Training in progress, step 400, checkpoint
Browse files- last-checkpoint/adapter_model.safetensors +1 -1
- last-checkpoint/global_step400/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step400/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step400/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step400/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step400/zero_pp_rank_0_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step400/zero_pp_rank_1_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step400/zero_pp_rank_2_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step400/zero_pp_rank_3_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/latest +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +739 -3
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13648688
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:58970c7b1499a51e52f13d34fc84f1be155b729c5dc01535b4fd3b471893cc7c
|
| 3 |
size 13648688
|
last-checkpoint/global_step400/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b15663e9002bfadcd2118544805c26092cb6ec0f666d9bc6587d6401abd21bd5
|
| 3 |
+
size 20450800
|
last-checkpoint/global_step400/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ef866156013d48c080d2301148283ca8543927968bc28db94f0b4473bdc8acc6
|
| 3 |
+
size 20450800
|
last-checkpoint/global_step400/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad9b9bb4dfcce0be4cee272860f5b8be7d9a8be61239e5282130fe726f4adae2
|
| 3 |
+
size 20450800
|
last-checkpoint/global_step400/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b5910c9d8c18991184c3eebf90eac60e2ca4a073d9aefb2105dd96e9b2836ac1
|
| 3 |
+
size 20450800
|
last-checkpoint/global_step400/zero_pp_rank_0_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5625e40f0bab8897b6c022d047ec4e1b309f6c5683c29a5d18f3f6b49d2fc2f
|
| 3 |
+
size 152238
|
last-checkpoint/global_step400/zero_pp_rank_1_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1e31e7163b707480ad478f8e6c8378875c9442dd275ed39a38db3f4141626ed
|
| 3 |
+
size 152238
|
last-checkpoint/global_step400/zero_pp_rank_2_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0e01df639f8ed07332533c55b3ba24300b71cc1b94368675bb8537e39c0bcbf7
|
| 3 |
+
size 152238
|
last-checkpoint/global_step400/zero_pp_rank_3_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3c88b6a29d249139ed50a0ee329aa72d539d54ce570568131b5fd790a11a890
|
| 3 |
+
size 152238
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step400
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:abaf0e70a6661521eb40188cdadbb09fcda9f6e1ac539eef99db2b1bc5a7ba52
|
| 3 |
size 14960
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:376c466d70aa79c6b0bb9fc6cc87d2e449a16493d5d1155107e37872dcdc22dc
|
| 3 |
size 14960
|
last-checkpoint/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:865c5a6a69a0b6acfd26560edcb10f0694871429483ae64bee81aba12e73a0b0
|
| 3 |
size 14960
|
last-checkpoint/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8848755180edfd38eee9486edbe1a58572435a9c200f8a462726bb43540dcbf5
|
| 3 |
size 14960
|
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:b217a5016949cb5cd10bb5c4b090e10b845f27963a84cf5bdc1f1d94facb5b3c
|
| 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": 40,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -2203,6 +2203,742 @@
|
|
| 2203 |
"learning_rate": 1.5922960037532057e-05,
|
| 2204 |
"loss": 0.281,
|
| 2205 |
"step": 300
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2206 |
}
|
| 2207 |
],
|
| 2208 |
"logging_steps": 1,
|
|
@@ -2222,7 +2958,7 @@
|
|
| 2222 |
"attributes": {}
|
| 2223 |
}
|
| 2224 |
},
|
| 2225 |
-
"total_flos":
|
| 2226 |
"train_batch_size": 4,
|
| 2227 |
"trial_name": null,
|
| 2228 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.49185367353212417,
|
| 5 |
"eval_steps": 40,
|
| 6 |
+
"global_step": 400,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 2203 |
"learning_rate": 1.5922960037532057e-05,
|
| 2204 |
"loss": 0.281,
|
| 2205 |
"step": 300
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 0.37011988933292345,
|
| 2209 |
+
"grad_norm": 0.36134533123850165,
|
| 2210 |
+
"learning_rate": 1.588827828521133e-05,
|
| 2211 |
+
"loss": 0.1997,
|
| 2212 |
+
"step": 301
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 0.37134952351675377,
|
| 2216 |
+
"grad_norm": 0.40175135868371165,
|
| 2217 |
+
"learning_rate": 1.585348777703486e-05,
|
| 2218 |
+
"loss": 0.2047,
|
| 2219 |
+
"step": 302
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 0.3725791577005841,
|
| 2223 |
+
"grad_norm": 0.3610033083703328,
|
| 2224 |
+
"learning_rate": 1.581858915557953e-05,
|
| 2225 |
+
"loss": 0.2037,
|
| 2226 |
+
"step": 303
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 0.3738087918844144,
|
| 2230 |
+
"grad_norm": 0.42325801209258107,
|
| 2231 |
+
"learning_rate": 1.5783583065419054e-05,
|
| 2232 |
+
"loss": 0.1871,
|
| 2233 |
+
"step": 304
|
| 2234 |
+
},
|
| 2235 |
+
{
|
| 2236 |
+
"epoch": 0.3750384260682447,
|
| 2237 |
+
"grad_norm": 0.28821395654955023,
|
| 2238 |
+
"learning_rate": 1.5748470153112093e-05,
|
| 2239 |
+
"loss": 0.1934,
|
| 2240 |
+
"step": 305
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 0.376268060252075,
|
| 2244 |
+
"grad_norm": 0.4052634070520442,
|
| 2245 |
+
"learning_rate": 1.57132510671903e-05,
|
| 2246 |
+
"loss": 0.2351,
|
| 2247 |
+
"step": 306
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 0.3774976944359053,
|
| 2251 |
+
"grad_norm": 0.48496008691724585,
|
| 2252 |
+
"learning_rate": 1.5677926458146327e-05,
|
| 2253 |
+
"loss": 0.2527,
|
| 2254 |
+
"step": 307
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 0.3787273286197356,
|
| 2258 |
+
"grad_norm": 0.4026838585087856,
|
| 2259 |
+
"learning_rate": 1.5642496978421842e-05,
|
| 2260 |
+
"loss": 0.2554,
|
| 2261 |
+
"step": 308
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 0.37995696280356595,
|
| 2265 |
+
"grad_norm": 0.34534876497819145,
|
| 2266 |
+
"learning_rate": 1.560696328239547e-05,
|
| 2267 |
+
"loss": 0.1656,
|
| 2268 |
+
"step": 309
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 0.38118659698739626,
|
| 2272 |
+
"grad_norm": 0.2751030687579328,
|
| 2273 |
+
"learning_rate": 1.5571326026370676e-05,
|
| 2274 |
+
"loss": 0.162,
|
| 2275 |
+
"step": 310
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 0.3824162311712266,
|
| 2279 |
+
"grad_norm": 0.38621022206094485,
|
| 2280 |
+
"learning_rate": 1.5535585868563688e-05,
|
| 2281 |
+
"loss": 0.2212,
|
| 2282 |
+
"step": 311
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 0.38364586535505685,
|
| 2286 |
+
"grad_norm": 0.3039979678378701,
|
| 2287 |
+
"learning_rate": 1.5499743469091303e-05,
|
| 2288 |
+
"loss": 0.2413,
|
| 2289 |
+
"step": 312
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 0.38487549953888717,
|
| 2293 |
+
"grad_norm": 0.31620852109871317,
|
| 2294 |
+
"learning_rate": 1.5463799489958727e-05,
|
| 2295 |
+
"loss": 0.1701,
|
| 2296 |
+
"step": 313
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 0.3861051337227175,
|
| 2300 |
+
"grad_norm": 0.3036007341488487,
|
| 2301 |
+
"learning_rate": 1.542775459504732e-05,
|
| 2302 |
+
"loss": 0.1718,
|
| 2303 |
+
"step": 314
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 0.3873347679065478,
|
| 2307 |
+
"grad_norm": 0.4288199829474487,
|
| 2308 |
+
"learning_rate": 1.5391609450102346e-05,
|
| 2309 |
+
"loss": 0.2237,
|
| 2310 |
+
"step": 315
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 0.3885644020903781,
|
| 2314 |
+
"grad_norm": 0.3798815943379263,
|
| 2315 |
+
"learning_rate": 1.5355364722720674e-05,
|
| 2316 |
+
"loss": 0.2177,
|
| 2317 |
+
"step": 316
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 0.38979403627420844,
|
| 2321 |
+
"grad_norm": 0.30645321053750946,
|
| 2322 |
+
"learning_rate": 1.5319021082338458e-05,
|
| 2323 |
+
"loss": 0.2356,
|
| 2324 |
+
"step": 317
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 0.3910236704580387,
|
| 2328 |
+
"grad_norm": 0.3377360946626629,
|
| 2329 |
+
"learning_rate": 1.5282579200218762e-05,
|
| 2330 |
+
"loss": 0.21,
|
| 2331 |
+
"step": 318
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 0.392253304641869,
|
| 2335 |
+
"grad_norm": 0.3645715594028636,
|
| 2336 |
+
"learning_rate": 1.5246039749439159e-05,
|
| 2337 |
+
"loss": 0.1733,
|
| 2338 |
+
"step": 319
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 0.39348293882569935,
|
| 2342 |
+
"grad_norm": 0.29596291916382467,
|
| 2343 |
+
"learning_rate": 1.5209403404879305e-05,
|
| 2344 |
+
"loss": 0.1505,
|
| 2345 |
+
"step": 320
|
| 2346 |
+
},
|
| 2347 |
+
{
|
| 2348 |
+
"epoch": 0.39348293882569935,
|
| 2349 |
+
"eval_accuracy": 0.8181818181818182,
|
| 2350 |
+
"eval_f1": 0.5853658536585366,
|
| 2351 |
+
"eval_loss": 0.40882813930511475,
|
| 2352 |
+
"eval_precision": 0.75,
|
| 2353 |
+
"eval_recall": 0.48,
|
| 2354 |
+
"eval_runtime": 23.2002,
|
| 2355 |
+
"eval_samples_per_second": 2.155,
|
| 2356 |
+
"eval_steps_per_second": 0.172,
|
| 2357 |
+
"step": 320
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 0.39471257300952967,
|
| 2361 |
+
"grad_norm": 0.4078103453153395,
|
| 2362 |
+
"learning_rate": 1.5172670843208477e-05,
|
| 2363 |
+
"loss": 0.2415,
|
| 2364 |
+
"step": 321
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 0.39594220719336,
|
| 2368 |
+
"grad_norm": 0.3270612364847978,
|
| 2369 |
+
"learning_rate": 1.5135842742873077e-05,
|
| 2370 |
+
"loss": 0.2059,
|
| 2371 |
+
"step": 322
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.3971718413771903,
|
| 2375 |
+
"grad_norm": 0.3309279944550533,
|
| 2376 |
+
"learning_rate": 1.5098919784084083e-05,
|
| 2377 |
+
"loss": 0.1569,
|
| 2378 |
+
"step": 323
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 0.3984014755610206,
|
| 2382 |
+
"grad_norm": 0.49875353395381333,
|
| 2383 |
+
"learning_rate": 1.5061902648804503e-05,
|
| 2384 |
+
"loss": 0.2477,
|
| 2385 |
+
"step": 324
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 0.3996311097448509,
|
| 2389 |
+
"grad_norm": 0.3787412609953064,
|
| 2390 |
+
"learning_rate": 1.502479202073678e-05,
|
| 2391 |
+
"loss": 0.179,
|
| 2392 |
+
"step": 325
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 0.4008607439286812,
|
| 2396 |
+
"grad_norm": 0.3919230580283582,
|
| 2397 |
+
"learning_rate": 1.4987588585310154e-05,
|
| 2398 |
+
"loss": 0.2249,
|
| 2399 |
+
"step": 326
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 0.4020903781125115,
|
| 2403 |
+
"grad_norm": 0.40414791953417883,
|
| 2404 |
+
"learning_rate": 1.4950293029668004e-05,
|
| 2405 |
+
"loss": 0.1772,
|
| 2406 |
+
"step": 327
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 0.40332001229634185,
|
| 2410 |
+
"grad_norm": 0.3778228440356831,
|
| 2411 |
+
"learning_rate": 1.4912906042655164e-05,
|
| 2412 |
+
"loss": 0.208,
|
| 2413 |
+
"step": 328
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 0.40454964648017216,
|
| 2417 |
+
"grad_norm": 0.4107607344666035,
|
| 2418 |
+
"learning_rate": 1.4875428314805195e-05,
|
| 2419 |
+
"loss": 0.1716,
|
| 2420 |
+
"step": 329
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 0.4057792806640025,
|
| 2424 |
+
"grad_norm": 0.3671542265326293,
|
| 2425 |
+
"learning_rate": 1.483786053832763e-05,
|
| 2426 |
+
"loss": 0.1805,
|
| 2427 |
+
"step": 330
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 0.40700891484783275,
|
| 2431 |
+
"grad_norm": 0.404957322338137,
|
| 2432 |
+
"learning_rate": 1.4800203407095194e-05,
|
| 2433 |
+
"loss": 0.1842,
|
| 2434 |
+
"step": 331
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 0.40823854903166307,
|
| 2438 |
+
"grad_norm": 0.48189780697782403,
|
| 2439 |
+
"learning_rate": 1.4762457616630972e-05,
|
| 2440 |
+
"loss": 0.2277,
|
| 2441 |
+
"step": 332
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 0.4094681832154934,
|
| 2445 |
+
"grad_norm": 0.30773052153479974,
|
| 2446 |
+
"learning_rate": 1.4724623864095595e-05,
|
| 2447 |
+
"loss": 0.1833,
|
| 2448 |
+
"step": 333
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 0.4106978173993237,
|
| 2452 |
+
"grad_norm": 0.4223729302065043,
|
| 2453 |
+
"learning_rate": 1.4686702848274328e-05,
|
| 2454 |
+
"loss": 0.2219,
|
| 2455 |
+
"step": 334
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 0.411927451583154,
|
| 2459 |
+
"grad_norm": 0.2902399210528883,
|
| 2460 |
+
"learning_rate": 1.4648695269564182e-05,
|
| 2461 |
+
"loss": 0.1785,
|
| 2462 |
+
"step": 335
|
| 2463 |
+
},
|
| 2464 |
+
{
|
| 2465 |
+
"epoch": 0.41315708576698434,
|
| 2466 |
+
"grad_norm": 0.297435572010037,
|
| 2467 |
+
"learning_rate": 1.461060182996098e-05,
|
| 2468 |
+
"loss": 0.2441,
|
| 2469 |
+
"step": 336
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 0.4143867199508146,
|
| 2473 |
+
"grad_norm": 0.4498936613858582,
|
| 2474 |
+
"learning_rate": 1.4572423233046386e-05,
|
| 2475 |
+
"loss": 0.1765,
|
| 2476 |
+
"step": 337
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 0.4156163541346449,
|
| 2480 |
+
"grad_norm": 0.476353253435308,
|
| 2481 |
+
"learning_rate": 1.4534160183974908e-05,
|
| 2482 |
+
"loss": 0.1711,
|
| 2483 |
+
"step": 338
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 0.41684598831847525,
|
| 2487 |
+
"grad_norm": 0.4136384364300188,
|
| 2488 |
+
"learning_rate": 1.4495813389460875e-05,
|
| 2489 |
+
"loss": 0.213,
|
| 2490 |
+
"step": 339
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 0.41807562250230557,
|
| 2494 |
+
"grad_norm": 0.457702192015087,
|
| 2495 |
+
"learning_rate": 1.4457383557765385e-05,
|
| 2496 |
+
"loss": 0.2056,
|
| 2497 |
+
"step": 340
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 0.4193052566861359,
|
| 2501 |
+
"grad_norm": 0.32316243545279294,
|
| 2502 |
+
"learning_rate": 1.4418871398683227e-05,
|
| 2503 |
+
"loss": 0.2445,
|
| 2504 |
+
"step": 341
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 0.4205348908699662,
|
| 2508 |
+
"grad_norm": 0.3952237252439469,
|
| 2509 |
+
"learning_rate": 1.4380277623529766e-05,
|
| 2510 |
+
"loss": 0.1789,
|
| 2511 |
+
"step": 342
|
| 2512 |
+
},
|
| 2513 |
+
{
|
| 2514 |
+
"epoch": 0.42176452505379647,
|
| 2515 |
+
"grad_norm": 0.45979858964325293,
|
| 2516 |
+
"learning_rate": 1.4341602945127806e-05,
|
| 2517 |
+
"loss": 0.225,
|
| 2518 |
+
"step": 343
|
| 2519 |
+
},
|
| 2520 |
+
{
|
| 2521 |
+
"epoch": 0.4229941592376268,
|
| 2522 |
+
"grad_norm": 0.456535265546009,
|
| 2523 |
+
"learning_rate": 1.4302848077794427e-05,
|
| 2524 |
+
"loss": 0.2245,
|
| 2525 |
+
"step": 344
|
| 2526 |
+
},
|
| 2527 |
+
{
|
| 2528 |
+
"epoch": 0.4242237934214571,
|
| 2529 |
+
"grad_norm": 0.3244747375904321,
|
| 2530 |
+
"learning_rate": 1.426401373732779e-05,
|
| 2531 |
+
"loss": 0.1801,
|
| 2532 |
+
"step": 345
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 0.4254534276052874,
|
| 2536 |
+
"grad_norm": 0.40300301462416604,
|
| 2537 |
+
"learning_rate": 1.422510064099391e-05,
|
| 2538 |
+
"loss": 0.2212,
|
| 2539 |
+
"step": 346
|
| 2540 |
+
},
|
| 2541 |
+
{
|
| 2542 |
+
"epoch": 0.42668306178911775,
|
| 2543 |
+
"grad_norm": 0.5264557343629197,
|
| 2544 |
+
"learning_rate": 1.4186109507513425e-05,
|
| 2545 |
+
"loss": 0.2202,
|
| 2546 |
+
"step": 347
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 0.42791269597294807,
|
| 2550 |
+
"grad_norm": 0.4350217170052354,
|
| 2551 |
+
"learning_rate": 1.4147041057048303e-05,
|
| 2552 |
+
"loss": 0.2061,
|
| 2553 |
+
"step": 348
|
| 2554 |
+
},
|
| 2555 |
+
{
|
| 2556 |
+
"epoch": 0.4291423301567784,
|
| 2557 |
+
"grad_norm": 0.6810095448654682,
|
| 2558 |
+
"learning_rate": 1.4107896011188546e-05,
|
| 2559 |
+
"loss": 0.1782,
|
| 2560 |
+
"step": 349
|
| 2561 |
+
},
|
| 2562 |
+
{
|
| 2563 |
+
"epoch": 0.43037196434060865,
|
| 2564 |
+
"grad_norm": 0.25807154769833757,
|
| 2565 |
+
"learning_rate": 1.4068675092938872e-05,
|
| 2566 |
+
"loss": 0.156,
|
| 2567 |
+
"step": 350
|
| 2568 |
+
},
|
| 2569 |
+
{
|
| 2570 |
+
"epoch": 0.43160159852443897,
|
| 2571 |
+
"grad_norm": 0.29519781987132376,
|
| 2572 |
+
"learning_rate": 1.4029379026705352e-05,
|
| 2573 |
+
"loss": 0.2078,
|
| 2574 |
+
"step": 351
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 0.4328312327082693,
|
| 2578 |
+
"grad_norm": 0.4580606963993864,
|
| 2579 |
+
"learning_rate": 1.3990008538282027e-05,
|
| 2580 |
+
"loss": 0.2024,
|
| 2581 |
+
"step": 352
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 0.4340608668920996,
|
| 2585 |
+
"grad_norm": 0.4014942247225734,
|
| 2586 |
+
"learning_rate": 1.3950564354837512e-05,
|
| 2587 |
+
"loss": 0.1801,
|
| 2588 |
+
"step": 353
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"epoch": 0.4352905010759299,
|
| 2592 |
+
"grad_norm": 0.40131332657814234,
|
| 2593 |
+
"learning_rate": 1.391104720490156e-05,
|
| 2594 |
+
"loss": 0.214,
|
| 2595 |
+
"step": 354
|
| 2596 |
+
},
|
| 2597 |
+
{
|
| 2598 |
+
"epoch": 0.43652013525976024,
|
| 2599 |
+
"grad_norm": 0.3560030996515948,
|
| 2600 |
+
"learning_rate": 1.387145781835161e-05,
|
| 2601 |
+
"loss": 0.2126,
|
| 2602 |
+
"step": 355
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"epoch": 0.4377497694435905,
|
| 2606 |
+
"grad_norm": 0.3729681090326996,
|
| 2607 |
+
"learning_rate": 1.3831796926399295e-05,
|
| 2608 |
+
"loss": 0.2055,
|
| 2609 |
+
"step": 356
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 0.43897940362742083,
|
| 2613 |
+
"grad_norm": 0.4172829333159122,
|
| 2614 |
+
"learning_rate": 1.3792065261576953e-05,
|
| 2615 |
+
"loss": 0.2326,
|
| 2616 |
+
"step": 357
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 0.44020903781125115,
|
| 2620 |
+
"grad_norm": 0.30998074509681783,
|
| 2621 |
+
"learning_rate": 1.3752263557724088e-05,
|
| 2622 |
+
"loss": 0.1633,
|
| 2623 |
+
"step": 358
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 0.44143867199508147,
|
| 2627 |
+
"grad_norm": 0.345268920537541,
|
| 2628 |
+
"learning_rate": 1.3712392549973814e-05,
|
| 2629 |
+
"loss": 0.1765,
|
| 2630 |
+
"step": 359
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 0.4426683061789118,
|
| 2634 |
+
"grad_norm": 0.3508481090946454,
|
| 2635 |
+
"learning_rate": 1.3672452974739278e-05,
|
| 2636 |
+
"loss": 0.1752,
|
| 2637 |
+
"step": 360
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 0.4426683061789118,
|
| 2641 |
+
"eval_accuracy": 0.786096256684492,
|
| 2642 |
+
"eval_f1": 0.42857142857142855,
|
| 2643 |
+
"eval_loss": 0.4385937452316284,
|
| 2644 |
+
"eval_precision": 0.75,
|
| 2645 |
+
"eval_recall": 0.3,
|
| 2646 |
+
"eval_runtime": 23.4097,
|
| 2647 |
+
"eval_samples_per_second": 2.136,
|
| 2648 |
+
"eval_steps_per_second": 0.171,
|
| 2649 |
+
"step": 360
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 0.4438979403627421,
|
| 2653 |
+
"grad_norm": 0.41404798323100117,
|
| 2654 |
+
"learning_rate": 1.3632445569700078e-05,
|
| 2655 |
+
"loss": 0.1745,
|
| 2656 |
+
"step": 361
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 0.44512757454657237,
|
| 2660 |
+
"grad_norm": 0.39572877694784087,
|
| 2661 |
+
"learning_rate": 1.3592371073788595e-05,
|
| 2662 |
+
"loss": 0.216,
|
| 2663 |
+
"step": 362
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 0.4463572087304027,
|
| 2667 |
+
"grad_norm": 0.3927819000312662,
|
| 2668 |
+
"learning_rate": 1.355223022717639e-05,
|
| 2669 |
+
"loss": 0.164,
|
| 2670 |
+
"step": 363
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 0.447586842914233,
|
| 2674 |
+
"grad_norm": 0.36355749807156507,
|
| 2675 |
+
"learning_rate": 1.3512023771260507e-05,
|
| 2676 |
+
"loss": 0.2439,
|
| 2677 |
+
"step": 364
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 0.4488164770980633,
|
| 2681 |
+
"grad_norm": 0.37694180103294717,
|
| 2682 |
+
"learning_rate": 1.347175244864979e-05,
|
| 2683 |
+
"loss": 0.2009,
|
| 2684 |
+
"step": 365
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 0.45004611128189365,
|
| 2688 |
+
"grad_norm": 0.3566619922589067,
|
| 2689 |
+
"learning_rate": 1.3431417003151162e-05,
|
| 2690 |
+
"loss": 0.2045,
|
| 2691 |
+
"step": 366
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 0.45127574546572397,
|
| 2695 |
+
"grad_norm": 0.36473757399388024,
|
| 2696 |
+
"learning_rate": 1.3391018179755886e-05,
|
| 2697 |
+
"loss": 0.1711,
|
| 2698 |
+
"step": 367
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 0.4525053796495543,
|
| 2702 |
+
"grad_norm": 0.3559479590045658,
|
| 2703 |
+
"learning_rate": 1.3350556724625809e-05,
|
| 2704 |
+
"loss": 0.2061,
|
| 2705 |
+
"step": 368
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 0.45373501383338455,
|
| 2709 |
+
"grad_norm": 0.34195723156033303,
|
| 2710 |
+
"learning_rate": 1.3310033385079589e-05,
|
| 2711 |
+
"loss": 0.1761,
|
| 2712 |
+
"step": 369
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 0.45496464801721487,
|
| 2716 |
+
"grad_norm": 0.5770960005405298,
|
| 2717 |
+
"learning_rate": 1.3269448909578866e-05,
|
| 2718 |
+
"loss": 0.227,
|
| 2719 |
+
"step": 370
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 0.4561942822010452,
|
| 2723 |
+
"grad_norm": 0.6125705002122824,
|
| 2724 |
+
"learning_rate": 1.3228804047714462e-05,
|
| 2725 |
+
"loss": 0.2351,
|
| 2726 |
+
"step": 371
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 0.4574239163848755,
|
| 2730 |
+
"grad_norm": 0.48363918457013794,
|
| 2731 |
+
"learning_rate": 1.3188099550192537e-05,
|
| 2732 |
+
"loss": 0.1847,
|
| 2733 |
+
"step": 372
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 0.4586535505687058,
|
| 2737 |
+
"grad_norm": 0.6610021081352014,
|
| 2738 |
+
"learning_rate": 1.31473361688207e-05,
|
| 2739 |
+
"loss": 0.2129,
|
| 2740 |
+
"step": 373
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 0.45988318475253614,
|
| 2744 |
+
"grad_norm": 0.4341786311855446,
|
| 2745 |
+
"learning_rate": 1.3106514656494147e-05,
|
| 2746 |
+
"loss": 0.2426,
|
| 2747 |
+
"step": 374
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 0.4611128189363664,
|
| 2751 |
+
"grad_norm": 0.3013643837535933,
|
| 2752 |
+
"learning_rate": 1.3065635767181748e-05,
|
| 2753 |
+
"loss": 0.1596,
|
| 2754 |
+
"step": 375
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.46234245312019673,
|
| 2758 |
+
"grad_norm": 0.29963722757790967,
|
| 2759 |
+
"learning_rate": 1.302470025591211e-05,
|
| 2760 |
+
"loss": 0.1821,
|
| 2761 |
+
"step": 376
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.46357208730402705,
|
| 2765 |
+
"grad_norm": 0.5856707955354147,
|
| 2766 |
+
"learning_rate": 1.2983708878759655e-05,
|
| 2767 |
+
"loss": 0.2024,
|
| 2768 |
+
"step": 377
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 0.46480172148785737,
|
| 2772 |
+
"grad_norm": 0.37812820220102683,
|
| 2773 |
+
"learning_rate": 1.2942662392830632e-05,
|
| 2774 |
+
"loss": 0.2049,
|
| 2775 |
+
"step": 378
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 0.4660313556716877,
|
| 2779 |
+
"grad_norm": 0.41966854124526104,
|
| 2780 |
+
"learning_rate": 1.290156155624914e-05,
|
| 2781 |
+
"loss": 0.227,
|
| 2782 |
+
"step": 379
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 0.467260989855518,
|
| 2786 |
+
"grad_norm": 0.39476761658009546,
|
| 2787 |
+
"learning_rate": 1.286040712814314e-05,
|
| 2788 |
+
"loss": 0.1552,
|
| 2789 |
+
"step": 380
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 0.46849062403934827,
|
| 2793 |
+
"grad_norm": 0.4381036125394883,
|
| 2794 |
+
"learning_rate": 1.2819199868630419e-05,
|
| 2795 |
+
"loss": 0.1686,
|
| 2796 |
+
"step": 381
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.4697202582231786,
|
| 2800 |
+
"grad_norm": 0.2834280308233097,
|
| 2801 |
+
"learning_rate": 1.2777940538804545e-05,
|
| 2802 |
+
"loss": 0.1292,
|
| 2803 |
+
"step": 382
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 0.4709498924070089,
|
| 2807 |
+
"grad_norm": 0.3708781081449464,
|
| 2808 |
+
"learning_rate": 1.2736629900720832e-05,
|
| 2809 |
+
"loss": 0.1575,
|
| 2810 |
+
"step": 383
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 0.4721795265908392,
|
| 2814 |
+
"grad_norm": 0.37089620983594307,
|
| 2815 |
+
"learning_rate": 1.2695268717382242e-05,
|
| 2816 |
+
"loss": 0.1923,
|
| 2817 |
+
"step": 384
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 0.47340916077466955,
|
| 2821 |
+
"grad_norm": 0.37850202105410397,
|
| 2822 |
+
"learning_rate": 1.2653857752725305e-05,
|
| 2823 |
+
"loss": 0.1751,
|
| 2824 |
+
"step": 385
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 0.47463879495849987,
|
| 2828 |
+
"grad_norm": 0.29465463630363,
|
| 2829 |
+
"learning_rate": 1.2612397771606015e-05,
|
| 2830 |
+
"loss": 0.1792,
|
| 2831 |
+
"step": 386
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 0.47586842914233013,
|
| 2835 |
+
"grad_norm": 0.35698339352314057,
|
| 2836 |
+
"learning_rate": 1.2570889539785683e-05,
|
| 2837 |
+
"loss": 0.1325,
|
| 2838 |
+
"step": 387
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 0.47709806332616045,
|
| 2842 |
+
"grad_norm": 0.43664575464965527,
|
| 2843 |
+
"learning_rate": 1.2529333823916807e-05,
|
| 2844 |
+
"loss": 0.1764,
|
| 2845 |
+
"step": 388
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 0.47832769750999077,
|
| 2849 |
+
"grad_norm": 0.4130236851643136,
|
| 2850 |
+
"learning_rate": 1.2487731391528919e-05,
|
| 2851 |
+
"loss": 0.2261,
|
| 2852 |
+
"step": 389
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 0.4795573316938211,
|
| 2856 |
+
"grad_norm": 0.37099838030443905,
|
| 2857 |
+
"learning_rate": 1.2446083011014389e-05,
|
| 2858 |
+
"loss": 0.187,
|
| 2859 |
+
"step": 390
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 0.4807869658776514,
|
| 2863 |
+
"grad_norm": 0.241662388149488,
|
| 2864 |
+
"learning_rate": 1.2404389451614253e-05,
|
| 2865 |
+
"loss": 0.1671,
|
| 2866 |
+
"step": 391
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 0.4820166000614817,
|
| 2870 |
+
"grad_norm": 0.40938373429888714,
|
| 2871 |
+
"learning_rate": 1.2362651483403985e-05,
|
| 2872 |
+
"loss": 0.1861,
|
| 2873 |
+
"step": 392
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 0.48324623424531205,
|
| 2877 |
+
"grad_norm": 0.3623538192567374,
|
| 2878 |
+
"learning_rate": 1.2320869877279297e-05,
|
| 2879 |
+
"loss": 0.1743,
|
| 2880 |
+
"step": 393
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 0.4844758684291423,
|
| 2884 |
+
"grad_norm": 0.4437845258381219,
|
| 2885 |
+
"learning_rate": 1.2279045404941883e-05,
|
| 2886 |
+
"loss": 0.2337,
|
| 2887 |
+
"step": 394
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 0.48570550261297263,
|
| 2891 |
+
"grad_norm": 0.3749955717955905,
|
| 2892 |
+
"learning_rate": 1.2237178838885168e-05,
|
| 2893 |
+
"loss": 0.2027,
|
| 2894 |
+
"step": 395
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 0.48693513679680295,
|
| 2898 |
+
"grad_norm": 0.3505048992054775,
|
| 2899 |
+
"learning_rate": 1.2195270952380052e-05,
|
| 2900 |
+
"loss": 0.1571,
|
| 2901 |
+
"step": 396
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 0.48816477098063327,
|
| 2905 |
+
"grad_norm": 0.3713688621968288,
|
| 2906 |
+
"learning_rate": 1.215332251946061e-05,
|
| 2907 |
+
"loss": 0.1915,
|
| 2908 |
+
"step": 397
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 0.4893944051644636,
|
| 2912 |
+
"grad_norm": 0.4834162644214324,
|
| 2913 |
+
"learning_rate": 1.2111334314909811e-05,
|
| 2914 |
+
"loss": 0.2389,
|
| 2915 |
+
"step": 398
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 0.4906240393482939,
|
| 2919 |
+
"grad_norm": 0.38954625041703195,
|
| 2920 |
+
"learning_rate": 1.2069307114245197e-05,
|
| 2921 |
+
"loss": 0.1682,
|
| 2922 |
+
"step": 399
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 0.49185367353212417,
|
| 2926 |
+
"grad_norm": 0.46232563297646273,
|
| 2927 |
+
"learning_rate": 1.2027241693704567e-05,
|
| 2928 |
+
"loss": 0.2382,
|
| 2929 |
+
"step": 400
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 0.49185367353212417,
|
| 2933 |
+
"eval_accuracy": 0.8128342245989305,
|
| 2934 |
+
"eval_f1": 0.5454545454545454,
|
| 2935 |
+
"eval_loss": 0.41859376430511475,
|
| 2936 |
+
"eval_precision": 0.7777777777777778,
|
| 2937 |
+
"eval_recall": 0.42,
|
| 2938 |
+
"eval_runtime": 22.7158,
|
| 2939 |
+
"eval_samples_per_second": 2.201,
|
| 2940 |
+
"eval_steps_per_second": 0.176,
|
| 2941 |
+
"step": 400
|
| 2942 |
}
|
| 2943 |
],
|
| 2944 |
"logging_steps": 1,
|
|
|
|
| 2958 |
"attributes": {}
|
| 2959 |
}
|
| 2960 |
},
|
| 2961 |
+
"total_flos": 393623730028544.0,
|
| 2962 |
"train_batch_size": 4,
|
| 2963 |
"trial_name": null,
|
| 2964 |
"trial_params": null
|