
|
| 910 |
"step": 60
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 3.2105263157894735,
|
| 914 |
+
"grad_norm": 2.3808937072753906,
|
| 915 |
+
"learning_rate": 0.0004561103900854401,
|
| 916 |
+
"loss": 0.5372,
|
| 917 |
+
"step": 61
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 3.2105263157894735,
|
| 921 |
+
"eval_loss": 0.535223662853241,
|
| 922 |
+
"eval_runtime": 0.8966,
|
| 923 |
+
"eval_samples_per_second": 33.459,
|
| 924 |
+
"eval_steps_per_second": 4.461,
|
| 925 |
+
"step": 61
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 3.263157894736842,
|
| 929 |
+
"grad_norm": 1.8272178173065186,
|
| 930 |
+
"learning_rate": 0.0004542005660466094,
|
| 931 |
+
"loss": 0.5399,
|
| 932 |
+
"step": 62
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 3.263157894736842,
|
| 936 |
+
"eval_loss": 0.5316082239151001,
|
| 937 |
+
"eval_runtime": 0.8994,
|
| 938 |
+
"eval_samples_per_second": 33.354,
|
| 939 |
+
"eval_steps_per_second": 4.447,
|
| 940 |
+
"step": 62
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 3.3157894736842106,
|
| 944 |
+
"grad_norm": 2.0635435581207275,
|
| 945 |
+
"learning_rate": 0.0004522542485937369,
|
| 946 |
+
"loss": 0.5531,
|
| 947 |
+
"step": 63
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.3157894736842106,
|
| 951 |
+
"eval_loss": 0.5134085416793823,
|
| 952 |
+
"eval_runtime": 0.8937,
|
| 953 |
+
"eval_samples_per_second": 33.567,
|
| 954 |
+
"eval_steps_per_second": 4.476,
|
| 955 |
+
"step": 63
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 3.3684210526315788,
|
| 959 |
+
"grad_norm": 2.268183708190918,
|
| 960 |
+
"learning_rate": 0.0004502717855601809,
|
| 961 |
+
"loss": 0.5291,
|
| 962 |
+
"step": 64
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 3.3684210526315788,
|
| 966 |
+
"eval_loss": 0.5419598817825317,
|
| 967 |
+
"eval_runtime": 0.8959,
|
| 968 |
+
"eval_samples_per_second": 33.486,
|
| 969 |
+
"eval_steps_per_second": 4.465,
|
| 970 |
+
"step": 64
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 3.4210526315789473,
|
| 974 |
+
"grad_norm": 1.8800358772277832,
|
| 975 |
+
"learning_rate": 0.0004482535312390058,
|
| 976 |
+
"loss": 0.5501,
|
| 977 |
+
"step": 65
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 3.4210526315789473,
|
| 981 |
+
"eval_loss": 0.5209227800369263,
|
| 982 |
+
"eval_runtime": 0.8927,
|
| 983 |
+
"eval_samples_per_second": 33.606,
|
| 984 |
+
"eval_steps_per_second": 4.481,
|
| 985 |
+
"step": 65
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 3.473684210526316,
|
| 989 |
+
"grad_norm": 3.1507558822631836,
|
| 990 |
+
"learning_rate": 0.00044619984631966527,
|
| 991 |
+
"loss": 0.5309,
|
| 992 |
+
"step": 66
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 3.473684210526316,
|
| 996 |
+
"eval_loss": 0.536996603012085,
|
| 997 |
+
"eval_runtime": 0.8951,
|
| 998 |
+
"eval_samples_per_second": 33.517,
|
| 999 |
+
"eval_steps_per_second": 4.469,
|
| 1000 |
+
"step": 66
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 3.526315789473684,
|
| 1004 |
+
"grad_norm": 3.5700478553771973,
|
| 1005 |
+
"learning_rate": 0.0004441110978235418,
|
| 1006 |
+
"loss": 0.7223,
|
| 1007 |
+
"step": 67
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 3.526315789473684,
|
| 1011 |
+
"eval_loss": 0.5140640139579773,
|
| 1012 |
+
"eval_runtime": 0.8962,
|
| 1013 |
+
"eval_samples_per_second": 33.474,
|
| 1014 |
+
"eval_steps_per_second": 4.463,
|
| 1015 |
+
"step": 67
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 3.5789473684210527,
|
| 1019 |
+
"grad_norm": 1.758971929550171,
|
| 1020 |
+
"learning_rate": 0.0004419876590383554,
|
| 1021 |
+
"loss": 0.6927,
|
| 1022 |
+
"step": 68
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 3.5789473684210527,
|
| 1026 |
+
"eval_loss": 0.47072505950927734,
|
| 1027 |
+
"eval_runtime": 0.9127,
|
| 1028 |
+
"eval_samples_per_second": 32.87,
|
| 1029 |
+
"eval_steps_per_second": 4.383,
|
| 1030 |
+
"step": 68
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 3.6315789473684212,
|
| 1034 |
+
"grad_norm": 1.5274709463119507,
|
| 1035 |
+
"learning_rate": 0.00043982990945145146,
|
| 1036 |
+
"loss": 0.4762,
|
| 1037 |
+
"step": 69
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 3.6315789473684212,
|
| 1041 |
+
"eval_loss": 0.4518219828605652,
|
| 1042 |
+
"eval_runtime": 0.8967,
|
| 1043 |
+
"eval_samples_per_second": 33.456,
|
| 1044 |
+
"eval_steps_per_second": 4.461,
|
| 1045 |
+
"step": 69
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 3.6842105263157894,
|
| 1049 |
+
"grad_norm": 1.7685797214508057,
|
| 1050 |
+
"learning_rate": 0.0004376382346819819,
|
| 1051 |
+
"loss": 0.5629,
|
| 1052 |
+
"step": 70
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 3.6842105263157894,
|
| 1056 |
+
"eval_loss": 0.40707579255104065,
|
| 1057 |
+
"eval_runtime": 0.8934,
|
| 1058 |
+
"eval_samples_per_second": 33.581,
|
| 1059 |
+
"eval_steps_per_second": 4.478,
|
| 1060 |
+
"step": 70
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 3.736842105263158,
|
| 1064 |
+
"grad_norm": 1.6618574857711792,
|
| 1065 |
+
"learning_rate": 0.00043541302641198946,
|
| 1066 |
+
"loss": 0.5877,
|
| 1067 |
+
"step": 71
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 3.736842105263158,
|
| 1071 |
+
"eval_loss": 0.3780651390552521,
|
| 1072 |
+
"eval_runtime": 0.9024,
|
| 1073 |
+
"eval_samples_per_second": 33.246,
|
| 1074 |
+
"eval_steps_per_second": 4.433,
|
| 1075 |
+
"step": 71
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 3.7894736842105265,
|
| 1079 |
+
"grad_norm": 1.542702317237854,
|
| 1080 |
+
"learning_rate": 0.00043315468231640834,
|
| 1081 |
+
"loss": 0.5222,
|
| 1082 |
+
"step": 72
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 3.7894736842105265,
|
| 1086 |
+
"eval_loss": 0.3732970356941223,
|
| 1087 |
+
"eval_runtime": 0.9166,
|
| 1088 |
+
"eval_samples_per_second": 32.73,
|
| 1089 |
+
"eval_steps_per_second": 4.364,
|
| 1090 |
+
"step": 72
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 3.8421052631578947,
|
| 1094 |
+
"grad_norm": 1.8039391040802002,
|
| 1095 |
+
"learning_rate": 0.00043086360599199516,
|
| 1096 |
+
"loss": 0.5238,
|
| 1097 |
+
"step": 73
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 3.8421052631578947,
|
| 1101 |
+
"eval_loss": 0.3568810820579529,
|
| 1102 |
+
"eval_runtime": 0.9031,
|
| 1103 |
+
"eval_samples_per_second": 33.218,
|
| 1104 |
+
"eval_steps_per_second": 4.429,
|
| 1105 |
+
"step": 73
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 3.8947368421052633,
|
| 1109 |
+
"grad_norm": 1.6215863227844238,
|
| 1110 |
+
"learning_rate": 0.0004285402068852002,
|
| 1111 |
+
"loss": 0.6504,
|
| 1112 |
+
"step": 74
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 3.8947368421052633,
|
| 1116 |
+
"eval_loss": 0.3885921835899353,
|
| 1117 |
+
"eval_runtime": 0.896,
|
| 1118 |
+
"eval_samples_per_second": 33.483,
|
| 1119 |
+
"eval_steps_per_second": 4.464,
|
| 1120 |
+
"step": 74
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 3.9473684210526314,
|
| 1124 |
+
"grad_norm": 1.5152952671051025,
|
| 1125 |
+
"learning_rate": 0.00042618490021899383,
|
| 1126 |
+
"loss": 0.5694,
|
| 1127 |
+
"step": 75
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 3.9473684210526314,
|
| 1131 |
+
"eval_loss": 0.38745489716529846,
|
| 1132 |
+
"eval_runtime": 0.8939,
|
| 1133 |
+
"eval_samples_per_second": 33.562,
|
| 1134 |
+
"eval_steps_per_second": 4.475,
|
| 1135 |
+
"step": 75
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 4.0,
|
| 1139 |
+
"grad_norm": 2.6989200115203857,
|
| 1140 |
+
"learning_rate": 0.00042379810691866064,
|
| 1141 |
+
"loss": 0.5849,
|
| 1142 |
+
"step": 76
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 4.0,
|
| 1146 |
+
"eval_loss": 0.42535698413848877,
|
| 1147 |
+
"eval_runtime": 0.9073,
|
| 1148 |
+
"eval_samples_per_second": 33.066,
|
| 1149 |
+
"eval_steps_per_second": 4.409,
|
| 1150 |
+
"step": 76
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 4.052631578947368,
|
| 1154 |
+
"grad_norm": 1.7381691932678223,
|
| 1155 |
+
"learning_rate": 0.00042138025353657407,
|
| 1156 |
+
"loss": 0.3779,
|
| 1157 |
+
"step": 77
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 4.052631578947368,
|
| 1161 |
+
"eval_loss": 0.37115439772605896,
|
| 1162 |
+
"eval_runtime": 0.9112,
|
| 1163 |
+
"eval_samples_per_second": 32.922,
|
| 1164 |
+
"eval_steps_per_second": 4.39,
|
| 1165 |
+
"step": 77
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 4.105263157894737,
|
| 1169 |
+
"grad_norm": 2.188385248184204,
|
| 1170 |
+
"learning_rate": 0.00041893177217596633,
|
| 1171 |
+
"loss": 0.44,
|
| 1172 |
+
"step": 78
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 4.105263157894737,
|
| 1176 |
+
"eval_loss": 0.2926563322544098,
|
| 1177 |
+
"eval_runtime": 0.8982,
|
| 1178 |
+
"eval_samples_per_second": 33.401,
|
| 1179 |
+
"eval_steps_per_second": 4.453,
|
| 1180 |
+
"step": 78
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 4.157894736842105,
|
| 1184 |
+
"grad_norm": 2.3652961254119873,
|
| 1185 |
+
"learning_rate": 0.0004164531004137049,
|
| 1186 |
+
"loss": 0.3639,
|
| 1187 |
+
"step": 79
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 4.157894736842105,
|
| 1191 |
+
"eval_loss": 0.2751067876815796,
|
| 1192 |
+
"eval_runtime": 0.9146,
|
| 1193 |
+
"eval_samples_per_second": 32.8,
|
| 1194 |
+
"eval_steps_per_second": 4.373,
|
| 1195 |
+
"step": 79
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 4.2105263157894735,
|
| 1199 |
+
"grad_norm": 2.165874719619751,
|
| 1200 |
+
"learning_rate": 0.0004139446812220924,
|
| 1201 |
+
"loss": 0.2683,
|
| 1202 |
+
"step": 80
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 4.2105263157894735,
|
| 1206 |
+
"eval_loss": 0.2685202360153198,
|
| 1207 |
+
"eval_runtime": 0.9124,
|
| 1208 |
+
"eval_samples_per_second": 32.881,
|
| 1209 |
+
"eval_steps_per_second": 4.384,
|
| 1210 |
+
"step": 80
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 4.2631578947368425,
|
| 1214 |
+
"grad_norm": 1.7391912937164307,
|
| 1215 |
+
"learning_rate": 0.0004114069628897006,
|
| 1216 |
+
"loss": 0.2993,
|
| 1217 |
+
"step": 81
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 4.2631578947368425,
|
| 1221 |
+
"eval_loss": 0.33646491169929504,
|
| 1222 |
+
"eval_runtime": 0.8952,
|
| 1223 |
+
"eval_samples_per_second": 33.51,
|
| 1224 |
+
"eval_steps_per_second": 4.468,
|
| 1225 |
+
"step": 81
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 4.315789473684211,
|
| 1229 |
+
"grad_norm": 3.65714693069458,
|
| 1230 |
+
"learning_rate": 0.0004088403989412559,
|
| 1231 |
+
"loss": 0.4252,
|
| 1232 |
+
"step": 82
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 4.315789473684211,
|
| 1236 |
+
"eval_loss": 0.2839888632297516,
|
| 1237 |
+
"eval_runtime": 0.9057,
|
| 1238 |
+
"eval_samples_per_second": 33.123,
|
| 1239 |
+
"eval_steps_per_second": 4.416,
|
| 1240 |
+
"step": 82
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 4.368421052631579,
|
| 1244 |
+
"grad_norm": 2.1762771606445312,
|
| 1245 |
+
"learning_rate": 0.00040624544805658794,
|
| 1246 |
+
"loss": 0.3304,
|
| 1247 |
+
"step": 83
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 4.368421052631579,
|
| 1251 |
+
"eval_loss": 0.27002134919166565,
|
| 1252 |
+
"eval_runtime": 0.8939,
|
| 1253 |
+
"eval_samples_per_second": 33.562,
|
| 1254 |
+
"eval_steps_per_second": 4.475,
|
| 1255 |
+
"step": 83
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 4.421052631578947,
|
| 1259 |
+
"grad_norm": 2.1018354892730713,
|
| 1260 |
+
"learning_rate": 0.00040362257398865713,
|
| 1261 |
+
"loss": 0.4506,
|
| 1262 |
+
"step": 84
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 4.421052631578947,
|
| 1266 |
+
"eval_loss": 0.2557659149169922,
|
| 1267 |
+
"eval_runtime": 0.8969,
|
| 1268 |
+
"eval_samples_per_second": 33.45,
|
| 1269 |
+
"eval_steps_per_second": 4.46,
|
| 1270 |
+
"step": 84
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 4.473684210526316,
|
| 1274 |
+
"grad_norm": 1.7509180307388306,
|
| 1275 |
+
"learning_rate": 0.00040097224548067613,
|
| 1276 |
+
"loss": 0.3731,
|
| 1277 |
+
"step": 85
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 4.473684210526316,
|
| 1281 |
+
"eval_loss": 0.26859304308891296,
|
| 1282 |
+
"eval_runtime": 0.9009,
|
| 1283 |
+
"eval_samples_per_second": 33.299,
|
| 1284 |
+
"eval_steps_per_second": 4.44,
|
| 1285 |
+
"step": 85
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 4.526315789473684,
|
| 1289 |
+
"grad_norm": 1.971816897392273,
|
| 1290 |
+
"learning_rate": 0.0003982949361823388,
|
| 1291 |
+
"loss": 0.38,
|
| 1292 |
+
"step": 86
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 4.526315789473684,
|
| 1296 |
+
"eval_loss": 0.2624681293964386,
|
| 1297 |
+
"eval_runtime": 0.8949,
|
| 1298 |
+
"eval_samples_per_second": 33.524,
|
| 1299 |
+
"eval_steps_per_second": 4.47,
|
| 1300 |
+
"step": 86
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 4.578947368421053,
|
| 1304 |
+
"grad_norm": 1.4714068174362183,
|
| 1305 |
+
"learning_rate": 0.0003955911245651726,
|
| 1306 |
+
"loss": 0.3944,
|
| 1307 |
+
"step": 87
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 4.578947368421053,
|
| 1311 |
+
"eval_loss": 0.23652420938014984,
|
| 1312 |
+
"eval_runtime": 0.8952,
|
| 1313 |
+
"eval_samples_per_second": 33.511,
|
| 1314 |
+
"eval_steps_per_second": 4.468,
|
| 1315 |
+
"step": 87
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 4.631578947368421,
|
| 1319 |
+
"grad_norm": 2.6970834732055664,
|
| 1320 |
+
"learning_rate": 0.0003928612938370292,
|
| 1321 |
+
"loss": 0.3374,
|
| 1322 |
+
"step": 88
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 4.631578947368421,
|
| 1326 |
+
"eval_loss": 0.2716277241706848,
|
| 1327 |
+
"eval_runtime": 0.8932,
|
| 1328 |
+
"eval_samples_per_second": 33.588,
|
| 1329 |
+
"eval_steps_per_second": 4.478,
|
| 1330 |
+
"step": 88
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 4.684210526315789,
|
| 1334 |
+
"grad_norm": 1.9066615104675293,
|
| 1335 |
+
"learning_rate": 0.00039010593185572867,
|
| 1336 |
+
"loss": 0.2442,
|
| 1337 |
+
"step": 89
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 4.684210526315789,
|
| 1341 |
+
"eval_loss": 0.2999991476535797,
|
| 1342 |
+
"eval_runtime": 0.8939,
|
| 1343 |
+
"eval_samples_per_second": 33.559,
|
| 1344 |
+
"eval_steps_per_second": 4.475,
|
| 1345 |
+
"step": 89
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 4.7368421052631575,
|
| 1349 |
+
"grad_norm": 2.6232354640960693,
|
| 1350 |
+
"learning_rate": 0.00038732553104187296,
|
| 1351 |
+
"loss": 0.2857,
|
| 1352 |
+
"step": 90
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 4.7368421052631575,
|
| 1356 |
+
"eval_loss": 0.2302989959716797,
|
| 1357 |
+
"eval_runtime": 0.8938,
|
| 1358 |
+
"eval_samples_per_second": 33.564,
|
| 1359 |
+
"eval_steps_per_second": 4.475,
|
| 1360 |
+
"step": 90
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 4.7894736842105265,
|
| 1364 |
+
"grad_norm": 2.0710129737854004,
|
| 1365 |
+
"learning_rate": 0.0003845205882908432,
|
| 1366 |
+
"loss": 0.4195,
|
| 1367 |
+
"step": 91
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 4.7894736842105265,
|
| 1371 |
+
"eval_loss": 0.21816590428352356,
|
| 1372 |
+
"eval_runtime": 0.9251,
|
| 1373 |
+
"eval_samples_per_second": 32.429,
|
| 1374 |
+
"eval_steps_per_second": 4.324,
|
| 1375 |
+
"step": 91
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 4.842105263157895,
|
| 1379 |
+
"grad_norm": 1.8006062507629395,
|
| 1380 |
+
"learning_rate": 0.0003816916048839979,
|
| 1381 |
+
"loss": 0.2859,
|
| 1382 |
+
"step": 92
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 4.842105263157895,
|
| 1386 |
+
"eval_loss": 0.21071405708789825,
|
| 1387 |
+
"eval_runtime": 0.8965,
|
| 1388 |
+
"eval_samples_per_second": 33.462,
|
| 1389 |
+
"eval_steps_per_second": 4.462,
|
| 1390 |
+
"step": 92
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 4.894736842105263,
|
| 1394 |
+
"grad_norm": 1.6352888345718384,
|
| 1395 |
+
"learning_rate": 0.0003788390863990875,
|
| 1396 |
+
"loss": 0.4275,
|
| 1397 |
+
"step": 93
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 4.894736842105263,
|
| 1401 |
+
"eval_loss": 0.20206846296787262,
|
| 1402 |
+
"eval_runtime": 0.9052,
|
| 1403 |
+
"eval_samples_per_second": 33.144,
|
| 1404 |
+
"eval_steps_per_second": 4.419,
|
| 1405 |
+
"step": 93
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 4.947368421052632,
|
| 1409 |
+
"grad_norm": 1.6399378776550293,
|
| 1410 |
+
"learning_rate": 0.00037596354261990007,
|
| 1411 |
+
"loss": 0.389,
|
| 1412 |
+
"step": 94
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 4.947368421052632,
|
| 1416 |
+
"eval_loss": 0.19467315077781677,
|
| 1417 |
+
"eval_runtime": 0.8973,
|
| 1418 |
+
"eval_samples_per_second": 33.435,
|
| 1419 |
+
"eval_steps_per_second": 4.458,
|
| 1420 |
+
"step": 94
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 5.0,
|
| 1424 |
+
"grad_norm": 1.5680173635482788,
|
| 1425 |
+
"learning_rate": 0.0003730654874451569,
|
| 1426 |
+
"loss": 0.395,
|
| 1427 |
+
"step": 95
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 5.0,
|
| 1431 |
+
"eval_loss": 0.19546455144882202,
|
| 1432 |
+
"eval_runtime": 0.91,
|
| 1433 |
+
"eval_samples_per_second": 32.968,
|
| 1434 |
+
"eval_steps_per_second": 4.396,
|
| 1435 |
+
"step": 95
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 5.052631578947368,
|
| 1439 |
+
"grad_norm": 1.0308386087417603,
|
| 1440 |
+
"learning_rate": 0.00037014543879667093,
|
| 1441 |
+
"loss": 0.1384,
|
| 1442 |
+
"step": 96
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 5.052631578947368,
|
| 1446 |
+
"eval_loss": 0.18969732522964478,
|
| 1447 |
+
"eval_runtime": 0.9021,
|
| 1448 |
+
"eval_samples_per_second": 33.258,
|
| 1449 |
+
"eval_steps_per_second": 4.434,
|
| 1450 |
+
"step": 96
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 5.105263157894737,
|
| 1454 |
+
"grad_norm": 1.4042502641677856,
|
| 1455 |
+
"learning_rate": 0.0003672039185267878,
|
| 1456 |
+
"loss": 0.2291,
|
| 1457 |
+
"step": 97
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 5.105263157894737,
|
| 1461 |
+
"eval_loss": 0.16800740361213684,
|
| 1462 |
+
"eval_runtime": 0.8938,
|
| 1463 |
+
"eval_samples_per_second": 33.563,
|
| 1464 |
+
"eval_steps_per_second": 4.475,
|
| 1465 |
+
"step": 97
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 5.157894736842105,
|
| 1469 |
+
"grad_norm": 1.6313552856445312,
|
| 1470 |
+
"learning_rate": 0.00036424145232512333,
|
| 1471 |
+
"loss": 0.1736,
|
| 1472 |
+
"step": 98
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 5.157894736842105,
|
| 1476 |
+
"eval_loss": 0.16714099049568176,
|
| 1477 |
+
"eval_runtime": 0.9009,
|
| 1478 |
+
"eval_samples_per_second": 33.301,
|
| 1479 |
+
"eval_steps_per_second": 4.44,
|
| 1480 |
+
"step": 98
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 5.2105263157894735,
|
| 1484 |
+
"grad_norm": 1.8922698497772217,
|
| 1485 |
+
"learning_rate": 0.0003612585696246158,
|
| 1486 |
+
"loss": 0.1677,
|
| 1487 |
+
"step": 99
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 5.2105263157894735,
|
| 1491 |
+
"eval_loss": 0.179762065410614,
|
| 1492 |
+
"eval_runtime": 0.9039,
|
| 1493 |
+
"eval_samples_per_second": 33.188,
|
| 1494 |
+
"eval_steps_per_second": 4.425,
|
| 1495 |
+
"step": 99
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 5.2631578947368425,
|
| 1499 |
+
"grad_norm": 2.409526824951172,
|
| 1500 |
+
"learning_rate": 0.0003582558035069091,
|
| 1501 |
+
"loss": 0.2379,
|
| 1502 |
+
"step": 100
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 5.2631578947368425,
|
| 1506 |
+
"eval_loss": 0.1902371197938919,
|
| 1507 |
+
"eval_runtime": 0.9097,
|
| 1508 |
+
"eval_samples_per_second": 32.98,
|
| 1509 |
+
"eval_steps_per_second": 4.397,
|
| 1510 |
+
"step": 100
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 5.315789473684211,
|
| 1514 |
+
"grad_norm": 2.084869146347046,
|
| 1515 |
+
"learning_rate": 0.0003552336906070838,
|
| 1516 |
+
"loss": 0.2165,
|
| 1517 |
+
"step": 101
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 5.315789473684211,
|
| 1521 |
+
"eval_loss": 0.17252177000045776,
|
| 1522 |
+
"eval_runtime": 0.8948,
|
| 1523 |
+
"eval_samples_per_second": 33.528,
|
| 1524 |
+
"eval_steps_per_second": 4.47,
|
| 1525 |
+
"step": 101
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 5.368421052631579,
|
| 1529 |
+
"grad_norm": 1.655718207359314,
|
| 1530 |
+
"learning_rate": 0.000352192771017753,
|
| 1531 |
+
"loss": 0.223,
|
| 1532 |
+
"step": 102
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 5.368421052631579,
|
| 1536 |
+
"eval_loss": 0.18867380917072296,
|
| 1537 |
+
"eval_runtime": 0.8956,
|
| 1538 |
+
"eval_samples_per_second": 33.495,
|
| 1539 |
+
"eval_steps_per_second": 4.466,
|
| 1540 |
+
"step": 102
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 5.421052631578947,
|
| 1544 |
+
"grad_norm": 2.672633409500122,
|
| 1545 |
+
"learning_rate": 0.0003491335881925407,
|
| 1546 |
+
"loss": 0.161,
|
| 1547 |
+
"step": 103
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 5.421052631578947,
|
| 1551 |
+
"eval_loss": 0.1944020837545395,
|
| 1552 |
+
"eval_runtime": 0.8924,
|
| 1553 |
+
"eval_samples_per_second": 33.616,
|
| 1554 |
+
"eval_steps_per_second": 4.482,
|
| 1555 |
+
"step": 103
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 5.473684210526316,
|
| 1559 |
+
"grad_norm": 1.9712008237838745,
|
| 1560 |
+
"learning_rate": 0.0003460566888489593,
|
| 1561 |
+
"loss": 0.2525,
|
| 1562 |
+
"step": 104
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 5.473684210526316,
|
| 1566 |
+
"eval_loss": 0.17671068012714386,
|
| 1567 |
+
"eval_runtime": 0.897,
|
| 1568 |
+
"eval_samples_per_second": 33.446,
|
| 1569 |
+
"eval_steps_per_second": 4.459,
|
| 1570 |
+
"step": 104
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 5.526315789473684,
|
| 1574 |
+
"grad_norm": 2.2153072357177734,
|
| 1575 |
+
"learning_rate": 0.00034296262287070335,
|
| 1576 |
+
"loss": 0.2105,
|
| 1577 |
+
"step": 105
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 5.526315789473684,
|
| 1581 |
+
"eval_loss": 0.1715732216835022,
|
| 1582 |
+
"eval_runtime": 0.8951,
|
| 1583 |
+
"eval_samples_per_second": 33.514,
|
| 1584 |
+
"eval_steps_per_second": 4.469,
|
| 1585 |
+
"step": 105
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 5.578947368421053,
|
| 1589 |
+
"grad_norm": 1.8106168508529663,
|
| 1590 |
+
"learning_rate": 0.0003398519432093782,
|
| 1591 |
+
"loss": 0.259,
|
| 1592 |
+
"step": 106
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 5.578947368421053,
|
| 1596 |
+
"eval_loss": 0.1465868353843689,
|
| 1597 |
+
"eval_runtime": 0.9077,
|
| 1598 |
+
"eval_samples_per_second": 33.051,
|
| 1599 |
+
"eval_steps_per_second": 4.407,
|
| 1600 |
+
"step": 106
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 5.631578947368421,
|
| 1604 |
+
"grad_norm": 2.1159439086914062,
|
| 1605 |
+
"learning_rate": 0.0003367252057856802,
|
| 1606 |
+
"loss": 0.2065,
|
| 1607 |
+
"step": 107
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 5.631578947368421,
|
| 1611 |
+
"eval_loss": 0.14219093322753906,
|
| 1612 |
+
"eval_runtime": 0.9049,
|
| 1613 |
+
"eval_samples_per_second": 33.154,
|
| 1614 |
+
"eval_steps_per_second": 4.42,
|
| 1615 |
+
"step": 107
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 5.684210526315789,
|
| 1619 |
+
"grad_norm": 1.4467761516571045,
|
| 1620 |
+
"learning_rate": 0.00033358296939004547,
|
| 1621 |
+
"loss": 0.2083,
|
| 1622 |
+
"step": 108
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 5.684210526315789,
|
| 1626 |
+
"eval_loss": 0.1406753957271576,
|
| 1627 |
+
"eval_runtime": 0.8954,
|
| 1628 |
+
"eval_samples_per_second": 33.505,
|
| 1629 |
+
"eval_steps_per_second": 4.467,
|
| 1630 |
+
"step": 108
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 5.7368421052631575,
|
| 1634 |
+
"grad_norm": 1.3671239614486694,
|
| 1635 |
+
"learning_rate": 0.00033042579558278717,
|
| 1636 |
+
"loss": 0.1825,
|
| 1637 |
+
"step": 109
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 5.7368421052631575,
|
| 1641 |
+
"eval_loss": 0.13007155060768127,
|
| 1642 |
+
"eval_runtime": 0.8998,
|
| 1643 |
+
"eval_samples_per_second": 33.342,
|
| 1644 |
+
"eval_steps_per_second": 4.446,
|
| 1645 |
+
"step": 109
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 5.7894736842105265,
|
| 1649 |
+
"grad_norm": 1.479944109916687,
|
| 1650 |
+
"learning_rate": 0.00032725424859373687,
|
| 1651 |
+
"loss": 0.2244,
|
| 1652 |
+
"step": 110
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 5.7894736842105265,
|
| 1656 |
+
"eval_loss": 0.12692232429981232,
|
| 1657 |
+
"eval_runtime": 0.901,
|
| 1658 |
+
"eval_samples_per_second": 33.298,
|
| 1659 |
+
"eval_steps_per_second": 4.44,
|
| 1660 |
+
"step": 110
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 5.842105263157895,
|
| 1664 |
+
"grad_norm": 1.5173969268798828,
|
| 1665 |
+
"learning_rate": 0.0003240688952214085,
|
| 1666 |
+
"loss": 0.2273,
|
| 1667 |
+
"step": 111
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 5.842105263157895,
|
| 1671 |
+
"eval_loss": 0.12454597651958466,
|
| 1672 |
+
"eval_runtime": 0.8987,
|
| 1673 |
+
"eval_samples_per_second": 33.382,
|
| 1674 |
+
"eval_steps_per_second": 4.451,
|
| 1675 |
+
"step": 111
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 5.894736842105263,
|
| 1679 |
+
"grad_norm": 2.7870988845825195,
|
| 1680 |
+
"learning_rate": 0.00032087030473170445,
|
| 1681 |
+
"loss": 0.2101,
|
| 1682 |
+
"step": 112
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 5.894736842105263,
|
| 1686 |
+
"eval_loss": 0.12002909928560257,
|
| 1687 |
+
"eval_runtime": 0.893,
|
| 1688 |
+
"eval_samples_per_second": 33.593,
|
| 1689 |
+
"eval_steps_per_second": 4.479,
|
| 1690 |
+
"step": 112
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 5.947368421052632,
|
| 1694 |
+
"grad_norm": 1.3659342527389526,
|
| 1695 |
+
"learning_rate": 0.00031765904875617973,
|
| 1696 |
+
"loss": 0.1882,
|
| 1697 |
+
"step": 113
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 5.947368421052632,
|
| 1701 |
+
"eval_loss": 0.10573837906122208,
|
| 1702 |
+
"eval_runtime": 0.8956,
|
| 1703 |
+
"eval_samples_per_second": 33.496,
|
| 1704 |
+
"eval_steps_per_second": 4.466,
|
| 1705 |
+
"step": 113
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 6.0,
|
| 1709 |
+
"grad_norm": 1.8464044332504272,
|
| 1710 |
+
"learning_rate": 0.00031443570118988356,
|
| 1711 |
+
"loss": 0.2285,
|
| 1712 |
+
"step": 114
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 6.0,
|
| 1716 |
+
"eval_loss": 0.10221625119447708,
|
| 1717 |
+
"eval_runtime": 0.8955,
|
| 1718 |
+
"eval_samples_per_second": 33.501,
|
| 1719 |
+
"eval_steps_per_second": 4.467,
|
| 1720 |
+
"step": 114
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 6.052631578947368,
|
| 1724 |
+
"grad_norm": 1.3894392251968384,
|
| 1725 |
+
"learning_rate": 0.00031120083808879663,
|
| 1726 |
+
"loss": 0.1115,
|
| 1727 |
+
"step": 115
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 6.052631578947368,
|
| 1731 |
+
"eval_loss": 0.09458151459693909,
|
| 1732 |
+
"eval_runtime": 0.8981,
|
| 1733 |
+
"eval_samples_per_second": 33.405,
|
| 1734 |
+
"eval_steps_per_second": 4.454,
|
| 1735 |
+
"step": 115
|
| 1736 |
}
|
| 1737 |
],
|
| 1738 |
"logging_steps": 1,
|
|
|
|
| 1752 |
"attributes": {}
|
| 1753 |
}
|
| 1754 |
},
|
| 1755 |
+
"total_flos": 4901149662148608.0,
|
| 1756 |
"train_batch_size": 1,
|
| 1757 |
"trial_name": null,
|
| 1758 |
"trial_params": null
|